Head First Learn To Code A Learner's Guide Coding And Computational Thinking

Head%20First%20Learn%20to%20Code%20-%20A%20Learner's%20Guide%20to%20Coding%20and%20Computational%20Thinking

User Manual: Pdf

Open the PDF directly: View PDF .
Page Count: 641

Download
Open PDF In Browser	View PDF

Head First

Learn to Code
A Brain-Friendly Guide

Watch out for
common traps
and pitfalls

A learner’s guide
to coding and
computational
thinking
Launch your
programming
career in
one chapter

Avoid embarrassing
syntax errors

Bend your mind
around 130 puzzles
& exercises

Learn why everything
your friends know about
computer science is
probably wrong

Eric FreemanElisabeth Robson & Eric Freeman

Praise for Head First Learn to Code
“This is one of the most surprising, entertaining and brilliantly-planned software books I’ve ever read.
It’s more interesting and better thought-out than all other beginning programming books I know put
together. I run into people all the time who need to learn programming and want to know what course
to take; I plan to tell them to forget about courses and get this book instead.”
— David Gelernter, Professor of Computer Science, Yale University
“Head First Learn to Code has humor, emotions, and step by step instructions. This book turns on your
brain, makes you laugh and teaches you to be a code master—it’s a keeper.”
— Smore Magazine, a science magazine to know more and be more
“The book is a great read, even for an experienced programmer, with new perspectives on how to teach
difficult concepts we take for granted. Whether your just starting out or trying to figure out ways to
bring more people to code, this book will serve you well. Eric has made learning to code accessible and
approachable without dumbing anything down and setting the expectations of effort and difficulty that
actually exist.”
— Avi Flombaum, Dean and Chief Product Officer of the Flatiron School
“As a high school computer science teacher, Head First Learn to Code is my new primary resource for
introducing topics to students. It has an immensely personal touch and its conversational nature, humor,
and general style make it feel as if you are learning from another human rather than merely reading a
piece of text.”
— Brandon Shufflebarger, Regents School of Austin
“This is the book I wish would have been around when I first started learning to code. Unlike other beginner’s
programming books, Freeman strikes a perfect balance of humor, concise tutorials, and helpful background
information—without sounding condescending or overly technical. Head First Learn to Code will certainly be a
valuable resource for beginning coders at our school fablab.”
— Patrick Benfield, Innovation Director, The Magellan International School
“It takes talent and creativity to write in such an accessible manner. I am looking forward to using the
book next school year in an introductory course. I became enthralled with the relevant and accessible
examples that were substantive rather than the typical irrelevant and superficial examples presented to
the reader of a typical text on the subject.”
— Josh Sharfman, Teacher, Shalhevet Advanced Studies, Computer Science
“In keeping with a book about a language named for Monty Python, there’s a lot of meta-humor here
about the software industry, about programmer culture, and about tutorial books themselves. It’s nice to be
reminded that coding takes place in the (sometimes illogical) world of humans.”
— Adjunct Professor of Music Technology at NYU
and Montclair State University

“As a National Champion Rock Band Coach, I understand that students must feel inspired by a subject
in order to consistently practice and enjoy themselves in the process. Without inspiration and joy, people
eventually lose focus and enthusiasm. That’s why good teachers are so valuable to our society. They
can break a potentially boring complex subject, like learning code, down into easily digestible, delicious
nuggets of fun. As and educator, Eric Freeman is masterful. He shows us how learning can be simple,
enjoyable and easily retainable.”
— James Mays, Director, Band Aid School of Music
“HFL2C is a pleasurable, entertaining and effective way to learn computational thinking and python
basics. The book has strong pedagogical underpinnings, an exciting array of learning activities
and is well written in a fun and conversational tone that is approachable and breaks down complex
computational concepts in an easy, digestible way.”
— Troy Welch, Coordinator, Innovations, Thompson Rivers University

Praise for other books by Eric Freeman
“I feel like a thousand pounds of books have just been lifted off of my head.”
— Ward Cunningham, inventor of the Wiki
“The admirable clarity, humor and substantial doses of clever make it the sort of book that helps even
non-programmers think well about problem-solving.”
— Cory Doctorow, co-editor of Boing Boing, Science Fiction author
“Freeman continues to use innovative teaching methods for communicating complex concepts to basic
principles.”
— Mark Arana, Strategy & Innovation, The Walt Disney Studios
“I can think of no better tour guide than Eric.”
— Miko Matsumura, VP of Marketing and Developer Relations at Hazelcast
Former Chief Java Evangelist, Sun Microsystems
“The definitive book on HTML5 for everyone from beginners to experienced developers.”
— Aaron LaBerge, CTO, ESPN
“The highly graphic and incremental approach precisely mimics the best way to learn this stuff...”
— Danny Goodman, author of Dynamic HTML: The Definitive Guide
“Eric clearly knows his stuff. As the Internet becomes more complex, inspired construction of web pages
becomes increasingly critical. Elegant design is at the core of every chapter here, each concept conveyed
with equal doses of pragmatism and wit.”
— Ken Goldstein, former CEO of Shop.com and author of
This is Rage: A Novel of Silicon Valley and Other Madness

Other O’Reilly books by Eric Freeman
Head First HTML and CSS
Head First JavaScript Programming
Head First HTML5 Programming
Head First Design Patterns
Other books in O’Reilly’s Head First series
Head First HTML and CSS
Head First JavaScript Programming
Head First HTML5 Programming
Head First Design Patterns
Head First Servlets and JSP
Head First Java
Head First Python

Head First
Learn to Code

Wouldn’t it be dreamy if there was
a book for learning to code that was
more fun than going to the dentist and
more revealing than an IRS form? It’s
probably just a fantasy...

Eric Freeman

Boston

Head First Learn to Code
by Eric Freeman
Copyright © 2018 Eric Freeman. All rights reserved.
Printed in the United States of America.
Published by O’Reilly Media, Inc., 1005 Gravenstein Highway North, Sebastopol, CA 95472.
O’Reilly Media books may be purchased for educational, business, or sales promotional use. Online editions are
also available for most titles (http://oreilly.com/safari). For more information, contact our corporate/institutional
sales department: (800) 998-9938 or corporate@oreilly.com.

Editors:

Jeff Bleiel, Dawn Schanafelt, Meghan Blanchette

Cover Designer:

Randy Comer

Production Editor:

Melanie Yarbrough

Indexer:		

Lucie Haskins

Proofreader:		

Rachel Monaghan

Printing History:
January 2018: First Edition.

Nutshell Handbook, the Nutshell Handbook logo, and the O’Reilly logo are registered trademarks of O’Reilly
Media, Inc. The Head First series designations, Head First Learn to Code, and related trade dress are trademarks of
O’Reilly Media, Inc.
Many of the designations used by manufacturers and sellers to distinguish their products are claimed as
trademarks. Where those designations appear in this book, and O’Reilly Media, Inc., was aware of a trademark
claim, the designations have been printed in caps or initial caps.
While every precaution has been taken in the preparation of this book, the publisher and the author assume no
responsibility for errors or omissions, or for damages resulting from the use of the information contained herein.
In other words, if you use anything in Head First Learn to Code to, say, run a nuclear power plant, you’re on your
own.
No turtles were harmed in the making of this book.

ISBN: 978-1-491-95886-5
[M]										
					

		

table of contents

Before KISS I had zero experience playing in a rock band
that wears makeup. —Gene Simmons

the author

Eric Freeman

Eric is described by Head First series co-creator Kathy Sierra as “one of those rare
individuals fluent in the language, practice, and culture of multiple domains from
hipster hacker, corporate VP, engineer, think tank.” And his background matches that
description well. By training, Eric is a computer scientist, having studied with industry
luminary David Gelernter during his Ph.D. work at Yale University. Professionally,
Eric is a former media company executive—having held the position of CTO of
Disney.com at The Walt Disney Company. Eric has also held positions at O’Reilly
Media, NASA, and several startups, and his IP is licensed and in use on every Mac
and PC. Over the last 15 years Eric has been one of the top-selling technical authors
on topics from beginning web development to high-level software design.
Eric is currently is a Principal at WickedlySmart, LLC, and lives with his wife and
young daughter in Austin, Texas.
Write to Eric at eric@wickedlysmart.com or visit http://wickedlysmart.com.

viii

table of contents

Table of Contents (summary)
1

Intro
xxv
Getting started: Thinking computationally				1

2

Know your value: Simple values, variables, and types		

3

Decisive code: Booleans, decisions, and loops			73

4

Providing some structure: Lists and iteration			125

5

Getting functional: Functions and abstraction			179

4b

Putting some order in your data: Sorting and nested iteration		

6

Putting it all together: Text, strings, and heuristics			245

7

Getting modular: Modules, methods, classes, and objects		 291

8

Beyond iteration and indices: Recursion and dictionaries		

9

Persistence: Saving and retrieving files				393

10

You really should get out more: Using web APIs			435

11

Getting interactive: Widgets, events, and emergent behavior		

12

A trip to Objectville: Object-oriented programming			523

Appendix: The Top Ten Topics (we didn’t cover): Leftovers		

33

225

341

467
575

Table of Contents (the real thing)
Intro
Your brain on coding.

Here you are trying to learn something, while here your

brain is doing you a favor by making sure the learning doesn’t stick. Your brain’s thinking,
“Better leave room for more important things, like which wild animals to avoid and whether
naked snowboarding is a bad idea.” So how do you trick your brain into thinking that your
life depends on knowing how to code?
Who is this book for ?					

xxvi

We know what you’re thinking				

xxvii

We think of a “Head First” reader as a learner			

xxviii

Metacognition: thinking about thinking				

xxix

Here’s what WE did					xxx
Here’s what YOU can do to bend your brain into submission

xxxi

Read Me							xxxii
Acknowledgments						xxxvii
The Review Team						xxxviii

ix

table of contents

1

thinking computationally

Getting Started
Knowing how to think computationally puts you in control.
It’s no secret the world around you is becoming more connected, more configurable,
more programmable, and more, well, computational. You can remain a passive
participant, or you can learn to code. When you can code, you’re the director, the
creator—you’re telling all those computers what they should be doing for you. When
you can code, you control your own destiny (or at least you’ll be able to program
your internet-connected lawn sprinker system). But how do you learn to code? First,
learn to think computationally. Next, you grab a programming language so you
can speak the same lingo as your computer, mobile device, or anything with a CPU.
What’s in it for you? More time, more power, and more creative possibilities to do the
things you really want to do.

Breaking it down						2
How coding works					6
Are we even speaking the same language?			

7

The world of programming languages			

8

How you’ll write and run code with Python			

13

A very brief history of Python				

15

Putting Python through its paces				

18

Saving your work						20
Congrats on coding your first Python program!		

21

Phrase-O-Matic						25
Getting the code in the machine				

x

26

table of contents

2

simple values, variables, and types

Know Your Value

Computers really only do two things well: store values and perform
operations on those values. You might think they’re doing a whole lot more, as you send
texts, shop online, use Photoshop, or rely on your phone to navigate in your car; however,
everything computers do can be broken down into simple operations that are performed on
simple values. Now, part of computational thinking is learning to use these operations and
values to build something that is much more sophisticated, complex, and meaningful—and
we’re going to get to that. First, though, we’re going to take a look at what these values are,
the operations you can perform on them, and just what role variables play in all this.

Coding the Dog Age Calculator				

34

Going from pseudocode to code				

36

Step 1: Getting some input					

37

How the input function works				

38

Using variables to remember and store values		

38

Assigning the user’s input to a variable			

39

Step 2: Getting more input				

39

It’s time to run some code					

40

Getting some code entered					43
A deep dive on variables					

44

Adding some expression					45
Variables are called VARY-ables for a reason			

46

Better living through operator precedence			

47

Computing with operator precedence			

48

Back away from that keyboard!				

51

Step 3: Computing the dog’s age				

52

Houston, we’ve got a problem!				

53

To err is human to code					

54

A little more debugging...					56
What are Python types, anyway?				

58

Fixing our code						59
Houston, we have liftoff 					60
Step 4: User-friendly output				

61
xi

table of contents

3

booleans, decisions, and loops

Decisive Code

Have you noticed how, so far, our programs aren’t
very, well, interesting? That is, all our code has strictly been a set of
statements the interpreter evaluates from top to bottom—no twists in the plot,
no sudden turns, no surprises, no independent thinking. For code to be more
interesting, it needs to make decisions, to control its own destiny, and to do
things more than once straight through. And in this chapter that’s exactly what
we’re going to learn to do. Along the way we’ll learn about the mysterious game
called shoushiling, meet a character named Boole, and see how a data type
with only two values could be worth our time. We’re even going to learn how to
deal with the dreaded infinite loop.
Would you like to play a game?				

74

First, a high-level design					76
The computer’s choice					77
How to use the random number				

78

Introducing the Boolean type					81
Making decisions						82
Decisions and more decisions					83
Back to Rock, Paper, Scissors					

84

Getting the user’s choice					85
Taking a look at the user’s choice				

88

Adding the code to detect a tie				

89

Who won?						90
How to implement the game logic				

92

More about Boolean operators				

93

Got documentation?					98

xii

How to add comments to your code				

99

We need to finish that game!					

100

How do we know if the user’s choice is invalid?			

101

How to continually prompt the user				

104

Doing things more than once					

105

How the while loop works					

106

How to use while to prompt the user until you get a valid choice

110

Congratulations on coding your first game!			

112

table of contents

4

lists and iteration

Providing Some Structure
There’s more to data types than numbers, strings, and Booleans.
So far you’ve been writing Python code using primitive types—those floats, integers, strings,
and of course Booleans—with values like 3.14, 42, “hey, it’s my turn”, and True. And

you can do a lot with primitives, but at some point you’ll want to write code that deals with lots
of data—say, all the items in a shopping cart, the names of all the notable stars, or an entire

product catalog. For that we need a little more ummph. In this chapter we’re going to look at a
new type, called a list, which can hold a collection of values. With lists, you’ll be able to provide
some structure for your data, rather than just having a zillion variables floating around your code
holding values. You’re also going to learn how to treat all those values as a whole as well as how
to iterate over each item in a list using that for loop we mentioned in the last chapter. After this
chapter, your ability to deal with data is going to grow and expand.
Can you help Bubbles-R-Us?				

126

How to represent multiple values in Python		

127

How lists work					128
How big is that list, anyway?				

131

Accessing the last item in the list			

132

Python makes this even easier				

132

Using Python’s negative indices			

133

Meanwhile, back at Bubbles-R-Us...			

135

How to iterate over a list				

138

Fixing the output glitch				

139

Really fixing the output glitch				

140

The for loop, the preferred way to iterate over a list

142

How the for loop works on a range of numbers		

145

Doing more with ranges				

146

Putting it all together				

148

Building your own list, from scratch			

156

Doing even more with lists				

157

Test drive the final report				

161

And the winners are...				

161

Testing the most cost-effective solution			

165

xiii

table of contents

5

functions and abstraction

Getting Functional
You already know a lot. Variables and data types and conditionals and
iteration—that’s enough to write basically any program you’d ever want to. In fact,
a computer scientist would tell you it’s enough to write any program that anyone
could ever conceive of. But you don’t want to stop now, because your next step
in computational thinking is learning how to create abstractions in your code.
That may sound complex, but it’s actually going to make your coding life simpler.
Creating abstractions gives you leverage; with abstraction, you can more easily
create programs of increasing complexity and power. You can put your code in neat
little packages that you can reuse over and over. And you can forget all the nittygritty details of your code and to start thinking at a higher level.
What’s wrong with the code, anyway?				

181

Turning a block of code into a FUNCTION			

183

We created a function, so how do we use it?			

184

But how does all this actually work?				

184

Functions can RETURN things too				

192

How to call a function that has a return value			

193

Getting a little refactoring under our belts			

195

Running the code						196
How to abstract the avatar code				

197

Writing the get attribute function body				

198

Calling get attribute					199
Let’s talk about variables a little more...				

201

Understanding variable scope					202

xiv

When variables get passed to functions				

203

Making the drink me function call				

204

What about using global variables in functions?			

207

Going further with parameters: default values and keywords

210

How default parameter values work				

210

Always list your required parameters first!			

211

Using arguments with keywords				

212

How to think about all these options				

212

table of contents

4

sorting and nested iteration

part 2

Putting Some Order
in Your Data
Sometimes the default ordering of your data doesn’t cut it.
You’ve got that list of high scores on 80s arcade games, but you really need it sorted
alphabetically by game name. Then there’s that list of the number of times your coworkers
have stabbed you in the back—it would be nice to know who’s at the top of that list. To do
that, though, we need to learn how to sort data, and to do that we’ll need to explore some
algorithms that are a little more involved than the ones we’ve seen so far. We’re also going
to have to explore how nested loops work as well as think a little about the efficiency of the
code we’re writing.
Understanding bubble sort				228
Starting with pass 1					228
Some bubble sort pseudocode				

231

Implementing bubble sort in Python			

234

Computing bubble solution numbers			

236

Built-in
sort! Maybe
you could have let
us know 10 pages
back?

Top Bubble
Software Maker

xv

table of contents

6

text, strings, and heuristics

Putting It All Together
You’ve already got a lot of superpowers. N
 ow it’s time to use
them. In this chapter we’re going to integrate what we’ve learned so far, bringing
it all together to build some increasingy cool code. We’re also going to keep
adding to your knowledge and coding skills. More specifically, in this chapter
we’ll explore how to write code that grabs some text, slices it, dices it, and then
does a little data analysis on it. We’re going to find out what a heuristic is too,
and implement one. Get ready—this is an all-out, heads-down, pedal-to-themetal, serious coding chapter!
Welcome to the data sciences					

246

How do you compute something like readability?			

247

The game plan						248
Writing some pseudocode					249
We need some text to analyze					

250

Setting up the function					252
First up: we need the total number of words in our text		

253

Computing the total number of sentences			

257

Writing the count sentences function				

258

Computing the number of syllables, or, learning to love heuristics

264

Setting up the heuristic					267
Writing the heuristic					268
How to count vowels					269
Ignoring consecutive vowels					269
Writing the code to ignore consecutive vowels			

270

Removing final e’s, y’s, and punctuation				

272

Putting slicing (substrings) to work				

274

Finishing off the heuristic code				

276

Implementing the reading ease formula				

278

Taking it even further					283

xvi

Definitely some
sophisticated writing in
this book.

table of contents

7

modules, methods, classes, and objects

Getting Modular

Your code is growing in size and complexity. As that happens you need better ways
to abstract, to modularize, to organize your code. You’ve seen that functions can be used to group lines
of code together into bundles you can reuse over and over. And you’ve also seen that collections of
functions and variables can be placed into modules so that they can be more easily shared and reused.
In this chaper we’ll revisit modules and learn how to use them even more effectively (so you’re all ready
to share your code with others) and then we’re going to look at the ultimate in code reuse: objects.
You’re going to see that Python objects are all around you, just waiting to be used.
A quick module review					294
The __name__ global variable				

296

Updating analyze.py					297
Using analyze.py as a module					

299

Adding docstrings to analyze.py				

301

Exploring other Python modules				

305

Wait, did someone say “turtles”?!				

306

Creating your very own turtle					

308

Turtle lab							309
Adding a second turtle					311
What are turtles, anyway?					314
What are objects?						315
Nice job, I was quickly
able to use the analyze
module, especially with the help
of the great documentation!

Okay, what’s a class then?					

316

How to use objects and classes				

318

What about those methods and attributes?			

319

Seeing classes and objects everywhere				

320

Get ready for some turtle races				

322

Planning the game						323
Let’s start coding						324
Setting up the game					324
Writing the setup code					325
Not so fast!						326
Starting the race						328

xvii

table of contents

8

recursion and dictionaries

Beyond Iteration and Indices
It’s time to take your computational thinking up a notch. A
 nd
this is the chapter to do it: we’ve been happily coding along with an iterative style of
programming—we’ve created data structures like lists and strings and ranges of numbers,
and we’ve written code to compute by iterating over them. In this chapter we’re going to look
at the world differently, first in terms of computation, and then in terms of data structures.
Computationally we’ll look at a style of computing that involves writing code that recurs,
or calls itself. We’ll expand the kinds of data structures we can work with by looking at a
dictionary-like data type that is more like an associative map than a list. We’ll then put them
together and cause all kinds of trouble. Be forewarned: these topics take a while to settle
into your brain, but the effort is going to pay off in spades.
A different way to compute					

342

And now the different way...					

343

Now let’s write some code for our two cases			

344

Let’s get some more practice					

347

Using recursion to detect palindromes				

348

Writing a recursive palindrome detector			

349

The Anti-Social Network					360
Introducing the dictionary					362
Let’s see how to create a dictionary				

362

Keys and values do not have to be strings			

363

Of course you can remove keys as well				

363

But you might want to test to see if it exists first			

363

What about iterating through a dictionary?			

364

Leveraging dictionaries at the Anti-Social Network		

366

But how do we add more attributes?				

368

Remember the Anti-Social Network’s killer feature?		

370

Finding the most anti-social user				

371

Can we just remember function call results?			

376

Using a dictionary to remember our Fibonacci results		

376

Memoization						377

xviii

A closer look at the koch function				

380

Truly exploring the koch fractal				

382

table of contents

9

saving and retrieving files

Persistence

You know you can save values in variables, but once your program ends
poof!—they’re gone forever. That’s where persistent storage comes in—storage that allows
your values and data to stick around a while. Most of the devices you’re going to run Python on also have
persisistent storage, like hard drives and flash cards, or they may have access to storage in the cloud as
well. In this chapter you’ll see how to write code to store and retrieve data from files. What good is that?
Oh, anytime you want to save a user’s configuration, store the results of your big analysis for the boss,
read an image into your code to process it, write some code to search a decade’s worth of email messages,
reformat some data to be used in your spreadsheet app—we could go on and on.
Ready for some Crazy Libs?					

394

How Crazy Libs is going to work				

396

Step 1: Read the text of the story from a file			

399

How to use file paths					

400

Absolute paths						401
Oh, and don’t forget to clean up when you’re done!		

402

Reading a file into your Python code				

403

Oh, give it a break already...					

406

Using the readline method of the file object			

407

How do we know when we’ve read the last line?			

409

Reading in a Crazy Lib template				

410

Processing the template text					411
Using a new string method to fix the bug			

413

Actually fixing the bug					414
Some code has real problems					

415

Handling exceptions					417
Explicitly handling exceptions				418
Updating Crazy Libs to handle exceptions			

420

Our last step: storing the Crazy Lib				

421

Updating the rest of the code					

421

xix

table of contents

10

using web apis

You Really Should Get Out More
You’ve been writing some great code, but you really need to get out
more.There’s a whole world of data just waiting for you on the web: Need weather data? Or
how about access to a huge database of recipes? Or are sports scores more your thing? Maybe
a music database of artists, albums, and songs? They’re all out there for the taking from Web
APIs. To use them all you need is to learn a bit more about how the web works, how to speak the
local web lingo, and how to use a couple of new Python modules: requests and json. In this

chapter we’re going to explore Web APIs and take your Python skills to new heights; in fact, we’re
going to take them all the way to outer space and back.
Extending your reach with Web APIs				

436

How Web APIs work					437
All Web APIs have a web address				

438

Time for a quick upgrade					

441

Doing the upgrade						442
All we need now is a good Web API...				

443

A closer look at the API					

444

Web APIs provide data using JSON				

445

Now let’s look at that request module again			

447

Putting it all together: making a request to Open Notify		

449

How to use JSON in Python					

450

Using the JSON module on our ISS data			

451

Let’s add some graphics					452
Meet the screen object					453

xx

Let’s add a turtle to represent the ISS				

455

Turtles can look like space stations too				

456

Forget the ISS—where are we?				

457

Finishing off the ISS code					

458

table of contents

11

widgets, events, and emergent behavior

Getting Interactive

You’ve certainly written some graphical applications, but you haven’t
created a real user interface yet.That is, you haven’t written anything that lets the user
interact with a graphical user interface (otherwise known as a GUI). To do that you need to adopt a
new way of thinking about how a program executes, one that is more reactive. Wait, did the user just
click on that button? Your code better know how to react and what to do next. Coding for interfaces is
quite different from the typical procedural method we’ve been using, and it requires a different way of
thinking about the problem. In this chapter you’re going to write your first real GUI, and no, we’re not
going to write a simple to-do list manager or height/weight calculator, we’re going to do something far
more interesting. We’re going to write an artificial life simulator with emergent behavior.
Enter the WONDERFUL WORLD of Artifical Life 		
468
A closer look at the Game of Life				
469
What we’re going to build					
472
Do we have the right design?					
473
Building the data model					477
Representing the grid					477
Computing a generation of the Game of Life			
478
Computing each cell’s fate					478
Completing the model code					482
Building the view						485
Creating your first widget					486
Adding the rest of the widgets				
487
Correcting the layout					488
Placing the widgets into a grid layout				
489
Translating the grid layout to code				
489
Moving on to the controller					
491
Adding an update function					491
Ready for another new style of computation?			
494
How the start/pause button is going to work			
497
Another kind of event					499
We have the technology: the after method			
501
How to enter and edit cells directly				
504
Writing the grid view handler					
505
Now it’s time to add some patterns				
506
Writing a handler for the OptionMenu				
507
Writing the pattern loader					510
Taking it Even Further!					517

xxi

table of contents

12

object-oriented programming

A Trip to Objectville

In this book you’ve used functions to abstract your code. And you’ve
approached coding in a procedural manner using simple statements, conditionals, and for/while

loops with functions—none of this is exactly object-oriented. In fact, it’s not object-oriented at all! We

have looked at objects and how to use them in our code, but you haven’t created any objects of your
own yet, and you haven’t really approached designing your code in an object-oriented way. So, the time
has come to leave this boring procedural town behind. In this chapter, you’re going to find out why using
objects is going to make your life so much better—well, better in a programming sense (we can’t really
help you with other areas of your life and your coding skills, all in one book).
Breaking it down, a different way				
524
What’s the point of object-oriented programming, anyway?
525
Designing your first class					527
Writing your first class					528
How the constructor works				
528
Writing the bark method					531
How methods work					532
Adding some inheritance					534
Implementing the ServiceDog class				
535
A closer look at subclassing				
536
A ServiceDog IS-A Dog					537
Testing IS-A in code					538
Overriding and extending behavior				
542
Welcome to Jargon City					544
Object can HAS-Another object				
546
Designing a Doggie Hotel					549
Implementing the Doggie Hotel				
550
Adding some hotel activities				
554
I can do anything you can, or Polymorphism			
555
It’s about time we teach the other dogs to walk			
556
The power of inheritance					558

xxii

table of contents

appendix: leftovers

The Top Ten Topics
(We Didn’t Cover)
We’ve covered a lot of ground, and you’re almost finished with this book.
We’ll miss you, but before we let you go, we wouldn’t feel right about sending you out into the world
without a little more preparation. We can’t possibly fit everything you’ll need to know into this relatively
small chapter. Actually, we did originally include everything you need to know about Python programming
(not already covered by the other chapters), by reducing the type point size to .00004. It all fit, but nobody
could read it. So we threw most of it away, and kept the best bits for this Top Ten appendix.
#1 List comprehensions				576
#2 Dates and times					577
#3 Regular expressions				578
#4 Other data types: tuples				

579

#5 Other data types: sets				

580

#6 Server-side coding				581
#7 Lazy evaluation					582
#8 Decorators					583
#9 Higher-order and first-class functions		

584

#10 A lot of libraries				

585

Server-side code
executes on a server on
the internet.
st

ue

q
re

Client-side code executes
on the client—that is, on
your computer.
xxiii

how to use this book

Intro

I can’t believe they
put that in a book on
learning to code!

estion:
swer the burning qu
code?”
on
In this section, we an
a book learning to
in
at
th
t
pu
ey
th
“So, why DID

xxv

how to use this book

Who is this book for ?
If you can answer “yes” to all of these:
1

Do you want to learn, understand, and remember how
to program?

2

Do you prefer stimulating dinner party conversation
to dry, dull, academic lectures?

this book is for you.

: this book is
[Note from marketcringedit card.]
for anyone with a

Who should probably back away f rom this book ?
If you can answer “yes” to any one of these:
1

Are you completely new to computers?
If you don’t know your way around your computer,
how to manage files and folders, how to install apps,
or how to use a word processor, you should probably
learn those first.

2

Are you a kick-butt programmer looking for a
reference book?

3

Are you afraid to try something different? Would you
rather have a root canal than mix stripes with plaid? Do
you believe that a technical book can’t be serious if we
have fun learning to code?

this book is not for you.

xxvi

intro

This is NOT a
reference book.
Head First Learn
to Code is a
book designed
for learning to
code. It’s not an
encyclopedia of
programming facts
(you have Google
for that, right?).

the intro

We know what you’re thinking.
“How can this be a serious book?”
“What’s with all the graphics?”
“Can I actually learn it this way?”

And we know what your brain is thinking.

Your brain
THIS is im thinks
portant.

Your brain craves novelty. It’s always searching, scanning, waiting for
something unusual. It was built that way, and it helps you stay alive.
Today, you’re less likely to be a tiger snack. But your brain’s still looking.
You just never know.
So what does your brain do with all the routine, ordinary, normal things
you encounter? Everything it can to stop them from interfering with the
brain’s real job—recording things that matter. It doesn’t bother saving the
boring things; they never make it past the “this is obviously not important”
filter.
How does your brain know what’s important? Suppose you’re out for a day
hike and a tiger jumps in front of you. What happens inside your head
and body?
Neurons fire. Emotions crank up. Chemicals surge.
And that’s how your brain knows…
This must be important! Don’t forget it!
But imagine you’re at home, or in a library. It’s a safe, warm, tigerfree zone. You’re studying. Getting ready for an exam. Or trying to
learn some tough technical topic your boss thinks will take a week,
10 days at the most.

in thinks
Your bran’t worth
THIinS gis.
sav

Great. Only 600
more dull, dry,
boring pages.

Just one problem. Your brain’s trying to do you a big favor. It’s
trying to make sure that this obviously non-important content doesn’t
clutter up scarce resources. Resources that are better spent storing
the really big things. Like tigers. Like the danger of fire. Like how
you should never again snowboard in shorts.
And there’s no simple way to tell your brain, “Hey brain, thank
you very much, but no matter how dull this book is, and how little
I’m registering on the emotional Richter scale right now, I really do
want you to keep this stuff around.”

you are here� xxvii

how to use this book

r.
rst” reader as a learne
We think of a “Head Fi

get it, then make
? First, you have to
ing
th
me
so
rn
lea
to
head. Based
So what does it take
ing facts into your
sh
pu
t
ou
ab
t
no
s
t it. It’
ational
sure you don’t forge
obiology, and educ
itive science, neur
gn
co
in
ch
what
ar
se
ow
re
a page. We kn
on the latest
more than text on
lot
a
s
ke
ta
ing
rn
psychology, lea
.
Your code
turns your brain on
inciples:
First lear ning pr
Some of the Head
ne,
morable than words alo
ages are far more me
in
ent
Make it visual. Im
vem
pro
im
89%
to
ch more effective (up
e.
and make learning mu
s more understandabl
ng
thi
kes
ma
o
als
It
s).
die
stu
er
nsf
tra
rather than
to,
and
recall
aphics they relate
in or near the gr
th
wi
s
rd
wo
likely to
e
as
th
Put
rs will be up to twice
ther page, and learne
ano
on
or
m
tto
bo
on the
d to the content.
The Python
solve problems relate
Intepreter
style. In recent
d
ze
ali
on
rs
pe
d
nal an
tests
g
nin
Use a conver satio
ear
st-l
po
on
r
rmed up to 40% bette
versational
studies, students perfo
ng a first-person, con
ectly to the reader, usi
dir
ke
’ll
spo
t
you
ten
. Use casual
nk
thi
con
ing
lly
if the
I rea
ries instead of lec tur
t
a formal tone. Tell sto
ing
tak
n
want to abstract tha
tha
attention
her
re
rat
mo
le
sty
. Which would you pay
code into a function.
yourself too seriously
e
tak
n’t
Do
.
age
gu
lan
a lec ture?
par ty companion, or
to: a stimulating dinner
r
Now that I have you
In other words,
, you should be
ly.
ion
ep
ent
de
att
re
mo
think
bal
Get the learner to
more careful using glo
ch happens in
r neurons, nothing mu
you
flex
.
ly
les
ive
iab
act
var
you
unless
aged, curious, and
to be motivated, eng
your head. A reader has
and generate new
ms, draw conclusions,
inspired to solve proble
s, exercises and
t, you need challenge
knowledge. And for tha
that involve
estions, and activities
thought-provok ing qu
and multiple senses.
both sides of the brain
arn to

le
Don’t justrn to think
ion. We’ve all
code—lea ionally.
he reader’s attent
Get—and keep—t
t
computat
I can’t stay awake pas

to learn this, but
had the “I really want
to things that are out
r brain pays attention
You
e.
enc
eri
,
exp
e”
on
page
ected. Learning a new
e, eye -catching, unexp
ang
str
g,
stin
re quick ly if it’s not.
ere
int
mo
ry,
ch
of the ordina
r brain will learn mu
You
g.
rin
bo
be
to
e
doesn’t hav
tough, technical topic
ething is largely
lity to remember som
now know that your abi
We
s.
ion
ot
remember when you
em
You
Touch their
r what you care about.
be
em
rem
You
t.
king
ten
otional con
y and his dog. We’re tal
dependent on its em
ng stories about a bo
chi
ren
rt-w
hea
you
g
en
kin
’re not tal
rule!” that comes wh
feel something. No, we
, and the feeling of “I
”
…?
the
hat
“w
t
,
tha
fun
ing
curiosity,
you know someth
emotions like surprise,
nks is hard, or realize
thi
e
els
y
od
ryb
eve
ething
solve a puzzle, learn som
ering doesn’t.
thou” Bob from Engine
n
tha
l
ica
hn
tec
“I’m more

xxviii

intro

the intro

Me tacognition: thinking about thinking
If you really want to learn, and you want to learn more quickly and more deeply,
pay attention to how you pay attention. Think about how you think. Learn how you
learn.
Most of us did not take courses on metacognition or learning theory when we were
growing up. We were expected to learn, but rarely taught how to learn.

I wonder how I
can trick my brain
into remembering
this stuff...

But we assume that if you’re holding this book, you really want to learn
how to code to create programs and apps. And you probably don’t want
to spend a lot of time. And you want to remember what you read, and be
able to apply it. And for that, you’ve got to understand it. To get the most
from this book, or any book or learning experience, take responsibility for
your brain.
The trick is to get your brain to see the new material you’re learning as
Really Important. Crucial to your well-being. As important as a tiger.
Otherwise, you’re in for a constant battle, with your brain doing its best
to keep the new content from sticking.
So how DO you get your brain to think coding is as
important as a tiger?
There’s the slow, tedious way, or the faster, more effective way. The
slow way is about sheer repetition. You obviously know that you are
able to learn and remember even the dullest of topics, if you keep pounding on the same
thing. With enough repetition, your brain says, “This doesn’t feel important to him, but he
keeps looking at the same thing over and over and over, so I suppose it must be.”
The faster way is to do anything that increases brain activity, especially different
types of brain activity. The things on the previous page are a big part of the solution,
and they’re all things that have been proven to help your brain work in your favor. For
example, studies show that putting words within the pictures they describe (as opposed to
somewhere else in the page, like a caption or in the body text) causes your brain to try
to make sense of how the words and picture relate, and this causes more neurons to fire.
More neurons firing = more chances for your brain to get that this is something worth
paying attention to, and possibly recording.
A conversational style helps because people tend to pay more attention when they
perceive that they’re in a conversation, since they’re expected to follow along and hold up
their end. The amazing thing is, your brain doesn’t necessarily care that the “conversation”
is between you and a book! On the other hand, if the writing style is formal and dry, your
brain perceives it the same way you experience being lectured to while sitting in a roomful
of passive attendees. No need to stay awake.
But pictures and conversational style are just the beginning.

you are here�

xxix

how to use this book

Here’s what WE did:

We used pictures, because your brain is tuned for visuals, not text. As far as your brain’s
concerned, a picture really is worth 1,024 words. And when text and pictures work together, we
embedded the text in the pictures because your brain works more effectively when the text is
within the thing the text refers to, as opposed to in a caption or buried in the text somewhere.
We used redundancy, saying the same thing in different ways and with different media types,
and multiple senses, to increase the chance that the content gets coded into more than one area of
your brain.
We used concepts and pictures in unexpected ways because your brain is tuned for novelty,
and we used pictures and ideas with at least some emotional content, because your brain is
tuned to pay attention to the biochemistry of emotions. That which causes you to feel something
is more likely to be remembered, even if that feeling is nothing more than a little humor,
surprise, or interest.
We used a personalized, conversational style, because your brain is tuned to pay more
attention when it believes you’re in a conversation than if it thinks you’re passively listening to a
presentation. Your brain does this even when you’re reading.
We included more than 120 activities, because your brain is tuned to learn and remember
more when you do things than when you read about things. And we made the exercises
challenging-yet-doable, because that’s what most people prefer.
We used multiple learning styles, because you might prefer step-by-step procedures, while
someone else wants to understand the big picture first, while someone else just wants to see a
code example. But regardless of your own learning preference, everyone benefits from seeing the
same content represented in multiple ways.

Be the Python
interpreter

BULLET POINTS

Puzzles

We include content for both sides of your brain, because the more of your brain you
engage, the more likely you are to learn and remember, and the longer you can stay focused.
Since working one side of the brain often means giving the other side a chance to rest, you can
be more productive at learning for a longer period of time.
And we included stories and exercises that present more than one point of view, because
your brain is tuned to learn more deeply when it’s forced to make evaluations and judgments.
We included challenges, by providing exercises and by asking questions that don’t always
have a straight answer, because your brain is tuned to learn and remember when it has to work
at something. Think about it—you can’t get your body in shape just by watching people at the
gym. But we did our best to make sure that when you’re working hard, it’s on the right things.
That you’re not spending one extra dendrite processing a hard-to-understand example,
or parsing difficult, jargon-laden, or overly terse text.
We used people. In stories, examples, pictures, and so on, because, well, you’re a person. And
your brain pays more attention to people than it does to things.
We used an 80/20 approach. We assume that if you’re going to be a kick-butt programmer,
this won’t be your only book. So we don’t talk about everything. Just the stuff you’ll actually need.

xxx

intro

They’re coming
along with us.

the intro

Here’s what YOU can do to bend
your brain into submission

So, we did our part. The rest is up to you. These tips are a
starting point; listen to your brain and figure out what works
for you and what doesn’t. Try new things.

Cut this out an
ick it
on your refrigerdatst
or.
1

Slow down. The more you understand,
the less you have to memorize.

6

Speaking activates a different part of the brain.
If you’re trying to understand something, or
increase your chance of remembering it later, say
it out loud. Better still, try to explain it out loud
to someone else. You’ll learn more quickly, and
you might uncover ideas you hadn’t known were
there when you were reading about it.

Don’t just read. Stop and think. When the
book asks you a question, don’t just skip to the
answer. Imagine that someone really is asking
the question. The more deeply you force your
brain to think, the better chance you have of
learning and remembering.
2

Do the exercises. Write your own notes.

We put them in, but if we did them for you,
that would be like having someone else do
your workouts for you. And don’t just look at
the exercises. Use a pencil. There’s plenty of
evidence that physical activity while learning
can increase the learning.
3

4

Drink water. Lots of it.

Your brain works best in a nice bath of fluid.
Dehydration (which can happen before you ever
feel thirsty) decreases cognitive function.

Listen to your brain.

Pay attention to whether your brain is getting
overloaded. If you find yourself starting to skim the
surface or forget what you just read, it’s time for a
break. Once you go past a certain point, you won’t
learn faster by trying to shove more in, and you
might even hurt the process.
8

Feel something!

Your brain needs to know that this matters. Get
involved with the stories. Make up your own
captions for the photos. Groaning over a bad joke is
still better than feeling nothing at all.

Make this the last thing you read before
bed. Or at least the last challenging thing.

Part of the learning (especially the transfer to
long-term memory) happens after you put the
book down. Your brain needs time on its own, to
do more processing. If you put in something new
during that processing time, some of what you
just learned will be lost.
5

7

Read the “There Are No Dumb Questions”

That means all of them. They’re not optional
sidebars—they’re part of the core content!
Don’t skip them.

Talk about it. Out loud.

9

Create something!

Apply this to something new you’re designing, or
rework an older project. Just do something to get some
experience beyond the exercises and activities in
this book. All you need is a pencil and a problem
to solve…a problem that might benefit from
programming.
10 Get sleep.
You’ve got to create a lot of new brain connections
to learn to program. Sleep often; it helps.
you are here�

xxxi

how to use this book

Re ad Me
This is a learning experience, not a reference book. We deliberately stripped out everything
that might get in the way of learning whatever it is we’re working on at that point in the
book. And the first time through, you need to begin at the beginning, because the book
makes assumptions about what you’ve already seen and learned.
We want you to learn the thinking process behind programming.
Some might call that computer science, but here’s a little secret: computer science isn’t a
science and it’s not even all that much about computers (any more than astronomy is about
telescopes). It’s a way of thinking, otherwise known these days as computational thinking,
and once you learn to think computationally, you’ll be in a good position to apply that to any
problem, environment, or programming language.
In this book we use Python.
Learning to drive without a vehicle is a little academic. And learning to think
computationally without a programming language is more of a thought experiment than
a marketable skill. So, in this book we use the very popular Python language. We’ll tell you
more about its accolades in Chapter 1, but whether you’re a hobbyist or hoping to land a sixfigure software development position, Python is a good place to start (and maybe end).
We don’t exhaustively cover every aspect of the Python language.
Not even close. There’s a lot you can learn about Python. This book is not a reference book,
it’s a learning book, so it doesn’t cover everything there is to know about Python. Our goal is
to teach you the fundamentals of coding and computational thinking so that you can pick up
a book on any programming language and not feel totally lost.
You can use a Mac or PC, or Linux for that matter.
As Python is our primary vehicle used in this book and it is cross-platform, you can use
whatever operating system you’re used to. Most of the screenshots in this book are from a
Mac, but they should look similar on your PC or Linux box.
This book advocates well-structured and readable code based
on best practices. You want to write code that you and other people can read and
understand, code that will still work in next year’s version of Python. In this book we’re going
to teach you to write clear, well-organized code from the get-go—code you can be proud of,
code you’ll want to frame and put on the wall (just take it down before you bring your date
over). The only thing that differs from what we’d write as professional code is that this book
uses handwritten annotations next to code to explain what the code is doing. We found this
works better in a learning book than traditional comments in code (if you have no idea what
we’re talking about, you will; just give it a few chapters). But don’t worry because we’ll teach
you how to document your code and we’ll show you examples of how we’d document our
own code. All that said, we’re interested in teaching you to write code in the most straightforward way so you can get the job done and move on to better things.

xxxii

intro

Annotations
like this

the intro

Programming is serious business. You’re going to have to work,
sometimes hard. A programmer has a different mindset, a different way of thinking
about the world. At times you’re going to find coding very logical, while at other times it
can be very abstract, if not downright mind bending. Some programming concepts take
time to sink into your brain—you actually do have to sleep on them before you’ll get it. But
no worries; we’re going to do all that in a brain-friendly way. Just take your time, give the
concepts time to sink in, and go over the material multiple times if needed.
The activities are NOT optional.
The exercises and activities in this book are not add-ons; they’re part of the core content of
the book. Some of them are to help with memory, some are for understanding, and some
will help you apply what you’ve learned. If you skip them you will be missing large parts of
the book (and you’ll probably be very confused). The crossword puzzles are the only things
you don’t have to do, but they’re good for giving your brain a chance to think about the
words in a different context.
The redundancy is intentional and important.
One distinct difference in a Head First book is that we want you to really get it. And we
want you to finish the book remembering what you’ve learned. Most reference books don’t
have retention and recall as a goal, but this book is about learning, so you’ll see some of the
same concepts come up more than once.
The examples are as lean as possible.
Our readers tell us that it’s frustrating to wade through 200 lines of an example looking for
the two lines they need to understand. Most examples in this book are shown within the
smallest possible context, so that the part you’re trying to learn is clear and simple. Don’t
expect all of the examples to be robust—they are written specifically for learning, and aren’t
always fully functional. That said, for the larger examples we also try to make them fun,
fascinating, and downright cool—something you’d want to show your friends and family.

Operators are NOT
standing by, but you can
get all the code and
samples files you’ll need at
http://wickedlysmart.com/
hflearntocode.

We’ve placed all the example files on the web so you can download them. You’ll find them
at http://wickedlysmart.com/hflearntocode.
The Brain Power exercises don’t usually have answers.
For some of them, there is no right answer, and for others, part of the learning experience
of the Brain Power activities is for you to decide if and when your answers are right. In
some of the Brain Power exercises, you will find hints to point you in the right direction.
Get the code examples, help, and discussion online.
You’ll find everything you need for this book online at http://wickedlysmart.com/
hflearntocode, including code sample files and additional support material.

you are here� xxxiii

how to use this book

You’re going to have to install Python
More than likely your computer is either not going to have Python installed, or it’s not going
to have the right version of Python installed. In this book we use Python 3, which at the
time of writing was version 3.6. So, you’ll need to install version 3.6 or later. Here’s how:

• For macOS, open your browser and enter:
https://www.python.org/downloads

On this page you should see the macOS download links. If not, look
under the Downloads menu on the page.
1. Click the Download button for Python 3.x (where x is the latest
version). Do not download version 2.7.
2. Once the installer is downloaded, open the installation package in
your downloads folder and follow the installation instructions.
3. After you’ve completed the install, navigate to your Applications
folder, under which you’ll find the Python 3.x folder. To test your
installation, double click the IDLE application in the Python 3.x
folder:

Note that you’ll need administrator privileges
to install Python—if you commonly install
apps, you should be fine; otherwise, ask your
administrator for help.

The IDLE app is located in the Python 3.x
folder, which you’ll find in the Applications
folder. We’ll talk more about what IDLE is in
the first chapter.
4. When the IDLE application appears on your screen, you should
see something similar to the screenshot below. If not, recheck your
installation for any errors that might have occurred.

It’s a good idea to add IDLE to your dock if
you haven’t already, as we’ll be using it a fair
bit in this book. To do that control-click the
icon in the dock and choose Options > Keep in
Dock from the pop-up menu.

You can choose the IDLE > Quit IDLE
menu option to exit the application.

xxxiv

intro

the intro

• For Windows, open your browser and enter:
https://www.python.org/downloads

1. Click the Download button for Python 3.x (where x is
the latest version). Do not download version 2.7.
2. Choose to either save or run the executable installer.
If the latter, click to run the installer after you’ve
downloaded it.
3. When you see the installer window appear on your
screen, make sure the “Add Python to PATH”
checkbox is checked at the bottom of the installer, then
click “Install Now.”
4. After you’ve completed the install, navigate using
the Start button to All Programs, and in your list of
apps you should see a menu option for Python 3.x
(with your version number in place of the x). Under
the Python menu you’ll see choices for Python 3.x,
documentation, and IDLE, which is an editor we will
also be using in this book.
5. To test, click the IDLE menu item; when the IDLE
application appears on your screen, you should see
something similar to the screenshot below. If not,
recheck your installation and any errors that might
have occurred.

You can choose the IDLE
> Quit IDLE menu option
to exit the application.
Note to Linux users: We’re not
worried about you; let’s be real,
you know what you’re doing. Just
grab the approriate distribution
from python.org

you are here� xxxv

how to use this book

A word on organizing your code
Your source code is all the code you’ll be writing with this book. We
recommend keeping your code organized on a per-chapter basis and
throughout the book we’ve assumed that you’ll be creating one folder
per chapter, like this:

We created a top-level
folder for all the code
in the book.
code

ch2

ch1

my_file =

my_text = my

print(my_text

my_file.
close(

phraseomatic.py

The first thing
that stands
between you and
VERB_ING your
first, real,
piece of NOUN,
is VERB_ING
the skill of
breaking

dogcalc.py

...

And a folder for each
chapter named ch1, ch2,
and so on, up to ch12.

ch12

The first thing
that stands
between you and
VERB_ING your
first, real,
piece of NOUN,
is VERB_ING
the skill of
breaking

dog.py

The first thing
that stands
between you and
VERB_ING your
first, real,
piece of NOUN,
is VERB_ING
the skill of
breaking

palindrome.py

Thoughout the book we’ll
guide you on what to call
your folders and files.

While you’re there don’t
forget to sign up for
the Head First Learn to
Code newsletter, where
we’ll do our best to keep
your brain engaged in
learning to code!

You should also visit:
http://wickedlysmart.com/hflearntocode

There you’ll find instructions for downloading the complete source
code for the book. In this code you’ll find our versions of the programs
you’re going to write as well as a few data files and images you’ll need.
We do ask that you take the time to type in the programs yourself (this
will help you develop your muscle memory for coding and help things
sink into your brain), but if you run into any issues you just can’t figure
out, you can always compare your code with ours to see where you
might have made a mistake.

xxxvi

intro

the intro

Acknowledgments*
A huge thanks goes out first
to my esteemed technical
reviewers: Elisabeth
Robson carefully and
expertly reviewed the
manuscript with a keen Head
First and computer science
David
Elisabeth Robson Josh Sharfman
eye. Josh Sharfman was
the MVP reviewer who added
depth and quality to every
corner of the book. David Powers, in his usual style,
rigorously scoured the technical text (his Harry Potter knowledge ain’t
too shabby either). And veteran Head First author Paul Barry provided a
much-needed Python critical eye. In addition, my review team (listed on the next
page) was invaluable across every aspect of reviewing the book.

Powers

Paul Barry

My biggest thanks to my editors, Jeff Bleiel, Dawn Schanafelt, and Meghan Blanchette. Meghan was
instrumental in getting this book off the ground, Dawn carefully saw it through its early developmental stages, and Jeff
drove the book through to its publication.
Also a big thanks to the entire O’Reilly team including Susan Conant, Rachel Roumeliotis, and Melanie
Yarbrough. At WickedlySmart, thanks to Jamie Burton for all her help, including early reader surveys and
managing the review team forum. And as always, thanks to Bert Bates and Kathy Sierra for inspiration,
interesting discussion, and all their help solving writing conundrums. Thanks to Cory Doctorow for his support and
for lending his writing to Chapter 7.
Finally, a number of individuals and organizations unknowingly inspired aspects of
this book, including Daniel P. Friedman, Nathan Bergey, the Raspberry Pi
Foundation, and Socratica.

Jamie Burton

Jeff Bleiel
Bert Bates

Kathy Sierra
Paul Barry

Dawn Schanafelt

Dawn Schanafelt
Meghan Blanchette

*The large number of acknowledgments is because we’re testing the theory
that everyone mentioned in a book acknowledgment will buy at least one copy,
probably more, what with relatives and everything. If you’d like to be in the
acknowledgment of our next book, and you have a large family, write to us.

you are here� xxxvii

how to use this book

The Revie w Te am
Meet the review team!
An amazing group of people took on
reviewing this book. With backgrounds
from newbie to expert, and professions as
diverse as architect, dentist, elementary
school teacher, real estate agent, and
AP computer science teacher, they
participated across the globe from Albania
to Australia, from Kenya to Kosovo,
from the Netherlands to Nigeria to New
Zealand.

driguez
Crystal Gilmore Ro

Ridvan Bunjaku

Hudson Read

Mark S. van der Linden

This group read every page, did every
exercise, and entered and executed every
line of code, providing feedback and
encouragement over 600 pages.
They also, on their own, worked as
a team, helping each other through
new concepts, double-checking
errors, and locating problems in the
text and code.

Every reviewer here made
significant contributions
to this book and vastly
improved its quality.

Chris Talent

Andrea Toston

Chris Griggs

Troy Welch

Mitch Johnson

David Oparanti

Thank you!
Johnny Rivera

Benjamin E. Hall

Alfred J. Speller

David Kinoti

Tiron Andric

Michael Peck

Dennis Fitzgerald

Mauro Caser

Abdul Rahman Shaik

Also a big thanks to Christopher Davies, Constance Mallon, and Wanda Hernandez for their
significant contributions to this book.

xxxviii intro

1 thinking computationally

Getting Started

Knowing how to think computationally puts you in
control. It’s no secret the world around you is becoming more connected,
more configurable, more programmable, and more, well, computational. You
can remain a passive participant, or you can learn to code. When you can code,
you’re the director, the creator—you’re telling all those computers what they
should be doing for you. When you can code, you control your own destiny (or
at least you’ll be able to program your internet-connected lawn sprinker system).
But how do you learn to code? First, learn to think computationally. Next,
you grab a programming language so you can speak the same lingo as your
computer, mobile device, or anything with a CPU. What’s in it for you? More
time, more power, and more creative possibilities to do the things you really
want to do. Come on, let’s get started...
this is a new chapter   1

breaking it down

Breaking it down
The first thing that stands between you and writing your first
real piece of code is learning the skill of breaking problems
down into achievable little actions that a computer can do for
you. Of course, you and the computer will also need to be
speaking a common language, but we’ll get to that topic in
just a bit.
Now breaking problems down into a number of steps may
sound like a new skill, but it’s actually something you do
every day. Let’s look at a simple example: say you wanted
to break the activity of fishing down into a simple set of
instructions that you could hand to a robot, who would do
your fishing for you. Here’s our first attempt to do that:

Let’s break the process of catching
fish down into a number of easily
understood steps.

We follow the steps in order.

1

Put worm on hook.

Some steps are simple instructions, or statements
if you will, like “cast line into pond.”

2

Cast line into pond.

3

Watch the bobber until it goes underwater.

4

Hook and pull in fish.

5

If done fishing, then go home; otherwise,
go back to step 1.

The bobber
The hook
The worm
2  Chapter 1

A statement might conditionally wait
before proceeding.
This statement only happens after the bobber has
gone underwater in the previous statement.

Notice that often statements repeat, like
here: if we don’t go home, we instead
go back to the beginning and repeat the
instructions to catch another fish.

Statements can also make
decisions, like is it time to go
home or should we keep fishing?

thinking computationally

You can think of these statements as a nice recipe for fishing.
Like any recipe, this one provides a set of steps that, when
followed in order, will produce some result or outcome (in our
case, hopefully, catching some fish).
Notice that most steps consist of a simple instruction, like
“cast line into pond,” or “pull in the fish.” But also notice that
other instructions are a bit different because they depend
on a condition, like “is the bobber above or below water?”
Instructions might also direct the flow of the recipe, like “if
you haven’t finished fishing, then cycle back to the beginning
and put another worm on the hook.” Or, how about a
condition for stopping, as in “if you’re done, then go home”?
You’re going to find that these simple statements or
instructions are the foundation of coding. In fact, every app
or software program you’ve ever used has been nothing more
than a (sometimes large) set of simple instructions to the
computer that tell it what to do.

Real recipes don’t just tell you what to do, they
also include objects that are used in making
a particular dish (like measuring cups, whisks,
food processors, and of course ingredients).
What objects are used in our fishing recipe?
Circle all the objects in the fishing recipe on
the previous page, and check your answer at
the end of the chapter before moving on.

This is a workbook.
You can write in it; in
fact, we encourage it.

One thing to understand right up front is that computers do exactly what
you tell them—nothing more, nothing less. Look at our recipe for fishing on
the previous page. If you were the robot and you followed these instructions
precisely, what problems might you encounter? Do you think we would really
be successful using this recipe?
A. If there are no fish, you’re
going to be fishing for a very
long time (like, forever).

D. Did we specify what
to do with a fish when we
pulled it in?

B. If the worm falls off the
hook, you’ll never know about
it or replace it.

E. What happened to the
fishing rod?

C. What happens if we run
out of worms?

Can you think
of more issues?

F. _______________
__________________
__________________
__________________
__________________

In case you are
not familiar with
fishing, this is a
bobber, otherwise
known as a fishing
float. When a
fish bites, it goes
underwater.

You’ll find the
answers to
Sharpen Your
Pencil exercises
at the end of
the chapter.
you are here 4  3

recipes, algorithms, and pseudocode

I bought this
expensive technical book
to learn to code and you’re starting by
telling me about recipes? That doesn’t
sound very promising or, well, technical.

Actually, a recipe is a perfectly good way
to describe a set of instructions to a computer. You might
even run into that term loosely used here and there in
more advanced programming books. Heck, you’ll even
find books on common software development techniques
that are called cookbooks. That said, if you want to
get technical we can—a computer scientist or serious
software developer would commonly call a recipe an
algorithm. What’s an algorithm? Well, not much more
than a recipe—it’s a sequence of instructions that solves
some problem. Often you’ll find algorithms are first
written in an informal form of code called pseudocode.
One thing to keep in mind is that, whether you’re talking
about a recipe, pseudocode, or an algorithm, the whole
point is to work out a high-level description of how to
solve a problem before you get into the nitty-gritty details
of writing code that the computer can understand and
execute.

are many
Just as therethe same dish,
recipes for hms, you'll find
with algorit any ways to solve
there are m oblem. Some
the same pr than others.
more tasty

4  Chapter 1

In this book, you’ll hear us interchange all these terms,
where appropriate—and, oh, in your next job interview
you might want to use the term algorithm or even
pseudocode to ensure that larger signing bonus (but there’s
still nothing wrong with the word recipe).

More on in
pseudocode
a bit...

As you’ll see,
this can make
the task of
coding more
straightforward
and less errorprone.

thinking computationally

Code Magnets
Let’s get a little practice with recipes algorithms. We put the Head First Diner’s algorithm for
making an three-egg omelet on the fridge to remember it, but someone came along and
messed it up. Can you put the magnets back in the right order to make our algorithm work?
Note that the Head First Diner makes two kinds of omelets: plain and cheese. Make sure you
check your answer at the end of the chapter.

Rearrange these magnets here
to make the algorithm work.

If the customer ordered cheese:

n top
eese o
Add ch

while eggs aren’t fully mixed:
Transfer eggs to plate

While eggs aren’t fully cooked:

m heat
pan fro
Remove

Stir e
ggs

Heat saute pan
Crac
k th
ree
eggs
into
bowl

Serve

ip
Wh

gs
eg

to pan
r eggs
Transfe

you are here 4  5

the process of coding

How coding works
So you’ve got a task you want the computer to do for you, and you know
you’ll need to break that task down into a number of instructions the
computer can understand, but how are you going to actually tell the computer
to do something? That’s where a programming language comes in—with a
programming language you can describe your task in terms that you and the
computer both understand. But before we take a deep dive into programming
languages, let’s look at the steps you’ll take to actually write code:

1

Craft your algorithm
This is where you take the problem or
task you want solved and turn it into
a high-level recipe, pseudocode, or
algorithm that describes the steps
that need to be performed by the
computer to achieve whatever result
you are after.

2

Put worm on hook.

2

Cast line into pond.

3

Watch the bobber until it goes underwater.

4

Hook and pull in fish.

5

If done fishing, then go home; otherwise, go back to
step 1.

This is the step where
we map out our solution
before doing the hard
work of translating it into
a programming language.

Write your program
Next, you take that recipe and
translate it into a specific set of
instructions that are written in a
programming language. This is the
coding stage, and we call the result a
program or just “your code” (or, more
formally, the source code).

3

1

def hook_fish():
print('I got a fish!')
def wait():
print('Waiting...')

print('Get worm')
print('Put worm on hook')
print('Throw in lure')

while True:
response = input('Is bobber underwater? ')
if response == 'yes':
is_moving = True
print('I got a bite!')
hook_fish()
else:
wait()

Run your program
Finally, you take your source
code and hand it to the computer,
which will start carrying out your
instructions. Depending on the
language you’re using, this process
might be called interpreting ,
running, evaluating, or executing
your code.

e of these
We often use somea
y as well.
terms interchang bl

6  Chapter 1

Get worm

Put worm on hook

Throw in lure
Is bobber underwater?

step
This is the “coding”ur
yo
rn
where you tu de
algorithm into co d for
(which is shorthan is
source code) that in the
ready to execute
next step.

When your source code is
complete, you’re ready to
execute it. If all goes well,
and you designed your code
well, you’ll get the result from
the computer you were looking
for.

thinking computationally

Are we even speaking
the same language?

Relax

Don’t worry; at
this point we don’t
expect you to read
or write code. For gosh
sakes, you’ve got the whole book
ahead of you—for now we’re just
getting familiar with code, what it
looks like, and how it works. The
important thing in this chapter is
just to take it all in.

Think of a programming language as a special-purpose language
created expressly for specifying tasks to a computer. Programming
languages give you a way to describe your recipes in a manner that is
clear and precise enough that a computer can understand it.
To learn a programming language there are two things you need to
nail down two things: what things can you say using the language, and
what do those things mean? A computer scientist would call these the
syntax and semantics of the language. Just stash those terms in the
back of your brain for now; we’ll be getting you up to speed on both
as the book progresses.
Now, as it turns out, just like spoken languages, there are many
programming languages, and, as you may already have figured out, in
this book we’re going to use the Python programming language. Let’s
get a little better feel for languages and Python...

You’ll find that the techniques you’re
learning in this book can be applied to
just about any programming language
you might encounter in the future.

YOU SAY TOMATO
On the left you’ll find some statements written in English, and on the right you’ll find statements written in a
programming language. Draw a line from each English statement to its corresponding code translation. We did the
first one for you. Make sure you check your work with the solution at the end of the chapter before proceeding.
Print “Hi there” on the screen.

for num in range(0, 5):
pour_drink()

If the temperature is more
than 72, then print “Wear
shorts” on the screen.

name = input('What is your name? ')

A grocery list with bread,
milk, and eggs on it.

if temperature > 72:
print('Wear shorts')

Pouring five drinks.

grocery_list = ['bread', 'milk', 'eggs']

Ask the user, “What is your name?”

print('Hi there')

you are here 4  7

a quick survey of programming languages

The world of programming languages
If you’re reading this book you may have, in passing, heard about various programming languages.
Just walking through the programming section of your local bookstore you might encounter Java,
C, C++, LISP, Scheme, Objective-C, Perl, PHP, Swift, Clojure, Haskell, COBOL, Ruby, Fortran,
Smalltalk, BASIC, Algol, JavaScript, and of course Python, to name just a few. You might also be
wondering where all these names came from. The truth is, programming language names are a lot
like the names of rock bands—they’re names that meant something to the people who created the
language. Take Java, for instance: it was named, not surprisingly, after coffee (the preferred name
Oak was already taken). Haskell was named after a mathematician, and the name C was chosen
because C was the successor of the languages A and B at Bell Labs. But why are there so many
languages and what are they all about? Let’s see what a few folks have to say about the languages
they use:

I’m for Objective-C all
the way. I build iPhone apps
all day long, and I love the way
Objective-C is like C, but is much
more dynamic and object-oriented.
I’m also learning Apple’s new
language called Swift.

Java keeps me thinking at
the level of objects, not low-level
code, and it takes care of a lot of lowlevel things, like memory management
and threading, for me.

I live in the world of
WordPress, which is written in
PHP, so PHP is my go-to language.
Some people call it a scripting language,
but it does everything I need.

8  Chapter 1

thinking computationally
I mostly use the
C programming language. I
write parts of operating systems
that have to be super efficient. In
my job, every CPU cycle and every
memory location counts.

Call me academic, but I
love Scheme- and LISP-style
languages. For me it is all about high-order
functions and abstraction. I’m glad to see
functional languages like Clojure getting
real industry use.

I’m a web developer
and JavaScript is my main
language. It’s the de facto language of
all browsers, and it’s being used to write
backend web services as well.

I’m a system administrator. I
use Perl quite a bit to write various
system scripts. It’s terse, but also very
expressive. Just a little code gets a lot
done.

you are here 4  9

why we chose python

We love Python.
It’s known for being a
very readable, clean language with
great library support that lets you
write code for all kinds of domains;
it also has a great community of
people involved in it.

Python’s
known as one of the best languages
for beginners, but it also grows with you
as your skills mature. And it’s a real language
too, with folks like Google, Disney, and NASA
using it to build serious systems.

Choices, choices...
As you can see, there are a lot of languages and opinions out there, and we’ve barely scratched
the surface of modern languages. You can also see there’s a lot of terminology that comes along
with these languages, and, as you progress, those terms are going to make a lot more sense to
you. For now, just know there’s a wide variety of languages out there, with more being created
every day.
So, in this book, what should we use? Here’s the thing: first and foremost, we want to learn how
to think computationally—that way, no matter what language you run across in the future, you’ll
be in a good position to learn it. But, that said, we have to start with some language, and, as you
already know, we’re going to use Python. Why? Our friends above said it well: it’s considered
one of the best languages for beginners because it’s such a readable and consistent language.
It’s also a powerful language in that no matter what you want to do with it (now or beyond this
book), you can find support in terms of code extensions (we call them modules or libraries) and a
supportive community of developers to give you a hand. Finally, some developers will even tell
you Python is just more fun than other languages. So how can we go wrong?

10  Chapter 1

thinking computationally

Look how easy it is
to write Python

You don’t know Python yet, but we bet you can make some good guesses about how
Python code works. Take a look at each line of code below and see if you can guess
what it does. Write in your answers below. If you get stuck, the answers are on the
next page. We did the first one for you.

customers = ['Jimmy', 'Kim', 'John', 'Stacie']

Make a list of customers.

winner = random.choice(customers)

flavor = 'vanilla'

print('Congratulations ' + winner +

' you have won an ice cream sundae!')

prompt = 'Would you like a cherry on top? '

wants_cherry = input(prompt)

order = flavor + ' sundae '
if (wants_cherry == 'yes'):
order = order + ' with a cherry on top'
print('One ' + order + ' for ' + winner +
' coming right up...')

Python Output

an ice cream sundae!
Congratulations Stacie you have won
Would you like a cherry on top? yes
on top for Stacie coming
One vanilla sundae with a cherry
right up...

This should help; it’s the output of this
code. Do you think this code has the
same output every time you run it?

you are here 4  11

your first exposure to python

Solution

Look how easy it is
to write Python

You don’t know Python yet, but we bet you can make some good guesses about how
Python code works. Take a look at each line of code below and see if you can guess
what it does. Write in your answers below. If you get stuck, the answers are on the
next page. We did the first one for you.

customers = ['Jimmy', 'Kim', 'John', 'Stacie']

winner = random.choice(customers)

flavor = 'vanilla'

print('Congratulations ' + winner +

' you have won an ice cream sundae!')

prompt = 'Would you like a cherry on top? '

wants_cherry = input(prompt)

order = flavor + ' sundae '

Make a list of customers.
Randomly choose one of those customers.
Set the name or variable called flavor to the text ‘vanilla'.
Print out a congratulations message to the screen that includes
the winning customer’s name. For instance, if Kim is the winner
this code prints “Congratulations Kim you have won an ice cream
sundae!”
Set the name or variable called prompt to the text “Would you
like a cherry on top? ”
Ask the user to type in some text, and assign it to wants_
cherry. Notice that when the user is asked for input, the
prompt is first displayed (as seen in the Python output).

if (wants_cherry == 'yes'):

Set order to the text ‘vanilla' followed by ‘sundae'.
If the user answered yes to ‘Would you like a cherry on top?',
then add the text “ with a cherry on top” to the order.

print('One ' + order + ' for ' + winner +

Print out that the winner’s order is coming right up.

order = order + ' with a cherry on top'

' coming right up...')

Python Output
Congratulations Stacie you have won an ice cream sundae!
Would you like a cherry on top? yes
One vanilla sundae with a cherry on top for Stacie coming
right up...

12  Chapter 1

P.S. If you can’t help yourself and you have to type
this code in, add import random at the very top of
the file before running it. We’ll get to what that does
later, but note that running the code at this point is
not required or all that useful. That said, we just know
someone’s going to have to try it. You know who you are!

thinking computationally

How you’ll write and run code with Python
Now that we’ve talked about and even looked at some code, it’s time to start thinking about how
you’d actually write and execute some real code. As we mentioned, depending on the language
and environment, there are a lot of different models for how you do that. Let’s get a sense for
how you’re going to be writing and running your Python code:

1

Writing your code

Your code

First you get your code typed into an editor and
saved. You can use any text editor, like Notepad on
Windows or TextEdit on the Mac, to write your Python
code. That said, most developers use specialized
editors known as IDEs (or Integrated Development
Environments) to write their code. Why? IDEs are a
bit like word processors—they give you lots of nice
features like autocompletion of common Python
keywords, highlighting of the language syntax (or of
errors), as well as built-in testing facilities. Python also
conveniently includes an IDE called IDLE, which we’ll
be looking at shortly.

This is Python’s
IDLE editor.

Your code
2

Running your code
Running your code is as easy as handing it to
the Python interpreter, a program that takes
care of everything needed to execute the code
you’ve written. We’ll step through the details of
this in a bit, but you can access the intepreter
through IDLE, or directly from your computer’s
command line.

The Python intepreter
3 How your code is interpreted
We’ve been describing Python as a language that
you and the computer both understand. And,
as we’ve learned, an interpreter does the job of
reading your code and executing it. To do this, the
interpreter actually translates your code behind the
scenes into a lower-level machine code that can
be directly executed by your computer hardware.
You won’t need to worry about how it does this;
just know that the interpreter will do the job of
executing each statement of your Python code.

Get worm

Put worm on hook
Throw in lure

Is bobber underwater?

The
interpreter
running your
code

you are here 4  13

more about python and programming languages

Q:

Why not use English to program computers? Then we
wouldn’t need to learn these special-purpose programming
languages.

A:

Yes, wouldn’t that be nice. As it turns out, English is full of
ambiguity, which makes creating such a translator extremely difficult.
Reseachers have made small inroads in this area, but we’re a
long way from using English, or any other spoken language as
a programming language. Also, in the few languages, that have
tried to be more English-like, we’ve found that programmers prefer
languages that are less like a spoken language and that are more
streamlined for coding.

Q:
A:

Why isn’t there just one programming language?

Technically all modern programming languages are equivalent
in that they can all compute the same things, so, in theory, we
could use one programming language to serve all our needs. Like
spoken languages, however, programming languages differ in their
expressive power—that is, you’ll find some types of programming
tasks (say, building websites) are easier in some languages than
others. Other times the choice of a programming language just
comes down to taste, using a particular methodology, or even the
need to use the language that your employer has adopted. Count
on one thing, though: there will be even more languages, as
programming languages continue to evolve.

Q:
A:

Is Python just a toy language for beginners? If I wanted to
be a software developer, is learning Python going to help me?

Python is a serious language and used in many products
you probably know and love. In addition, Python’s one of the
few professional languages that is also considered an excellent
language for beginners. Why? Compared to many existing languages,
Python approaches things in a straightforward and consistent way
(something you’ll have a better understanding of over time, and as
you gain experience with Python and other languages).

Q:

What’s the difference between learning to code and
thinking computationally? Is the latter just a computer science
thing?

A:

Computational thinking is a way of thinking about problem
solving that grew out of computer science. With computational
thinking we learn how to break problems down, to create algorithms
to solve them, and to generalize those solutions so we can solve
even bigger problems. Often, though, we want to teach a computer
to execute those algorithms for us, and that’s where coding comes
in. Coding is the means by which we specify an algorithm to a
computer (or any computational device, like your smartphone). So
the two really go hand in hand—computational thinking gives us a
way to create solutions to problems that we want to code, and coding
provides a means of specifying our solutions to a computer. That said,
computational thinking can be valuable even if you aren’t coding.

Where do you think the name Python most likely came from?
Check the most likely answer below:
A. Python’s creator loved snakes,
and had previously created the less
successful Cobra language.
B. The creator of Python loved pie…
thus, Pie-thon.

C. The Python name is inspired by the name of
a British surreal comedy group.
D. Python is an acronym for Programming Your
Things, Hosted On the Network
E. The Python name was inspired by the
Anaconda runtime system, on which it is built.

C. Monty Python is the British comedy group who created the show Monty Python’s Flying Circus. Feel free to watch a few of their
shows as homework—whether it is your cup of tea or not, you’ll often see Monty Python references in the Python community.

14  Chapter 1

thinking computationally

A very brief history of Python

Over in the Netherlands, at the
National Research Institute for
Mathematics and Computer
Science, they had a big problem:
their scientists found programming
languages difficult to learn. Yes,
even to these highly educated,
skilled scientists, the most current
programming languages were
confusing and inconsistent. To the
rescue, the Institute developed a new
language called “ABC” (you thought
we were going to say “Python,”
didn’t you?), which was designed
to be much easier to learn. While
ABC was somewhat successful, an
enterprising young developer named
Guido van Rossum, after a weekend
of binge-watching Monty Python
reruns, thought he could take things
further—so, using what he’d learned
from ABC, Guido created Python.
And the rest is history.

Python 3.0

Python 2.0

Python 1.0

Python came of age with version
2.0 and a whole new set of features
aimed at supporting its growing
community of developers. For
instance, in recognition that Python
was truly now a global language, 2.0
was extended to handle character
sets from languages far beyond
the typical English letters. Python
also also improved many technical
aspects of the language, like better
handling of computer memory as
well as better support for common
types of data like lists and character
strings.
The Python development crew also
worked hard to make Python open
to a whole community of developers
who could help improve the
language and implementation.

Actually, we just ma
about the binge weekde up the part
end.

Nobody’s perfect, and there
came a time when Python’s
creators looked back and saw a
few things in Python they wanted
to improve. While Python was
known for its adherence to
keeping things straightforward,
experience with the language had
revealed a few parts of its design
that could be improved, and a
few things that hadn’t aged well
that needed to be removed.
All these changes meant that
some aspects of Python 2 would
no longer be supported. That
said, Python’s creators made sure
there were ways to keep the 2.0
code running. So, if you have
code written in Python 2, don’t
worry—Python 2 is still alive and
well, but know that Python 3 is
the future of the language.

We fully expect our
flying car is going to
be Python enabled.

Note from editor: umm, “the rest” s.is
actually in the next few paragraph

1994

2000

2008

The Future!
you are here 4  15

the versions of python

So, it sounds like there are two versions of
Python out there, version 2 and version 3. Which
are we going to use and how different are they?

Good question. And you’re right, there are two
versions of Python—to be a little more specifc, at
the time this book was printed, the current versions
are 3.6 and 2.7.
Here’s how we’re going to think about the versions.
As it turns out, when you view the two languages,
from, say, 10,000 feet, they are remarkably similar,
and in fact you might not even be able to see any
differences. That said, there are differences, and
if you aren’t paying attention to the version you
are using, they could trip you up. We’ll be using
the newest version of Python in this book—that
is, version 3. We’d rather get you started with the
version that is going to carry on into the future.
Now there is a lot of code in the world already
written in Python 2, and you certainly might
encounter it in a module you’ve downloaded online
or, if you happen to become a software developer,
as part of some old code you become responsible
for. After this book you’ll be in a good position to
study the small differences between Python 3 and 2,
should you need to.

When we say Python 3 or 2, we
mean in each case, the latest
version of 3 or 2 (at the time of
writing, 3.6 and 2.7).
16  Chapter 1

thinking computationally

Got Python?

We’re not going to get much further if you haven’t
installed Python yet. If you haven’t taken the time to install Python,
now is the time. Check out the “You’re going to have to install Python”
section in the introduction to get started. Remember, if you’re using a Mac
or Linux there’s a good chance you already have Python installed; however,
there’s also a good chance you have version 2, not version 3. So whether
you’re on a Mac, a Windows, or a Linux machine, at this point you may
need to install version 3 of Python.
So, go do that and once you’ve got Python up and running, you’re ready to
dive into some real code.

you are here 4  17

a small test program

Putting Python through its paces
Now that you’ve got Python installed, let’s put it to work. We’re going to
start with a small test program just to make sure everything is up and
running. To do that, you need to use an editor to enter your program
before you can execute it; that’s where IDLE, the Python editor (or
Integrated Development Environment if you prefer), comes in. Go ahead
and open IDLE like we did in the introduction. As a reminder, on the Mac
you’ll find IDLE under the Applications > Python 3.x folder. On
Windows navigate using the Start button to All Programs, and you’ll find
IDLE under the Python 3.x menu option.

When you first open IDLE, it
displays an interactive interpreter
called the Python Shell. If you’re
really curious, type 1+1 (that’s
one plus one) into it and hit return.
More on this in the next chapter.

When you first run IDLE it displays an interactive window, called the
Python Shell, that you can type Python statements directly into. You can
also use an editor to type your code into; to do that, choose File > New
File from the IDLE menu. You’ll see a new, empty editor window appear.

Use the File > New File menu
option to get a new window
ready for entering Python code.

18  Chapter 1

thinking computationally

IDLE works just like a word processor, only it knows about Python code
and assists you by highlighting Python language keywords, helping you with
formatting, and even, when appropriate, autocompleting common Python
keywords to make code entry even easier.

After choosing New
File, you should see a
new, empty window
on top of the Python
Shell window.

After creating a new file and getting a new blank editing window, you’re going to
type in one line of code to test things out; go ahead and enter this code:
print('You rock!')

IDLE uses color
to highlight the
different aspects
of your Python code.
You will not be able
to see this if you’re
viewing this page in
black and white.

Notice your file won’t
have a meaningful name
until you save it.

Also, depending on your
version and configuration
of IDLE, you may see
different colors, font
sizes, and so on than you
see here.

Pay careful attention to
every detail of spelling and
punctuation, as Python and
other programming languages
are quite intolerant of
mistakes.

you are here 4  19

how to save your programs

Saving your work
Now that you’ve typed in your first line of code, let’s save it. To do that just
choose File > Save from the IDLE menu:

Before we execute our code we need to
save it. In IDLE, the Save menu item is
under the File menu. Simply choose this
option and give your source code a name.
When we’re writing Python code, we use
an extenion of “.py” on the end of the
filename.
Source code, source file,
code, and program are all
common names for your
files with code in them.

And then give your code a name with a .py extension. We chose the name
rock.py. Note that, although we didn’t show it, we also created a folder for
Chapter 1 code, called ch1, and we recommend you do too.

You’ll want to follow the same code
organization scheme we do in this
book. If you haven’t done so already,
take a look at page xxxvi in the
book’s introduction.

Finally, when you have a folder for
your code and a name, click Save.

20  Chapter 1

thinking computationally

A Test Drive

This is where it all comes together. After you’ve got the code saved,
choose the Run > Run Module menu item. After that, look at the
Python Shell window and you should see your program’s output there.

Execute your code by using
the Run Module menu item
in the Run menu. If you
haven’t saved your code,
IDLE will ask you to.

Congrats on coding your first Python program!
You’ve installed Python, you’ve entered a short bit of real Python code using IDLE,
and you’ve even executed your first Python program. Now it’s not a very complex bit
of code, but you’ve gotta start somewhere. And the good news is, we’re all set up and
ready to move on to a serious business application!

run your code,
Every time yostuart the Python
IDLE will re and then execute
er
, interpret
ut
tp
ou
e
th
's
re
e.
he
d
And
your co
!
just as we expected

you are here 4  21

more about python and output

Did you get something other than “You rock!”?
The process of writing and testing code can be error-prone. If you
weren’t successful on your first try, get used to it: all of us developers
are constantly fixing errors in our code. Here are a few things to try:
• If Python reported an error like invalid syntax check your code for incorrect
punctuation, like missing parentheses. If you look at the code highlighting in IDLE,
usually you can quickly determine where these errors are.
• If Python reported an error like Python NameError: name 'prin' is not
defined check your code for typos, like a name that isn’t spelled correctly—in
this case, the word print.
• If Python reported an error like EOL while scanning that typically means you
left out a single quote on one of the 'You rock' string of characters—make
sure you’ve surrounded it with two single quotes, like 'You rock!'.
• If you’re really stuck, try the book’s community page on wickedlysmart.com.

Q:
A:

Why do we use Run Module to execute our code?

Python calls a file of Python code a module. So, all this means
is “run all the Python code in my file.” Modules are also a way to
further organize code that we’ll get to later.

Q:
A:

What exactly do you mean by input and output?

Right now we’re just dealing with simple kinds of input and
output. Our output is the text that is generated from your program
and displayed in the Python Shell window. Likewise, input is any text
your program gets from you in the shell window. More generally, all
kinds of input and output are possible, such as mouse input, touch
input, graphics and sound output, and so on.

Q:

I’ve figured out that print gives you a way to print text
to the user, but why the name print? When I first saw that, I
thought it was something for the printer.

A:

Think way, way back to when computers were more likely
to output to a printer than a screen. Back in those days, the name
print made a lot more sense than it does now. Of course, Python is

22  Chapter 1

young enough to not have an excuse for using the name print, but
print has traditionally been the way to output to the screen in a lot
of languages. So, print outputs to the screen (by way of the Python
Shell window), and on a related note, you saw an example earlier of
input, which gets user input from the Python Shell.

Q:
A:

Is print the only way to output from Python?

No, it’s just the most basic way. Computers and programming
languages excel at output (and input), and Python is no exception.
Using Python (or most languages), you can output to web pages, the
network, files on storage devices, graphics devices, audio devices,
and many other things.

Q:

Okay, well, when I use print, I type something like print('hi
there'). What exactly is going on there?

A:

What you’re doing is using some functionality built into Python
to print. More specifically, you’re asking a function called print to
take your text in quotes and output it to the Python Shell. Now, we’re
going to be getting more into exactly what functions are, what text
is, and so on later, but, for now, just know you can use this function
anytime you want to print to the shell.

thinking computationally

Python Exposed

This week’s interview: are you serious?
Head First: Welcome, Python! We’re looking forward to
digging in and finding out what you’re all about.
Python: Glad to be here.
Head First: Let’s see, you’re named after a comedy
troupe, and you’re known for being a beginner’s language;
honestly, should anyone take you seriously?
Python: Well, let’s see, I’m used for everything from
running silicon chip production lines to applications that
help create major motion pictures (you didn’t hear it
from me, but can you say, “George Lucas”?) to powering
interfaces for air traffic control systems. I could go on...
does that sound serious?
Head First: Well, then, if you’re such a serious language,
how on earth can newbies easily use you? I mean, the
projects you’re describing sound quite complex. Would
you have us believe it doesn’t take a hardcore, complex
language to pull all that off ?
Python: One of the reasons newbies and professionals
appreciate me is because my code is quite straightforward
and readable. Ever look at a language like, say, Java?
Blech. My gosh, the effort you have to go to just to say,
“Hello World!” That takes a single line of Python code.
Head First: Okay, you’re readable, great, but what
exactly does that mean, anyway?
Python: Given I mentioned Java, let me just give you a
little example. Let’s say you want to tell your user “Hello!”
Here’s how you do it with Java:
class HelloWorldApp {
public static void main(String[] args) {
System.out.println("Hello!");
}
}

That’s a lot to take in. I’d call it totally unreadable,
especially to someone just learning to program. What
the heck does all that mean, anyway? Is all that really
necessary? Now let’s look at my version, which I’ve written
in Python of course:

print('Hello!')

I think you’d have to agree that is more straightforward
and readable—anyone can look at that line and have
a decent idea of what it does. But that’s just a simple
example. Overall, Python strikes people as clear, almost
English-like, and consistent...
Head First: Consistent? What does that mean?
Python: One way to think about consistency is that
there aren’t a lot of surprises in the language. In other
words, once you understand a bit of the language, other
things tend to work as you might guess, or expect. Not all
languages are like that.
Head First: I want to go back to something you said
earlier—you mentioned some rather esoteric examples,
like air traffic control, chip manufacturing, being the main
software for the space shuttle, and so on. That all sounds
very industrial and special purpose. I’m not sure Python is
going to be the best language for our readers.
Python: The space shuttle? You made that up. For
the others, I was giving you examples of things you
might consider serious, given that you claimed Python was
otherwise. Some of the most common uses of Python are
for things like creating websites, writing games, and even
creating desktop apps.
Head First: Can we switch gears? Someone just handed
me a note: our sources tell us that there are actually two
versions of Python, and what’s more, they are actually…
gosh, how do I say it, incompatible with each other. How on
earth is that being consistent?
Python: Like anything, languages tend to grow and
evolve, and yes, there are two version of Python, version
2 and version 3. Version 3 has new things in it that
were not part of version 2, but there are ways to make
things backward compatible. Let me walk your readers
through…
Head First: ...on that note, we’re out of time. We look
forward to our next ambush, er, I mean opportunity, to
speak with you.
Python: Thanks, my pleasure…I think.

you are here 4  23

writing a serious application

Try my new Phrase-O-Matic
and you’ll be a slick talker just like
the boss or those guys in marketing...

1

# let python know we’ll be using some random
# functionality by importing the random module
import random

2
Okay, it’s time to get serious and
to write a real-world business
application using Python. Check out
this Phrase-O-Matic code—you’re
going to be impressed.

# make three lists, one of verbs, one of adjectives,
# and one of nouns
verbs = ['Leverage', 'Sync', 'Target',

'Gamify', 'Offline', 'Crowd-sourced',
'24/7', 'Lean-in', '30,000 foot']

adjectives = ['A/B Tested', 'Freemium',

'Hyperlocal', 'Siloed', 'B-to-B',

Relax

Yes, we mean
it! Your job right
now is to take it all
in by osmosis. Take a
e,
at
look every line of cod
what
read the description of
r in
it does, and let it registe
some
your brain. We’ve added
are
ich
wh
,
nts
me
Python com
rks,
just notes after the # ma
a of
so you have a better ide
nk of
Thi
s.
doe
e
cod
s
thi
at
wh
l
pfu
hel
as
those comments
’ll
We
nt.
wa
you
if
e
cod
pseudo
bit
a
nts
talk about comme
you
later in the book. When
e
think you’ve made som
on to
sense of the code, move
walk
’ll
we
and
e
pag
t
nex
the
through it in more detail.

24  Chapter 1

'Oriented', 'Cloud-based',
'API-based']

nouns = ['Early Adopter', 'Low-hanging Fruit',

'Pipeline', 'Splash Page', 'Productivity',
'Process', 'Tipping Point', 'Paradigm']

3

# choose one verb, adjective, and noun from each list
verb = random.choice(verbs)

adjective = random.choice(adjectives)
noun = random.choice(nouns)

4

# now build the phrase by "adding" the words together

5

# output the phrase

phrase = verb + ' ' + adjective + ' ' + noun

print(phrase)

thinking computationally

Phrase-O-Matic
In a nutshell this program takes three lists of words, randomly picks one
word from each list, combines the words into a phrase (suitable for your
next startup’s slogan), and then prints out the phrase. Don’t worry if
you don’t understand every aspect of this program; after all, you’re on
page 25 of a 600-page book. The point here is just to start to get some
familiarity with code:

1

2

The import statement tells Python we’re going to be using
some additional built-in functionality that’s in Python’s random
module. Think of this as extending what your code can do—in
this case, by adding the ability to randomly choose things. We’ll
get into the details of how import works later in the book.
Next we need to set up three lists. Declaring a list is
straightforward—just place each item in the list within quotes
and surround it by square brackets, like this:
verbs = ['Leverage', 'Sync', 'Target',

'Gamify', 'Offline', 'Crowd-sourced',
'24/7', 'Lean in', '30,000 foot']

Notice we’re assigning each list to a name, like verbs, so we can
refer to it later in the code.

3

Next we need to choose one word randomly from each list. To
do that we’re using random.choice, which takes a list and
randomly chooses one item. We then take that item, and assign it
to the corresponding name (verb, adjective, or noun) so we
can refer to it later.

4

We then need to create the phrase, and we do this by gluing the
three items together (the verb, adjective, and noun)—in Python
we can glue these together using the plus sign. Notice also that
we have to insert spaces between the words; otherwise, we’d end
up with phrases like “Lean-inCloud-basedPipeline.”

5

Finally, we output the phrase to the Python Shell with the print
statement and...voila! We’re in marketing.

Slogans for Your Next
Python-based Startup
24/7 Freemium Productivity
Lean-in Hyperlocal Splash Page
Gamify Siloed Early Adopter
Offline API-based Process
Crowd-sourced Cloud-based
Pipeline

random.choice is another built-in
function from Python. We'll learn
more about these later in the book.

e to
Computer scientists lik
er
th
call gluing text toge a pretty
“concatenation.” It’su’ll see it more
handy word, so yo
in this book.

you are here 4  25

how to enter the phrase-o-matic code

Getting the code in the machine
You’ve entered one program into IDLE, but let’s step through it again—we’re going to
create a new Python file by choosing File > New File from the menu. Go ahead and
do that, and then type the code from two pages back into the editor.

Here’s an
example of a
typo; notice
how it is
highlighted.
If you get
this, doublecheck that
your code is
correct.

Pay special attention to the words
and punctuation that are being used.
We’ll get to all this detail in later
chapters, but for now it’s great to
get accustomed to it.

Note that IDLE highlights common
errors. You might see errors
highlighted while you’re typing your
code or when you run your code,
depending on the error. Here a
missing quote error is highlighted.
If you see any errors, you'll want to
double-check the code and fix them.
We purposefully added
error here to show
an
Now that you’ve typed in all the code, let’s save it. Remember, to do that just choose
IDLE does; as long
what
File > Save from the IDLE menu and then save your code as phraseomatic.py.
as you faithfully type in
the code from two pages
back, you should not see
this error.
Remember, IDLE
will color your code
according to its function.

IDLE will also insert indenting
when appropriate after you hit
the Return key.
Notice that extra whitespace—that is,
spaces and return characters—can be used
to make your code more readable; Python
largely ignores whitespace (we’ll talk about
an exception to that later).
26  Chapter 1

thinking computationally

A Test Drive

Now it’s time to run the Phrase-O-Matic. Let’s explicitly go through this once
more to really set it in—after you’ve got the code saved, choose the Run >
Run Module menu item. After that, look for the output in the shell window, and
get ready for your next startup’s slogan!

Remember, to execute your
code, use the Run Module
menu item in the Run menu.
If you haven’t saved your
code, IDLE will ask you to.

We’re going to Gamify A/B
Tested Splash Pages!

We ran Phrase-O-Matic a few
times; check out what we got.

To re-run Phrase-O-Matic, first click on your code window, then
choose Run > Run Module again.
you are here 4  27

a review of computational thinking and coding

 To write code, you first have to break a
problem down into a simple set of actions
that solves the problem.
 We call this set of actions an algorithm, or
less formally a recipe for solving a problem.
 Actions are in the form of statements that
can carry out very simple tasks, make
decisions, or control the flow of an algorithm
by repeating parts of the code.
 Computational thinking is a way of thinking
about problem solving that grew out of
computer science.
 Coding is the act of taking an algorithm and
translating its steps into a programming
language that can be executed on a
computer.
 Algorithms are sometimes expressed in a
more human-readable pseudocode before
being translated to an actual programming
language.
 Programming languages are special-purpose
languages created expressly for describing
tasks to computers.
 English makes for a poor programming
language because it is quite ambiguous.

28  Chapter 1

 There are many programming languages,
each created with inherent advantages
and disadvantages, but all have the same
computational power.
 The Python name doesn’t come from a
snake, but from the creator’s love of the
Monty Python comedy troupe.
 New and experienced programmers
appreciate Python’s clean and consistent
design.
 There are two versions of Python, 2 and 3,
and this book focuses on Python 3 (although
the differences are minor in a lot of cases).
 Python code is executed by an interpreter,
which translates high-level Python code into
low-level machine code that your computer
can execute directly.
 Python provides an editor called IDLE that is
tailored to writing Python code.
 You can use whitespace in a Python program
to increase readability.
 input and print are two functions provided
by Python for simple, shell-based input and
output.

thinking computationally

Coding Crossword
Let’s give your right brain something to do.
It’s your standard crossword, but all of the
solution words are from Chapter 1.
1

2

3
4

5

6

7

8

9

10
11

12

13
14

15

16

17

Across

Across

1. Layman’s term for algorithm.
Layman5.term
forIDE.
algorithm.
Python
Python 6.
IDE
Input to interpreter or compiler.
Language named
for coffee.
Input to7.interpreter
or compiler.
8. Human-readable
code.
Language
named for coffee.
Flying _______.
Human 10.
readable
code.
11.
Technical
Flying __________.name for recipe.
12. One of best beginner languages.
Technical
name for recipe.
14. Low-level code.
One of best
beginner
16. Python
is thislanguages.
type of language.
Low level
code.
17. Python is one of them.

1.
5.
6.
7.
8.
10.
11.
12.
14.
16. Python is this type of language
17. Python is one of them.

Down
Down

2. Running a program.

2. Running
a word
program.
3. Another
for source code.
3. Another
forthis
source
4. Kind ofword
thinking
bookcode.
teaches.
9. Tooofambiguous
for code.
4. Kind
thinking this
book teaches.
10. Supportive
side
Python.
9. Too
ambiguous
forofcode.
13.
Head
First
Diner
serves
them.
10. Supportive side of Python.
15.
One
company
using
Python
13. Head First Diner serves them.
15. One company using Python.

you are here 4  29

exercise solutions

Real recipes don’t just tell you what to do, they also include objects that are used in
making a particular food (like utensils, appliances, and ingredients). What objects are
used in our fishing recipe?

1

Put worm on hook.

2

Cast line into pond.

3

Watch the bobber until it goes underwater.

4

Hook and pull in fish.

5

If done fishing, then go home; otherwise, go back to step 1.

One thing to understand right up front is that computers do exactly what you tell them—nothing more,
nothing less. Look at our recipe for fishing on page 2. If you were the robot and you followed these
instuctions precisely, what problems might you encounter? Do you think we would really be successful using
this recipe?

A. If there are no fish, you’re
going to be fishing for a very
long time (like, forever).
B. If the worm falls off the
hook, you’ll never know about
it or replace it.
C. What happens if we run
out of worms?

It looks like the
answer was “all
of the above.”

30  Chapter 1

D. Did we specify what to do with a fish when we pulled it in?
E. What happened to the fishing rod?
F. Are

there any specifics of what a good cast is? If the worm
lands on a lily pad, do we need to redo the cast?
Typically when the bobber goes underwater, you try to “hook” the
fish before pulling it in. This doesn’t say anything about that.
How do we know if we’re done fishing? By the time? When we’re
out of worms? Something else?
There’s a lot we assumed in our recipe; we’re sure you probably
thought of a lot of other instructions that were unspecificed in
addition to these.

thinking computationally

Code Magnets Solution
We put the Head First Dinner’s recipe algorithm for making an three-egg omelet on the fridge
to remember it, but someone came along and messed it up. Can you put the magnets back in
the right order to make our algorithm work? Note that the Head First Dinner makes two kinds
of omelets: plain and cheese.

There are several correct variations you
could come up with; just make sure you
understand our solution, and that yours,
if different, makes logical sense.

Here are the unscrambled magnets!
Heat saute pan
Crack three eggs into bowl
while eggs aren’t fully mixed:

Notice we indented whipping the eggs to indicate this
is done as long as the eggs aren’t yet whipped. If you
indicated this in another way, that is fine.
Next we move the eggs to
the preheated pan.

Whip eggs

Transfer eggs to pan
While eggs aren’t fully cooked:

Stir eggs
If the customer ordered cheese:

Add cheese on top

Remove pan from heat

Transfer eggs to plate
Serve

We start by setting everything up:
heating the pan, cracking the eggs.
til they
Then we whip the eggs un
are thoroughly mixed.

And cook them until they are done.
Notice we just indented stirring the eggs to indicate
the eggs are stirred as long as they aren’t done. If
you indicated this in another way, that is fine.
If the customer wanted cheese,
then add it.
We also indented adding cheese, because that is only
done if the customer ordered cheese.

And in either case, remove the pan
from the heat, and transfer it to
a plate.

Finally, we serve the omelet.

you are here 4  31

exercise solutions

YOU SAY TOMATO
solutioN
On the left you’ll find some statements written in English, and on the right you’ll find statements written in a
programming language. Draw a line from each English statement to its corresponding code translation. We did
the first one for you.
Print “Hi there” on the screen.

for num in range(0,5):
pour_drink();

If the temperature is more
than 72, then print “Wear
shorts” on the screen.

name = input('What is your name? ')

A grocery list with bread,
milk, and eggs on it.

if temperature > 72:
print('Wear shorts')

Pouring five drinks.

grocery_list = ['bread', 'milk', 'eggs']

Ask the user, “What’s your name?”

print('Hi there')

1
3

6

Coding
Crossword
Solution

7

J

A
9

P

4

R
S O U R C E

V

D

L

U

P

T

8

A

A

P

S

E

U D O C O D E
T

10

A
L G O R

I

R C U S

N E

C O D E

O
M

L

O

M

12

P

Y

N

I

C

I
13

T H O N
14

M A

I

E
C

T H M

H

E
X

G

16

G H

T E

L

R P

E

V

R E

E

U
C H

L

Across
1. Layman term for algorithm. [RECIPE]
5. Python IDE [IDLE]
6. Input to interpreter or compiler.

I

15

I

I

L

T

S

E

Y

N

T E

R

S

32  Chapter 1

I

2

M

I

I

P

R

S

17

I

G

11

N

5

C

C

C O D E

M
E

R E

E
Y

Down
2. Running a program. [EXECUTE]
3. Another word for source code. [PROGRAM]
4. Kind of thinking this book teaches.

2

simple values, variables, and types

Know Your Value

Computers really only do two things well: store values and
perform operations on those values. You might think they’re doing a whole lot
more, as you send texts, shop online, use Photoshop, or rely on your phone to
navigate in your car; however, everything computers do can be broken down
into simple operations that are performed on simple values. Now, part of
computational thinking is learning to use these operations and values to build
something that is much more sophisticated, complex, and meaningful—and
we’re going to get to that. First, though, we’re going to take a look at what
these values are, the operations you can perform on them, and just what role
variables play in all this.

this is a new chapter   33

coding a dog age calculator

Coding the Dog Age Calculator
You didn’t think we were going to take you through 50 pages of
Python specifications on values and operations before writing some
real code, did you? Of course not—we’ve got real work to do!
Up next, the Dog Age Calculator. You already know what
the calculator does: you enter a dog’s chronological age and the
calculator tells you the dog’s age in relative human years. To perform
that calcuation you simply multiply the dog’s chronological age by
the number 7. Or, is it that simple? We shall see...

Especially given
you don’t know
how to code yet!

But where do we even start? Do we just start trying to write some
code? Well, remember the concept of pseudocode, which we briefly
discussed in the last chapter? Recall that pseudocode allows you to
work out a high-level solution of your problem before you have to
get into all the specifics required to write code. That’s a good place
to start.
So what exactly is pseudocode? Think of it as nothing more than
your algorithm written in human-readable form. With pseudocode,
you typically spell out, step by step, everything your solution needs
to do to solve your problem (in our case, figuring out a dog’s age in
human years).

Remember, you’ve already seen
one example of pseudocode in
the fishing recipe on page 2
of Chapter 1.

Codie, 12
34  Chapter 2

They say 12 is
the new 9...

simple values, variables, and types

ars

We’re going to write some pseudocode. First, think about how you’d write an
algorithm or recipe that computes a dog’s age in human years. Once you have
an idea, write it down as a set of steps in plain English (or your chosen language).
Make your solution user-friendly by asking the user for the dog’s name and age.
You’ll also want to generate some nice output at the end, like “Your dog, Rover, is
72 years old in human years.”
Again, just write your pseudocode in English. Important: make sure that you
compare your answer with ours at the end of the chapter before you move on.

Or, if you get
stuck, feel free
to peek ahead.

Write your
pseudocode here.

Here is an example of how the
Dog Age Calculator will work.
Python 3.6.0 Shell

Sparky, 1

What is your dog's name? Codie
What is your dog's age? 12
Your dog Codie is 84 years old in human years
>>>

Fido, 5
you are here 4  35

from pseudocode to code

Going from pseudocode to code
Now that we’ve written our pseudocode, we have a good sense of the steps our code
will need to take to implement the Dog Age Calculator—of course the pseudocode
doesn’t provide every detail, but it will provide us with a nice guide to follow as we
implement each step in code.
So let’s do just that: we’ll take our pseudocode step by step, and implement as we go.

As always, if you get stuck,
feel free to peek ahead,
but try not to until you’ve
made a first pass through
every line.

As a first step to transforming our pseudocode into code, step through
each line of your pseudocode and make some notes about what you think
the code will have to do. Just keep it high level. Make sure and check your
answers with ours before you move on. We did the first one for you.

Dog Age Calculator Pseudocode
1.

When we translate our
ideas, algorithms, or
pseudocode to real
code, we often say we’re
implementing them.

1. Prompt the user to get the dog’s name
and then have the user type it in. We’ll
presumably need to save the name somewhere
so we can use it in step 4.

Ask the user for the dog’s name.

2.
2.

Ask the user for the dog’s age.

3.

Multiply the dog’s age by the number 7 to get
the dog’s age in human years.

4.

3.

Output to the user:

		
		
		
		
		

“Your dog”
the dog’s name
“is”
the dog’s age in human years
“years old in human years”

4.

Put your notes for
each step here.
36  Chapter 2

simple values, variables, and types

Q:

Why bother with something English-like when the
computer only understands a programming language?

A:

Working out things in pseudocode gives you the
opportunity to think about your algorithm without the burden
and complexity of actual computer code. It also gives you the
opportunity to study your solution, possibly improving it, before
committing it to code.

Q:
A:

Do experienced software developers use
pseudocode?

Yes. Planning out how you are going to attack a problem
before committing it to more complex coding is always a
good idea. Some developers are good enough to do a lot of
that in their head, but many still use pseudocode or similar
techniques to map out things before they code. Pseudocode is
also commonly used for communicating coding ideas to other
developers.

You are here.

Step 1: Getting some input

Dog Age Calculator Pseudocode

Now we’re ready to tackle step 1—asking the user for the dog’s
name. As we’ve indicated in our notes, we’re going to need to
prompt the user to give us their dog’s name, and then remember
that name so we can use it in step 4 (when we print out the name
and dog’s age in human years). So there’s really two things we
need to do here: prompt the user to get the dog’s name, and then
store that name for later use. Let’s focus on prompting the user
and getting the name first.
You may have already noticed in a couple of the code exercises
from Chapter 1 that we used a Python function named input
to solicit input from the user. While a function may sound like
something from math class, think of it as just a way to call upon
built-in functionality provided by Python.
Let’s look at the syntax for calling the input function and then
we’ll look at how it works:

Remember, the syntax tells us how we
write things in a computer language.

Start with the name of the
function, input.

Follow it
with a left
parenthesis.

1.

Ask the user for the dog’s name.

2.

Ask the user for the dog’s age.

3.

Multiply the dog’s age by the number 7 to get
the dog’s age in human years.

4.

Output to the user:

		
		
		
		
		

“Your dog”
the dog’s name
“is”
the dog’s age in human years
“years old in human years”

d a lot of time on
We’re going to spenbook, and before long
functions in this exactly how they work,
you’ll understand think of them as a way
but for now just to do some work for
of asking Pythonrry too much about the
us and don’t wo
details.

Then place the text you’d
like to prompt the user with
between quotes.
And end the statement with a

input ("What is your dog's name?" )

right parenthesis.

you are here 4  37

the input function

How the input function works
Alright, we now know how to type in the input function (in other
words we know the syntax), but how does it actually work? Like this:
1

When the interpreter sees your call to the input
function, it takes your prompt text and displays
it for the user in the Python Shell.

2

The interpreter then waits for the user to type in
a response, which the user completes by pressing
the Return key.

3

Finally, the text the user typed in is passed back to
your code.

illk

Buy M

Make it stick

Syntax tells us how to write
Python statements.
Semantics tell us what
Python statements mean.

Okay, so the text the user typed in is passed back to our code, but
what does that mean? Well, when you call the input function, it goes
off and gets text from the user and then, as a result of calling the
function, returns that text for your code to make use of.
Now, that text isn’t going to be too useful if we can’t remember it
for later, because we’ll need it in step 4 when we print out our userfriendly output. So, how do we remember things with Python?

Using variables to remember and store values
One of the most common things you’ll do when programming
is to store a value so you can use it later. To do that, we use a
variable. Think of a variable as a name you can always use to
retrieve a value you’ve previously stored. Here’s how we store, or
assign, a value to a variable:
First name your variable. Almost
any name will do, but we’ll talk
more about legal names in a bit...

Next add an equals sig
n, followed
by the value you want
stored and
assigned to your variab
le.

dog_name = 'Codie'

Again, on the lefthand side we
have a variable, which you can
think of as a name you can refer
to over and over to recall a value.

38  Chapter 2

Pretty much every programming
language you’re ever going to
encounter works this way too.

And on the righthand side
we have the value—in this
case, the text ‘Codie’.

We refer to text as a string.
Think of this like a string of
characters. You’ll find this
terminology is common across
practically every programming
language. There are a lot of
other types of values you can use
in Python too, such as numbers,
which we’ll talk about soon.

Python is
interpreteran

simple values, variables, and types

Assigning the user’s input to a variable
Now that you know a little bit about variables (don’t worry, we’re going
to dive in deep in a few pages), let’s store the user’s input in a variable.
To do that we simply call the input function and assign its return value
to a variable. Here’s how we do that:

Let’s use the
variable dog_name.

Then we call the input function,
which prompts the user with
“What is your dog’s name?”

dog_name = input("What is your dog's name? ")

Wondering about how to name
variables? Or how to correctly use
single and double quotes? Hang on,
we'll discuss both shortly.

When the user finishes entering a name, the
input function then passes that name back to
your code in the form of a return value.

We’re moving on to step 2.

And that return value is then
assigned to the variable dog_name.

Dog Age Calculator Pseudocode

Step 2: Getting more input
We also need to get the dog age from the user. How? We’ll do this
just like we did with the dog name: first we make use of the input
function and supply it with a prompt like “What is your dog’s age?”
Then we will take the age the user types in and store this in a variable
called, say, dog_age. When we say “we will” we actually mean YOU
will, in this next exercise:

1.

Ask the user for the dog’s name.

2.

Ask the user for the dog’s age.

3.

Multiply the dog’s age by the number 7 to get
the dog’s age in human years.

4.

Output to the user:

		
		
		
		
		

“Your dog”
the dog’s name
“is”
the dog’s age in human years
“years old in human years”

It’s your turn. Write the code to get the dog’s age using the input function, just as we did with the
dog’s name. Prompt the user with “What is your dog’s age?” and store the result in a variable called
dog_age. Make some notes as well about what each piece of your code does. Check your answer in
the back of the chapter before moving on.

you are here 4  39

using the python shell

It’s time to run some code
It’s one thing to look at code on paper, and it’s another to actually execute real code. Let’s take the
Using the shell is a
code from the last couple of pages and get some experience executing it in a different way—rather
grea
t way to test
than entering code into IDLE’s editor, we’re going to use the Python Shell instead. Why? Well, you’re
out
small
pieces of
going to see that the shell is a great way to experiment with and test small pieces of code. Don’t
code
to
see
how it
worry: when we get to writing longer programs, we’ll be right back in the editor.
To use the shell, run idle3 as usual, only this time we’ll be typing into the Python Shell window:

works.

Python 3.6.0 Shell
Python 3.6.0 (default, Feb 22 2017, 13:46:35)
[GCC 4.2.1 Compatible Apple LLVM 8.0.0] on darwin
Type "copyright", "credits" or "license()" for more information.
>>>

This is the shell prompt; it’s waiting
for you to type something.

Your version numbers and
startup messages may differ
slightly depending on your ,
version and operating system
as may your window’s look
and feel and color.

After you’ve located the shell, click in the window and locate the prompt, which looks like >>>. Type

1 + 1 and hit return; Python will evaluate your expression and print its value (in this case 2) before

displaying another prompt. Congrats, you’ve successfully used the Python Shell! Now let’s try some of
our code:

1

Let’s start by entering the code to get the dog’s name and then hitting return.

Python 3.6.0 Shell
>>> dog_name = input("What is your dog's name? ")
Type this in
What is your dog's name?
and hit return.

Python will
display this.

40  Chapter 2

simple values, variables, and types

2

Next, enter your favorite dog name and press return again.

Notice that an
assignment statement
doesn’t evaluate to a
value like, say, 1 + 1
does. Rather, as you
already know, the
assignment statement
takes the value on
the righthand side
and assigns it to
the variable on the
lefthand side.

Python 3.6.0 Shell
>>> dog_name = input("What is your dog's name? ")
Type a dog’s name here
What is your dog's name? Rover
and hit return.
>>>

Behind the scenes Python takes your dog's name and
stores it, and assigns it to the dog_name variable.
After that, you’ll get another command prompt.

The variable dog_name should now hold the value 'Rover', or
whatever dog name you entered. How can you show it does?

3

We’ve looked at print a few times; let’s use print to see the value of the
variable dog_name. You can also determine the value of dog_name more
directly in the shell by just entering the variable name.

Python 3.6.0 Shell
>>> dog_name = input("What is your dog's name? ")
What is your dog's name? Rover
>>> print(dog_name)
We can use print to display the
Rover
current value of the variable
>>> dog_name
dog_name, which is Rover.
'Rover'
Or enter any variable and
>>>
the shell will evaluate it,

displaying its value.

Note that print
does not display
quotes around
a string value,
while evaluating
it directly in the
shell does.
you are here 4  41

dealing with single and double quotes

I noticed you can’t keep your single and
double quotes straight—sometimes you use text
surrounded by single quotes and sometimes by double
quotes. What’s the deal?

Good catch. First of all, remember we call text

in quotes strings, like a string of characters. And you
are exactly right: we have mostly used single quotes but
then changed to double quotes with the input function.
As it turns out, Python doesn’t care—either single or
double quotes are just fine, as long as you are consistent.
In other words, if you start a string with a single quote,
then you have to end it with a single quote. Likewise, if
you start a string with a double quote, then be sure to
end it with a double quote.
So, why did we choose one over the other? In general a
lot of Python developers prefer to use single quotes, and
so do we; however, there is a case where you’ll be forced
to use double quotes, and that is when you need to use a
single quote as part of your string.
Notice that our prompt needs to use a single quote in
the word dog's:
dog_name = input("What is your dog's name? ")

To use a single quote as part of
your string, just surround the
text with double quotes.
The same is true if you need to use a double quote in
your string’s text, in which case you’ll need to surround
your string by single quotes.

42  Chapter 2

simple values, variables, and types

Getting some code entered
Alright, enough playing around in the shell; we’ve got a real app to build here!
Next, we’ll enter the existing two lines of code (the ones that prompt the user for
the dog’s name and age), then give that code a quick test before moving on to
finish converting our pseudocode to code.
So, in IDLE, choose File > New File and let’s get our first two lines of the
dog calculator typed in. Here’s the code:

dog_name = input("What is your dog's name? ")
dog_age = input("What is your dog's age? ")
After you’ve entered and double-checked the code, choose File > Save and
save your work in a new folder ch2 with the filename dogcalc.py.

A Test Drive
The Test Drive is where we’ll
ask you to stop what you’re
doing, make sure all the new
code is up to date, and then
execute and test your new
code.
Keep in mind at this point
all your code will do is ask
you for your dog’s name
and age, and then display
another prompt.

Note the extra space, so there
is space between the prompt
and where the user types.

We’re not done with the calculator yet, but it’s a good idea to test code
as you go to make sure everything is working as you’d expect. After
you’ve saved the code, choose the Run > Run Module menu item.
Look for the output in your shell window. Remember, the code is going
to ask you to enter your dog’s name and age.

Python 3.6.0 Shell

er
What is your dog's name? Rov
What is your dog's age? 12
>>>

If you get a “SyntaxError: EOL while scanning string literal” when
executing this code, double-check your use of single and double quotes with
the code above.
you are here 4  43

a closer look at variables

A deep dive on variables
We’ve gotten a little experience by creating a variable name and assigning a value
to it, but how do these variables really work and how can you make use of them?
Let’s take a little time over the next few pages to better understand variables. After
that, we’ll finish off our Dog Age Calculator. Let’s start by looking at what is
happening behind the scenes when we assign a value to a variable.

dog_name = 'Codie'
1

2

We often just call
this “assignment.”

'Codie'

The first thing Python does is evaluate the
righthand side of the assignment, which evaluates
to the string 'Codie', and then it finds a free spot
in your computer’s memory where it stores the
string. You can think of this like taking an empty
cup and throwing the text 'Codie' in it.
With the string 'Codie' stored, Python then
creates a label—think of it as a sticky note if you
want—with the name dog_name, and puts it on
the cup.

Python creates a spot
in your device’s memory
for a string, and places
a ‘Codie' there.
e'

i
od

'C

dog_name
e

3

phrase = "Your dog's name is "

r

u
Yo

Of course we can create and store as many values
as we need. How about two more:

"

s
g'

m
na

is"

Next, a label, dog_name,
is made and associated
with the value’s location
in memory.

do

phrase

4

This retrieves the value for
phrase and dog_name, and
then prints...

44  Chapter 2

12

dog_age

We can create as many values assigned to
variables as we need, which will be stored in
memory for us until we need them.

Anytime we need to retrieve the stored values, we
can use the variables:

print(phrase)
print(dog_name)

'C

dog_name

dog_age = 12

Note we can store numbers
as well as strings.

e'

i
od

Python
Your dog’s name is
Codie

simple values, variables, and types

Adding some expression
So far our values have been pretty simple, but there’s no need for them to be—instead of using
simple values, you can use expressions to compute values. If you’ve ever seen any form of mathematical
expression, Python’s expressions should look familiar to you: they just consist of simple values
combined with operators like +, -, *, and /. For instance, what if it is Codie’s birthday?

dog_age = 12 + 1

erator,
We call + thetop
operands.
and 12 and 1 he

This evaluates to a
single value, 13.
After the expression is evaluated, the
variable dog_age is assigned the value 13.

13

dog_age

52

Here’s the expression.

weight = 38 * 0.454

17

.2

Or we need to compute Codie’s weight in kilograms?

g the
Here we’re usinop
erator.
multiplication

Almost without
exception,
programming
languages use an
asterisk for the
multiplication
symbol.

weight

After this statement completes, weight
has been assigned the value 17.252.

avg = (12 + 5 + 1) / 3

6.

0

Or we need to compute the average age of Codie, Fido, and Sparky?

avg

We can group operations
together using parentheses.

Anywhere we put a variable, it

codie = 12
is replaced with its value to
fido = 5
compute the expression.
sparky = 1
avg = (codie + fido + sparky) / 3

6.

0

In these expressions we can use variables anywhere you’d use a value. As
an example, let’s rewrite the average of Codie, Fido, and Sparky’s age:

avg

you are here 4  45

changing a variable’s value

greeting = 'Hi'
name = 'Codie'
message = greeting + ' '

+ name

And add (or rather,
concatenate) them
together.

'H

Let’s create a
couple strings.

iC
od
ie
'

Of course we aren’t limited to simple math and numbers in our expressions. Remember
concatenation from Chapter 1? We use concatenation to add strings together.

message

This expression
evaulates to ‘Hi Codie’,
which is assigned to
the variable message.

Variables are called VARY-ables for a reason
Variables are called variables because their values usually vary over time. Let’s take Codie’s height
for example, which we start at 22:

Here we’re creating a new variable
and assigning it the value 22.

dog_height = 22

22

dog_height

Let’s say Codie has grown and we need to update his height by one inch, which changes the value
that dog_height is holding.

dog_height = 22 + 1

As usual we evaluate the righthand
side, which evaluates to 23, and then
assign that value to dog_height. So,
dog_height changes from 22 to 23.

23

dog_height

But there’s a better way to update Codie’s height. Let’s do that when Codie grows another two
inches:

dog_height = dog_height + 2
2

We then take 25 and make it
the new value of dog_height.

46  Chapter 2

1

anywhere
Remember we can use a variableand
side we
we use a value, so on the righth
eight,
_h
add 2 to the current value of dog
or, 23 + 2 = 25.

dog_height

25

simple values, variables, and types

We need to talk about your
operator precedence...

Better living through
operator precedence
Evalute this expression:

mystery_number = 3 + 4 * 5
Is mystery_number 35? Or is it 23? Well, it depends on whether you added three
to four first, or if you multiplied four time five first. What’s the right answer? 23.
How did we know the right order of evaluation? Operator precedence. Operator
precedence tells you the order in which operations should be applied. This isn’t
a computer programming thing, mind you, it’s a mathematics thing, and in your
algebra class you most likely covered precedence at some point (either recently or
long ago). If you’ve forgotten it, no worries—we’re going to cover it now.
To understand operator precedence, think of operators being listed from highest to
lowest precedence, like this:

Highest
**
-

* / %
+

-

If you remember your math,
Two asterisks give us
3
2
**3 is the same as 2.
exponentiation, which has the
highest precedence.
Next highest is negation (in other words, just
The % is the modulus operator—modulus
putting a negative sign in front of a value).
gives you the remainder of a division.
For instance, 7 % 3 is 1 because
Followed by multiplication, division, and modulus.
3 divides into 7 twice, leaving a
remainder of 1.
And then addition and subtraction.

Lowest
you are here 4  47

what is operator precedence

Computing with operator precedence
illk

Buy M

Let’s evaluate an expression to understand how operator precedence is
applied. Here’s an expression to evalute:

3 / 1 + 2 * 3 ** 2
3

18

9

The highest precedence
in this expression is 2
exponentiation. 3 is 9.

Next we have a division and a
multiplication. They have equal
21
precedence, so do them in whatever
order you like (the result will be
the same).
Finally, the lowest precedence
If you try this in the Python
operator we have is addition. So
Shell, you’ll actually get 21.0;
let’s add 3 plus 18, giving us 21.
more on that in a bit.

Make it stick

Python is an
interpreter

Remember the mnemonic PEMDAS
for Parentheses, Exponents,
Multiplication, Division, Addition, and
Subtraction. If you follow that order of
evaluation, left to right, you’ll always
evaluate expressions correctly. If you
have trouble remembering PEMDAS, you
could always try “Please Excuse My Dear
Aunt Sally.”

You can determine how any Python numeric expression is evaluated by simply applying its operations in
the order of their precedence (highest first). In these examples we’re using numbers to keep things simple,
but of course we could replace any of these numbers with variables and the same rules apply.
But what if this isn’t the way you wanted to evaluate this expression? What if you really wanted to add 1 to
2 before the division and multiplication occurred? That’s where parentheses come in. With parentheses you
can dictate the order:

3 / (1 + 2) * 3 ** 2
9

The highest precedence in this2
expression is exponentiation. 3 is 9.

Next we have to take care of the addition in
parens before we can divide or multiply: 1+2 = 3
1
We can then choose division or
multiplication in any order because
they have the same precedence. We
9
chose division. 3/3 = 1
You can also add parens that
Finally we multiply 1 * 9,
don’t necessarily change
resulting in 9.
evaluation order, but improve
You can add as many parentheses as you need to evaluate your expression in the
your code’s readability.
appropriate order. For instance:
This expression first does division,
then addition, then multiplication,
(((3 / 1) + 2) * 3) ** 2
then exponentiation, and evaluates to
225.

3

48  Chapter 2

simple values, variables, and types

I?
m
a
ho

A bunch of Python expressions lost their values. Can you help
them get back together? For each expression on the left, draw a
line to the value it evaluates to on the right. Careful—there may
be an impostor hanging out too.

W

Here’s modulus again—it
may seem esoteric, but it
has a lot of uses in coding.
Some folks like to call this
remainder.

Expressions

Values

'kit e' + ' ' + 'cat'

1

(14 - 9) * 3

2

3.14159265 * 3**2

15

Who’s the
impostor?

21

42

28.27433385

'h' + 'e' + 'l' + 'l' + 'o'

42

8%3

-13

7-2*3

'kit e cat'

(7 - 2) * 3

'hello'

Get ready for the classic shell game, usually played with cups and balls; in this game we’re going
to use variables and values. Using what you know about variables, values, and assignment, see
if you can beat the cup game. Work through the code and see which cup has the number 1 in it at
the end of the game. Will it be in cup 1, cup 2, or cup 3? Place your bet now!

cup1
cup2
cup3
cup1
cup2
cup3
cup1
cup2
cup3
cup1
cup3
cup2

=
=
=
=
=
=
=
=
=
=
=
=

0
1
0
cup1
cup1
cup1
cup1
cup3
cup1
cup2
cup2
cup3

+ 1
- 1

Use your brain to evaluate this code and see, when it
completes, which cup has the 1 in it. When you’re done,
check your answer at the end of the chapter (you can
type the code in as well, but only to check yourself).

* 0
% 1
- cup3

1

2

3

you are here 4  49

an exercise with variables

this is an error in Python
What does this expression evaluate to? Or do you think
string?
because we’re multiplying a number times a

3 * 'ice cream'

You might want to
try typing it into
the Python Shell.

Crack the Code Challenge
You’re all ready for your first spy assignment, but to get started you need your allimportant passcode. You’ll find the passcode below in code—that is, literally, in code.
Work through the code in your head to figure out the passcode, but be careful—get it
wrong, and it may be the last code you’ll ever look at. We wish you luck.
word1 = 'ox'
word2 = 'owl'

Here’s your passcode; all you
have to do is work through
the code to get it.

word3 = 'cow'
word4 = 'sheep'
word5 = 'flies'
word6 = 'trots'
word7 = 'runs'
word8 = 'blue'
word9 = 'red'
word10 = 'yellow'
word9 = 'The ' + word9
passcode = word8
passcode = word9
passcode = passcode + ' f'
passcode = passcode + word1
passcode = passcode + ' '
passcode = passcode + word6
print(passcode)

50  Chapter 2

This prints your passcode.

simple values, variables, and types

Back away from that keyboard!
You know variables have a name, and you know they have a value.
But what can you call your variables? Is any name okay? Well, no, but the
rules around creating variable names are simple: just follow the two rules below to
create valid variable names:

1

Start your variables with a letter or an underscore.

2 After that, use as many letters, numeric digits, or
underscores as you like.

Oh, and one more thing: we really don’t want to confuse Python by using any
of the built-in keywords, like False or while or if, so consider those off-limits
for your own variable names. We’ll get to these keywords and what they mean
throughout the rest of the book, but here’s a list just to take a quick look at:
False

as

continue

else

from

in

not

return

None

assert

def

except

global

is

or

try

True

break

del

finally

if

lambda

pass

while

and

class

elif

for

import nonlocal

raise

with

Q:
A:

What’s a keyword?

A keyword is basically a word that
Python reserves for its own use. Keywords
are part of the core Python language, and so
using them as variables in code would result
in confusion.

Q:

What if I used a keyword as part
of my variable name? For instance, can I
have a variable named if_only (that is, a
variable that contains the keyword if)?

A:

You sure can—just don’t match the
keyword exactly. It’s also good to write clear
code, so in general you wouldn’t want to
use something like elze, which might be
confused with else. Again, we’ll look at what
these keywords mean throughout the book.

yield

Q:

Does Python consider myvariable
and MyVariable the same thing?

A:

Python treats these as two different
variables names. More technically, Python is
said to be case sensitive, meaning it treats
upper- and lowercase letters differently.
Today, most common programming
languages are case sensitive, but not all are.

Q:

Are there any conventions for
naming variables? Do I use myVar, MyVar,
or my_var, or does it matter?

A:

These rules are specific to
Python; other programming
languages have their own rules,
which can be quite different.

There are conventions that Python
programmers follow. Python programmers
prefer to use lowercase letters for variables.
If a variable name has multiple words, then
put underscores between the words, like

Expect Python to complain or at least
be confused if you use any of these
reserved keywords as variable names.
max_speed, height, or super_turbo_mode.
As we’ll see later, there are some additional
conventions for naming other things in
Python. Also, these conventions are specific
to Python; you’ll find each language has its
own conventions.

Q:

Okay, but what makes a good
variable name? Or does it matter?

A:

To Python it doesn’t matter at all, as
long as your variable names follow the rules.
To you, your variable names could matter
a lot. When you choose names that are
clear and meaningful, your code becomes
more readable and understandable. Short,
terse names can be hard to read, as can
long, cumbersome names. In general, call
variables what they are; for instance, don’t
name your variable num when you should
have named it number_of_hotdogs.
you are here 4  51

the dog age calculation

Step 3: Computing the dog’s age
We’ve got the first two steps in our pseudocode out of the
way, so let’s move on to step 3. In this step we simply need to
multiply the dog’s age by 7 to get the dog’s age in human years.
It seems like that would be straightforward, so let’s do a little
experimentation back in the shell:

Python 3.6.0 Shell

First let’s define our
variable dog_age and set
it to the number 12.

>>> dog_age = 12
>>> human_age = dog_age * 7
Now let’s just try to multiply
>>> print(human_age)
it by 7 and assign it to a
new variable, human_age.
84
>>>
And then let’s print human_age.

We get 84. Perfect, just what we wanted!

We’re here now in the
pseudocode.

then

Could we
get back to more
pressing matters,
namely the dog age
code?

Jackson, 9
Dog Age Calculator Pseudocode
1.

Ask the user for the dog’s name.

2.

Ask the user for the dog’s age.

3.

Multiply the dog’s age by the number 7 to get
the dog’s age in human years.

4.

Output to the user:

		
		
		
		
		

“Your dog”
the dog’s name
“is”
the dog’s age in human years
“years old in human years”

Below you’ll find the Dog Age Calculator code so far. Using the experiment above as
a guide, add the code to compute the dog age in human years.

dog_name = input("What is your dog's name? ")
dog_age = input("What is your dog's age? ")

Add your new
code here, then
check your
answer before
moving on.

52  Chapter 2

simple values, variables, and types

A Test Drive

Let’s test drive our code now that we’ve got step 3 coded. Go ahead and
get the new code into your dogcalc.py file, save your code, and choose
the Run > Run Module menu item. After that, head to the shell and enter
your dog’s name and age, and then check the output to see that we’re
calculating the human age correctly.
Here’s the code again:

Add this new code
to your file.

dog_name = input("What is your dog's name? ")
dog_age = input("What is your dog's age? ")
human_age = dog_age * 7
print(human_age)

Could you have printed dog_age * 7
directly? Like print(dog_age * 7)?
Give it a try.

Throughout the book you’ll find the gray
background signifies new code additions.

Python 3.6.0 Shell

Houston, we’ve got a problem!

What is your dog's name?
Codie
What is your dog's age? 12
12121212121212
>>>
Okay, that’s not right!

Did you get what we got? We entered that Codie was
12 and expected to see an output of 84, but we got
12121212121212! What on earth! Where did we go
wrong? We’ve double-checked our code, and everything
looks good. Why would this output be different than our
test run in the command-line interpreter?

thing
121212121212. Is there any
Think about the number 12
?
ber
num
this
s how Python got
you can think of that explain
Hint: how many 12s are in 121

21212121212?

you are here 4  53

different kinds of errors

To err is human to code
Finding and fixing errors is part of coding. No matter how well you plan
your pseudocode and how diligent you are writing your code, errors are a
part of coding life; in fact, most programmers view debugging as a natural
part of the coding process. What’s debugging? Debugging is the process of
removing errors from the software you’re writing. We formally call these
errors defects, and informally everyone just calls them bugs.
As you code you’ll actually encounter three different kinds of errors. Let’s
have a look at them:

They’re called bugs
because one of the first
programming errors was
caused by a moth that
became stuck in an early
computer relay. So, we
ended up calling the
process of removing errors
debugging. More at: https://
en.wikipedia.org/wiki/
Software_bug.

syntax errors
get the
You’ll know quickly if you have a syntax error in Python when you
errors
Syntax
eter.
interpr
the
from
e
messag
dreaded 'Syntax Error'
are the equivalent of making grammatical errors—in other words,
correct
you’ve typed something that violates the conventions for writing
t find the
fix—jus
to
Python. The good news is syntax errors are usually easy
offending line of code and double -check your syntax.

runtime errors

Runtime errors occur when you
’ve written a syntac tically cor
rec t program, but
Python encounters a problem
running your program. An exa
mple of a runtime
error would be if at some point
in your code you accidentally
divided a number
by zero (an invalid mathematic
al operation in any language).
To fix runtime errors,
look at the specific error you rece
ived and then track down whe
re in your code
you’re causing the runtime con
dition to occur.

semantic errors
Semantic errors are also known as logic errors. With a semantic error
your program will appear to operate normally—the interpreter won’t
complain that you’ve made a syntax error and at runtime you won’t
encounter any issues, but your program won’t give you the results you
expected. These always occur because what you think you’ve told
your program to do isn’t actually what you’re telling it to do. Semantic
errors can be some of the toughest errors to debug.

54  Chapter 2

simple values, variables, and types

We currently have a bug in our Dog Age Calculator because we’re getting
121212121212 instead of 84. What kind of error is this?
A. Runtime error

D. Calculation error

B. Syntax error

E. None of the above

C. Semantic error

F. All of the above
Answer C because there are no syntax or runtime errors.

I noticed there are seven 12s,
so somehow when we multiplied 12 by 7,
Python didn’t use multiplication, but rather just repeated
the 12 seven times. I did a little experimenting in the Python
Shell and I discovered if I multiply the string "12" by 7, I get the
result of 12121212121212, but if I multiply the number 12 by
7, I get 84 like we’d expect.

Great debugging. The reason you got the

result 12121212121212 instead of 84 is that Python,
as you guessed, treated the dog age as a string and not a
number. Why does that matter? Well, let’s take a closer
look at how Python deals with multiplication...

you are here 4  55

doing some debugging

A little more debugging...
Let’s do a little investigative debugging and see exactly what is happening. Using our
handy Python Shell let’s try a few things:

Python 3.6.0 Shell
Note this is
a string.

>>> '3' * 7

As reported, if Python multiplies a string and a
number, it just repeats the string that many times.

'3333333'
>>> 3 * 7

And this is
a number.

If Python multiplies two numbers, it works as you’d expect.

21
>>> 'ice cream ' * 3
'ice cream ice cream ice cream'

And it looks like we can multiply ANY string by
a number and it evaluates to a string that is the
repetition of that string.

>>> num = input('input a number: ')
input a number: 12
>>> num
'12'
>>> num * 7
'12121212121212'

This evaluates to a string that contains
‘ice cream' three times.
Let's get a number from the user
using the input function.

Yes, it looks like we're getting a string back from input.
And multiply by 7.
As suspected.

>>>
So why is Python treating the number we entered using the input function as a string? After all, we
clearly typed in a number when prompted. The reason is because the input function always returns
a string. How do we know? Well, while our experiment seems to bear this out, there are actually more
direct ways to know. First of all, we could just look at the Python specification, which says:
input(prompt) After the prompt argument is written to output, the function then
reads a line from input, converts it to a string, and returns that.
There are also ways, in code, we could examine the value and determine it is a string. Both of
these topics, specifications and examining values, are something we’ll be getting to later, but for
now just know that we’ve uncovered the problem: input is returning a string when we need a number.
56  Chapter 2

This definition is
from the Python
specification (we
abbreviated it
slightly).

simple values, variables, and types

Aren’t these
computers? Python can’t even figure
out that when I’m multiplying a number
with a number in a string, I want to do
real multiplication rather than just
repeat the string over and over? How
dumb is that?

Careful what you ask for...
It certainly feels like Python could just figure this
out and decide our string is actually a number, but
that might lead to surprises down the road when
you really meant for a number to be a string. That’s
the thing: it’s hard for an interpreter to decide the
right thing every time, and that is why Python tries
NOT to make any assumptions. Python, as we said
early on, tries to avoid surprises by taking you at your
word (er, code)—if you say something is a string, then
Python treats it as a string. If you instead represent
something as a number, then Python’s good with that
too. What Python doesn’t like to do is start guessing
what you intend, because it’s going to get it wrong a
lot of the time, leading to more debugging by you.
What this all means is that as a programmer you’re
better served by understanding the type of data
you’re using and operating on. So, maybe we need to
take a closer look at types and what they’re all about.

you are here 4  57

introducing python types

What are Python types, anyway?
Every piece of data in Python has a type. Types are important
because they determine how Python operates on our data. As
we’ve seen with multiplication, it makes a big difference if we’re
multiplying a number by a string or a number by a number. So,
as you code, you want to be very conscious of the types you’re
using and operating on.
We’ve already seen a few examples of different data types. We
have strings, which, as you already know, you can think of as
a sequence of characters, like “What is your dog’s age?” We’ve
also seen a few number types in passing. When it comes to
numbers, Python supports two main types: integers and floats.
Let’s take a closer look:
Also known as an int.

Integer Numbers
0

1
-452

10001

2

-2

Has no decimal point.

-1

1200000

42

3
999

Floating-Point Numbers
0.0
-2.1
10.3+e4

3.1415

50.919103

-4.e10

Whole numbers like
–3, –2, –1, 0, 1, 2, 3, ...
Notice we don’t
use commas with
Python numbers
greater than 999.

Also known as a float.

I get that a string is a sequence of
characters, but what can I do with them other than
concatenate them and print them out?

A:

Don’t be fooled by our simple beginnings: string
processing is an important part of our computational
world—just think of all the text being processed by
Google or Facebook as an example. We’re going to
explore strings a lot more as the book progresses
and find out how we can search them, change them,
format them, and create all kinds of interesting
algorithms that operate on them.

Q:
A:

Do I have to worry about mixing integers
and floats like I do with integers and strings?

Not really. Anytime you use ints and floats
together in an expression, Python will typically
convert everything to a floating-point value.

Q:

What if a variable is assigned to an int
value and then I switch it to be assigned to a
string value?

A:

Python is totally fine with that. Remember the
value has a type, not the variable. So you can assign
variables to values and change those values over
time, even if it means the type changes as well. Note,
however, that this is not considered a good practice,
as code quickly becomes confusing when variables
change types in a piece of code.

Has a decimal point.
Fractional numbers
like 3.1415, 42.0,
and 1.6667.
Supports scientific
notation.

Python has other types too, like booleans, lists, dictionaries,
and objects, that we’ll be getting to later in the book.

58  Chapter 2

Q:

Serious Coding
In need of some rocket-scientist-style
numbers? Python has you covered
there too; with Python you can use
scientific notation or imaginary
numbers should you need to.

simple values, variables, and types

Fixing our code
Well, it looks like our little bug was caused by an assumption we
made about types: we assumed that if we supplied a number when
prompted by the input function, then we’d have a number returned
to us. But, as we’ve discovered, the input function always returns
a string. No worries, though—this is exactly why we test as we go, to
uncover bugs like this. And, as it turns out, Python has a handy way
to convert a string to a numeric value.
Here’s how it works:
1

First, let’s create a string representation of a number.

answer = '42'

2

Here the string ‘42’
is assigned to the
variable answer.

Next, call the int function and pass it the string.

answer = int('42')
Look,
Ma, I’m an int!

Here the string ‘42’ is
converted to an integer and
assigned to the variable answer.

Q:

What happens if my string
doesn’t have a number in it and I call
the int function? Like int('hi')?

A:

If you call the int function and
pass it a string that does not contain a
number, then you’ll get a runtime error that
indicates your value is not a number.

Q:

Does the int function work only
with integers values; for instance,
could I use int('3.14')?

A:

The int function only works for
integers (or rather, strings that represent
integers), but if you need to convert a
string that contains a float value, like
'3.14', you can use the float function, like
float('3.14'), to convert it to a floatingpoint value.

So we know that the bug is caused by our assumption that dog_age is a number,
when in fact Python is treating it as a string. Take the code below and add an int
function to fix this problem.

There’s a few ways to approach this, so be sure and check
your work with the answer at the end of the chapter.

dog_name = input("What is your dog's name? ")
dog_age = input("What is your dog's age? ")
human_age = dog_age * 7
print(human_age)

you are here 4  59

testing the calculator

_

A Test Drive

So now that we’ve figured out the code to convert the string to an int (in the
last exercise), let’s get it added. First make the changes below to update
your code to handle the conversion of the dog age to an integer, then save
your code and choose the Run > Run Module menu item. After that, head
to the console and enter your dog’s name and age, and check the output to
see that we’re finally calculating the human age correctly.

We added the int function
here so that the dog_age
value is converted to an
integer just before it is
multiplied by 7.

Here’s the code (changes are highlighted):
dog_name = input("What is your dog's name? ")
dog_age = input("What is your dog's age? ")
human_age = int(dog_age) * 7
print(human_age)

Houston, we have liftoff
Much better this time! Now our calculation works; we got
84 just as we’d expect. Give this version a few more test

runs with some different values. After that, we’re ready to
implement the last step of our pseudocode and finish the
Dog Age Calculator.

Python 3.6.0 Shell

What is your dog's name? Codie
What is your dog's age? 12
84
Here’s what we got.
>>>

If your code isn’t working as expected...

syntactic, or semantic, can
As you already know, errors, whether they are runtime,
tly. If Python is complaining
get in the way of your Dog Age Calculator working correc
ation, like missing
punctu
ct
about invalid syntax check your code for incorre
read the code backward
people
many
issue,
the
parentheses. If you’re still not seeing
friends look at the code
their
having
by
review
peer
some
do
or
,
to look for typos they aren’t seeing
int type conversion
the
got
sure you’ve
and compare it with the book’s source code. Also make
well and directly
as
code
source
the
ad
function and its parentheses. You can always downlo
com/hflearntocode.
smart.
ickedly
http://w
at
found
be
can
code
compare it with yours. Remember the

60  Chapter 2

simple values, variables, and types
Dog Age Calculator Pseudocode

Step 4: User-friendly output
Well, we’ve made it to the last step of our pseudocode—we have input
from the user, we’ve computed the dog’s age in human years, and now
all we need to do is provide some nice output. To do that, we just need
to survey our current variables, which hold all the values we need, and to
use some print statements to output them as specified by our algorithm:
ie
'
'C
od

Ask the user for the dog’s name.

2.

Ask the user for the dog’s age.

3.

Multiply the dog’s age by the number 7 to get
the dog’s age in human years.

4.

Output to the user:

		
		
		
		
		

12

dog_name

1.

“Your dog”
the dog’s name
“is”
the dog’s age in human years
“years old in human years”

dog_age
84

human_age

We’ve got all the values
we need to finish the last
step of our pseudocode.

Our pseudocode already does a nice job of specifying the exact output
we need to code. Before we write the code, though, let’s look at another
way you can use the print function that will make your job easier. So far
we’ve just been providing print with one argument, like:
We refer

Until now we've only been
passing single values to
print, either strings or
numbers.
We've also used string
print(42)
concatenation to build up
strings in the print function,
but
that still, ultimately,
print('Good' + 'bye')
evaluates to just one argument.
print('Hi there')

or:

or:

to the values we provide
to a function as its arguments. We
also typically say that we pass
those arguments to a function.

But you can actually pass multiple arguments to print, like:

You can pass as many arguments to print
as you like, separated by commas.

print('Hi there', 42, 3.7, 'Goodbye')

An argument is just a fancy name for the values you pass
to a function. More on this topic as we progress...

When you pass multiple
arguments, print will display
each one with a one-space
separator in between.

Python 3.6.0 Shell

Hi there 42 3.7 Goodbye
>>>
you are here 4  61

testing the calculator output

Output to the user:
“Your dog”
the dog’s name
“is”
the dog’s age in human years
“years old in human years”

With your newfound print functionality, write the code to call to
the print function that will produce the user-friendly output:

Here's what you
need to output.

A FinalTest Drive
Okay, we’ve got the step 4 code sketched out in the Sharpen exercise, which we’ve
reproduced below. Go ahead and make the changes, then save your code and
choose the Run > Run Module menu item. After that, head to the shell and give
your Dog Age Calculator a try!
Here’s the code:

Remove this
code.
Add this
code to your
dogcalc.py file.

dog_name = input("What is your dog's name? ")
dog_age = input("What is your dog's age? ")
human_age = int(dog_age) * 7
print(human_age)
print('Your dog',
dog_name,
'is',
human_age,
'years old in human years')

Python 3.6.0 Shell
Our final output

62  Chapter 2

What is your dog's name? Codie
What is your dog's age? 12
Your dog Codie is 84 years old in human years
>>>

Just what we’d
planned for!

simple values, variables, and types

Brain Building
Well, the chapter is almost over; you’ve got just the bullet points and a
crossword to go. Why not do a few extra reps for that brain before you go?
Here’s one for you: say you have two variables, first and last, like below.
Can you swap their values? See if you can write the code. As usual, you’ll find
our solution at the end of the chapter.

first = 'somewhere'
last = 'over the rainbow'
print(first, last)

Write your
code here.

print(first, last)

Here's the output
you should get
when you tune
this code.

Python 3.6.0 Shell

somewhere over the rainbow
over the rainbow somewhere
>>>

you are here 4  63

review of values, variables, and types

What if you left the quotes off of Codie’s name? What would this evaluate to?

dog_name = Codie

 Computers do two things well: store values
and perform operations on those values.

 Keywords are reserved words in a
programming language.

 Using pseudocode is a great way to start the
design of your program.

 A string is a Python type that consists of a
sequence of characters.

 Syntax tells us how to form Python
statements; semantics tells us what they
mean.

 Python supports two numeric types, integers
and floating-point numbers, in addition to a
couple other types indended for scientific
computation.

 A file containing Python code is also called a
module.
 You can use the Python Shell to enter and
evaluate one statement or expression at a
time.
 The input function prompts the user, gets a
line of input, and returns it as a string.
 We call functions and they can return values
as a result.
 We can store a value for later use by
assigning it to a variable.
 When you assign a value to a variable, Python
stores the value in memory, and labels it with
the variable name for later use.

 The int function will convert a string to an
integer value. The float function converts a
string to a floating-point value.
 We test our code often to detect issues early
on.
 There are three kinds of errors: syntax,
runtime, and semantic.
 If you multiply a string times a number, n,
Python will repeat the string n times.
 You can combine strings with the
concatenation operator.
 Concatenation just means adding strings
together.

 A variable can vary, with its value changing
over time.

 For strings, Python uses + as the
concatenation operator.

 Python provides simple rules for naming
variables and doesn’t allow them to be named
the same name as a keyword.

 When we pass values to functions, we call
those values arguments.

64  Chapter 2

simple values, variables, and types

Coding Crossword
It’s time for something completely different. It’s
your standard crossword, but all of the solution
words are from Chapter 2.
1
3

2

4

5
6

7
8

9

10

11
12

13

14

15
16
17

18

19

20

21

Across Across

3. What we’re coding.
3. What6.we're
coding.
Can’t name
a variable after one of these.
6. Can't7.name
a variable
after
onestatements
of these. in
Determines
how we
write
7. Determines
how
we
write
statements
in
Python.
Python.
8. Stores a value for later use.
8. Stores
a value
later use.
9. The
otherfor
numeric
type.
9. The other
numeric
11. 12 years
old. type.
11. 12 years
old. of values passed to a function.
13. Name
13. Name
of
values
passed
14. Input
returns
this.to a function.
14. Input17.returns
this.
Special character you can use in your
17. Special variable
character
you can use in your
names.
variable names.
19. Another dog.
19. Another dog.
20. Another name for a file of Python.
20. Another name for a file of Python.
21. Assign a variable to a value with this.

Down
Down

1. Has a decimal point.

1.2.Has
decimal
point. can do this.
Theavalue
of a variable
2.4.The
value
of
a
Adding stringsvariable
together.can do this.
4.5.Adding
strings
together.
What statements
mean.
5.10.
What
statements
Input shows this mean.
to the user.
10.12.
Input
shows
to the
user.to integers.
Function
to this
convert
strings
12.15.
Function
toyour
convert
to integers.
Evaluates
codestrings
line by line.
15.16.
Evaluates
your
code
line
by
line.
Values are stored in this.
16.18.
Values
are
stored
in
this.
Error encountered before your code is
18. Error
encountered before your code is run.
run.

21. Assign a variable to a value with this.

you are here 4  65

exercise solutions

ars
We’re going to write some pseudocode. First, think about how you’d write
an algorithm or recipe that computes a dog’s age in human years. Once you
have an idea, write it down as a set of steps in plain English (or your chosen
language). Make your solution user-friendly by asking the user for the dog’s
name and age. You’ll also want to generate some nice output, like “Your dog,
Rover, is 72 years old in human years.”
Again, just write your pseudocode in English.

First we need to get some
information from the user.

Ask the user for the dog’s name.
Ask the user for the dog’s age (in dog years).

Multiply the dog’s age by the number 7 to get the dog age in human years.
Output to the user:

Next we'll need to
calculate the dog's age in
human terms.

“Your dog”
then the dog’s name
“is”
then the dog’s age in human years

Finally, we need to output
our results in a user-friendly
manner.

“years old in human years”

Sparky, 1

Fido, 5
66  Chapter 2

simple values, variables, and types
As a first step to transforming our pseudocode into code, step through each line of your pseudocode
and make some notes about what you think the code will have to do. Just keep it high level.

Dog Age Calculator Pseudocode
1.

Ask the user for the dog’s name.

2.

Ask the user for the dog’s age.

3.

Multiply the dog’s age by the number 7 to get
the dog’s age in human years.

4.

Output to the user:

		
		
		
		
		

“Your dog”
the dog’s name
“is”
the dog’s age in human years
“years old in human years”

1. Prompt the user to get the dog’s name
and then have the user type it in. We'll
presumably need to save the name
somewhere so we can use it in step 4.
2. Prompt the user to get the dog's age and
then have the user type it in. We'll also
need to save this somewhere so we can use
it in step 4.

3. Take the age from step 2 and multiply it
by the number 7. We'll also need to store
this somewhere for use in step 4.

4. First print to the console “Your dog”,
then print the value from step 1, then
print “is”, then print the value from step
3, then print “years old in human years”.

It’s your turn. Write the code to get the dog’s age using the input function, just as
we did with the dog’s name. Prompt the user with “What is your dog’s age?” and
store the result in a variable called dog_age. Make some notes as well about what
each piece of your code does.

dog_age = input("What is your dog's age? ")

Here’s our new
variable, dog_age.

And here’s the call to input, which
will prompt the user with “What is
your dog’s age?” and then return
the input to be stored in dog_age.

you are here 4  67

exercise solutions

Who

am I?

ion

Solut

A bunch of Python expressions lost their values.
Can you help them get back together? For each
expression on the right, draw a line to the value it
evaluates to. Careful—there may be some impostors
hanging out too.

Expressions

Values

'kit e' + ' ' + 'cat'

1

(14 - 9) * 3

2
15

3.14159265 * 3**2

21

42

28.27433385

'h' + 'e' + 'l' + 'l' + 'o'

42

8%3

-13

7-2*3

'kit e cat'

(7 - 2) * 3

Impostor!

'hello'

Get ready for the classic shell game, usually played with cups and balls; in this game we’re
going to use variables and values. Using what you know about variables, values, and
assignment, see if you can beat the cup game. Work through the code and see which cup has
the number 1 in it at the end of the game. Will it be in 1, 2, or 3? Place your bet now!

cup1
cup2
cup3
cup1
cup2
cup3
cup1
cup2
cup3
cup1
cup3
cup2

3
68  Chapter 2

=
=
=
=
=
=
=
=
=
=
=
=

0
1
0
cup1
cup1
cup1
cup1
cup3
cup1
cup2
cup2
cup3

cup1 is 0
cup1 is 0, cup2 is 1

+ 1
- 1
* 0
% 1
- cup3

cup1 is 0, cup2 is 1, cup3 is 0
cup1 is 1, cup2 is 1, cup3 is 0
cup1 is 1, cup2 is 0, cup3 is 0
cup1 is 1, cup2 is 0, cup3 is 1
cup1 is 0, cup2 is 0, cup3 is 1
cup1 is 0, cup2 is 1, cup3 is 1
cup1 is 0, cup2 is 1, cup3 is 0
cup1 is 0, cup2 is 1, cup3 is 0
cup1 is 0, cup2 is 1, cup3 is 1
cup1 is 0, cup2 is 0, cup3 is 1

Winner!

simple values, variables, and types

Crack the Code Challenge SOLUTION
You’re all ready for your first spy assignment, but to get started you need your allimportant passcode. You’ll find the passcode below in code—that is, literally in code. Work
through the code in your head to figure out the passcode, but be careful—get it wrong,
and it may be the last code you’ll ever look at. We wish you luck.

Python 3.6.0 Shell

The red fox trots
>>>

word1 = 'ox'
word2 = 'owl'
word3 = 'cow'
word4 = 'sheep'
word5 = 'flies'
word6 = 'trots'
word7 = 'runs'
word8 = 'blue'

If you type the code in
and run it, you’ll get this.

word9 = 'red'
word10 = 'yellow'
word9 = 'The ' + word9

word9 is 'The red'

passcode = word8

passcode is 'blue'

passcode = word9

passcode is 'The red'

passcode = passcode + ' f'

passcode is 'The red f'

passcode = passcode + word1

passcode is 'The red fox'

passcode = passcode + ' '

passcode is 'The red fox '

passcode = passcode + word6

password is 'The red fox trots'

The red fox trots
The passcode!

print(passcode)

Below you’ll find the Dog Age Calculator code so far. Using the experiment above as
a guide, add the code to compute the dog age in human years.

Throughout the
book you’ll find the
gray background
signifies new code
additions.

dog_name = input("What is your dog's name? ")
dog_age = input("What is your dog's age? ")
human_age = dog_age * 7
Let’s take the dog_age the user
print(human_age)

Let’s also print the human age
when we’re done so we can see the
final value.

entered and mulitply it by 7. We’ll
then take the result and assign it
to the variable human_age.

you are here 4  69

exercise solutions

So we know that our bug is caused by our assumption that dog_age is a number,
when in fact Python is treating it as a string. Take the code below and add an int
function to fix this problem.

There are quite a few ways to add
the int function. Here’s a few:
dog_name = input("What is your dog's name? ")
dog_age = input("What is your dog's age? ")
dog_age = int(dog_age)
human_age = dog_age * 7
First get dog_age as a string and then call the
int function on it, and then reassign the result to
print(human_age)

the dog_age variable.

dog_name = input("What is your dog's name? ")
dog_age = int(input("What is your dog's age? "))
human_age = int(dog_age) * 7
print(human_age)
Just wait until you are going to use dog_age as an

integer and call int there.

dog_name = input("What is your dog's name? ")
dog_age = int(input("What is your dog's age? "))
human_age = dog_age * 7
print(human_age)

Or finally, use int around the call to input. This
way, dog_age is always assigned to an integer.

If you find these confusing, we’ll be getting into how to use and call
functions later in the book. For now, just know you can pass a string to
the int function and have it convert the string to an integer.
70  Chapter 2

simple values, variables, and types

With your newfound print functionality, write the call to the
print function that will produce the user-friendly output:

Here's what you
need to output.

Output to the user:
“Your dog”
the dog’s name
“is”
the dog’s age in human years
“years old in human years”

print('Your dog',
Here we have each part of the output we need,
dog_name,
separated in the print function by commas.
'is',
human_age,
'years old in human years')

Note you can add vertical space to your
code like this for readability.

Brain Building Solution
Well, the chapter is almost over; you’ve just got the bullet points and a crossword to go. Why not do a few
extra reps for that brain before you go?
Here’s one for you: say you have two variables, first and last, like below. Can you swap their values?
See if you can write the code:

first = 'somewhere'
last = 'over the rainbow'
print(first, last)
A common technique is to use a temporary variable to store the first
temp = first
item while you set it to the second item. Then you set the second item
first = last
to the temporary variable.
last = temp
Trace through this until you understand how it works.
print(first, last)

you are here 4  71

crossword solutions

Coding Crossword Solution
1
3

C A
7

9

S

L
Y

4

C U

L

O

O

N

T A

C

T

A
F

L O A

12

S

T R

I

V

A

6

K

S
E

R

I

A

N G P O

A
Y W O R D
Y
B

L

E

I

N

C O D

I

T
I

11

R G U M E

N

O

15

T
18

R S

16

21

H

C O R E

E

M

L

20

S O N

M O D U

T

E Q U A

S

M

Y
C K

E

T S

P
N G

V

N

10

U N D E
A

5

M

O

T

J

8

N

N

19

X

T O R

R

A

17

I

A

E
13

I

14

T

2

F

L

E

R
L

S

Y

X

Across
3. What we're coding. [CALCULATOR]
6. Can't name a variable after one of these.
[KEYWORD]
7. Determines how we write statements in
Python. [SYNTAX]
8. Stores a value for later use. [VARIABLE]
9. The other numeric type. [FLOATINGPOINT]
72  Chapter
2
11. 12 years
old. [CODIE]
13. Name of values passed to a function.
[ARGUMENTS]

Down
1. Has a decimal point. [FLOAT]
2. The value of a variable can do this. [VARY]
4. Adding strings together. [CONCATENATION]
5. What statements mean. [SEMANTICS]
10. Input shows this to the user. [PROMPT]
12. Function to convert strings to integers. [INT]
15. Evaluates your code line by line. [SHELL]
16. Values are stored in this. [MEMORY]

3

booleans, decisions, and loops

Decisive Code

Have you noticed how, so far, our programs aren’t
very, well, interesting? That is, all our code has strictly been
a set of statements the interpreter evaluates from top to bottom—
no twists in the plot, no sudden turns, no surprises, no independent
thinking. For code to be more interesting, it needs to make decisions,
to control its own destiny, and to do things more than once straight
through. And in this chapter that’s exactly what we’re going to learn
to do. Along the way we’ll learn about the mysterious game called
shoushiling, meet a character named Boole, and see how a data type
with only two values could be worth our time. We’re even going to learn
how to deal with the dreaded infinite loop. Let’s get started!

We may even
create one just
for the fun of it!
this is a new chapter   73

the rock, paper, scissors game

Would you like to play a game?
Passed down from the ancient Chinese Han dynasty, the game shoushiling has been used
to settle court case decisions, to decide multimillion-dollar deals, and perhaps most
importantly, to determine who gets to sit in the front seat of the car.
Today you know the game as Rock, Paper, Scissors, and we’re going to implement it so that
you can play the game against a rather tough opponent: your computer.

ROCK

PAPER

SCISSORS

There are three possible hand
positions in the game Rock,
Paper, Scissors.

How Rock, Paper, Scissors works
If you’re never heard of Rock, Paper, Scissors we’re going to go over its simple rules now;
and if you do know the game, this will be a good review for you. The game is played by two
players, who each, upfront, secretly choose either rock, paper, or scissors. The two players
then typically count out loud to three (or shout “rock-paper-scissors!”) and then show their
choice through their hand position, which is either, you guessed it, a rock, paper, or scissors.
The winner can be determined by this chart:

WIN
WRAPS

SCISSORS

CUTS

ROCK

BLUNTS

PAPER

74  Chapter 3

SCISSORS PAPER

If both players make
the same choice, then
the game is a tie.

LOSE
ROCK

If one player chooses
scissors and the other
chooses rock, the rock
blunts the scissors, and
rock wins.

If one player
chooses paper
and the other
chooses rock,
the paper wraps
the rock, and
paper wins.
If one player
chooses paper and
the other chooses
scissors, the scissors
cut the paper, and
scissors wins.

booleans, decisions, and loops

If you’ve never played Rock, Paper, Scissors before, grab a
friend and play a few games. Even if you know the game,
grab a friend and refresh your memory.

How you’re going to play against the computer
Given the computer doesn’t have hands, we’ll have to change the way the game works,
at least a little—what we’ll do is have the computer preselect its choice of rock, paper,
or scissors, but not tell us. We’ll then enter our choice, and the computer will compare
the two before revealing the winner.
It helps to see an example of how the game is going to be played. Below in the Python
Shell you’ll find a few rounds of Rock, Paper, Scissors being played that show each
possible outcome: the user wins, the computer wins, and a draw.

Python 3.6.0 Shell
rock, paper or scissors? rock
User won, I chose scissors.
rock, paper or scissors? rock
We both chose rock, play again.
rock, paper or scissors? paper
Computer won, I chose scissors.

Computer prompts you for rock, paper,
or scissors and you respond with rock.
Computer determines you’ve won.
Let's run it again.
Looks like a tie, so try again.
Computer wins this time.

you are here 4  75

designing the game

First, a high-level design
The first thing we need to do is figure out the flow of the
game. Let’s put our pseudocode skills to use by reviewing
this high-level design for the game. Notice we’ve added
something new this time too: a diagram that helps map
out the flow of the game in the form of a flowchart.

eans
A circle mfinish.
start or

Here’s the basic idea:

1

A

2

The computer determines what
its choice is going to be: rock,
paper, or scissors.

Computer
makes
its choice

A

Get user
choice

A

Examine
choice

same

Game play begins.

A

Get the user’s choice.

B

Examine the user’s choice. If
it is invalid (not rock, paper,
or scissors), go back to step
2A.
If it is the same as the
computer’s, set the winner to
a tie and move on to step 3.

C

3

le is
A rectanrgepresent
used to n.
an actio

Start

1

User starts the game.

2

invalid

A diamond
represents a
decision point.

Tell the user who won along with
what the computer’s choice was.

Now we have a high-level idea of the kinds of
things the program needs to do. Next we’ll dig into
each step and figure out a few more details.

different

Determine
winner by rules

Determine who wins by the
rules of the game.

Game finishes.

B

3

C

Display winner

Game
over
Whoa! A real flowchart.

76  Chapter 3

Set winner
to tie

Think of the
flowchart as an
alternative way
to get a feel for
the flow of your
algorithm.

booleans, decisions, and loops

The computer ’s choice
Looking at our high-level design, the first thing we need to do is have the
computer make its choice in the game; that is, it needs to choose rock,
paper, or scissors. To make the game interesting, that choice should be
random and not predictable by the user.
Making random choices is a task many programs need to perform, and
you’ll find practically every programming language provides a way to
generate random numbers. Let’s see how we can get a random number in
Python, and how we turn that into a choice of rock, paper, or scissors.

How to generate a random number
Python ships with a lot of prebuilt code—that is, code you don’t have to
write yourself. You’ll often find prebuilt code supplied in the form of a module
(sometimes called a library), and we’ll be discussing modules in detail later in
the book. But for now, we’d like to use the random module, and to do that we
import it into our code using Python’s import statement.
Here’s how we do that:

Follow that by the name of the
module you want to use—in this
case, the random module.
Typically we place import statements
at the top of the code file so that
import random
you can easily keep track of all the
modules you’re importing.
another
just
is
module
a
r,
Remembe
We’ll be looking much more
it.
in
code
Python
with
file
seriously at functions (and
modules) later in the book, as
After you’ve imported the random module, you’re all ready to make use
but for now, think of them in
of the many random functions it provides. We’re going to use just one of
a way to make use of builtthem for now:
functionality in Python.
Use the import
keyword first.

We start with the
name of the module—
in this case random.

Then we add a period Then comes the
(otherwise known to function name, randint.
coders as a dot).

random.randint(0,2)
We pass randint
two numbers....

...this is a range, so randint
will give us random integers
between 0 and 2.

So randint will return 0, ,1 or 2.
We’re going to dive into all the specifics
of this notation later in the book, but
for now, just take it all in.
you are here 4  77

using random numbers

See for yourself. Get into the Python Shell, import the

random module, and then generate some random

integers with random.randint(0.2).

Keep calling
randint to see
what you get!

Q:
A:

How are random numbers going to help?

Think of generating a random number like
throwing the dice. In this case we have three
choices (rock, paper, scissors), so generating a
random number of 0, 1, or 2 is sort of like having a
dice with three sides. Once we generate a random
number, we’ll then associate that number with our
choices, so 0 = rock, 1 = paper, and 2 = scissors.

Python 3.6.0 Shell

>>> import random
>>> random.randint(0, 2)

What will you get?

Q:

Why do you start with 0 for the random
numbers? Why not generate the numbers 1, 2,
and 3? That makes more sense.

A:
How to use the random number
Alright, we know how to generate a random number of either 0, 1,
or 2, and we’re going to use that number to represent the computer’s
game choice. So, if we generate a 0, then the computer’s choice is
rock; if it’s a 1, it will be paper; and if it’s 2, the choice is scissors.
Let’s write that code:

Remember to import the
random module.
import random

We added an extra line (also
known as “whitespace”) just to
separate the imported module
from the actual code we’re
writing. That should help with
readability as our code grows.

random_choice = random.randint(0,2)

And we can’t forget to
assign our random number
to a variable so we can
actually make use of it
later in code.

78  Chapter 3

Here’s where we
generate the
random number
to represent the
computer’s choice.

Ah, not to a computer scientist.
Programmers usually think of sequences of
numbers starting at zero. This will start to feel
more natural (and sensical) as you see it used in
a variety of ways in code. For now, just go with
the flow.

Q:
A:

Are random numbers truly random?

No, random numbers generated by a digital
computer are pseudorandom, meaning not truly
random. Pseudorandom numbers, at some level,
have patterns that are predictable, whereas
truly random numbers do not. To generate true
random numbers, we have to make use of natural
phenomena like radioactive decay—not a very
convenient method for everyday use. For most
programming applications, though, pseudorandom
numbers are generally sufficient.

Q:
A:

So import gives me a way to access
Python code written by someone else?

Python developers take useful code and
make it available in modules. By using import,
you can make use of their code and use it along
with your own. For instance, the random module
includes many functions you can use to generate
random numbers. Right now we’re just making
use of the randint function, but we’ll be seeing
more of this module as the book progresses.

booleans, decisions, and loops

A Test Drive
Add this new
code to your file.

We added an extra
line just to provide
some output.

Just to get things rolling, go ahead and get this code into a file called rock.py,
save your code, and choose the Run > Run Module menu item to make sure
everything’s working.
Here’s the code so far:
import random
random_choice = random.randint(0, 2)
print('The computer chooses', random_choice)

Python 3.6.0 Shell
Here's what we got.
You might want to
try it a few times
to see the choices
are random.

The computer chooses 2
>>>

Taking things further...
By using the random module we’ve now implemented a way for the computer to randomly
make its choice, but it’s a little unsatisfying. Why? Well, our goal was to have the computer
choose rock, paper, or scissors, and we’ve done that by mapping those choices to the integers
0, 1, and 2, but wouldn’t it be nicer if we had a variable that was set to a string "rock",
"paper", or "scissors" instead? Let’s make that happen. But to do that we’re going to
have to step back and learn about how to make decisions in Python.
We’re

here.

Start

still

Computer
makes
its choice

Assume random_choice is already set to 0, 1, or 2 and write some pseudocode to set the variable
computer_choice to "rock", "paper", or "scissors" based on random_choice’s value.

You don’t know how to code
this in Python yet, but
remember pseudocode uses
English-like language. Don’t
overthink it.
you are here 4  79

true and false values

True? Or False?
Python makes decisions by asking questions with yes or no answers, only in
Python we call those true or false answers. The questions themselves are just
expressions, similar to the expressions you’ve already learned, but instead of
evaluating to strings or integers or floating-point numbers, they evaluate to True
or False. Here’s an example:

We call this a relational operator;
in
this case it's the greater than
Here’s an expression
operato
r, which is True if the
o
tw
comparing the values of s.
first
operand
is greater than the
ber
variables that hold num
second, and False if not.
We’re going to look at otherss
relational operators like “le
than,” “equal to,” and so on
bank_balance > ferrari_cost
in a bit.
The result of this expression is
either True or False depending on
whether the bank balance is greater
than the cost of the ferrari.

You can read this as
“is the bank_balance
greater than the
ferrari_cost?”

You can assign the result of this expression to a variable as well, and you can
even print it if you want.

Again, this is the expression.

decision = bank_balance > ferrari_cost
print(decision)

The decision variable holds
the (True or False) value
of the comparison.

Here’s the
output.

Python 3.6.0 Shell
False
>>>

No Ferrari today.

The values True and False belong to their own data type, the Boolean type.
Let’s take a look at it...
80  Chapter 3

lse

Fa

decision

booleans, decisions, and loops

Introducing the Boole an t ype
Oh, forgive us, we’ve been talking about a brand new data type, but we haven’t
formally introduced you. The Boolean data type is a simple one; it has only
two values, and, as you can guess, they are True and False.
Has only two values, True or False.

Booleans

True

True

False

Uses no quotes around values.
Always capitalized.
Expressions can evaluate to
True or False.

You can treat Booleans like any other type in that you can store them in a
variable, print them, or use them in expressions. Let’s get some practice in
with them, and then we’re going to see how to use them to make decisions.

Q:
A:

Boolean?

Right, pronounced Boo-lee-an.
Strange name, huh? Unlike numbers
and strings, Booleans are named
after a person, George Boole, a
19th-century mathematician who
formalized much of the logic we
use with computers today. While it
may sound overly formal, you’ll find
the word Boolean in common use
amongst programmers today. We’re
sure in no time it will be rolling right
off your tongue as well.

Get out your pencil and put some Boolean expressions through their paces. For each expression below,
compute its value and write in your answer. Be sure to check your answers at the end of the chapter.
Remember, Boolean expressions always evaluate to either True or False.

This tests if the first value is greater than the second. You can also use
>= to test if the first value is greater than or equal to the second.
your_level > 5

The == operator tests if two
values are equal to each other. It's
True if they are and False if not.
color == "orange"

The != operator tests if two values
are NOT equal to each other.

color != "orange"

When your_level is 2, what does this evaluate to? ________
When your_level is 5, what does this evaluate to? ________
When your_level is 7, what does this evaluate to? ________

Is this expression True or False when color has the value
"pink"? __________
Or has the value "orange"? __________

Is this expression True or False when color has the value
"pink"? __________
you are here 4  81

how to make decisions with code

That's two equals signs put together.

Serious Coding

y, while
Did you notice in the last exercise that the == operator tests for equalit
assign
to
sign
equals
one
use
we
is,
That
we’ve been using = for assignment?
equal to
values to variables, and we use two equals signs to test if two values are
mistake
coding
n
commo
a
is
other
the
for
one
each other. Accidentally substituting
nced).
experie
more
for beginners (and sometimes the

Making decisions
Now that we know about Boolean expressions and relational operators, like > and
< and ==, we can use them to make decisions in code. To do that we use the if
keyword combined with a Boolean expression. Here’s an example:

Start with the keyword if.

Then we have a Boolean
expression, sometimes called a
conditional expression, that
evaluates to True or False.

Next we have a colon.

if bank_balance >= ferrari_cost:
Then we have one or more
print('Why not?')
print('Go ahead, buy it') statements that will be executed
Notice that all the statements we want executed
when the conditional is True are indented.

if the condition is True.
The convention in Python is
to indent four spaces.

But we don’t have to stop there: we can supply an alternative set of statements to
execute if the conditional expression is False.

if bank_balance >= ferrari_cost:
print('Why not?')
We add an
else keyword.
print('Go ahead, buy it')
Next we have a colon.
else:
print('Sorry')
print('Try again next week')
Notice that all the statements we want executed when
the conditional is False are indented four spaces too.
82  Chapter 3

You'll find anytime we have a colon
in Python, it is followed by an
indented set of statements.

Then we have one or more
statements that will be executed
if the condition is False.

booleans, decisions, and loops

Decisions and more decisions
But there’s more: we can even set up a whole series of conditions, by using
the elif keyword. Admittedly, elif is a strange keyword, but it’s just a
contraction of “else if,” so don’t let it throw you. Let’s see how elif works:

Start with your first condition,
using an if keyword.

if number_of_scoops == 0:

print("You didn't want any ice cream?")
print('We have lots of flavors.')

Follow that with an elif
keyword and a second condition.

elif number_of_scoops == 1:

And then add any number of
print('A single scoop for you, coming up.') other elifs with their own
conditions.

elif number_of_scoops == 2:

print('Oh, two scoops for you!')

elif number_of_scoops >= 3:

print("Wow, that's a lot of scoops!")

else:

Remember, for each
if, elif, and else,
we can supply as
many statements to
execute as we like.

print("I'm sorry I can't give you negative scoops.")

And finally, you can supply a final else, which
acts as a catch-all if all previous conditions fail.

Note that only the code of the first True
condition will be executed, or if no conditions
are True, the else’s code will be executed.

Below you’ll find several potential values for the number_of_scoops
variable. Using each value, write down what the output of the code above
would be with number_of_scoops set to the indicated value. We did the
first one for you.

When number_of_scoops has this value, the output is...

number_of_scoops = 0

You didn’t want any ice cream?
We have lots of flavors.

Take a moment and READ the
code above out loud to yourself in
English (or your chosen language).
After you’ve gotten comfortable
speaking the code, jot down the
words you use.

number_of_scoops = 4
number_of_scoops = 1
number_of_scoops = 3
number_of_scoops = 2
number_of_scoops = -1

you are here 4  83

determining the computer’s choice

Back to Rock, Paper, Scissors

Start

We’re still finishing up the first stage of our Rock, Paper, Scissors game.
Remember, before our Boolean diversion, we wanted to improve our code so
that the computer could pick a string, "rock", "paper", or "scissors",
instead of a number 0, 1, or 2. Now that you’ve learned how to use Python’s
if statement, you’re all set to do that. What we’re going to do now is write
some code that, depending on the value of random_choice, sets a new
variable, computer_choice, to one of those three strings.

We’re still here.

Get user
choice

Here’s our code so far.
import random

Computer
makes
its choice

We don’t need this anymore, so
you can delete it.

random_choice = random.randint(0,2)

print('The computer chooses', random_choice)

if random_choice == 0:

Add this new code to
your file rock.py.
Check to see if random_choice is
0 and if so, set the computer’s
choice to the string “rock”.
Otherwise, check to see if
random_choice is 1 and if so, set
the computer’s choice to the
string “paper”.

computer_choice = 'rock'

elif random_choice == 1:

computer_choice = 'paper'

else:

computer_choice = 'scissors'

print('The computer chooses', computer_choice)

Otherwise, the only choice left
is “scissors”.

And just to test things, let’s
print out computer_choice.

A Test Drive
Make sure you have this new
code added to your file, and
then give it a test run.

84  Chapter 3

Here’s our first run; remember to try it a few times.

Python 3.6.0 Shell
The computer chooses scissors
>>>

booleans, decisions, and loops

Getting the user’s choice
Now that we have the computer’s choice, it’s time to get the user’s choice. After
Chapter 2, you’re a pro at getting user input. Let’s start by prompting the user
and storing the response in a variable called user_choice.

Now we’re here.

Computer
makes
its choice

Get user
choice

import random
random_choice = random.randint(0, 2)

invalid

Examine
choices

if random_choice == 0:

same

different

computer_choice = 'rock'

elif random_choice == 1:

computer_choice = 'paper'

else:

computer_choice = 'scissors'

We're assigning the string returned from the input
function to the variable user_choice.

We don’t need this debugging
print statement anymore.

print('The computer chooses', computer_choice)

We’re using the input function
again, and prompting for the
user’s choice in the game.

user_choice = input('rock, paper or scissors? ')

print('You chose', user_choice, 'and the computer chose', computer_choice)

Let's add a print statement just to keep
track of things as we’re coding this.

A Test Drive
Can you say quick turnaround?
Make sure you have this even
newer code added to your rock.py
file, and then give it a test run too.

Here’s our first run; remember to try this a few times.

Python 3.6.0 Shell
rock, paper or scissors? rock
You chose rock and the computer chose paper
>>>

you are here 4  85

a question on making a random choice

In Chapter 1 we used this random.choice
function in one of the examples. Would that
have been useful in picking the computer’s
choice?

Good eye.
We said there were a lot of other useful functions in the
random module, and there are. One of those is the function
choice, and here’s how it works:

First create a list of choices,
which is just a list of strings.

Dont’ be confused by the
square brackets; we'll be
learning about lists in the
next chapter.

choices = ['rock', 'paper', 'scissors']
computer_choice = random.choice(choices)

Then we pass our list to the
choice function, which will
randomly choose one item for us.

NOTE: for those Type A’s out
there dying to update their code
to use random.choice, go for it,
if you must. All you need to do
is replace every line between the
import and the input statement
with the code above. That said,
this isn’t necessary at this point,
but you know how you are, so
we’re letting you know.
86  Chapter 3

So now that you mention it, this is exactly what we’d want to
use if we were to implement this again because this approach
uses less code and is more readable. That said, if we’d used
this from the beginning of this chapter, we would have had no
reason to talk about decisions, or Boolean values, or relational
operators, or conditionals, or data types...well, you see the point.
But we’re glad you asked because choice is a great function to
use to do just this kind of thing, especially after the next chapter
when you totally understand lists.

booleans, decisions, and loops

Booleans Exposed
This week’s interview:

Getting to know Booleans
Head First: Welcome Boolean. We know
you’re super-busy out there, working in all those
Python programs, so we’re glad you could take
time out to talk to us.
Boolean: No problem. It’s true, I am busier
than ever these days. People are using Boolean
values all over the place. It’s nuts!
Head First: That’s amazing given you are, well,
just two values, yes or no.
Boolean: I’m actually the two values True or
False, not yes or no.
Head First: Right, of course, but either way,
you’re two values. Hardly enough to call you a
data type, no?
Boolean: Actually, Boolean values are pretty
core to any algorithm or program. Every
language has them, not just Python. I’m
everywhere.
Head First: Okay, big guy, you tell us then—
what’s the big deal?
Boolean: Think of a Boolean value
representing any current condition in your
code: Is the temperature above 98.6? Is the data
fully loaded? Is the list sorted yet? Did your
credit card charge go through? All of those are
Boolean values, True or False, that determine
where you code goes from there.
Head First: You’re talking about conditional
expressions. We test those conditions to
determine the code that we should evaluate next.
Boolean: The official term is Boolean expression,
and that is part of it, sure. You can test a
condition and then, using something like the if
statement, specify what is to come next. We call
it controlling the flow of the program.

Head First: Well, if that is just part of it, what’s
the rest of it?
Boolean: As your readers are going to see
you can also repeat code over and over while
a Boolean expression is True. Say you want to
keep prompting for a password while the user
is entering it incorrectly; say you want to keep
retrieving data while you haven’t reached the end
of it; say you want to keep updating the screen as
long as a game’s player is moving.
Head First: Okay, but I’m still not over the two
values thing. It’s hard to take you seriously.
Boolean: I’m a lot more than two simple
values. There’s a whole algebra named after
me—Boolean Algebra, maybe you’ve heard of
it?
Head First: No.
Boolean: Oh boy. Boolean Algebra is a whole
branch of mathematics that studies how we deal
with Boolean values.
Head First: Again, just the values True and
False?!
Boolean: Here’s an example: take the game
your audience is working on. Now there’s a
whole bit of logic that needs to determine who
wins. How do you take that graph of rock versus
paper versus scissors and translate that to code
that figures out the winner?
Head First: You got me.
Boolean: Okay, to solve that we need to...
Head First: Well, thank you so much, Boolean;
as always it’s been a pleasure to have you, but
it looks like we’re out of time. We’re looking
forward to the next time we can have you on.
Boolean: Yeah, I can, uh, barely wait.
you are here 4  87

examining the user’s choice

Taking a look at the user’s choice

Get user
choice

Now that we’ve got the user’s choice, let’s examine it.
According to our flowchart, we’ve got three possibilities:
invalid

The user and the computer made the same
choice, and we have a tie.

Now we’re
here.

The user and the computer made different
choices, and we need to determine who won.
The user entered an invalid choice, and
needs to enter another choice.

We’ll tackle these in order (saving the last one for a bit later in
the chapter), but first, let’s observe that no matter who wins,
or if there is a tie, we need some kind of variable to hold that
information. So let’s use a variable, winner, that will hold the
outcome of the game, which will be either 'Tie', 'User', or
'Computer'. Create that variable and give it an initial value
like this:

winner = ''

This is an empty string;
there is no space
between the two quotes.

same

Set winner
to tie

different

This possibility is where the user
enters a word that isn’t rock, paper,
or scissors; we’re going to come back
and handle this case a bit later.

We’ll add this to our
code in just a sec...

Here we’ve assigned the empty string to the new variable
winner as an initial value. An empty string is a string that has
no characters in it (yet it’s still a string). You might think of it like
this: a laundry basket is still a laundry basket even if it currently
has no laundry in it. Right? You’ll find this kind of thing pops
up all over programming languages: empty strings, empty lists,
empty files, and so on. For us, setting winner to an empty
string gives us a way to indicate that winner is going to be a
string, even if we’re not yet in a position to put any meaningful
characters in it (because we haven’t computed the outcome).
Now that we’ve created the winner variable to hold the
outcome of the game, let’s proceed with implementing the
possibilities at the top of the page. Looking at the first item
above, where the user and the computer make the same choice,
we’ll need to set winner to 'Tie'. To do that, we need to first
write the code to compare the user’s and computer’s choices,
and, again, if they are the same, then we’ll set our new winner
variable to 'Tie'.
88  Chapter 3

Examine
choices

Although there's nothing wrong with this
approach, later in the book we'll see an
alternative, that, for Python, is a better
way to provide an initial value for winner.

booleans, decisions, and loops

Your turn again. Based on our plan on the previous page, finish the code fragment below. Your code
should determine if there is a tie, and if so, set the winner variable to 'Tie'. After you’ve completed
this exercise, we’ll get this code into the rock.py file in the next step.

if ___________ == ___________:
winner = _______

Adding the code to detect a tie
We’ve got a new variable to add, winner, and we’ve got some new
code that compares the user’s and computer’s choice to see if they
are the same, in which case we have a tie. Let’s take a look at all the
code together:

import random
winner = ''

Here's our new variable, winner.
Right now it's just going to be set
to an empty string, but later it
will be set to the winner, which
will be either ‘User’, ‘Computer’, or
‘Tie’.

random_choice = random.randint(0,2)
if random_choice == 0:
computer_choice = 'rock'
elif random_choice == 1:
computer_choice = 'paper'
else:
computer_choice = 'scissors'

Go ahead and make these changes to
rock.py. We’ll test things thoroughly in
a bit when we’ve added more code. For
now, just run the code to make sure
you don’t have any syntax errors. Note,
because we’ve removed the print
statements, you won’t currently see any
output when you run your code.

You can go ahead and

remove this code.
user_choice = input('rock, paper or scissors? ')
print('You chose', user_choice, 'and the computer chose', computer_choice)
if computer_choice == user_choice:
winner = 'Tie'

And if the computer and the user make the same
choice, we're going to set the winner to 'Tie'.

you are here 4  89

determining who won

Who won?
So now that we’ve written the code to deal with a tie, we’re
ready for the interesting part of the code: figuring out who won. We
already have everything we need to decide a winner—we’ve got
the computer’s choice in the variable computer_choice, and
we’ve got the user’s choice in the variable user_choice. So
what we need at this point is to figure out the logic of determining
who won. To do that it really pays to study our Rock, Paper,
Scissors diagram to see if we can break the process of
determining the winner down into a simple set of rules. Here’s
another insight too: if we pick a side, say, by figuring out the ways
the computer can win, then we know if the computer doesn’t win,
the user does. That can really simplify our logic because we only
need to look at one set of cases.

To Do:
Now we need
to tackle this.

The user and the computer
made the same choice, and we
have a draw or tie.
The user and the computer
made different choices, and we
need to determine who won.
The user entered an invalid
choice, and needs to enter
another choice.

So with that in mind, let’s take a look at all the cases where the
computer wins:

If the computer
chooses paper and the
user chooses rock, the
computer wins.

If the computer chooses rock
and the user chooses scissors,
the computer wins.

PAPER

ROCK

If the computer
chooses scissors and
the user chooses paper,
the computer wins.

SCISSORS

CUTS

BLUNTS

WRAPS

Computer

User
ROCK
90  Chapter 3

SCISSORS PAPER

In all other cases
the user wins or we
have a tie.

booleans, decisions, and loops

So you’re saying we only need to
consider the cases where the computer
wins because if the computer doesn’t
win, we know the user wins?

Right.
Let’s step through this in more detail. We’ve already
determined if there is a tie, so we can rule out that case.
So, assuming there isn’t a tie, who won? The computer or
the user? Well, let’s start by looking at the cases where the
computer wins:
• The computer wins if it chooses paper and the user chooses rock.
• The computer wins if it chooses rock and the user chooses scissors.
• The computer wins if it chooses scissors and the user chooses paper.

In all other cases the computer does not win.
So what about the user? What are the cases where the user
wins? Well, we could enumerate those just like we did with
the computer, but do we really need to? We’ve assumed
there isn’t a tie (and our code already takes care of that),
and we’ve gone through the cases where the computer is
going to win, and if none of those applies, guess what: the
user wins. So we don’t need to write any code to determine
if the user wins, we just need to know the computer doesn’t
win in order to call the user a winner.
Let’s work through the actual logic to bring this all together.

you are here 4  91

implementing the game logic

How to implement the game logic
As you can see there are three ways the computer can win, and for each way we have to
test two conditions, like “did the computer choose paper?” AND “did the user choose
rock?” But, so far, in our coding, we’ve never had to test two conditions at once. That said,
we do know how to test for a single condition, like, if the computer chose paper:

computer_choice == 'paper'
And if the user chose rock:

user_choice == 'rock'

A simple Boolean expression
that we’re familiar with at
this point, which asks if the
computer’s choice is paper
And another expression asking
if the user’s choice is rock

But how do we test for both conditions?
To do that we can use a Boolean operator. It sounds fancy, but Boolean operators are
just a way to combine Boolean expressions together, and, for now, there are only three of
them to know about: and, or, and not.

We’ll see one
additional Boolean
operator in a bit.

To test if the computer chose paper AND the user chose rock, we can use the and Boolean
operator and combine our expressions, like this:

Here’s our first condition.

Here’s our second condition.

computer_choice == 'paper' and user_choice == 'rock'

Placing an and operator between them means this entire
expression will be True if and only if both conditions are True.

This entire phrase
is a Boolean
expression and will
evaluate to either
True or False.

And we can use this Boolean expression with an if statement:

Now the if statement's conditional expression is
the entire combined Boolean expression.

if computer_choice == 'paper' and user_choice == 'rock':
winner = 'Computer'

So this code handles one of the three
cases where the computer wins.
92  Chapter 3

If the expression is True, then
we execute the if's code block.

booleans, decisions, and loops

More about Boolean operators
As you’ve already seen, the and operator is True if, and only if, both of its
conditions (or we can call them operands) are True. But what about or
and not; how do they work? Like and, or is used to combine two Boolean
values or expressions, and is True if either of those values or expressions
evaluates to True.

You've got the money for the Ferrari
if you have enough money in the bank
balance...

OR, you have a loan that is equal
to the cost of the Ferrari.

if bank_balance > ferrari_cost or loan == ferrari_cost:
print('Buy it!')

Note that only one of these conditions needs to be True
to get the Ferrari, but both can be True as well. If both
are False, then you'll have to wait for the Ferrari.

Serious Coding
Wondering about precedence
with Boolean operators? The
relational operators (>, <,
==, and so on) are highest,
followed by not, or, and
then and. You can also add
parentheses to your Boolean
expressions to override the
default precedence or to add
clarity to your expression.

The not operator, on the other hand, when placed before any single Boolean value
or expression, gives you the opposite of the Boolean value—in other words, if not’s
operand evaluates to True then not evaluates to False, and if its operand is False
then not evaluates to True. We like to say that not negates its operand.

Here we’ve put a not in front of a
Boolean expression.

First we evaluate this relational operator to
True or False, and then the not operator is
applied, evaluting to the opposite Boolean value.

if not bank_balance < ferrari_cost:
print('Buy it!')
You can read this as “if the bank account is

NOT less than the Ferrari cost,” then buy it.

Take out your pencil and put some more Boolean expressions through their paces.
For each expression below, compute its value and write in your answer.
age > 5 and age < 10
age > 5 or age == 3

Notice we added parens here, which
makes this more readable.

not (age > 5)

When age is 6, what does this evaluate to?

________

When age is 5, what does this evaluate to?

________

When age is 6, what does this evaluate to?

________

When age is 11, what does this evaluate to? ________

When age is 2, what does this evaluate to?

________

When age is 3, what does this evaluate to?

________

When age is 6, what does this evaluate to?

________

When age is 2, what does this evaluate to?

________

you are here 4  93

figuring out the game logic

Code Magnets

We’d already worked out the code for the game
logic on the fridge with code magnets, but someone
came along and knocked most of it on the floor. Can
you reconstruct the code snippets so that we can
determine the winner? Notice, there may be some
extra code magnets, so you may not use all of them.
Check your answer at the end of the chapter.

computer_choice == user_choice

if

winner
elif

=

'Tie'

winner = 'Computer'

else

or

and

and

if

elif

if

:

:

This is the first case, where the
computer and user made the same choice,
so we have a tie.

:

computer_choice == 'paper'

or

and

and

elif

if

elif

:

and

user_choice == 'rock'

:

the
You’re lucky they didn’t knockputoff
n, and
wo
er
first test to see if the com user wins.
the very last part where the

:

Place your magnets here.

winner == 'User'

winner = 'Computer'

r'
winner = 'Compute

user_choice == 'sciss
ors'

94  Chapter 3

user_choice == 'paper'

computer_choice == 'rock'

computer_choice == 'scissors'

winner = 'User'

=

booleans, decisions, and loops

A Test Drive

Alright, we’ve got our game logic all figured out. Let’s get it typed into rock.py and
then give the game a few tries. Of course we haven’t added our user-friendly output
yet, so we’re just going to find out who wins, not what the computer’s choice was.
That’s next on our list.

import random

Examine
choice

winner = ''
random_choice = random.randint(0,2)
if random_choice == 0:
computer_choice = 'rock'
elif random_choice == 1:
computer_choice = 'paper'
else:
computer_choice = 'scissors'

You are here.

user_choice = input('rock, paper or scissors? ')

Determine
winner by rules

Here's the code for the
game logic. Go ahead and
enter this code.

if computer_choice == user_choice:
winner = 'Tie'
elif computer_choice == 'paper' and user_choice == 'rock':
winner = 'Computer'
elif computer_choice == 'rock' and user_choice == 'scissors':
winner = 'Computer'
elif computer_choice == 'scissors' and user_choice == 'paper':
winner = 'Computer'
else:
winner = 'User'
print('The', winner, 'wins!')

Play it a few times and make sure
it appears to be working. We'll add
more user-friendly output next!
You may even see “The
Tie wins!” but we’ll fix
that in a sec.

Python 3.6.0 Shell
rock, paper or scissors? rock
The Computer wins!
>>>

you are here 4  95

displaying the winner

Display the winner
Now it’s time to display the winner. If you look at the sample output again,
either the user or the computer wins, or there’s a tie.

Now we’re
here.

Display winner

Python 3.6.0 Shell

Game
over

rock, paper or scissors? rock

User won, I chose scissors.

rock, paper or scissors? rock

We both chose rock, play again.

rock, paper or scissors? paper

Computer won, I chose scissors.
Let’s first take care of the code to handle the tie. Looking at our existing code,
if there’s a tie then the winner variable will be assigned to the value 'Tie'.
So, let’s set up a condition for this case:

If winner has been
set to ‘Tie' then...

Remember that print adds a space
betweens values separated by commas.
So, sometimes we’ll use string
concatenation if we don’t want
spaces (like if we need to follow a
word immediately by punctuation).

if winner == 'Tie':
print('We both chose', computer_choice + ', play again.')

Print the message along with
the choice both players made.

Note we could have used user_choice here
too. Because there is a tie, they both
have the same value.

If there isn’t a tie, we need to announce the winner, which is conveniently
stored in the winner variable.

This code will be run,
only if there isn't a tie.

else:
print(winner, 'won, I chose', computer_choice + '.')

We announce the winner.
96  Chapter 3

And then report what the
computer's choice was.

booleans, decisions, and loops

A Test Drive

We should have a fully functioning game now! Let’s get the new
code typed into rock.py and then give it a spin.

import random
winner = ''
random_choice = random.randint(0,2)
if random_choice == 0:
computer_choice = 'rock'
elif random_choice == 1:
computer_choice = 'paper'
else:
computer_choice = 'scissors'
user_choice = input('rock, paper or scissors? ')
if computer_choice == user_choice:
winner = 'Tie'
elif computer_choice == 'paper' and user_choice == 'rock':
winner = 'Computer'
elif computer_choice == 'rock' and user_choice == 'scissors':
winner = 'Computer'
elif computer_choice == 'scissors' and user_choice == 'paper':
winner = 'Computer'
else:
Here's the code for
winner = 'User'
Remove this.

game
output. Go ahead and
enter this code.

print('The', winner, 'wins!')

if winner == 'Tie':
print('We both chose', computer_choice + ', play again.')
else:
print(winner, 'won. The computer chose', computer_choice + '.'

Alright, we have a fully
functioning game! Give it a whirl.

Python 3.6.0 Shell

rock, paper or scissors? rock
Computer won. The computer chose paper.
>>>

you are here 4  97

documenting code

Got documentation?
It’s a good time to step back and look at all the code you’ve written. There’s actually
enough code that if you revisited it in the future you might have to remind yourself
of what each piece does and how it all fits together. You might also have to study the
code to remember the design decisions you made and why you made them.
Also notice that the code has an inherent structure and is pretty well organized in
that it’s divided into pieces that handle the parts of our algorithm (or the actions in
the corresponding flowchart). Let’s mark these sections and also add some notes to
remind us in the future of how all this works.

Here we’re doing some setup
by importing the random
module and setting up the
winner variable.
The computer randomly
chooses rock, paper, scissors
by generating a random
number from 0 to 2 and
then mapping that to a
corresponding string.

Get the user’s choice with
a simple input statement.
Here’s our game logic,
which checks to see if
the computer wins (or
not), and makes the
appropriate change to
the winner variable.
Here we announce the game
was a tie, or the winner along
with the computer's choice.

import random
winner = ''
random_choice = random.randint(0,2)
if random_choice == 0:
computer_choice = 'rock'
elif random_choice == 1:
computer_choice = 'paper'
else:
computer_choice = 'scissors'
user_choice = input('rock, paper or scissors? ')
if computer_choice == user_choice:
winner = 'Tie'
elif computer_choice == 'paper' and user_choice == 'rock':
winner = 'Computer'
elif computer_choice == 'rock' and user_choice == 'scissors':
winner = 'Computer'
elif computer_choice == 'scissors' and user_choice == 'paper':
winner = 'Computer'
else:
winner = 'User'
if winner == 'Tie':
print('We both chose', computer_choice + ', play again.')
else:
print(winner, 'won, I chose', computer_choice + '.')

But isn’t this silly that we’re documenting this code in a book? After all, you’ve got real,
live code on your computer. Why don’t we document the actual code so the documentation
is right there when you need it? Let’s see how to do that.

98  Chapter 3

It’s also handy to document
your code for anyone else who
might want to take a look at
it, like another programmer.

booleans, decisions, and loops

How to add comments to your code
With Python, and pretty much any programming language, you can add human-readable
comments right into your code. To add a comment with Python, type a hash character
(#) on any line, and then your comment. Python will conveniently ignore anything you
type after the hash. With comments, the general idea is to add remarks to your code that
are going to be read by you, or other programmers, with the goal of providing additional
information about the design, structure, or approach you used in the code. Let’s look at
an example:

Start your comment with a hash character and then type your
human-readable text after it. Each new line needs its own hash.

Comments are one form
of documentation; later
in the book we’ll look
at help documentation,
which is meant for
coders who just want
to use your code, not
necessarily understand it.

# This code supports my weekly habit of seeing if I can
# afford a ferrari. The general algorithm is to compare

# my bank account amount to the current cost of a ferrari.
if bank_balance >= ferrari_cost:

# If it's greater I can finally buy one.
print('Why not?')
print('Go ahead, buy it')

else:

print('Sorry')
print('Try again next week')

# bummer

Comments can start
anywhere on the line.

They can even start
after code on a line.

Do as we say, not as we do.
In the IDLE editor add comments to your
rock.py file. Feel free to use our comments
on the previous page as a starter, but also
make your own comments. You’ll want your
comments to be meaningful to anyone who,
in the future, might read your code (including
you!). Check our solution at the end of the
chapter to see how we did it.

Documenting code is an important part of
coding, but you’re going to notice in this
book we don’t do a lot of it. That’s because
our method of using handwritten code
annotations is more effective in Head First
book form (and all those comments tend
to kill lots of trees when we have to include
them in every code listing).

Annotations like these.

you are here 4  99

dealing with invalid input

Thinking about the Rock, Paper, Scissors game again, what happens if you don’t enter rock, paper, or
scissors correctly? Say you enter “rack” instead of “rock”: How does your program behave? Do you think
that behavior is correct?

We need to finish that game!
You realize that we haven’t quite finished our game, right? Check
out the To Do list: we haven’t dealt with that possibility of invalid
user input. Now the user is supposed to enter “rock” or “scissors”
or “paper,” but they might not; they might mistype, like “scisors,”
or they might just be troublemakers who decide to enter “dog,”
“hammer,” or “no.” So, when you’re creating an app or program
that’s going to be used by actual people, you want to keep in mind
that they often make mistakes, and your code needs to deal with
that.

To Do:
The user and the computer
made the same choice, and we
have a tie.
The user and the computer
made different choices, and
we need to determine who won.
The user entered an invalid
choice, and needs to enter
another choice.

We forgot
something.

So, let’s deal with it.

Get user
choice

But first we have to figure out how we want the game to behave
when the user enters an invalid answer. Looking back at the
flowchart, our original intent was to have the program reprompt the
user if the input was invalid.
invalid

Perhaps something like this:

Python 3.6.0 Shell
rock, paper or scissors? rack

rock, paper or scissors? papper
rock, paper or scissors? rock
User won, I chose scissors.
>>>

If we get
invalid input, we
can just keep
prompting until
we get valid
input.

We can always make it more elaborate later, but for now we’ll just reprompt
the user until we get a valid input.
Are you ready to get this coded and finish this game? We just need to make
sure we know how to approach coding two aspects of this:
1. How do we detect invalid input?
2. How do we continually prompt the user until we get a valid answer?
100  Chapter 3

Examine
choices

same

Remember
this part we
left out?

Users often make
mistakes. Make sure
your code anticipates
and handles these
mistakes—even if the
only user is you.

booleans, decisions, and loops

How do we know if the user’s choice is invalid?
How do we detect if the user’s input is invalid? Well, you probably know we’re going to make
use of our new Boolean logic skills, but what does an expression that detects invalid answers look
like? Sometimes it’s good to just talk things out: we know if the user’s choice is invalid if:

The user’s choice is NOT
EQUAL TO ‘rock’ AND...

The user’s choice is NOT
EQUAL TO ‘paper’ AND...

The user’s choice is NOT
EQUAL TO ‘scissors’

Convert the spoken English above into its Boolean expression equivalent. We’ve done the
first part for you.
user_choice != 'rock' and

Complete the Boolean expression.
you are here 4  101

improving readability

Checking out and cleaning up the expression
Hopefully your Boolean expression in the last Sharpen exercise was close to
our solution. Here it is again, this time as part of an if statement:

Wow, long and hard to read!

if user_choice != 'rock' and user_choice != 'paper' and user_choice != 'scissors':

This statement says, basically, if
the user’s input is invalid...

That looks like a perfectly acceptable statement. But sometimes really long
lines like this are quite unwieldy once we start typing them into an editor,
or if we have to go back and read them later. It would be nice if we could
reformat the statement a bit and make it look more like:

if user_choice != 'rock' and

Ah, much better and
easier on the eyes!

user_choice != 'paper' and
user_choice != 'scissors':

Uh oh, not
good

Python 3.6.0 Shell

>>>

if user_choice != 'rock' and
^
SyntaxError: invalid syntax

The only problem is when we try to break the code into more than one
line, Python complains about our syntax.
There is another way—we can wrap a set of parentheses around the
expression, like this:
if (user_choice != 'rock' and

Get user
choice

user_choice != 'paper' and

user_choice != 'scissors'):

Wrap parens around your expression and
then you can break it into multiple lines.
And Python is just fine with the reformatting of the code.
Okay, now that we know how to detect an invalid user choice, we still
need to figure out how to reprompt the user.. Let’s spend a little time
thinking through how that might work....

102  Chapter 3

invalid

We’re here now:
if the input is
invalid, we need to
reprompt the user.

Examine
choices

same

booleans, decisions, and loops

Before you go on to solve the
reprompting problem, I had a question
about the user input. Are we considering
'ROCK' or 'Rock' or other variations of
'rock' as valid entries too?

No, but we could, and you bring up a good
point. First of all, what is the issue here? Well, the strings
'rock' and 'ROCK', for example, are different strings because
Python treats strings as case sensitive. In other words, in Python
(and almost every programming language), the following
equality test would evaluate to False:
'rock' == 'ROCK'

False

So if the user enters Rock instead of rock, right now our code
would say that entry was invalid (and our logic code, for that
matter, wouldn’t know what to do with Rock).
That said, the suggestion does seem very reasonable—after
all, if you enter the word rock no matter the capitalization, it
should count as a valid answer.
So what do we do? Well, we could just add in additional logic
to test all permutations of upper- and lowercase letters for the
words rock, paper, and scissors, and that would work.
However, it would make our code very complex, and there are
better ways to approach this problem that we’re going to learn
about later in the book.
But right now, let’s just assume that the user needs to enter an
answer in lowercase, and we’ll point out how this could have
been more easily solved when we get to it later in the book.

you are here 4  103

how to reprompt the user

While you were discussing upper- and lowercase strings, we went ahead and wrote
the code to reprompt the user, only our approach was flawed. Can you take a look
and see what you think the flaw might be? You’ll find the code, some notes, and a
sample run of the code below.
user_choice = input('rock, paper or scissors? ')

if (user_choice != 'rock' and

user_choice != 'paper' and

First we get the
user’s choice.

Then we test to see
if the input was valid.

user_choice != 'scissors'):

user_choice = input('rock, paper or scissors? ')

print('User chose', user_choice)

If not, then we get the
user’s choice again.

Python 3.6.0 Shell

We thought we were on the
right track, but what do you
think we’re missing?

rock, paper or scissors? rack
rock, paper or scissors? papper
User chose papper

>>>

Ugh!

We're getting
reprompted.
But then we can
just enter invalid
input again.

How to continually prompt the user
Our first attempt failed. We tried to test the user input and then if it wasn’t
valid, prompt again. The problem is, this solution only works once. If the user
enters “rocknroll” on the second try, then that string will be accepted as the
valid user input.
Now we could keep adding if statements for a second and third and fourth try,
but that would lead to a coding mess, and our requirements are to reprompt
the user as many times as it takes.
The problem is, given our Python knowledge, we only know how to do things
once. What we really need to be able to do is write code in a way that it can
repeat over and over, as many times as needed. We need a way to do things
more than once.
104  Chapter 3

Get user
choice

invalid

Examine
choice

We’re here; we need to keep
getting the user’s choice
until it’s valid.

booleans, decisions, and loops

while juggling:

keep_balls_in_air()

Doing things more than once
You do a lot of things more than once:
Lather, rinse, repeat…
Wax on, wax off…
Keep turning the pages of the book, until it’s done.
Of course you’ll often need to do things in code more than once, and Python
gives you a couple of ways to repeatedly execute code in a loop using its
while and for statements. We’ll look at both of these ways of looping, but
let’s focus on while for now.
We’ve talked a lot about expressions that evaluate to Boolean values, like
scoops > 0, and these kinds of expressions are the key to the while
statement. Here’s how the while statement works:

A while statement starts
with the keyword while.

while uses a Boolean
expression as a conditional
test (just like the if
statement).

If the conditional is true,
everything in the code block
is executed.

while scoops > 0:

print('Another scoop!')
scoops = scoops - 1

After we execute the code block, we loop back around
and do the conditional test again. If the conditional is
still True, we execute the code block again. When the
conditional is False, we’re done executing the block.

What’s a code block? We haven’t
introduced the term formally, but it’s
everything indented after the colon
While Python uses indenting to identify
a block of code, this is actually a bit
unusual, as most programming languages
use matched pairs of curly braces or
parentheses.
you are here 4  105

understanding the while loop

How the while loop works
Seeing as this is your first while loop, let’s trace through a round of its
execution to see exactly how it works. Notice we’ve added a declaration for the
variable scoops at the top of the code, and initialized it to the value 5.
Now let’s start executing this code. First we set scoops to 5.
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

A note from readers who
have read this before you:
read the next several pages
slowly and carefully. There’s
a lot to take in and you
really want to get how this
works into your brain.

After that we encounter the while statement. When we evaluate a
while statement, the first thing we do is evaluate the conditional to
see if it’s True or False.

Is scoops greater
than zero? Looks
like it to us!

scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

Because the conditional is True, we start executing the block of code.
The first statement in the body prints the string "Another scoop!" to
the shell.
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

106  Chapter 3

Python 3.6.0 Shell

Another scoop!

booleans, decisions, and loops

1 scoop gone,
4 left!

The next statement subtracts one from the number of
scoops and then sets scoops to that new value, 4.
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

That’s the last statement in the block, so we loop back up
to the conditional and start over again.
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

Observe that our code never
changes. But the variables
in our code, like scoop, are
changing throughout the
computation. At this stage,
scoops is equal to 4.

Evaluating our conditional again, this time scoops is 4. But that’s
still more than zero.

Still plenty left!

scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

Once again we write the string "Another scoop!" to the shell.
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

Python 3.6.0 Shell

Another scoop!
Another scoop!

you are here 4  107

understanding the while loop, continued

2 scoops gone,
3 left!

The next statement subtracts one from the number of
scoops and sets scoops to that new value, which is 3.
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

That’s the last statement in the block, so we loop back up
to the conditional and start over again.
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same")

Evaluating our conditional again, this time scoops is 3. But that’s
still more than zero.

Still plenty left!

scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

Once again we write the string "Another scoop!" to the shell.
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

108  Chapter 3

Python 3.6.0 Shell

Another scoop!
Another scoop!
Another scoop!

booleans, decisions, and loops

3 scoops gone,
2 left!

And as you can see, this continues. Each time we loop,
we decrement (reduce scoops by 1), write another
string to the browser, and keep going.
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

And continues...

Python 3.6.0 Shell

Another scoop!
Another scoop!
Another scoop!

Another scoop!

4 scoops gone, 1 left!

scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

Until the last time...this time something’s different. scoops is
0, and so our conditional evaluates to False. That’s it, folks;
we’re not going to go through the loop anymore, and we’re
not going to execute the block. This time, we bypass the
block and execute the statement that follows it.
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

Now we execute the other print statement, and write the
string "Life without ice cream isn't the same". We’re done!
scoops = 5
while scoops > 0:
print('Another scoop!')
scoops = scoops - 1
print("Life without ice cream isn't the same.")

5 scoops gone, 0 left!

Python 3.6.0 Shell

Another scoop!
Another scoop!
Another scoop!

Another scoop!
Another scoop!

Python 3.6.0 Shell

Another scoop!
Another scoop!
Another scoop!
Another scoop!

Another scoop!
Life without
ice cream isn'
t the same.

you are here 4  109

using while to reprompt the user

w it

Write a quick game. Here’s how it works: you prompt the
player with “What color am I thinking of?” and you see
how many guesses it takes the player to guess it.
color = 'blue'
guess = ''
guesses = 0
while
:
guess = input('What color am I thinking of? ')
guesses = guesses + 1
print('You got it! It took you', guesses, 'guesses')

How to use while to prompt the
user until you get a valid choice
Now that you know how to use while, you’re all ready to get this
code reprompting the user. To do that we just need to make a
couple simple changes to the previous attempt: we’re going to start
by initializing user_choice to the empty string, and then we’re
going to replace the if keyword with while.
Like this:

First set user_choice to the empty string.
user_choice = ''

while (user_choice != 'rock' and

user_choice != 'paper' and

user_choice != 'scissors'):

We’re setting user_choice to
the empty string because when
it first enters the while loop
it needs a value.

While the user_choice isn’t
valid, we’ll keep executing the
code body.

user_choice = input('rock, paper or scissors? ')

Each time through the loop, our code
prompts the user for a choice, and assigns
the input to the user_choice variable.
So, when we finally get a valid choice,
the while loop stops and we have that
choice in the user_choice variable.
110  Chapter 3

booleans, decisions, and loops

A Test Drive
import random

In your rock.py file, replace the input statement with
the new while loop, and then give it a final test run.
At this point, the game should be complete!

winner = ''
random_choice = random.randint(0,2)
if random_choice == 0:
computer_choice = 'rock'
elif random_choice == 1:
computer_choice = 'paper'
else:
computer_choice = 'scissors'

Delete the old input
statement.

user_choice = input('rock, paper or scissors? ')
user_choice = ''
while (user_choice != 'rock' and
user_choice != 'paper' and
user_choice != 'scissors'):
user_choice = input('rock, paper or scissors? ')

Here’s our new code to
handle the user input. This
code will keep prompting
the user until they enter
rock, paper, or scissors.

if computer_choice == user_choice:
winner = 'Tie'
elif computer_choice == 'paper' and user_choice == 'rock':
winner = 'Computer'
elif computer_choice == 'rock' and user_choice == 'scissors':
winner = 'Computer'
elif computer_choice == 'scissors' and user_choice == 'paper':
winner = 'Computer'
else:
winner = 'User'
if winner == 'Tie':
print('We both chose', computer_choice + ', play again.')
else:
print(winner, 'won. The computer chose', computer_choice + '.')

Now we have a fully functional game. Our
game randomly picks a computer choice,
prompts the user until it receives a valid
choice, and then figures out the winner (or
if the game is a tie)!

Python 3.6.0 Shell

rock, paper or scissors? scisors
rock, paper or scissors? rock
User won. The computer chose scissors.
RESTART: /ch3/rock.py
rock, paper or scissors? papper
rock, paper or scissors? rocker
rock, paper or scissors? paper
Computer won. The computer chose scissors.
you are here 4  111

some extra credit

Congratulations on coding your first game!
What’s the best thing to do after coding your new game? Play a few rounds,
of course! Sit back, relax, and let everything in this chapter sink in as you try
to defeat the computer at Rock, Paper, Scissors. Of course, you’re not quite
done yet—you’ve still got the extra credit, the bullet points, and a crossword
to do, but take some time and enjoy the game first.

extra credit
w it

Remember the color guessing game? It has a bug. Have you
noticed if you guess the color correctly on the first try, it prints:
"You got it! It took you 1 guesses". Can you fix
the bug so that it prints “guess” for one guess and “guesses” for
multiple guesses?
color = 'blue'
guess = ''
guesses = 0

Your code
fixes here

112  Chapter 3

while guess != color:
guess = input('What color am I thinking of? ')
guesses = guesses + 1
print('You got it! It took you', guesses, 'guesses')

booleans, decisions, and loops

Beware of the

The Dreaded Infinite Loop
Before we wrap up this chapter we need to talk about infinite
loops. You see, when you write code without loops it just goes
straight through—you know it’s going to end, someday. But with
loops, things get more interesting.
Let’s say you’ve just written your latest code, you feel good
about it, and you confidently run it. What happens next? Nothing.
Any output slows to a crawl. Your program seems to be doing
something, but what, you’re not quite sure. Whatever it is, it’s
taking a long time.
You just encountered an infinite loop—that is, a loop that is
looping and looping and is never going to end, ever.
It’s easier to get into this situation than it sounds. In fact, sooner
or later you’re going to encounter one, so it might as well be
now. Let’s create one:
counter = 10
while counter > 0:
print('Counter is', counter)
counter = counter + 1
print('Liftoff!')

Be our guest: type it
in and run it.

Python 3.6.0 Shell

Counter is 814304
Counter is 814305
Counter is 814306
Counter is 814307
Counter is 814308
Counter is 814309
Counter is 814310

Uh oh, this is
out of control.

So what do you do when you have an out-of-control program
running on your computer? If you’re using IDLE, simply close
the shell window to terminate the program. If you’re using your
computer’s command line, then typically a tap of Control+C
(Ctrl+C on some keyboards) will terminate the program as well.
And what do you do with your code? Well, infinite loops are
logic errors. You’ve create some logic that never lets the loop
end, so examine the conditional of your loop (or loops) and
trace through the execution of your code until you determine
what about the conditional logic is wrong. In our case, we
simply need to rewrite the counter + 1 as counter - 1, so
that the code counts down.

you are here 4  113

review of booleans, decisions, and loops

 Many programs make use of random
numbers.

 The elif keyword is a contraction of “else if.”

 Practically all languages provide a means of
generating random numbers.

 The else keyword can be used to provide
a final alternative or catch-all for an if
statement.

 Python provides a random module for
generating random numbers.

 Boolean expressions can be combined with
the Boolean operators and and or.

 Use the import statement to include
Python’s random functionality in your code.

 The Boolean operator not can be used to
negate a Boolean expression or value.

 The Boolean data type has two values,
True and False.

 The while statement evaluates a Boolean
expression and executes a code block while
the expression remains True.

 Boolean or conditional expressions evalute
to True or False.
 Relational operators, like ==, >, and <,
compare two values.
 Relational operators are provided for
numbers and strings.
 Boolean expressions provide the foundation
of the if statement.
 The if statement evaluates a Boolean
expression and then, if True, executes a
code block.
 A code block is a set of Python statements
that are executed together.
 Code blocks are indented sections of code.
 The elif keyword can be used to test
additional conditionals in an if statement.

114  Chapter 3

 We call a string without any characters an
empty string.
 You use = for assignment and == for
equality testing.
 You can add comments to your code by
using the hash character (#) followed by
arbitrary text.
 It’s a good idea to use comments to add
documentation to your code so you can
remember your design decisions later (or so
others can understand your code).
 Anticipating user error is an important part
of designing a user-centered program, like
a game.
 Logic errors can lead to infinite loops.

booleans, decisions, and loops

Coding Crossword
Let’s give your right brain something to do.
It’s your standard crossword, but all of the
solution words are from Chapter 3.
1
2
3
4

5

6

7
8
10

9
11

13

12

14
15
16
18

17

19

20

21

22

23

Across

Down

2. == is for.
4. >, <, and == are for.
6. Keeps executing as long as True.
7. Has two values.
8. If its Boolean expresssion is True it executes
a code block.
10. Character for comments.
11. = is for.
13. Boolean is another one of these.
14. Cuts paper.
15. A string with nothing in it.
19. Many programs need to generate these.
20. Paper wraps ______.
21. Rock blunts.
22. Alternative for if.

1. Can't afford.
3. Foundation of if and while.
5. Ancient Chinese game.
9. True or _____.
12. Catch-all.
16. How you include the random module.
17. Boole's first name.
18. We often say repeating code is in one of
these.

you are here 4  115

exercise solution

See for yourself. Get into the Python Shell, import the random module, and then
generate some random integers with random.randint(0.2):

Your output should look slightly
different because these are
random numbers! Perhaps you’ll
see 2, 2, 0, 0, or maybe 1, 2,
0, 1, or maybe 1, 1, 1, 1 (it could
happen!) or who knows—that’s the
fun of random numbers!

Python 3.6.0 Shell

>>> import random
>>> random.randint(0,2)
1
>>> random.randint(0,2)
0
>>> random.randint(0,2)
1
>>> random.randint(0,2)
2
>>>

Assume random_choice is already set to 0, 1, or 2 and write some pseudocode to
set the variable computer_choice to "rock", "paper", or "scissors" based on
random_choice’s value.

if random_choice equals 0, then set computer_choice to “rock”
otherwise, if random_choice equals 1, then set computer_choice to “paper”
otherwise, set computer_choice to “scissors”

116  Chapter 3

booleans, decisions, and loops

Get out your pencil and put some Boolean expressions through their paces. For each
expression below, compute its value and write in your answer.

This tests if the first value is greater than the second. You can also use
>= to test if the first value is greater than or equal to the second.
When your_level is 2, what does this evaluate to? False
________

your_level > 5

This is a Boolean expression. The
== operator tests if two values
are equal to each other.

When your_level is 5, what does this evaluate to? False
________
True
When your_level is 7, what does this evaluate to? ________

color == "orange"

Is this expression True or False when color has the value
False
"pink"? __________
True
Or has the value "orange"? __________

color != "orange"

True
"pink"? __________

The != operator tests if two values
are NOT equal to each other.

Is this expression True or False when color has the value

Below you’ll find several values of number_of_scoops. Using each value, write down what
the output of the code above would be when run with the given value. We did the first one for
you.

When number_of_scoops has this value, the output is...

number_of_scoops = 0

You didn’t want any ice cream?
We have lots of flavors.

number_of_scoops = 4

Wow, that’s a lot of scoops!

number_of_scoops = 1

A single scoop for you, coming up.

number_of_scoops = 3

Wow, that’s a lot of scoops!

number_of_scoops = 2

Oh, two scoops for you!

number_of_scoops = -1

I’m sorry I can’t give you negative scoops

you are here 4  117

exercise solution

Take a moment and READ the
code above out loud to yourself in
English (or your chosen language).
After you’ve gotten comfortable
speaking the code, jot it down
using your pencil.

if the number of scoops is zero, then print ‘You didn’t want
any ice cream? We have lots of flavors.’
Else if the number of scoops is one, then print ‘A single scoop
for you, coming up.’
Else if the number of scoops is two, then print ‘Oh, two scoops
for you!’
Else if the number of scoops is greater than or equal to three,
then print ‘Wow, that’s a lot of scoops!’
Else, print “I’m sorry I can’t give you negative scoops.”

Your turn again. Based on our plan on the previous page, finish the code fragment below. Your code
should determine if there is a tie, and if so, set the winner variable accordingly. After you’ve done this
we’ll get this code into the rock.py file.

user_choice
if computer_choice
___________ == ___________:
winner = _______
‘Tie’

Get out your pencil and put some more Boolean expressions through their paces. For
each expression below, compute its value and write in your answer.
age > 5 and age < 10
age > 5 or age == 3

not age > 5

118  Chapter 3

When age is 6, what does this evaluate to? True
________

False
When age is 11, what does this evaluate to? ________
When age is 5, what does this evaluate to? ________
False
When age is 6, what does this evaluate to? True
________

False
When age is 2, what does this evaluate to? ________
When age is 3, what does this evaluate to? True
________

False
When age is 6, what does this evaluate to? ________
True
When age is 2, what does this evaluate to? ________

booleans, decisions, and loops

Code Magnets Solution

We'd already worked out the code for the game logic on the fridge with code magnets, but someone
came along and knocked most of it on the floor. Can you reconstruct the code snippets so that we can
determine the winner? Notice, there may be some extra code magnets, so you may not use all of them.

if

computer_choice == user_choice
winner

elif

=

:

'Tie'

computer_choice == 'paper'

user_choice == 'rock'

and

:

winner = 'Computer'
computer_choice == 'rock'

elif

user_choice == 'scissors'

and

:

winner = 'Computer'
computer_choice == 'scissors'

elif

and

user_choice == 'paper'

:

winner = 'Computer'
else

:

winner = 'User'

you are here 4  119

exercise solution

In the IDLE editor add
comments to your rock.py
file. Feel free to use our
comments on the previous
page as a starter, but
make the comments your
own; that is, make sure
they are meaningful to you.

#
#
#
#
#

ROCK, PAPER, SCISSORS
Passed down from the ancient Chinese Han dynasty, the game
shoushiling is now better known as Rock, Paper, Scissors.
This code implements a version of the game that is you against
the computer.

# Here we're doing some setup by importing the random module
# and setting up the winner variable.
import random
winner = ''
# The computer randomly chooses rock, paper, scissors by
# generating a random number from 0 to 2 and then mapping that
# to a corresponding string.
random_choice = random.randint(0,2)

Here’s our attempt
to document Rock,
Paper, Scissors.

if random_choice == 0:
computer_choice = 'rock'
elif random_choice == 1:
computer_choice = 'paper'
else:
computer_choice = 'scissors'
# Get the user's choice with a simple input statement.
user_choice = input('rock, paper or scissors? ')
# Here's our game logic, which checks to see if the computer wins
# (or not), and makes the appropriate change to the winner variable.
if computer_choice == user_choice:
winner = 'Tie'
elif computer_choice == 'paper' and user_choice == 'rock':
winner = 'Computer'
elif computer_choice == 'rock' and user_choice == 'scissors':
winner = 'Computer'
elif computer_choice == 'scissors' and user_choice == 'paper':
winner = 'Computer'
else:
winner = 'User'
# Here we announce the game was a tie, or the winner along
# with the computer's choice.
if winner == 'Tie':
print('We both chose', computer_choice + ', play again.')
else:
print(winner, 'won. The computer chose', computer_choice + '.')

120  Chapter 3

booleans, decisions, and loops

Convert the spoken English above into its Boolean expression equivalent. We’ve
done the first part for you.
user_choice != 'rock' and user_choice != 'paper' and user_choice != 'scissors'

While you were turning the page we went ahead and wrote the code to reprompt the user,
only our approach was flawed. Can you take a look and see what you think the flaw might
be? You’ll find the code, some notes, and a sample run of the code below.

user_choice = input('rock, paper or scissors? ')

if (user_choice != 'rock' and

user_choice != 'paper' and

First we get the
user’s choice.

Then we test to see
if the input was valid.

user_choice != 'scissors'):

user_choice = input('rock, paper or scissors? ')

print('User chose', user_choice)

If not, then we get the
user’s choice again.
We thought we were on the
right track, but what do you
think we’re missing?

Python 3.6.0 Shell

rock, paper or scissors? rack
rock, paper or scissors? papper
User chose papper

>>>

Ugh!

We're getting
reprompted.
But then we can
just enter invalid
input again.

ANALYSIS: On the good side, the code to detect invalid input is working great. However, the
attempt to reprompt the user, while it works, only works the first time. We really need a way
for this to work as many times as the user enters invalid input.

you are here 4  121

exercise solution

w it

Write a quick game. Here’s how it works: you prompt the player with
“What color am I thinking?” and you see how many guesses it takes the
player to guess it.
color = 'blue'
guess = ''
guesses = 0
while guess != color:
guess = input('What color am I thinking of? ')
guesses = guesses + 1
print('You got it! It took you', guesses, 'guesses')

extra credit
w it

Remember the color guessing game? It has a bug. Have you noticed if you guess the color
correctly on the first try, it prints: "You got it! It took you 1 guesses". Can you fix
the bug so that it prints “guess” for one guesses and “guesses” for multiple guesses?
color = 'blue'
guess = ''
guesses = 0
while guess != color:
guess = input('What color am I thinking of? ')
guesses = guesses + 1
if guesses == 1:
print('You got it! It took you 1 guess')
else:
print('You got it! It took you', guesses, 'guesses')

122  Chapter 3

booleans, decisions, and loops

1

F

2

E Q U A

3

B

4

R

C O M P
O

6

W H

Coding
Cross
Solution

I
10

L

13

T

E

R

8

S

I

H

P

I
7

9

11

14

S

X

S
C

E

R A

L

S

I

I
L

12

S

15

16

I

S

O

P

S

R

23

I

I

S

S O R S

T

S O R S

O
C

N

L

N D O M N U M B
G

N

G N M E

E

S

A

H

O

21

E

S O N

I

19

L

Y

H

P

18

T

U

L

E

I

5

B O O

F

A

N
Y

R

R

E

H A

A

L

P

T

E M P

17

T

Y

G
E

R S

O

20

R O C K

G

22

E

L

I

F

O
N D E

N

T

E

D

Across

Down

2. == is for. [EQUALITY]
4. >, <, and == are for. [COMPARISON]
6. Keeps executing as long as True. [WHILE]
7. Has two values. [BOOLEAN]
8. If its Boolean expresssion is True it executes
a code block. [IF]
10. Character for comments. [HASH]
11. = is for. [ASSIGNMENT]
13. Boolean is another one of these. [TYPE]
14. Cuts paper. [SCISSORS]
15. A string with nothing in it. [EMPTY]
19. Many programs need to generate these.
[RANDOMNUMBERS]
20. Paper wraps ______. [ROCK]
21. Rock blunts. [SCISSORS]

1. Can't afford. [FERARRI]
3. Foundation of if and while.
[BOOLEANEXPRESSION]
5. Ancient Chinese game. [SHOUSHILING]
9. True or _____. [FALSE]
12. Catch-all. [ELSE]
16. How you include the random module.
[IMPORT]
17. Boole's first name. [GEORGE]
18. We often say repeating code is in one of
these. [LOOPS]

you are here 4  123

4

lists and iteration

Providing Some Structure
You come here often?
Only when pushed.

There’s more to data types than numbers, strings, and
Booleans. So far you’ve been writing Python code using primitive types—
those floats, integers, strings, and of course Booleans—with values like 3.14,
42, "hey, it’s my turn", and True. And you can do a lot with primitives,
but at some point you’ll want to write code that deals with lots of data—say, all
the items in a shopping cart, the names of all the notable stars, or an entire
product catalog. For that we need a little more ummph. In this chapter we’re
going to look at a new type, called a list, which can hold a collection of values.
With lists, you’ll be able to provide some structure for your data, rather than
just having a zillion variables floating around your code holding values. You’re
also going to learn how to treat all those values as a whole as well as how to
iterate over each item in a list using that for loop we mentioned in the last
chapter. After this chapter, your ability to deal with data is going to grow and
expand.

this is a new chapter   125

the bubbles-r-us company

Can you help Bubbles-R-Us?
Check out the Bubbles-R-Us company. Their tireless research
makes sure bubble wands and machines everywhere blow the best
bubbles. Today they’re testing the “bubble factor” of several different
formulations of their new bubble solution—that is, they’re testing how
many bubbles can be blown with a given solution. Here’s their data:

of
Each bubble solution was tested for the number
bubbles it can create.

Each test tube
is labeled 0 to
9 and contains a
different bubble
solution.

0

1

2

3

60

50

60

58

4

54

5

6

7

8

9

54

58

50

52

54

Of course you want to get all this data into Python so you can write
code to help analyze it. But that’s a lot of values. How are you going
to construct your code to handle all these values?
126  Chapter 4

And here’s the
bubble factor
score for each
solution.

lists and iteration

Many programming
languages call their
ordered data type an
array instead of a list.

How to represent multiple values in Python
You know how to represent single values like strings, numbers, and Booleans with Python,
but how do you represent multiple values, like all the bubble factor scores from the 10
bubble solutions? To do that we use Python lists. A list is a Python data type that can hold
many values. Here’s a Python list that holds all the bubble factor scores:

scores = [60, 50, 60, 58, 54, 54, 58, 50, 52, 54]

We placed numbers in
this list, but you can put
whatever values you want.

Here’s all 10 values, grouped together into a
list, and assigned to the scores variable.
Once you have your data in a list, you can access the individual scores when you need
to. Each individual score, or item, has an index. Computer scientists like to number things
starting at zero, so the first item has an index of 0. You can retrieve any item in the list
using its index, like this:

To access an item in the list we use
this syntax: the variable name of the
list followed by the index of the item,
surrounded by square brackets.

The index is zero-based, so the first
item in the list is at index 0, the
second items is index 1, and so on.

score = scores[3]

We also often call
types like lists a data
structure because
they provide a way to
organize values or data.

Remember we said
computer scientists
like to start things
from zero.

print('Solution #3 produced', score, 'bubbles.')

My
bubble solution #3
is definitely going to be
the best.

Python 3.6.0 Shell

bles.
Solution #3 produced 58 bub
>>>

One of the
Bubbles-R-Us
bubbleologists
you are here 4  127

understanding lists

How lists work
It looks like we might have some interesting
work to do for Bubbles-R-Us, but before we
start, let’s make sure we’ve got lists down, and
rather than using bubble factor scores, let’s put another kind
of value in our lists: strings, or even better, smoothie flavors! After we understand
lists a bit better, we’ll get right back to helping Bubbles-R-Us.
So, once you have a bunch of values you want to group together, you can create a list that holds
them, and then access those values in the list whenever you need them. Most often you’ll use lists
when you want to group together similar things, like bubble factor scores, ice cream flavors, daytime
temperatures, or even the answers to a set of true/false questions. Let’s look again at how to create
a list, paying a little more attention to the syntax this time.

How to create a list
Let’s say you wanted to create a list that holds the name of a bunch
of smoothies. Here’s how you’d do that:

Notice that each
in the list is separaitem
ted
by a comma.

smoothies = ['coconut', 'strawberry', 'banana', 'pineapple', 'acai berry']

Let’s assign the list
to a variable named
smoothies.

To begin the list, use the And then we have
each item of the
[ character, otherwise
list...
known as an opening
square bracket.

t
...and we end theualisre
sq
with a closing
bracket.

As we already said, every item in a list resides at a location, or index. With the smoothies list, the
first item, “coconut,” is at index 0; the second, “strawberry,” is at index 1; and so on. Here’s a
conceptual look at how lists are stored:

The list collects
all these values
together.

'
ry
'
r
ut
be
a'
on
aw
an
c
r
n
o
t
a
'c
's
'b
0

smoothies
smoothies

The list is assigned to
a variable.

128  Chapter 4

1

2

Each value has
an index number,
starting at 0.

'
e'
ry
l
r
pp
be
ea
i
n
a
i
c
'p
'a
3

4

Every index in the list
holds a value.

lists and iteration

How to access a list item
Each item in the list has an index, and that’s your key to both accessing and changing the
values in a list. We’ve already seen how to access an item by starting with the list’s variable
name and then adding on an index, surrounded by square brackets. You can use that
notation anywhere you’d use a variable:

This evaluates to the value of the smoothies list at
index 2 (which is ‘banana’), which is then assigned
to the variable favorite.

favorite = smoothies[2]

To get an element from a list,
you need both the variable name
of the list, and the index of
the value you want.
favorite is assigned the value in smoothies[2].
a'
an
n
a
'b

'
ry
t'
er
u
b
a'
on
aw
an
oc
tr
an
c
s
b
'
'
'

favorite

0

1

2

'
e'
ry
pl
er
p
b
ea
i
in
ca
'p
'a
3

4

And remember, because the index starts at 0,
smoothies[2] is the third item in the list.

smoothies

Updating a value in the list
You can also change the value of an item in a list using its index:
smoothies[3] = 'tropical'

This sets the value of the item at index 3
(previously ‘pineapple’) to a new value, ‘tropical’.

So, after this line of code, the smoothie list will look like this:

'
y'
l'
'
ry
rr
a
t
r
'
e
b
ic
nu
be
na
op
co
aw
na
ai
r
o
t
a
c
't
'c
's
'b
'a

smoothies

0

1

2

3

4

The value of the item at
index 3 has been changed.
you are here 4  129

an exercise with lists

It’s a good time to get some practice with lists. Pretend you’re the
Python interpreter and trace though this code, figuring out what
the final output is. After you’ve completed the exercise, we’ll take
our list knowledge a little further.

eighties = ['', 'duran duran', 'B-52s', 'muse']

newwave = ['flock of seagulls', 'postal service']

remember = eighties[1]
eighties[1] = 'culture club'
band = newwave[0]
eighties[3] = band
eighties[0] = eighties[2]
eighties[2] = remember
print(eighties)

130  Chapter 4

lists and iteration

How big is that list, anyway?
Say someone hands you a nice big list with important data in it. You know how to
get what’s in the list, but you have no idea exactly how big it is (in other words, how
many items it has). Luckily, Python provides a built-in function to tell you, called
len. Here’s how you use the len function:

Use the len function to tell you
the number of items currently
in the list.

Just pass the len function
the list you'd like to know
the length of.

length = len(smoothies)

After this statement
is executed, the length
variable will hold the
number of items in the list,
in this case 5.

The length is 5 because there are five items.
'
ry
t'
er
u
b
a'
on
aw
an
c
r
n
o
t
a
'c
's
'b
0

flavors

smoothies

1

2

'
ry
l'
er
ca
b
i
p
i
ro
ca
't
'a
3

4

the
Notice that because we start numberiofngthe
list
th
leng
the
indices of the list at 0,
x.
inde
last
the
n
tha
ter
will always be one grea
This important fact
crops up in a lot of
other programming
languages as well.

Now that you know how to get the length of a list, how can you use the length’s value to obtain the
last item of a list?

you are here 4  131

how to access the last item

Accessing the last item in the list
Accessing the last item of a list is something you’ll do often when
coding. Say you’ve got a list that holds the most recent scores of a
sports game and you need to display the latest score. Or say you have
a list of current wind speeds of an approaching hurricane, and you
need to report the lastest speeds. You get the point: lists often have data
arranged with the latest, and often most important, values at the end
(that is, at the largest index), so accessing the last item of the list is a
common task.
The conventional way to do this, across many programming languages,
is to use the length of the list as an index. But remember, lists are
indexed starting at zero, so the index of the last item is actually one less
than the length of the list. To get the last item of our smoothies list, we
do this:
length = len(smoothies)

last = smoothies[length-1]
print(last)

This is a common technique
in most languages: figure out
the length of the list and
then subtract one to get
the index of the last item.

Python 3.6.0 Shell
acai berry
>>>

Python makes this even easier
Finding the last item of a list is such a common task that Python actually
provides an easier way to do it. Here’s how it works: you can use a
negative index, starting at –1, to specify the items in a list in reverse
order. So an index of –1 is the last item in the list, an index of –2 is the
second to last, and so on.

flavors
smoothies

132  Chapter 4

Python supports negative indices as
offsets from the END of the list. So, -1
is the last item, -2 is the item before
that, -3 is the third to the last item,
and so on. Note that a lot of languages
don't offer the convenient negative indices
syntax like Python does.

y'
'
rr
t
e
u
b
a'
on
aw
an
c
r
n
o
t
a
'c
's
'b

y'
'
rr
al
e
c
b
pi
ai
ro
c
t
'
'a

-5

-2

-4

-3

-1

Here’s what negative
indices look like,
which start from
the end of the list.

lists and iteration

Using Python’s negative indices
Let’s give Python’s negative indices a try. Let’s say we want to take the
last three smoothies on our list and print them:

With an index of -1 we
get the last item.
last = smoothies[-1]
second_last = smoothies[-2]
And -2 gets us the
third_last = smoothies[-3]
second to the last item.
print(last)
Likewise -3 gets us the
print(second_last)
third to the last item.
print(third_last)
Now let’s print them.

Python 3.6.0 Shell
acai berry
tropical
banana
>>>

The Thing-A-Ma-Jig

The Thing-A-Ma-Jig is quite a contraption—it clanks and clunks and even thunks,
but what it really does, well, you’ve got us stumped. Coders claim they know how it
works. Can you uncrack the code and find its quirks?
characters = ['t', 'a', 'c', 'o']
output = ''
length = len(characters)
i = 0
while (i < length):
output = output + characters[i]
i = i + 1

View this as a character-building
exercise—spend some real time on
this and make your brain work; it
will thank you later.
When you think you know what this code
does, write your answer here and check
the solution at the end of the chapter.

length = length * -1
i = -2
while (i >= length):
output = output + characters[i]
i = i - 1
print(output)

Try these as an alternative for the characters list above:
characters = [‘a’, ‘m’, ‘a’, ‘n’, ‘a’, ‘p’, ‘l’, ‘a’, ‘n’, ‘a’, ‘c’]
or
If you need hints, look at
characters = [‘w’, ‘a’, ‘s’, ‘i’, ‘t’, ‘a’, ‘r’]
the code comments at the
end of the chapter.
you are here 4  133

more about lists

Q:

Does the order of items in a list
matter?

A:

The list is an ordered data type.So,
most of the time, it matters, but not always.
In the Bubbles-R-Us scores list, the ordering
matters a lot, because the index of the score
in the list tells us which bubble solution got
that score—bubble solution 0 got score 60,
and that score is stored at index 0. If we
mixed up the scores in the list, then we’d
ruin the experiment! However, in other cases,
the order may not matter. For instance, if
you’re using a list just to keep track of
grocery items you need to pick up, the order
probably doesn’t matter much. So it really
depends on how you’re using the list. You’ll
probably find that ordering matters more
often than not when you use a list. Python
also has other data types—for instance,
dictionaries and sets—that are unordered.
More on those later in the book.

Q:

How many things can you put into
a list?

A:

Theoretically, as many as you
want. Practically, however, the number is
limited by things like the memory on your
computer. Each list item takes up a little
bit of space in memory and if you keep
adding items to a list, eventually you’ll run

out of memory. However, depending on the
kind of items you’re putting in your list, the
maximum number of items you can put into
a list is probably in the many thousands or
hundreds of thousands. Once you get into
the millions there are other solutions (like
databases) that are probably going to be
more appropriate.

Q:

Can you have a list without any
elements?

A:

Remember when we talked about
empty strings? Yes, you can have empty lists
too. In fact, you’ll see an example of using
an empty list in this chapter. To create an
empty list, just write:
empty_list = []

If you start with an empty list, you can add
things to it later. We’ll see how shortly.

Q:

Can values in a list have different
types, or do they all have to be the same?

A:

There is no requirement in Python that
all the values in a list be of the same type.
We call lists with items of different types
heterogeneous lists. Here’s one:
heterogenous = ['blue', True, 13.5]

Q:

What happens if I try to access an
item in a list that doesn’t exist?

A:

You mean like you have a list of 10
items and you try to access item at index
99? If you do that you’ll get a runtime error,
like this:
IndexError: list index out of

Q:

range

A:

A:

So far we’ve seen strings and
numbers in a list; can you put other
things in lists too?
You can; in fact, you can put values
from any Python type (including ones you
haven’t seen yet) into a list.

Or even another list!

Q:

Okay, well, can I assign a new value
to a list index that doesn’t exist?

No, you can reassign an item to a new
value, but you can’t assign a value to an item
that doesn’t exist—if you do you’ll get a runtime
“out of bounds” error. Note that some languages
do allow this, but not Python. In Python we
have to first add a new item to the list instead.

The smoothie flavors below were added to the list in the order
of their creation. Finish the code to determine the most recent
smoothie flavor created.
smoothies = ['coconut', 'strawberry', 'banana', 'pineapple', 'acai berry']
most_recent = ________________

recent = smoothies[most_recent]

134  Chapter 4

There are two ways to do this;
one uses len, the other doesn’t.
Can you get them both?

lists and iteration

Meanwhile, back at Bubbles-R-Us...

Hey, glad you guys
are here. We just got a lot of
new bubble tests run. Check out all
the new bubble scores! I really need
some help understanding this data. I’d
love for you to code up what I
sketched below.

scores = [60, 50, 60, 58, 54, 54,

58, 50, 52, 54, 48, 69,
34, 55, 51, 52, 44, 51,
69, 64, 66, 55, 52, 61,
46, 31, 57, 52, 44, 18,
41, 53, 55, 61, 51, 44]

The Bubbles-R-Us CEO

Bubbles-R-Us

What we need to build

New bubble scores

le to make quick decisioneds?
ab
be
to
ort
rep
is
th
d
nee
lly
Hey, I rea
produce! Can you get this cod
about which bubble solution to
- Bubbles-R-Us CEO

60
Bubble solution #0 score:
50
Bubble solution #1 score:
60
Bubble solution #2 score:

rest of scores here...
Bubbles tests: 36
Highest bubble score: 69
score: [11, 18]
Solutions with highest

you are here 4  135

understanding the report

Let’s take a closer look at what the CEO is looking for:

Bubbles-R-Us

We need to start
by listing all the
solutions and their
corresponding scores.
Then we need to print
the total number of
bubble scores…

…followed by the
highest score and
each solution that
has that score.

Hey, I really need this report to be able to make quick decisions
about which bubble solution to produce! Can you get this coded?
- Bubbles-R-Us CEO

Bubble solution #0 score: 60

Bubble solution #1 score: 50

Bubble solution #2 score: 60

rest of scores here...
Bubbles tests: 36

Highest bubble score: 69

Solutions with highest score: [11, 18]

Time once again to put your pseudocoding skills to use! Go ahead and write some
pseudocode that will create the bubble score report. Take each item in the report
separately and think of how you’d break it down and generate the right output. Put any
notes and your pseudocode here.

far
Do your best and take thiswoasrk
’ll
we
en
th
as you can, and
through the entire bubble score
report together.
136  Chapter 4

lists and iteration

Cubicle conversation
Let’s take a look at the
CEO’s mockup and see how
we can tackle coding it...

Judy: The first thing we need to do is display every score
along with its solution number.
Joe: And the solution number is just the index of the score
in the list, right?
Judy: Oh, yeah, that’s totally right.
Frank: Slow down a sec. So we need to take each score,
print its index, which is the bubble solution number, and
then print the corresponding score.

Frank

Judy

Joe

Judy: You’ve got it, and the score is just the corresponding
item in the list.
Joe: So, for bubble solution #10, its score is just

scores[10].

Judy: Right.
Frank: Okay, but there are a lot of scores. How do we write code to output all of them?
Judy: Iteration, my friend.
Frank: Oh, you mean like a while loop?
Judy: Right, we loop through all the values from zero to the length...oh, I mean the length
minus one, of course.
Joe: This is starting to sound very doable. Let’s write some code; I think we know what we’re
doing.
Judy: That works for me! Let’s do it, and then we’ll come back to the rest of the report.

you are here 4  137

iterating over a list

How to iterate over a list
Your goal is to produce some output that looks like this:
Bubble
Bubble
Bubble
.
.
.
Bubble

solution #0 score: 60
solution #1 score: 50
solution #2 score: 60

solution #35 score: 44

Scores 3 through 34 will be
here...we’re saving some trees
(or electrons depending on which
version of the book you have).

0

1

2

3

60

50

60

58

4

54

5

6

7

8

9

54

58

50

52

54

We’ll do that by outputting the score at index 0, and then we’ll do the
same for index 1, 2, 3, and so on, until we reach the last index in the list.
You already know how to use a while loop; let’s see how we can use that
to output all the scores:
we’ll

show you a
And then
sec...
a
in
way
better

scores = [60, 50, 60, 58, 54, 54, 58, 50, 52, 54, 48, 69,

34, 55, 51, 52, 44, 51, 69, 64, 66, 55, 52, 61,
46, 31, 57, 52, 44, 18, 41, 53, 55, 61, 51, 44]

Create a variable to keep track of the
current index, which we'll start at 0.
i = 0
Get the length of the scores list.
length = len(scores)
Loop over the items while our index is
less than the length of the list.
while i < length:

Note we didn’t need
length-1 here because
we’re using < (less than
length).

print('Bubble solution #', i, 'score:', scores[i])
i = i + 1

And finally, increment the index i
by one before looping again.

138  Chapter 4

Here we create the report output
by using the variable i to represent
our solution #. The variable i is
also being used as the index into
the scores list.

lists and iteration

A quick Test Drive
Here’s what we got; not bad!

It’s about time we get some real code written. Go ahead
and get the code on the previous page into a file called
bubbles.py before running it.

Python 3.6.0 Shell

Bubble solution # 0 score: 60
Bubble solution # 1 score: 50
Bubble solution # 2 score: 60
...

Bubble solution # 34 score: 51
Bubble solution # 35 score: 44
>>>

Hmm, we have
an extra space
here.

Just one small nit:
did you notice how
the output has an
extra space after
the hash character?
The CEO’s version
doesn’t have that.

Fixing the output glitch
Let’s look at the print statement to identify where the extra space is
coming from:

Bubbles-R-Us

Bubble solution #0 score: 60
Bubble solution #1 score: 50
Bubble solution #2 score: 60

As you know, when we provide multiple values
to print, separated by commas, by default
behavior is to add a space between each one.

print('Bubble solution #', i, 'score:', scores[i])
To fix this we could just do something like:

bubble_string = 'Bubble solution #' + i
print(bubble_string, 'score:', scores[i])

Concatenate the string
“Bubble solution#” to i
first, then hand it to
the print function.

It looks like by just concatenating "Bubble solution #" to the i variable we can easily
get rid of the extra space, but this isn’t going to work. Can you spot where we went
wrong?

you are here 4  139

a quick fix

Really fixing the output glitch
Did you figure out where we went wrong? Well, you can’t concatenate a string to an
integer. D’oh! But how do we change an integer into a string? Well, remember when we
did the opposite? We changed a string into an integer using the int function. As it turns
out, there is also a str function that does the opposite: give it an integer, and it will give
you back a string representation of that integer.
Given that, we can rework our code like this:

All we need to do is
pass the integer i to
the str function and it
will convert it into a
string representation.

bubble_string = 'Bubble solution #' + str(i)
print(bubble_string, 'score:', scores[i])
Let’s get that into our code, only we’ll do it without the extra bubble_string variable.
Instead, we’ll make our code more concise and add the call to str right in with the print
arguments. Check out the Test Drive below for the changes.

A quick fix Test Drive
Let’s make a quick update to the code and then you’ll have
the CEO’s spec implement, so far, just like he wanted it.

scores = [60, 50, 60, 58, 54, 54, 58, 50, 52, 54, 48, 69,
34, 55, 51, 52, 44, 51, 69, 64, 66, 55, 52, 61,
46, 31, 57, 52, 44, 18, 41, 53, 55, 61, 51, 44]
i = 0

length = len(scores)
while i < length:

Here we just concatenate “Bubble solution #” with i before it is
passed to print. And we make sure to use the str function so we
have a string representation of i.

print('Bubble solution #' + str(i), 'score:', scores[i])
i = i + 1

Python 3.6.0 Shell
Bubble solution #0 score: 60
Bubble solution #1 score: 50

Much better

140  Chapter 4

Bubble solution #2 score: 60
...

Bubble solution #34 score: 51
Bubble solution #35 score: 44
>>>

lists and iteration

Code Magnets

Time for a quick exercise. We wrote some
code to see which smoothies have coconut
in them. We had all the code nicely laid out
on our fridge using fridge magnets, but the
magnets fell on the floor. It’s your job to put
them back together. Be careful; a few extra
magnets got mixed in. Check your answer at
the end of the chapter before you go on.

)
while i < len(has_coconut

:

:

:

i = i + 2

i = i + 1

i = 0
if has_coconut[i]

while i > len(has_coconut)
conut',
smoothies = ['co
'strawberry',
'banana',

'tropical',
'acai berry']

has_coconut = [True,

False,

False,
True,

False]

nut')
i],'contains coco
print(smoothies[

Here’s the output
we’re expecting.
Python 3.6.0 Shell

Rearrange the magnets here.

coconut contains coconut
tropical contains coconut
>>>

you are here 4  141

iterating with a for loop

The for loop, the preferred way to iterate over a list
So, you can use a while loop to iterate over your lists, but the preferred method is actually using a for
loop. Think of the for loop as the while loop’s cousin—the two basically do about the same thing,
except we typically use a while loop when we’re looping over some condition, and a for loop when
we’re iterating over a sequence of values (like a list). Let’s return to our smoothies to see how we loop, or
iterate, over a list with the for loop. After we’ve done that, we’ll nail down the Bubbles-R-Us code.

Notice the convention of
using a plural variable name
for the list and a singular
name for the variable.

Next we have a variable that is going
to be assigned to each item in our list
as we iterate through it.

A for loop
starts with the
keyword for.

Next we have
the in keyword.

Here's the smoothies list
we're going to iterate over.

colon
And as always with Python we have a
before the code block below.
for smoothie in smoothies:
The code block is
output = 'We serve ' + smoothie
executed once for
each item in the list.
print(output)

You can use the smoothie variable anywhere in the
block to access the value of the current item.

How the for loop works

Let’s execute the code above. The first time through the loop, the first item in the list smoothies is
assigned to the variable smoothie. After that the body of the for loop is executed.
'
ut
on
c
o
'c

The smoothie variable (not to be
confused with smoothies) is assigned
to the first item in the list.

Then we execute the
code block of the for
statement.

smoothie
'
y'
l'
'
ry
rr
ca
ut
er
a'
be
i
n
b
n
p
o
w
a
i
ro
oc
ta
an
ca
't
'c
's
'b
'a

smoothies

142  Chapter 4

0

1

2

3

4

Python 3.6.0 Shell

We serve coconut

lists and iteration

Next time through the loop, the next item, “strawberry,” in the list smoothies is assigned
to the variable smoothie. After that the code block is executed.
y'
rr
be
w
a
tr
's

Because smoothie now is set to
strawberry, we now get the output
“We serve strawberry”.

Next, ‘strawberry’ is then assigned
to the smoothies variable.

smoothie

Python 3.6.0 Shell

We serve coconut
We serve strawberry

y'
'
y'
rr
al
t'
rr
'
be
ic
nu
be
na
p
o
w
a
i
ro
oc
ta
an
ca
't
'c
's
'b
'a
0

1

2

3

4

smoothies

The third time through the loop, the next item, “banana,” in the list smoothies is assigned
to the variable smoothie. After that the code block of the for loop is executed.
'
na
na
a
'b

This time we
get banana.

smoothie

We serve coconut
We serve strawberry
We serve banana

'
'
y'
ry
al
t'
rr
'
er
c
u
e
a
b
pi
on
wb
an
i
ro
oc
ta
an
ca
't
'c
's
'b
'a
0

1

2

3

Python 3.6.0 Shell

4

smoothies

And by now you can see the pattern—the fourth time through the loop, the next item,
“tropical,” is assigned to the variable smoothie before we execute the code block.
'
al
ic
p
ro
't

Tropical, anyone?

smoothie
'
y'
l'
'
ry
rr
ca
ut
er
a'
be
i
n
b
n
p
o
w
a
i
ro
oc
ta
an
ca
't
'c
's
'b
'a
0

1

2

3

Python 3.6.0 Shell

We serve coconut
We serve strawberry
We serve banana
We serve tropical

4

smoothies

you are here 4  143

understanding the for loop

And as you can guess at this point, the fifth, or last time, through the loop, the next
item, “acai berry,” in the list smoothies is assigned to the variable smoothie. After
that the code bock of the for loop is executed for the last time.
i
ca
'a

y'
rr
be

Python 3.6.0 Shell

smoothie
'
l'
'
ry
ca
ut
er
a'
pi
on
wb
an
i
o
c
a
n
r
o
t
a
ca
't
'c
's
'b
'a
0

1

2

3

y'
rr
be

4

smoothies

I like the for
loop; that’s nice. But how
do we print the score numbers—it looks like we just
have the actual scores?

Judy: Oh, you’re saying when we used a while loop we had the
counter i, which we used for the score number and as an index
to get the scores.
Frank: Exactly, and when we’re using a for statement, we just
seem to have the item of the list. Where’s the index?
Judy: Uh, good question.
Joe (shouting from across the room): Guys, I did some research,
there’s another way to use for. The way you’re talking about
is great for sequences when you don’t care about an index, but
you can use for with a range of indices to iterate through the
bubble solutions.
Frank: Say what?
Joe: It’s almost easier to show you...

144  Chapter 4

We’ve now iterated
through every item in
the smoothies list.

We serve coconut
We serve strawberry
We serve banana
We serve tropical
We serve acai berry

lists and iteration

How the for loop works on a range of numbers
There’s another kind of sequence the for loop works on: a range of numbers. In fact,
Python gives you a built-in function called range that you can use to generate different
sequences of numbers. After you’ve generated a sequence of numbers, you can use the
for loop to iterate through them.
Here’s how you generate a range from 0 to 4:

range(5)

Creates the sequence 0, 1, 2, 3, 4.

The range starts at
0 and gives you a
sequence of 5 numbers.
You can combine range with for like this:

range makes the sequence 0, 1, 2, 3, 4.
for i in range(5):

print('Iterating through', i)

The i variable is assigned to each item of the
sequence before the body is executed.

Python 3.6.0 Shell

Iterating through 0
Iterating through 1
Iterating through 2
Iterating through 3
Iterating through 4
>>>

So say you want to iterate through our smoothies and print the index of each. Here’s
how you can do that:

Create a range from zero to
the length of smoothies.

length = len(smoothies)

for i in range(length):

print('Smoothie #', i, smoothies[i])

Each time we iterate, we print
“Smoothie#", the index, and the
smoothie at that index.

Python 3.6.0 Shell

Smoothie # 0 coconut
Smoothie # 1 strawberry
Smoothie # 2 banana
Smoothie # 3 tropical
Smoothie # 4 acai
>>>

you are here 4  145

exploring ranges

Doing more with ranges
With a range you don’t have to create sequences from
zero to some number; you can create all kinds of ranges
of numbers. Here’s a few examples:

Q:

Try a starting and ending number

A:

Start of range

Up to end of range
(but not including
this number)

range(5, 10)

Creates a sequence
starting at 5 and
going up to 10, so:
5, 6, 7, 8, 9

Add a step size
You can add a “step size”
as well, which tells Python
to count by increments.
range(3, 10, 2)

Count backward

Creates a sequence starting
at 3 and going up to 10,
but counting by steps of 2,
so: 3, 5, 7, 9

We can even count backward by making the first argument
larger than the second, and using a negative step size.
Creates a sequence
starting at 10 and going
range(10, 0, -1)
down to 0, but by steps
of -1, so: 10, 9, 8, 7, 6, 5,
4, 3, 2, 1
Or start from negative numbers
You can start at a negative number too.
range(-10, 2)

146  Chapter 4

Creates a sequence starting at -10
counting to 2, so: -10, -9, -8, -7,
-6, -5, -4, -3, -2, -1, 0, 1

Does range(5) just create a list,
like [0, 1, 2, 3, 4]?

No, it doesn’t, although we can
easily see how you’d think that. The
reason it doesn’t is Python actually
creates something a lot more efficient
than a list. For now, though, it is fine to
think of it that way; just know you can’t
substitute range for a list in your code.
Oh, and if you ever want to use range to
create a list, you can do that like this:

list(range(5))

to create the list you mentioned in your
question.

Q:

You used a variable name
called i. That doesn’t seem very
good for readability. Why not index
or smoothie_index or something like
that?

A:

Good catch. You’re right, the
variable i may not be the most readable
variable name, but when a variable is
used as an index in a iteration, there is
a long history of using variables like i, j,
and k—so much so, it is almost blindly
followed by programmers and in fact it
might strike them as odd to use a longer
variable name. So, we encourage you,
for this exception, to use short variable
names, and before long it will feel like
second nature to you.

lists and iteration

We had our range function calls figured out, and then they got all mixed
up. Can you help us figure out who does what? Be careful, we’re not sure
if each range matches zero, one, or more sequences. We’ve already figured
one out, which is marked below:
range(8)

-3, -2, -1, 0

range(0, 8)

3, 6
range(0, 8, 1)

0, 1, 2, 3, 4, 5, 6, 7
range(0, 8, 2)

range(0, 7)

range(-3, 1)

range(3, 8)

range(3, 8, 3)

range(3, 8, 2)

0, 1, 2, 3, 4, 5, 6

0, 2, 4, 6
3, 4, 5, 6, 7

3, 5, 7
1, 2, 3, 4

you are here 4  147

implementing the report

I think we’ve got all the
pieces we need for the first
part of the report, so let’s put this
all together...

Putting it all together
Let’s now use our knowledge of ranges and the for loop to rework the
while loop we previously wrote to generate the bubble solution numbers
plus their scores.

Here’s our bubble scores list.

scores = [60, 50, 60, 58, 54, 54, 58, 50, 52, 54, 48, 69,
34, 55, 51, 52, 44, 51, 69, 64, 66, 55, 52, 61,
46, 31, 57, 52, 44, 18, 41, 53, 55, 61, 51, 44]

First, get length of
the scores list, as
before.
length = len(scores)
while i < length:

for i in range(length):

You can delete the while loop.
Then create a range from the scores
length, and iterate over those values from
zero to the length of scores minus one.

print('Bubble solution #' + str(i), 'score:', scores[i])

Then we create our output. Notice this is
exactly the same print statement we used
with the while loop—nothing changed!

148  Chapter 4

lists and iteration

Te st dri ve the bubble report
Type the new code in and save it in the file bubbles.py, and
then give it a test run. Check out the brilliant report you
just generated for the Bubbles-R-Us CEO.

Just what the CEO ordered
It's nice to see all the bubble
scores in a report, but it's
still hard to find the highest
scores. We need to work
on the rest of the report
requirements to make it a
little easier to find the winner.

Python 3.6.0 Shell
Bubble solution #0 score: 60
Bubble solution #1 score: 50
Bubble solution #2 score: 60
Bubble solution #3 score: 58
Bubble solution #4 score: 54
Bubble solution #5 score: 54
Bubble solution #6 score: 58
Bubble solution #7 score: 50
Bubble solution #8 score: 52
Bubble solution #9 score: 54
Bubble solution #10 score: 48
Bubble solution #11 score: 69
Bubble solution #12 score: 34
Bubble solution #13 score: 55
Bubble solution #14 score: 51
Bubble solution #15 score: 52
Bubble solution #16 score: 44
Bubble solution #17 score: 51
Bubble solution #18 score: 69
Bubble solution #19 score: 64
Bubble solution #20 score: 66
Bubble solution #21 score: 55
Bubble solution #22 score: 52
Bubble solution #23 score: 61
Bubble solution #24 score: 46
Bubble solution #25 score: 31
Bubble solution #26 score: 57
Bubble solution #27 score: 52
Bubble solution #28 score: 44
Bubble solution #29 score: 18
Bubble solution #30 score: 41
Bubble solution #31 score: 53
Bubble solution #32 score: 55
Bubble solution #33 score: 61
Bubble solution #34 score: 51
Bubble solution #35 score: 44

you are here 4  149

more with loops and lists

Another quick exercise. Remember the fridge
magnet code from a few pages back? Update
that code so that it uses a for loop instead of a
while loop. If you need a hint, revisit the way we
reworked the while loop for Bubbles-R-Us.

smoothies = ['coconut',

'strawberry',
'banana',

You don’t need to turn back a few pages
because we’ve reproduced the solution
for you here.

'tropical',

'acai berry']
has_coconut = [True,

False,

Your code goes here.

False,
True,

False]
i = 0
while i < len(has_coconut)
if has_coconut[i]

:

:

print(smoothies[i],
'contains coconut')
i = i + 1

150  Chapter 4

lists and iteration

Tonight’s talk: The WHILE and
FOR loop answer the question
“Who’s more important?”

The WHILE loop

The FOR loop

What, are you kidding me? Hello? I’m the general
looping construct in Python. I don’t need a
sequence or a range, as I can be used with any type
of conditional. Did anyone notice I was taught first
in this book?
I don’t appreciate that tone.
And that’s another thing: have you noticed that the
FOR loop has no sense of humor? I mean if we all
had to do skull-numbing iteration all day, I guess
we’d all be that way.
Cute. But have you noticed that 9 times out of 10,
coders use FOR loops?
Oh, I don’t think that could possibly be true.
Not to mention, doing iteration over, say, a list that
has a fixed number of items with a WHILE loop is
just a bad, clumsy practice.
This book just said that FOR and WHILE loops are
pretty much the same thing, so how could that be?
Ah, so you admit we’re more equal than you let on,
huh?
I’ll tell you why...
When you use a WHILE loop you have to
initialize your counter and increment it in separate
statements. If, after lots of code changes, you
accidentally moved or deleted one of these
statements, well, then things could get ugly. But with
a FOR loop, everything is packaged right in the
FOR statement for all to see and with no chance of
things getting changed or lost.

you are here 4  151

comparing while and for

The WHILE loop

The FOR loop

Well, isn’t that nice and neat of you. Hey, most of
the iteration I see doesn’t even include counters; it’s
stuff like:
while (input != ''):
try that with a FOR loop!
So that’s all you got? You’re only better when you’ve
got a condition to loop over?
Not only better, prettier.
Oh, I didn’t realize this was a beauty contest. I’d
argue people iterate over sequences way more than
they write loops over general conditionals.
Hey, I can iterate over a sequence too.
I think we’ve already covered that ground. Sure
you can, but it’s, well, it ain’t pretty. Don’t forget I’m
quite general too, I don’t just work on lists.
Like what?
There are lots of sequences in Python. We’ve seen
lists and ranges and strings, but there’s even more
you can iterate over, like files, and quite a few other
more advanced data types the readers haven’t even
looked at it in this book.
I’m sure I can work with them too.

Oh sure, you’re the tough guy. Next time you need
to iterate while a condition is True, don’t call me,
and then we’ll see how heavy duty you are.

Perhaps, but, again, wouldn’t be pretty. Face it,
when it comes to heavy-duty iteration, I’m designed
for it.

Likewise, don’t call me when you need to iterate
over a sequence!
152  Chapter 4

lists and iteration
We’ve got all
the bubble solution
scores displaying, so now we
just need to generate the
rest of the report.

Cubicle conversation
continued...

Judy: Right, and the first
thing we need to do is
determine the total number of
bubble tests. That’s easy; it’s
just the length of the scores list.

Bubbles-R-Us

Hey, I really need this
report to be
about which bubble
to make quick decisions
solution to producable
e! Can you get thi
s coded?
- Bubbles-R-Us CEO

Bubble so
lution #0
score: 60
Bubble so
lution #1
score: 50
Bubble so
lution #2
score: 60

Joe: Oh, right. We’ve got to
find the highest score too, and
then the solutions that have the
highest score.

Bubbles te
sts: 36
Highest bu
bble scor
e: 69
Solutions
with high
est score:

rest of scores

here...

[11, 18]

Judy: Yeah, that last one is going to be the toughest. Let’s work out
finding the highest score first.
Joe: Sounds like a good place to start.
Judy: To do that I think we just need to maintain a highest score variable that keeps
track as we interate through the list. Here, let me write some Python-like pseudocode:
DECLARE a variable high_score and set to 0.
FOR i in range(length)
		PRINT i and the bubble solution score[i]
		IF scores[i] > high_score:
high_score = scores[i];
PRINT high_score

score.
Add a variable to hold the high
Check each time through the loop to see if we have
a higher score, and if so that's our new high score.

If we have a new high score, then assign it to high_score.
After the loop we just display the high score.

Joe: Oh nice; you did it with just a few lines added to our existing code.
Judy: Each time through the list we look to see if the current score is greater than

high_score, and if so, that’s our new high score. Then, after the loop ends we just

display the high score.

you are here 4  153

taking the implementation further

Go ahead and implement the pseudocode on the previous page to find the
highest score by filling in the blanks in the code below. Once you’re done,
add the code to bubbles.py and give your code a test. Check the results in the
shell, and fill in the blanks in our shell below with the number of bubble tests
and the highest score. As always, check your answer at the end of the chapter
before you go on.
scores = [60, 50, 60, 58, 54, 54,

58, 50, 52, 54, 48, 69,
34, 55, 51, 52, 44, 51,
69, 64, 66, 55, 52, 61,
46, 31, 57, 52, 44, 18,
41, 53, 55, 61, 51, 44]

Fill in the blanks to complete the code here...

high_score = ______
length = len(scores)

for i in range(length):

print('Bubble solution #' + str(i), 'score:', scores[i])
if _____________ > high_score:
_____________ = scores[i]

print('Bubbles tests:',

____________)

print('Highest bubble score:',

_____________)

Python 3.6.0 Shell

...and then fill in the blanks showing the
output you get in the console.

154  Chapter 4

Bubble solution #0 score: 60
Bubble solution #1 score: 50
Bubble solution #2 score: 60
...
Bubble solution #34 score: 51
Bubble solution #35 score: 44
Bubbles tests: ________
Highest bubble score: _______

lists and iteration

Hey, you guys are almost
there! All you’ve got left is collecting
up all the solutions with the highest score
and printing them. Remember, there
might be more than one.

More than one? When we need to store
more than one thing, what do we use? A list, of
course. So, can we iterate through our existing
scores list looking for scores that only match
the highest score, and then add those to a new
list that we can later display in the report? You
bet we can, but to do that we’ll have to learn
how to create a brand new, empty list, and then
understand how to add new elements to it.

Remember here's
what we have left.

Bubbles-R-Us

Hey, I really need this report to be able to make quick decisions
about which bubble solution to produce! Can you get this coded?
- Bubbles-R-Us CEO

Bubble solution #0 score: 60

Bubble solution #1 score: 50

Bubble solution #2 score: 60

rest of scores here...
Bubbles tests: 36

Highest bubble score: 69

Solutions with highest score: [11, 18]

you are here 4  155

how to create and add to lists

Building your own list, from scratch
Before we take on finishing this code, let’s get a sense for how to create a
new list, and how to add items to it. You already know how to explicitly
create a list with values, like this:
menu = ['Pizza', 'Pasta', 'Soup', 'Salad']

But you can also omit the initial items and just create an empty list:
menu = []

A new list, all ready to go with no
items and a length of zero.

Once you’ve created an empty list you can add new items with append,
like this:
menu = []

menu.append('Burger')

As before, a new list item is created.
The string ‘Burger' is added to the list.

menu.append('Sushi')
print(menu)

And a second item is created that
holds the string ‘Sushi'.
Python 3.6.0 Shell

['Burger', 'Sushi']
>>>

You can also
create a new,
empty list in
Python by calling
list(), but more on
this later in the
book. Just tuck
that in the back
of your brain for
now.

You’ve seen how to create an empty list with two square backets,
but what do you think the code below does? Feel free to type the
code into the shell to experiment. Record your observations here.
mystery = ['secret'] * 5

mystery = 'secret' * 5

156  Chapter 4

Multiplication of a
number and a list? What
on earth does this do?
How is it different
from this?

lists and iteration

Doing even more with lists
There’s a lot more you can do with lists, like insert new items, delete items, add
lists together, and search for items in a list—here are a few examples to whet
your appetite.

Delete an item from a list
Need to get rid of an item in a list? Python provides a built-in
function called del to do just that. Here’s how it works:
del menu[0]

Delete the item at
index 0.

print(menu)

Python 3.6.0 Shell

The item at index 0 is
removed, leaving only one
item, ‘Sushi’, which is now
at index 0.

['Sushi']
>>>

After you’ve deleted an item from a list, all the items with great
er
indices move down one. So if you delete the item at index
2,
then
the item previously at index 3 takes its place. The item at index
4
becomes 3, and so on.

Add one list to another
Let’s say you have a list, and someone hands you another list and
you want to add all those items to your list. No worries, here’s
how you do that:

Extend our menu with the
items in another list.

Our current list, menu, is
extended with the items
that are in the supplied list.

What is the difference
between using append
and using extend on
lists?

menu.extend(['BBQ', 'Tacos'])
print(menu)

Python 3.6.0 Shell
['Sushi', 'BBQ', 'Tacos']
>>>

you are here 4  157

doing more with lists

There’s another way to combine lists as well—you can just add the
lists together using the + operator, like this:
menu = menu + ['BBQ', 'Tacos']

If we execute this
instead of our
previous use of the
extend function, we
get the same result.

Python 3.6.0 Shell

', 'Tacos']
['Sushi', 'BBQ
>>>

Note: extend extends an existing list. When you use + you get a
brand new list with the items of both lists in it.

Or insert items into your list
Let’s say you really need to add an item in the middle of your list.
Use the insert function to do that.

Here’s the index where we
want an item inserted.

And here’s
the item.

menu.insert(1, 'Stir Fry')

Python 3.6.0 Shell

['Sushi', 'Stir Fry', 'BB
Q', 'Tacos']
>>>

insert adds a new
item at the supplied
index, in this case 1.

print(menu)

As we said, we’ll be seeing even more list operations as the book progresses, but
these are some good operations to get you started.

Q:
A:
Q:

What happens if I insert an item after an index that
doesn’t exist, like menu.insert(100, 'French Fries')?
If you try to insert an item beyond the end of your list, it will
simply add the item in the last position in your list.
What does the syntax mylist.append(value) actually
mean? It looks similar to the random.randint(0,2) syntax we
used in the last chapter.

A:

Yes, they are related; both are an example of something we’ll
get to later in the book: the use of functions and objects (actually
we’ll make our use of terminology even more precise at that stage
of the book). Now that all won’t mean a lot to you right now, but
we’re going to see how data types, like lists, can provide their own

158  Chapter 4

special behavior to do things like append items. So, mylist.append
is using the behavior append, which is provided by the list. For now,
go with the syntax, and down the road you’ll better understand the
true meaning behind it as we explore objects and functions.

Q:

Well, why do we have menu.append and menu.insert, but
del menu[0]? Why isn’t it menu.delete(0) or something similar?
I thought Python was consistent?

A:

It’s a very good question. It turns out the designers of Python
thought common operations, like len and del, deserved a bit of
special treatment. They also thought that, for example, len(menu)
was more readable than menu.length(). The reasoning behind this
has been debated at great length, but that’s the way it is in Python.
And, as in the last question, you’re asking all the right things, and
the madness behind the method will be clearer once we get to
talking about functions and objects.

lists and iteration

Now that we know how to
add items to a list, we can finish up this
report. We can just create the list of the
solutions with the highest score as we iterate
through the scores list to find the highest
bubble score, right?

Judy: Yes, we’ll start with an empty list to hold
the solutions with the highest scores, and add each
solution that has that high score one at a time to it
as we iterate through the scores list.
Frank: Great, let’s get started.
Judy: But hold on a second…I think we might
need another loop.
Frank: We do? It seems like there should be a way
to do that in our existing loop.
Judy: Actually, I’m sure we do. Here’s why: we
have to know what the highest score is before we can
find all the solutions that have that highest score.
So we need two loops: one to find the highest score,
which we’ve already written, and then a second one
to find all the solutions that have that score.
Frank: Oh, I see. And in the second loop, we’ll
compare each score to the highest score, and if it
matches, we’ll add the index of the bubble solution
score to the new list we’re creating for the solutions
with the highest scores.
Judy: Exactly! Let’s do it.

you are here 4  159

finding the high scores

Can you help write the loop to find all the scores that match the high score?
You’ll find all the code up until now listed below. Give it a real shot below before
you check the solution.

scores = [60, 50, 60, 58, 54, 54,

58, 50, 52, 54, 48, 69,

Here's all the
code so far.

34, 55, 51, 52, 44, 51,
69, 64, 66, 55, 52, 61,
46, 31, 57, 52, 44, 18,
41, 53, 55, 61, 51, 44]
high_score = 0
length = len(scores)

for i in range(length):

print('Bubble solution #' + str(i), 'score:', scores[i])
if scores[i] > high_score:
high_score = scores[i]

print('Bubbles tests:',

length

print('Highest bubble score:',
best_solutions = []

________________________

_________________________
_________________________

Remember, the variable
high_score has the highest
score in it; you can use that
in your code.

160  Chapter 4

high_score)

Here’s the new list you'll use to store the bubble
solutions with the highest score.

Write your code here, using
more lines if necessary.

lists and iteration

Test drive the final report
Go ahead and add your code to generate the bubble solutions with the highest score to your
code in bubbles.py and run another test drive. All of our code is shown below:
scores = [60,
58,
34,
69,
46,
41,

50,
50,
55,
64,
31,
53,

60,
52,
51,
66,
57,
55,

58,
54,
52,
55,
52,
61,

54,
48,
44,
52,
44,
51,

54,
69,
51,
61,
18,
44]

high_score = 0
length = len(scores)
for i in range(length):
print('Bubble solution #' + str(i), 'score:', scores[i])
if scores[i] > high_score:
high_score = scores[i]
print('Bubbles tests:', length)
print('Highest bubble score:', high_score)
best_solutions = []
for i in range(length):
if high_score == scores[i]:
best_solutions.append(i)
print('Solutions with the highest score:', best_solutions)

And the winners are...
Bubble solutions #11 and #18 both have a high score of 69, so
they are the best bubble solutions in this batch of test solutions!

If this code looks new to
you, then you may not have
studied the solution to the
last Sharpen your pencil
exercise. You’ll want to do
that now.

Python 3.6.0 Shell
Bubble solution #0 score: 60
Bubble solution #1 score: 50
...
Bubble solution #34 score: 51
Bubbles tests: 36
Highest bubble score: 69
Solutions with the highest score: [11,18]

you are here 4  161

finding the most cost-effective solution

Great job! Just one more thing...
can you figure out the most cost-effective
bubble solution? With that final bit of data,
we’ll definitely take over the entire bubble
solution market. Here’s a list with the cost of
each solution you can use to figure it out.

a
Here’s the new cost list. Notice that it has
in
ions
solut
g
ndin
cost for each of the correspo
the scores list.
costs = [.25, .27, .25, .25, .25, .25,
.33, .31, .25, .29, .27, .22,
.31, .25, .25, .33, .21, .25,
.25, .25, .28, .25, .24, .22,
.20, .25, .30, .25, .24, .25,
.25, .25, .27, .25, .26, .29]

So, what’s the job here? It’s to take the leading bubble solutions—that is, the ones
with the highest bubble scores—and choose the lowest-cost one. Now, luckily, we’ve
been given a costs list that mirrors the scores list. That is, the bubble solution
score at index 0 in the scores list has the cost at index 0 in the costs list (.25), the
bubble solution at index 1 in the scores list has a cost at index 1 in the costs list
(.27), and so on. So, for any score you’ll find its cost in the costs list at the same index.
Sometimes we call these parallel lists:
Scores and costs

are parallel lists because
e is a corresponding
ther
e
scor
for each
x.
inde
same
cost at the

costs = [.25, .27, .25, .25, .25, .25, .33, .31, .25, .29, .27, .22, ..., .29]

And likewise for the other cost
and score values in the lists.

The cost at 0 is the cost of
the bubble solution at 0...
scores = [60,

162  Chapter 4

50,

60,

58,

54,

54,

58,

50,

52,

54,

48,

69,

..., 44]

lists and iteration

This seems a little tricky. How do
we determine not only the scores
that are highest, but then pick the one
with the lowest cost?

Judy: Well, we know the highest score already.
Frank: Right, but how do we use that? And we have these two lists, but
how do we get those to work together?
Judy: I’m pretty sure either of us could write a simple for loop that
goes through the scores list again and picks up the items that match
the highest score.
Frank: Yeah, I could do that. But then what?
Judy: Anytime we hit a score that matches the highest score, we need
to see if its cost is the lowest we’ve seen.
Frank: Oh, I see, so we’ll have a variable that keeps track of the index of the “lowest cost
high score.” Wow, that’s a mouthful.
Judy: Exactly. And once we get through the entire list, whatever index is in that variable
is the index of the item that not only matches the highest score, but has the lowest cost as
well.
Frank: What if two items match in cost?
Judy: Hmm, we have to decide how to handle that. I’d say, whatever one we see first is the
winner. Of course we could do something more complex, but let’s stick with that unless
the CEO says differently.
Frank: This is complicated enough I think I want to sketch out some pseudocode before
writing anything.
Judy: I agree; whenever you are managing indices of multiple lists things can get tricky.
Let’s do that; in the long run I’m sure it will be faster to plan it first.
Frank: Okay, I’ll take a first stab at it…

you are here 4  163

understanding the pseudocode

I’m pretty sure I nailed the pseudocode.
Check it out below. Make sure you’ve got
it down, and then we’ll go ahead and translate
it into real code.

Create a variable to hold the cost of
the most cost-effective solution. We'll
make it larger than any item in costs
and also make it a float to match the
type of each item in the costs list.
DECLARE a variable cost and set to 100.0

Create a variable to hold the index
of the most cost-effective solution.
Iterate over every solution, and if
a solution has the highest score...

DECLARE a variable most_effective
FOR i in range(length):

		IF the bubble solution at scores[i] equals high_score AND bubble solution at costs[i] is less than cost:
SET the value of most_effective to the value of i
SET the value of cost to the cost of the bubble solution

At the end of the loop most_effective holds
the index of the solution with the highest score
and lowest cost. And the variable cost holds
the cost of that solution. Note, if there is a
tie between one or more solutions, this code will
always pick the solution it sees first in the list.

…and a lower cost than
previous solutions…

…then record the index
and cost of current
solution.

As Judy suggested in the cubicle conversation, if there are multiple high score solutions with the
same cost, this code favors the first solution it finds. But why is that? What aspect of the code
makes that happen? What if you wanted to favor the last solution you found instead; how would
you do that?

Answer: Because this code is always comparing the current low cost by using less than, so once a low cost is
established, it takes an even lower cost (not an equal cost) to pick a new winner. You could favor the last solution it
sees by changing the less than comparison to a less than or equal to, <=, comparison.

164  Chapter 4

lists and iteration

Testing the most cost-effective
solution
scores = [60,
58,
34,
69,
46,
41,

50,
50,
55,
64,
31,
53,

60,
52,
51,
66,
57,
55,

costs = [.25,
.33,
.31,
.25,
.20,
.25,

.27,
.31,
.25,
.25,
.25,
.25,

58,
54,
52,
55,
52,
61,

.25,
.25,
.25,
.28,
.30,
.27,

54,
48,
44,
52,
44,
51,

.25,
.29,
.33,
.25,
.25,
.25,

54,
69,
51,
61,
18,
44]

.25,
.27,
.21,
.24,
.24,
.26,

We should have everything coded below for the Bubbles-R-Us CEO.
Check out the code and see how it matches the pseudocode, and
then enter the new code into bubbles.py and give it another test run.
All the code is shown below. When you’ve got a winning solution,
turn to page to see if it matches ours.
.25,
.22,
.25,
.22,
.25,
.29]

Don’t forget the new
costs list.

high_score = 0
length = len(scores)
for i in range(length):
print('Bubble solution #' + str(i), 'score:', scores[i])
if scores[i] > high_score:
high_score = scores[i]
print('Bubbles tests:', length)
print('Highest bubble score:', high_score)
best_solutions = []
for i in range(length):
if high_score == scores[i]:
best_solutions.append(i)
print('Solutions with the highest score:', best_solutions)

d Frank'sode
e
t
la
s
n
a
r
We t -like pseudoc
Python to Python.
directly

cost = 100.0
most_effective = 0
for i in range(length):
if scores[i] == high_score and costs[i] < cost:
most_effective = i
cost = costs[i]
print('Solution', most_effective,
'is the most effective with a cost of', costs[most_effective])

We also added some
output to add the most
cost-effective solution
to the report.
you are here 4  165

the winning solution

The WINNER: SOLUTION #11
The last bit of code you wrote really helped determine the TRUE
winner; that is, the solution that produces the most bubbles at the
lowest cost. Congrats on taking a lot of data and crunching it down
to something Bubbles-R-Us can make real business decisions with.
Now, if you’re like us, you’re dying to know what is in Bubble
Solution #11. Look no further: the Bubbles-R-Us CEO said he’d be
delighted to give you the recipe after all your unpaid work.
So, you’ll find the recipe for Bubble Solution #11 below. Take some
time to let your brain process lists by making a batch, getting out,
and blowing some bubbles before you begin the next chapter. Oh,
but don’t forget the bullet points and the crossword before you go!

tion #11

Bubble Solu

ap
washing so
2/3 cup dish
er
1 gallon wat
e at the
ine (availabl
ons of glycer
po
es
bl
e)
ta
us
3
2 to
pply ho
chemical su
pharmacy or
gether in a

edients to
NS: Mix ingr
INSTRUCTIO
d have fun!
large bowl an

#11

166  Chapter 4

DO try this at HOME!

Python 3.6.0 Shell
...
Bubble solution #34 score: 51
Bubbles tests: 36
Highest bubble score: 69
Solutions with the highest score: [11,18]
Solution 11 is the most effective with a cost of 0.22
>>>

lists and iteration

I have to ask:
we already knew the highest
score solutions, they are in the
best_solutions list, so why did
we need to go through EVERY
score again?

You’re right: we didn't need to.
We could have found the lowest-cost solution from just the list in
best_solutions, because that list is the result of already figuring
out one or more solutions with the highest bubble scores. The only
reason we didn’t was to keep things simple on our first attempt.
Others might be asking, though: what’s the difference? Who cares?
It works! Well, it is all about the efficiency of the code. How much
work is your code doing? And, for a list as small as ours, there really
isn’t much of a difference; however, if you had a huge list of data
you’d want to avoid iterating over it multiple times if you had a more
efficient way. And we do.
To determine the lowest-cost solution (with the highest score), all
we need to do is consider the solutions in the best_solutions list.
Doing that is a little more complex, but not much.

This time we'll iterate over
the best_solutions list instead
of the scores lists
And we use each best_solutions
item as an index into the cost list.
cost = 100.0
So, in this code we're using
most_effective = 0
the values in best_solutions as
our indices.
for i in range(len(best_solutions)):
We're rewritten the
code to compute the
most cost-effective
solution.

COMPARE this code to
the previous version; can
you see the differences?
Think through how each
executes; can you see how
much less work this version
does to compute the most
cost-effective solution?
It's worth some time to
see the difference.

index = best_solutions[i]
if cost > costs[index]:
most_effective = index
cost = costs[index]

We examine the cost of each
solution in the best_solutions
list, and look for the lowest.

print('Solution', most_effective,
'is the most efftive with a cost of',
costs[most_effective])

And as before we
print the results.

you are here 4  167

review of lists and iteration

 Lists are a data structure for ordered data.
 A list holds a set of items, each with its own index.
 Lists use a zero-based index, where the first item is
at index 0.
 You can use the len function to get the number of
items in a list.
 You can access any item using its index. For
example, use my_list[1] to access the second item
in the list.
 You can also use negative indices to identify items
starting at the end of the list.
 Trying to access an item beyond the end of the list
will result in a runtime index error.
 Assigning a value to an existing item will change its
value.
 Assigning a value to an item that doesn’t exist in
the list results in an index “out of bounds” runtime
error.
 List items can hold values of any type.
 Not all the values in a list need to be the same type.
 Lists that hold values of different types are called
heterogeneous.
 You can create an empty list with
my_list = [ ].
 You can add a new value to a list using append.

168  Chapter 4

 You can extend a list with the items in another list
with extend.
 You can create a new list from two existing lists by
simply adding them together with +.
 Use insert to add a new item at an index in an
existing list.
 The for loop is commonly used to iterate through
sequences, like lists.
 The while loop is most often used when you don’t
know how many times you need to loop, and you’re
looping until a condition is met. The for loop is most
often used when you know the number of times the
loop needs to execute.
 The range function creates a range of integers.
 You can iterate over ranges with the for loop.
 The str function converts a number to a string.

lists and iteration

Coding cross
Let lists sink into your brain as you do the crossword.
1
3

7

1

34

4

5

5

2

6

7

8
9

10

12
14

14

16

16

18

6

8

9

10

11

11

1213

13
15

17

Across
Across

2

15

17

18

DownDown

4. Author
4. Author
probably
probably
likes this
likes
kind
thisofkind
music.
of music. 1. An ordered
1. An ordered
data structure.
data structure.
5. Another
word for
loop.
2. A list2.with
of different
types. types.
5. Another
word
for loop.
A listvalues
with values
of different
7. Gets7.the
length.
3. Iterates
over a over
condition.
Gets
the length.
3. Iterates
a condition.
8. Creates
a sequence
of numbers.
5. Puts5.anPuts
item
a list.
8. Creates
a sequence
of numbers.
aninto
item
into a list.
9. Computer
scientists
like to like
starttolists
atlists
this at this6. How6.toHow
delete
an item.
9. Computer
scientists
start
to delete
an item.
index. index.
12. A list
items.
12.with
A listno
with
no items.
10. Accessing
an item
doesn't
exist results
13. Iterates
over a over
sequence
or range.
10. Accessing
anthat
item
that doesn't
exist results
13. Iterates
a sequence
or range.
in this in
error.
15. Adds
oneof
listone
to another.
this error.
15.items
Adds of
items
list to another.
11. Best
company.
11.bubble
Best bubble
company.
14. We14.
made
lot ofathese
We amade
lot of drinks.
these drinks.
16. Makes
numbers
into strings.
16. Makes
numbers
into strings.
17. Every
one.
17. item
Everyhas
item
has one.
18. Adds
item
a list.
18.an
Adds
antoitem
to a list.

you are here 4  169

exercise solution

It’s a good time to get some practice with lists. Pretend you’re the
Python interpreter and trace though this code, figuring out what
the final output is. After you finish, we’ll take our list knowledge a
little further.

r
Let’s keep track of the eighties and othe
The
.
code
this
ugh
thro
e
variables as we trac
code.
newwave list never changes throughout this
eighties = ['', 'duran duran', 'B-52s', 'muse']

newwave = ['flock of seagulls', 'postal service']

remember = eighties[1]

eighties

remember

band

[‘ ’, ‘duran duran', ‘B-52s', ‘muse']

‘duran duran'

eighties[1] = 'culture club'

[‘ ’, ‘culture club', ‘B-52s', ‘muse']

‘duran duran'

band = newwave[0]

[‘ ’, ‘culture club', ‘B-52s', ‘muse']

‘duran duran'

‘flock of seagulls'

[‘ ’, ‘culture club', ‘B-52s', ‘flock of seagulls']

‘duran duran'

‘flock of seagulls'

[‘B-52s’, 'culture club', ‘B-52s', ‘flock of seagulls']

‘duran duran'

‘flock of seagulls'

eighties[3] = band
eighties[0] = eighties[2]
eighties[2] = remember

[‘B-52s', 'culture club', ‘duran duran', ‘flock of seagulls'] ‘duran duran'

print(eighties)

Python 3.6.0 Shell
['B-52s', 'culture club', 'duran duran', 'flock of seagulls']
>>>

The final output

170  Chapter 4

‘flock of seagulls'

lists and iteration

The Thing-A-Ma-Jig

The Thing-A-Ma-Jig is quite a contraption—it clanks and clunks and even thunks,
but what it really does, well, you’ve got us stumped. Coders claim they know how it
works. Can you uncrack the code and find its quirks?
What does it all mean? The Thing-A-Ma-Jig takes a sequence of characters and
it creates and prints out a palindrome. You might remember that a palindrome is a
word that reads the same forward as it does backward, like “tacocat”. So if we give
the Thing-A-Ma-Jig a sequence of t-a-c-o, it will turn that into the palindrome
tacocat for us. For a lot of sequences, like t-a-r, the results aren’t so interesting:
‘tarat’. But for others, like a-m-a-n-a-p-l-a-n-a-c, they are quite impressive: a
manaplanacanalpanama (or, “a man a plan a canal panama”).
The important thing, though, is how does this code take half a palindrome and create
a full one. Let’s study the code:
characters = ['t', 'a', 'c', 'o']
output = ''

We’ll start with output as the empty string.

Let’s get the length of the characters list.
And set i to 0.
Now we’ll start at 0 and
while (i < length):
go through each item in
output = output + characters[i]
the list, adding it to the
i = i + 1
output list.
Now let’s reset things a bit, setting the length
length = length * -1
to its corresponding negative value (so, for
i = -2
example, a length of 8 would become -8).
And we’ll set i to -2; you’ll see how this is used
while (i >= length):
in the code block.
Why backward? Because
output = output + characters[i]
our indices are now
negative, not positive.
i = i - 1
Now we’re looping over the characters
backward! We’re also skipping the very last
print(output)
character so it doesn’t get repeated in the
middle of the string.
Finally, we print the results.
Study this until you understand it! Go through
each
pass of the loop and write down the values
We’ll be seeing more of
of
the
variables and lists that are changing if
palindromes in Chapter 8.
necessary (and it may be!).
length = len(characters)
i = 0

you are here 4  171

exercise solution

The smoothie flavors were added to this list in the order of their
creation. Finish the code to determine the most recent smoothie
flavor created.

smoothies = ['coconut', 'strawberry', 'banana', 'pineapple', 'acai berry']

-1
most_recent = ________________

We can make use of Python negative
indexes to grab the last item in the
list with -1.

recent = smoothies[most_recent]

smoothies = ['coconut', 'strawberry', 'banana', 'pineapple', 'acai berry']

len(smoothies) - 1
most_recent = ________________

Or we can get the length of the list,
and subtract one.

recent = smoothies[most_recent]

You can take this a little further.
Note that while breaking down things into clear
steps can often help with the clarity of our code,
for simple, common operations more succinct code
can actually be more readable. Let’s take the first
attempt at the code above and make it more succinct:
smoothies = ['coconut', 'strawberry', 'banana', 'pineapple', 'acai berry']

-1
most_recent = ________________

recent = smoothies[-1]

We can do this with the second
code attempt as well.

We can just get rid of the most_recent
variable and use the -1 value directly in
the list index.

smoothies = ['coconut', 'strawberry', 'banana', 'pineapple', 'acai berry']

len(smoothies) - 1
most_recent = ________________

recent = smoothies[len(smoothies)-1]

Here we’ve also gotten rid of the intermediate most_recent variable and just moved the
len(smoothies)-1 calcuation into the index of the list. This is arguably not as readable, but to an
experienced programmer it most likely would be. Code clarity is more an art than a science, so go
with what you find most clear and readable—and keep in mind that may change over time!

172  Chapter 4

lists and iteration

Code Magnets Solution

We’ve got code for testing to see which
smoothies have coconut in them. We had all
the code nicely laid out on our fridge using
fridge magnets, but the magnets fell on the
floor. It’s your job to put them back together.
Be careful; a few extra magnets got mixed in.

:

smoothies = ['coconut',

i = i + 2

)
while i > len(has_coconut

'strawberry',
'banana',

'tropical',

'acai berry']
has_coconut = [True,

False,
False,
True,

False]
i = 0
while i < len(has_coconut)
if has_coconut[i]

:

:

print(smoothies[i],'contains coconut')
i = i + 1

Here’s the output
we’re expecting.
Python 3.6.0 Shell

Rearrange the magnets here.

coconut contains coconut
tropical contains coconut
>>>

you are here 4  173

exercise solution

SOlUTion
We had our range function calls figured out, and then they got all mixed
up. Can you help us figure out who does what? Be careful, we’re not sure
if each range matches zero, one, or more sequences. We’ve already figured
one out, which is marked below:

range(8)
range(0, 8)
range(0, 8, 1)
range(0, 8, 2)
range(0, 7)
range(-3, 1)
range(3, 8)

-3, -2, -1, 0
3, 6
0, 1, 2, 3, 4, 5, 6, 7
0, 1, 2, 3, 4, 5, 6
0, 2, 4, 6
3, 4, 5, 6, 7

range(3, 8, 3)

3, 5, 7

range(3, 8, 2)

1, 2, 3, 4

174  Chapter 4

lists and iteration

Rewrite your fridge magnet code (from a few
pages back) so that it uses a for loop instead of a
while loop. If you need a hint, revisit the way we
reworked the while loop for Bubbles-R-Us.

smoothies = ['coconut',

'strawberry',
'banana',

'tropical',

'acai berry']
has_coconut = [True,

Your code goes here.

False,
False,
True,

smoothies = ['coconut',

False]

'strawberry',

i = 0

'banana',

while i < len(has_coconut)
if has_coconut[i]

:

print(smoothies[i],
'contains coconut')
i = i + 1

:

'tropical',

'acai berry']
has_coconut = [True,

False,
False,
True,

False]
length = len(has_coconut)
for i in range(length):
if has_coconut[i]:

Let’s get the length
of the list.

And iterate from 0 to the
length of the list (minus 1).

print(smoothies[i], 'contains coconut')

Test the has_coconut list at index i to see if the item has coconut
and if so, print its name from the smoothies’ list at i.
Notice we could have written “if has_coconut[i] == True”, but
has_coconut[i] evaluates to a Boolean value, so it isn’t really necessary.

you are here 4  175

exercise solution

Go ahead and implement the pseudocode on the previous page to find the
highest score by filling in the blanks in the code below. Once you’re done,
add the code to bubbles.py and give your code a test. Check the results in the
shell, and fill in the blanks in our shell below with the number of bubble tests
and the highest score.

scores = [60, 50, 60, 58, 54, 54,

58, 50, 52, 54, 48, 69,
34, 55, 51, 52, 44, 51,
69, 64, 66, 55, 52, 61,
46, 31, 57, 52, 44, 18,
41, 53, 55, 61, 51, 44]

Fill in the blanks to complete the code here...

0
high_score = ______
length = len(scores)

for i in range(length):

print('Bubble solution #' + str(i), 'score:', scores[i])

scores[i]
if _____________
> high_score:

high_score = scores[i]
_____________

print('Bubbles tests:',

____________)
length

print('Highest bubble score:',

high_score
_____________)
Python 3.6.0 Shell

Here’s what we got.

176  Chapter 4

Bubble solution #0 score: 60
Bubble solution #1 score: 50
Bubble solution #2 score: 60
...
Bubble solution #34 score: 51
Bubble solution #35 score: 44
Bubbles tests: 36
Highest bubble score: 69

lists and iteration

You’ve seen how to create an empty list with two square backets,
but what do you think the code below does? Feel free to type the
code into the shell to experiment. Record your observations here.
mystery = ['secret'] * 5

mystery = 'secret' * 5

This syntax creates a list with ‘secret' as an item repeated
five times: [‘secret', ‘secret', ‘secret', ‘secret', ‘secret']. This
is a Python special feature, so you won't find it in too many
other programming languages. It does come in handy at times
(for instance, in Chapter 11).

Multiplication and strings? Well, we saw this already when
Codie's “12” was repeated seven times, remember? So if you
multiply a number by a string, you get a new string with the
original string repeated.

Can you help write the loop to find all the scores that match the high score?
You’ll find all the code up until now listed below. Give it a real shot below before
you check the solution.

Again, we're starting by creating a
new list that will hold all the bubble
solutions that match the highest score.
best_solutions = []

for i in range(length):

if high_score == scores[i]:

best_solutions.append(i)

Next, we iterate through the entire
scores list, looking for those items
with the highest score.
Each time through the loop,
we compare the score at
index i with the high_score
and if they are equal, then
we add the score at that
index to the best_solutions
list using append.

We're just
showing you the
new code. You
know, saving
trees and all
that.

print('Solutions with the highest score:', best_solutions)

And finally, we can display the bubble solutions with the highest scores. Notice we’re
using print to display the best_solutions list. We could have created another loop to
display the list items one by one, but, luckily, print will do this for us (and, if you look
at the output, it also adds commas between the list values; just what we need).
you are here 4  177

exercise solution

Coding cross Solution
Let lists sink into your brain as you do the crossword.

1
3

W

E

H
7

5

I

L

E

E

2

L

4

I

H

G H

T

I

S
I

T

E

R A
9

S

T

R A

R O

T

I

M E

U B

B

L

11

B

S

T

T

H

R
R

I

18

A

S

R U S

I

15

E

U
S

X
17

Y

E
O

F

P

L

N

13

S M O O

16

N G E

G

R U N

14

D

8

E

E

E

6

E

Z

10

12

S

T

N

T

E

T
N D E

X

N
P

P

E

N D

Across

Down

4. Author probably likes this kind of music.
[EIGHTIES]
5. Another word for loop. [ITERATE]
7. Gets the length. [LEN]
8. Creates a sequence of numbers. [RANGE]
9. Computer scientists like to start lists at this
index. [ZERO]
10. Accessing an item that doesn't exist results
in this error. [RUNTIME]
11. Best bubble company. [BUBBLESRUS]
14. We made a lot of these drinks. [SMOOTHIES]
16. Makes numbers into strings. [STR]
17. Every item has one. [INDEX]
18. Adds an item to a list. [APPEND]

1. An ordered data structure. [LIST]
2. A list with values of different types.
[HETEROGENEOUS]
3. Iterates over a condition. [WHILE]
5. Puts an item into a list. [INSERT]
6. How to delete an item. [DEL]
12. A list with no items. [EMPTY]
13. Iterates over a sequence or range. [FOR]
15. Adds items of one list to another. [EXTEND]

178  Chapter 4

5

functions and abstraction

Getting Functional
Now that we
know how to abstract
our code, the future’s
so bright we gotta wear
shades.

You already know a lot. Variables and data types and
conditionals and iteration—that’s enough to write basically any program
you’d ever want to. In fact, a computer scientist would tell you it’s enough
to write any program that anyone could ever conceive of. But you don’t
want to stop now, because your next step in computational thinking is
learning how to create abstractions in your code. That may sound
complex, but it’s actually going to make your coding life simpler. Creating
abstractions gives you leverage; with abstraction, you can more easily
create programs of increasing complexity and power. You can put your
code in neat little packages that you can reuse over and over. And you
can forget all the nitty-gritty details of your code and to start thinking at a
higher level.

this is a new chapter   179

some code analysis

Do a little analysis of the code below. How does it look? Choose as many of
the options below as you like, or write in your own analysis:
dog_name = "Codie";
dog_weight = 40
if dog_weight > 20:
print(dog_name, 'says WOOF WOOF')
else:
print(dog_name, 'says woof woof')
dog_name = "Sparky"
dog_weight = 9
if dog_weight > 20:
print(dog_name, 'says WOOF WOOF')
else:
print(dog_name, 'says woof woof')

Discussing my
age wasn’t enough?

dog_name = "Jackson"
dog_weight = 12
if dog_weight > 20:
print(dog_name, 'says WOOF WOOF')
else:
print(dog_name, 'says woof woof')
dog_name = "Fido"
dog_weight = 65
if dog_weight > 20:
print(dog_name, 'says WOOF WOOF')
else:
print(dog_name, 'says woof woof')

A. It’s the same code over and
over, so it seems very redundant.

D. Not the most readable code I’ve
ever seen.

B. It looks tedious to type in!

E. If we wanted to change how the
dogs bark, that would be a lot of
changes to make!
F. ___________________________
_____________________________

C. It doesn’t seem to do a lot for
that much code.

180  Chapter 5

Codie,

functions and abstraction

What’s wrong with the code, anyway?
Okay, so we have some code that is repeated over and over. What’s wrong
with that? Well, at face value, nothing. After all, it works, right? Let’s
have a closer look at the code in question:
dog_name = "Codie"
dog_weight = 40
if dog_weight > 20:
print(dog_name, 'says WOOF WOOF')
else:
print(dog_name, 'says woof woof')
dog_name = "Sparky"
dog_weight = 9
if dog_weight > 20:
print(dog_name, 'says WOOF WOOF')
else:
print(dog_name, 'says woof woof')

Sure, this code looks innocent enough, but it’s tedious to write, it’s
hard to read, and it will be problematic if your code needs to change
over time. That last point will become more and more clear as you
gain experience in programming—almost all code does change over
time, and the code above is a nightmare waiting to happen because
we’ve got the same logic repeated over and over.
Say you need to add a new bark called “yip yip,”for tiny dogs that are
less than 2 lb. How many places do you need to make that change in
the existing code?

What we're doing here is comparing
the dog’s weight to 20, and if it's
greater than 20, we're outputting a
big WOOF WOOF. If it's less than 20
or equal to 20, we’re outputting a
smaller woof woof.
And this code is...wait, it's
doing EXACTLY the same
thing! And the same is true of
the rest of the code.
dog_name = "Jackson"
dog_weight = 12
if dog_weight > 20:
print(dog_name, 'says WOOF WOOF')
else:
print(dog_name, 'says woof woof')
dog_name = "Fido"
dog_weight = 65
if dog_weight > 20:
print(dog_name, 'says WOOF WOOF')
else:
print(dog_name, 'says woof woof')

Again, only
the details
of the dog
are different.
And again,
different
dog, same
code.

Oh boy, so what do we do?

How can you improve this code? Take a few
minutes to think of a few possibilities.

you are here 4  181

reusing code

If only I could find a way to
reuse code so that anytime I needed it,
I could just use it rather than retyping it. And
a way to give it a nice memorable name so that I
could remember it. And a way to make edits in just
one place instead of many if something changes. That
would be dreamy. But I know it’s just a fantasy...

182  Chapter 5

functions and abstraction

Turning a block of code into a FUNCTION
What if we told you that you could take a block of code, put a name on it, and then use and
reuse that code anytime you like? You’d probably say, “What took you so long?!”
Python gives us a way to do exactly that, and it’s called defining a function. You’ve already used a
few functions, like, print, str, int, and range. Let’s look at how they work again:

We call or invoke a The name is
followed by a
function by using
parenthesis.
its name in code.

And then we provide
zero or more
arguments we pass
along to the function.

print("I'm a function")

And then finally we have
an ending parenthesis.

When we call a function, we’ve assumed the function goes off and does a
bunch of work (like printing the values we passed it) and then at some point
returns and our code resumes where it left off. And that’s exactly what a
function does.
You don’t have to be satisfied calling prebuilt functions; you can create your
own. Here's how you do that:

To create a function, start
with the def keyword, and
follow it with a name to
remember the function by.

You'll find all
programming languages
(at least languages that
you'd want to use) give
you a way to define
functions.

You might have also
noticed that sometimes
a function returns
something to us, like with
the str function, which
returns a string that
represents an integer.

Next you can have zero or
more parameters. Think of
these like variables that hold
values that you will pass into
the function when you call it.
And then we have a colon, which, as you
know, starts a block of code in Python.

def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof')

When the Python interpreter sees this code, it simply defines
the function and stashes the code away for later use—the body
of the function is not executed at this point. The body is only
executed when the function is called.

}

Here’s the block of code
we are going to reuse.
In Python and pretty
much all languages, we call
this code block the body
of the function.

So now that you know how to define a function, let’s see how you use it.
you are here 4  183

how to call functions

We created a function, so how do we use it?
Now that you’ve created a function by taking a block of code, giving it a
name, and defining some parameters (the dog’s name and weight) that
you’ll be supplying when you call it, you’re all ready to use your new
function.
And given your experience with functions like print, str, random, and
range, you already know what to do. Let’s call bark a few times to see
how it works:
bark('Codie', 40)
bark('Sparky', 9)
bark('Jackson', 12)
bark('Fido', 65)

Let’s test bark
with all the dogs
we know about.

Great, that’s the output
we were expecting!

Python 3.6.0 Shell
Codie says WOOF WOOF
Sparky says woof woof
Jackson says woof woof
Fido says WOOF WOOF
>>>

But how does all this actually work?
We’ve created a function and we’ve put it to work, and everything seems
like it is working as expected, but what is really going on behind the
scenes? How does this all really work? Let’s take things step by step and
see. To do that, let’s step through the following code:

Here we start with a print
statement, just to get things going.
print('Get those dogs ready')

And then we have
our bark function
definition.

def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof')
bark('Codie', 40)
print("Okay, we're all done")

And finally we use a
print statement to
say we’re done.
184  Chapter 5

And then we’re
going to call the
bark function with
the arguments
Codie and 40.

Pretend you are the Python
interpreter. Start at the top of
the code and mentally step
through and execute each
line. Are there any steps that
are unclear to you?

functions and abstraction

Like any good interpreter, let's
start at the beginning. So our

first line of code is a print statement. This
executes as you’d expect and outputs ‘Get those
dogs ready’ to the Python Shell.

We start here.

Behind
the Scenes

Python 3.6.0 Shell
Get those dogs ready

So far we just have a print statement,
which sends its output to the shell.

print('Get those dogs ready')
def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof')
bark('Codie', 40)
print("Okay, we're all done")

Up next: the function definition.

After executing the print statement, the interpreter
encounters the function definition for the bark
function. At this point, the Python interpreter
doesn’t execute the code in the function; rather, it
creates a name, bark, and stores away the function
parameters and body (after giving the body a quick
syntax scan) for later use. After the definition is
processed by the interpreter, we can use the name
bark to invoke the function any time we want.

The whole point of this bit of code isit.
to create the function, not to invoke
So after this code is evaluated, we’re
all ready to invoke this function at any
time with the function’s name, bark.

print('Get those dogs ready')
def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof')
bark('Codie', 40)
print("Okay, we're all done")

you are here 4  185

closer look at calling functions

Next we have a call to the bark
function. After the function definition is complete,

the interpreter next encounters a function call.
This time it’s a call to the bark function with two
arguments: the string ‘Codie’ and the number 40.

The interpreter retrieves the function definition named
bark from memory, and assigns the arguments ‘Codie’
and 40 to the parameters name and weight, respectively.

Here we’re passing two
arguments, which are
‘Codie’ and 40...
bark('Codie', 40)

Behind
the Scenes
It’s time to invoke
our function.
print('Get those dogs ready')
def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof')
bark('Codie', 40)
print("Okay, we're all done")

'Codie'

40

def function bark(name, weight)
...

...which are assigned
to the parameters
name and weight.

illk

Buy M

Make it stick

Python is an
interpreter

We call, or invoke, a function.
You pass arguments into
your function calls.

You can think of name and weight as brand new
variables that are around as long as your function
body is being executed. Anytime you refer to them in
your function body, you can count on them to hold the
values of the arguments you passed into your function.
Next, after the parameters have been set, the
interpreter takes bark’s code and begins executing it.

186  Chapter 5

A function has zero or more
parameters that accept values from
your function call.

functions and abstraction

Now we execute the function
body. An important thing to notice here is that

the flow of the program has now gone from the
call to the bark function to the body of the bark
function—we’re no longer proceeding straight
through our code, but rather, we’ve jumped to the
code in the function body. Keep that in mind.

Behind
the Scenes

In the function body we first have a conditional test
to see if the weight parameter is greater than 20,
and given we passed in 40, it is, so we execute the
code block of the conditional.

uments have been
After the values of your arg
en it’s time to
th
assigned to each parameter,of your function.
start executing the body

The block of the conditional prints the parameter
name, which has the value of ‘Codie’ and then
‘says WOOF WOOF’.

print('Get those dogs ready')

Python 3.6.0 Shell
Get those dogs ready
Codie says WOOF WOOF.

def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof')
bark('Codie', 40)
print("Okay, we're all done")

And with that, we complete the code in the body of
the bark function. So where do we go from here?
Well, when a function completes, the control of the
program returns back to where the function was
called, and the interpreter resumes execution there.
print('Get those dogs ready')

Remember, when a
function completes, the
control of the program
returns back to where the
function was called, and
the interpreter resumes
execution there.

def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof')
bark('Codie', 40)
print("Okay, we're all done")

With the call to bark finished, the interpreter
resumes execution right after the call to the
bark function. So, we’ll pick back up here.

you are here 4  187

testing the bark function

Now we execute the code after
the function call. Finally, we return

from the call to the bark function and the only
code remaining is a simple print statement that
outputs Okay, we’re all done to the Python
Shell.

Python 3.6.0 Shell
Get those dogs ready
Codie says WOOF WOOF.
Okay, we’re all done

Behind
the Scenes
print('Get those dogs ready')
def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof')
bark('Codie', 40)
print("Okay, we're all done")

Finally we reach the last line of

A Test Drive

our program.

That’s a lot to take in. Take a deep breath, and then get this code into a file called
bark.py, save your code, and choose the Run > Run Module menu item to make sure
everything’s good.
Here’s the code:

Add all this new
code to your file.

def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof')
bark('Codie', 40)
bark('Sparky', 9)
bark('Jackson', 12)
bark('Fido', 65)

Here's what we
got when we
tested the code.

188  Chapter 5

Python 3.6.0 Shell
Codie says WOOF WOOF
Sparky says woof woof
Jackson says woof woof
Fido says WOOF WOOF
>>>

WOW, look how much more
readable and clear this code is!

functions and abstraction

Here’s a little connect-the-dots exercise for you. Using your
pencil, start at step 1 and draw a line from each step to the next
as the program executes. Feel free to annotate your lines where
necessary to explain what is happening in the code’s execution. We
already drew the first couple lines for you.

We start here.
Print, and then
proceed to the
next line.

1

print('Get those dogs ready')

2

def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof')

a
b

3

bark('Codie', 40)

4

print("Okay, we're all done")

We’ve got some more calls to bark below. Next to each call, write what you think the
output should be, or if you think the code will cause an error. Check your answer at the
end of the chapter before you go on.
bark('Speedy', 20)

Write what you
think the shell
will display here.

bark('Barnaby', -1)
bark('Scottie', 0, 0)

Hmm, any ideas
what these do?

bark('Lady', "20")
bark('Spot', 10)
bark('Rover', 21)

you are here 4  189

functions and abstraction

I thought you
said we were going to
learn about abstraction? So
far it seems this chapter is
just about functions.

Functions are a way to abstract your code.
Think about the first dog example in this chapter. We had
a lot of messy code, and if you looked at that code it was a
bit hard, at least at first, to figure out what its purpose was.
But after some study, it was pretty obvious: we wanted each
dog to bark. And we wanted it to bark in a way that was
representative of the dog’s size—WOOF WOOF for big
dogs, and woof woof for small dogs.
So, what we did was pull the code that does the barking
out and abstracted it into a function. Once we did that, we
could simply write code like this:
bark('Codie', 40)

We no longer had to worry about how the code barks (as
we did with every single case of barking before we had this
function), we just need to use the bark function. And, two
months from now when you need to make some more dogs
bark, you can reuse this function with very little knowledge
of how it works. That allows you to focus on what you’re
coding (say, a dog show simulator), not the low-level details
of barking.
So we are taking code, abstracting it into a function, and
then making use of that abstraction in our code.

190  Chapter 5

Or your
coworker
does.

functions and abstraction

Code Magnets

,

,

This working code is all scrambled up on the fridge.
Can you reconstruct the code snippets to make a
working program that produces the output listed
below? Notice, there may be some extra code on the
fridge, so you may not use all the magnets. Check
your answer at the end of the chapter.

:
:

,

}
}

(

def

)

}
{

how_should_I_get_there(800.3)
else:

print('Walk')

how_should_I_get
_there
elif miles >= 2.0:

print('Take a car')

miles

kilometer

if miles > 120.0:
print('Take a plane')

how_should_I_get_there(.5)

Python 3.6.0 Shell

how_should_I_get
_there(2.0)

Take a plane
Take a car
Walk

you are here 4  191

more about functions

Q:

Do I need to define my function
before the code that calls it? Or can I put
my functions at the end of the file?

A:

Yes, you need to define functions
before they are called in your code. One
thing to consider: say we have two functions,
f1 and f2, where f1 calls f2 in its body. In this
case, it is perfectly fine to define f2 after f1 in
your code, so long as f1 does not get called
before f2 is defined. That’s because defining
the function body of f1 does not invoke
f2, until f1 is actually called. In terms of
where to put functions, we suggest defining
functions at the top of your file for better
organization and clarity.

Q:

What kind of values can I pass to a
function?

A:

In Python you can pass any of the
data types you’ve learned (as well as all the
ones you will learn) to a function: Booleans,
strings, numbers, and lists. Believe it or not,
you can even pass a function to another

function—we will get to why you’d want to
and what it gets you later in the book (in the
Appendix).

Q:
A:

Sorry, I’m still confused on the
whole argument versus parameter thing.

Don’t overthink it. They are both terms
for simple things. Arguments are the values
you pass to a function when you make a
function call. Parameters are part of the
function definition, and get initialized to the
arguments passed to the function call.

Q:

What happens if I mix up the
order of my arguments, so that I’m
passing the wrong arguments into the
parameters?

A:

All bets are off; in fact, we’d guess
you’re pretty much guaranteed either an
error at runtime or incorrectly behaving code.
Always take a careful look at a function’s
parameters, so you know what arguments
the function expects to be passed and in
what order.

That said, there are some other ways to
approach passing arguments that we’ll look
at before the chapter is over.

Q:

What are the rules for function
names?

A:

The rules for naming a function are
the same as the rules for naming a variable,
which we saw in Chapter 2. Just start with
an underscore or letter, and continue with
letters, underscores, or numbers. Most
Python programmers, by convention, keep
their function names all lowercase with
undescores between words, like get_name
or fire_cannon.

Q:
A:

Can functions call other functions?

Yes, happens all the time. Note you’re
already doing this when you call the print
function within the bark function code. Your
own functions are no different; you can call
them from your other functions.

Functions can RETURN things too
So far you’ve only passed things to a function; that is, you know how to call a function and pass
arguments to it. But you can also get a value back from a function, with the return statement:

Here's a new function, get_bark, that returns
the appropriate bark, given a dog's weight.
def get_bark(weight):
if weight > 20:

return 'WOOF WOOF'

else:

return 'woof woof'

192  Chapter 5

But only one will be
executed, because
as soon as a return
we
20
than
If the weight is greater
statement is
WOOF'.
‘WOOF
return the string
executed, the code
immediately returns
from the function.

Otherwise, we return the string ‘woof woof'.
A function can have zero, one,
or more return statements.

functions and abstraction

How to call a function that has a return value
So we’ve got a function, get_bark, that we just need to pass the weight of our dog, and it
returns the appropriate bark for that dog. Let’s see how to use it:

You call the function like any other function,
only this function returns a value, so let's
set that value to a variable, codies_bark.

Python 3.6.0 Shell

Codie's bark is WOOF WOOF
>>>

codies_bark = get_bark(40)
print("Codie's bark is", codies_bark)

Get some practice to make return values sink in. Compute the
return value of each function call.
g(name):
'
_greetin
ame + '!
def make
Hi ' + n
return '
def compute(x, y):
total = x + y
if (total > 10):
total = 10
return total

Defined on the
previous page

get_bark(20)
make_greeting('Speedy')

def allow_acce
ss(person):
if person ==
'Dr Evil':
answer = True
else:
answer = False
return answer

Write what
each call
returns here.

compute(2, 3)
compute(11, 3)
allow_access('Codie')
allow_access('Dr Evil')

you are here 4  193

understanding local variables

I noticed in that last exercise
that you declared some new
variables right inside your
functions, like total and answer.

Good catch.
We sure did. You can declare new
variables right inside of your function,
just as we did. It’s often handy to have
variables that hold the result of an
intermediate calculation that you’ll
need in your function. We call these
local variables, because they are local to
the function and only exist as long as
the function execution does. That’s in
comparison to the global variables we’ve
been using so far, which exist as long as
your entire program does.
But hold that thought. Given our new
knowledge of functions, parameters,
return values, and so on, we’ve just
received a request to help “refactor”
some code. We never turn down a
chance to do real coding, so we’re going
to do that and then circle back and
dig into variables (local, global, and
otherwise) some more.

194  Chapter 5

functions and abstraction

Getting a little refactoring under our belts
A new startup around the corner has been working on some code to help
their users choose an avatar—you know, an onscreen representation, real or
imagined. They’ve just started, and so far all their code does is ask for the
user’s preferences for hair color, eye color, gender, and so on—once they get
this all working, they’ll presumably take all these preferences and generate a
nice avatar image for each user.
But they feel like their code is already getting too complex for such a simple
task. You’ll find the code below; notice that they’ve tried to make this easy on
the user by supplying some default values—the user can either type in a value,
or just hit return to accept the default.

For each attribute of the avatar, we
prompt the user, including a default
choice, like brown hair.
hair = input("What color hair [brown]? ")
if hair == '':
hair = 'brown'
print('You chose', hair)

If the user simply hits the Return key, then
we assign the variable to the default value.
Otherwise, we use what they typed in.

hair_length = input("What hair length [short]? ")
if hair_length == '':
hair_length = 'short'
print('You chose', hair_length)
eyes = input("What eye color [blue]? ")
if eyes == '':
eyes = 'blue'
print('You chose', eyes)

We also print each user's choice.

We do this over and over for
each attribute.

gender = input("What gender [female]? ")
if gender == '':
gender = 'female'
print('You chose', gender)
has_glasses = input("Has glasses [no]? ")
if has_glasses == '':
has_glasses = 'no'
print('You chose', has_glasses)
has_beard = input("Has beard [no]? ")
if has_beard == '':
has_beard = 'no'
print('You chose', has_beard)
you are here 4  195

testing the avatar code

Running the code
Let’s do a little run-through of the code so far to see how it
works. Before we do, though, make sure you’ve studied the
code a bit so you get the gist of it.

The program prompts us for
each attribute. We either
type out an attribute, or
accept the default (like short
hair length and no glasses) by
hitting return.

And for each attribute, the
program confirms the user's
choice by printing it back out.

Python 3.6.0 Shell
What color hair [brown]? blonde
You chose blonde
What hair length [short]?
You chose short
What color eyes [blue]? brown
You chose brown
What gender [female]? male
You chose male
Has glasses [no]?
You chose no
Has beard [no]? yes
You chose yes

Obviously our avatar code is in need of some abstraction. Use this space to work
out how you think this code should be abstracted into a function (or functions) and
what your function calls might look like. When you’re done, we’ll walk through it
together on the next page (but make sure you do your own work first!).

196  Chapter 5

functions and abstraction

How to abstract the avatar code
Now that you’ve done your own thinking, let’s take a look and find
the commonality in the code that we can abstract into a function:

For each attribute we're prompting
the user and getting their response.

Each time we prompt we
ask a different question.

And we have a differ
ent default
value, like brown, shor
t, blue.

hair = input("What color hair [brown]? ")
if hair == '':
hair = 'brown'
print('You chose', hair)
hair_length = input("What hair length [short]? ")
if hair_length == '':
hair_length = 'short'
print('You chose', hair_length)

For every
attribute we do
exactly the same
thing.

eyes = input("What eye color [blue]? ")
if eyes == '':
eyes = 'blue'
print('You chose', eyes)

We're just showing
the code for three
attributes to save
a little paper.
Each attribute is also
assigned to a variable, like
hair or eyes.
.
.
.

And we print the
attribute back to
the user.

There are two things that vary over each bit of code, and that’s the
question and the default value. Those are going to be our parameters
because each time we call our function they will be different. Let’s
start there:

Let’s call our function get_attribute.
def get_attribute(question, default):

And we’ll use two parameters, question
for the question to ask, like “What
color hair”, and default, to supply the
default value, like “brown”.

you are here 4  197

writing the get_attribute function

Writing the get attribute function body
Now let’s work on the function body. Based on the original code, the
first thing we need to do is create a string that acts as a question, so
that we can prompt and get input from the user.
def get_attribute(query, default):
question = query + ' [' + default + ']? '
answer = input(question)

Let’s put together the question for
the user by using the query parameter
along with the default parameter, and
we’ll format the string just like the
previous code did.

Then we’ll prompt the user and get
their input. We’ll assign the answer
variable to their input.
Now, just as the previous code did, we need to check to see if the user
chose the default value by hitting return (in which case, the answer
will be the empty string). Then we need to print out their choice:
def get_attribute(query, default):
question = query + ' [' + default + ']? '
answer = input(question)
Check the answer against the empty
if (answer == ''):
string, and if it is the empty string, set
answer = default
answer to the default parameter.
the
print('You chose', answer)
Finally, we just have one thing left to do: we need to get the answer
back to the code that called get_attribute. How do we do that?
With the return statement, of course:
def get_attribute(query, default):
question = query + ' [' + default + ']? '
answer = input(question)
if (answer == ''):
answer = default
print('You chose', answer)
return answer

We’ve got the user’s answer, so the only
thing left to do is return it.

198  Chapter 5

functions and abstraction

Calling get attribute
Now, for each attribute, we just need to write the appropriate call to
the get_attribute function.
def get_attribute(query, default):
question = query + ' [' + default + ']? '
answer = input(question)
if (answer == ''):
answer = default
print('You chose', answer)
return answer

We’ve taken each of the original
attributes and created a call to
get_attribute for each one.

hair = get_attribute('What hair color', 'brown')
hair_length = get_attribute('What hair length', 'short')
eye = get_attribute('What eye color', 'blue')
gender = get_attribute('What gender', 'female')
glasses = get_attribute('Has glasses', 'no')
beard = get_attribute('Has beard', 'no')

A Test Drive

Get this code above into a file called avatar.py, save your code, and choose the
Run > Run Module menu item to give the avatar code a try.

her look at
Make sure you take anatothow much
the code above; look rstandable
more concise and unde it’s more
it is! Not to mention ed to change
maintainable if you nere.
something in the futu

Python 3.6.0 Shell

Exactly as it looked and
operated before our rework

What color hair [brown]? blonde
You chose blonde
What hair length [short]?
You chose short
What color eyes [blue]? brown
You chose brown
What gender [female]? male
You chose male
Has glasses [no]?
You chose no
Has beard [no]? yes
You chose yes

Serious Coding
Reworking code to make it more concise, readable, and well structured is a
common activity of good coders. Often this activity is called refactoring code.

you are here 4  199

understanding more about local variables

Functions Exposed
This week’s interview:

Sitting down with a Local Variable
Head First: Welcome, Local Variable. We’re glad you
joined us.
Local Variable: Glad to be here.
Head First: So, we really don’t know a lot about you.
Maybe you could fill us in.
Local Variable: Ever declare a variable inside a
function? That’s me.

Local Variable: Oh, Parameters! She’s like a sister to
me. Parameters are essentially local variables, but special
ones that get set up when the function is called. Basically
each parameter is created and set to the value of the
corresponding incoming arguments. And, like a local
variable, a parameter is available to you throughout the
function body, but then it goes away when the function
ends.

Head First: And so how is that any different than any
other time we declare a variable?

Head First: Ah, I didn’t know you knew each other.
Well, the other interviewee that comes to mind is Global
Variable.

Local Variable: I’m meant to be used only within the
function body; in fact, that’s the only place you can use
me!

Local Variable: Ah yes, overrated if you ask me; that
guy’s a pain.

Head First: What do you mean?
Local Variable: Say you’re writing a function and you
need a variable to hold something you’re calculating in
that function, say a variable called position. You declare
position within the function and use it anywhere in
the function body. Now, that variable doesn’t exist until
the function is invoked, and then it exists as long as the
function body is being evaluated, but as soon as your
function returns, the variable disappears.
Head First: What about the next time the function is
invoked?
Local Variable: You get a fresh new local variable
called position, which is around as long as the function
invocation is, and then it goes away again.
Head First: So what good are you if you’re just around
for a bit and you keep disappearing as soon as the
function ends?
Local Variable: Oh, I’m hugely useful. I’m hugely
useful. In a function you need variables to hold the
temporary results as you compute something. So you have
me to do that. And when the function is done, I take care
of all the cleanup; you don’t even have to worry about me
cluttering up your program with unneeded variables.
Head First: So, I don’t think this audience has heard
her, but we’ve also had Parameters in for an interview in
the past; how are you different from her?
200  Chapter 5

Head First: Come again?
Local Variable: A global variable is a variable you
declare outside a function. It has global scope.
Head First: Scope?
Local Variable: Scope is the extent of where your
variables are visible. It’s an area of the program where
you can access a variable, to read or change its value. A
global variable is visible everywhere in your code.
Head First: Well, what’s your scope?
Local Variable: Like I said, just the function body, same
as Parameters.
Head First: Well, what’s the issue with Global Variable
being, well, global scope?
Local Variable: It’s not considered the best design
decision to use lots of global variables.
Head First: Why, they sound like a convenient thing to
me.
Local Variable: I think your readers will hear more
about this later, but it can lead to some issues with large
programs.
Head First: Maybe next time we can have you and
Globals in for a fireside chat?
Local Variable: Oh, I can hardly wait.

functions and abstraction

What does this code output? Are you sure?
Maybe you better test it for real. Why did you
get that result? Was it what you expected?
def drink_me(param):

msg = 'Drinking ' + param + ' glass'
print(msg)

param = 'empty'
glass = 'full'

drink_me(glass)

print('The glass is', glass)

Don’t even think about
skipping this one!

Let’s talk about variables a little more...
We’re finding out there’s more to variables than just declaring them, setting their values, and changing them.
By adding functions into the mix, we’ve introduced a bunch of new concepts like local variables and, by
contrast, global variables. And don’t forget parameters, which act like local variables (except they are set up
for us when the function is invoked). And then we have this concept of scope as well.

Make sure we can recognize each type of variable. Annotate this code by identifying
local and global variables as well as any parameters. Check your answers.
def drink_me(param):

msg = 'Drinking ' + param + ' glass'
print(msg)

param = 'empty'
glass = 'full'

drink_me(glass)

print('The glass is', glass)

you are here 4  201

overview of variable scope

Understanding variable scope
The term scope describes where a variable is visible, or accessible, within your code. The rules
are pretty straightforward:
•

Global variable: visible anywhere in your program, although we’ll see one minor
exception to this in just a sec.

•

Local variable: only visible within the function body where it is declared.

•

Parameter: only visible within the function body where it is declared.

Notice local
variables are
created only
within functions.

Let’s check out this idea of variable scope with some code we’re already familiar with:

The local variables question and
answer have local scope within
the get_attribute function.

The parameters query and default
also have local scope within the
get_attribute function.

def get_attribute(query, default):

question = query + ' [' + default + ']? '
answer = input(question)
if (answer == ''):

answer = default

print('You chose', answer)
return answer

Notice we return
the value in a local
variable, answer,
just before the
function ends.

hair = get_attribute('What hair color', 'brown')

hair_length = get_attribute('What hair length', 'short')

eye = get_attribute('What eye color', 'blue')

gender = get_attribute('What gender', 'female')
glasses = get_attribute('Has glasses', 'no')
beard = get_attribute('Has beard', 'no')

The variables hair, hair_length, eye, gender,
glasses, and beard are all global variables and are
visible anywhere in this code.
202  Chapter 5

functions and abstraction

Q:

How can you return a local variable
from a function if it goes away when the
function completes?

A:

When you return a local variable from
a function, you are returning the value of
the local variable, not the local variable
itself. Think of it like this: if you’re running a
relay race, you pass the baton (value), not
yourself (the variable holding the value), to
the next runner. So given only the value is
being returned, this is not a problem when
the function ends and the variable goes
away (because the value the variable held is
still around).

The same is true of local variables;
they can shadow globals within a
function.

Q:

Q:

What happens if I name a
parameter the same name as a global
variable, or is that even allowed?

A:

It is allowed. What happens is that
within your function body any references to
that variable will refer to the local version
(the parameter), not that global one. So
effectively, that particular global variable
won’t be visible within that function. When
this happens we call it shadowing a
variable (because the local parameter is
overshadowing the global variable). This
is quite common and isn’t necessarily
something to avoid as long as your code
is well structured. If a parameter name
makes sense, go for it—your function really
shouldn’t be referring to global variables in
the function body anyway (more on that topic
later).

I get why we might want local
variables to hold temporary parts of
things we’re computing in a function, but
why do we need parameters. Can’t we
just always refer to global variables with
the values we want?

A:

Technically that could be done,
but it would lead to code that is difficult
to understand and error prone. Using
parameters allows us to write general
functions that don’t need to depend on
specific global variables; instead, they are
parameterized, allowing the calling code
to determine the arguments to pass to the
function.

Take our bark function, for example; if it
relied on globals, how would you make it
work for various dogs?

Spoiler alert: here's what the
drink_me code outputs.

When variables get passed to functions

Python 3.6.0 Shell

Are you still wondering about the output for the drink_me code? Or,
wondering more specifically why the glass didn’t get emptied? Most
people do. The reason has to do with how variables and values are
treated when they are passed to a function. Let’s take a look:

Drinking full glass
The glass is full

>>>

ll'

Here's our drink_me code again.
Let's step through it and see
how the glass remains full after
drink_me is called.
msg = 'Drinking ' + param + ' glass'
print(msg)

param = 'empty'
glass = 'full'

drink_me(glass)

print('The glass is', glass)

'fu

def drink_me(param):

glass

In this code, after the function is
defined, we assign the string value
‘full' to the variable glass.
you are here 4  203

understanding the drink_me function

'fu

ll'

Making the drink me function call
Now let’s call the drink_me function and see what happens with our
variables and parameters.

param

msg = 'Drinking ' + param + ' glass'
param = 'empty'

glass = 'full'

drink_me(glass)

print('The glass is', glass)

'fu

The value ‘full’ is assigned
to the parameter param.

def drink_me(param):
print(msg)

ll'

When drink_me is called, the value of the argument
glass is evaluated, and its value ‘full’ is passed along,
and assigned to the parameter param.

glass

Remember from Chapter 2, when we assign a value to a
variable (and you can think of param as a variable), we
create a location for the value, and then we label the
location (a cup in our diagram) with the variable name.
NOTE: the way Python handles passing arguments
is a little more complex than this, in particular
when we start talking about objects, but for now
this is a good working model for us.

Next we construct the msg using the
value of the parameter param.

Python 3.6.0 Shell

Drinking full glass

def drink_me(param):

'fu

ll'

msg = 'Drinking ' + param + ' glass'
print(msg)

param = 'empty'

And then we print it.

glass

print('The glass is', glass)

The value of param at
this point is ‘full'.

'fu

drink_me(glass)

ll'

glass = 'full'

param
204  Chapter 5

functions and abstraction

Nothing about the glass
variable has changed; it is
still labeling the value ‘full'.

'fu

ll'

Now for the important part. We're going
to assign ‘empty' to param.

mp
ty
'

glass

def drink_me(param):

msg = 'Drinking ' + param + ' glass'

'e

print(msg)

param = 'empty'

drink_me(glass)

print('The glass is', glass)

param

Now we’ve changed the
value of the param
parameter to ‘empty’.

Finally we return from the
function call to drink_me, and we
execute the print statement.

Our parameter param no longer
exists because the function call
has completed. Sad.

'fu

def drink_me(param):

msg = 'Drinking ' + param + ' glass'
print(msg)

param = 'empty'
glass = 'full'

drink_me(glass)

print('The glass is', glass)

The print statement prints
the value of the glass variable,
which is ‘full'.

ll'

glass = 'full'

glass

Again, nothing about the glass
variable has changed; it is still
labeling the value ‘full'.
Python 3.6.0 Shell

Drinking full glass
The glass is full

>>>

you are here 4  205

passing values to functions

So, when we pass a variable to a function, we’re
actually passing the value of that variable, not the
variable itself.

You got it. It’s common for coders to say, “When
I pass the variable x to the do_it function,” but that
is a coder colloquialism, and what they really mean
is “When I pass the value of the variable x to the do_it
function.” Think about it this way—when you make
a function call, each argument is evaluated before it is
passed to the function, so if we have:
x = 10

do_it('secret', 2.31, x)

The string 'secret' evaluates to the string
'secret' and is passed along, the number 2.31
evaluates to the floating-point number 2.31 and is
passed along, and finally, the variable x is evaluated
to the value 10 before it is passed along. So your
function doesn’t even know there is a variable x, it
just gets the value 10 and sets its corresponding
parameter to that value.

Now, we’re going to see that things get slightly more
complicated when we start working with objects, but
think of your arguments as always being passed to
your functions, not your variables.

206  Chapter 5

functions and abstraction

What about using global variables in functions?
Global variables are global, right? So they should be visible both inside and outside
functions, and they are. To use a global variable in a Python function, we first let
Python know we’re going to use a global variable with the global keyword.
greeting = 'Greetings'
def greet(name, message):
global greeting

Create a global variable greeting.
Tell our function we're
going to use a global.

print(greeting, name + '.', message)
greet('June', 'See you soon!')

Python 3.6.0 Shell

Greetings June. See you soon!
>>>

And use it.

Of course you can change the value of a global variable in
your function if you want to:
greeting = 'Greetings'
def greet(name, message):
global greeting

Python 3.6.0 Shell

Change the global
greeting to ‘Hi’.

greeting = 'Hi'

Hi June. See you soon!
Hi

>>>

print(greeting, name + '.', message)
greet('June', 'See you soon!')
print(greeting)

Print the value of greeting after
the call to greet.

We changed the value of
the global greeting within
our function to ‘Hi'.

Make sure you use the global keyword as a best practice
This all seems simple enough: to use a global variable in a function, just use the global keyword
to state your intent and use it at will. But watch out—this one has bitten many coders before you. If
you don’t use the global keyword you will still be able to read the values of global variables in your
functions, but if you try to change the value of this global, one of two things will happen: either it
will have been the first time you’ve used the variable in your function, and Python will assume it is a local variable
instead of a global, or you will have already read the value once, and Python will throw an UnboundLocalError.
Whenever you see this, look for cases where you are unintentionally mixing local and global variables.

you are here 4  207

local versus global variables

Tonight’s talk: LOCAL and GLOBAL variables answer
the question “Just what is GLOBAL’s problem,
anyway?”

The LOCAL variable

The GLOBAL variable

Look, I get I don’t have as big a job as Global, but
it’s a nice job—anytime I’m needed to hold a value
inside a function, I’m there. And as soon as that
function is done, I clean up and get out of the way.
I’ve got global scope. I’m everywhere; enough said.
Just because you’re everywhere doesn’t mean folks
should go out of their way to use you.
No one needs to go out of their way. Just declare a
variable outside a function, I’m there.
What I mean is, folks shouldn’t jump right into a
global variable when they don’t really need one.
What are you talking about—who doesn’t like a
global variable? Especially for simple programs,
they are actually one of the clearest and most
straightforward ways to store important values.
I don’t disagree with that for simple code, but
for more complex code apparently you haven’t
spent much time on the top coder discussion site
stackoverflow.com, because people say, “Avoid
globals!” all the time.

One problem is that if you use a lot of globals,
sooner or later you’re going to accidentally reuse the
same global name. Or imagine giving someone else
your code—they won’t even know without a lot of
study what globals you’re using. In fact, they might
already have code using the same global names.

208  Chapter 5

Oh, you can’t be serious; I’m like a basic part of
almost every programming language. You think all
those smart computer scientists would have put me
in if I wasn’t totally useful?

Oh sure, blame people being sloppy in their code
on me. Who are you going to blame next, the for
loop?

functions and abstraction

The LOCAL variable

The GLOBAL variable

Another issue is when you’re looking at a piece of
code that uses a global variable, you have no idea
what other parts of the code could be using and
changing that value, especially in a large program.
At least with a local variable you can typically see
all the code that deals with it; after all, it’s all right
there in the function.
What’s the big deal? I don’t see the issue.
Say you’re building some code for a chocolate
factory.

You’ve got a global Boolean variable controlling
whether the boiling vat’s drain is closed, and you
always want set it to True before pouring chocolate
into the vat.

Okaaaaaay...

Exactly; see how well that works? You have a global
variable and you just look to see if it is closed or not.
It’s in one place, easy to find.

Ah, but then in all your hundreds of lines of
chocolate factory code, a new hire comes along
and sets the variable to False, without checking
to see if there is chocolate in the vat. Can you say
inherently dangerous?
Well, you just need to train people better. And how
does a local variable solve that?
Actually, a really good solution to that problem is
objects. Which I think the readers will learn about
much later in the book.
And guess what, you still use local variables within
objects as well.
Have you considered looking for another line of
work? Once these readers get more advanced, they
may not need you anymore.

Objects, so I’ve heard.

Oh, well good for you.

Oh, I’m pretty sure they’ll be using me for quite a
while, at least until they learn...oh, nevermind.
you are here 4  209

using default values and keywords

Going further with parameters:
default values and keywords
Earlier we’d said you need to be careful about argument
order: if you don’t pass the correct arguments, in the
correct order, then all bets are off in terms of how a
function is going to operate. If you’ve got a function that
has speed and altitude parameters and you switch the
order when you pass your arguments, watch out!
You’ll find this is true across most programming languages;
to alleviate this potential ordering problem, Python
provides another, more flexible way to provide parameters.
With Python, parameters can have default values and
keywords, which allows you to pick and choose your
arguments and the order you want to supply them. You’ll
find parameter keywords and default values used in many
Python modules and libraries (not to mention you might
want to use keywords and default values in your own code).

You want a speed of 35,000 and
an altitude of 580? That could be
a bit of a problem. You might want
to check your argument order.

We’ll be seeing them used in
practice, later in the book,
as we explore more Python
modules.

How default parameter values work
Your function parameters can have default values. Let’s use a simpler version
of our greet function, without the global variable:

Here’s a normal,
everyday parameter
waiting for an argument.
def greet(name,
print('Hi',

And here’s a parameter that
doesn’t really care if it gets a
corresponding argument or not
because it has a default value.
message='You rule!'):
This is the default value for
name + '.', message)
message if the calling code
doesn’t pass an argument for
message.

Now that we’ve got a parameter with a default value, let’s
see how we use this:
without an

greet('John')

If we call greet
argument for message, then the
greet function will use the default
value for the message parameter.

greet('Jennifer', 'How are you today?')

And if we supply an argument for message,
then our greet function will gladly accept
that as the message argument.
210  Chapter 5

Python 3.6.0 Shell

Hi John. You rule!
Hi Jennifer. How are you today?
>>>

functions and abstraction

Always list your required
parameters first!
If you’re providing defaults for some of your parameters, you
need to be careful and put all your required parameters first. What’s
a required parameter? Well, if your function has a parameter
without a default value, then when your function is called, it must
supply an argument for that parameter, so it’s required. So say we
were to expand our greet function like this:

Here's a new parameter,
but it's required because
there is no default value.

def greet(name, message='You rule!', emoticon):
print('Hi', name + '.', message, emoticon)

What not to do!

Our new function definition is not
allowed. Python is complaining that
a required (non-default as Python
calls it) argument follows a default
argument.

So what’s the big deal? Why can’t we do this? The short story: if you
work through more complex examples, you’ll see the interpreter can
easily get into a situation where it has no idea which arguments go
with which parameters. While we encourage you to think through
such a situation, for now just remember to list all your required
(non-default in Python lingo) parameters before the parameters with
defaults. So, fixing our code above, we get:

Now the non-defaults
(required) parameters are first…

…followed by the
optional ones.

Python 3.6.0 Shell

File "defaults.py", line 1
def greet(name, message='You
rule!', emoticon):
^

SyntaxError: non-default argument
follows default argument

If you've been keeping track of
the parameter versus argument
definitions, you might be
asking why the interpreter is
calling these arguments and
not parameters. We have the
same question, but who are
we to argue with the Python
interpreter?

def greet(name, emoticon, message='You rule!'):
print('Hi', name + '.', message, emoticon)

you are here 4  211

more on keywords and default values

Using arguments with keywords
So far every time we’ve called a function, the arguments have
been positional. That is, the first argument is mapped to the first
parameter, the second argument to the second parameter, and so
on. You can also use the parameter names as keywords and specify
the arguments in a different order if you like.
To understand how this works, let’s take our newly expanded greet
function and call it with some keywords:

Python 3.6.0 Shell

Hi Jill. Where have you been?
thumbs up
>>>

greet(message='Where have you been?', name='Jill', emoticon='thumbs up')

To use a keyword, use the
parameter name, followed by
an equals sign and the value for
that argument.

Using keywords we can mix and match the
order of our arguments and even omit them if
they have defaults. Just make sure your calls
provide any required arguments before the
keywords arguments.

You can mix and match positional and keyword arguments as well.
greet('Betty', message='Yo!', emoticon=':)')

Here we're positionally specifying
the name parameter, but using
keywords for the others.

How to think about all these options
Parameter defaults and argument keywords are fairly specific to the
Python language. In some cases you’ll find that when you have a
function with many parameters, they are quite convenient in allowing
you to specify common default values for those who might use your
code. We’re not going to focus heavily on the use of either in this book,
but if you’re going to continue Python programming you will come
across them, and as we said, we’ll need knowldege of them later in the
book for some of Python’s modules we make use of.

212  Chapter 5

Python 3.6.0 Shell

Hi Betty. Yo! :)
>>>

functions and abstraction

Cement how those keyword arguments work into your brain before this chapter
ends by working through the code below and figuring out what it outputs. Write your
answer in the Python Shell window below.
def make_sundae(ice_cream='vanilla', sauce='chocolate', nuts=True,
banana=True, brownies=False, whipped_cream=True):
recipe = ice_cream + ' ice cream and ' + sauce + ' sauce '
if nuts:
recipe = recipe + 'with nuts and '
if banana:
recipe = recipe + 'a banana and '
if brownies:
recipe = recipe + 'a brownie and '
if not whipped_cream:
recipe = recipe + 'no '
recipe = recipe + 'whipped cream on top.'
return recipe
sundae = make_sundae()
print('One sundae coming up with', sundae)
sundae = make_sundae('chocolate')
print('One sundae coming up with', sundae)
sundae = make_sundae(sauce='caramel', whipped_cream=False, banana=False)
print('One sundae coming up with', sundae)
sundae = make_sundae(whipped_cream=False, banana=True,
brownies=True, ice_cream='peanut butter')
print('One sundae coming up with', sundae)
Python 3.6.0 Shell

Output goes here.
you are here 4  213

the none data type

I know the chapter is about to end, but I
just have one more question about functions.
If a function doesn’t have a return statement,
does it return anything?

It returns None. This is a bit of a can of worms to open
up right before the chapter ends, but it’s our fault for not
bringing it up sooner.
When you have a function that doesn’t explictly return
a value with a return statement, it returns the value
None. Not the string "None", but the value None. If you’re
saying “what on earth is that?” it’s one of those strange
computational objects, sort of like the empty string or an
empty list, or maybe more like True or False. In this case
it’s just a value that means no value or perhaps undefined.
You might also be asking if None is a value, what’s its type?
The type of None is NoneType. We know, now it all makes
sense, right? Here’s the trick: don’t think too hard about
None; it’s a value that signifies the lack of a value, and we’ll
see a bit here and there how this can be used. For now, here’s
more about the type:
Has a single value None.

NoneType
None

True

Uses no quotes around None.
First letter always
capitalized.
Expressions can evaluate to
None.

So add that to the stack of things you’ve got in the back of
your brain and we’ll return to this topic later.

You can file NoneType under
“truly weird types.”
214  Chapter 5

You'll find many programming languages haveand nil.
similar values with names like NULL, null,

functions and abstraction

The case of the attempted robbery not worth investigating

Five Minute
Mystery

Sherlock finished his phone call with the bumbling chief of police, Lestrade, and sat
down in front of the fireplace to resume reading the newspaper. Watson looked at him
expectantly.
“What?” said Sherlock, not looking up from the paper.
“Well? What did Lestrade have to say?” Watson asked.
“Oh, he said they found a bit of rogue code in the bank account where the
suspicious activity was taking place.”
“And?” Watson said, trying to hide his frustration.
“Lestrade emailed me the code, and I told him it wasn’t worth pursuing. The
criminal made a fatal flaw and will never be able to actually steal the money,”
Sherlock said.
“How do you know?” Watson asked.
“It’s obvious if you know where to look,” Sherlock exclaimed. “Now stop bothering me
with questions and let me finish this paper.”
With Sherlock absorbed in the latest news, Watson snuck a peek at Sherlock’s phone and
pulled up Lestrade’s email to look at the code.
balance = 10500

camera_on = True

This is the real, actual bank
balance in the account.

def steal(balance, amount):
global camera_on

camera_on = False

if (amount < balance):

balance = balance - amount

return amount

camera_on = True
proceeds = steal(balance, 1250)

print('Criminal: you stole', proceeds)

Why did Sherlock decide not to investigate the case? How could he
know that the criminal would never be able to steal the money just by
looking at the code? Is there one problem with the code? Or more?
you are here 4  215

review of functions and abstraction

 Functions give us a way to package up,
abstract, and reuse code.

 You can create local variables within a
function.

 A function has a name, zero or more
parameters, and a body.

 Local variables exist only as long as the
function execution does.

 You call or invoke a function and pass it
zero or more arguments.

 The places in your code where a variable
is visible define its scope.

 You can pass a Python function any valid
Python value.

 Variables that aren’t created inside
functions are called global variables.

 The number and order of arguments
in a function call needs to match the
parameters in the function. You can use
keyword arguments to specify a subset of
arguments in an arbitrary order.

 A function’s parameters are treated as
local variables in the function body.

 When a function is invoked, it assigns the
arguments to parameter variables and
then executes its code block.
 A function’s code block is also known as
the body of the function.

 When we name a parameter the same
name as a global variable, we say that
parameter is shadowing the global
variable.
 The global keyword is used within a
function to signify you’d like to refer to a
global variable in the function body.

 Functions can return values by using the
return statement.

 Abstracting code often makes your code
more readable, well structured, and
maintainable.

 To capture the value returned from a
function, simply assign the result of the
call to a variable.

 Abstracting code also allows you to focus
at a higher level, and to forget about the
low-level details a function implements.

 Functions can themselves call builtin functions or other functions you’ve
defined.

 Reworking code is often called refactoring.

 You can declare functions in any order as
long as they are defined before they are
called.

216  Chapter 5

 You can use parameter defaults to provide
default values for missing arguments.
 You can use parameter names as keyword
arguments when calling a function.

functions and abstraction

Coding cross
Get more functional as you do the crossword.
1

2

3

4
5
6

7
8

9
10

11
12
13
14
15
16
17

18

Across

Down

3. Solved the case.
5. Parameters can do this to global variables.
6. An online representation.
9. Another name for the function's block of
code.
10. When you call a function, you supply these.
11. Arguments are passed to these.
13. Reworking your code.
14. Functions are great for this.
16. Keyword for using globals in function.
17. Parameter names can be used as these.
18. Another word for calling a function.

1. Parameters are also these.
2. How long a variable lasts.
4. All functions have one.
5. We made some of these with lots of default
values.
7. You can use functions to do this.
8. We usually use this to match arguments to
parameters.
12. Variables declared in a function.
13. How you pass values back.
15. When we pass a variable to a function, we
really pass its ______.

you are here 4  217

exercise solution

Do a little analysis of the code below. How does it look? Choose as
many of the options below as you like, or write in your own analysis:
A. It’s the same code over and
over, so it seems very redundant.

D. Not the most readable code I’ve
ever seen.

B. It looks tedious to type in!

E. If we wanted to change how the
dogs bark, that would be a lot of
changes to make!
Should we be reusing the same
F. ___________________________
_____________________________
variables over and over for

C. It doesn’t seem to do a lot for
that much code.

different dogs?

Here’s a little connect-the-dots exercise for you. Using your
pencil, start at step 1 and draw a line from each step to the
next as the program executes. Feel free to annotate your lines
where necessary to explain what is happening in the code’s
execution. We already drew the first couple lines for you.

We start here.
Print, and then
proceed to the
next line.
ction't
n
u
f
e
n
ss thut we do
e
c
o
r
We p inition, b
def cute it.
exe

1

print('Get those dogs ready')

2

def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof') bark('Codie', 40)

a
b

3

bark('Codie', 40)

4

print("Okay, we're all done")

SOLUTION CONTINUES ON NEXT PAGE...
218  Chapter 5

functions and abstraction

o
rk, s 2.
a
b
l
l
we ca o step
Thenre back t
we'
Then we execute the
function’s body with
steps 2a and 2b…

…before returning at
from the function
3 and then executing
step 4.

1

print('Get those dogs ready')

2

def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof') bark('Codie', 40)

a
b

3

bark('Codie', 40)

4

print("Okay, we're all done")

1

print('Get those dogs ready')

2

def bark(name, weight):
if weight > 20:
print(name, 'says WOOF WOOF')
else:
print(name, 'says woof woof') bark('Codie', 40)

a
b

3

bark('Codie', 40)

4

print("Okay, we're all done")

We’ve got some more calls to bark below. Next to each call, write what you think
the output should be, or if you think the code will cause an error. Here’s our solution.

bark('Speedy', 20)

Speedy says woof woof
Barnaby says woof woof
Our bark function doesn’t check to make sure dog weights are
greater than 0. So this works because -1 is less than 20.

bark("Barnaby", -1)

bark('Scottie', 0, 0)
bark('Lady', "20")

ERROR, bark() takes 2 positional arguments but 3 were given.
ERROR, ‘>' not supported between instances of ‘str' and ‘int'

bark('Spot', 10)

Spot says woof woof

bark('Rover', 21)

Rover says WOOF WOOF

you are here 4  219

exercise solution

Code Magnets Solution

,

This working code is all scrambled up on the fridge.
Can you reconstruct the code snippets to make a
working program that produces the output listed
below? Notice, there may be some extra code on the
fridge, so you may not use all the magnets.

,

,

:
,

(

def

how_should_I_get_there
if miles > 120.0:

print('Take a plane')

(

miles

)

:

The function takes miles
as a parameter and
then...

elif miles >= 2.0:

print('Take a car')

else:

print('Walk')

how_should_I_get_there(800.3)
how_should_I_get_there(2.0)
how_should_I_get_there(.5)

Python 3.6.0 Shell
Take a plane
Take a car
Walk

220  Chapter 5

…checks the distance to
see if a plane, a car, or
walking is appropriate.

kilometer

functions and abstraction

Get some practice to make return values sink in. Compute the
return value of each function call.
get_bark(20)
make_greeting('Speedy')
compute(2, 3)
compute(11, 3)

‘woof woof’
‘Hi Speedy!’

5
10

False
allow_access('Dr Evil') True
allow_access('Codie')

Let’s make sure we can recognize each type of variable. Annotate this code by
identifying local and global variables as well as any parameters. Here’s our solution.

local variable
parameter
global variable

def drink_me(param):

parameter

msg = 'Drinking ' + param + ' glass'
print(msg)

param = 'empty'
glass = 'full'

drink_me(glass)

local variable

global variable

print('The glass is', glass)

you are here 4  221

exercise solution

Cement how those keyword arguments work into your brain before this chapter
ends by working through the code below and figuring out what it outputs. Write your
answer in the Python Shell window below.
def make_sundae(ice_cream='vanilla', sauce='chocolate', nuts=True,
banana=True, brownies=False, whipped_cream=True):
recipe = ice_cream + ' ice cream and ' + sauce + ' sauce '
if nuts:
recipe = recipe + 'with nuts and '
if banana:
recipe = recipe + 'a banana and '
if brownies:
recipe = recipe + 'a brownie and '
if not whipped_cream:
recipe = recipe + 'no '
recipe = recipe + 'whipped cream on top.'
return recipe
sundae = make_sundae()
print('One sundae coming up with', sundae)

Did you notice this
code has a bug if
there are no nuts?

sundae = make_sundae('chocolate')
print('One sundae coming up with', sundae)

sundae = make_sundae(sauce='caramel', whipped_cream=False, banana=False)
print('One sundae coming up with', sundae)
sundae = make_sundae(whipped_cream=False, banana=True,
brownies=True, ice_cream='peanut butter')
print('One sundae coming up with', sundae)
Python 3.6.0 Shell

One sundae coming up with vanilla ice cream and chocolate sauce with nuts and a banana and whipped
cream on top.
One sundae coming up with chocolate ice cream and chocolate sauce with nuts and a banana and whipped
cream on top.
One sundae coming up with vanilla ice cream and caramel sauce with nuts and no whipped cream on top.
One sundae coming up with peanut butter ice cream and chocolate sauce with nuts and a banana and a
brownie and no whipped cream on top.
>>>

Here’s our output.

222  Chapter 5

functions and abstraction

Five Minute
Mystery
Solution

balance is a global variable...

balance = 10500

camera_on = True

...but it’s shadowed by
this parameter.

def steal(balance, amount):
global camera_on

camera_on = False

So when you change the
balance in the function steal,
you’re not changing the actual
bank balance!

if (amount < balance):

We’re returning the
amount stolen...
...but we’re not using it to
update the real balance in
the account. So the balance
of the bank account is the
same as it was originally.

balance = balance - amount

return amount

camera_on = True

proceeds = steal(balance, 1250)

print('Criminal: you stole', proceeds)

The criminal thinks he stole
the money, but he didn’t!

And, in addition to not actually stealing any money, the criminal forgets to turn the camera back on,
which is a dead giveaway to the police that something nefarious is going on. Remember, when you return
from a function, the function stops executing, so any lines of code after the return are ignored!
1
3

S

H E

R

A

V

A

L O C K
T

8

P

10

O

A

S

11

I

14

P

A

5

M
E

V

D

A

A

R G U M E

L
R E
S
K

C
D O W
P
E

7

A

9

B O D Y

N

T

S

T

E

R S

I

I

17

H A

N

B

N

S
U

T
C O D E

R

N

L

A

S

4

C

6

Coding
Cross
Solution

2

L

T
R A M E
13

R E

U S

A
F

A

E

L

C

T O R

T

C

T

16

U
E

12

G

A
L O B

Y W O R D S
N

3. Solved the case. [SHERLOCK]
5. Parameters can do this to global variables.
[SHADOW]
6. An online representation. [AVATAR]

V
A

L

L
18

I

Across

15

U
N V O K

E

Down
1. Parameters are also
these.
you
are here
[LOCALVARIABLES]
2. How long a variable lasts. [SCOPE]
4. All functions have one. [NAME]

4  223

4, part 2, sorting and nested iteration

in which we return to the lists and add some superpowers

Putting Some Order
in Your Data

Sometimes the default ordering of your data
doesn’t cut it. You’ve got that list of high scores on 80s arcade
games, but you really need it sorted alphabetically by game name. Then
there’s that list of the number of times your coworkers have stabbed you
in the back—it would be nice to know who’s at the top of that list. To do
that, though, we need to learn how to sort data, and to do that we’ll need
to explore some algorithms that are a little more involved than the ones
we’ve seen so far. We’re also going to have to explore how nested loops
work as well as think a little about the efficiency of the code we’re writing.
Come on, let’s take that computational thinking up a level!
this is a new chapter   225

another bubbles-r-us challenge

I’m back! My bubbleologists did
such a great job in Chapter 4 that I’d love to
give out some awards. Check out my idea below; I
just need for you to write some code to generate one
more report. After all the code you’ve written, I’m
sure this will be a walk in the park for you.

Bubbles-R-Us

the
give awards to
to
ve
lo
d
I’
g,
st
in
li
a of the
e th
e
m
or
m
ve
e
gi
on
u
yo
ed
ne
an
C
I just
e top solutionsns.in descending order by their
inventors ofttbhub
io
t
lu
ble so
ow.
top five bese? Something like the report bel
bubble scor
s CEO
- Bubbles-R-U
Top Bubble Solutions
score: 68
1) Bubble solution #10
score: 60
2) Bubble solution #12
re: 57
3) Bubble solution #2 sco
score: 50
4) Bubble solution #31
re: 34
5) Bubble solution #3 sco

is
real scores; tanhist.
't
en
ar
e
es
h
t
e
Not
ion of what I w
just an illustrat
Thanks!!

226  Chapter 4B

sorting and nested iteration

We just need to sort
them? That doesn’t
sound too bad.

Cubicle conversation

Frank: Well, it doesn’t sound bad, but how do we do it?
Judy: We’ve been coming up with some great algorithms, so
I’m sure we can tackle sorting.
Joe: Actually, sorting algorithms are a topic some computer
scientists spend their entire careers researching. They’re not
something we should “come up with,” but rather we need
to research the existing algorithms and choose one that is
appropriate for our problem.

Frank

Judy

Joe

Frank: It sounds like we’re going to be researching for a
while. Should we postpone the softball game this afternoon?

Joe: I don’t think so, Frank. I saw this coming and I’ve already taken a look.
Frank: Oh, why didn’t you say so? What kind of sort are we going to use?
Joe: Bubble sort.
Judy, Frank (chuckling): Very funny, Joe. We know we’re sorting bubbles. Do you have a sort
that is appropriate or not?
Joe: I’m actually being completely serious. We’re going to use a bubble sort. It’s not the most
efficient sorting algorithm ever invented, but it is one of the easiest to understand.
Frank: I don’t get it; was it invented by one of our competitors?
Joe: Oh, you mean the name. It’s called bubble sort because as you run the algorithm, the larger
(or smaller) items in the list tend to “bubble up” to one end. You’ll see how that looks when we
start implementing it.
Judy: Joe, you seem to have done your homework, so we’ll take your lead on this.
Joe: Okay, let’s do it.
Judy: That reminds me, I told Greg to get on this too. I didn’t realize you’d already done all this
research. I need to remember to tell him not to bother. Don’t let me forget!

you are here 4  227

how bubble sort works

Understanding bubble sort
We’re going to look at some pseudocode for bubble sort shortly, but before we
do, let’s get an intuitive feel for how this algorithm works. To do that, let’s sort a
list of numbers:

[6, 2, 5, 3, 9]

Here’s our
unsorted list.

We typically think of sorting lists into ascending order, so we’re expecting our list
to look like this after the bubble sort algorithm has finished:

[2, 3, 5, 6, 9]

The same list sorted in
ascending order.

One thing to know up front: bubble sort works by making a number of passes over a
list. As you’ll see, in each pass, if we end up swapping any values in the list, we’ll need
to make another pass. When no values are swapped in a pass, we’re done. Keep that in
mind, as it’s key to how the algorithm works.

Starting with pass 1
We begin by comparing the first and second items (the items at index
0 and 1). If the first item is larger than the second, we swap them:

Next we compare the values of index 1 and index 2. If the first value,
6, is larger than the second value, 5 (and it is), we swap them.

[6, 2, 5, 3, 9]

Compare these. The
first is larger, so we
need to swap them.
[2, 6, 5, 3, 9]

Swap.
Moving along, we compare the next values at index 2 and index 3,
and once again, the first is larger. So we swap.

Next we compare the value at index 3 to the value at index 4, and it is
not larger. So, we do nothing and pass 1 is complete.

Notice the value 6 used to be the first in the list, and it is
slowly making its way (bubbling) toward the end of the list.
At this point we’re done with pass 1, but we swapped some values, so
we need to do another pass. On to pass 2!

228  Chapter 4B

[2, 5, 6, 3, 9]

Swap.
[2, 5, 3, 6, 9]

No change.
[2, 5, 3, 6, 9]

sorting and nested iteration

Pass 2
On pass 2 we start over, comparing the values of index 0 and index 1;
index 1 has a larger value, so we don’t need to swap values.

Moving on, we compare the values of index 1 and index 2; in this case,
the index 2 value is larger than index 3, so we swap values.

[2, 5, 3, 6, 9]

No change.
[2, 5, 3, 6, 9]

Swap.
You’re staring to understand the method now: we now compare the
values of index 2 and index 3, and the value 5 is not larger than 6, so
we do not swap values.

Moving on, we compare the 3rd and 4th index values, and the 4th
value is larger, so we do not need to swap.

Now pass 2 is over, but we swapped index 2 and 3 during this pass, so
we need to do another pass. On to pass 3.

Pass 3
On pass 3 we start over once again, comparing index 0’s value to
index 1’s; index 1 has a larger value, so we don’t need to swap values.

Now we compare the values of index 1 and 2; index 2’s value is larger,
so we don’t need to swap.

[2, 3, 5, 6, 9]

No change.
[2, 3, 5, 6, 9]

No change.
[2, 3, 5, 6, 9]

[2, 3, 5, 6, 9]

No change.
[2, 3, 5, 6, 9]

No change.

Moving on we compare the values of index 2 and 3; index 3’s value is
larger, so we don’t need to swap.

[2, 3, 5, 6, 9]

Likewise, we compare the values of index 3 and 4; index 4’s value is
larger, so we don’t need to swap.

[2, 3, 5, 6, 9]

Now pass 3 is over, and we didn’t swap any values. So our
algorithm is complete and we have a sorted list!

No change.
[2, 3, 5, 6, 9]

you are here 4  229

some practice with bubble sort

We’re not asking you to write
any code, we’re just asking
you to use the bubble sort
algorithm to sort this list.

Hopefully you now have a good feel for how to apply bubble sort’s
algorithm. Let’s practice. Take the list below and bubble sort it, showing
your work just like we did on the previous two pages. This will help to lock
in your understanding. If you get stuck, take a peek at our solution at the
end of the chapter.

['coconut', 'strawberry', 'banana', 'pineapple']

Pass 1

230  Chapter 4B

Start pass 1 here.

Here's the list. Don't
let the strings throw
you; just compare them
alphabetically (what a
computer scientist would
call lexicographically,
otherwise known as
dictionary order).

sorting and nested iteration

Some bubble sort pseudocode
Now that you have an intutive feel for the bubble sort, let’s actually look at some
pseudocode for the algorithm. You’ve already got all the knowledge you need
to understand this code; it uses basic Boolean logic and some loops. However,
it does use loops in a more complex way than you’ve seen before by using two
loops together—what we call a nested loop.
Nested loops can make you stop and think the first time you see one, but here’s
the trick: you just mentally did a nested loop when you worked through the last
Sharpen exercise (as well as when we walked through sorting the list of numbers
together). The outer loop represents each pass in the algorithm. The inner
loop goes through each item in the list and performs the comparisons (and any
needed swapping). So with that in mind, let’s have a look:

We’re going to use a
while loop that loops as
long as swapped is True.
Think of the while loop
as performing each pass.
Each time through the
loop is another pass.

Let’s put all that
Chapter 5 function
knowledge to work!
We’re going to have a function, bubble_sort,
which takes a list as a parameter.
And we’re going to have a
DEFINE a function bubble_sort(list):
variable to track if we’ve
swapped any values in the
DECLARE a variable swapped and set to True.
current pass. We set this
to True initially to kick off
WHILE swapped:
the first pass.
SET swapped to False.
The first thing we
FOR variable i in range(0, len(list)-1)
do in the loop is set
IF list[i] > list[i+1]:
swapped to False.
DECLARE a variable temp and set to list[i].
SET list[i] to list[i+1]
SET list[i+1] to temp

Pseudocode can take many
forms. This form is closer to
code, but still not quite code,
at least not Python.

SET swapped to True.

If we ever swap any values, we set swapped
to True, meaning we’re going to have to do
another pass when the for loop finishes.

In the for loop, you
iterate through each
item in the list (except
the last item) and
compare it to the next.
If it’s greater, then you
swap the values.

This pseudocode sorts a list in ascending order. What change would you need to make to sort in
descending order?

you are here 4  231

understanding the pseudocode

What do you
think of the pseudocode;
does it make sense?

Cubicle conversation
continued...

Judy: I think so, if I understand the two
loops.
Joe: Right, so there is a while loop and a
for loop that runs within it.
Judy: Okay, so stepping through this—the
outer while statement loops until the
swapped variable has the value False.
Joe: Yes, we go one time through the

while loop for each pass we need to make

over the list.

Judy: But in all our examples wasn’t that
always three times?
Joe: Oh, that was a total coincidence; the algorithm can have any number of passes. Actually, that
isn’t quite accurate—in the worst case, it has the same number of passes as there are items in the list.
Judy: So if I have 100 items, the while loop could loop 100 times?
Joe: Yup.
Judy: Why so many?
Joe: That’s the worst case; if the list is in totally backward order, it takes that many passes.
Judy: Got it. Okay, now the inner loop, the for loop. What’s happening there?
Joe: The for loop walks through every item and compares it to the next item. If the first item is of
greater value, then we swap the two.
Judy: So the for loop is also iterating through all items in the list as well, only not in the worst case,
but always.
Joe: Technically the number of items minus one, but yes.
Judy: For a big list that is a lot of iterating.
Joe: It sure is. If you use your example of 100 items, the worst case is roughly 100 passes comparing
100 items in each pass. So we’re talking 100 * 100 = 10,000 comparisons.
Judy: Whoa!
Joe: Yes, bubble sort is known more for its simplicity than its efficiency. Why do you think so many
people spend their time trying to create fast sorting algorithms? But our lists are very small, so it’s
really not a problem in our case, and a perfectly acceptable way for us to sort.
Judy: Okay, and the only other thing happening in the for loop is that if we swapped some values,
we have to set the swapped variable back to True, which means we’ll do another pass.
Joe: Bingo.
232  Chapter 4B

sorting and nested iteration

BE the
Interpreter
For these bits of code, your job is to
play like you’re the Python interpreter.
Step through each bit of code and evaluate
it (in your brain). After you’ve done the
exercises, look at the end of the chapter to
see if you got it right.

for i in
range(0,
4):
for j in
range(0,
4):
print(i
* j)

:
'lion', 'tiger', 'bobcat']
for word in ['ox', 'cat',
for i in range(2, 7):
letters = len(word)
if (letters % i) == 0:
print(i, word)

Remember modulus is like
finding the remainder of a
division. So, 4%2 would be 0,
while 4%3 would be 1.

full = False
donations = []
full_load = 45
toys = ['robot', 'doll',
'ball', 'slinky']
while not full:
for toy in toys:

donations.append(toy)
size = len(donations)
if (size >= full_load):
full = True

print('Full with', len(do
nations), 'toys')
print(donations)

you are here 4  233

implementing bubble sort

Implementing bubble sort in Python
Our pseudocode is fairly close to code, and translating it into Python is
straightforward. Let’s take a look:

Here’s our function,
which takes a
Python list.

def bubble_sort(scores):
swapped = True

We set the swapped variable to True
to kick off the first pass.
while swapped:
We make our passes with the while
swapped = False
loop as long as swapped is True…
Our nested loop: a
for i in range(0, len(scores)-1):
…and then step through the entire list
for loop within a
if scores[i] > scores[i+1]:
comparing and swapping values where needed.
while loop
temp = scores[i]
Just like the variable swap code
scores[i] = scores[i+1]
we looked at in Chapter 2!
scores[i+1] = temp
swapped = True

Note there is nothing to return because
we've swapped the actual values of the list.
In other words, we've changed the original
list to be in sorted order.

A Test Drive
scores = [60,
58,
34,
69,
46,
41,

Let’s start a new file called sort.py. Copy the code above into it, and add the test
code below:
50,
50,
55,
64,
31,
53,

60,
52,
51,
66,
57,
55,

bubble_sort(scores)
print(scores)

58,
54,
52,
55,
52,
61,

54,
48,
44,
52,
44,
51,

54,
69,
51,
61,
18,
44]

Python's > comparison works with
strings too, so we can also have the
sort function sort our smoothies.

Python 3.6.0 Shell
[18, 31, 34, 41, 44, 44, 44, 46, 48, 50,
50, 51, 51, 51, 52, 52, 52, 52, 53, 54,
54, 54, 55, 55, 55, 57, 58, 58, 60, 60,
61, 61, 64, 66, 69, 69]
['banana', 'coconut', 'pineapple',
'strawberry']
>>>

smoothies = ['coconut', 'strawberry', 'banana', 'pineapple']
bubble_sort(smoothies)
print(smoothies)

234  Chapter 4B

Nice, those look
sorted!

sorting and nested iteration

A Test Drive

We’d like to have the solutions with the highest bubble scores first (in other words
we want descending order, not ascending). We can do that by simply changing the
comparison in the sort function from a > to a < comparison. Make the change
and do another test drive.
There we go; sort in

def bubble_sort(scores):
swapped = True

This is all you need to change
to have a descending sort.
Why does this work?

while swapped:
swapped = False
for i in range(0, len(scores)-1):
if scores[i] < scores[i+1]:
temp = scores[i]
scores[i] = scores[i+1]
scores[i+1] = temp
swapped = True

descending bubble score
order.

Python 3.6.0 Shell
[69, 69, 66, 64, 61, 61, 60, 60, 58, 58,
57, 55, 55, 55, 54, 54, 54, 53, 52, 52,
52, 52, 51, 51, 51, 50, 50, 48, 46, 44,
44, 44, 41, 34, 31, 18]
['strawberry', 'pineapple', 'coconut',
'banana']
>>>

I think we’re almost there; we just
need to generate the report with
the top 5 solutions, scores along with
their number.

Frank: Yay, Joe wrote all the sort code! But something is missing. We’re sorting the
list of scores, but we have no idea what the original index of the score was, so how do
we know the score number? We need that for the report.
Judy: That’s a good point. How can we do that?
Frank: Well, Joe may be the rock star here, but I had an idea. What if we make
another list, a parallel list called solutions_numbers, where each value in the list
is the same as its index, like [0, 1, 2, 3, ..., 35]? Then, when we sort the
scores, we sort this list exactly the same way. So, at the end, each score number will
be in the same relative position as its corresponding score.
Judy: How do we create a list like that where the values match the indices?
Frank: Remember, you can use range and list together to do that. Like this:

Get the length of the list.
Create a range from 0 to the
length of the list (minus 1) and then
use the list function to convert the
understand your direction.
range into a list [0, 1, 2, ... ].

number_of_scores = len(scores)
solution_numbers = list(range(number_of_scores))

Judy: Interesting. Okay, I think I sort of

Frank: Some things are actually easier to explain over code. Here, check this out...
you are here 4  235

sorting the bubble solution numbers too

Serious Coding

Computing bubble solution numbers
Frank was right. We have a way to sort the bubble scores, but
when we do sort them we lose the identifying number of that
score (because we’ve always used the index of the solution
as the bubble number). So what we’re going to do is create
a second list that contains each solution’s corresponding
bubble number, and then, when we sort the solution’s scores,
we’re going to, in sync, sort the bubble numbers list the same
exact way. Let’s check out his code:

50,
50,
55,
64,
31,
53,

60,
52,
51,
66,
57,
55,

58,
54,
52,
55,
52,
61,

54,
48,
44,
52,
44,
51,

Everything else works
exactly the same as
before...

…except, when we swap two values of
the score list, we swap the same two
values in the numbers list.
If you think this looks like
duplicated code, it is. We’ll find out
in the next chapter how to remove
duplicate code.

54,
69,
51,
61,
18,
44]

number_of_scores = len(scores)
solution_numbers = list(range(number_of_scores))
bubble_sort(scores, solution_numbers)

236  Chapter 4B

Evalutes to the integer
length of scores: 36

Evaluates to the list
containing the numbers 0
through 35

def bubble_sort(scores, numbers):
swapped = True

scores = [60,
58,
34,
69,
46,
41,

list(range(number_of_scores))

Evaluates to the range 0
through 35

First, we’re going to accept two lists
in our bubble_sort function, the scores
and the corresponding bubble numbers.

while swapped:
swapped = False
for i in range(0, len(scores)-1):
if scores[i] < scores[i+1]:
temp = scores[i]
scores[i] = scores[i+1]
scores[i+1] = temp
temp = numbers[i]
numbers[i] = numbers[i+1]
numbers[i+1] = temp
swapped = True

How do you go about evaluating an
expression like the one below? Evaluate
from the inside out:

Here we’re just creating the
solution_numbers list, which holds
the number of each solution (and
corresponds to its original index in
the scores list).

Now when we call sort, we pass both lists.

sorting and nested iteration

Now our code produces two lists, one with the
scores sorted and another with the corresponding
solution numbers. Write the code to generate the
report given these two lists.

Bubbles-R-Us

Top Bubble Solutions

1) Bubble solution #10 score: 68
2) Bubble solution #12 score: 60
3) Bubble solution #2 score: 57

4) Bubble solution #31 score: 50
5) Bubble solution #3 score: 34

Remember the report looks like this;
also remember this is not the actual
data your report will generate, just a
mockup from the CEO.

A Test Drive

Take the code from the last Sharpen exercise (repeated below) and add to your
sort.py file, replacing the previous test code. Give it a test drive.

print('Top Bubble Solutions')
for i in range(0,5):
print(str(i+1) + ')',
'Bubble solution #' + str(solution_numbers[i]),
'score:', scores[i])

Exactly like the CEO requested. Boy,
is he going to be happy!

Python 3.6.0 Shell
Top Bubble Solutions
1) Bubble solution #11
2) Bubble solution #18
3) Bubble solution #20
4) Bubble solution #19
5) Bubble solution #23
>>>

score:
score:
score:
score:
score:

69
69
66
64
61

you are here 4  237

review of sorting

Not only am I giving my top
five bubble solution engineers awards,
I’m giving YOU an award. We could not
have made Bubbles-R-Us successful without
all the business intelligence your code
provided!

Nice job.
It’s been a short but challenging
chapter. You’ve had to wrap your head
around lots of new concepts. Give
your brain some time to let it all sink
in. Of course make some more of that
bubble solution #11; that’s a great way
to relax and do something completely
different. Get some rest too, and after
that, maybe even quickly review this
chapter before you move on.
Of course you aren’t quite done yet.
You’ve got a brain twister ahead, and
the bullet points and the crossword...

 There are many sorting algorithms with
various tradeoffs in complexity and space/
time considerations.
 Bubble sort is a simple algorithm that
makes passes through a list, comparing
and swapping values as it goes.
 Bubble sort is complete when a pass
through the list finds no items that are out
of order.

238  Chapter 4B

Top Bubble
Software Maker

 Most languages and libraries provide sort
functionality.
 When we have a loop within a loop, we call
it a nested loop.
 Nested loops often increase the runtime
and complexity of an algorithm.
 It’s worth studying sorting algorithms
and the sort algorithms provided by your
language’s libraries, which, speaking of...

sorting and nested iteration

Uh, you’re not going to believe
this. I just went to tell Greg we
had solved the problem and he’d
beat us to it. In fact he hardly had to
write any code; he did it using Python’s
built-in sort!

You’ll find sort provided by most languages.
Yes, you’ll find many modern languages and many libraries
provide sort functionality. Given that, there’s ususally no
reason to write your own sort—not only would you be
reinventing the wheel, but off-the-shelf sort algorithms tend
to be more sophisticated (and more efficient) than bubble
sort and take significant time and effort to implement.
So, unless your passion turns out to be sorting algorithms,
wouldn’t you rather be working on your probem at hand?
Our time on the last 10 pages is not wasted, though,
because the techniques you just learned in implementing
bubble sort, like using nested loops, are core to
implementing many algorithms. And, should you need to
study sorting or a related algorithm, bubble sort is where
most folks start.
Built-in
sort! Maybe
you could have let
us know 10 pages
back?

Back to Python, you can sort a list by simply calling sort:
scores.sort()

And there are lots of ways to customize Python’s sort to your
needs, but you’ll want to learn a few more programming
concepts to make full use of that.
Finally, if you’re interested in sorting, you’ll want to study up
on the many sorting algorithms and their advantages and
disadvantages. With every algorithm you’ll encounter, there
are tradeoffs in time (how long it takes) and space (the amount
of memory or resources it uses). Some rather well-known
sorts are insertion sort, merge sort, and quicksort, just to
name a few. Oh, and Python interally uses Timsort, a hybrid
of merge and insertion sort. Here’s a good place to start your
studies: https://en.wikipedia.org/wiki/Sorting_algorithm.

you are here 4  239

a deep thinking exercise

ter 4, part 2...

Try a deep thinking exercise before you finish off Chap

The Thing-A-Ma-Jig2

The Thing-A-Ma-Jig is a crazy contraption—it clanks and clunks and even thunks,
Whether you give it lists or strings, it still does things. But how exactly does it
work? Can you uncrack the code and find its quirks?
characters = 'taco'

All we changed was the
list—now it's a string. But
the code still works. How?

output = ''
length = len(characters)
i = 0
while (i < length):
output = output + characters[i]
i = i + 1
length = length * -1
i = -2
while (i >= length):
output = output + characters[i]
i = i - 1
print(output)

Try these as a replacement for characters’ value above:
characters = ‘amanaplanac’
or
characters = ‘wasitar'

How does this code work on lists AND strings?
We'll leave you with that deep question, and we'll
uncover some answers in Chapter 6 (and beyond).

240  Chapter 4B

sorting and nested iteration

Hopefully you now have a good feel for how to apply bubble sort’s
algorithm. Let’s practice. Take the list below and bubble sort it, showing
your work just like we did on the previous two pages. This will help to lock
in your understanding.

['coconut', 'strawberry', 'banana', 'pineapple']

Pass 1
['coconut', 'strawberry', 'banana', 'pineapple']

No change.

['coconut', 'strawberry', 'banana', 'pineapple']

Swap.

['coconut', 'banana', 'strawberry', 'pineapple']

Swap.

['coconut', 'banana', 'pineapple', 'strawberry']

Pass 2
['coconut', 'banana', 'pineapple', 'strawberry']

Here's the list. Don't
let the strings throw
you; just compare them
alphabetically (what a
computer scientist would
call lexicographically,
otherwise known as
dictionary order).

Compare each index to the
next, iterating through the list.
Swap if the first value is > the
second.
We had swaps in pass 1, so
we need to do a pass 2.

Swap.

['banana', 'coconut', 'pineapple', 'strawberry']

No change.

['banana', 'coconut', 'pineapple', 'strawberry']

No change.

['banana', 'coconut', 'pineapple', 'strawberry']

We had swaps in pass 2, so
we need to do a pass 3.

Pass 3
['banana', 'coconut', 'pineapple', 'strawberry']

No change.

['banana', 'coconut', 'pineapple', 'strawberry']

No change.

['banana', 'coconut', 'pineapple', 'strawberry']

No change.

['banana', 'coconut', 'pineapple', 'strawberry']

We had no swaps in pass 3,
so we're done and the list
is sorted.

you are here 4  241

exercise solution

Python 3.6.0 Shell

BE the Interpreter
SOLUTION
For these bits of code, your job is to
play like you’re the Python interpreter.
Step through each bit of code and evaluate
it (in your brain). After you’ve done the
exercises, look at the end of the chapter to
see if you got it right.

(0,4):
n range
for i i
(0,4):
n range
i
j
r
o
f
* j)
print(i

:
'lion', 'tiger', 'bobcat']
for word in ['ox', 'cat',
for i in range(2, 7):
letters = len(word)
Python 3.6.0 Shell
if (letters % i) == 0:
2 ox
print(i, word)
3 cat
2 lion
4 lion
5 tiger
2 bobcat
3 bobcat
6 bobcat
>>>

0
0
0
0
0
1
2
3
0
2
4
6
0
3
6
9
>>>

full = False
[]
donations =
45
=
d
oa
full_l

nky']
'ball', 'sli
t', 'doll',
bo
ro
['
=
ys
to

Python 3.6.0 Shell

Full with 48 toys
['robot', 'doll', 'ball', 'slinky', 'robo
'ball', 'slinky', 'robot', 'doll', 'ball t', 'doll',
', 'slinky',
'robot', 'doll', 'ball', 'slinky',
'robot', 'doll',
'ball', 'slinky', 'robot', 'doll', 'ball
'robot', 'doll', 'ball', 'slinky', 'robo', 'slinky',
t', 'doll',
'ball', 'slinky', 'robot', 'doll
', 'ball', 'slinky',
'robot', 'doll', 'ball', 'slinky',
'robot', 'doll', 'ball',
'slinky', 'robot', 'doll', 'ball', 'slin
ky']
>>>

242  Chapter 4B

ll:
while not fu
in toys:
y
to
for
pend(toy)
donations.ap
nations)
size = len(do
full_load):
if (size >=
full = True

s')
tions), 'toy
th', len(dona
wi
l
ul
'F
t(
prin
ons)
print(donati

sorting and nested iteration

Now our code produces two lists, one
with the scores sorted and another with
the corresponding solution numbers.
Write the code to generate the report
given these two lists.

Bubbles-R-Us

Top Bubble Solutions

1) Bubble solution #10 score: 68
2) Bubble solution #12 score: 60
3) Bubble solution #2 score: 57

4) Bubble solution #31 score: 50
5) Bubble solution #3 score: 34

Print the heading.
print('Top Bubble Solutions')
Iterate through five times
for i in range(0,5):
for the top five scores.
print(str(i+1) + ')',
'Bubble solution #' + str(solution_numbers[i]),
'score:', scores[i])

For each line of output, print the final standing of
the solution with the value i plus 1,the solution number
from the solution_numbers list, and the score from the
scores list.

you are here 4  243

6 text, strings, and heuristics

Putting It All Together
Not so fast.
Sure, you know variables
and data types and
iteration and functions,
but can you put it all
together?

You’ve already got a lot of superpowers. N
 ow it’s time to
use them. In this chapter we’re going to integrate what we’ve learned so far,
bringing it all together to build some increasingy cool code. We’re also going
to keep adding to your knowledge and coding skills. More specifically, in this
chapter we’ll explore how to write code that grabs some text, slices it, dices
it, and then does a little data analysis on it. We’re going to find out what a
heuristic is too, and implement one. Get ready—this is an all-out, heads-down,
pedal-to-the-metal, serious coding chapter!

And by the end of the chapter
you'll realize how much you've
learned about coding!
this is a new chapter   245

intro to data sciences
Definitely
some sophisticated
writing in this book.

Welcome to the data sciences
Heard of data science? It’s all about extracting insight and
knowledge from data, and we’re about to get into the business. Our
data? Any text: news articles, blog posts, books, anything written by
an author. And what we’re going to do with all that text is tell the
world how readable it is. That is, could a 5th grader read it, or does it
take a Ph.D.? With our app, we’ll soon know.
Now to perform such an analysis, we’re going to have to dig deep;
we’re going to examine, as you’ll soon discover:
every sentence…
and every word…
and every syllable...
not to mention every single character of the text in question!
With all that analysis, we’re going to come up with a score that
maps to a reading level, from 5th grade all the way up to a college
graduate. Let’s dig in and see more precisely how this is going to
work...

246  Chapter 6

text, strings, and heuristics

How do you compute
something like readability?
Luckily someone else has already answered this question—namely,
the US Navy and Army, who have been testing a key formula for
years in the military to assess the readability of things like (we’re
assuming) tank training manuals. Here’s how it works: take the text
you want to analyze and run it through the following formula to get
back a reading ease score. We’ll talk in a second about what that score
means. Here’s the formula:

The formula makes use of three
different numbers we’re going
to need to compute.

206.835 – 1.015

(

total words
total sentences

We start with the total number
of words in the text, which is
used in two places in the formula.

)– 84.6 (

)

We’re not kidding; you can read up on
it here: https://en.wikipedia.org/wiki/
Flesch–Kincaid_readability_tests.
To give due credit, the formula
was originally developed by
Rudolph Flesch in 1948. Dr.
Flesch was an author who
earned a Ph.D. in English at
Columbia University.

We also need to compute the total
number of syllables in the text.

total syllables
total words

We also need to compute the
total number of sentences.

Once we’ve computed those values, the
formula just consists of standard
multiplication, division, and subtraction,
with a few floating-point values like 84.6.

If you're wondering where numbers like 206.835
come from, all we can say is years of research (see
the Wikipedia article above).
Say you were writing some text
for advertising; how would you
After we’ve used this formula to compute a score, we map the score to
want it to score?
the reading ease of the text using this chart:
Score

School Level

Notes

100.00–90.00

5th grade

Very easy to read. Easily understood by an average 11-year-old student.

90.0–80.0

6th grade

Easy to read. Conversational English for consumers.

80.0–70.0

7th grade

Fairly easy to read.

70.0–60.0

8th & 9th grade

Plain English. Easily understood by 13- to 15-year-olds.

60.0–50.0

10th–12th grade

Fairly difficult to read.

50.0–30.0

College

Difficult to read.

30.0–0.0

College Graduate Very difficult to read. Best understood by university graduates.

The higher the score, the easier your text is to read.

You'll find the original chart at
the Wikipedia URL above.
you are here 4  247

planning the implementation
One of the first rules for a guide in polite conversation, is to avoid

The game plan

political or religious discussions in general society. Such discussions
lead almost invariably to irritating differences of opinion, often to

At first glance computing readability just comes down to computing a
formula. But, if we look closer, it’s the key values that we need for that
formula where the computing comes in. That’s because, to complete
the formula, we need:

open quarrels, and a coolness of feeling which might have been avoided
by dropping the distasteful subject as soon as marked differences of
opinion arose. It is but one out of many that can discuss either
political or religious differences, with candor and judgment, and yet so
far control his language and temper as to avoid either giving or taking
offence.

One of the first rules for a guide in polite conversation, is to avoid

1

2

3

To compute the total number of words in our text: that
means we’ll have to take our text and break it down into
words and count them.
To compute the total number of sentences in our text: that
means we’ll have to take our text and break it down into its
individual sentences and count them.
To compute the total number of syllables in our text: that
means we’ll have to take each word, figure out how many
syllables it has, and then add up all the syllables in the
entire text.

political or religious discussions in general society. Such discussions
lead almost invariably to irritating differences of opinion, often to
open quarrels, and a coolness of feeling which might have been avoided
by dropping the distasteful subject as soon as marked differences of
opinion arose. It is but one out of many that can discuss either
political or religious differences, with candor and judgment, and yet so
far control his language and temper as to avoid either giving or taking
offence.

to irritating differences of
open quarrels, and a coolness
by dropping the distasteful
opinion arose. It is but one
differences, with candor and

Can you come up with an algorithm to compute the number of syllables in an
arbitrary word? Assume you don’t have a large dictionary at your disposal. Write
your algorithm here in pseudocode.

This is a difficult task, but do your best and think through the
problem. Come up with the best answer you can in five minutes or so.
248  Chapter 6

text, strings, and heuristics

Writing some pseudocode
We’re going to write some pseudocode to drive our implementation. For now, we’re going to keep the pseudocode
high level, and we’ll fill in a lot of the details as we progress. There isn’t a lot of tricky logic in this code (like a
game); rather, our task will be computing the numbers we need for the readability formula (which will nonetheless
give us plenty to do). Follow along with our thinking as we write the pseudocode. As we’ve said, there are many
forms of pseudocode; you’ll find this one a little more formal and closer to code.

This function expects to be passed the
text of the book or article...we’ll have
Let’s start out right and
We already know of quite a few
to figure out how to do that.
put this code in a function.
important values we’re going
to need to compute for our
formula. Let’s go ahead and
DEFINE a function compute_readability(text):
create local variables for those.
DECLARE a variable total_words and set to 0.
SETUP: here we’re
We’ll also create a variable,
DECLARE a variable total_sentences and set to 0.
declaring all the local
score, to hold the final score.
variables we’ll be using
DECLARE a variable total_syllables and set to 0.
We need to compute those
throughout the function.
DECLARE a variable score and set to 0.
values next, so let’s rely on a
function to compute each one.
ANALYSIS: we compute all
ASSIGN
variable
total_words
to
result
of
calling
the
function
count_words(text)
the values we need, and we’ll
ASSIGN variable total_sentences to result of calling the function count_sentences(text)
be diving into each one of
these separately.
ASSIGN variable total_syllables to result of calling the function count_syllables(text)
FORMULA: we have all the
values, so we can now use
the formula to compute the
score.
RESULTS: we’ve got a
score and we’re mapping
that to a reading level by
using the table two pages
back.

ASSIGN variable score to
206.835 - 1.015 * (total_words / total_sentences) - 84.6 * (total_syllables / total_words)
IF score >= 90.0:
PRINT ‘Reading level of 5th Grade’
ELIF scores >= 80.0:
PRINT ‘Reading level of 6th Grade’
ELIF scores >= 70.0:
PRINT ‘Reading level of 7th Grade’

With all the values
computed, we can evaluate
the reading ease score…
…and then figure out the
reading level.

ELIF scores >= 60.0:

Computing readability
seems like a generally
useful piece of code,
so let’s put it in a
function—we may be able
to reuse it later!

PRINT ‘Reading level of 8-9th Grade’
ELIF scores >= 50.0:
PRINT ‘Reading level of 10-12th Grade’
ELIF scores >= 30.0:

This is a big if/elif/else
statement that prints the
appropriate reading level based
on the computed score.

PRINT ‘Reading level of College Student’
ELSE:
PRINT ‘Reading level of College Graduate’

We’ll probably put this code in its own function.
you are here 4  249

inputting multiline text

We need some text to analyze
Before we get coding, we’re going to need some interesting text
to put through our analysis. Now the truth is, you can analyze any
text you want: blog posts, your own writing, news articles, books,
whatever—and half the fun of creating this code is analyzing your
favorite news outlets and writers. That said, as we build and test this
code, it’ll help if you use the same text we use, so we’re seeing the same
results. So let’s find some text we can all test together.
And why not put ourselves to the test? We’re going to use the first
couple pages of this book, which you’ll find in the file ch6/text.txt.

For testing purposes we’ll
put ourselves to the test
by using the text from
Remember, instructions for the book downloads are also in Chapter 1 of this book.
the introduction section of this book, but this is Chapter
6, so of course you’ve already downloaded the files.

How to get multiline text into Python
If you look at the file ch6/text.txt, you’ll see we have a big text file, but
how do we get that into Python? Well, you already know how to add text
to your code with strings:
text = 'The first thing that stands between you'
And we’re going to do the same with the text in ch6/text.txt. To do that,
we’re going to use a Python convention for entering strings that span
multiple lines by using triple quotes, like this:

ree quote characters,
Start your string with th
will do.
either single (') or double (")

Don't enter anything into
IDLE yet. We'll enter it
on the next page.

Then enter your entire string,
including new lines.

text = """The first thing that stands between you and writing your first, real,
piece of code, is learning the skill of breaking problems down into
achievable little actions that a computer can do for you."""

Obviously make sure your text
within the string doesn't have
any triple quotes, as unlikely
as that is.

After you’ve done this, use the string as you would any other; it’s just a
normal string with a bit more text in it. Now let’s get the text from the
first couple of pages of Chapter 1 into a string.
250  Chapter 6

And then end your string
with triple quotes as well.

text, strings, and heuristics

Remember, a
Python file is also
known as a module.
Take the text in the ch6/text.txt file and get it into a Python file. To do that, in IDLE,
create a new file and add the code below. For the Chapter 1 text, you’ll want to
open the text.txt file in IDLE as well, and then copy and paste the text into your
new file. When you’re done, save your file as ch1text.py.
Finally, run your code, and you should see the entire article printed to the shell.

Start with the string’s variable name and follow it with an equals sign, and then with
triple quotes. Next, paste in the text from the text.txt file.
text = """The first thing that
piece of code, is learning the
achievable little actions that
you and the computer will also
but we'll get to that topic in

stands between you and writing your first, real,
skill of breaking problems down into
a computer can do for you. Of course,
need to be speaking a common language,
just a bit.

Now breaking problems down into a number of steps may sound a new
skill, but its actually something you do every day. Let’s look at an
example, a simple one: say you wanted to break the activity of fishing
down into a simple set of instructions that you could hand to a robot,
who would do your fishing for you. Here’s our first attempt to do that,
check it out:
.
.
.

We're saving a few trees and omitting
some of the text here in the book.

You’re going to find these simple statements or instructions are the
first key to coding, in fact every App or software program you’ve ever
used has been nothing more than a (sometimes large) set of simple
instructions to the computer that tell it what to do."""

Don’t forget the
final triple quotes.

print(text)

Let’s print the text just to make sure
everything is working.

you are here 4  251

setting up the compute readability function

Setting up the function
To get started let’s go ahead and set up the compute_readability
function. Let’s then translate the setup portion of the pseudocode
into Python:

def compute_readability(text):
total_words = 0
total_sentences = 0
total_syllables = 0
score = 0

Just like in our pseudocode,
we have a function that takes
text as a parameter.
We also set up four
local variables that are
going to hold important
values in this function.

Go ahead and enter this code in a file called analyze.py.
Now let’s call compute_readability and pass it that text in ch1text.py.
But how? After all, it’s in another file. Well, remember any file with a .py
extension is a Python module, and also remember that in the past we’ve
imported Python modules into our code using the import statement.
So let’s import the ch1text.py file into our analyze.py file. Once we’ve done
that, we can access the variables and functions within the module by
prepending their names with the module name. Take a look:

We use import to include
the ch1text.py file.

Remember, a Python module is
just a file with a .py extension
and Python code inside.
We'll go into a lot more
detail on how modules
work in the next chapter.

import ch1text
def compute_readability(text):
total_words = 0
total_sentences = 0
We call compute_readability and pass
total_syllables = 0
string from the ch1text file.
text
score = 0
compute_readability(ch1text.text)

To access the variable named
text in the ch1text file, we
prepend it with the module
name, ch1text.
252  Chapter 6

And just as a bit of review,
note we’ve defined a function
called compute_readability
it the that has a parameter, text,
and sets up some local variables
And we’re calling that
function and passing it the
text variable from the
ch1txt module (in other
words, the ch1text.py file).

text, strings, and heuristics

A Test Drive
import ch1text

Let’s give this a quick test to make sure
everything is working. To do that, move the
print statement from your ch1text.py file
to your analyze file. You’ll want to place it
in the compute_readablility function.
The output should be exactly the same as
the test run of the ch1text.py file.

def compute_readability(text):
total_words = 0
total_sentences = 0
total_syllables = 0
score = 0
Add this to
print(text)

Python 3.6.0 Shell
into pond", or "pull in the fish." But also notice that other
instructions are a bit different because they depend on a condition,
like “is the bobber above or below water?". Instructions might also
direct the flow of the recipe, like "if you haven’t finished fishing,
then cycle back to the beginning and put another worm on the hook."
Or, how about a condition for stopping, as in “if you’re done” then go
home.
You’re going to find these simple statements or instructions are the
first key to coding, in fact every App or software program you’ve ever
used has been nothing more than a (sometimes large) set of simple
instructions to the computer that tell it what to do.
>>>

compute_readability(ch1text.text)

You should again see the
Chapter 1 text printed in
the shell.
DON’T FORGET to remove the print
statement from your ch1text.py file.

First up: we need the total
number of words in our text
Referring to our pseudocode, the first number we need to compute for
our readability formula is the total number of words in our text, and
to compute that you need to extend your coding knowledge a little bit.
You already know how to combine strings with concatenation, but what
you don’t know how to do is break them apart. Strings provide a handy
function called split that will split a string into words (which we
usually call substrings) and then place those substrings into a list.
Here’s how to use split:
1

Take any string with some
text and whitespace.

It looks like split does a good job
of separating our text into its
individual words.

Python 3.6.0 Shell
['I', 'heard', 'you', 'on',
'the', 'wireless', 'back',
'in', 'fifty', 'two']
>>>

lyrics = 'I heard you on the wireless back in fifty two'
words = lyrics.split()
2 And then call the
print(words)
string’s split function.
4

Let's print the list to see
what split did.

your
compute_readability
function.

3

The list holds all the words
(or substrings), each from
the original string.

Using whitespace (spaces, tabs, newlines) as a
separator, split breaks the string up into a number
of substrings, and places each one in the list.
you are here 4  253

figuring out the number of words

Cubicle conversation
Well, it looks like
we’re going to use the split
function to help figure out the
number of words?

Frank: I’m already lost. We want to count the number of
words, but we’re going to split up our text into a list first?
Joe: It’s a two-step process, Frank. We’re going to
essentially extract all the words into a list. After that we’ll
count how many words are in the list.
Frank: I understand the counting part, but how is the

split function giving us back words?

Judy: Frank, that’s what split does—it breaks a string up
into words, using whitespace as a separator.

Frank

Judy

Joe

Frank: Oh, so anytime it sees a space or tab or return, it
uses that to figure out how to break out the words.
Judy: Exactly.
Joe: And once we’ve split our string up into words and put
them in a list, we can easily count them.
Frank: Oh right, we can just use the len function for that,
correct?
Joe: That’s what I’m thinking.
Judy: Okay, guys, I think we have a plan: we’ll use the

split function to break the Chapter 1 text into words,

which we’ll get back in a list, and then we’ll call the built-in

len function to count how many words there are.

Frank: I like it!

254  Chapter 6

text, strings, and heuristics

Q:
A:

So split takes a string and breaks it up into words?

Close. Split takes a string and breaks it up into a number of
substrings. The way it knows how to break the string up is by treating
whitespace as a separator, or as a computer scientist would say, a
delimiter. In other words, if it sees a delimiter, like a space, a tab, or
a newline in your string, it knows that is a place it can split the string.
Note the resulting substrings don’t have to technically be words; they
could be dates, tabular formatted data, numbers, and so on.

Q:

What if my string has, say, commas rather than
whitespace separating the items?

A:

split allows you to specify a set of characters you want to use
to delimit your substrings. So you could set the comma character
as your delimiter. That said, split is not flexible enough to combine,
say, whitespace and commas. You’ll find all the details in the
documentation for the split function. A bit later we’ll be looking at how
to make use of Python’s documentation. You can also find a more
sophisticated way to match words in text using what are known as
regular expressions, which are discussed in the Appendix.

Q:

Help me with the syntax again: how does lyrics.split()
work exactly?

A:

For now just know that you can attach functions to data types,
and the string data type has a function called split. When you see
lyrics.split(), that’s saying use the function split that is associated with
strings. We’re going to be taking a close look at exactly how this
works later in the book.

Now that you know how to use the split function, let’s return to the compute_readability function
and take it a little further. In our pseudocode we were going to write a function count_words, but as it
turns out, with split we can handle the job of computing the total number of words with only two lines
of code, so let’s forgo the function. Go ahead and complete that code below, and then check with our
version in the solution at the end of the chapter. After that, give it a test run.

import ch1text
def compute_readability(text):
total_words = 0
total_sentences = 0
total_syllables = 0
score = 0

Let’s take our text
and split it into words.

words = text.split()
total_words = __________________
print(words)
print(total_words, 'words')
print(text)
compute_readability(ch1text.text)

Hint: Frank, Joe, and
Judy already figured
out how to do this.

Computing the total number of
words is one line of code. Finish
this one up.
Let’s print all the words, and
then the total count. And
let’s also remove the old print
statement.
you are here 4  255

more about split

I noticed in the shell output that we’re
not doing a perfect job of extracting
words. I have words like ‘book!,’ and
‘fire.’ and 'is,'.

Right. The split function, by default, splits up text only
using whitespace as a separator. And, while you can also pass
split a custom separator as a second parameter, unfortunately
split isn’t implemented in a way where you can easily tell it to
use “whitepace and commas and periods and semicolons and
exclamation points and question marks.” The result: we have
punctuation at the end of some of our words in the words list.
That’s okay, though, as it doesn’t really affect our count of
words. That said, it may cause some issues later in our coding,
but nothing we can’t handle. You’ll see in a bit...
Python 3.6.0 Shell
['The', 'first', 'thing', 'that', 'stands', 'between', 'you', 'and', 'writing', 'your',
'first,', 'real,', 'piece', 'of', 'code,', 'is', 'learning', 'the', 'skill', 'of', 'breaking',
'problems', 'down', 'into', 'achievable', 'little', 'actions', 'that', 'a', 'computer',
'can', 'do', 'for', 'you.', 'Of', 'course,', 'you', 'and', 'the', 'computer', 'will', 'also',
'need', 'to', 'be', 'speaking', 'a', 'common', 'language,', 'but', "we'll", 'get', 'to',
'that', 'topic', 'in', 'just', 'a', 'bit.', 'Now', 'breaking', 'problems', 'down', 'into',
'a', 'number', 'of', 'steps', 'may', 'sound', 'a', 'new', 'skill,', 'but', 'its', 'actually',
'something', 'you', 'do', 'every', 'day.', 'Let’s', 'look', 'at', 'an', 'example,', 'a',
'simple', 'one:', 'say', 'you', 'wanted', 'to', 'break', 'the', 'activity', 'of', 'fishing',
'down', 'into', 'a', 'simple', 'set', 'of', 'instructions', 'that', 'you', 'could', 'hand',
'to', 'a', 'robot,', 'who', 'would', 'do', 'your', 'fishing', 'for', 'you.', 'Here’s', 'our',
'first', 'attempt', 'to', 'do', 'that,', 'check', 'it', 'out:', 'You', 'can', 'think',
'of', 'these', 'statements', 'as', 'a', 'nice', 'recipe', 'for', 'fishing.', 'Like', 'any',
'recipe,', 'this', 'one', 'provides', 'a', 'set', 'of', 'steps,', 'that', 'when', 'followed',
'in', 'order,', 'will', 'produce', 'some', 'result', 'or', 'outcome', 'in', 'our', 'case,',
'hopefully,', 'catching', 'some', 'fish.', 'Notice', 'that', 'most', 'steps', 'consists',
'of', 'simple', 'instruction,', 'like', '"cast', 'line', 'into', 'pond",', 'or', '"pull',
'in', 'the', 'fish."', 'But', 'also', 'notice', 'that', 'other', 'instructions', 'are',
'a', 'bit', 'different', 'because', 'they', 'depend', 'on', 'a', 'condition,', 'like',
'“is', 'the', 'bobber', 'above', 'or', 'below', 'water?".', 'Instructions', 'might', 'also',
'direct', 'the', 'flow', 'of', 'the', 'recipe,', 'like', '"if', 'you', 'haven’t', 'finished',
'fishing,', 'then', 'cycle', 'back', 'to', 'the', 'beginning', 'and', 'put', 'another',
'worm', 'on', 'the', 'hook."', 'Or,', 'how', 'about', 'a', 'condition', 'for', 'stopping,',
'as', 'in', '“if', 'you’re', 'done”', 'then', 'go', 'home.', 'You’re', 'going', 'to', 'find',
'these', 'simple', 'statements', 'or', 'instructions', 'are', 'the', 'first', 'key', 'to',
'coding,', 'in', 'fact', 'every', 'App', 'or', 'software', 'program', 'you’ve', 'ever',
'used', 'has', 'been', 'nothing', 'more', 'than', 'a', '(sometimes', 'large)', 'set', 'of',
'simple', 'instructions', 'to', 'the', 'computer', 'that', 'tell', 'it', 'what', 'to', 'do.']
300 words
>>>

256  Chapter 6

Do you see what he's talking
about? The way this was split
up, some words have periods,
commas, and double quotes at
the end of them. We could
potentially have exclamation
points, colons, question marks,
and semicolons as well.

text, strings, and heuristics

If you wanted to go through all the words to remove any extraneous characters
like periods and commas, how might you write the code to do that?

Even if you don’t have any
idea how to write the code,
think about what the process
might be like.

Computing the total number of sentences
The next step in our pseudocode is computing the number of sentences in the text.
It would be great if there was a built-in function for counting sentences, but there
isn’t, so it’s going to be up to us to figure out how to do this.
Here’s a suggestion: if we count the number of periods, semicolons, question
marks, and exclamation points in the text, that should provide a good
approximation of the number of sentences. Now it might not be perfect if, say,
an author uses punctuation in a non-standard way, but it should provide a close
approximation of the number of sentences. And often when working with messy
data like text, we can’t be perfect, at least not without considerable effort—more
on this topic in a bit.
So how do we figure out the number of terminal characters (that is “.”, “;”, “?”, or “!”)
in the text? Well, why don’t we just iterate through every character in the text and
keep a running count of the terminal characters we encounter? Sounds like a plan.
But we don’t know how to iterate through the characters of a string, or do we?
Remember we said that you can use the for statement on sequences? Well, as
it turns out, a string is just a sequence of characters. So, you can use the for
statement to iterate through all the characters in a string. Here’s an example:

Take any Python string…
lyrics = 'I heard you on the wireless back in fifty two'

…and iterate over each
character in the string.
for char in lyrics:
Each time through the loop, the next
print(char)
character in the string will be assigned
to the variable char.
Let’s print the character so
we can see how this works.

Python 3.6.0 Shell
I
h
e
a
r
d
y
o
u
o
n
t
h
e
w
i
r
e
l
e
s
s
b
a
c
k
i
n
f
i
f
t
y

t
w
o
>>>
you are here 4  257

writing some skeleton code

Writing the count sentences function
Now that we know how to iterate through a string, let’s get a skeleton of
the count_sentences function down before we write the code to count
the terminal characters:

Just like in the
pseudocode, we're
expecting text to be
passed to us.

def count_sentences(text):
count = 0
for char in text:

return count

Here's our count local
variable.
Here we’re iterating over each
character in the text.
And we need to figure out the code
for this part. Is the character a
terminal character? If so, we increase
the count.

Serious Coding
Skeleton code is a style of
coding where you lay down
the major parts of your code
without yet filling in all the
details. It’s sort of a step
between pseudocode and
completed code. Writing
skeleton code can help you
rough in a code idea before
you’ve figured out all the
actual low-level code.

Finally, we return the
count as the result of
the function.

Can you write the code to test each character to see if it is a period, semicolon, question mark, or
exclamation point, and then, if it is, increase the value of the count variable by one?

def count_sentences(text):
count = 0
for char in text:

return count

258  Chapter 6

Your code goes here.

text, strings, and heuristics

A Test Drive

We’re overdue for a test drive. Let’s get all this code together, get it in our analyze.py
file, and give it a run.

Just a reminder that we need to define functions before we
call them. Given that, think through where you could define
count_sentences (and where you couldn’t).

import ch1text
def count_sentences(text):
count = 0

for char in text:
if char == '.' or char == ';' or char == '?' or char == '!':
count = count + 1
return count
def compute_readability(text):
total_words = 0
total_sentences = 0
Make sure you’ve
total_syllables = 0
changed this since
score = 0
last Test Drive.

the

words = text.split()
total_words = len(words)
total_sentences = count_sentences(text)
print(words)
print(total_words, 'words')
print(total_sentences, 'sentences')

Here's our new
function for counting
sentences.
Let's make sure we
call our new function,
passing it the text.
We'll add some output to show us
the number of sentences.

compute_readability(ch1text.text)

Make sure you’ve
changed this since the
last Test Drive.

Here's the output
we got on the
Chapter 1 text.

Python 3.6.0 Shell

300 words
12 sentences
>>>

you are here 4  259

using the in operator

You know, you could be
a lot more concise in the
way you’re testing those
terminal characters.

It’s true.
While the way we wrote our test for terminal characters is perfectly acceptable,
there’s a more concise way to compare each character to the set of terrminal
characters, and it relies on a Boolean operator we haven’t seen yet: the in
operator. Using the in operator, you can test to see if a value is contained in a
sequence. For instance, if you remember the smoothies from Chapter 4, we could
test to see if a specific smoothie is in our list. Like this:

Here's our
sequence, a list.
smoothies = ['coconut', 'strawberry', 'banana', 'pineapple', 'acai berry']

if 'coconut' in smoothies:

print('Yes, they have coconut!')

else:

print('Oh well, no coconut today.')

And here’s what we get.

260  Chapter 6

And here we're testing if ‘coconut'
is in the list of smoothies.

Python 3.6.0 Shell

Yes, they have coconut!
>>>

text, strings, and heuristics

But as we just saw, a string is a sequence as well, and given Python tries to keep things
consistent, we can use the in operator with a string. Like this:

Here's our
sequence, a string.

Do you remember the
Thing-A-Ma-Jig mystery
at the end of Chapter 4?
That used a for loop to
iterate over a string too.

lyrics = 'I heard you on the wireless back in fifty two'
if 'wireless' in lyrics:

Python 3.6.0 Shell

print('Yes, they have wireless!')

else:

print('Oh well, no wireless today.')

And here we're testing
if ‘wireless' is in the
lyrics string.

Yes, they have wireless!
>>>

Let’s see if we can make the count_sentences function more concise (and readable) by using
the in operator. Below we’ve removed the existing code that checks for terminal characters.
We’ve also added a new local variable, terminals, that holds a string containing all the terminal
characters. Complete the if statement, using the in operator, so that it determines if the current
character is a terminal.

def count_sentences(text):
count = 0
terminals = '.;?!'
for char in text:
if ____________________________
count = count + 1
return count

Add the code using the in
operator here.

you are here 4  261

a question on strings and characters

A Test Drive

Update your code before we implement the next part of the pseudocode. The only
change you need to make is to the count_sentences function, shown below. Get
the change in and give it a test drive; your output should be the same as the last
test drive.

def count_sentences(text):
count = 0

Add these two
lines of code.

terminals = '.;?!'
And get rid of the old comparison.
for char in text:
if char in terminals:
if char == '.' or char == ';' or char == '?' or char == '!':
count = count + 1
return count

You should still see this
as output.

Python 3.6.0 Shell

300 words
12 sentences
>>>

Given that you say a string is a
sequence of characters, does that mean
characters are another data type?

In many languages, yes.
But not in Python. It’s a good question
because many languages do treat characters
as their own data type. Python, however, treats
everything as a string. So, the character ‘A’
in Python is just a string of length one that
happens to hold an A.

262  Chapter 6

text, strings, and heuristics

Given a string is a sequence, can I use
the index syntax on a string? Like my_
string[1] to get the first character?

You sure can.
Only remember my_string[1] would
get you the second character of a string
because indices start at 0. You can also use
my_string[-1] to get the last character
in the string. As you’ll soon see, you can
even use a similar syntax to access substrings
of a string, but more on that in a bit.

Nice! And I assume I can change
the string too, like with a
mystring[1] = 'e' statement?

Not so fast.
You can’t actually alter a string. One difference between lists
and strings as sequences is that lists are mutable, and strings
are immutable. That is, you can change the items in a list, but
you can’t change the characters in a string—as it turns out,
this is the case with almost every modern programming
language. You’re probably wondering why. Well, as you’ll
see as you gain more experience, being able to alter strings
can lead to unreliable code; it can also makes implementing
efficient interpreters quite difficult. That said, you can
effectively “change” a string any time just by creating a
new, altered string, and that is the common practice across
almost every modern programming language.

you are here 4  263

understanding heuristics

Computing the number of syllables, or,
learning to love heuristics
Are you ready to implement your algorithm for finding syllables?
Remember, the one you tried to write back at the beginning
of the chapter in the BRAIN POWER exercise? We should
probably apologize, as that exercise is a bit tricky—you see, it
isn’t easy to write an algorithm for finding syllables; in fact,
we can’t claim there is a definitive algorithm, short of using a
large database of words.
Finding syllables is a messy business because the English language is messy. For instance, why does “walked” have
one syllable while “loaded” is two? English is full of such inconsistencies.
For problems like this we don’t write algorithms, we write heuristics. A heuristic is a lot like an algorithm, only it’s
not a 100% solution. It might, for instance, solve a problem with a good answer, but not necessarily a perfect
answer.
So what good is a heuristic? Why not just write an algorithm with a perfect answer? For many reasons—in our
case it is because there may be no method of finding syllables that is perfect given the inconsistencies of the
English language (again, outside of using a very large database of words). In other cases, it may be that providing
a 100% solution requires so much computing time (or memory) that it makes an algorithm impractical, while
providing only a good answer requires much less time (or memory). Or, it may be because not all aspects of the
problem are known to the implementors, and so the best they can do is to provide a partial solution.
Back to our problem, though...because computing syllables is messy, our goal will be to do enough work to get a
decent estimate of syllables in our text. It won’t be perfect, but it will be close, and if you’re interested, there are
plenty of ways you can improve our heuristic further on your own (and we’ll give you some ideas for that).

Here’s your second chance: can you come up with some ideas for how to count
the syllables of an arbitrary word? Study the words in our text as a test case.
See if you can develop some general rules for determining how many syllables
a word has. Make your notes here. We’ll get you started...
If a word has three characters or fewer it typically has only one syllable.

264  Chapter 6

We came up with this:
can you come up with
some more possible rules?

text, strings, and heuristics

I don’t get why
this is so hard...why don’t you
just use a dictionary? It has all the
syllables spelled out for any word.

You have a point, but... loading
an entire English dictionary so that
words can be quickly retrieved is
actually a pretty big undertaking. It
requires not only a lot of data and
storage, but also additional technologies
like a database or a search engine in
order to have quick enough response
time to make the program usable. It
probably would require you to license
the dictionary data as well at some nontrivial cost.
Compare that to implementing a few
simple rules, which may get us up to
80–90% or greater accuracy, and the
heuristic starts to look pretty attractive
as a technology direction.

At least for the
purposes of this book!

you are here 4  265

brainstorming how to count syllables

Cubicle conversation
continued...

Alright, syllables look a
little more challenging
than counting words.

Joe: That’s what I thought too. I did
notice that if a word has fewer than
three characters, you might as well go
ahead and call it one syllable.
Judy: I’m sure there’s an exception
to that rule, but sounds good. I also
noticed that if you just look at the
number of vowels in a word, it’s a
pretty good indicator of the number of
syllables it has.
Joe: So, take “chocolate”...
Judy: I don’t mind if I do!
Joe: ...take the WORD “chocolate”—it
three vowels and three syllables.

has

Judy: Well, I think you skipped the silent e? I see a few other caveats too.
Joe: Like what?
Judy: Well, take the word “looker.” It has consecutive vowels, like oo, and we just need to count
the first vowel we see in those cases.
Joe: Oh, so we count the first o and then the e, giving us two syllables.
Joe: ...and, like you said, a lot of words have a silent e at the end as well. Like “home.”
Judy: Right, so we count the o in “home” but not the final e. So one syllable.
Judy: And related to that, a y at the end of a word often adds a syllable, like “proxy,” so we
need to consider that perhaps as a vowel as well?
Joe: Or at least that it adds a syllable to our count.
Judy: Sure. The other cases I came up with are really just special cases, like “walked” is only
one syllable, but “loaded” is two.
Joe: Right, we could build a word list of special cases and then scan all our words for them.
Judy: We could, but we only have so much time, so let’s go with the general rules we have and
then come back to the special cases if we need to.
Joe: Sounds like a plan.
266  Chapter 6

text, strings, and heuristics

Setting up the heuristic
It looks like we’ve got some good ideas for developing our syllable counting heuristic.
Of course, we’re sure you came up with even more. Let’s summarize the ideas
before we start implementing them:
 If a word has few than three characters, then count it as one syllable.
 Otherwise, count the number of vowels and let that represent the number of syllables.
 To make the previous step more accurate, remove any consecutive vowels in a word.
 Remove the final e from words, in order to account for silent e’s.
 Treat the character y as a vowel if it’s the last character.
Let’s also set up the count_syllables function so that we are all ready to code
these. Here’s how we’re going to do that:

def

,
Here’s our new funcoftionwords.
t
lis
which takes a
We’ll use a count local variable to
count_syllables(words):
hold the total number of syllables.
count = 0
We’re going to iterate over every word in our words list.
We’re then going to call another function
called count_syllables_in_word that we’ll
for word in words:
next. It returns the number
word_count = count_syllables_in_word(word) going to write
of syllables in a single word.
count = count + word_count
The local variable word_count holds the number
And then we’ll add the of syllables for the current word. Perhaps not
return count
current word’s syllable the clearest variable name, but syllable_count as a
choice would be too similar to the function name.
count to the total.
This
is a place where you’d want to document your
Finally, we’ll return the total number
choice
to avoid confusion.
of syllables across all words.

While we’re at it, let’s set up the count_syllables_in_word function too:

def count_syllables_in_word(word):
count = 0

The count_syllables_in_word function takes a single
string, a word, as a parameter.
Once again, we’ll use a local count variable to keep
track of the number of syllables.

All our heuristic code will be going here.
return count

And we’ll return it when we’re done.
you are here 4  267

writing the code to count syllables

A Test Drive

Go ahead and add the new code in your analyze.py file. At this point, just make sure
you don’t have any syntax errors. You’ll get the same output as the previous run.

You can add this code at the top of your
file, right below the import statement.

def count_syllables(words):
count = 0

for word in words:

word_count = count_syllables_in_word(word)
count = count + word_count

return count
def count_syllables_in_word(word):
count = 0

return count

We’re here.

Writing the heuristic

The first thing on our heurstic punch list is that we need to treat any word
with fewer than three characters as having one syllable. But how do we get
the length of a string? It turns out, you already know the way: with the len
function. Previously we’ve used len on a list to get its length, but it can be
applied to any sequence in Python, and a string is a sequence of characters.
Let’s try it:
def count_syllables_in_word(word):
count = 0

if len(word) <= 3:
return 1

return count

268  Chapter 6

A common theme in
this chapter

 If a word has fewer than three characters,
then count it as one syllable.
 Otherwise, count the number of vowels and
let that represent the number of syllables.
 To make the previous step more accurate,
remove any consecutive vowels in a word.
 Remove the final e from words, in order to
account for silent e’s.
 Treat the character y as a vowel if it’s the
last character.

We use the len function to get the length of the word,
which is a string. If it is less than or equal to three,
then we execute the code block of the if statement.
If the word has a length of three
or less, then we've finished the
computation and we can go ahead and
return 1 from the function.

text, strings, and heuristics

How to count vowels

 If a word has fewer than three characters,
then count it as one syllable.

Next, we need to count the vowels in each word. To do that, we’re
going to do something quite similar to when we looked for terminal
characters in a string: define a string containing all the vowels, and
then use the in operator to see if a character from our word is in that
string of vowels. We’ll compare to both lower- and uppercase vowels
in case someone typed in words in all uppercase:

 Otherwise, count the number of vowels and
let that represent the number of syllables.
 To make the previous step more accurate,
remove any consecutive vowels in a word.
 Remove the final e from words, in order to
account for silent e’s.
 Treat the character y as a vowel if it’s the
last character.

def count_syllables_in_word(word):
count = 0

if len(word) <= 3:
return 1

vowels = "aeiouAEIOU"
for char in word:

if char in vowels:

count = count + 1

return count

First we're going to create a local
variable, vowels, that holds all the vowels,
lower- and uppercase.
Then let's step through every character in the word.

We then use the in operator to see if the
acter
currently in the char variable matches any char
of
the
characters in the vowels string.

And if that character is
in the vowel string, then
we have a vowel. Let's
update our count by one.

Ignoring consecutive vowels
But we’re not done, because we need to take into account two
or more consecutive vowels. For instance, if you have the word
“book,” then you want to count the first o and ignore the second, for
a total of one syllable. Or if you have the word “roomful,” then
you want to count the first o, ignore the second, and then count the
u for a total of two syllables.

 If a word has fewer than three characters,
then count it as one syllable.
 Otherwise, count the number of vowels and
let that represent the number of syllables.
 To make the previous step more accurate,
remove any consecutive vowels in a word.
 Remove the final e from words, in order to
account for silent e’s.
 Treat the character y as a vowel if it’s the
last character.

So our goal will be to scan each character of each word for vowels,
just like our code does now, but after we encounter a vowel, we
need to make sure we ignore any vowels that follow it, until we see
another consonant. After that we can repeat the process, until we
reach the end of the word.
you are here 4  269

ignoring consecutive vowels

Writing the code to ignore consecutive vowels
Alright, let’s implement our consecutive vowel code. Here’s how we’re going to approach this:
we’ll use a Boolean variable to track if the previous character was a vowel, and if so, we know not
to count the current character if it’s a vowel. Let’s take a look:
def count_syllables_in_word(word):
count = 0

First, we’re going to add
a new local variable called
prev_char_was_vowel, and
set it to False.

if len(word) <= 3:
return 1

vowels = "aeiouAEIOU"

prev_char_was_vowel = False
for char in word:

if char in vowels:

if not prev_char_was_vowel:
count = count + 1

prev_char_was_vowel = True

else:

prev_char_was_vowel = False

return count

BE the
Interpreter
Using the count_
syllables_in_word function
above, your job is to play like you’re
the interpreter. The function is being
called with the argument “roomful”.
Step through each iteration of the loop
and fill in the local variable values
as they change. After you’ve done the
exercise, look at the end of the chapter
to see if you got it right.

270  Chapter 6

There’s a lot going on here.
Give this code some quality
time and make sure you
understand exactly how it
operates.

Our code is going to iterate over each
character in a word, just as before.
And we’re going to test if the current
character is in the vowels string.
If the current character is a vowel, and
the previous character wasn’t, then
increment the syllable count.
In either case, we then set the prev_char_
was_vowel to True before we process the
next character.
If the current character isn’t a vowel, then
we just set prev_char_was_vowel to False
before we process the next character.

char

prev_char_was_vowel

count

r

False

0

o
o
m
f
u
l

Record changes
to the local
variables at the
end of each step
through the
iteration.

text, strings, and heuristics

A Test Drive

We need to update two functions at this point, the count_syllables_in_word
function to add our new code, and the compute_readability function to call
the count_syllables function. We’re only showing the new code here, which is
highlighted below. Go ahead and make the changes and give it a test drive.

def count_syllables_in_word(word):
count = 0
if len(word) <= 3:
return 1
vowels = "aeiouAEIOU"
prev_char_was_vowel = False
for char in word:
if char in vowels:
if not prev_char_was_vowel:
count = count + 1
prev_char_was_vowel = True
else:
prev_char_was_vowel = False
return count

def compute_readability(text):
total_words = 0
total_sentences = 0
total_syllables = 0
score = 0
words = text.split()
total_words = len(words)
total_sentences = count_sentences(text)
total_syllables = count_syllables(words)
print(total_words, 'words')
print(total_sentences, 'sentences')
print(total_syllables, 'syllables')

Double-check to make sure you’re
getting the same number of words,
sentences, and syllables. If not,
double-check your code, and you
can also compare it to the book’s
source code to find any mistakes.

Python 3.6.0 Shell
300 words
12 sentences
450 syllables
>>>

you are here 4  271

count syllables finishing touches

Removing final e’s, y’s, and punctuation
Now we just have two more tasks left: removing final e’s and counting final y’s.
Actually, we’re forgetting another issue as well—that we still have some words
with punctuation at their ends, which may complicate testing for final e’s and
y’s. Let’s deal with removing the punctuation first.
Looking at our words in the words list, you can see words like “first,”, “you.”,
and “out:”. What we need to do is replace them with “first”, “you”, and
“out”, and to do that we need to get the substring of each word, excluding the
punctuation. With Python we can do that by slicing. Let’s see how that works.

 If a word has fewer than three characters,
then count it as one syllable.
 Otherwise, count the number of vowels and
let that represent the number of syllables.
 To make the previous step more accurate,
remove any consecutive vowels in a word.
 Remove the final e from words, in order to
account for silent e’s.
 Treat the character y as a vowel if it’s the
last character.

He who controls the slice, controls the...
We’ve talked about substrings, but let’s talk for a moment about exactly what
one is. A substring is just a string that occurs in another string. So if I have a
string:
lyrics = 'I heard you on the wireless back in fifty two'

then 'I' is a substring, as is 'I heard', as is 'on the wire', as is 'o',
and so on. Given a string, Python gives us a way to extract substrings from a
string using its slice syntax, which works like this:

Slice the string starting
at this index.

And stop at this index,
BUT don’t include it.

my_substring = lyrics[2:7]
print(my_substring)

Note this is a colon,
not a comma!

Python 3.6.0 Shell

heard
>>>

That means this evaluates to a string
containing the characters at index 2
through 6, or ‘heard’.
Let’s look for a few other ways you can use slice as well. If you omit the
beginning index, then slice assumes you mean to start at the beginning of
the list:

Omitting the first index
my_substring = lyrics[:6] means you want to start at
the beginning of the string.
print(my_substring)
272  Chapter 6

Python 3.6.0 Shell

I hear
>>>

text, strings, and heuristics

Or, likewise, if you omit the ending index, then slice assumes you mean to
stop the end of the list:

Python 3.6.0 Shell

Omitting the last index means
you
want to end at the end of back in fifty two
my_substring = lyrics[28:]
the string.
>>>
print(my_substring)
You can even use the negative indices like you did with lists, like this:

This says end just before the
last character in the string.
my_substring = lyrics[28:-1]
print(my_substring)

Python 3.6.0 Shell

back in fifty tw
>>>

Or, you can even use a negative index in the first index:

Start 17 characters
from the end of the
my_substring = lyrics[-17:] string, and then go all
the way to the end of
print(my_substring)
the string.

Python 3.6.0 Shell

back in fifty two
>>>

Surprise, slice isn't just for strings!
We said Python tries to be consistent, and consistent it is—you can use slice on your lists too!
smoothies = ['coconut', 'strawberry', 'banana', 'pineapple', 'acai berry']

Here's a familiar list.
smoothies[2:4]
smoothies[:2]
smoothies[3:-1]

And here's a few examples of slicing your lists. The
indexing works as it does on strings.
evaluates to
['banana', 'pineapple']

evaluates to
evaluates to

['coconut', 'strawberry']
['pineapple']

you are here 4  273

more about slicing strings

We had our slice operation and substrings all figured out, and then they
got all mixed up. Can you help us figure out who does what? Be careful,
we’re not sure if each slice has zero, one, or more matches. We’ve already
figured one out, which is marked below:
str = 'a man a plan panama'
str[:]

'man a plan panama'

str[:2]

'a man a plan panam'

str[2:]

'a '

str[1:7]

'a man a plan panama'

str[3:-1]

'an a plan panam'

str[-2:-1]

'a ma'

str[0:-1]

' man a'

str[0:4]

'm'

Putting slicing (substrings) to work
So with slicing we can take a word like “out:” and easily produce the substring
"out". In fact, based on all our slicing training over the last few pages, we know

exactly what that code would look like:

process_word = word[0:-1]

Get the substring that starts at index
0 (the beginning of the string) and
go until, but not including, the last
character.
This variable will hold a new string
containing only the string without the
ending character.
274  Chapter 6

itten
Note we could have wr
o!
to
1]
this as word[:-

text, strings, and heuristics

But we still need the logic of knowing when to remove the final character of a word. We want
to remove the final character if it is a period, comma, semicolon, exclamation point, or question
mark, and based on our work in this chapter, we have all the pieces we need to do ths:
def count_syllables_in_word(word):
count = 0

endings = '.,;!?:'

last_char = word[-1]

Let's set up a string that contains
all the word endings.
And get the last character of the
current word.
Then check to see if the last
character is one of the endings.

if last_char in endings:

processed_word = word[0:-1]
else:
processed_word = word
if len(processed_word) <= 3:
return 1
vowels = "aeiouAEIOU"
prev_char_was_vowel = False

If so, we'll set the processed_word to be
the word, without the last character.
If not, we're going to set the
processed_word to the entire word.
And then we're going to use
processed_word rather than word in
the rest of the code.

for char in processed_word:
if char in vowels:
if not prev_char_was_vowel:
count = count + 1
prev_char_was_vowel = True
else:
prev_char_was_vowel = False

Don’t worry about entering this code
yet:; we'll do it on the next page.

return count

Assuming the variable process_word already has any punctuation removed from the last
character, can you write a couple lines of code (or so) to remove a final e character, if it has one?
Just write the code fragment here.

you are here 4  275

testing count syllables

A Test Drive

Let’s get all these additions based on slicing into the code. You’ll find all the
new code from the last few pages highlighted here. Go ahead and get it entered
and try a test run to see your syllable count change.
def count_syllables_in_word(word):
count = 0
endings = '.,;!?:'
last_char = word[-1]
if last_char in endings:

Make sure
you've made
these additions
we covered on
the previous
page.

processed_word = word[0:-1]

else:

processed_word = word

if len(processed_word) <= 3:
return 1
if processed_word[-1] in 'eE':
processed_word = processed_word[0:-1]
vowels = "aeiouAEIOU"
prev_char_was_vowel = False
for char in processed_word:
if char in vowels:
if not prev_char_was_vowel:
count = count + 1
prev_char_was_vowel = True
else:
prev_char_was_vowel = False
return count

Finishing off the heuristic code
We’re on the last step of implementing the heuristic—all we need to do
now is count the y’s at the end of words as syllables. At this point we know
how to examine the end of a word, and we know that if we see a y at the
end, we just need to increment our count local variable by one. Let’s put
it all together and test the final version of count_syllables_in_word,
not to mention count_syllables.
276  Chapter 6

We've put the code to
remove silent e characters
here, after the test for
words that are length
three or less.
Notice the number of syllables
calcuation has changed with
those code changes!
Python 3.6.0 Shell
300 words
12 sentences
410 syllables
>>>

 If a word has fewer than three characters,
then count it as one syllable.
 Otherwise, count the number of vowels and
let that represent the number of syllables.
 To make the previous step more accurate,
remove any consecutive vowels in a word.
 Remove the final e from words, in order to
account for silent e’s.
 Treat the character y as a vowel if it’s the
last character.

text, strings, and heuristics

A Test Drive

It's time to finish off the count_syllables_in_word function. Below you’ll find
the code to increase the syllable count if there is a y at the end of the word.
Give it a test run and check out the final syllable number.
def count_syllables_in_word(word):
count = 0
endings = '.,;!?:'
last_char = word[-1]
if last_char in endings:

processed_word = word[0:-1]

else:

processed_word = word

if len(processed_word) <= 3:
return 1
if processed_word[-1] in 'eE':
processed_word = processed_word[0:-1]
vowels = "aeiouAEIOU"
prev_char_was_vowel = False
for char in processed_word:
if char in vowels:
if not prev_char_was_vowel:
Check the word to see if
count = count + 1
it ends in y or Y, and if
prev_char_was_vowel = True
so, increase the syllable
else:
count.
prev_char_was_vowel = False

Looks like we found some
words ending in y, so our
syllable count increased.

if processed_word[-1] in 'yY':
count = count + 1
return count

You’ll notice we’ve written quite a long function here, which
is perfectly acceptable. A bit like a paragraph, you’ll find a
function is probably too long when it becomes hard to understand
and keep in your head as you study it.

Python 3.6.0 Shell
300 words
12 sentences
416 syllables
>>>

you are here 4  277

the reading ease formula

Implementing the reading ease formula
We’re getting near the end; all we have left is to implement the formula
to get a reading ease score, and then to provide some output to report the
results. In terms of the formula, we have all the pieces we need: a count
of words, a count of sentences, and a count of syllables. Let’s look at the
original formula:

206.835 – 1.015

(

total words
total sentences

)– 84.6 (

)

total syllables
total words

and let’s translate it into Python:
score = (206.835 - 1.015 * (total_words / total_sentences)
- 84.6 * (total_syllables / total_words))

A Test Drive

Remember we can wrap the
formula in parentheses so we can
write it across multiple lines.

And now let’s get the formula in our compute_readability function. Go ahead
and add the formula as well as the new print statement and give it a try.

def compute_readability(text):
total_words = 0
total_sentences = 0
total_syllables = 0
score = 0
words = text.split()
total_words = len(words)
total_sentences = count_sentences(text)
total_syllables = count_syllables(words)

Add the formula below
to our word, sentence,
and syllable count code.

score = (206.835 - 1.015 * (total_words / total_sentences)
- 84.6 * (total_syllables / total_words))
print(total_words, 'words')
print(total_sentences, 'sentences')
print(total_syllables, 'syllables')
print(score, 'reading ease score')

278  Chapter 6

And add a print
statement so you can
see the results.

Nice, we have
a reading ease
score!
Python 3.6.0 Shell

300 words
12 sentences
416 syllables
64.14800000000001 reading ease score
>>>

text, strings, and heuristics

You have the reading ease score, so all you need to do is output the results. Define a
function, output_results, that takes a score as a parameter and outputs the reading
level as specified in the pseudocode. We’ve reproduced pseudocode below to make
things easier.
IF score >= 90.0:
PRINT ‘Reading level of 5th Grade’
ELIF score >= 80.0:
PRINT ‘Reading level of 6th Grade’
ELIF score >= 70.0:
PRINT ‘Reading level of 7th Grade’
ELIF score >= 60.0:
PRINT ‘Reading level of 8-9th Grade’
ELIF score >= 50.0:
PRINT ‘Reading level of 10-12th Grade’
ELIF score >= 30.0:
PRINT ‘Reading level of College Student’
ELSE:
PRINT ‘Reading level of College Graduate’

.

Here’s the pseudocode

Write your Python code here.

you are here 4  279

the finished code

A Test Drive

This is it—we just need to add in the output_results function and we should
have everything we need to complete the code! Go ahead and make the additions
and changes below. You’ll find the complete code over the next two pages.

import ch1text
def count_syllables(words):
count = 0
for word in words:
word_count = count_syllables_in_word(word)
count = count + word_count
return count
def count_syllables_in_word(word):
count = 0
endings = '.,;!?:'
last_char = word[-1]
if last_char in endings:
processed_word = word[0:-1]
else:
processed_word = word

Here’s all the code we’ve already
written. This should be in your
analyze.py file already.

if len(processed_word) <= 3:
return 1
if processed_word[-1] in 'eE':
processed_word = processed_word[0:-1]
vowels = "aeiouAEIOU"
prev_char_was_vowel = False
for char in processed_word:
if char in vowels:
if not prev_char_was_vowel:
count = count + 1
prev_char_was_vowel = True
else:
prev_char_was_vowel = False
if processed_word[-1] in 'yY':
count = count + 1
return count

280  Chapter 6

There's more code on
the next page.

text, strings, and heuristics

def count_sentences(text):
count = 0
terminals = '.;?!'
for char in text:
if char in terminals:
count = count + 1
return count
def output_results(score):
if score >= 90:
print('Reading level
elif score >= 80:
print('Reading level
elif score >= 70:
print('Reading level
elif score >= 60:
print('Reading level
elif score >= 50:
print('Reading level
elif score >= 30:
print('Reading level
else:
print('Reading level

of 5th Grade')

Add the new
output_results
code.

of 6th Grade')
of 7th Grade')
of 8-9th Grade')
of 10-12th Grade')
of College Student')
of College Graduate')

def compute_readability(text):
total_words = 0
total_sentences = 0
total_syllables = 0
score = 0
words = text.split()
total_words = len(words)
total_sentences = count_sentences(text)
total_syllables = count_syllables(words)

So our writing is at the
8th- to 9th-grade level.
Scores in that range are
considered acceptable for
most books and articles, so
we’re not doing too bad!

score = (206.835 - 1.015 * (total_words / total_sentences)
- 84.6 * (total_syllables / total_words))
print(total_words, 'words')
print(total_sentences, 'sentences')
print(total_syllables, 'syllables')
print(score, 'readability score')
output_results(score)

compute_readability(ch1text.text)

Python 3.6.0 Shell
Reading level of 8-9th Grade
>>>

And make sure we’re
calling our new function,
passing it the score.
you are here 4  281

explaining scores

Are you trying to tell
me an 8th- or 9th-grade reading
level is good? Doesn’t the author of
this book have a Ph.D. or something?

The goal is not to get a really high
or really low score. A low score on

Flesch’s readability test indicates that the text is quite
challenging to read. For example, a score in the 30–50
range would indicate that the text is written like an
advanced college textbook or research paper in a
scientific field. Not exactly the score you’d want for
the opening text of a Head First book; in fact, it’s just
the opposite of what you’d want.

Hopefully this book is accessible to those entering high
school, so this score doesn’t seem too off the mark. In
fact, think of another example: say you wrote a novel
for the young adult market. You’d probably want a
score that indicates an even lower reading level (say,
7th grade) and higher score (say, 70–80) so that it is
accessible to those, say, in junior high school.

282  Chapter 6

Score

School Level

Notes

100.00–90.00

5th grade

Very easy to read. Easily understood by an average 11-year-old student.

90.0–80.0

6th grade

Easy to read. Conversational English for consumers.

80.0–70.0

7th grade

Fairly easy to read.

70.0–60.0

8th & 9th grade

Plain English. Easily understood by 13- to 15-year-olds.

60.0–50.0

10th–12th grade

Fairly difficult to read.

50.0–30.0

College

Difficult to read.

30.0–0.0

College Graduate Very difficult to read. Best understood by university graduates.

text, strings, and heuristics

Taking it even further
Are you all ready to test the writing of your favorite author,
news sources, or even your own writing? Here’s instructions
for how to do that:
We'll
 Create a new file of your own, with a name ending in .py.
 Using the three-quote syntax, add your own multiline text
and assign it to a variable—we recommend naming it
text for consistency.

see
even
an
better way
in the next
chapter.

 In your analyze.py file, import your new file (otherwise
known as a module).
 Call the compute_readability function and pass it
your module name followed by the variable name text.
And don’t forget our heuristic code could still be improved.
Just take a look at words in the word list; there’s still some
issues, like the use of double quotes that show up in words,
that could be cleaned up. Also, we could still build that
special-case word list, right?
As you can see, a coder’s job is never truly done. But in terms
of this chapter you are done—oh, except for the bullet points
and the crossword. Anyway, congrats, this was a tough chapter
that really demonstrates the coding chops you’re developing!

Nice job!

 You can include multiline text in your code
using Python’s three-quote syntax.
 You can determine the length of a string
using the len function, just like you can
with lists.
 The string split function uses a separator
to split a string into words that are placed
into a list.
 A delimiter is another word for a separator.
 Strings are immutable. Lists are mutable.
 Mutable means you can change it;
immutable means you can’t.
 Almost all modern languages have
immutable strings.
 We can iterate through the characters of a
string using the for statement.
 We can also use the index notation to
access characters in a string.
 The Boolean operator in allows you to
check for inclusion of values in lists or
strings.
 A heuristic provides a best estimate or
good answer but not necessarily a perfect
one.
 Heuristics are often used when an
algorithm isn’t computationally practical,
or known.
 Strings support a slice notation that
returns a substring. You can use this
notation on lists as well.
 In Python there is no separate character
type as there is in some programming
languages. In Python all text is a string.

you are here 4  283

coding crossword

Coding cross
This was a character-building chapter. Relax
with a little crossword.
1
2

3

1

2

3

4

4

5
5

6
6

7
7
8

8

10

12

10

11

11

12
13

13
14

15

15

Across

Across
2. We removed these.

3.
test text was
from.
2.What
We removed
these.
5.
The
in
operator
works
on these.
3. What test text was from.
6. At the end of some words.
5. The in operator works on these.
7. Another name for separator.
6.We
At ignored
the endthese
of some
words.
10.
vowels.
7.
Another
name
for
separator.
12. Subset of a string.
10.Heard
We ignored
these
vowels.
13.
you on this
in 52.
12.Came
Subset
a string.
15.
upof
with
the formula.
16.
needed
heuristic
13.We
Heard
youaon
this infor
52.this.

15. Came up with the formula.
16. We needed a heuristic for this.

284  Chapter 6

9

9

16

14

16

Down

Down
1. Our
new business.
4. A solution
not 100%.
1. Our that's
new business.
8. Means
you
can't
4. A solutionchange
that's it.
not 100%.
9. May have made computing syllables easier.
8. Means you can't change it.
11. Python's way to get substrings.
9. May
have apart.
made computing syllables easier.
14. Breaking
a string
11. Python's way to get substrings.
14. Breaking a string apart.

text, strings, and heuristics

We’re going to take the text in the ch6/text.txt file and get it into a Python file. To
do that, in IDLE, create a new file and add the code below. For the Chapter 1 text,
you’ll want to open the text.txt file in IDLE as well, and then copy and paste the
text into your new file. When you’re done, save your file as ch1text.py.
Finally, run your code, and you should see the entire article printed to the shell.
Python 3.6.0 Shell

The code printed the entire
article; here’s the tail end.

into pond", or "pull in the fish." But also notice that other
instructions are a bit different because they depend on a condition,
like “is the bobber above or below water?". Instructions might also
direct the flow of the recipe, like "if you haven’t finished fishing,
then cycle back to the beginning and put another worm on the hook."
Or, how about a condition for stopping, as in “if you’re done” then go
home.
You’re going to find these simple statements or instructions are the
first key to coding, in fact every App or software program you’ve ever
used has been nothing more than a (sometimes large) set of simple
instructions to the computer that tell it what to do.
>>>

Now that you know how to use the split function, let’s return to the compute_readability
function and take it a little further. In our pseudocode we were going to write a function count_words,
but as it turns out, with split we can handle the job of computing the total number of words with only
two lines of code, so let’s forgo the function. Go ahead and complete that code below.

import ch1text

Python 3.6.0 Shell

def compute_readability(text):
To get the total number
total_words = 0
total_sentences = 0
of words we use the len
total_syllables = 0
function on the list of
score = 0
words = text.split()
total_words = len(words)

words.

print(words)
print('total words', total_words)
print(text)
compute_readability(ch1text.text)

['The', 'first', 'thing', 'that', 'stands', 'between', 'you', 'and', 'writing', 'your',
'first,', 'real,', 'piece', 'of', 'code,', 'is', 'learning', 'the', 'skill', 'of', 'breaking',
'problems', 'down', 'into', 'achievable', 'little', 'actions', 'that', 'a', 'computer',
'can', 'do', 'for', 'you.', 'Of', 'course,', 'you', 'and', 'the', 'computer', 'will', 'also',
'need', 'to', 'be', 'speaking', 'a', 'common', 'language,', 'but', "we'll", 'get', 'to',
'that', 'topic', 'in', 'just', 'a', 'bit.', 'Now', 'breaking', 'problems', 'down', 'into',
'a', 'number', 'of', 'steps', 'may', 'sound', 'a', 'new', 'skill,', 'but', 'its', 'actually',
'something', 'you', 'do', 'every', 'day.', 'Let’s', 'look', 'at', 'an', 'example,', 'a',
'simple', 'one:', 'say', 'you', 'wanted', 'to', 'break', 'the', 'activity', 'of', 'fishing',
'down', 'into', 'a', 'simple', 'set', 'of', 'instructions', 'that', 'you', 'could', 'hand',
'to', 'a', 'robot,', 'who', 'would', 'do', 'your', 'fishing', 'for', 'you.', 'Here’s',
'our', 'first', 'attempt', 'to', 'do', 'that,', 'check', 'it', 'out:', 'You', 'can', 'think',
'of', 'these', 'statements', 'as', 'a', 'nice', 'recipe', 'for', 'fishing.', 'Like', 'any',
'recipe,', 'this', 'one', 'provides', 'a', 'set', 'of', 'steps,', 'that', 'when', 'followed',
'in', 'order,', 'will', 'produce', 'some', 'result', 'or', 'outcome', 'in', 'our', 'case,',
'hopefully,', 'catching', 'some', 'fish.', 'Notice', 'that', 'most', 'steps', 'consists',
'of', 'simple', 'instruction,', 'like', '"cast', 'line', 'into', 'pond",', 'or', '"pull',
'in', 'the', 'fish."', 'But', 'also', 'notice', 'that', 'other', 'instructions', 'are',
'a', 'bit', 'different', 'because', 'they', 'depend', 'on', 'a', 'condition,', 'like',
'“is', 'the', 'bobber', 'above', 'or', 'below', 'water?".', 'Instructions', 'might', 'also',
'direct', 'the', 'flow', 'of', 'the', 'recipe,', 'like', '"if', 'you', 'haven’t', 'finished',
'fishing,', 'then', 'cycle', 'back', 'to', 'the', 'beginning', 'and', 'put', 'another',
'worm', 'on', 'the', 'hook."', 'Or,', 'how', 'about', 'a', 'condition', 'for', 'stopping,',
'as', 'in', '“if', 'you’re', 'done”', 'then', 'go', 'home.', 'You’re', 'going', 'to', 'find',
'these', 'simple', 'statements', 'or', 'instructions', 'are', 'the', 'first', 'key', 'to',
'coding,', 'in', 'fact', 'every', 'App', 'or', 'software', 'program', 'you’ve', 'ever',
'used', 'has', 'been', 'nothing', 'more', 'than', 'a', '(sometimes', 'large)', 'set', 'of',
'simple', 'instructions', 'to', 'the', 'computer', 'that', 'tell', 'it', 'what', 'to', 'do.']
300 words
>>>

you are here 4  285

exercise solution

Can you write the code to test each character to see if it is a period, semicolon, question mark, or
exclamation point, and then, if so, increase the value of the count variable by one?

def count_sentences(text):
count = 0
for char in text:
if char == '.' or char == ';' or char == '?' or char == '!':
count = count + 1
return count

Check if char is
any of the terminal
characters, and if so,
increment count by one.

Let’s see if we can make the count_sentences function more concise (and readable) by using
the in operator. Below we’ve removed the existing code that checks for terminal characters. We’ve
also added a new local variable called terminals, which holds a string containing all the terminal
characters. Complete the if statement, using the in operator, so that it determines if the current
character is a terminal.

def count_sentences(text):
count = 0
terminals = '.;?!'
for char in text:
if char in terminals:
count = count + 1
return count

286  Chapter 6

Wow, that is a lot more
concise and readable!

text, strings, and heuristics

BE the Interpreter
SOLUTION
Using the
count_syllables_in_word
function above, your job is to
play like you’re the interpreter.
The function is being called with
the argument “roomful”.

char

prev_char_was_vowel

count

r

False

0

o

True

1

o

True

m

False

f

False

u

True

l

False

2

SOlUTion
We had our slice operation and substrings all figured out, and then they
got all mixed up. Can you help us figure out who does what? Be careful,
we’re not sure if each slice matches zero, one, or more substrings. We’ve
already figured one out, which is marked below:
str[:]

'man a plan panama'

str[:2]

'a man a plan panam'

str[2:]
str[1:7]
str[3:-1]
str[-2:-1]

'a '
'a man a plan panama'
'an a plan panam'

str[0:-1]

'a ma'

str[0:4]

' man a'
'm'
you are here 4  287

exercise solution

Assuming the variable process_word alread has any punctuation removed from the last character,
can you write the code to remove a final e character, if it has one?

..

If the last character of process_word is either ‘e' or ‘E' then.
if processed_word[-1] in 'eE':
processed_word = processed_word[0:-1]

…set processed_word to itself, without the last character.

You have the reading ease score, so all you need to do is output the results.
Define a function, output_results, that takes a score as a parameter and
outputs the reading level as specified in the pseudocode.

def output_results(score):
if score >= 90:
print('Reading level
elif score >= 80:
print('Reading level
elif score >= 70:
print('Reading level
elif score >= 60:
print('Reading level
elif score >= 50:
print('Reading level
elif score >= 30:
print('Reading level
else:
print('Reading level

288  Chapter 6

of 5th Grade')
of 6th Grade')
of 7th Grade')
of 8-9th Grade')
of 10-12th Grade')
of College Student')
of College Graduate')

text, strings, and heuristics

1
2

F

I

D

N A

L

T

5

A
6

4

C H A

P

U N C

8

S

E Q U E

I

12

S

15

F

T U A
L

I

10

M
11

C O N S

M

E

U B

S

T R

I
C

A

E

B
E

S

N C E

S

T

I

1

O N

I

T E

R

I
E

C U

T

L

T

L

R

S

N

M

T E

R

I
D E

P

E

S

7

Coding
Cross
Solution

3

E

C H

9

D
I

V

E

C
N G

13
14

16

S

Y

E

W

T
I

S

O

P

N

L

L

A

I

R

T

Y

R E

L

E

B

E

S

L

S

S

Across

Down

2. We removed these. [FINALE]
3. What test text was from. [CHAPTER1]
5. The in operator works on these.
[SEQUENCES]
6. At the end of some words. [PUNCTUATION]
7. Another name for separator. [DELIMITER]
10. We ignored these vowels. [CONSECUTIVE]
12. Subset of a string. [SUBSTRING]
13. Heard you on this in 52. [WIRELESS]
15. Came up with the formula. [FLESCH]
16. We needed a heuristic for this. [SYLLABLES]

1. Our new business. [DATASCIENCES]
4. A solution that's not 100%. [HEURISTIC]
8. Means you can't change it. [IMMUTABLE]
9. May have made computing syllables easier.
[DICTIONARY]
11. Python's way to get substrings. [SLICE]
14. Breaking a string apart. [SPLIT]

you are here 4  289

7

modules, methods, classes, and objects

Getting Modular

Your code is growing in size and complexity. As that
happens you need better ways to abstract, to modularize, to organize
your code. You’ve seen that functions can be used to group lines of code
together into bundles you can reuse over and over. And you’ve also seen
that collections of functions and variables can be placed into modules
so that they can be more easily shared and reused. In this chaper we’ll
revisit modules and learn how to use them even more effectively (so
you’re all ready to share your code with others) and then we’re going
to look at the ultimate in code reuse: objects. You’re going to see that
Python objects are all around you, just waiting to be used.

this is a new chapter   291

reusing the readability analyzer

Hey guys, I’d love to use your
readability analyzer on my book.
How can I reuse your code
from the previous chapter?

Cory Doctorow

292  Chapter 7

Text from his book
“Little Brother"

modules, methods, classes, and objects

Cubicle conversation
Whoa! Someone wants
to use our code?

Greg: That’s easy; just ship him analyze.py. He can just import
it and call the compute_readability function. Done!
Frank: Totally. Modules are just Python files; just ship it.
Joe: I don’t think that is quite right. I know we’ve said a file is
just a module and all that, and we’ve certainly used import,
but I think there is more to it than that.
Greg: Like what?

Greg

Frank

Joe: Well, for instance, right now, if you’ll remember, our
analyze.py file has the code to compute the readability of the
Chapter 1 text.

Joe

Frank: That was great for testing out the analyzer, but Cory
will need to rewrite that to use his text.
Joe: Right, but anyone should be able to use our code without
having to open the code and edit it; and I’d love to leave in
our test in case we want to improve our heuristic in the future.
There’s gotta be a better way.
Greg: What do you have in mind?
Joe: I’ve been doing some research. There’s a convention for
organizing modules so that we can still have our test code, and
Cory can use it for his own analysis.
Greg: I’d love to hear more.
Frank: Wait, is this going to be a lot of work?
Joe: Frank, this isn’t difficult, and I think you’re going to like
the end result. These changes are going to make our code
more reusable by other people and us.

you are here 4  293

a review of python modules

A quick module review
Now, as you also already know, to import a module, we use the import
statement, like this:
import random

Import the random module.

And then you can refer to any of that module’s functions or variables by
prepending their name with the module name, like this:

We would like to invoke the randint function,
which is found in the random module.
num = random.randint(0,9)

…and then the function orle.
We start with
variable name in the modu
the module name…
…followed by the
dot operator…

We’re going to see that
the dot operator has a
few more tricks up its
sleeve when it comes to
objects; more on that later.

Notice that we’ve already used the dot operator quite few times in this book
without really talking about it—think of the dot operator as just a way of
saying “look for randint in the module random.”

Q:

to tell Python to look in other directories too,
should you need that functionality.

A:

I’ve seen the term “Python library”;
is that related to a module?

How does Python know where to
look for a module when I import it?

Good question, because when you
import a module you specify only the
module’s name and not, say, a directory path.
So how does Python find modules? Python
starts by looking at its internal list of built-in
modules (random is a built-in module) and if
it doesn’t find the module in that list, it looks
in the local file directory where you ran your
code from. There are more advanced ways

294  Chapter 7

Q:

A:

A library is a generic term sometimes
used with Python modules (or collections of
Python modules). Typically the term library
just means that the modules have been
published for others to use. You’ll hear the
term package too, which often means a set
of Python modules that work together.

Q:

What happens if I import, say, the
random module into my code, but I also
import another module that already
imports random? Is there going to be a
problem or conflict?

A:

No, Python keeps track of the modules
that have been imported so that it doesn’t
have to reimport modules over and over. It’s
also perfectly fine for your code and another
module to both import the same module.

modules, methods, classes, and objects

Cubicle conversation continued...

So are we gonna ship
this analyze code?

Frank: We are; we just have to make sure we have it in a form that
other people can use.
Joe: Right now we still have our test of the Chapter 1 text in the
code.
Judy: Can’t you just pull it out?
Joe: We could, but we really want to keep it around for testing as we
improve our own heuristic.
Judy: Oh, well, how are you going to do that?
Frank: Joe’s been studying Python modules. You can structure a
module file so that it knows if it is running as the main program, or
if it has been imported by another Python file.
Judy: What does that get us?
Joe: Think about it like this, Judy: if someone is running analyze.py
directly, then it’s probably us and we’ll want to execute our test code.
If not, then someone is importing analyze.py, so they’d want to skip
the test code.
Judy: Oh, that makes a lot of sense. We can do that?
Joe: Yes, we use a convention that allows us to check if analyze.py is
being executed directly as the main program, and, if so, we’ll invoke
the test code; otherwise, we’ll ignore it. Let me show you...

you are here 4  295

using the __name__ variable

The

name

global variable

Serious Coding

Whenever a Python file is executed, behind the scenes the Python interpreter
creates a global variable __name__, which consists of two underscores, the
word “name,” and two more underscores. After it creates the variable, Python
then sets this variable to one of two things: if the Python file is being executed
directly as the main program, the variable is set to the string "__main__".
Otherwise, __name__ is set to the name of your module, like "analyze".
Given how __name__ works, there is a long-standing Python convention of
using the following code to test if your code is running as the main program:

if __name__ == '__main__':

Python coders pronounce
the __name__ variable as
dunder name, which is short
for “double underscore name,
double underscore.”

Test to see if this is the main
program, and if so, supply
whatever code you want here.

print("Look, I'm the main program y'all.")

The important thing to note is this
print statement will be ignored if this
file is being imported.

A Test Drive

Let’s put __name__ to the test. Below you’ll find two Python files; type them both
into IDLE, store them in the same folder, and run them both. Check your output.

if __name__ == '__main__':

import just_a_module

else:

print('Greetings from main.py')

print("Look, I'm the main program y'all.")
print("Oh, I'm just a module.")

just_a_module.py
Python 3.6.0 Shell
Look, I’m the main program y’all.
>>>

Here just_a_module knows
it’s the main program.
296  Chapter 7

main.py

Here’s what we got
executing these two files.

Python 3.6.0 Shell
Oh, I’m just a module.
Greetings from main.py
>>>

Here main.py does an import, and just_a_module
knows it’s being executed from an import.

modules, methods, classes, and objects

Cubicle conversation continued...
Okay, I think I got it.

Greg: Yeah, we just need to test to see if the __name__
variable is set to "__main__", and if so, we’ll run our test
code; otherwise, we’ll do nothing. After we’ve done that, we
can give it to Cory.
Frank: Let’s do it... Wait, how is Cory going to know which
functions to use?
Greg: One problem at a time, Frank; we’ll come back to that
one. Let’s get the code in there first.
Frank: Fair enough!

Updating analyze.py
At this point copy the analyze.py file from your Chapter 6 folder to your
Chapter 7 folder and open it. You’ll want to make these changes:
import ch1text
def count_syllables(words):
count = 0

Go ahead and make a copy of
ch1text.py from the Chapter 6
folder as well and place it in your
Chapter 7 folder.

At the top of the file, delete
the import of the ch1text file.
We’re going to move this to the
bottom of the file.

Here’s the top of
your analyze.py file.

for word in words:
word_count = count_syllables_in_word(word)
count = count + word_count
return count
def count_syllables_in_word(word):
count = 0
endings = '.,;!?:'
last_char = word[-1]

you are here 4  297

running as the main program

Here’s the bottom of
your analyze.py file.

def compute_readability(text):
total_words = 0
total_sentences = 0
total_syllables = 0
score = 0
words = text.split()
total_words = len(words)
total_sentences = count_sentences(text)
total_syllables = count_syllables(words)
score = (206.835 - 1.015 * (total_words / total_sentences)
- 84.6 * (total_syllables / total_words))

Here’s the conditional test for the
__main__ value in the __name__ variable.
If true we’re going to import the ch1text
__name__ == "__main__":
file, and compute its readability.
import ch1text
print('Chapter1 Text:')
compute_readability(ch1test.text) We added a print statement too.
output_results(score)

if

Don’t forget to indent the compute_readability
line four spaces under the if statement.

A Test Drive

Get the changes made to analyze.py and then give it a test drive. Because we’re
running it as the main program, you should get the same output you did in Chapter 6.
Don’t forget to make sure you have a copy of ch1text.py in the same folder.

Works as advertised. We’re running
analyze.py as the main program, so
it goes ahead and does a test of
the text in ch1text.py.

298  Chapter 7

Yes, you can use the
import statement
anywhere in your code!

Python 3.6.0 Shell
Chapter 1 Text:

Reading level of 8-9th Grade
>>>

modules, methods, classes, and objects

Using analyze.py as a module
So how are others going to reuse our code? Well, they’re first going to import the
analyze module into their own code, and then they’ll call compute_readability
in the analyze module with a text string. Let’s create a new file called cory_analyze.py

and write the code to do that:

Import the module.
import analyze

And call analyze.compute_readability with the text.

analyze.compute_readability("""
If you've never programmed a computer, you should. There's nothing like it in the
whole world. When you program a computer, it does exactly what you tell it to do.
It's like designing a machine: any machine, like a car, like a faucet, like a gas
hinge for a door using math and instructions. It's awesome in the truest sense it
can fill you with awe.
A computer is the most complicated machine you'll ever use. It's made of billions
of micro miniaturized transistors that can be configured to run any program you
can imagine. But when you sit down at the keyboard and write a line of code, those
transistors do what you tell them to.
Most of us will never build a car. Pretty much none of us will ever create an
aviation system. Design a building. Lay out a city.""")

Remember, you can use
three double quotes to
create a multiline string.

A Test Drive
You don’t actually
have to type in
Cory’s text; you
can find it in the
Chapter 7 source
files under cory.txt.

Get cory_analyze.py typed in and then give it a test drive as well. Because we’re
running analyze.py as a module now, our test code is not going to be run. Give it a try
and see what you get.

Python 3.6.0 Shell

Cory's text gets a 7th-grade
rating. That seems perfect for a
book with teen subject matter!

Reading level of 7th Grade
>>>

you are here 4  299

the python help function

Okay, I’d still love
to know how Cory is going to
know what functions to call in
the analyze module?

That’s where Python help comes in.
With Python you can add help documentation right
into your source code. Of course, you’ve already
seen how to add comments to your code—those
are for the purpose of documenting your low-level
code—but Python also allows you to add docstrings to
provide higher-level documentation for programmers
who are using your modules (but aren’t interested in
digging through your code).

Frank

The docstring format is simple: you just add a string
at the top of your module as a general description,
and a string after each function definition (as well as
object definitions, which you haven’t learned about
yet).
So how do you use these docstrings? Do they require
others to open your module file and read them?
No, there’s a better way: Python provides a help
function, which anyone can use in the interpreter to
see the documentation.
Let’s add some docstrings to analyze.py and see how
this works.

300  Chapter 7

You’ll find
most modern
programming
languages
provide some
form of
docstrings.

modules, methods, classes, and objects

Adding docstrings to analyze.py
Let’s add in some docstrings so other coders can make use of Python’s help
system to learn how to use our module.
"""The analyze module uses the Flesch–Kincaid readability test to analyze text and
produce a readability score. This score is then converted into a
grade-based readability category.
"""

You can add a multiline
string at the top of
your module...

def count_syllables(words):
"""This function takes a list of words and returns a total
count of syllables across all words in the list.
"""
count = 0
for word in words:
word_count = count_syllables_in_word(word)
count = count + word_count

…and below any definition.

return count
def count_syllables_in_word(word):
"""This function takes a word in the form of a string
and returns the number of syllables. Note this function is
a heuristic and may not be 100% accurate.
"""
count = 0

endings = '.,;!?:'
# these are the word terminals we care about
last_char = word[-1]
if last_char in endings:
processed_word = word[0:-1]
else:
processed_word = word
if len(processed_word) <= 3:
return 1

Only docstrings are included in
Python help. Comments added to the
code, like this one, are not included.

if processed_word[-1] in 'eE':
processed_word = processed_word[0:-1]
vowels = "aeiouAEIOU"
prev_char_was_vowel = False

for char in processed_word:
if char in vowels:
if not prev_char_was_vowel:
count = count + 1
prev_char_was_vowel = True
else:
prev_char_was_vowel = False
if processed_word[-1] in 'yY':
count = count + 1
return count

Continued on next page...
you are here 4  301

adding docstrings

def count_sentences(text):
"""This function counts the number of sentences in a string of text
using period, semicolon, question mark and exclamation mark as
terminals.
"""
count = 0
terminals = '.;?!'
for char in text:
if char in terminals:
count = count + 1

Even more docstrings!

return count
def output_results(score):
"""This function takes a
corresponding reading
"""
if score >= 90:
print('Reading level
elif score >= 80:
print('Reading level
elif score >= 70:
print('Reading level
elif score >= 60:
print('Reading level
elif score >= 50:
print('Reading level
elif score >= 30:
print('Reading level
else:
print('Reading level

Flesch-Kincaid score and prints the
level.
of 5th Grade')
of 6th Grade')
of 7th Grade')
of 8-9th Grade')
of 10-12th Grade')
of College Student')
of College Graduate')

def compute_readability(text):
"""This function takes a text string of any length and prints out a
grade-based readability score.
"""
total_words = 0
total_sentences = 0
total_syllables = 0
score = 0
words = text.split()
total_words = len(words)
total_sentences = count_sentences(text)
total_syllables = count_syllables(words)
score = (206.835 - 1.015 * (total_words / total_sentences)
- 84.6 * (total_syllables / total_words))
output_results(score)
if __name__ == "__main__":
import ch1text
print('Chapter 1 Text:')
compute_readability(ch1text.text)

302  Chapter 7

Note that you can make
your documentation as
elaborate as you like or as
is needed. Python allows
you to do this in a freeform manner. Some other
programming languages have
quite sophisticated systems
for specifying docstrings
that are more standardized
and less free-form.

modules, methods, classes, and objects

A Test Drive
Depending on your
operating system
and version of
Python

Python 3.6.0 Shell
>>> help(analyze)
Help on module analyze:
NAME

Go ahead and document the analyze.py file; get creative if you like and expand
your documentation. After you’ve done that, you need to follow some very specific
instructions. Remember we said that Python, for efficiency reasons, won’t import the
same module over and over? Instead, it keeps a cached version around in memory.
So you’ve changed your analyze module, but Python is still going to rely on the
cached version (the one without docstrings). To get around that, you need to exit
IDLE. Quit all the way out. Then, run IDLE again, open your cory_analyze.py file
and run it to ensure you didn’t introduce any errors when you documented the code.
After that, locate IDLE’s shell window and follow the instructions below to test your
new help documentation:

Don’t skip running cory_analyze.py,
as it imports your new analyze.py
file and ensures IDLE is using the
right directory paths.
Here's how you import
and call help on the
analyze module.

analyze
DESCRIPTION
Kincaid readability test to analyze text and
The analyze module uses the Flesch–
This score is then converted into a
produce a readability score.
grade-based readability category.
FUNCTIONS
compute_readability(text)a text string of any length and prints out a
This function takes
grade-based readability score.

Don’t skip the test of
running cory_analyze.py,
count_sentences(text)
in a string of text
the number
n countsyour
asThis
itfunctio
imports
new of sentencesexclama
tion mark as
using period, semicolon, question mark and
ls.
termina
analyze.py file.
count_syllables(words) a list of words and returns a total
This function takes
list.
count of syllables across all words in the
)
count_syllables_in_word(word
in the form of a string
This function takes a word
es. Note this function is
and returns the number ofbesyllabl
100% accurate.
a heuristic and may not
output_results(score) a Flesh-Kincaid score and prints the
This function takes
corresponding reading level.

Here we're asking for help on
the entire analyze module.

Python 3.6.0 Shell
>>> help(analyze.compute_readabi
lity)
Help on function compute_readabil
ity in module analyze:
compute_readability(text)
This function takes a text stri
ng of any length and prints out
a
grade-based readability score.

You can also ask for help on a
specific function in a module.

you are here 4  303

further thinking on the analyze code

Nice job! I was quickly
able to use the analyze
module, especially with the help
of the great documentation!

If you think about how another programmer might want to use the analyze
module, are there ways you might restructure it? For instance, might another
programmer want to get direct access to the score value? Think through how you
might refactor the code.

304  Chapter 7

modules, methods, classes, and objects

Exploring other Python modules
Now that you have a better understanding of modules and how to use Python’s help system,
there are plenty of interesting modules out there for you to explore. We’ll be tackling some of
the more interesting ones throughout the rest of the book, and you’ll find a few discussed in the
appendix as well. Here’s a few that are popular just to get your creative juices flowing...

We use
Python’s datetime
module in everything we do;
we’re a shipping company, and the
datetime module gives us all the tools
we need to create, manipulate, and
compare dates and times.

We use the requests
module to connect to
and retrieve web pages
that we monitor.

I love Python’s turtle module. I
teach children and beginners, and
it gives them a great way to create and play
with graphics on the screen. It’s inspired
by Seymour Papert’s work at MIT.

In my computational
physics group, we rely
on the math module heavily!

In our
health care app,
the entire interface
is written using the
Tkinter module, which
gives you everything you
need to build desktop
user interfaces.

I like programming
turtles with Python.

you are here 4  305

the turtle module

Wait, did someone say “turtles”?!
Ah, one of our favorite topics, because, as you’ll see, they’re a great way to play with
computation, and they are built right into Python. All you need to do is import the turtle
module and then you’re all ready to create your own turtles. But, before you start creating
turtles, let’s take a look at what a turtle living in Python-land looks like:

On the grid every turtle has
a location, or x, y coordinate.

In Python, turtles live on a grid. The center
of the grid is at coordinate 0,0.
Turtles in Python
also have a pen
they can draw with.

Their pen can be up or
down, and have a color
and size (thickness).

Every turtle has a heading
or direction it is pointing.

Think of these as the
turtle's attributes.
A turtle can
move forward,
back, or to a
specific grid
location.

A turtle can also
turn right or left,
which changes a
turtle's heading.

If the turtle's pen
is down, then it
leaves a mark when
it moves.

Think of these as the
turtle's behaviors.
306  Chapter 7

Turtles can also
take on different
shapes, like this
turtle shape, or a
circle, or an arrow
and so on.

modules, methods, classes, and objects

Turtles,
really? You started
the book with recipes and
now we’re learning about
turtles?

Sorry, were you expecting an
MIT-level treatment? Well, guess
what? That’s exactly what you’re getting!
Turtle graphics were invented at MIT (the
Massachusetts Institute of Technology) by
pioneering computer scientist Seymour
Papert in the 1960s. Since that time, turtle
graphics have had a big influence on many
programming languages that followed and
they’ve also helped to educate computer
scientists and mathematicians alike (not
to mention lots of kids). We’ll ask you to
withhold judgment until you’ve finished
this chapter (and a few other chapters in
this book), when we think you’ll be quite
pleased with what you’ve accomplished
using turtles.

you are here 4  307

how to create a turtle

Creating your very own turtle
What are we waiting for? Let’s create a turtle (or two). First
we’re going to need to import the turtle module.

Don’t start entering
code yet; we’ll do
that in just a bit.

import turtle
And then we can create a turtle like this:

turtle
module

The dot
operator

What looks like a function
call, only we usually don’t name
functions with uppercase letters

slowpoke = turtle.Turtle()
And let’s assign the new turtle
to the variable slowpoke.

This creates a new turtle.

Now let’s make slowpoke do something:

One of the turtle's
behaviors An argument, like
when making a
slowpoke.forward(100) function
call

A variable referencing The dot
operator
our new turtle

This is one of
slowpoke's behaviors.

This tells our slowpoke turtle
to move forward 100 units.

So slowpoke moved straight 100 units and drew a line
with its pen as it did. But did you notice slowpoke doesn’t
look much like a turtle? Let’s fix that by setting one of its
attributes and rerunning the code.

slowpoke.shape('turtle')
This sets the slowpoke turtle's
shape attribute to be ‘turtle'.
That's better!

308  Chapter 7

When you create a
turtle, this window
will pop up.

modules, methods, classes, and objects

Turtle lab
Okay, now it’s time for you to join in: let’s write a little code to see if we can do something
interesting with Python turtles. Go ahead and put this code in a file called turtle_test.py:

import turtle
slowpoke = turtle.Turtle()
slowpoke.shape('turtle')
slowpoke.forward(100)
slowpoke.right(90)
slowpoke.forward(100)
slowpoke.right(90)
slowpoke.forward(100)
slowpoke.right(90)
slowpoke.forward(100)
slowpoke.right(90)
turtle.mainloop()

Again, first we import the turtle module.
Next let's create slowpoke as we did before. We're
going to talk more about how turtles get created in a
bit, but for now just know that this line of code gives
you your very own turtle to play with.
Next we make sure the turtle’s shape is that of a turtle,
and not the default triangle shape we saw earlier.
Then let’s tell the turtle to go forward 100 steps using
its forward function, and then use the right function to
have it turn right 90 degrees.

Then let’s make it go forward and
turn three more times.
We're going to talk more about this line later in the book, but it basically
allows the turtle module to monitor everything going on in the window, including
shutting everything down when you click the close button. You just need to put
it at the end of your code.

With that, go ahead and give this a test run and you’ll see your
turtle move forward and turn right four times, each time leaving
a trail behind it to form the shape of a square.

Here’s what we got!
If you’re not
seeing this window
and you have no
errors in the shell,
look behind your
Python windows.
On some systems
the turtle
window does not
automatically
appear on top of
the others.

Don’t name your
file tur tle.py
Careful, if you name
your test file the same
name as a module
you’re importing, then you’re going
to
run into trouble: when Python goes
looking for the turtle module it’s
going to find your file first. So, just
make sure you don’t name your
Python files the same names as
common modules, especially ones
you’re importing.

you are here 4  309

more about turtles

Q:
A:

So when we tell a turtle to
turn, we do it with degrees?

Correct. For instance, turning
right 360 degrees would mean it would
turn all the way around clockwise.
Turning 90 degrees would turn a
quarter of a turn, and so on.

Q:

When the turtle is going
forward, what does the argument
100 mean?

A:

100 units. What’s a unit? A pixel
on your screen. So turtle.forward(50)
would move a turtle 50 pixels in the
direction it is heading.

Q:

Why did we have to set the
shape to be a turtle? I thought it
was a turtle!

A:

For historical reasons, by default
the turtle displays a shape similar to a
triangle. You can also set it to shapes
like a square, a circle, an arrow, or
even your own images. But what could
be more fun than the shape of a real
turtle?

Hey, it’s Chapter 7 already, so we’re confident you’re in good shape to
take the turtle square code (on the previous page) and get it wrapped up
in a nice function; call it make_square, which will take one parameter, a
turtle. Write your code here and see how much you can clean things up by
removing any duplicated code. Of course you’ll find our version at the end
of the chapter if you need any hints.

Make sure you take a look at
our version, because that’s
what we’ll work from on the
next few pages.

310  Chapter 7

modules, methods, classes, and objects

Adding a second turtle
How about we give slowpoke a friend? Let’s add another turtle
to our code:

import turtle
slowpoke = turtle.Turtle()
slowpoke.shape('turtle')
pokey = turtle.Turtle()
pokey.shape('turtle')
pokey.color('red')
def make_square(the_turtle):
for i in range(0,4):
the_turtle.forward(100)
the_turtle.right(90)
make_square(slowpoke)
pokey.right(45)
make_square(pokey)

In this code, slowpoke draws a square
as before, but pokey turns slightly
and then draws its own red square.

Here’s the nice new code
we generalized in the last
Sharpen exercise.

Create a second turtle
and assign it to the
variable pokey. We're
going to set pokey's shape
attribute to a turtle as
well, and also set its color
attribute to red.

Then let's turn pokey slightly, 45
degrees to the right.

turtle.mainloop()

And then pass pokey to make_square.

If the make_square function looks
new to you, make sure you go through
the answer to the last Sharpen your
pencil on the previous page.

Let’s use your newfound superpower of creating and using
functions to take this further. Let’s write a function that uses
make_square and see what kind of interesting graphic we can
generate:
import turtle
slowpoke = turtle.Turtle()
slowpoke.shape('turtle')
slowpoke.color('blue')
pokey = turtle.Turtle()
pokey.shape('turtle')
pokey.color('red')

To make it interesting, let's change
slowpokes’ pen color to blue.

Rest of the code continued
on the next page...

you are here 4  311

creating graphics with a turtle

def make_square(the_turtle):
for i in range(0,4):
the_turtle.forward(100)
the_turtle.right(90)

Let’s add a new function
called make_spiral.

def make_spiral(the_turtle):
for i in range(0, 36):
make_square(the_turtle)
the_turtle.right(10)
make_square(slowpoke)
pokey.right(45)
make_square(pokey)
make_spiral(slowpoke)
pokey.right(5)
make_spiral(pokey)
turtle.mainloop()

make_spiral is going to call make_square
36 times, and then, each time, have the
turtle turn an additional 10 degrees.

Get rid of these calls
to make_square.
Now we’re going to call
make_spiral intead of
make_square.

pokey is going to turn right 5
degrees before making a spiral.

Update turtle_test.py and give it a test drive. Is this the output
you expected?

A Test Drive

Here’s what we got! Are you
having Spirograph flashbacks?
Feel free to play with all the attributes
and parameters and see what you can come
up with. We’ve also got some more turtle
experiments for you on the next page.
312  Chapter 7

Notice that each turtle
has its own position, heading,
color, and so on.

modules, methods, classes, and objects

MORE TURTLE EXPERIMENTS
We’ve got even more turtle experiments for you below. Take a look at each one, guess what it
does, and then run it to see if you were right. Change a few values; how does the output change?

Experiment #1
for i in range(5):
slowpoke.forward(100)
slowpoke.right(144)

What happens if you
change this number?
Or this one?

Experiment #2

Just take your
turtle_test.py
file and delete
everything but
the first three
lines (and the
final mainloop
line), then add
in this code in
the middle.
Draw your output here.

slowpoke.pencolor('blue')
slowpoke.penup()
slowpoke.setposition(-120, 0)
slowpoke.pendown()
slowpoke.circle(50)
slowpoke.pencolor('red')
slowpoke.penup()
slowpoke.setposition(120, 0)
slowpoke.pendown()
slowpoke.circle(50)

Experiment #3

We’re using some new
turtle functions
here: we’re setting
the color of the
pen, lifting it up,
moving to a position,
and putting it down
before drawing a
circle.
What happens if
you remove the
calls to penup?

def make_shape(t, sides):
angle = 360/sides
for i in range(0, sides):
t.forward(100)
t.right(angle)
make_shape(slowpoke,
make_shape(slowpoke,
make_shape(slowpoke,
make_shape(slowpoke,

3)
5)
8)
10)

Try some more
values, like 1, 2?
50?
As usual, you’ll find our
answers in the back of
the chapter.
you are here 4  313

understanding turtles

What are turtles, anyway?
Let’s return to the code where we first created our turtle. At first glance it looks like
we called a function named Turtle that is located in the turtle module:

turtle
module

The dot
operator

What looks like a
function call

slowpoke = turtle.Turtle()
So does the turtle module have a function that creates turtles? And what is a turtle,
anyway? We know about integers and strings and lists and booleans, but what’s a
turtle? Is it a new type? To dig a little deeper, we could always ask Python for help:

Python 3.6.0 Shell
>>> import turtle
>>> help(turtle.Turtle)
Help on class Turtle in module turtle:

Ask for help on the
turtle.Turtle function.
This doesn’t look a
function.

class Turtle()
|
| When a Turtle object is created or a function derived from some
| Turtle method is called a TurtleScreen object is automatically created.
|
|
|
|
|
|
|
|
| Methods defined here:
|
| __init__(self, shape='classic', undobuffersize=1000, visible=True)
|
Initialize self. See help(type(self)) for accurate signature.

Class?
Methods?

Object?

Relax, because this isn't the
clearest description in the
world for us, either.

So reading this not-so-clear docstring, you’d think that Turtle is either a class or
an object, or both (whatever those are), and it has something called methods. And
you’d be right. But is it a class or an object? Well, although we haven’t made much
of a point of it yet, we should tell you that Python is a very object-oriented language—
as are most modern languages—and given how far you’ve come in your coding,
it’s about time you learn what objects, classes, and methods are.

314  Chapter 7

We trimmed this down to
save some trees (or electrons
if you’re reading the digital
copy).

modules, methods, classes, and objects

What are objects?
Of course you intutively know about objects; they’re all
around you. Cars, iPhones, radios, toasters, kitchen appliances,
you name it. And one thing that all these objects have in
common is they have some internal state and they have some
behavior. Take a car, for example—it has state:
• Make
• Model
• Fuel level
• Speed
• Mileage

Just a few off the top
of our heads; we’re sure
you can think of even
better ones.

• Engine state (on/off )

And it has behavior. A car can:
• Start
• Turn Off

’57 Chevy

• Drive
• Brake

With programming, objects are no different. The whole point
of software objects is that we can bundle state and behavior together.
Think about a Boolean value: it has state but no behavior.
Think about a Python function: it has behavior but no state.
With Python objects, we can have both working together. For
example, when you start the car, the engine state is
changed from off to on. Likewise, if the brake behavior is
applied, then the speed state will naturally decrease.
So what’s the big deal? We could do all that with functions
and variables, right? But when you start thinking in terms of
objects, you can approach solving computational problems at
an even higher level—you can think about programming as
composing a set of objects and managing their interactions,
instead of herding a large number of variables and functions.
Our turtles are a good example—we could write code to
manage locations and colors and coordinates to draw graphics
(always a difficult task). Or we can use our turtle objects, which
inherently keep track of much of this state internally, freeing
us to just think about the larger issues, like how to get two
turtles to draw a spiral together. It’s a simplistic example, but
you have to start somewhere when thinking about objects.

State and
behavior bundled
together in an
object

make: 'Chevy'
model: 'Bel Air'
fuel: 8
speed: 0
mileage: 1211
engine_on: False
def start():
def turn_off():
def brake():

Objects aren't
unique to Python;
practically all
modern languages
provide objects.

make: 'Mini'
model: 'Cooper'
fuel: 2
speed: 14
mileage: 43190
engine_on: True
def start():
def turn_off():
def brake():

Mini

make: 'Pontiac'
model: 'Fiero'
fuel: 10
speed: 56
mileage: 196101
engine_on: True
def start():
def turn_off():
def brake():

Fiero

make: 'DeLorean'
model: 'DMC-12'
fuel: 6
speed: 88
mileage: 10125
engine_on: True
def start():
def turn_off():
def brake():

DeLorean
you are here 4  315

understanding classes

Okay, what’s a class then?
If we’re going to have lots of objects, say lots of turtles, we want all our turtles to share common behaviors
(after all, who wants to reinvent the wheel every time we need a turtle to move forward), but we also want
each turtle to have its own state (because if every turtle was in the same location, had the same heading,
and had the same color, we couldn’t get much interesting work out of them). A class gives us a template or
blueprint for creating objects of the same type.

Here’s a handy way
to diagram a class.
Turtle

color
xcor
ycor
heading
forward()
backward()
turn()

State

From one class we can
create many different
objects, with each having
its own state.

Behavior

penup()
pendown()

Q:
A:

Okay, but what is the point of
creating all these objects from classes?
Without classes and objects you’ll
always be stuck solving any problem in
terms of the Python basic types, like strings
or numbers or lists, and so on. With classes
and objects, you can use types that are
higher level and closer to the problem you’re
trying to solve. For example, if you were
creating a fishing game, having fish objects
and a pond object would be a lot easier than
managing a whole collection of variables and
functions.

316  Chapter 7

A class is a blueprint for making objects. A
class tells Python how to make an object of that
particular type. Each object made from that class
can have its own values for its state; for example,
you might use the Turtle class to make dozens of
different turtles, and each turtle has its own color,
size, shape, location and pen settings (up or down),
and so on.
That said, all turtles created from the same class
share the same behaviors, like turning, going
forward and backwards, and controlling the pen.

shape()

Turtles have different
state, but share common
behaviors.

A class is not an object, it’s
used to construct them.

Q:
A:

Is a class like a type?

Exactly. Think of each class as a
Python type, just like the string and list and
number types that you already know.

Q:
A:

So a class is a blueprint for making
many objects. Can I make my own class?

You sure can, and that is the beauty
of object-oriented programming: you can
extend Python (or any language) with
your very own classes, or even extend the
functionality of anyone else’s classes.

In this chapter we’ll be looking at how to use
pre-existing classes, and in Chapter 12 we’ll
start making our own.

Q:

Objects have a lot in them (data
and functions). How does a variable like
slowpoke actually hold an object?

A:

Remember when we talked about how
a variable is assigned to a list? We said that
a variable holds a reference to where the list
is stored, like a pointer to the list. Objects
are assigned to variables in the same way;
the variable holds a reference to the object
in Python’s memory, not the object itself.

modules, methods, classes, and objects

A class tells us what an object knows
and what an object can do
On the previous page we saw a diagram for sketching out a class.
Let’s take a little closer look at what it provides. A class diagram
tells you two things (for starters): what a object knows and what
a object can do.

Here’s our Turtle
class diagram again.
Turtle
color
xcor
ycor
heading
forward()
backward()
turn()
penup()

The things the
Turtle object knows
(the state).

The things the
Turtle object does
(the behavior).

Fill in what a radio
object might need to
know and do.

pendown()
shape()

The things the object knows about itself are called its:

• attributes
The things the object can do are called:

• methods
Attributes represent an object’s state (its data), and each object has its own
attribute values. Attributes are similar to local variables, only they
live in an object. Also like variables, attributes can be assigned
to any of the Python types you’re already familiar with. You’ll
also hear the term instance variable around coding circles. An
instance variable is the same as a Python object attribute. In fact,
anytime you hear the word instance, just substitute the word object.
So, an instance variable is the same as an object variable, which
is the same as a Python attribute.
Things an object can do are called methods. Think of methods as
functions that belong to an object. The difference between a
method and a function is that methods are typically getting,
setting, altering, and making decisions based on an object’s
attributes.

Radio

attributes

methods

you are here 4  317

python classes and objects

How to use objects and classe s
In this chapter we’re learning how to use objects and classes. As it turns out, there
are lots of classes out there, written by other developers, all ready for you to use.
To use them all you really have to know is what an object can do (in other words,
its methods) along with any attributes you might want to make use of. Of course
you also need to know how to create an object (otherwise known as an instance)
before you can use one. We’ve created a couple Turtle objects, but let’s take
another look at how to do this:

In Chapter 12 you’re going to
learn how to make your own
classes and objects.

dot
turtle The
operator
module

The Turtle
class name
And this calls the Turtle class’s
internal initialization method, which
slowpoke = turtle.Turtle()
creates the new object and gets any
initial values set up for you.
In the object-oriented
The result of calling the initialization method is a brand new
world,
we don’t just
Turtle object created from the Turtle class blueprint. The new
create
new objects, we
Turtle object is assigned to the variable slowpoke.
instantiate them. And
So there’s a lot going on here in this little line of code. Let’s talk through it. The
we call each object an
first thing to know is that we’re accessing the Turtle class from the turtle
instance.
module. That’s the reason for the dot notation here: the dot has nothing to do with
attributes or methods in this statement.

turtle.Turtle

Notice that, by convention, class names start
with an uppercase letter. We’ll see some
exceptions to this later in the chapter.

Here’s where we’re getting the Turtle
CLASS from the turtle module.
Next, we’re invoking the class like it was a function. What’s going on here?

turtle.Turtle()

Invoking the class like a function?

Here’s what’s going on: every class has a special method known as its constructor. The
constructor gets the object all set up with whatever default attribute values it needs
(among other things it might do). The constructor always has another important role:
it always returns the newly created object (again, we call it the instance) to you.

In the object-oriented world,
we call these initialization
methods constructors. Any
time you hear “constructor,”
just think “the method that
initializes the object.”

So when the object is created, initalized, and returned, it is assigned to the variable
slowpoke.

slowpoke = turtle.Turtle()

318  Chapter 7

Returns a new object that
is instantiated from the
Turtle blueprint

Constructors can take
arguments too, although we’ll
see these a bit later

modules, methods, classes, and objects

What about those methods and attributes?
As you’ve already seen, once you have an object in hand, you are free to call its
methods:

To call a method, start with the name of an object,
followed by a dot, and then the method name—in other
you call them just like you call functions, only you
words,
slowpoke.forward(100)
precede the function with an object name (and a dot).
You always need an object to call a method
on. Calling a method without an object
doesn’t compute (which object’s state is the
method supposed to act on?).
slowpoke.turn(90)

So what about attributes? We haven’t seen any code where we’re getting or
setting an attribute’s value. To access an attribute in an object, you use dot
notation. Say the Turtle class had an attribute named shape (it doesn’t, but
hold that thought). Then you’d access or set the value of the shape attribute like
this:

slowpoke.shape = 'turtle'
print(slowpoke.shape)

If this were valid code,
this would print ‘turtle’.

You can set or get the value of an
object's attribute using dot notation
(it's the same syntax as accessing a
variable in a module).

Okay, but as you’ve already seen we’ve been using the shape method to set a turtle’s
shape. So why doesn’t the Turtle object have a shape attribute? Well, it could, but
there is a common strategy used in object-oriented programming where we rely on
a method to get or set an attribute’s value. The reason stems from an idea called
encapsulation that we’ll talk more about in Chapter 12. Encapsulation often gives the
object developer more control over an object (than just letting any code change an
attribute’s value). Again, we’ll return to this point in Chapter 12.
For now just know that many of the attributes you’ll want to get your hands on are
accessed through methods, rather than directly. Here’s an example: to get or set the
state of the shape attribute, we use a method, shape, instead:

slowpoke.shape('circle')
print(slowpoke.shape())

We can call the shape method to change
slowpoke's internal shape attribute to a circle.

And we can call the shape method without
arguments to get its current value.
you are here 4  319

other classes and objects in python

Seeing classes and objects everywhere
With that background behind us, let’s take a new look at the Python world—we
did say Python was a very object-oriented language, and in fact objects are all
around you. Check this out:

A list is
actually a
Python class.

Here we're calling the list
constructor to instantiat
e a new,
empty list.

my_list = list()
my_list.append('first')

…and the reverse
method, which you
haven't seen before.

print(my_list)

All these methods change the
list object's internal state.
Here’s another class that may look familiar. Check this one out:

It turns out strings are really classes. Here's how
you'd instantiate one with the constructor.
greeting = str('hello reader') Notice this constructor
takes an argument; we’ll
shout = greeting.upper()
see more as the book
print(shout)
progresses.
Calling one of the
string methods.
Because lists and strings
and floats are built into
What about this one?
Python, you don’t need
That’s right,
to explicitly call the
another class…
classes' constructor like
we’re doing here. Python
handles it for you behind
pi = float(3.1415)
the scenes.
is_int = pi.is_integer()
print(pi, is_int)

320  Chapter 7

['second', 'first']
>>>

Here we're calling
the append method…

my_list.append('second')
my_list.reverse()

Python 3.6.0 Shell

…and another
method.

Remember we said class names
start with an uppercase letter?
Unfortunately, this isn’t the case
for some Python built-in classes
(for historical reasons). So just know
that some built-in classes start with
a lowercase letter, and any other
classes you run into should begin with
an uppercase letter.
Python 3.6.0 Shell
HELLO READER
>>>

Python 3.6.0 Shell
3.1415 False
>>>

modules, methods, classes, and objects

Are you telling
me we’ve been using objects
the whole time and you’re just now
telling us?

Pretty much... if you forget about
modules for a moment, anytime you see a
variable followed by dot notation, you can
bet that what follows the dot is either an
object attribute or a method. Not only that,
but almost everything in Python, including
the built-in types, is actually a class. The
reason it didn’t initially appear so was
because Python goes to a lot of trouble to
make things easy—for instance, when you
type:
my_list = []
to create an empty list. Behind the scenes,
Python is effectively rewriting that as:
my_list = list()
So until you start using methods on your
lists, like:
my_list.append(42)
you really have no reason to think you’re
dealing with an object.
But now that you know the truth, objects
are all around you in Python. So, it’s
probably time we get back to using them.

you are here 4  321

building a turtle game

Get ready for some turtle races
We’ve seen that the selling point of an object is that each object maintains
its own state, but objects also benefit from sharing their behavior (their
methods) with all other objects of the same class.
You already know that each turtle is its very own object—its very own
independent instance, complete with its own set of attributes. That means
each turtle has its own color, position, heading, and shape (to name a few
attributes). Let’s leverage that by creating a little game. Would you believe
racing turtles?

We’re going to create a bunch of turtles, each with
its own color and position, and then let them duke
it out, racing across the screen. Place your bets!

The starting line for the
turtles is going to be on the
left side the screen.
322  Chapter 7

And they’ll race toward the right
side of the screen. The first
turtle across the finish line wins.

modules, methods, classes, and objects

Planning the game
You’re going to see that treating our turtles as objects will make this game a lot easier to
implement. Without them we’d presumably have to create and update a lot of variables
to track the turtles and their positions, which would be messy business. But with turtle
objects, we’re just going to use the turtle’s methods to move them on the screen, relying
on each turtle to track its own state.
So, let’s hash out a little pseudocode for how this racing game is going to work:

First we need to create our
turtles, and assign each one
to a different color and
position on the starting line.
Each turtle is a different
instance of a Turtle, so we
can do this!

1

Set up game.

2

Start the race.

Create a few turtles, each with
its own color and position on the
starting line.

Set a variable winner to False.
While winner is false:

A

For each turtle:

i

Pick a random amount to move
forward, say between 0 and
2 pixels.
Move forward.

ii

Check to see if turtle’s
position is across finish line.
If so:

a

3

Set winner to True.

Game finishes.
Announce the winner’s name!

We’re going to let the race run until a turtle wins by
crossing the finish line. But we need a way to know if
a turtle has won. Let’s create a variable called winner
that is initially set to False, and then set it to True
when a turtle crosses the line.
Now we’re just going to cycle over each
turtle, over and over, letting it make a
move—that is, until one of them wins.
To make a move, we’ll generate a random
number between 0 and 2 and move the
turtle by that amount—and we’ll do this
for each turtle.
Each time we move we need to check
to see if the turtle crossed the
finish line. And if so, we’ll set winner
to True.
Once winner is set to
True, the loop stops and
we’ll announce a winner.

you are here 4  323

writing some setup code

Let’s start coding
Let’s first set up the game board and all the turtles, and then we’ll move on
to implementing the game logic. And, just to get the coding kicked off, we
already know two modules we’re going to need, namely the turtle module
and the random module, so let’s import those into our code. We’re also going
to use one global variable to hold the turtle racers, so let’s add that too. Go
ahead and create a new file called race.py and enter this code:

import turtle
import random
turtles = list()

We’ll need these modules.
And we’ll use a list to hold all our turtles.
Just to rub in the constructor syntax,
we’re creating an empty list by calling the
list constructor. We could have used the
shorthand [] like we have in the past.

Setting up the game
Referring to our pseudocode, most of the setup consists of creating the set
of turtles, each with its own attributes. To do that we need some idea of the
turtles we want to create. How about these turtle objects:

color: 'red'

color: 'pink'

color: 'cyan'

ycor: -20

ycor: 20

ycor: 0

Note, xcor and ycor are the
x and y coordinates of the
turtle in the window.
Turtle
color
shape
xcor
ycor
heading
forward()
backward()
turn()
penup()
pendown()

Pink Turtle

Red Turtle

color: 'orange'
ycor: -40

Orange Turtle

Cyan Turtle

color: 'gray'
ycor: 40

Gray Turtle

Each turtle has a name, a col
and a ycor, which is its vertical
position on the start line. Youor,
'll
see
course the turtle has all its oth how this is used in a bit. Of
the ones we want to set up. er attributes too, but these are

324  Chapter 7

These are objects
instantiated from
the Turtle class.

modules, methods, classes, and objects

Writing the setup code

1

Now let’s write the code to instantiate and set up those turtles. If
you look at the turtle attributes on the previous page, we need a
place to store those so we can intialize our turtles with those values.
To do that, let’s create a list that contains the values of each turtle’s
attributes—actually we’ll need two lists, one for the y positions
and one for color. Then we’ll instantiate each turtle, and set its
attributes to the appropriate values. Of course, let’s put all this in a
setup function like this:

import turtle
import random

def setup():
global turtles
startline = -480

Create a few turtles, each
with its own color and position
on the starting line.

In the next chapter we’re going
to discover a data structure
that would work better than
two lists for storing these values.

Let's define a function called
setup to create and position our
turtles.

turtles = list()

Set up game.

The startline variable just holds the
x coordinate of the starting line;
you'll see how this is used below.

turtle_ycor = [-40, -20, 0, 20, 40]
turtle_color = ['blue', 'red', 'purple', 'brown', 'green']

Here are the initial
values we need for
each turtles attribute,
stored in two separate
parallel lists.

Let's iterate over the number of turtles.
And for each we'll instantiate
a new turtle, make its shape a
turtle, and set its position on the
grid from our list.
set its color.

for i in range(0, len(turtle_ycor)):
new_turtle = turtle.Turtle()
new_turtle.shape('turtle')
new_turtle.setpos(startline, turtle_ycor[i])
new_turtle.color(turtle_color[i])
We'll also
turtles.append(new_turtle)
setup()
turtle.mainloop()

Don't forget to call
the setup function, and
remember we need our
turtle.mainloop.

Then let's add this new turtle
to the global list of turtles.
Remember the append method from
Chapter 4? It adds an item to an
existing list, in this case the list
turtles. We of course weren’t calling
them methods yet in Chapter 4.

A reminder that the
setpos method sets the
x and y position for
a turtle. Here we’re
moving the turtle to
the starting line.

you are here 4  325

testing the setup code

Not so fast!
It would be easy to blow through that code on the last page. And,
on the other hand, we could step you through every single line, but this
is Chapter 7, and you’re definitely up to the task of studying a little code.
So, given there’s a lot happening on the previous page, go back and
make sure you understand every line until you know exacty what the
code does. Then proceed...

A Test Drive

Update your code and let’s give this a test drive to see how it is progressing
before going further.

A step in the right direction, but this
looks a little weird. What happened?

Our turtles are the right color and
seem to be in the right position. But
remember, they start life at coordinate
0,0, in the center of the window; so it
looks like they drew lines on their way
from the center to the start position,
which isn't what we wanted. Let's add
some code to pull their pen up on their
way to the new position.

326  Chapter 7

You may find the default window size on
your machine is smaller than shown here. If
you see the turtles fly off the left side
of your window, adjust the width of your
window until you can see them.

And while we're talking about the look and feel
of the game, let's make the window a little
bigger and add a nice background too.

modules, methods, classes, and objects

Check out the updated fixes: we’re lifting up the turtle pen so it doesn’t draw on the way
to the starting position; we’re also enlarging the window and adding a background image.
Make these changes and try again. You’ll want to grab the file pavement.gif from the
import turtle chapter source files and place it in the same folder as race.py.

A Test Drive

import random

turtles = list()

Don't miss this tiny change; we're going to move the turtles
more to the left. Remember the center is at 0,0, so an x
coordinate of -620 is far left on the screen.

def setup():
global turtles
startline = -620
screen = turtle.Screen()
screen.setup(1290,720)
screen.bgpic('pavement.gif')

These three lines use another object we haven't
seen yet, the Screen object. With the Screen
object we can make the window bigger and add
a background image.

turtle_ycor = [-40, -20, 0, 20, 40]
turtle_color = ['blue', 'red', 'purple', 'brown', 'green']
for i in range(0, len(turtle_ycor)):
Let's pull up the
new_turtle = turtle.Turtle()
new_turtle.shape('turtle')
pen before we move.
new_turtle.penup()
new_turtle.setpos(startline, turtle_ycor[i])
new_turtle.color(turtle_color[i])
new_turtle.pendown()
turtles.append(new_turtle)
And put it back down
setup()
turtle.mainloop()

Here's what we got! Now
we're talkin’. We've got
each turtle lined up on
the starting line ready
to go. Looks like we're in
great shape to start on
the actual game now.

For now you can probably
tell what this code is doing,
but we're going to explore
when the Screen object more in
we're done moving the turtle Chapter 11.
to the starting line.
Note the finish line is just part of
the background image graphic.

Think about what's going on here:
you've got five Turtle objects, all
with their own internal state,
including color and position.

you are here 4  327

implementing the game

Starting the race

2

Set a variable winner to False.
While winner is false:

Now it’s time to get these turtles racing. We’ve got some nice
pseudocode to guide us, so let’s just work through it:

2

Start the race.

A

For each turtle:

i

First we need to get the winner variable set up. To
do that let’s create a new function called race. The winner
variable is easily implemented as a Boolean initially set to
False. We’ll also be using the global turtles variable in this
function too, so let’s remember that in this code.
def race():
global turtles
winner = False

Move forward.

ii

a

def race():
global turtles
winner = False
finishline = 590
Now we are going to keep the game going until
there is a winner. To do that we’re going to start with a
while statement, which will loop until the variable winner is
set to True.

We’re going to endlessly
loop until the winner
variable is set to True.

while not winner:

A

Next we need to give the turtles a chance to move
forward. To do that we simply use a for/in statement over
our global turtles lists.
def race():
global turtles
winner = False
finishline = 590
while not winner:
for current_turtle in turtles:

328  Chapter 7

Check to see if turtle’s
position is across finish line.
If so:
Set winner to True.

The x value of 590 is located
about here on the grid.

We’re also going to add a local variable that holds the x
position of the finish line:

def race():
global turtles
winner = False
finishline = 590

Pick a random amount to move
forward, say between 0 and
2 pixels.

h
And each time througgoing
the while loop we’re
ery
to iterate throughchevance
turtle, giving it a across
to move forward
the screen.

modules, methods, classes, and objects

i

Now we use random numbers to move the turtles
forward. So let’s compute a number between 0 and 2 and

move the turtle forward that many units.

def race():
global turtles
winner = False
finishline = 590
while not winner:
for current_turtle in turtles:
move = random.randint(0,2)
current_turtle.forward(move)

ii

Compute a random number
between 0 and 2 and move
the turtle forward that
many units.

Finally we just need to check for a winner. Note that
we have a winner if a turtle’s xcor attribute is greater than
or equal to finishline, which is set to 590. So, let’s get the
turtle’s xcor attribute and do the comparison. If the turtle is
over the finish line, we set winner to True and announce the
winner.
def race():
global turtles
winner = False
finishline = 590

Let’s get the turtle's x coordinate to
see if it’s crossed the finish line. To do
that, we use the xcor method, which
returns a turtle’s x coordinate.
Compare it to the finish line.
xcor = current_turtle.xcor()
if (xcor >= finishline):
And if greater we
winner = True
have a winner.
winner_color = current_turtle.color()
Set the winner
print('The
winner
is',
winner_color[0])
to True.
while not winner:
for current_turtle in turtles:
move = random.randint(0,2)
current_turtle.forward(move)

Note the color method
returns two values, a pen
color and a fill color.
We're interested in the
first one, the pen color,
so we use the index 0 to
get it.

Use the color method
to get the winning
turtle's color.

3

Game finishes
Announce the winner’s name!

you are here 4  329

testing the game

We should be ready to race now. Get all the code additions made to your
race.py file. You’ll find the complete code below. So, place your bets and
give it a true test drive.

A Test Drive
import turtle
import random
turtles = list()

def setup():
global turtles
startline = -620
screen = turtle.Screen()
screen.setup(1290,720)
screen.bgpic('pavement.gif')

Remember the
amount each
turtle moves is
random, so your
mileage may vary.

turtle_ycor = [-40, -20, 0, 20, 40]
turtle_color = ['blue', 'red', 'purple', 'brown', 'green']
for i in range(0, len(turtle_ycor)):
new_turtle = turtle.Turtle()
new_turtle.shape('turtle')
new_turtle.penup()
new_turtle.setpos(startline, turtle_ycor[i])
new_turtle.color(turtle_color[i])
new_turtle.pendown()
turtles.append(new_turtle)

Brown wins!

def race():
global turtles
winner = False
finishline = 590
while not winner:
for current_turtle in turtles:
move = random.randint(0,2)
current_turtle.forward(move)
xcor = current_turtle.xcor()
if (xcor >= finishline):
winner = True
winner_color = current_turtle.color()
print('The winner is', winner_color[0])
setup()
Don’t forget
race()
call race!
turtle.mainloop()

330  Chapter 7

to

Python 3.6.0 Shell
The winner is brown
>>>

modules, methods, classes, and objects

Look at the code on the previous page. Say there is a photo
finish and two or more turtles cross the finish line during
the same iteration of the while loop. What happens? Who
wins? Do you think this is the correct behavior?

With object-oriented programming there’s a lot of new jargon flying around. In this game of
who does what, match each piece of jargon to its description.

class

Things an object knows about.

object

Blueprint for an object.

methods

Act of creating an object from a class.

instantiate

The behavior an object can do.

attributes

Created from the blueprint.

instance

Another name for an object.

you are here 4  331

more about objects

I feel like we’ve just only
started and the chapter’s
already over.

Don’t worry, this isn’t
the end of objects.

Remember you can use Python’s help on
a class, like help(Turtle). Or even on a
module, like help(turtle). Don’t forget to
import the turtle module first! We’ll be
getting experience with more classes and
objects ahead in the book as well.

You’re right, we’ve barely scratched the surface.
Object-oriented programming is a huge topic, one
that would fill this entire book, and this chapter
is just the beginning. Next time you encounter a
class in a module you’re going to know that you
can instantiate it using its constructor. You’re
going to know it has methods and attributes
that are accessible to you. And you’re going to
understand that each instance of any object you
create has its own attributes.
You’ve also gained a general awareness now that
all Python types are, in fact, classes, which is great
because in the remaining chapters we’re going
to meet a lot of new classes and objects. And in
Chapter 12 we’ll break the surface and explore
object-oriented programming, including topics like
how to create your own classes. We’ll see you when
you get there.
Oh, and this chapter’s not really over, anyway.
You’ve got a mystery to solve, some bullet points to
read, a crossword to do. But then you really should
move on—we don’t want things to get awkward.

332  Chapter 7

modules, methods, classes, and objects

Odd goings-on at the turtle races.
Since you’ve released your turtle racing code, something strange has
started to happen: the green turtle is always winning, and by a large
margin. The police are thinking someone has hacked the code. Can
you take a look and see what is going on?
import turtle
import random

turtles = list()

class SuperTurtle(turtle.Turtle):
def forward(self, distance):
cheat_distance = distance + 5
turtle.Turtle.forward(self, cheat_distance)

Whoa! Green is
really fast!

def setup():
global turtles
startline = -620
screen = turtle.Screen()
screen.setup(1290,720)
screen.bgpic('pavement.gif')

turtle_ycor = [-40, -20, 0, 20, 40]
turtle_color = ['blue', 'red', 'purple', 'brown', 'green']
for i in range(0, len(turtle_ycor)):
if i == 4:
new_turtle = SuperTurtle()
else:
new_turtle = turtle.Turtle()
new_turtle.shape('turtle')
new_turtle.penup()
new_turtle.setpos(startline, turtle_ycor[i])
new_turtle.color(turtle_color[i])
new_turtle.pendown()
turtles.append(new_turtle)

def race():
global turtles
winner = False
finishline = 590

Study the code carefully;
what has changed? What
does this new code do? You
won’t be able to fully solve
this mystery until Chapter
12, but do your best and
think through what this
code might be doing.

while not winner:
for current_turtle in turtles:
move = random.randint(0,2)
current_turtle.forward(move)

setup()
race()

xcor = current_turtle.xcor()
if (xcor >= finishline):
winner = True
winner_color = current_turtle.color()
print('The winner is', winner_color[0])

turtle.mainloop()

you are here 4  333

review of modules, methods, classes, and objects

 Modules are collections of Python
variables, functions, and classes.

 Some languages use the name instance
variables or properties for object attributes.

 Using the __name__ variable you can
determine if your code is being imported or
being run as the main program (by looking
for a value of "__main__").

 An attribute can be assigned any valid
Python value.

 In the Python Shell you can use the help
function to see documentation on functions,
modules, and classes.
 In your own code, add docstrings to supply
help for programmers using your code.
 You’ll find plenty of Python modules
to explore in the areas of math, user
interfaces, interacting with web services,
dates and time, and pedagogy, to name a
few.
 The turtle module provides an
implementation of a turtle graphics system,
originally developed at MIT for teaching.
 With turtle graphics, turtle objects live on a
grid and can move and draw.
 Turtles are Python objects and include data
and behavior.
 We call the data in Python objects
attributes.

334  Chapter 7

 The behavior in Python objects is known as
methods.
 A method is a Python function that belongs
to an object.
 You can access attributes and methods by
using the dot notation on an object.
 Objects are created from classes. Classes
supply a blueprint to create objects from.
 When we create a new object, we say that
we instantiate it.
 An object instantiated from a class is known
as an instance.
 An object is instantiated using a constructor
method, defined in the class.
 Constructors do all the setup and
initialization needed for an object.
 All types in Python are classes, including
numbers, strings, lists, and so on.

modules, methods, classes, and objects

Mod coding cross
1

2

1

3

4

Get more modular as you do the crossword.

2

3

6

4
5

7

6

5

7

8

9
8

10

10

9

11

11

12
13

12

14

13

14

15
17

15

16
16

17

Across

Across
1. Behavior of object.

Behavior
object.
3.1.Who
wonofthe
first race?
Who wonname
the first
6.3.Another
forrace?
instance variable.
Another
name for
instance variable.
8.6.Unit
of turning
a turtle.
Unit of turning
a turtle.
11.8.Method
that sets
up object.
11.
Method
that
sets
12. Data of object. up object.
12.Command
Data of object.
14.
to move ahead.
14. Command to move ahead.
16. Function to get documentation.
16. Function to get documentation.
17.
Another word for object.
17. Another word for object.

Down

Down
2. Graphics system from MIT.

2. 4.
Graphics
system
MIT.
Another
wordfrom
for instance.
4. 5.
Another
forPython
instance.
Every word
type in
is one.
5. 7.
Every
type inan
Python
is one.
Creating
object.
7. 8.
Creating
anscience
object. fiction author.
Famous
8. 9.
Famous
science
Type of stringfiction
used author.
in help.
9.10.
Type
of string used in help.
Blueprint.
10. Blueprint.
13. Which turtle did the hacker hack?
13. Which turtle did the hacker hack?
15. Value of __name__ when main program is
15. Value of __name__ when main program is
(excluding
underscores).
runrun
(excluding
underscores).

you are here 4  335

exercise solution

Hey, it’s Chapter 7 already, so we’re confident you’re in good shape to
take the turtle square code (on the previous page) and get it wrapped up
in a nice function; call it make_square, which will take one parameter, a
turtle. Write your code here and see how much you can clean things up by
removing any duplicated code.

Notice we can pass
any turtle into
make_square, not just
slow_poke.

import turtle

import turtle

slowpoke = turtle.Turtle()
slowpoke.shape('turtle')

slowpoke = turtle.Turtle()
slowpoke.shape('turtle')

def make_square(the_turtle):
the_turtle.forward(100)
the_turtle.right(90)
the_turtle.forward(100)
the_turtle.right(90)
the_turtle.forward(100)
the_turtle.right(90)
the_turtle.forward(100)
the_turtle.right(90)
make_square(slowpoke)
turtle.mainloop()

Define a make_square
function.

make_square(slowpoke)
turtle.mainloop()

And make sure you call
the function.

Step 1: get the
code in a function
and use it.

336  Chapter 7

def make_square(the_turtle):
for i in range(0,4):
the_turtle.forward(100)
the_turtle.right(90)

Step 2: look at all that
duplicated code in our function
let’s just iterate over the
forward/right calls four times.

We don’t
need
all that
duplicate
code, we
can just
iterate
over it
four times.

modules, methods, classes, and objects

MORE TURTLE EXPERIMENTS SOLUTIONS
We’ve got even more turtle experiments for you below. Take a look at each one, guess what it
does, and then run it to see if you were right. Change a few values; how does the output change?

Experiment #1
for i in range(5):
slowpoke.forward(100)
slowpoke.right(144)

Here’s what we
got. Did you get
anything cool by
tweaking the various
variable values?

Experiment #2
slowpoke.pencolor('blue')
slowpoke.penup()
slowpoke.setposition(-120, 0)
slowpoke.pendown()
slowpoke.circle(50)
slowpoke.pencolor('red')
slowpoke.penup()
slowpoke.setposition(120, 0)
slowpoke.pendown()
slowpoke.circle(50)

Experiment #3
def make_shape(t, sides):
angle = 360/sides
for i in range(0, sides):
t.forward(100)
t.right(angle)
make_shape(slowpoke,
make_shape(slowpoke,
make_shape(slowpoke,
make_shape(slowpoke,

3)
5)
8)
10)

you are here 4  337

exercise solution

Fill in what a radio object might need to know and do.

There are many reasonable
answers for this. Here’s
what we came up with.

Radio

attributes

frequency
volume
power

methods

turn_on()
turn_off()
tune()
set_volume()

SOlUTion

With object-oriented programming, there’s a lot of new jargon flying around. In this game of
who does what, match each piece of jargon to its description.

class
object
methods
instantiate
attributes
instance
338  Chapter 7

Things an object knows about.
Blueprint for an object.
Act of creating an object from a class.
The behavior an object can do.
Created from the blueprint.
Another name for an object.

modules, methods, classes, and objects

Mod cross Solution
1

M E

2

3

T H O D

B

4

R O W N

U
6

R
A

T

T R

L
E

10

9

7

I

O

A
T

T R

I

S

R

A

S

I

T

N

I

T

G

17

I

J

L
A

T

T

C

T E

S

S

A

U

C

8

S
D E G R E

E

S

O
C

C O N S

L

12

B

N
D

11

C

5

B

T R U C

T O R
O

B

U

14

R

13

O

G

F O R W A

15

O
N S

T E

16

M

T A

R D
E

H E

N C E

L

P

N

I
N

Across

Down

1. Behavior of object. [METHOD]
3. Who won the first race? [BROWN]
6. Another name for instance variable.
[ATTRIBUTE]
8. Unit of turning a turtle. [DEGREES]
11. Method that sets up object.
[CONSTRUCTOR]
12. Data of object. [ATTRIBUTE]
14. Command to move ahead. [FORWARD]
16. Function to get documentation. [HELP]
17. Another word for object. [INSTANCE]

2. Graphics system from MIT. [TURTLE]
4. Another word for instance. [OBJECT]
5. Every type in Python is one. [CLASS]
7. Creating an object. [INSTANTIATION]
8. Famous science fiction author.
[DOCTOROW]
9. Type of string used in help. [DOCSTRING]
10. Blueprint. [CLASS]
13. Which turtle did the hacker hack?
[GREEN]
you
are here 4  339
15. Value of __name__ when main program is
run (excluding underscores). [MAIN]

8

recursion and dictionaries

Beyond Iteration and Indices
It’s time for a
few new tools in
that toolbelt.

It’s time to take your computational thinking up a notch.
 nd this is the chapter to do it: we’ve been happily coding along with an iterative
A
style of programming—we’ve created data structures like lists and strings and
ranges of numbers, and we’ve written code to compute by iterating over them. In this
chapter we’re going to look at the world differently, first in terms of computation, and
then in terms of data structures. Computationally we’ll look at a style of computing
that involves writing code that recurs, or calls itself. We’ll expand the kinds of data
structures we can work with by looking at a dictionary-like data type that is more like
an associative map than a list. We’ll then put them together and cause all kinds of
trouble. Be forewarned: these topics take a while to settle into your brain, but the
effort is going to pay off in spades.

this is a new chapter   341

thinking about computing differently

A different way to compute
It’s time for some mind-bending activity—you’ve been thinking
about the same, iterative style of programming for too long. So
let’s expose your brain to a totally different way of thinking about
solving problems.
Before we get there, though, let’s take a simple problem and think
it through the way we have throughout this book. For instance,
take a handy list of numbers you want to sum up; it could be
any numbers, say the number of marbles you and each of your
friends has in his or her pockets. Now, Python does have a sum
function that can be used to sum a list of numbers:

We’ve got a list with each friend’s
count of marbles.
marbles = [10, 13, 39, 14, 41, 9, 3]
print('The total is', sum(marbles))

Python 3.6.0 Shell
The total is 129
>>>

Here we use Python's built-in sum
function to tally up the marbles.
But we’re still learning about computation, so let’s compute the sum the
old-fashioned way (again, using what we’ve learned so far in this book) by
writing code that uses iteration to tally the list. Like this:

Let’s define a function to compute a sum of numbers.
with a local
To compute the sum of a list of numbers, we start ng
variable, sum, set to zero, which will hold the runni total.
Then we iterate through the list, and
list:
add each number to sum.
+ number
Good, we got the
Finally, we return the sum.
same result.

def compute_sum(list):
sum = 0
for number in
sum = sum
return sum

print('The total is', compute_sum(marbles))

Let’s test this by calling
compute_sum on our marbles list.
342  Chapter 8

Python 3.6.0 Shell

The total is 129
>>>

functions and abstraction

No! You can’t usme
the built-in su .
function either

Pretend the folks who developed the Python language decided to
remove any form of iteration (like the for and while loops). But
you still needed to sum a list of numbers, so could you do it without
iteration?

And now the different way...
There’s another approach that computer scientists (and some in-the-know coders) use to break down
problems. At first, this approach may seem a little like magic (or sleight of hand), but let’s get a feel for
it by revisiting our problem of summing our marbles. Here’s how the approach works: we come up with
two cases for summing our list of numbers: a base case, and a recursive case.

base
case

The base case is the simplest case you can think of. So what is the simplest
list of numbers you can take the sum of ? How about an empty list? What
is its sum? Zero, of course!

An empty list

compute_sum([])
recursive

case

Here’s the simplest case: if we
have an empty list, then we know
the sum is going to be 0.

Now for the recursive case. With the recursive case we’re going to solve a
smaller version of the same problem. Here’s how: we take the first item in the
list, and add it to the sum of the rest of the list...

[10, 13, 39, 14, 41, 9, 3]

How can we reduce the problem a
little? How about we just worry
about the sum of a list that is one
item smaller?

10 + compute_sum([13, 39, 14, 41, 9, 3])
We’ve made our problem a little smaller: to compute the sum of the list, we’re
going to add 10 to the sum of a slightly smaller list.
you are here 4  343

our first recursive function

Now let’s write some code for our two cases
Now that we have our base case and our recursive case, we’re ready to code this new way of
computing a sum. As we said up front, doing so is a little mind-bending for most, at least at first.
So let’s very slowly step through coding our new recursive sum function.

base
case

For the base case our job is easy: we just need to see if the list is empty, and if so,
return 0 as the sum of the list:
def recursive_compute_sum(list):
if len(list) == 0:
Here
return 0

we check to see if the list is
empty (in other words, if its size is
0), and if so, we return 0.

recursive

case

The recursive case is less obvious. Let’s take it a step at a time. We know that we’re
going to take the first item of the list and add it to the sum of the rest of the list. For
clarity, let’s first set up some variables to hold the first item and the remainder of the
list (without the first item):
def recursive_compute_sum(list):
Here’s our base case again.
if len(list) == 0:
return 0
Let’s set a variable to the first
else:
item in the list, and another one to
first = list[0]
the rest of the list.
rest = list[1:]

Remember your list notation? This
returns a list starting at index 1
through the last element in the list.

Now we need to add the first item to the sum of the remainder of the list:
def recursive_compute_sum(list):
if len(list) == 0:
return 0
else:
first = list[0]
rest = list[1:]
sum = first + Sum of rest of

The sum is the first
item plus the sum of
the rest of the list.
344  Chapter 8

list

But how do we code this?

What is the value
of rest if the list
only has one item?
It’s the empty list.

We need to sum the rest
of the list, but isn’t that
exactly what we’re coding?
A way to sum lists? It feels
like a conundrum.

functions and abstraction

If only we knew how to compute the sum of the rest of the list, we’d be set. But
how? Well, do you know of any good functions sitting around ready to compute the
sum of a list? How about recursive_compute_sum?
def recursive_compute_sum(list):
if len(list) == 0:
return 0
else:
first = list[0]
rest = list[1:]
sum = first + recursive_compute_sum(rest)
return sum

Let’s not forget to
return the sum after
we’ve computed it!

A Test Drive

Our assumption up front was that the
recursive_compute_sum computes the sum
of lists, so let’s call it to finish the job
on the slightly smaller list.

Whether you believe this code will work or not, go ahead and get the recursive_
compute_sum code (repeated below, including some test code) into a file called sum.
py. Save your code and choose the Run > Run Module menu item. After that, head to
the console to see the sum magically computed.

marbles = [10, 13, 39, 14, 41, 9, 3]
def recursive_compute_sum(list):
if len(list) == 0:
return 0
else:
first = list[0]
rest = list[1:]
sum = first + recursive_compute_sum(rest)
return sum
sum = recursive_compute_sum(marbles)
print('The total is', sum)

We got the same
result we did from
iteration!
Python 3.6.0 Shell

The total is 129
>>>

you are here 4  345

calling a function within a function

Aren’t we violating
that “define your functions before
you call them” rule we talked about?
After all, the recursive_compute_
sum function is called from within its
own definition!

No. Remember a function body is not evaluated

until the function is called. So, in this code, the
function recursive_compute_sum is first defined.
Then, when the recursive_compute_sum function
is called with:
sum = recursive_compute_sum(marbles)

the function’s body is then evaluated and calls itself
to recur. When that happens, the recursive_
compute_sum function is already defined, so we are
not violating that rule.
If you’re finding this takes a bit to wrap your head
around, that’s normal. The trick is deliberate
practice: write as many recursive functions as you
can. Trace through the execution and understand
how and why recursive functions work.
On that topic, let’s get some more practice...

346  Chapter 8

We’ll be tracing through
some recursive code in
just a bit.

functions and abstraction

Let’s get some more practice
Have no fear: getting your brain to think recursively takes a little extra effort, but
it’s well worth the blood, sweat, and tears (you think we’re kidding). Now, we could
stop at this point and analyze the recursive_compute_sum to death, but the
best way to get your brain thinking more recursively is deliberate practice: take
problems and solve them recursively, and, of course, write the code.
Let’s practice on another problem. Remember those palindromes from Chapter
4? You’ll recall that palindromes are words that read the same forward as they do
backward, like “tacocat”:

Reads the same forward...

tacocat
…as it does backward.

Want some more examples? How about “madam” or “radar” or “kayak,” or
there are even whole phrases (assuming you remove the punctutation and
whitespace), like “a nut for a jar of tuna” or “a man, a plan, a canal: panama” or
even more impressive, “a man, a plan, a cat, a ham, a yak, a yam, a hat, a canal:
panama.” Don’t believe the last few? Try them; they’re palindromes, alright.

Forget recursion for a bit, and think through how you might write a function to check if a word
is a palindrome. Do that using the skills you learned in Chapters 1 through 7. Write some
simple pseudocode to summarize your thoughts. Or, if you just had that cup of java and feel
like writing some code, don’t let us get in your way.

you are here 4  347

a recursive palindrome detector

Using recursion to detect palindromes
So can we write a recursive function to detect palindromes? And if so, will we have gained anything?
Let’s give it a try and find out. Do you remember what to do? To write a recursive function we need
a base case, and then we need a case that recurs by reducing the problem and then calling the same
function recursively. Let’s figure out the base and recursive cases:

base
case

The base case is the simplest case we can think of. We can actually think
of two simple cases. First, how about an empty string? Is the empty string
a palindrome? It reads the same front and backward, so yes.

The empty string

is_palindrome('')

Here’s the simplest case: if we
have an empty string, then we
know it’s a palindrome.

But there’s another really simple case to consider: the case of a single letter. Is a
single letter a palindrome? It’s the same forward and backward, so yes.

is_palindrome('a')
recursive

case

A single letter is a palindrome
too; after all, it is the same read
forward and backward.

Now for the recursive case. This is where things always get interesting.
Remember, we want to reduce the problem size a little before asking our
is_palindrome function to finish the job for us. How about we compare the
outer two characters and if they are the same, we can then check to see if the
middle of the word (which is a little smaller) is a palindrome?

Check the outermost characters
to see if they are the same.

'tacocat'
'acoca'

And let our is_palindrome function
worry about whether the middle
is a palindrome.
348  Chapter 8

functions and abstraction

Writing a recursive palindrome detector
We’ve got our base case and our recursive case, so once again we’re ready to write our
recursive code. As is typical, our base case is going to be fairly trivial to implement. Then we
just need to wrap our minds around the recursive case. As with computing sums, the trick is
always to reduce the problem a little and to rely on a recursive call to solve the problem.

base
case

For the base case our job is easy: we just need to see if the word is the empty
string or has one character:
def is_palindrome(word):
if len(word) <= 1:
return True

recursive

case

Let’s check our base case to see if
the word is the empty string (len
is 0) or has one character (len is 1),
and if so, return True.

Now for the recursive case. First we’re going reduce the problem by checking the
outer two characters. If they match, we have a palindrome if all the rest of the
letters (inside the two characters) make a word that is a palindrome. If not, we’re
going to return False:
def is_palindrome(word):
Here’s
if len(word) <= 1:
return True
else:
if word[0] == word[-1]:

else:
return False

our base case again.
Compare the first character to the
last character to make sure they
are equal. If not, we’ll return False.

We’ll think through the
recursive call in the next step.

Now we need to finish the recursive case. At this point the code has determined the
two outer characters are equal, so we have a palindrome if the middle of the word is a
palindrome, and that’s exactly what we need to code.

If the two ends match, then we need to see if the
we

def is_palindrome(word):
middle of the word is a palindrome. Good thing
if len(word) <= 1:
have
a function to do that—let’s call it.
return True
else:
if word[0] == word[-1]:
return is_palindrome(word[1:-1])
else:
Note that we need to return the result
return False

of calling is_palindrome, which will
ultimately return True or False.

you are here 4  349

more about recursive functions

A Test Drive

Go ahead and get the is_palindrome code (repeated below, including some test code)
into a file called palindrome.py; save your code and choose the Run > Run Module menu
item. After that head to the console to see if it is correctly detecting palindromes. Feel
free to add your own palindrome candidates to the test as well.

def is_palindrome(word):
Take a look through the code again.
if len(word) <= 1:
Is this clearer than the iterative
return True
version? What do you think?
else:
if word[0] == word[-1]:
Looks
return is_palindrome(word[1:-1])
else:
return False
words = ['tacocat', 'radar', 'yak', 'rader', 'kayjak']
for word in words:
print(word, is_palindrome(word))

Q:
A:

How do you know a recursive function will ever end?

In other words, if a function keeps calling itself, over and over,
how does it ever stop? That’s where the base condition comes in.
The base condition acts as a piece of the problem we know we can
solve directly, without the help of recursively calling the function
again. So, when we hit the base condition, we know we’ve reached
the point where the recursive calls stop.

Q:

case?

Okay, but how do we know if we’ll ever get to the base

A:

Remember each time we call the recursive case, we make
the problem a little smaller before calling the function again. So, if
you designed your code correctly, you can see that at some point,
by making the problem repeatedly smaller, you will eventually
reach the base case.

350  Chapter 8

like it works!

Python 3.6.0 Shell

tacocat True
radar True
yak False
rader False
kayjak False
>>>

Q:

I kinda get how we could call a function from itself—after
all, it is just like any other function call—but how do all the
parameters not get messed up? That is, each time I recursively
call the function, the parameters are reassigned to a new set of
arguments, right?

A:

This is a very good question. You are right; each time you
call a function, the parameters are bound to a set of arguments.
To make matters worse, if it is a recursive call, we’re calling the
same function, and so those parameters are going to get rebound
to other arguments—you’d think the whole thing would go haywire
when those parameter values get overridden, right? Ah, but that
isn’t what happens. You see, Python and all modern languages
keep track of every call to a function along with its corresponding
set of parameters (and local variables). Hang tight; we’re going to
look at this in a sec.

functions and abstraction

Behind
the Scenes
How is Python handling recursion and keeping track
of all those calls to the same function? Let’s take a look
behind the scenes and see how is_palindrome is
being computed by the Python interpreter.

def is_palindrome(word): 1
if len(word) <= 1: 2
return True
else:
3
first = word[0]
last = word[-1]
middle = word[1:-1]
if first == last: 4
return is_palindrome(middle) 5
else:
return False

Let's evaluate this statement.

Here's the code again. Notice that we
added some local variables that make the
code a bit clearer. It also allows us to see
how the variables work behind the scenes.

is_palindrome('radar')
1

The first thing Python (or practically any
language) does when it sees a function call is
to create a data structure to hold its parameters
and local variables. This is typically called a frame.
Python first puts the value for the parameter
word in the frame.

2

Next we see if the word has a length of 1 or less,
which it doesn’t.

3

Next we have three local variables that are
created and set to the first, last, and middle
portions of the word passed in. These values are
added to the frame as well.

4

Next we check to make sure the first and last
characters are equal, which they are, so we then
recursively call is_palindrome:
return is_palindrome(middle)

1

2

frame 1

word = 'radar'

frame 1

word = 'radar'
first = 'r'
last = 'r'
middle = 'ada'

Referring to frame 1,
middle is ‘ada'.

We’re back to another function call, so we need
a new frame to hold the parameters and local
variables. Python stores the multiple frames like
a stack of plates, putting one on top of the other.
We refer to the set of frames as a stack or call stack.
That name kinda makes sense, doesn’t it?
Okay, next we can see this word is not <= 1
characters, so we move on to the else statement.

frame 2

word = 'ada'

frame 1

word = 'radar'
first = 'r'
last = 'r'
middle = 'ada'
you are here 4  351

understanding the call stack

3

Once again, we compute our local variables and
add them to the frame.

4

And we can see that the first and last characters
are equal.

frame 2

word = 'ada'
first = 'a'
last = 'a'
middle = 'd'

So, all that remains is to call is_palindrome
again.

frame 1

return is_palindrome(middle)

Referring to frame 2,
middle is ‘d'.

frame 3

word = 'd'
first = 'd'
last = 'd'
middle = 'd'

We’re back to another function call, so we need
a new frame to hold the parameters and local
variables. At this point the word parameter is just
the string 'd'.

1

frame 2

word = 'ada'
first = 'a'
last = 'a'
middle = 'd'

Finally, the length of the word parameter is finally
less than or equal to 1, meaning we return True.
When the call returns, we remove, or pop, the top
frame off the stack.

2

5

word = 'radar'
first = 'r'
last = 'r'
middle = 'ada'

frame 1

word = 'radar'
first = 'r'
last = 'r'
middle = 'ada'

frame 2

Now we need to return the result of calling
is_palindrome, which is True. So we pop
another frame off the stack as we return True.

word = 'ada'
first = 'a'
last = 'a'
middle = 'd'

frame 1

word = 'radar'
first = 'r'
last = 'r'
middle = 'ada'

5

Again, now we need to return the result of calling
is_palindrome, which is True. Note this is the only
remaining frame that resulted from calls to is_
palindrome, so we’re done (again, with a result of True).
When we return from the initial call to is_
palindrome, we return the value True.

is_palindrome('radar')
352  Chapter 8

When we
return from
the function
its frame is
popped off
the stack.

Evaluates to True!

frame 1

word = 'radar'
first = 'r'
last = 'r'
middle = 'ada'

When we
return from
the function
its frame is
popped off
the stack.
Again, we pop
a frame off
the stack.

The stack is
now clear of
all calls to
is_palindrome.

functions and abstraction

Try evaluating some recursive code yourself. How about using our
recursive_compute_sum function?

Here’s the code again.

def recursive_compute_sum(list):
if len(list) == 0:
return 0
else:
first = list[0]
rest = list[1:]
sum = first + recursive_compute_sum(rest)
return sum
recursive_compute_sum([1, 2, 3])

recursive_compute_sum([1, 2, 3])

recursive_compute_sum([2, 3])

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

frame 2

list =
first =
rest =

And here we’re calling
the function.

We did the first one for you. The
parameter list is bound to the list
[1,2,3] and then the local variables
first and rest get computed and
added to the frame.
Trace through the rest of the
computation and fill in the stack
details.

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

recursive_compute_sum([3])

frame 3

list =
first =
rest =

frame 2

list =
first =
rest =

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

you are here 4  353

practice with call stacks

recursive_compute_sum([])

frame 4

list =

frame 3

list =
first =
rest =

frame 2

list =
first =
rest =

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

recursive_compute_sum([3])

frame 3

list =
first =
rest =
sum =

frame 2

list =
first =
rest =

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

recursive_compute_sum([2, 3])

frame 2

list =
first =
rest =
sum =

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

recursive_compute_sum([1, 2, 3])

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]
sum =

354  Chapter 8

functions and abstraction

Tonight’s talk: Iteration and
Recursion answer the question
“Who’s better?”

Iteration

Recursion

To know I’m better, all you have to do is look at how
many times coders use iteration over recursion.
I think that depends on the language you’re talking
about.
What do you mean? Any modern language supports
recursion, and yet coders opt to use me.

Take a language like LISP or Scheme or Clojure, for
instance—way more recursion is used than iteration.
Last time I looked, this book was in Python.

Hah! Efficient? Ever heard of a call stack?

That’s not the point. The point is, some
programmers know and understand recursion very
well, and see the beauty and efficiency of using it.

Well, yes, and so have the readers, but please, do
educate us.
Every time a function calls itself, the Python
interpreter has to create a little data structure
to hold all the parameters and local variables of
the current function. As the function gets called
recursively, it has to maintain a whole stack of those
data structures, which goes on and on as you keep
calling the function over and over. Call it enough
times and that adds up to a lot of memory, and then
BOOM, your program goes bye bye.

you are here 4  355

iteration versus recursion

Iteration

Recursion
That’s actually the way any modern (or ancient,
for that matter) language, including Python, works.
Anytime you call a function, that is happening.

Right, but as I said, when you do it recursively, it’s
like abusing the system, and sooner or later there’s
going to be trouble.

Not true. For many recursive algorithms, that isn’t
an issue and there are techniques for dealing with
that, anyway. The point is, look at the clarity of
using a recursive solution. Palindromes were a good
example; look how ugly and unclear the iterative
code was.

Hasn’t stopped millions of coders from writing
palindromes iteratively.
My point is, for some algorithms the recursive one is
easier to think about and code.
Sure, for those brainiacs who get recursion.
Oh please, as we’ve seen it just takes a little practice.
You have to admit, for a lot of problems, iterative
solutions are better.
I wouldn’t say better, I’d say more natural, but
I’d also say for some problems recursion is more
natural.
I say why bother for a little clarity?

Oh, you mean for those Earth-Shattering-GrandChallenge-type problems like finding palindromes?

It’s not just that the code is more readable, it’s
that there are algorithms that are downright hard
to code iteratively, but that work out easily and
naturally with recursion.
Of course not; however, maybe we’ll see one before
the end of this book.
By the way, you don’t find the fact we’re talking
about the book in the book slightly recursive?
Recursion is everywhere.

You think talking about talking about the book in
the book is...oh dear.
356  Chapter 8

functions and abstraction

RECURSION LAB
Today we’re testing the code for a recursive
algorithm that computes the Fibonacci sequence.
The sequence produces a set of numbers that
appear often in nature and can describe shapes,
like the pattern of seeds in a sunflower or the
shape of galaxies.
It works like this:
fibonacci(0) = 0
fibonacci(1) = 1

If you evaluate the function with
0 you get 0, and if you evaluate
it with 1 you get 1.

fibonacci(n) = fibonacci(n-1) + fibonacci(n-2)

And for any other number, n, we
produce the Fibonacci number by adding
fibonacci(n-1) to fibonacci(n-2).

The Fibonacci sequence is related
to the Golden Ratio, which
appears often in nature and is
considered by many artists to be
related to good design.

Here are a few values from the sequence:
fibonacci(0) is 0
fibonacci(1) is 1
fibonacci(2) is 1
fibonacci(3) is 2
fibonacci(4) is 3
fibonacci(5) is 5
fibonacci(6) is 8

Every number in the sequence
is computed by adding the two
Fibonacci numbers before it.

and continuing from there... 13, 21, 34, 55, 89, 144, 233,
377, 610, 987, 1597, 2584, 4181...and so on.
In the lab we’ve developed an algorithm to compute
Fibonacci numbers. Let’s take a look:

Working from the
definition above...

def fibonacci(n):
If n is 0 or 1, we just
return that number.
base
if n = = 0:
case
return 0
Otherwise, we return the sum
elif n = = 1:
of the two previous Fibonacci
numbers in the sequence, by
return 1
recursively calling fibonacci.
case else:
return fibonacci(n-1) + fibonacci(n-2)
recursive

Did you notice the
recursive case calls
fibonacci not once
but twice!

you are here 4  357

testing fibonacci

Now it’s time to test the code. Here in the Recursion Lab we need it to be
correct and fast. To do that we’ve developed a little test code using a new
module, time, which is going to help us time our code’s execution.
import time

We’re going to use Python’s time module to
time our code execution; see below.

def fibonacci(n):
if n = = 0:
Here’s the recursive
return 0
Fibonacci code.
elif n = = 1:
return 1
else:
return fibonacci(n-1) + fibonacci(n-2)

As a test we’re going to compute the Fibonacci
numbers 20 through 50, counting by fives. If that
goes well, we’ll compute all 100.

Test code.

for i in range(20, 55, 5):
start = time.time()
Start timer.
result = fibonacci(i)
Compute the Fibonacci.
end = time.time()
End timer.
duration = end - start
Compute duration.
print(i, result, duration)

Print results.

You job is to get this code entered and to perform the test run.
When you get the data, record it below, including the value
of n, the Fibonacci number, and how long it took to compute,
in seconds. For this code to be used in production, it has to
compute the first 100 Fibonacci numbers in less than 5 seconds.
Based on this test run, would we pass?

Fibonacci Test Data
Number

20

Answer

6765

Time to compute

We’re also going to time each
computation of Fibonacci. To do
that we’re going to use a module
called time. See Appendix A for
more on date and time modules.
If this program is taking too long to
execute, you can always stop it by closing
the shell window it’s running in.
Here’s what we got
for the first test,
n=20. Your timings
may differ depending
on the speed of your
computer.

.002 seconds

So far it
looks nice
and fast!
Our results are on the next page;
compare them with yours!

358  Chapter 8

functions and abstraction

RECURSION LAB FAIL
To meet Recusion Lab standards, this code has to compute the first 100
Fibonacci numbers in less than 5 seconds. How did you do? What? You
had lunch and they are still computing? No worries—we went ahead
and computed the results. Our numbers are below, but they don’t look
encouraging at all.

Fibonacci Test Data
Number

20
25
30
35
40
45
50

Answer

6765
75025
832040
9227465
102334155
1134903170
12586269025

Time to compute

0.002 seconds
0.04 seconds
0.4 seconds
4.8 seconds
56.7 seconds
10.5 minutes
1.85 hours

But while the execution time started very
fast, it is getting slower and slower the larger
n is. At 50 we’re taking almost 111 minutes to
compute just that one Fibonacci number!

The code is working
great in that we’re
getting the right
answers.

Your own results
may vary
depending on the
speed of your
machine.

Uh oh. It doesn't look good. We're hoping we
could really nail this Fibonacci code so that
we could compute the first 100 numbers in
the sequence in less than 5 seconds, but our
sample test run shows the 50th number on its
own takes over an hour!
Are we doomed? What on earth is taking so
long? Give it some thought, and we'll come
back to this after learning about an interesting
data structure (maybe it will help us?).

you are here 4  359

working on your next startup

The Anti-Social Network
You’ve really started to develop some coding chops. In fact, we think you’ve got the skills to
create a prototype to launch your first startup. And we even have an idea for the startup: a new
social network. Wait, what? Oh, you’ve had quite enough of Facebook and its competitors? No
worries—actually the Anti-Social Network is just what you’ve been looking for. With the AntiSocial Network it’s easy to communicate sentiments like “turn that smile upside down” or “if
you’re happy and you know it, then get away from me” to your network of, umm, friends.
The Anti-Social Network also has this killer feature where you can see the top anti-social user
at any time. Sounds like a billion-dollar idea to us; all you need to do is get started.
Let’s start simple: the first thing we need is to maintain a list of users. For each user we’ll store
a name and an email address.

To store a name and an email address for a set of users, let’s use what we know: let’s have
one list that holds the names, and another list that holds an email address for each name.
These are parallel lists (remember parallel lists from Chapter 4). In other words, if we store
a name at index 42 in the name list, you’ll find the matching email address at index 42 in
the email list.
names = ['Kim', 'John', 'Josh']
emails = ['kim@oreilly.com', 'john@abc.com', 'josh@wickedlysmart.com']

Can you think of any disadvantages of this approach?
Inserting new names and
addresses requires keeping
things consistent between
two lists.
We have to search the entire
list to find someone.
What if we need to add
another attribute, say, their
gender or phone number? It
gets more complex.

Can you think
of more issues?

360  Chapter 8

This is very similar to our approach
for storing the racing turtles’ color
and locations in the last chapter.

No issues; this totally
works.
As written, it is hard to see
how data relates to each
other because the name
and address are on two
different lists.
Deleting users requires we
keep our lists in sync.

__________________
__________________
__________________
__________________

functions and abstraction

If only I could find a data
structure that actually allowed me to give
each item a nice memorable name without having to
deal with indices in a list. And a data structure I could
just add to without having to worry about where in
a list I’m putting it. And a data structure in which I
could look up values quickly, without having to search the
whole list. That would be dreamy. But I know it’s just a
fantasy...

you are here 4  361

the dictionary data type

Introducing the dictionary
Meet Python’s dictionary data type. A dictionary, often called a map or an associative array (by those
computer science types), is a powerful and versatile data structure. Given its versatility, we’re
going to cover a few different ways you can use one, including using one to solve our Anti-Social
Network problems (that is, finding a better way to store and retrieve our users). Before that,
though, let’s first get acquainted with the dictionary and how it works.
The first thing to know is that, unlike Python lists, dictionaries are an unordered data type. With a
list each value is stored in order according to its index. If we want the third value in the list, we
can ask for it by its index. Dictionary items don’t have an inherent order; rather, each value stored
in a dictionary is accessed by a key. To access a value in a dictionary, we present its key.

Let’s see how to create a dictionary
my_dictionary = {}

And then add items

Use two curly braces to create an empty dictionary
that is all ready to hold keys and values.
Remember, use curly braces
for dictionaries and square
brackets for lists.

A dictionary stores items as key/value pairs. Say we have a phone number 867-5309
that we need to store for our friend Jenny:
my_dictionary['jenny'] = '867-5309'

Here we're storing this value, a phone
number in the form of a string, under
the key ‘jenny', also a string.

And you can store any number of key/value pairs. Let’s store a few more:
my_dictionary['paul'] = '555-1201'

my_dictionary['david'] = '321-6617'

We can store as many key/
value pairs as we like.

my_dictionary['paul'] = '443-0000'

Storing a value at a key that already
exist, overwrites its previous value.

my_dictionary['jamie'] = '771-0091'

And then retrieve values by their keys
To retrieve a value from a dictionary, we just use the key:
phone_number = my_dictionary['jenny']

print("Jenny's number is", phone_number)
362  Chapter 8

Python 3.6.0 Shell

Jenny's number is 867-5309
>>>

functions and abstraction

Keys and values do not have to be strings
For keys you can use numbers, strings, or Booleans. For dictionary values you can
use any valid Python value. Here are some examples:
my_dictionary['age'] = 27

my_dictionary[42] = 'answer'

We can have a string key and an integer value, an integer
key and a string value, a string key and a list value, and so
on, using any of Python's data types as a value.

my_dictionary['scores'] = [92, 87, 99]

Of course you can remove keys as well
del my_dictionary['david']

There are a few other types
that can act as keys as well, but
we haven’t covered them yet.

This removes the key ‘david'
along with its value from
the dictionary.

Remember, we can use the
del statement on other
things too, like Python lists.

But you might want to test to see if it exists first
Python is quite consistent in how you test to see if an item is part of a set of things
(which coders often call a collection), like a list or string, and the same holds for dictionaries.
You can test to see if a key is in a dictionary like this:
if

You can also use the pop
method on a dictionary,
which removes the key and
returns the value to you.

Use the in operator to check to see if
the
key exists in the dictionary. So we should have
'jenny' in my_dictionary:
written the code to
print('Found her', my_dictionary['jenny'])
delete an item above as:

else:

print('I need to get her number')

Q:

What happens if I try to delete a key
that doesn’t exist??

A:

Python will raise a runtime exception
known as a KeyError. We’ll be talking about
how to handle exceptions later in the book,
but you can avoid them by testing to see if
the key exists first.

Q:

So there’s only one of every key in
a dictionary?

if 'david' in my_dictionary:

del(my_dictionary['david'])

A:

A:

Q:

Q:

Right. In other words, each key is
unique within a dictionary. For instance,
there is only one key 'Kim' in my_dictionary;
if you were to assign a value to that key a
second time, it would overwrite the previous
value.

I can see the convenience of using
a dictionary, but I’m guessing that not
having indices is going to make things
less efficient. Am I going to have to
worry about performance when I use
dictionaries for a lot of data?

Remember how we promised this
chapter would be a little mind-bending?
Well, as it turns out, for many applications,
dictionaries are way more efficient than
using a list to store the same data. Hold on
to that thought; we’re going to get to it soon.

Given we can use the built-in
operator del on a dictionary, can we use
len too?

A:

You sure can. The len operator will tell
you the total number of keys in the dictionary.
you are here 4  363

iteration and literal syntax

What about iterating through a dictionary?
The one thing to remember about dictionaries is they are unordered.
You can iterate through the keys, just don’t expect them to be in any
particular order.

As with lists and strings, use
the for/in statement to iterate
through the keys of a dictionary.
for key in my_dictionary:

print(key, ':', my_dictionary[key])

Let’s print the key…

…followed by its value
in the dictionary.

Python 3.6.0 Shell
jenny : 867-5309
paul : 443-0000
jamie : 771-0091
age : 27
42 : answer
scores : [92, 87, 99]
>>>

Okay, but can we get literal for a sec?
As with lists, there’s also a literal notation for creating dictionaries,
which looks like this:
harry = {'firstname': 'Harry',
'lastname': 'Potter',

'house': 'Gryffindor',

Your order may be different
and that's just fine.

Each key/value is separated by a colon,
and followed by a comma (except for the
final pair).
This creates a full-fledged
dictionary.

'friends': ['Ron', 'Hermione'],
'born': 1980}

You can also print a dictionary to see its literal form:
print(my_dictionary)

Python 3.6.0 Shell
{'jenny': '867-5309', 'paul': '4430000', 'jamie': '771-0091', 'age':
27, 42: 'answer', 'scores': [92,
87, 99]}
>>>

364  Chapter 8

Never count
on dictionary
order
Even if you find
the keys in your dictionary are
always in the same order, you
won’t be able to count on that
fact across different operating
systems and implementations
of Python. Just remember, if
you’re counting on the order of
a dictionary, you’re going to run
into trouble sooner or later.

functions and abstraction

Now that you’ve got some book
knowledge of dictionaries, it’s time
to put that knowledge to work. Work
through the code below to see what it
computes.

List

movies = []
movie = {}
movie['name'] =
movie['year'] =
movie['rating']
movie['year'] =

Dictionary
'Forbidden Planet'
1957
= '*****'
1956

movies.append(movie)
movie2 = {'name': 'I Was a Teenage Werewolf',
'year': 1957, 'rating': '****'}
movie2['rating'] = '***'
movies.append(movie2)
movies.append({'name': 'Viking Women and the Sea Serpent',
'year': 1957,
'rating': '**'})
movies.append({'name': 'Vertigo',
'year': 1958,
'rating': '*****'})
print('Head First Movie Recommendations')
print('--------------------------------')
for movie in movies:
if len(movie['rating']) >= 4:
print(movie['name'], '(' + movie['rating'] + ')', movie['year'])

you are here 4  365

applying dictionaries to the anti-social network

Leveraging dictionaries at the Anti-Social Network
Having read the last few pages, you know 95% of everything there is to know
about using Python dictionaries, except, well, how to actually use them. Sure, you
can store and retrieve values, but there has to be more to it than that, right? Yes,
simplicity can be deceptive. Let’s look at how you might use dictionaries at the
Anti-Social Network.
Recall that when we last left off, we needed to store a set of names with email
addresses. We started by using two lists, which turned out to be rather clumsy
because adding new names required adding to both lists, as did deleting names;
looking up a name meant we had to search over the entire list; and adding more
properties, like say the user’s gender, was going to require an entirely new list to
manage. Yuck. Well, let’s see if we can do better with dictionaries.

Here are some users and their
emails stored in two lists.

names = ['Kim', 'John', 'Josh']
emails = ['kim@oreilly.com', 'john@abc.com', 'josh@wickedlysmart.com']

Let’s see if a dictionary can help us.

Well, that is certainly
more readable!

users = {'Kim' : 'kim@oreilly.com',
'John': 'john@abc.com',
'Josh': 'josh@wickedlysmart.com'}

What about adding new users, or removing them?
users['Avary'] = 'avary@gmail.com'
del users['John']

Adding

Removing

Okay, that was easy—no worrying
about having two lists to keep in sync.

What about quickly getting a user’s email address? Let’s say
we’re after Josh’s email address:
if 'Josh' in users:
print("Josh's email address is:", users['Josh'])

Let's first check to make sure the
key ‘Josh' exists, and if it does, we
grab the email address.

366  Chapter 8

Python 3.6.0 Shell
Josh's email address is: josh@wickedlysmart.com

>>>

functions and abstraction

I’ve gotta admit, from a coding
perspective this is so much nicer than
dealing with two lists. But is this really as
efficient as using explicit indices?

The dictionary is more efficient, by a long shot.
To understand why, you have to understand how dictionaries
actually work behind the scenes. Let’s start with a list as an
example: if you looked up “Josh” in a list of users, you’d have to
go through every single user in the list until you found Josh. In the
worst case, that means you’d have to go through the entire list.
Dictionaries use a data structure behind the scenes that’s known as
a hash map. With a hash map the dictionary stores all the values in a
list or array-like data structure, but it uses a special function, known
as a hash function, to compute where the value is in that list, based
on the key. So, rather than brute-force searching the entire list, a
dictionary, by way of the hash function, can go right to the index
where the value is located. Lucky for you, the dictionary does all
this for you.
Now hash functions aren’t perfect, and sometimes multiple values
may be mapped to one location (the dictionary knows how to deal
with this case), but this isn’t likely to happen often, so the average
time to look up a value based on a key is constant time (you can think
of constant time as the time it takes to perform a single operation).
So, when searching for keys, dictionaries are blindingly fast.

A computer
science term

you are here 4  367

dictionaries within dictionaries

But how do we add more attributes?
Managing the Anti-Social Network usernames and email addresses with a
dictionary seems like a real win, but remember we also wanted to potentially add
some attributes, like a user’s gender. Now, we know if we’re using parallel lists then
we’d need another list to hold the genders, but with dictionaries, don’t we need to
use another dictionary too? Something like:
email = {'Kim' : 'kim@oreilly.com',
'John': 'john@abc.com',
'Josh': 'josh@wickedlysmart.com'}
genders = {'Kim' : 'f',
'John': 'm',
'Josh': 'm'}

We could do it this way, but now
we’re back to managing two data
structures anytime we add, delete,
or look up a user. Yuck again.

Sure, that would work, but then we’re back to managing two data structures—
that’s obviously not what we want. To solve this problem, we need to think a little
deeper about how to use dictionaries. What if we use a dictionary to hold all the
attributes for each user, like this:
attributes = {
'email' : 'kim@oreilly.com',
'gender': 'f',
'age': 27,
'friends': ['John', 'Josh']
}

Here's a dictionary holding the
attributes for Kim. We can create a
dictionary like this for every user.
Notice we’ve added a new list to each
user as well, a list of friends. This will
play into our killer feature.

With this new scheme, let’s start over, create a brand new users dictionary, and
store Kim’s attributes:
users = {}
users['Kim'] = attributes

Let’s add John and Josh too:

We're setting the users
variable to an empty
dictionary, and then adding
the attributes dictionary
under the key ‘Kim'.

Don’t move too quickly and
overlook what is going on in
this code. We’re making the
value of the ‘Kim’ key another
dictionary, the attributes
dictionary.

users['John'] = {'email' : 'john@abc.com','gender': 'm', 'age': 24, 'friends': ['Kim', 'Josh']}
users['Josh'] = {'email' : 'josh@wickedlysmart.com','gender': 'm', 'age': 32, 'friends': ['Kim']}

Same here: don’t move too quickly
and overlook what is going on in this
code. We’re assigning dictionaries to
‘John’ and ‘Josh’ as well.
368  Chapter 8

We’re assigning dictionaries to the keys
‘John’ and ‘Josh’. We’re specifying those
dictionaries with the literal syntax.

functions and abstraction

Dictionaries in dictionaries—it’s a
common arrangement. See how it plays
out at the cinema by using your brain to
execute the code below.
movies = {}
movie = {}
movie['name'] =
movie['year'] =
movie['rating']
movie['year'] =

'Forbidden Planet'
1957
= '*****'
1956

movies['Forbidden Planet'] = movie
movie2 = {'name': 'I Was a Teenage Werewolf',
'year': 1957, 'rating': '****'}
movie2['rating'] = '***'
movies[movie2['name']] = movie2
movies['Viking Women and the Sea Serpent'] = {'name': 'Viking Women and the Sea Serpent',
'year': 1957,
'rating': '**'}
movies['Vertigo'] = {'name': 'Vertigo',
'year': 1958,
'rating': '*****'}
print('Head First Movie Recommendations')
print('--------------------------------')
for name in movies:
movie = movies[name]
if len(movie['rating']) >= 4:
print(movie['name'], '(' + movie['rating'] + ')', movie['year'])
print('Head First Movie Staff Pick')
print('---------------------------')
movie = movies['I Was a Teenage Werewolf']
print(movie['name'], '(' + movie['rating'] + ')', movie['year'])

you are here 4  369

implementing the killer feature

Brain Building

Now that we know how we’re going to store our users on the Anti-Social Network, it’s
time to write a little code. Let’s create a function we might need for the startup, called
average_age, that takes a name and returns the average age of that user’s friends.
We’re letting you tackle this on your own—after all, it is Chapter 8—although don’t
forget to write some some pseudocode or do similar planning, as it goes a long ways
toward writing correct code the first time.

users = {}
users['Kim'] = {'email' : 'kim@oreilly.com','gender': 'f', 'age': 27, 'friends': ['John', 'Josh']}
users['John'] = {'email' : 'john@abc.com','gender': 'm', 'age': 24, 'friends': ['Kim', 'Josh']}
users['Josh'] = {'email' : 'josh@wickedlysmart.com','gender': 'm', 'age': 32, 'friends': ['Kim']}

Write your
average_age
function here.

average_age('Kim')
average_age('John')
average_age('Josh')

Here’s the
output you
should get with
this test code.

Python 3.6.0 Shell
Kim’s friends have an average age
of 28.0
John’s friends have an average age
of 29.5
Josh’s friends have an average age
of 27.0
>>>

Remember the Anti-Social
Network’s killer feature?
You didn’t forget, did you? It’s going to be the crucial feature to demo
when raising your seed round of investor cash. The killer feature for the
Anti-Social Network is finding the least social person at any time—that is,
the user with the fewest friends. All we need to do is write the code.
So let’s take stock: we’ve got a dictionary of users, and each entry in that
dictionary consists of a key, which is the user’s name, and a value, which
is another dictionary full of the user’s attributes. One of those attributes
is a list of friends. So, we’re going to have to iterate through all the users
and keep track of who has the least friends. Let’s give it a shot.

370  Chapter 8

Who will it be?

functions and abstraction

Finding the most anti-social user
The code for finding the most anti-social user isn’t going to be the most complex we’ve written, but
let’s just hash it out in a little pseudocode so we have a clear picture of what we’re doing:

1

Set max to large number

2

For each name in users

3

A

Get the user attribute
dictionary.

B

Get the list of friends from
the attribute dictionary.

C

If number of friends is less
than max

We’re going to create a variable to hold the
current most anti-social count. We’ll call it max
and initialize it to something large for starters.
Iterate through every key in
the users dictionary.
For each user, grab the dictionary of attributes
associated with the name.
Use the ‘friends’ key to grab the list of friends.

i

set variable most_anti_
social to name.

ii

set variable max to number of
friends

If the number of friends is less than any we’ve seen so far
(max), then that’s our current least candidate.
Set a variable most_anti_social to the name…
…and max to the new number of friends.

Print user with key most_anti_social

When we’ve been through all the keys in names, we then
have the most_anti_social user. Let’s print it, and so on.

Now that we have some pseudocode, use your new knowledge of dictionaries to
complete this code. After that we’ll get it running.

max = 1000

for name in ___________:

user = ______[________]

friends = user[_________]
if len(_________) < max:

most_anti_social = _______
max = len(________)

print('The most_anti_social user is', __________________)
you are here 4  371

testing the anti-social network

A Test Drive

Let’s get this code in and test it (it's been a few pages since we ran real code).
Put the code below in a file called antisocial.py. Give it a run, and double-check
your output.

users = {}

users['Kim'] = {'email' : 'kim@oreilly.com','gender': 'f', 'age': 27, 'friends': ['John', 'Josh']}
users['John'] = {'email' : 'john@abc.com','gender': 'm', 'age': 24, 'friends': ['Kim', 'Josh']}

users['Josh'] = {'email' : 'josh@wickedlysmart.com','gender': 'm', 'age': 32, 'friends': ['Kim']}
max = 1000

for name in users:

user = users[name]

friends = user['friends']
if len(friends) < max:

most_anti_social = name
max = len(friends)

print('The most_anti_social user is', most_anti_social)

Python 3.6.0 Shell
The most_anti_socia
l user is Josh
>>>

That Josh!

Now, it’s all you!
That’s about as far as we’re going to take the Anti-Social Network because we
think you’re in a good position to take it forward on your own. The idea and
the code are yours, so run with it and send us a postcard when you’re rich and
famous (although maybe postcard is a little too social for the founders of the
Anti-Social Network).
We’re not quite done with dictionaries, though, as we’ve got another hoop we’re
going to make them jump through...
372  Chapter 8

functions and abstraction

Meanwhile, back at the...

RECURSION LAB
We last left the Recursion Lab in a bit of a demoralized state. While we
needed to compute the first 100 Fibonacci numbers in less than 5 seconds,
it took us much longer. So long, in fact, that the 50th alone took over an
hour. Given that, is there any hope? Of course.
But we can’t make progress if we don’t figure out why our computation is
so slow. Let’s start by looking at the recursive calls that are made when we
compute, say, fibonacci(50):

To compute fibonacci(50), we need to
compute fibonacci(49) and fibonacci(48).
But computing fibonacci of
49 itself has to compute
fibonacci(48) too.

fibonacci(50)

fibonacci(49)
fibonacci(48)

fibonacci(48)

fibonacci(47)

fibonacci(47) fibonacci(46) fibonacci(46) fibonacci(45)

We stopped here,
but obviously we
could have gone
on and on...

fibonacci(47)
fibonacci(46)

And notice at each level of
recursion, we’re computing the
same Fibonacci numbers over
and over!

So while our code is logically correct and clear, it is very inefficient. For
any number, the algorithm has to compute all the Fibonacci numbers
less than that number. That leads to a lot of unnecessary computation
because we’re repeatedly computing the same Fibonacci numbers
over and over again. Every time we need fibonacci(5), we have to
recompute fibonacci(4), fibonacci(3), and fibonacci(2) as
well.

fibonacci(46)

fibonacci(45) fibonacci(45)

fibonacci(44)

By only the third level of our
graph we’re computing 48
twice, 47 three times, 46
four times, 45 three times,
and it’s only going to get
worse.

you are here 4  373

understanding fibonacci’s inefficiency

Try drawing a graph of all the function calls it takes to compute fibonacci(7). When
you’re done, total up how many duplicated calls you find for each value 0 through 7.

fibonacci(7)
fibonacci(6)

Put your totals here. We’ve
already filled in the answer to
fibonacci(7), which is called
one time, as is fibonacci(6).

fibonacci(5)

1 time
fibonacci(7): _____

1 time fibonacci(5): _____
fibonacci(6): _____

fibonacci(4): _____

fibonacci(3): _____ fibonacci(2): _____

fibonacci(1): _____

fibonacci(0): _____

Warning: you may find
the answer shocking.

Any guess how many times fibonacci(50) has to compute fibonacci(3)? ___________

We’re computing the same Fibonacci numbers over and over
again. Without drastically changing the code, think about how we
might make this code way more efficient. Put your ideas here.

374  Chapter 8

functions and abstraction

If only there was a way
to store previous function call results
so that if the function gets called with
the same value again, we just remember
its previous result rather than having to
recompute it every time. I know it’s just
a fantasy...

you are here 4  375

remembering the results of function calls

Can we just remember function call results?
It’s not a bad idea. So, for instance, if fibonacci is called with the argument 49,
the result is computed and stored, so the next time fibonacci is called with 49, the
function can look up the result, rather than recomputing it.
With the Fibonacci code recomputing so many function calls, if we could just store
the results rather than recomputing them, we’d probably shave a lot of time off the
execution. But can we really save hours of computation time? We shall see.
So what would we need to store the results of the calls to the fibonacci function?
We need a way to store the value of the n argument, and also the result that
gets computed for fibonacci(n). Oh, and we’d need to be able to access the
computed result quickly, for a given value of n.
Any ideas? Is this sounding familiar?

Using a dictionary to remember our Fibonacci results
Does this sound like a job made for a dictionary? Let’s take a look at how it might
work.

A cache is a common name
for a place to store data you
need quick access to.

1

Create a dictionary, and name it cache.

2

Every time fibonacci is called with a number n:

A

Check if there is a key, n, in the cache.

i

If so, return the value of key n.

B

Otherwise, compute the Fibonacci number of n.

C

Store the result in the cache under the key n.

D

Return the value of the Fibonacci number.

So each time the Fibonacci code function is called with a value n, we first check
our dictionary, cache, to see if a key n is already in the cache. If it is, great, that’s
the Fibonacci number for n, so just return it as the result of calling the function.
If not, we compute the Fibonacci number as we always do, but before we return
the result, we first stash it in the cache for the key n.

376  Chapter 8

functions and abstraction

We might as well use the $5 words: memoization
If you thought we just made up this brilliant idea of storing function call results, well, we
wish we did, but we didn’t. In fact, it’s a simple but powerful technique called memoization.
Memoization can be thought of as a way to optimize a program by storing the results
of expensive function calls. What’s expensive? Well, how much you got? Just kidding. In
computer science terms, expensive typically means a computation takes a lot of time or
consumes a lot of space (typically memory). Whether you’re trying to optimize time or
space really depends on what your code does, but in our case it’s mostly time we’re trying to
optimize.
So how does memoization work? You already know, becasue the pseudocode we wrote on
the previous page is an implementation of memoization.
To understand it better, let’s rework our existing code using the pseudocode as a guide.
Doing so is pretty straightforward:
import time
cache = {}

Here’s the dictionary we’re
going to use as a cache.

def fibonacci(n):
global cache
if n in cache:
return cache[n]

Now the first thing we do is check to see if n is a
key in the cache dictionary. If it is, then we return
the value stored for that key.

if n == 0:
Now rather than returning immediately if n equals 0 or 1, we
result = 0
assign n to a local variable named result.
elif n == 1:
result = 1
If we end up recursively computing
else:
the result, we then assign that to
result = fibonacci(n-1) + fibonacci(n-2)
result local variable as well.
cache[n] = result
return result
Before we return the result, we
start = time.time()
for i in range(0, 101):
result = fibonacci(i)
print(i, result)

the

store it in the cache under the key n.
Remember in Python a dictionary key
can be any value, including integers.

We must be confident because we’ve changed the
timing code to capture the duration of computing
the entire first 100 Fibonacci numbers.

finish = time.time()
duration = finish - start
print('Computed all 100 in', duration, 'seconds')

you are here 4  377

testing the optimized fibonacci

RECURSION LAB Success
What are you waiting for? Are you skeptical a couple lines of code using a
dictionary could cut hours of computation down to less than 5 seconds?
Well, get these code updates made and give it a run!

Python 3.6.0 Shell

46 18363119033
47 297121507 6
48 480752697 9
49 777874204 25
50 125862690 74
51 203650110 99
52 329512800 73
53 533162911 72
54 862675712 445
55 139583862 717
56 225851433 162
57 365435296 879
58 591286729 041
59 956722026 5920
60 154800875 1961
61 250473078 7881
62 405273953 9842
63 655747031 57723
64 106102098 77565
65 171676801 35288
66 277778900 12853
67 449455702 48141
68 727234602 460994
69 117669030 709135
70 190392490 170129
71 308061521 879264
72 498454011 049393
73 806515533 4928657
74 130496954 7978050
75 211148507 2906707
76 341645462 0884757
77 552793970 3791464
78 894439432 24676221
79 144723340 48467685
80 234167283 73143906
81 378890623 21611591
82 613057907 94755497
83 991948530 816367088
84 160500643 911122585
85 259695496 727489673
86 420196140 638612258
87 679891637 8366101931
88 110008777 6004714189
89 177997941 4370816120
90 288006719 0375530309
91 466004661 4746346429
92 754011380 15121876738
93 122001604 19868223167
94 197402742 34990099905
95 319404346 54858323072
96 516807088 89848422977
97 836211434 344706746049
98 135301852 834555169026
99 218922995 179261915075
100 354224848 100 in 0.0005749298095703125 seconds
Computed all
>>>

378  Chapter 8

Impressive! You got
it computed in way
under 1 second!

Optimizing your
code can affect its
runtime behavior in a
HUGE way.

functions and abstraction

Some more brain building
Hopefully this has been the mind-bending chapter we promised, but before you
go we wanted to show you that recursive code isn’t just for computing Fibonacci
numbers and palindromes. In fact, we’re going to apply recursion to generating
some computer graphics, more specifically fractals. Now everyone’s familiar
with the term, but what does it actually mean? Think of a fractal as a geometry
shape that is similar at any scale. So if you zoom out and look at a fractal, you
see the same general shape as when you zoom way in. The best way to get the
feel for fractals is to make some. To do that let’s look at some code.
import turtle
def setup(pencil):
pencil.color('blue')
pencil.penup()
pencil.goto(-200,100)
pencil.pendown()

Our turtles are back.
All this setup function does is set the turtle
(we're calling it a pencil in this code) color and
move it to a location that will make the drawing
more centered.

def koch(pencil, size, order):
if order == 0:
pencil.forward(size)
else:
for angle in [60, -120, 60, 0]:
koch(pencil, size/3, order-1)
pencil.left(angle)
def main():
pencil = turtle.Turtle()
setup(pencil)
order = 0
size = 400
koch(pencil, size, order)
if __name__ == '__main__':
main()
turtle.tracer(100)
turtle.mainloop()

This is the recursive function; we'll
come back to this and study it.

The main function creates a turtle (the
pencil); defines two variables, order and
size; and then calls the recursive function,
sending it all three arguments.
Notice the order starts at 0, and size
starts at 400. You’ll see how these are
used in a bit.
About what we'd expect in this bit of
code. We call main and then make sure
the turtle mainloop is running. We also
use a function we haven't seen before,
tracer, which accelerates the speed on
the turtle.
you are here 4  379

the koch recursive function

A closer look at the koch function

We’ve pulled out the koch
function to take a look, so
for the entire context refer
to the previous page.

Our code (on the previous page) is pretty basic: it creates a turtle, repositions
it, and sets its color. Other than that, it just calls the koch function. But what
does this koch function do? Let’s look at it again, here:

koch takes a pencil, a size, and an order.
def koch(pencil, size, order):
base
case
recursive

if order == 0:

pencil.forward(size)

Our base case: if order equals
0 we draw a straight line of
length size.

else:

case

for angle in [60, -120, 60, 0]:

koch(pencil, size/3, order-1)
pencil.left(angle)

ch was a Swedish
Niels Fabian Helge von veKohis name to the fractal
mathematician who gaowflake. See more at:
known as the Koch sn rg/wiki/Helge_von_Koch.
https://en.wikipedia.o

Otherwise, call koch four times,
passing it the size parameter divided
by three and reducing order by one.

After each call to koch, we adjust
the angle of the turtle.

The best way to understand this
is to see some examples.

We have some abstract idea of what this function does, but we have no idea
what it really does. We do know that the order parameter seems to play a big
role. Let’s start with order equal to 0 and see what this draws. Given 0 is the
base case, it should just draw a line.
Type in the code and give it a test drive. Here’s what we got:

A Test Drive

The base case, a line of 400 pixels
(because we’re passing it a size of 400)
If we increase order to 1, then we should see the recursive case used. Just
change the order local variable to 1 and test it again.

A Test Drive

When order is 1, we use the recursive case in
the code. In the recursive case we draw four
segments at different angles. The drawing
is actually done calling the koch function
recursively, where for each segment koch hits
the base case and draws a line of size 400/3

380  Chapter 8

functions and abstraction

If you can see what koch does from order 0 to order 1, can you imagine what it might do
when we increase the order to 2? Or 3? Remember, at every scale a fractal is the same. See
if you can try to work out how the graphic at order 1 will change for order 2, and order 3.
We didn’t say it was easy to think about, but it’s great brain building.

Order 0:

Order 1:

Order 2:

Order 3:

Draw your best
guesses here.

Hint: look at what
happened between
order 0 and 1.
What if you did
that to each small
segment of order 1
to get order 2?

Hint: can you take
that one step
further on order 3?

you are here 4  381

creating a snowflake with koch

Truly exploring the koch fractal
Did that exercise help you think a little more recursively, or fractally? Let’s take
this all the way and run the code at orders 4 and 5. Go ahead and change your
order and do a test drive for each.

A Test Drive

Order 4

Order 5

They don’t call it the Koch snowflake for nothin’
Here’s one last tweak for you. Edit your main function to call the koch function three
times, rotating 120 degrees between each call.
def main():
pencil = turtle.Turtle()
setup(pencil)
turtle.tracer(100)

Not bad for a
six-line recursive
function. That’s a
superpower!

order = 5
size = 400
koch(pencil, size, order)
for i in range(3):
koch(pencil, size, order)
pencil.right(120)

382  Chapter 8

Call koch three
times, and rotate
120 degrees in
between.

Learn more about the
Koch snowflake at rg/
https://en.wikipedia.oe.
wiki/Koch_snowflak
A Test Drive

functions and abstraction

With that we’ll leave you to ponder the
power of recursion and fractals, not
to mention dictionaries. Great job—
you’ve done some mind bending and
growing in this chapter. Now it’s time to
go let your brain do something else for
a while, but only after the bullets and
crossword.

 Iteration and recursion can both be
used to solve problems.
 Recursion defines a solution in terms
of itself.
 Recursion typically consists of a base
case and a recursive case.
 With the recursive case we make the
problem a little smaller and then make
a recursive function call.
 Programming languages handle
recursive calls by putting parameters
and local variables on a call stack.
 Recursion can sometimes lead to
problems with the size of the call stack
growing too large.
 Some solutions are more natural using
a recursive style, and some solutions
are more natural with iteration.
 For some problems recursion can
produce very clear and straightforward
solutions.
 Python dictionaries are a type of
associative array or map.
 Dictionaries store key/value pairs.

 In Python, dictionary keys can be
strings, numbers, or Booleans;
dictionary values can be any type of
Python value.
 Keys within a dictionary are unique.
 Assigning a value to an existing key
overwrites the previous value.
 You can programmatically create
dictionaries or specify them with a
literal syntax.
 Keys and their associated values may
be deleted from a dictionary.
 Key lookup in a dictionary happens in
constant time.
 Dictionaries can be used with
memoization as an optimization
technique.
 Memoization is a technique used to
remember previous function calls.
 By not recomputing expensive function
calls, we can sometimes significantly
speed up computations.
 Fractals are shapes that have similar
structure at every scale.

you are here 4  383

coding crossword

Coding cross
1

2

3

4

Don’t worry: it’s not a recursive crossword,
just a normal5 one.
1

2

3

4

5

6

7

8

6

7

9
8

10
9
10

12

11

13
11

12

13

14
14

15
15

16
16

17

17

Across
Across

1. And another name for a dictionary.
1. And
another name for a dictionary.
4. Snowflake.
4. Snowflake.
5. Functions calling themselves.
5. Functions
6. Another calling
name forthemselves.
dictionary.
8. Just typing
a dictionary
in.
6. Another
name
for dictionary.
9. Shape
same
at every level
8. Just
typing
a dictionary
in.of scale.
10.
Our
idea
is
this
kind
of
social.
9. Shape same at every level of scale.
11. Keys are this.
10. 12.
OurMakes
idea aisdictionary
this kindlookup
of social.
fast.
11. 14.
Keys
are this.
Tacocat
is one.
12. 15.
Makes
a dictionary
lookup fast.
Easiest
case.
Call stack
is made up of.
14. 16.
Tacocat
is one.
Computes
15. 17.
Easiest
case. a sequence seen in nature.

.
16. Call stack is made up of.
17. Computes a sequence seen in nature.

384  Chapter 8

Down
Down

2. Dictionaries store key/value _____.
2. Dictionaries store key/value _____.
3. Remembering function calls.
3. Remembering
function calls.
4. Used
to look up a value.
4. Used to
to recursion.
look up a value.
7. Alternative
11. Dictionaries
are ________.
7. Alternative
to recursion.
13. Parameters
are stored
it.
11. Dictionaries
areon________.

13. Parameters are stored on it.

functions and abstraction

Forget recursion for a bit, and think through how you might write a function to
check if a word is a palindrome. Do that using the skills you learned in Chapters 1
through 7. Write some simple pseudocode to summarize your thoughts. Or, if you
just had that cup of java and feel like writing some code, don’t let us get in your
way.

In general we’d want to compare the
outside characters and move inward until
we find a character that doesn’t match,
or if we get to the middle and they all
match, then we have a palindrome.

We could have an index called i
that starts at position 0…

…and another called
that starts at the j
end.

tacocat

We start at each end of
the word and compare
pairs of letters from the
outside in.
When i >= j, then we’ve reached the middle of
the string and we’ve compared everything.

is_palindrome(word):
set i to 0
set j to length of word (minus 1)
while i < j:
if characters at i and j are not equal then return False
increment i
If at any point our outer
decrement j
characters aren’t equal,
then the word is not a
if the loop completes, return True
palindrome.

def is_palindrome(word):
i = 0
j = len(word) - 1
while i < j:
if word[i] != word[j]:
return False
i = i + 1
j = j - 1
return True

Okay, we just had to
write the code. Did you?
This probably works great, but
it requires a lot of thinking
about indices. It’s also not the
clearest code in the world. We
can probably do better.

Study this until you’ve convinced yourself it works.

you are here 4  385

exercise solution

Try evaluting some recursive code yourself. How about using our
recursive_compute_sum function?
def recursive_compute_sum(list):
if len(list) == 0:
return 0
else:
first = list[0]
rest = list[1:]
sum = first + recursive_compute_sum(rest)
return sum
recursive_compute_sum([1, 2, 3])

recursive_compute_sum([1, 2, 3])

recursive_compute_sum([2, 3])

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

frame 2

list = [2, 3]
first = 2
rest = [3]

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

recursive_compute_sum([3])

frame 3

list = [3]
first = 3
rest = []

frame 2

list = [2, 3]
first = 2
rest = [3]

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

386  Chapter 8

Here’s the code again.

And here we’re calling
the function.

We did the first one for you. The
parameter list is bound to the list
[1,2,3] and then the local variables
first and rest get computed and
added to the frame.
Now we’re recursing by calling
recursive_compute_sum again, so we
add a new frame and add list as a
parameter. This time it has the value
[2, 3].
Like before, we have to compute first
and rest.
Again, we’re recursing by calling
recursive_compute_sum, so we add
a third frame and add list as a
parameter. This time it has the value
[3].
Like before, we have to compute first
and rest.

functions and abstraction

recursive_compute_sum([])

frame 4

list = []

frame 3

list = [3]
first = 3
rest = []

frame 2

list = [2, 3]
first = 2
rest = [3]

This will be our last time calling
recursive_compute_sum. We’ll hit our
base case because list is the empty list.
The function just returns, so we never
compute first or rest.

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

recursive_compute_sum([3])

frame 3

list = [3]
first = 3
rest = []
sum = 3

frame 2

list = [2, 3]
first = 2
rest = [3]

frame 1

In the last step we hit the base case
and returned 0. So we pop that frame
off the stack.
Our code takes this and adds it to
first and then stores the result in sum.

Our code then returns the value of sum.

list = [1, 2, 3]
first = 1
rest = [2, 3]

recursive_compute_sum([2, 3])

frame 2

list = [2, 3]
first = 2
rest = [3]
sum = 5

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]

recursive_compute_sum([1, 2, 3])

frame 1

list = [1, 2, 3]
first = 1
rest = [2, 3]
sum = 6

In the last step the value 3 was
returned, so that is added to first,
which is 2 and stored in sum.
We then return the sum of 5 and the
stack is popped again.

Finally we get back the value 5 and
add it to first, which is 1, resulting
in 6, so the stack is popped and 6
is returned as the result of calling
recursive_compute_sum([1, 2, 3]).
you are here 4  387

exercise solution

Now that you’ve got some book
knowledge of dictionaries, it’s time
to put that knowledge to work. Work
through the code below to see what it
computes.

Python 3.6.0 Shell

Is this what you got?

Head First Movie Recommendations
-------------------------------Forbidden Planet (*****) 1956
Vertigo (*****) 1958
>>>

Dictionaries in dictionaries—it’s a
common arrangement. See how it plays
out at the cinema by using your brain to
execute the code below.

Python 3.6.0 Shell

A little trickier with
two dictionaries?

388  Chapter 8

Head First Movie Recommendations
-------------------------------Forbidden Planet (*****) 1956
Vertigo (*****) 1958
Head First Movie Staff Pick
--------------------------I Was a Teenage Werewolf (***) 1957
>>>

functions and abstraction

Brain Building Solution

Now that we know how we’re going to store our users on the Anti-Social
Network, it’s time to write a little code. Let’s create a function we might
need for the startup, called average_age, that takes a name and returns
the average age of that user’s friends.

users = {}

users['Kim'] = {'email' : 'kim@oreilly.com','gender': 'f', 'age': 27, 'friends': ['John', 'Josh']}
users['John'] = {'email' : 'john@abc.com','gender': 'm', 'age': 24, 'friends': ['Kim', 'Josh']}

users['Josh'] = {'email' : 'josh@wickedlysmart.com','gender': 'm', 'age': 32, 'friends': ['Kim']}

Here’s our function, which takes a
username in the form of a string.

def average_age(username):

Let’s grab the user’s attributes
dictionary from the users dictionary.

global users

user = users[username]

And then let’s grab the friends list
from the user’s attributes dictionary.
Here’s a local variable to keep track of
sum = 0
the sum of the friend’s ages.
Now let’s iterate through all the friends.
for name in friends:
Now we need to take the friend’s name and
grab their attributes dictionary out of the
friend = users[name]
users dictionary.
sum = sum + friend['age']
And then we can get their age from the attributes
dictionary. Here we add it to the sum.
With the ages totaled up, let’s compute the
average = sum/len(friends)
average…
friends = user['friends']

print(username + "'s friends have an average age of", average)
average_age('Kim')

…and print the result.

average_age('John')
average_age('Josh')

And here’s our test code again.

Here’s the
output you
should get with
this test code.

Python 3.6.0 Shell
Kim’s friends have an average age
of 28.0
John’s friends have an average age
of 29.5
Josh’s friends have an average age
of 27.0
>>>

you are here 4  389

exercise solution

Now that we have some pseudocode, use your new knowledge of dictionaries to
complete this code. After that we’ll get it running.

max = 1000

users
for name in ___________:

user = users
______[________]
name

friends = user[_________]
'friends'
if len(_________)
friends < max:

name
most_anti_social = _______
friends
max = len(________)

most_anti_social
print('The most_anti_social user is', ________________)

Try drawing a graph of all the function calls it takes to compute fibonacci(7). When
you’re done, total up how many duplicated calls you find for each value 0 through 7.

fibonacci(7)
fibonacci(6)
fibonacci(5)
fibonacci(4) fibonacci(3)

fibonacci(5)

fibonacci(4)

fibonacci(4)

fibonacci(3)

fibonacci(3) fibonacci(2) fibonacci(3) fibonacci(2) fibonacci(2) fibonacci(1)

fibonacci(3) fibonacci(2) fibonacci(2) fibonacci(1)

fibonacci(1) fibonacci(0)

fibonacci(2) fibonacci(1) fibonacci(1) fibonacci(0)
fibonacci(2) fibonacci(1) fibonacci(1) fibonacci(0)
fibonacci(1) fibonacci(0)

Put your totals here. We’ve
already filled in the answer to
fibonacci(7), which is called
one time, as is fibonacci(6).

fibonacci(1) fibonacci(0)

fibonacci(2) fibonacci(1) fibonacci(1) fibonacci(0)

fibonacci(1) fibonacci(0)

fibonacci(1) fibonacci(0)

1 time
fibonacci(7): _____
fibonacci(4):
fibonacci(1):

3 times
_____
13 times
_____

1 time fibonacci(5): _____
2 times
fibonacci(6): _____
5 times fibonacci(2): _____
8 times
fibonacci(3): _____
8 times
fibonacci(0): _____

Okay, that
is shocking!

4,807,526,976 times
Any guess how many times fibonacci(50) has to compute fibonacci(3)? _________________
390  Chapter 8

functions and abstraction

If you can see what koch does from order 0 to order 1, can you imagine what
it might do when we increase the order to 2? Or 3? Remember, at every scale a
fractal is the same. See if you can try to work out how the graphic at order 1 will
change for order 2, and order 3. We didn’t say it was easy to think about, but it’s
great brain building.

Order 0:

In this step we drew
a shape like this…
Order 1:

Order 2:

…to each side (there
was only one side).

Same thing; for
each straight side in
the previous round,
we added another.

And as the sides
get smaller, we keep
adding…
Order 3:

…reducing the scale
each time.

you are here 4  391

exercise solution

1

M A

2

3

P

5

A

6

Coding
Cross
Solution

8

A

R E

I

M

R

O

S

4

M

S O C

I

C U R S

I

T E
10

A

12

H A

S

H

R A

A

T

P

15

B

L

T

F

I

E

A

9

F R A

C

T A

A
T

13

T

A
L

I

11

U N

I

O N

O

S

E

C A

O

S

B O N A

C C

I

L
Q U E

D
E

16

E

F R A M E

A
I

Y

N D R O M E

L
A

R R A

E

I

A

E

V

T
17

7

O

F U N C
14

K O C H

T

T
N

V

Y

Z
L

I

S

E
I

D

K

392  Chapter 8

Across

Down

1. And another name for a dictionary. [MAP]
4. Snowflake. [KOCH]
5. Functions calling themselves. [RECURSIVE]
6. Another name for dictionary.
[ASSOCIATIVEARRAY]
8. Just typing a dictionary in. [LITERAL]
9. Shape same at every level of scale.
[FRACTAL]
10. Our idea is this kind of social. [ANTI]
11. Keys are this. [UNIQUE]
12. Makes a dictionary lookup fast.
[HASHFUNCTION]
14. Tacocat is one. [PALINDROME]
15. Easiest case. [BASECASE]
16. Call stack is made up of. [FRAMES]

2. Dictionaries store key/value _____. [PAIRS]
3. Remembering function calls.
[MEMOIZATION]
4. Used to look up a value. [KEY]
7. Alternative to recursion. [ITERATION]
11. Dictionaries are ________. [UNORDERED]
13. Parameters are stored on it. [CALLSTACK]

9 saving and retrieving files

Persistence

He was right
behind me and then
poof!—he was gone.
If we’d only saved him
out to storage before
everything ended.

You know you can save values in variables, but once
your program ends poof!—they’re gone forever. T hat’s
where persistent storage comes in—storage that allows your values and data to
stick around a while. Most of the devices you’re going to run Python on also have
persisistent storage, like hard drives and flash cards, or they may have access to
storage in the cloud as well. In this chapter you’ll see how to write code to store
and retrieve data from files. What good is that? Oh, anytime you want to save a
user’s configuration, store the results of your big analysis for the boss, read an
image into your code to process it, write some code to search a decade’s worth
of email messages, reformat some data to be used in your spreadsheet app—we
could go on and on, but we should probably start the chapter now.

this is a new chapter   393

the crazy libs game

Ready for some Crazy Libs?
No, that’s not a political statement! We’re being serious here. We’re talking about
creating our own version of a game that is in the spirit of the popular Mad
Libs™ games—we’ll call our version Crazy Libs.
If you didn’t play a Mad Libs–style game as a kid, here’s how it works:

To create a Crazy Lib, just
remove arbitrary words and
replace them with blanks
that indicate their parts of
speech (noun, verb, etc.).

You can create a Crazy Lib
from any text story you
want. We used some text we
just had lying around.

Crazy Libs
How to Play
1

Ask a friend to give you a word for
each blank noun, verb, and adjective,
but don’t show your friend the story.

2

Read the story back to your friend,
adding in your friend’s words.

3

Hilarity ensues.

The first thing that stands between you and ____

___________ your

VERB ENDING IN ‘ING’

first, real, piece of ________, is _______________
the skill of
NOUN

VERB ENDING IN 'ING'

breaking problems down into achievable _______

ADJECTIVE

____________ can do for you. Of course, you and
the computer will
NOUN

also need to be _________________ a common
VERB ENDING IN 'ING'

we’ll get to that topic in just a bit.

The text from the
first paragraph of
Chapter 1

394  Chapter 9

actions that a

___________, but
NOUN

saving and retrieving files

Okay, I’m going with
VERB ENDING IN ING: buying, NOUN:
pudding, VERB ENDING IN ING: forgetting,
ADJECTIVE: crazy, NOUN: monkey, VERB
ENDING IN ING: eating, NOUN: pizza.
So, “The first thing that stands
between you and buying your first real piece of
pudding is forgetting the skill of breaking problems down
into achievable crazy actions that a monkey can do for you. Of
course, you and the computer will also need to be eating a common
pizza, but we’ll get to that topic in just a bit.”
Nice!

you are here 4  395

how the game works

How Crazy Libs is going to work
To turn Crazy Libs into a computer game, we’re going to have the computer
retrieve a story stored in a text file, prompt the user for all the missing words,
and then create a new text file that contains the completed Crazy Libs. Let’s go
through that in a little more detail:

Game code

Let’s start with a template

1

By text file, we just mean a
file containing text that is
stored on your device.

We need a template that has all the
text of the story, with placeholders for
the words that need to be provided by
the user; we also need some indication
of each placeholder’s part of speech.
To signify a placeholder, we’ll just use
uppercase words like NOUN, VERB, and
ADJECTIVE right in the text.

Here’s an example template.

The Python intepreter

1

The first thing that stands between you
and VERB_ING your first, real, piece of NOUN,

is VERB_ING the skill of breaking
problems down into achievable ADJECTIVE
actions that a NOUN can do for you. Of
course, you and the computer will also
need to be VERB_ING a common NOUN, but

A template is just a text
file on your computer that
contains a story along with
placeholders for nouns, verbs,
verbs with ‘ing’ endings,
adjectives, and so on.

we’ll get to that topic in just a bit.

lib.txt

396  Chapter 9

You’ll find lib.txt in the ch9
folder in the book’s source code.

Feel free to take
your favorite story
and create your own!

saving and retrieving files

2

Reading the template
Next, when you run the game, the code is going
to read in the template from the disk and then
locate all the placeholders—remember the
placeholders are just the uppercase words, like
NOUN, VERB, and so on, embedded in the text.
2

The first thing that stands between you
and VERB_ING your first, real, piece of NOUN,
is VERB_ING the skill of breaking

3

problems down into achievable ADJECTIVE

Prompting the user

actions that a NOUN can do for you. Of

Give me a VERB: buying

For each placeholder your code is going to
prompt the user to provide a word that matches
the needed part of speech. Your code will then
take their word, and substitute it into the text in
place of the placeholder.

4

3
Give me a NOUN: pudding

Give me a VERB_WITH_ING:

course, you and the computer will also
need to be VERB_ING a common NOUN, but
we’ll get to that topic in just a bit.

Give me a ADJECTIVE: cra
Give me a NOUN: monkey

lib.txt

Giveme a VERB_WITH_ING:
Give me a NOUN: pizza

Finally, we’ll write out a new file that
contains the completed Crazy Lib.

4

After getting all
the user input, our
code creates a new
file and writes out
the updated crazy
version of the story.

After all the placeholders have been collected from
the user, we’re going to create a new file and then
write the completed Crazy Lib to it.
The first thing that stands between you

Here’s the completed
Crazy Lib, which has each
placeholder replaced by the
user’s input.

and buying your first, real, piece of pudding,
is forgetting the skill of breaking
problems down into achievable crazy
actions that a monkey can do for you. Of
course, you and the computer will also
need to be eating a common pizza, but
we’ll get to that topic in just a bit.

For the output filename,
we’ll just prepend “crazy_”
to the original filename.

crazy_lib.txt
you are here 4  397

writing the pseudocode

Hey, it’s Chapter 9 already, so we have no doubt you can tackle the task
of creating a little pseudocode based on the previous two pages. Writing
up a little pseudocode now is going to give you a clearer idea of how to
approach writing the Crazy Libs game. Don’t skip this! But of course you’ll
find our version at the end of the chapter.

Write your pseudocode
here. Remember the story
is in a file, so you’ll need
to read it, figure out
where the placeholders are,
prompt the user for the
placeholder replacements,
and then write it all out
to a new file.

P.S. We know you don’t
know how to read or
write data from files yet,
so just focus on the logic
of the code, and don’t
worry about the specifics
(which is the point of
pseudocode!).

398  Chapter 9

saving and retrieving files

Step 1: Read the text of the story from a file
The first thing you need to do is get your hands on the text file that contains the
example story. You’ll find this file, lib.txt, in the book source files under the ch9 folder.
Take a quick look at it. We also encourage you to create your own story files, but for
testing let’s use lib.txt so we’re testing the same thing.

1

Read the text for the story
from file.

2

Process text.
For each word in text

A

B
The first thing that stands between you
and VERB_ING your first, real, piece of NOUN,
is VERB_ING the skill of breaking
problems down into achievable ADJECTIVE
actions that a NOUN can do for you. Of

Our test story
file, complete with
placeholders

3

If word is a placeholder (NOUN
or VERB or VERB_WITH_ING or
ADJECTIVE):

1

Prompt user for placeholder
part of speech.

2

Substitute user’s word for
the placeholder.

Otherwise, the word is fine;
keep it in the story.

Store results.
Take the processed text with
the placeholders filled, and write
it out to a file with the filename
prepended by “crazy_”.

course, you and the computer will also
need to be VERB_ING a common NOUN, but
we’ll get to that topic in just a bit.

lib.txt

To read a file in Python, you first have to open it
If we’re going to grab the data out of a file, we first need to open it. What
does opening a file actually do? From your perspective, not much, but
behind the scenes, Python has to locate your file, make sure it exists, and
ask the operating system for access to the file—after all, you might not have
permission to access it.

You’ll find this is the
case with practically
every programming
language—to read or
write to a file, you
have to open it first.

To open a file you use Python’s built-in open function, which takes a
filename and a mode as arguments.

It takes a filename
open to
's
on
th
Py
e
th
e
be opened…
us
We
le.
…and a mode.
fi
a
en
op
to
on
ti
func

my_file = open('lib.txt', 'r')
The filename can be a simple filename in
The open function returns a file the same directory as your code, or a
object, which we're assigning to more explicit path to a file. Right now
the variable my_file.
we're using a simple filename. More on
paths on the next page.

The mode can either be
‘r' for reading or ‘w' for
writing to the file.
We use the terms “folder” and
“directory” interchangeably in this
chapter, although when we’re talking
about files and file paths, we tend
to use the word “directory.”
you are here 4  399

learning about file paths

How to use file paths
When we used the open function on the previous page, we
specified a filename of 'lib.txt', and, because we didn’t
indicate where the file was located, we were assuming the file
was located in the current directory. But what if the file you
want to open is located somewhere else? How do you then
open it? To do that we add a path to our filename, which tells
open where to look for the file. There are actually two kinds
of paths: a relative path or an absolute one. A relative path is a
description of the file’s location that is relative to the folder
(we usually say directory instead when we talk about paths)
you ran your Python code from.

my_file = open('lib.txt', 'r')
ch9

my_file =

The first thing
that stands
between you and
VERB_ING your
first, real,
piece of NOUN,
is VERB_ING
the skill of
breaking

my_text = my

print(my_text

my_file.
close(

crazy.py

lib.txt

Same directory.

When you supply a
filename, you’re counting
on that file being in the
same directory as the
directory you ran your
main Python file from.
We’ll assume that is our
ch9 directory.

Relative paths
Let’s look at relative paths first, which are always specified in
relation to your current directory (the directory you ran your
program from). Let’s say you put the file lib.txt in a subfolder
(again, we typically say subdirectory) called libs; well, you
can prepend your filename with the path to the file, like this:
'libs/lib.txt'. So the open function will first go to the libs
subdirectory before looking for the file lib.txt.
You can list as many subdirectories as you need, separating
each one with a path separator, otherwise known as the
forward slash character, /.

Windows users, check out
the Watch it! below.

my_file = open('libs/lib.txt', 'r')
ch9

my_file =

my_text = my

print(my_text

my_file.
close(

Relative to our ch9
folder, you first go to
the libs directory and
then you’ll find lib.txt.

libs

crazy.py
The first thing
that stands
between you and
VERB_ING your
first, real,
piece of NOUN,
is VERB_ING
the skill of
breaking

lib.txt

Mac and Windows use different separators.
Mac (and Linux) machines use a forward slash / as a separator,
while Windows uses a backslash, \. That said, Python allows you
to enter path separators uniformly using a forward slash. So if you
have a Windows path of:
C:\Users\eric\code\hfcode\ch9\lib.txt

then, instead, use:

C:/Users/eric/code/hfcode/ch9/lib.txt

400  Chapter 9

saving and retrieving files

So, what if you want to open a file in a directory that’s in an
adjacent folder (like ch9 and libs are adjacent folders), or a
directory higher in your filesystem (like hfcode is higher than
libs in the filesystem)? No problem. You can use the .. (two
periods) notation to specify a directory one level higher. If
you’re running your code from the ch9 directory and you
want to get to the lib.txt file in the adjacent libs directory, then
you can use a path of '../libs/lib.txt'. So we go up
one directory and then down in to the libs directory before
finding lib.txt.

my_file = open('../libs/lib.txt', 'r')
code

hfcode

ch9

Or, if you’re after the lib.txt in the directory oldlibs, you can
use the path '../../oldlibs/lib.txt'.

oldlibs

libs

my_file =

Absolute paths

lib.txt

The first thing
that stands
between you and
VERB_ING your
first, real,
piece of NOUN,
is VERB_ING
the skill of
breaking

my_text = my

print(my_text

my_file.
close(

Here we go up two directories, before going
down into the oldlibs directory and finally,
finding the lib.txt we’re after.

The first thing
that stands
between you and
VERB_ING your
first, real,
piece of NOUN,
is VERB_ING
the skill of
breaking

crazy.py

lib.txt

my_file = open('../../oldlibs/lib.txt', 'r')

The root of a filesystem is, perhaps
counterintuitively, the very top level of
your filesystem.

An absolute path is a path that always starts from the root of
your filesystem. Absolute paths tell you exactly where a file is in
a filesystem. Now, while that may sound like a more definitive
way to identify your files, and it is, often absolute paths are less
flexible because if you move all your code and files to another
machine, you have to update the absolute paths in your code.
That said, you’ll find times when using an absolute path is more
convenient than a relative path.

usr

Here's my toplevel folder. Note
this is on a macOS
X filesystem.

eric

code

Let’s say we want to specify an absolute path to the lib.txt file.
That looks like this on our machine:

hfcode

my_file = open('/usr/eric/code/hfcode/ch9/lib.txt', 'r')

On the Mac and Linux systems, start at the root
with a forward slash, and then add the rest of the
file path separated by forward slashes.
And on a Windows machine, assuming you have your
code on the C: drive, your path would look like this:

ch9

my_file =

my_text = my

print(my_text

my_file.
close(

crazy.py

The first thing
that stands
between you and
VERB_ING your
first, real,
piece of NOUN,
is VERB_ING
the skill of
breaking

lib.txt

my_file = open('C:/Users/eric/code/hfcode/ch9/lib.txt', 'r')

On Windows systems, start with the drive and then a colon
and then a forward slash. Then add the rest of the file path
separated by forward slashes.
you are here 4  401

how to close a file

Trace out each path to find the file it identifies. For any relative paths, start
in the code directory. We’ve done one for you already.

/usr/harry/code/find.py

usr

../inventory/map.pdf

harry

hermione

../../hermione/spells/leviosa.txt
code

inventory

find.py

spells

data/students.txt
cloak.gif

data

map.pdf

oculus.txt

leviosa.txt

find.py

horcrux.db

students.txt

Oh, and don’t forget to clean up when you’re done!
Before we go on, one point of etiquette: if you’re going to open a file, when you’re
done with it, you need to close it. Why? Open files take up resources in your machine’s
operating system, and, especially for long-running programs, leaving files open that
you aren’t using can lead to code that eventually crashes and burns. So, if you’re
going to open a file, just make sure you close it too. Here’s how:

variable,
Here’s our my_filefile object.
which is set to a

Just call the file object’s close
method to close the file.

my_file.close()

Keep this method in mind, as we’ll use it in a bit...
402  Chapter 9

saving and retrieving files

Reading a file into your Python code

We couldn’t fit the
whole thing on the page:
but you get the point,
this string contains all
the text in lib.txt.

First of all, what does it mean to read a file into Python? Well, with Python, the
contents of the file are taken and placed into a Python string, which you can
examine or process as you wish using standard Python string operations.

nand VERB_ING your first real

een you\
‘The first thing that stands betw

piece of NOUN\n is VERB_ING

the skill of...’

Here’s our sample
text file.

my_text

The first thing that stands between you
and VERB_ING your first, real, piece of NOUN,

Here’s a standard
Python string holding
the contents of the
file lib.txt.

is VERB_ING the skill of breaking

When reading a file,
Python takes the
contents and places it
into a string.

problems down into achievable ADJECTIVE
actions that a NOUN can do for you. Of
course, you and the computer will also
need to be VERB_ING a common NOUN, but
we’ll get to that topic in just a bit.

Did you notice those weird
\n characters in the string?
More on those in a sec...

Using the file object to read a file
There are actually two ways to read the contents of your file into
a string: you can read the entire file all at once, or you can read
the file a line at a time. Let’s start by reading the entire file all at
once; to do that we’re going to use the file object’s read method:

Technically there are a few other ways
too, but we’re going to focus on the
most common methods of reading files.

Let’s use the read method to get the ...which we’ll assign to
entire contents of the lib file...
the string my_text.
my_file = open('lib.txt', 'r')
my_text = my_file.read()

print(my_text)

my_file.close()

Don’t forget to
close the file!

Then we print
the string out,
which displays
everything in
the file to the
Python Shell.

Python 3.6.0 Shell
The first thing that stands betwee
n you
and VERB_ING your first, real, piece
of NOUN,
is VERB_ING the skill of breaking
problems down into achievable ADJECT
IVE
actions that a NOUN can do for you.
Of
course, you and the computer will
also
need to be VERB_ING a common NOUN,
but
we’ll get to that topic in just a
bit.
>>>

you are here 4  403

searching over files

Code Magnets

Can you help find the needle in the haystack? We’ve been given a directory of 1,000 files with
the filenames 0.txt through 999.txt, and only one has the word needle in it. We had the code to
find the needle all written on the refrigerator, but then someone came along and messed it all
up. Can you help us put it back together? Notice, there may be some extra code magnets, so
you may not use all of them. Check your answer at the end of the chapter.

1000):
for i in range(0,

file.close()
xt:
if 'needle' in te

print('Found needle in file ' + str(
i) + '.txt')

Rearrange the
magnets here.

filename = str(i) + '.txt'
999):
for i in range(0,

text = file.read()
ame, 'r')
file = open(filen

404  Chapter 9

saving and retrieving files

I was looking at the
code in the previous
exercise and something is bothering
me. Even when we find a file with the
needle string in it, the code continues on,
reading every remaining file. I guess that
would be good if we wanted to check every
file, but we said it was only in one file.

Good catch. You’re right; say the file

with the needle is 512.txt. With this code we’d
unnecessarily check files 513 through 999 before
the code completed. Opening files can be a
relatively time-consuming activity, so this isn’t an
optimal way to do things.

The problem is, if you look at the code again,
we’re iterating with a for loop, so we can’t just
stop checking files in the middle. Or can we?
This does bring up something we haven’t
discussed yet, which is a way to break out of
an iteration, in the middle, if we decide we
don’t need to complete it. If we had something
like that, we could easily fix this code so that it
stopped the search after finding the needle. Of
course we do have something: Python and most
programming languages have a break statement
for just this purpose. Let’s take a look at how it
works...

you are here 4  405

the break statement

Oh, give it a break already...
When we use a for statement, say a for statement that is
iterating over a range of numbers, we know that the for
statement will iterate over every number in the range before
it completes. But there are times when in a computation
you’ll decide partway through that continuing is pointless,
and there is no reason to complete the iteration—our
needle in a haystack code is a good example, because once
we find the needle, there is no reason to continue the for
loop to open the remaining files.

Don’t forget us while
loops! The break statement
works with us too!

w hi le lo op

The break statement gives you a way out. With the break
statement we can push the eject button and stop the for
loop at any time. Let’s have a look:
for i in range(0, 1000):

When we hit the
break statement
we immediately
jump out of the
for loop.

filename = str(i) + '.txt'
file = open(filename, 'r')

Here’s our code to find
the needle again.

text = file.read()

if 'needle' in text:

print('Found needle in file ' + str(i) + '.txt')
break

file.close()

print('Scan complete')

So now, when we find the needle, we’re going to
use the break statement and bypass the rest of
the code block as well as breaking out of the loop
altogether.

Uh oh. By adding the break statement to the code above we
actually introduced a small bug. Can you find it? Can you
determine what it is? How would you fix it?

Hint: is every file still being closed?

406  Chapter 9

saving and retrieving files

Hey, we’ve got a Crazy Lib game to finish!
Short excursions are great for learning new things about coding, but we’ve got a game
to finish here. When we last left off, we were reading the entire contents of a file at once,
which is certainly easy, but it has a drawback: for large files it can consume significant
resources. Think about processing a file with, say, hundreds of thousands of lines—you
probably don’t want to read all that into memory at once, or the result could be out-ofmemory errors for your program or even your operating system.
A more common approach is to read the contents one line at a time. But what is a line?
Exactly what you’d think: it is all the text on each line, up until you encounter a new line.
In fact, if we could open our lib.txt file in a special editor that showed us all the characters
in the file, we’d see something like this:

The first thing that stands between you\n
and VERB your first, real, piece of NOUN,\n
is VERB_WITH_ING the skill of breaking\n
problems down into acheivable ADJECTIVE\n
actions that a NOUN can do for you. Of\n
course, you and the computer will also\n
need to be VERB_WITH_ING a common NOUN, but\n
we’ll get to that topic in just a bit.

Each line ends with
an actual new line
character, otherwise
known as a newline,
which is represented by
the escape character
sequence ‘\n’.

Think of an old-style
typewriter with a
carriage return. Every
time you hit the
carriage return, that’s
like a new line.
You never see newlines
in editors because
rather than show them,
editors just treat
them as a directive to
display the next bit of
text on the next line.

lib.txt

Using the readline method of the file object
To read one line from a file, we use the readline method of the file object. Let’s try
reading the lib.txt file again using readline:
my_file = open('lib.txt', 'r')

line1 = my_file.readline()
print(line1)

line2 = my_file.readline()
print(line2)

The file object tracks its position as
it reads the file. So each time you call
the readline method, it picks up where
it last left off.

This time we’re using the readline method
and reading the first two lines into the
variables line1 and line2, respectively.
Python 3.6.0 Shell
The first thing that stands between you
and VERB_ING your first, real, piece of NOUN,

you are here 4  407

escape sequences

Python 3.6.0 Shell

Why do you think the output from the code on the
previous page has an extra newline between the two
lines “The first thing that stands between you” and
“and VERB your first, real, piece of NOUN,”?

The first thing that stands between you
and VERB_ING your first, real, piece of NOUN,

Do you see the
extra newline?

Back up the bus.
A page back you used the term
“escape sequence.” Excuse me?
You didn’t really define it.

Fancy name, simple concept.

When you want to create a string with a newline in it, there’s no way to
directly type in that newline character. Think about it: when you’re typing in
a string, you type a quote and then some text and if you typed a newline by
hitting the Return key, the editor would just go to the next line. So, instead we
use a sequence of characters that represents a newline character, and those
characters happen to be \n. So when you see the characters \n don’t think
of them as two characters, a backslash and the letter n, but rather as a way to
represent a newline character. So, for example, if you want to print some text
followed by five new lines, you’d do this:
print('Get ready for new lines:\n\n\n\n\n')

Newlines, of course, aren’t the only
escape sequence. There’s \t for
tabs, \b for backspace, and \v for a
vertical tab.

This explains the
answer to the
Brain Power above.
408  Chapter 9

Here we have the text “Get
ready for new lines:” followed by
how many new lines? Five? Nope!
Six! We get five from the string,
but, as it turns out, the print
function adds a newline to its
output by default.

You might also hear some
coders refer to ‘\n’ as a
line feed.

Python 3.6.0 Shell
Get ready for new lines:

>>>

saving and retrieving files

How do we know when we’ve read the last line?
Think of readline as keeping a pointer to the position it last read. Each time you
read another line, readline picks up where it last left off, and then reads the next line.
When there are no more lines to read, readline then returns the empty string. Here’s
how you can use that to read all the lines in a file a line at a time:

When we use a while True,
that means we’re iterating
FOREVER.

my_file = open('lib.txt', 'r')

while True:

line = my_file.readline()

Each time we loop we read the next line, and
if it isn’t the end of the file (signified by an
empty string), then we print it.

if line != '':

print(line)

else:

Note a blank line
would have a newline
chararacter in it, while
the end of file sets
line to an empty string.

If it is the end of the file, well, then we
use our handy break statement to break
out of the while loop. Thank goodness—
forever is a long time.

break

my_file.close()

Ah, but there's a easier way using the
power of Python sequences
We’re glad you’ve adapted so quickly to using the break statement,
but in this case there’s a nicer and cleaner way to iterate over the
lines in a file. Remember how we’ve used the for statement to iterate
over sequences? We’ve used it to iterate over the items in a list, and
we’ve used it to iterate over the items in a string. Well, isn’t a file just a
sequence of lines? Ah, so can we use for to iterate over the lines of a
file as well? You bet. Let’s rewrite the code above using for:
my_file = open('lib.txt', 'r')

for line in my_file:
print(line)

my_file.close()

Wow, not only shorter,
but much more readable!

Serious Coding
Anytime you’re using for with the
in keyword, whether it’s on a list
or a string or a file or a dictionary,
we’re using a concept called an
iterator.

With an iterator we assume the
data type we’re iterating over
gives us a standard way to iterate
through its sequence of values. We
don’t worry about how—the for
loop does that—we just know we
can step through each value until
we’re out of values.
You’ll find iterators in many
modern languages—they’re
based on a high-level software
design practice called Design
Patterns.

you are here 4  409

reading a template

Reading in a Crazy Lib template
You now know just about everything there is to know about opening and reading a text file, line
by line. Let’s put that knowledge to use to write the first part of our Crazy Lib code. Let’s start
by defining a function, make_crazy_lib, that is going to take a filename and return the text
of the Crazy Lib in a string, complete with the user’s word choices already substituted.
We’re also going to create a helper function, process_line, which is going to process each
line for any placeholders. The process_line function will also be responsible for getting each
replacement word from the user, and substituting it into the text.

Here's our function; it takes a filename and
then opens it for reading.
def make_crazy_lib(filename):
file = open(filename, 'r')

We're going to use a variable, text, to build
up the text as we process it.

text = ''
for line in file:
text = text + process_line(line)

A helper function
is just a common
name for a function
that takes on a
subtask for another
function. In this case
our make_crazy_lib
function is going to
rely on process_line
to look at a single
line and process it
for placeholders.

For each line in the file we’re going
to process it with the process_line
function, and then add it to the text
we’re building up.

After we've gone through each line
of the file, we just need to close
file and return the text.
the
return text
For testing, let’s have process_line return the text we
process_line(line):
pass it. That way we can test that all the file reading and
return line
string concatenation is working correctly in make_crazy_lib.
file.close()

def

def main():
lib = make_crazy_lib('lib.txt')
print(lib)
if __name__ == '__main__':
main()

A Test Drive

And of course we need to call the make_crazy_lib
function. Let's add the main function to do that.

Make sure the file “lib.txt” is in
the same directory as your code.

Get the code above into a file called crazy.py,
save your code and choose the Run > Run
Module menu item. Double-check your output.

Once again you should just see the entire file,
complete with placeholders. Remember, all we did
was process the entire file, line by line, putting
everything back together into the variable, text,
but without any real processing.
410  Chapter 9

Python 3.6.0 Shell
The first thing that stands betwee
n you
and VERB_ING your first, real, piece
of NOUN,
is VERB_ING the skill of breaking
problems down into achievable ADJECT
IVE
actions that a NOUN can do for you.
Of
course, you and the computer will
also
need to be VERB_ING a common NOUN,
but
we’ll get to that topic in just a
bit.
>>>

saving and retrieving files

Processing the template text

1

Read the text for the story
from file.

So, moving on in our pseudocode, we now need to process the text, which
means we need to get the process_line function actually doing something.
For starters, in step 2A, we need to iterate through each word in each line.
Luckily we learned how to do that back in Chapter 6, when implementing
the readability app. Let’s give the same technique a try; here’s a first cut at the
structure of the process_line function:

2

Process text.

def

Okay, the processed line
takes a line of the text.
And we’ll probably need another string to
process_line(line):
hold the processed version of the line.

processed_line = ''

words = line.split()
for word in words:

return processed_line

For each word in text

A

B

3

If word is a placeholder (NOUN
or VERB or VERB_WITH_ING or
ADJECTIVE):

1

Prompt user for placeholder
part of speech.

2

Substitute user’s word for
the placeholder.

Otherwise, the word is fine;
keep it in the story.

Store results.
Take the processed text with
the placeholders filled, and write
it out to a file with the filename
prepended by “crazy_”.

Let’s split the line into a list of words…
…and then iterate over the list.
All our processing code
is going to go here.

And when we’re all
done, we’ll return the
processed line.

Now let’s process the text
Okay, that’s some good skeleton code, but now what? Following on with our
pseudocode, now we need to check for our placeholders, and if we have one
prompt the user for a replacement word.
placeholders = ['NOUN', 'ADJECTIVE', 'VERB_ING', 'VERB']
def process_line(line):
global placeholders
processed_line = ''
words = line.split()

Then we can check each word to
see if it is a placeholder.

for word in words:
if word in placeholders:
answer = input('Enter a ' + word + ":")
processed_line = processed_line + answer + ' '
else:
processed_line = processed_line + word + ' '
return processed_line + '\n'

We also need to put a newline back on
the line because split removes them.

Let’s just make a list
of the placeholders.
We will make it global
in case it is needed in
more than one function.
If it is, let’s
prompt the user.

And add their answer
to the processed line.
Otherwise, just add the word
(which isn’t a placeholder) to
the processed line.
you are here 4  411

testing our code

A Test Drive

It’s time to test Crazy Libs. Other than writing the text back to a file, we
should almost have everything working. Get the changes to the code in crazy.
py and choose the Run > Run Module menu item. We’ve reproduced all the
code below. Double-check your output.

def make_crazy_lib(filename):
file = open(filename, 'r')
text = ''
for line in file:
text = text + process_line(line) + '\n'
file.close()
return text
placeholders = ['NOUN', 'ADJECTIVE', 'VERB_ING', 'VERB']
def process_line(line):
global placeholders
processed_line = ''
words = line.split()

It’s worth one more look at the
structure of this code. The for
loop is stepping through every word
in a single line. And for each word
we see if the word is a placeholder,
and if so, ask the user for a word
to substitute in the processed line.

for word in words:
if word in placeholders:
answer = input('Enter a ' + word + ":")
processed_line = processed_line + answer + ' '
else:
processed_line = processed_line + word + ' '
Python 3.6.0 Shell
return processed_line + '\n'
def main():
lib = make_crazy_lib('lib.txt')
print(lib)
if __name__ == '__main__':
main()

Here’s what we got.
Hmm, the code seems to
be skipping some of our
placeholders.

We have a problem!

412  Chapter 9

Enter a VERB_ING:buying
Enter a VERB_ING:pudding
Enter a ADJECTIVE:forgetting
Enter a NOUN:monkey
Enter a VERB_ING:eating
The first thing that stands between
you
and buying your first, real, piec
e of NOUN,
is pudding the skill of breaking
problems down into achievable forg
etting
actions that a monkey can do for
you. Of
course, you and the computer will
also
need to be eating a common NOUN,
but
we’ll get to that topic in just a
bit.
>>>

saving and retrieving files

Why is the code mysteriously skipping some of the nouns? Is it something
to do with the nouns or something more systemic? Take a look at the input
and output (and of course the code) and see if you can figure out the
problem. Write your observations here.

We have three noun
placeholders, but we’re only
being asked for one.
Is there anything different
about the noun placeholders?
Or is something wrong
in the code?

Python 3.6.0 Shell
Enter a VERB_ING:buying
Enter a VERB_ING:pudding
Enter a ADJECTIVE:forgetting
Enter a NOUN:monkey
Enter a VERB_ING:eating
The first thing that stands between you
and buying your first, real, piece of NOUN,
is pudding the skill of breaking
problems down into achievable forgetting
actions that a monkey can do for you. Of
course, you and the computer will also
need to be eating a common NOUN, but
we'll get to that topic in just a bit.
>>>

Using a new string method to fix the bug
We’re going to solve this problem a bit differently than we did in Chapter 6 by
making use of a string method: strip. The strip method returns a new string
that has characters stripped off the beginning and ending of the string. Let’s see
how the strip method works:
hello = '!?are you there?!'
goodbye = '?fine be that !way!?!!'
hello = hello.strip('!?')
goodbye = goodbye.strip('!?')
print(hello)
print(goodbye)

All occurrences of !
and ? are stripped
from the beginning
and ending of the
strings hello and
goodbye.

Let's create a
couple strings.
The strip method takes a string of characters, and
removes any occurrence of each character from
the beginning and ending of the string.
Note that goodbye
Python 3.6.0 Shell
still has a !
are you there
character in it
fine be that !way
because it wasn't
at the end of the
string, so it wasn’t
stripped.
you are here 4  413

using strip to fix the bug

Using the strip method, spend some time reworking the process_
line function to correctly handle placeholders that end in punctuation
(commas, periods, semicolons, question marks, and so on). You not
only need to recognize placeholders correctly, but you also need to
preserve the punctuation in the output. Give your brain a real chance to
work on this, and then we’ll solve it together. No peeking ahead!

def process_line(line):
global placeholders
processed_line = ''
words = line.split()
for word in words:
if word in placeholders:
answer = input('Enter a ' + word + ":")
processed_line = processed_line + answer + ' '
else:
processed_line = processed_line + word + ' '

Here's the code
we need to
improve.

return processed_line + '\n'

Actually fixing the bug
Fixing this bug requires a few tweaks to our code. This is a great example of how a small
change to the functionality of your code, whether that is fixing a bug or adding a new feature,
can introduce a lot of changes to your code. We would have been more effective if we’d
caught this issue when we were writing our pseudocode.
In terms of tweaks, we’re going to strip each word of all punctuation, and then compare the
stripped version of the word against the placeholder. We also have to be careful to add any
punctuation back on to the user’s chosen word when we’re done. Here’s how you do that:

You'll find anytime
you are making
lots of changes to
existing code, you're
introducing the
possibility of even
more bugs.

def process_line(line):
First let’s strip the word of all periods,
global placeholders
commas, semicolons, and so on.
processed_line = ''
words = line.split()
We’ll test the stripped version against the placeholders.
for word in words:
stripped = word.strip('.,;?!')
We want to display the stripped placeholder
if stripped in placeholders:
text, not the one with punctuation.
answer = input('Enter a ' + stripped + ":")
processed_line = processed_line + answer
if word[-1] in '.,;?!':
If there was punctuation,
processed_line = processed_line + word[-1] + ' '
add it back, and then add
else:
the space. Otherwise, just
processed_line = processed_line + ' '
add the space.
else:
processed_line = processed_line + word + ' '
return processed_line + '\n'

414  Chapter 9

saving and retrieving files

A Test Drive

Make the changes from the previous page to your crazy.py file and then
choose the Run > Run Module menu item. Double-check your output.
Python 3.6.0 Shell

To test your own
story, just call
make_crazy_lib and
pass it the name of
your story’s file.

Finally, this is what
we’re looking for! We
hope you tried your
own set of verbs,
nouns, and adjectives.
Try your own story
now and see how it
works as well.

Enter a VERB_ING:buy
ing
Enter a NOUN:pudding
Enter a VERB_ING:for
getting
Enter a ADJECTIVE:cr
azy
Enter a NOUN:monkey
Enter a VERB_ING:eat
ing
Enter a NOUN:pizza
The first thing tha
t stands between you
and buying your fir
st, real, piece of
pudding,
is forgetting the ski
ll of breaking
problems down into
achievable crazy
actions that a monkey
can do for you. Of
course, you and the
computer will also
need to be eating a
common pizza, but
we’ll get to that top
ic in just a bit.
>>>

Some code has real problems
Give this a try: open your crazy.py file and change the filename from 'lib.txt' to

'lib2.txt', and then execute your code.

If you try to open a file for reading
and it doesn't exist, you'll get a
FileNotFoundError exception.
Python 3.6.0 Shell
Did our last bug fixes really
take care of all issues
we might encounter with
punctuation? Or can you
think of other ways the code
still is a bit buggy?
Hint: One issue—would ‘VERB!!!’ be
handled correctly in the output, or would
only one ! be added? How would you fix
that?

Traceback (most recent call last):
File "crazy.py", line 45, in 
main()
File "crazy.py", line 41, in main
crazy_lib = make_crazy_lib(filename)
File "crazy.py", line 2, in make_crazy_lib
file = open(filename, 'r')
FileNotFoundError: [Errno 2] No such file or directory: 'lib2.txt'
>>>

you are here 4  415

exploring exception

Let’s take a moment for a little exercise that is going to show you some
of the problems the Python interpreter has to deal with on a daily basis.
Then we’ll see what we can do about them with code. Your job is to act
like you’re the Python interpreter. You need to parse each piece of code
and execute it. Write any output or errors in the Python Shell window.

list = [1, 2, 3, 4]
item = list[5]

Actually executing
the code might help
too. A lot.
Write your
results here.

Python 3.6.0 Shell

Be the Python
Interpreter
.
filename = 'document' + 1 + '.txt'
Python 3.6.0 Shell

msg = 'hello'
def hi():
print(msg)
msg = 'hi'
hi()
Python 3.6.0 Shell

416  Chapter 9

int('1')
int('2')
int('E')
int('4')
int('5')
int('6')

Python 3.6.0 Shell

firstname = 'Beethoven'
print('First name: ' + name)
Python 3.6.0 Shell

saving and retrieving files

Handling exceptions
We’ve discussed errors before: they can be syntax errors (essentially,
typos in your code), semantic errors (logic problems with your
code), or runtime errors (things that go wrong during the execution
of your code). For runtime errors, so far we’ve just let them fall on
the floor with the Python interpreter spitting out an error and your
program coming to a grinding halt. But it doesn’t have to end this
way, especially when a runtime error occurs naturally in the course of
running your code—say that file you’re trying to open for reading just
isn’t there anymore. We’ll see how we can handle that in a sec.
Let’s talk a little more about what a runtime error, or exception,
actually is. An exception is an event that occurs during the execution
of your code when the Python interpreter encounters something it
can’t handle. When that happens, the interpreter stops the execution
of your code and creates an exception object that has information about
the error that occurred. By default this information is then displayed in
the form of a traceback error message, which you’re already familiar
with from the Python Shell.
As we said, there is a way to deal with these errors when they arise
without having the interepter just give up; in fact, in your code you can
tell the interpreter you’d like to take over when certain exceptions occur,
and handle them your own way. To do this wrap any code at risk in a
try/except block, like this:

Start with a try keyword…
try:

…then the block of code
where the error could
occur…
…then the except keyword.

filename = 'notthere.txt'

file = open(filename, 'r')

except:

print('Sorry, an error occured opening', filename)

else:

print('Glad we got that file open')
file.close()

You can also add an optional finally
block that runs whether an exception
occurs or not. More on that in a sec.

Next, add a block
of code to execute
if an exception
occurs.

And then add an
optional else block,
with code that will
execute only if no
exceptions occurred.

you are here 4  417

how to handle expectations

Explicitly handling exceptions
When we use the except keyword followed by a code block, you get a exception catchall—that is, the except code block will catch any exception that occurs in the try block.
You can be more precise by giving a specific exception name, like this:

Here’s our try
block again.
filename = 'notthere.txt'
This time we’ve added a specific
exception to the except statement.
file = open(filename, 'r')
So this block will
except FileNotFoundError:
execute only if a
print('Sorry,', filename, 'could not be found.')
FileNotFoundError
occurs.
except IsADirectoryError:
We can add more exceptions too;
print("That's a directory not a file!")
this one executes only if you try to
open a directory instead of a file.
else:
try:

print("It's a good thing we could open that file.")

As before, if all is good, this
block will execute.

file.close()

finally:

print("I'm running no matter what happens")

Here’s where you can add a
finally block; this block will
execute whether there is
an exception or not!

Q:

A:

Q:

A:

Q:

A:

Is there a limit on the number of
open files I can have?

Python does not place a limit on the
number of open files, but your operating
system does—that’s another reason you
should close files as you finish with them.
If is often possible to increase this limit,
although you’ll need to reconfigure your
operating system’s process limits to do so
(and you should ask yourself if you really
need to).

Q:

Is there a list of all the exceptions
I can get?

418  Chapter 9

Of course. As always, check out
python.org. Or, more specifically, https://
docs.python.org/3/library/exceptions.html.
Is it possible to create your own
exception types?

A:

Yes. Exceptions are just objects, so
you can extend Python by creating your
own, new exception objects. That’s a litte
beyond the scope of this chapter, but you’ll
find plenty on this online and at python.org.
You’ll want to read Chapter 11 too.

Can I catch multiple exceptions
using a single block? How do I do that?

After your except keyword, you can
have one or more exceptions separated by
commas within parentheses, like:

except (FileNotFoundError, IOError):

As you already know, you can also omit the
exception name and your except clause will
catch any exception that occurs.

saving and retrieving files

Trace through the code below three times. The first time enter any number
except 0. The next time, enter 0. The last time, enter the string "zero".
Write in the output you expect.

try:
num = input('Got a number? ')
result = 42 / int(num)
except ZeroDivisionError:
print("You can't divide by zero!")
except ValueError:
print("Excuse me, we asked for a number.")
else:
print('Your answer is', result)
finally:
print('Thanks for stopping by.')

Input a number other than 0
Python 3.6.0 Shell

Input 0
Python 3.6.0 Shell

Input the string "zero"
Python 3.6.0 Shell

you are here 4  419

adding exception handling

Updating Crazy Libs to handle exceptions
With your new knowledge of exceptions, let’s update the code that opens a file to
handle some file exceptions:
def make_crazy_lib(filename):
try:
file = open(filename, 'r')
text = ''

Most of this code is making
use of a file, so let's put it
all in a try block.

for line in file:
text = text + process_line(line)
file.close()
return text
except FileNotFoundError:
print("Sorry, couldn't find", filename + '.')
except IsADirectoryError:
print("Sorry", filename, 'is a directory.')
except:
print("Sorry, could not read", filename)

Notice we don't have an else or finally clause.
That takes care of the make_crazy_lib function, but there’s something
important to notice in this code: what gets returned from the function if we get an
exception? Because we are not explicitly returning anything, the value None will
get returned. Keep this in mind, because you’re going to need to remember this when we
write the code to save the Crazy Libs.

Let's see if we have a
file not found error or if
the user tried to open a
directory (assuming they
changed the filename in
the code).

And then let's catch any
other exceptions that
happen when processing
the file.
We told you we’d be
bumping into the None
value now and then.
Python 3.6.0 Shell

A Test Drive
Your errors
here may differ
depending on your
operating system.
420  Chapter 9

Get the code above into the file crazy.py and save
your code. Give it a run and double-check your
output. Change the template filename to a file
that doesn’t exist or a directory name and see
how it works.

You should see the same output as before.
Feel free to change the filename in
crazy.py to test out the exceptions.

Enter a VERB_ING:buy
ing
Enter a NOUN:pudding
Enter a VERB_ING:for
getting
Enter a ADJECTIVE:cr
azy
Enter a NOUN:monkey
Enter a VERB_ING:eat
ing
Enter a NOUN:pizza
The first thing tha
t stands between you
and buying your fir
st, real, piece of
pudding,
is forgetting the ski
ll of breaking
problems down into
achievable crazy
actions that a monkey
can do for you. Of
course, you and the
computer will also
need to be eating a
common pizza, but
we’ll get to that top
ic in just a bit.
>>>

saving and retrieving files

Our last step: storing the Crazy Lib
Saving your Crazy Lib or any text to a file is straightforward with
the file object’s write method. To use it we need to first open a file
in write mode. Let’s write a function save_crazy_lib, which takes a
filename and a string as parameters. When called, save_crazy_lib
will create the file, save the string to it, and then close the file. Check
out this code:

1

Read the text for the story
from file.

2

Process text.
For each word in text

A

B

Our save function expects a filename and some text
in the form of a string.

3

If word is a placeholder (NOUN
or VERB or VERB_WITH_ING or
ADJECTIVE):

1

Prompt user for placeholder
part of speech.

2

Substitute user’s word for
the placeholder.

Otherwise, the word is fine;
keep it in the story.

Store results.
Take the processed text with
the placeholders filled, and write
it out to a file with the filename
prepended by “crazy_”.

def save_crazy_lib(filename, text):
file = open(filename, "w")
file.write(text)
file.close()

Finally, we close
the file.

First we open the file in ‘w’ mode so we can
write to it. Python will create this file if
it doesn't already exist.
Then we use the file object write
method, and pass it the text string to
save to the file.

With a lot of programming languages,
if you don’t close a file you’ve written
to, then you can’t guarantee that the
data will be fully written to the file.

Updating the rest of the code

Be careful
opening files
in 'w' mode.

When you open
an existing file
using 'w' mode, its contents
will be erased and replaced by
whatever you write to the file. So
be careful!

Now that we have save_crazy_lib written, we need to call it
from the main function, and to do that we need to check to make
sure make_crazy_lib returned a Crazy Lib—remember, if it
encounters a file exception it will return None.

Let's store the crazy lib
filename in a variable to
make things more convenient.

def main():
filename = 'lib.txt'
lib = make_crazy_lib(filename)
print(lib)
if (lib != None):
save_crazy_lib('crazy_' + filename, lib)

If an exception occurs when opening
or reading the file, then lib will
have the value None. So, we need
to test lib to see if it has a value
before passing it to save_crazy_lib.

The filename we're going to save
the lib under is going to be “crazy_"
prepended to the original filename.
you are here 4  421

testing crazy libs

A Test Drive
It’s time to get a little crazy with
Crazy Libs, because we’re all
done; we just need to get these
last changes into the crazy.py file.
We’ve reproduced the entire game
on this page, and we’ve even
added in some exception code for
you as well. Go ahead and update
your code and give Crazy Libs a
try.

def make_crazy_lib(filename):
try:
file = open(filename, 'r')
text = ''

for line in file:
text = text + process_line(line)
file.close()
return text

except FileNotFoundError:
print("Sorry, couldn't find", filename + '.')
except IsADirectoryError:
print("Sorry", filename, 'is a directory.')
except:
print("Sorry, could not read", filename)
placeholders = ['NOUN', 'ADJECTIVE', 'VERB_ING', 'VERB']
def process_line(line):
global placeholders
processed_line = ''
words = line.split()
for word in words:
stripped = word.strip('.,;?!')
if stripped in placeholders:
answer = input('Enter a ' + stripped + ":")
processed_line = processed_line + answer
if word[-1] in '.,;?!':
processed_line = processed_line + word[-1] + ' '
else:
processed_line = processed_line + ' '
else:
processed_line = processed_line + word + ' '
return processed_line + '\n'

We went ahead and added the
exception code to the save_crazy_
lib for you. Check it out.

def save_crazy_lib(filename, text):
try:
file = open(filename, 'w')
file.write(text)
file.close()
except:
print("Sorry, couldn’t write file.", filename)
def main():
filename = 'lib.txt'
lib = make_crazy_lib(filename)
print(lib)
if (lib != None):
save_crazy_lib('crazy_' + filename, lib)
if __name__ == '__main__':
main()

422  Chapter 9

saving and retrieving files

I’d love to use this
on different templates without
having to open the Python source
file to change the filename. Could I
supply a template filename using my
command line?

Ah, yes, that would polish this off nicely.
We haven’t used Python on the command line, but
if you open your command line and change your
directory to the location of your crazy.py file, then you
can run your code like this on the Mac:
python3 crazy.py

and like this on your Windows machine:
python crazy.py

Let’s add a command-line argument to specify the
Crazy Libs template, like this:
python3 crazy.py lib.txt

Run the program crazy.py
and use the lib.txt template.

To get the lib.txt argument from the command line, we’ll
use Python’s sys module that includes an attribute
called argv. The argv attribute is a list that holds the
terms you typed into the command line (excluding the
Python command). For instance, if you type:
python3 crazy.py lib.txt

then item 0 of argv will hold crazy.py and item 1 will
hold lib.txt. Let’s see if we can use this, and polish off
our Crazy Lib game.

you are here 4  423

adding command line arguments

Let’s add that final piece of polish and allow your user to specify a filename on the
command line. You’ll do that with two simple additions; go ahead and work through
the two additions and then give your code another, final, test drive.

A Test Drive
1

First import the sys module by adding an import sys to the top
of your crazy.py file.

Add this to the top
of your file.

import sys

2

Then let’s make a few changes to the main function in crazy.py.

We want there to be two arguments
or the user hasn’t provided a
filename. So let them know if there
aren’t two arguments.

def main():

if len(sys.argv) != 2:

print("crazy.py ")

else:

filename = sys.argv[1]

lib = make_crazy_lib(filename)

Otherwise, the filename is the
command-line argument at
item index 1.

if (lib != None):

save_crazy_lib('crazy_' + filename, lib)

Terminal

This time we'll need to
run our Python from the
command line. If you're
using a Mac, it’s called
the Terminal, which
you can find in the
Applications > Utilities
folder.
If you’re on a PC, it’s
called the command
prompt, which you can
find by clicking the Start
button, typing “cmd”, and
then choosing Command
Prompt.
424  Chapter 9

$ cd /Users/eric/code/ch9
$ python3 crazy.py lib.txt

You'll need to navigate to the correct directory that
contains your crazy.py file using your operating system's
cd (short for change directory) command.

Enter a VERB_ING:running
Enter a NOUN:hotdog
Enter a VERB_ING:eating
Enter a ADJECTIVE:spicy
Enter a NOUN:taco
Enter a VERB_ING:breaking
Enter a NOUN:glass
$ cat lib.txt
The first thing that stands between you
and running your first, real, piece of hotdog,
is eating the skill of breaking
problems down into achievable spicy
actions that a taco can do for you. Of
course, you and the computer will also
need to be breaking a common glass, but
we'll get to that topic in just a bit.
$

Here's our command line on a Mac.

On Windows you’ll
need to run python,
not python3.

Remember to use the “python”
command on a Windows machine.
Here's what's in the file
crazy_lib.txt.

saving and retrieving files

 You have to open a file to access it from
Python.

 The break statement prematurely stops
the execution of a for or while loop.

 To open a file use Python’s open function.

 Most text files have a newline character
between lines of text.

 When you open a file you supply a mode:
'r' for reading or 'w' for writing.
 The open function can take either a
relative path or an absolute one.
 Call the file object’s close method when
you are done reading it or writing to it.
 The open function supports several forms
of reading standard text files.
 Use the read method to read the entire
contents of the file at once.
 Reading an entire file may be resource
intensive for large files.
 Use the readline method to read a single
line at a time.
 The empty string signifies when readline
has read the last line.
 Or you can treat the file as a sequence
and use it in a for statement, as in for
line in file:
 An iterator lets us iterate over a sequence
with the for/in statement.

 The escape sequence \n represents a
newline character.
 The strip method removes zero or more
occurrences of the specified characters
from the beginning and end of a string. If
no character is specified, strip defaults to
removing whitespace.
 Use try/except to capture exceptions.
Place the code that may generate an
error in the try block and then one or
more except statements to catch the
exceptions.
 An except without an explicit exception
acts as a catch-all.
 The finally statement is always executed,
whether or not an exception occurs.
 The sys module has an attribute argv that
contains your program’s command-line
arguments.
 The argv attribute holds a list that contains
each word used on the command line.

you are here 4  425

coding crossword

Coding cross
1

2
3

Let
files sink into your brain as you do the crossword.
5

4

6

1

7
9

4

8

5

10

9

2

3

7

11

6

8

10
11 12

12

13

14

13

14

15

16

15

18

17

16
17

19

18
19

Across
Across
3. Gets
you you
thethe
entire
file.
3. Gets
entire
file.
Use try/except
catchthese.
these.
4. Use4. try/except
to to
catch
9. Windows
Mac
disagree.
9. Windows
andand
Mac
disagree.
If your
isn't
relativeititmust
must be.
10. If 10.
your
pathpath
isn't
relative
be.
11. A huge file might lead to these errors.
11. A12.
huge
file
might
lead
to
these
errors.
End of file when you hit this.
12. End
fileyour
when
you hitline
this.arguments.
13. of
Holds
command
13. Holds
your
command
line
arguments.
14. Another name for \n.
15. How name
to access
file.
14. Another
fora \n.
17. Our
new game.
15. How
to access
a file.
18. new
You can
'r' or 'w'.
17. Our
game.
19. Cleans up the ends of strings.
18. You can 'r' or 'w'.
19. Cleans up the ends of strings.

426  Chapter 9

Down
Down
1. Type
of argv
1. Type
of argv
2. An
except
without
an explicit
exception.
2. An
except
without
an explicit
exception.
5. What's
\n? \n?
5. What's
6. Always
do this
you're you're
done. done.
6. Always
dowhen
this when
7. A design pattern.
7. A design pattern.
8. Just need one line?
8. Just with
needargv.
one line?
9. Module
9.
Module
with
argv. is executed.
13. When a finally statement
16.13.
When
you're
done. statement is executed.
When
a finally
16. When you're done.

saving and retrieving files

Hey, it’s Chapter 9 already, so we have no doubt you can tackle the task
of creating a little pseudocode based on the previous two pages. Writing
up a little pseudocode now is going to give you a clearer idea of how to
approach writing the Crazy Libs game.

Here’s what we came up with; depending
on the level of detail you went into, your
pseudocode may differ, but just make
sure we’re aligned on the general logic
before moving on.

1

Read the text for the story
from file.

2

Process text.
For each word in text

A

B

3

If word is a placeholder (NOUN
or VERB or VERB_ING or
ADJECTIVE):

1

Prompt user for placeholder
part of speech.

2

Substitute user’s word for
the placeholder in text.

Otherwise, the word is fine;
keep it in the story.

Store results.
Take the processed text with
the placeholders filled, and write
it out to a file with the filename
prepended by “crazy_”.

Somehow we’re going to read the
contents of the story into Python.
Then we’re going to start processing the
text by examining each word.
If we find a word that is
a placeholder, we need to
ask the user for a word
to replace it with.
And then we need to
substitute that word for
the placeholder in the
story text.
And if the word isn’t a
placeholder, then we just keep
the original word.

Finally, when we’re done
processing every word, we write
all the text of the story into
a new file.

you are here 4  427

exercise solution

Draw a line from each file path to the file it identifies. For any relative paths,
all code is run from the code directory.

/usr/harry/code/find.py
../inventory/map.pdf
../../hermione/spells/leviosa.txt
find.py

usr

data/students.txt

harry

code

inventory

cloak.gif

hermione

spells

data

map.pdf

oculus.txt

find.py

horcrux.db

428  Chapter 9

students.txt

leviosa.txt

saving and retrieving files

Code Magnets Solution

Can you help find the needle in the haystack? You’ve been given a directory of 1,000 files with the
filenames 0.txt through 999.txt, and only one has the word needle in it. Can you find the needle? We had
the code all written and on the refrigerator, but then someone came along and messed it all up. Can you
help us put it back together? Notice, there may be some extra code magnets, so you may not use all of
them.

Rearrange the
magnets here.
Iterate over all 1,000 files, constructing a
filename for each one of 0.txt, 1.txt, 2.txt...

for i in range(0, 1000):
filename = str(i) + '.txt'
file = open(filename, 'r')

Open the file for reading.
Read in the entire file, which
returns a string, assigned to text.

text = file.read()
if 'needle' in text:

Check to see if the needle
in the string, and...

print('Found needle in file ' + str(i) + '.txt')
file.close()

And don’t forget to
close each file.

…if found, print out
the filename.

Give it a try; look at the needle
directory in the ch9 source code
directory for the files 0.txt through
999.txt. Which file has the needle?

you are here 4  429

exercise solution

Why is the code mysteriously skipping some of the nouns? Is it something
to do with the nouns or something more systemic? Take a look at the input
and output (and of course the code) and see if you can figure out the
problem.

It looks like we’re skipping the
first noun, and the last.

The first thing that stands between you
and VERB_ING your first, real, piece of NOUN,
is VERB_ING the skill of breaking
problems down into achievable ADJECTIVE
actions that a NOUN can do for you. Of
course, you and the computer will also
need to be VERB_ING a common NOUN, but
we’ll get to that topic in just a bit.

If we look at the skipped noun
placeholders in the input (or
output) file, we can see what
is different is that they have
punctuation at the end of them.

430  Chapter 9

Python 3.6.0 Shell
Enter a VERB_ING:buying
Enter a VERB_ING:pudding
Enter a ADJECTIVE:forgetting
Enter a NOUN:monkey
Enter a VERB_ING:eating
The first thing that stands between you
and buying your first, real, piece of NOUN,
is pudding the skill of breaking
problems down into achievable forgetting
actions that a monkey can do for you. Of
course, you and the computer will also
need to be eating a common NOUN, but
we’ll get to that topic in just a bit.
>>>

Argh! It’s the same old bug we saw
in Chapter 6 with the text analysis.
We need to account for the comma
(and we’ll assume period as well)
when comparing text against these
placeholders.

saving and retrieving files

Let’s take a moment for a little exercise that is going to show you some
of the problems the Python interpreter has to deal with on a daily basis.
Then we’ll see what we can do about them with code. Your job is to act
like you’re the Python interpreter. You need to parse each piece of code
and execute it. Write any output or errors in the Python Shell window.
list = [1, 2, 3, 4]
item = list[5]

Python 3.6.0 Shell

Be the Python
Interpreter
Solution

Traceback (most recent call last):
File "/Users/eric/code/ch8/errors/list.py", line 2, in 
item = list[5]
IndexError: list index out of range
>>>

filename = 'document' + 1 + '.txt'

Python 3.6.0 Shell
Traceback (most recent call last):
File "/Users/eric/code/ch8/errors/filename.py", line 1, in 
filename = "document" + 1 + ".txt"
TypeError: must be str, not int
>>>

msg = 'hello'
def hi():
print(msg)
msg = 'hi'

int('1')
int('2')
int('E')
int('4')
int('5')
int('6')

hi()
Python 3.6.0 Shell
Traceback (most recent call last):
File "/Users/eric/code/ch8/errors/function.py", line 7, in 
hi()
File "/Users/eric/Documents/code/ch8/errors/function.py", line 4, in hi
print(msg)
UnboundLocalError: local variable 'msg' referenced before assignment
>>>

Python 3.6.0 Shell
Traceback (most recent call last):
File "/Users/eric/code/ch8/errors/ints.py", line 3, in 
int('E')
ValueError: invalid literal for int() with base 10: 'E'
>>>

firstname = 'Beethoven'
print('First name: ' + name)

Python 3.6.0 Shell
Traceback (most recent call last):
File "/Users/eric/code/ch8/errors/print.py", line 2, in 
print('First name: ' + name)
NameError: name 'name' is not defined
>>>

you are here 4  431

exercise solution

Trace through the code below three times. The first time, enter any number
except 0. The next time, enter 0. The last time, enter the string "zero".
Write in the output you expect.

try:
num = input('Got a number? ')
result = 42 / int(num)
except ZeroDivisionError:
print("You can't divide by zero!")
except ValueError:
print("Excuse me, we asked for a number.")
else:
print('Your answer is', result)
finally:
print('Thanks for stopping by.')

Input a number other than 0
Python 3.6.0 Shell

Got a number? 2
Your answer is 21.0
Thanks for stopping by.
>>>

Input the string "zero"
Python 3.6.0 Shell

Got a number? zero
Excuse me, we asked for a number.
Thanks for stopping by.
>>>

432  Chapter 9

Input 0
Python 3.6.0 Shell

Got a number? 0
You can’t divide by zero!
Thanks for stopping by.
>>>

saving and retrieving files

File i/o cross Solution
1

4

E

X

5

C E

P

T
7

C
S

E

Y
S

P
11

A

R A

P

8

I
T O R

E

A

6

C

C

R

L

H

E

O

A

S O

L

E

L

10

A

E M P

T Y
O

13

A

E

L

N

W

17

C R A

M O D E

Y
S

B

D

A

Q
U

18

T

E

S

O P

R E

O N S

M E M O R Y
12

15

I

C

3

I
S

9

2

L

D

T

U

T E

L
S

T R
14

R G V

I

N G

N

N E W L

I

N E

16

C

Z Y

L

I

B

S

T R

I

O

19

S

P

E

Across
3. Gets you the entire file. [READ]
4. Use try/except to catch these. [EXCEPTIONS]
9. Windows and Mac disagree. [SEPARATOR]
10. If your path isn't relative it must be.
[ABSOLUTE]
11. A huge file might lead to these errors.
[MEMORY]
12. End of file when you hit this.
[EMPTYSTRING]
13. Holds your command line arguments.
[ARGV]
14. Another name for \n. [NEWLINE]
15. How to access a file. [OPEN]
17. Our new game. [CRAZYLIBS]

Down
1. Type of argv [LIST]
2. An except without an explicit exception.
[CATCHALL]
5. What's \n? [ESCAPESEQUENCE]
6. Always do this when you're done. [CLOSE]
7. A design pattern. [ITERATOR]
8. Just need one line? [READLINE]
9. Module with argv. [SYS]
13. When a finally statement is executed.
[ALWAYS]
16. When you're done. [CLOSE]
you are here 4  433

10

using web apis

You Really Should
Get Out More

Yes, please do bring the
car around; we really should
get out and see some of these
Web APIs we’ve been hearing so
much about.

You’ve been writing some great code, but you really
need to get out more.There’s a whole world of data just waiting for
you on the web: Need weather data? Or how about access to a huge database
of recipes? Or are sports scores more your thing? Maybe a music database of
artists, albums, and songs? They’re all out there for the taking from Web APIs.
To use them all you need is to learn a bit more about how the web works, how
to speak the local web lingo, and how to use a couple of new Python modules:
requests and json. In this chapter we’re going to explore Web APIs and take
your Python skills to new heights; in fact, we’re going to take them all the way to
outer space and back.

You think we’re kidding!
this is a new chapter   435

using web apis

Extending your reach with Web APIs
You’ve seen a recurring pattern in this book: first you took your code and abstracted it
away into functions. Once you did that, you could make use of the functionality in a
function without having to worry about all the details of how the function does what
it does. This allowed you to elevate your thinking above the level of simple statements.
Next you packaged things up into modules that contained a whole set of functions
and values. Remember the code you handed to Cory? He was able to quickly look at
your documentation, see what functions were available, and make use of them. That’s
an API. Think of an API, otherwise known as an application programming interface, as a set
of documented functions that anyone can use in their code.
You’ve also seen there are modules created by other developers, which you can
think of as APIs, available to you in Python. These modules—modules for math
and random numbers and graphics and turtles and a lot more—really allow you to
expand the capabilities of your own code.
And now we’re going to take abstraction one step further by considering code on the
internet that you can use to extend the capabilities of your code. Only this code isn’t
necessarily written in Python, it’s code that runs on the web and it’s accessible through
a Web API.
You might be asking, though, what kind of API would we find on the web? How
about an API that computes and returns weather current conditions, or information
about songs and music artists? Or how about an API that gives you the actual
location of objects in outer space?
Though are all actual Web APIs and they’re just a few examples from a vast catalog
of information available to your code over the internet.

Pretend your boss just walked in and asked you to quickly write an
application that shows the current local weather conditions—and he
needs it for a demo tomorrow. How much would it help to use a Web
API that gives you the actual current weather conditions? How would
you write the app without such a service? How long do you think it
would take to write your app in each case?

436  Chapter 10

We’ve seen another
example of abstraction
with object-oriented
programming. We’ll be seeing
even more of that coming
up in Chapter 12 when we
learn how to create our
own objects.

web services

How Web APIs work
With Web APIs we don’t call a function in a module or library; instead, we
issue a request over the web. So, rather than looking at a set of functions that
a module provides, we look at the different web requests we can issue to a
web server to understand its Web API.
Issuing a web request is something you already know how to do if you
know how to use a web browser. And who doesn’t? You do it every time
you request a web page.The main difference with a Web API is that your
code makes the request to a web server, and that server sends back data,
rather than a page, to your code.

Sure!
67.2 degrees with
a 40% chance of rain

In this chapter we’re going to take a close look at how all this works, and
where Python fits into it—but for now, let’s take a look at how an actual
Web API request works:

To use a Web API, send it a request
Think of a Web API as just a web server, only it serves
up data rather than pages. To make use of a Web API,
your code sends a request to a web server. The web
server in turn generates a response, and sends it
back to your code.

Can I get the weather
conditions for London?

st

e
requ

resp

{

.2,
": 67 0%",
"4
ture
pera ob":
"tem cip_pr {
,
:
"pre ation" ondon"
"L
"loc ity":
"UK"
":
"c ntry
"cou
}

Your Python code

onse

A web server,
waiting for API
requests

}

Here's the
response; more
on this in a sec.
you are here 4  437

web apis and web addresses

All Web APIs have a web address
With Web APIs we use web addresses similar to the ones you type into
your browser—they contain a server name along with a description of
a resource. With a web browser, that resource is typically a web page.
With a Web API, the resource is data, which can vary greatly. The
way we specify the resource for a Web API can also vary quite a bit,
depending on the Web API. Let’s take a look at a few examples:

First of all, we call web addresses
“Uniform Resource Locators,” or URL
for short. URL is pronounced “Earl”
or “U-R-L.”
http://api.open-notify.org/iss-now.json

Then we have a description
Next we have
the
h
wit
rt
sta
ays
alw
of the resource we’re after.
They
the actual
in
ol,
toc
pro
n
tio
ica
Here we’re interested in
commun
server’s name
is
is
Th
.
TP
HT
e
cas
current location of the
the
this
on the internet.
by
d
use
ol
toc
pro
e
Space Station.
International
sam
the
r.
wse
bro
your
HTTP stands for the Hyptertext Transfer Protocol. Find out more about it at
https://en.wikipedia.org/wiki/Hypertext_Transfer_Protocol.
Let’s look at another example, this time weather data for the
city of London from the Open Weather Map organization:

Here’s another
URL to a web
service for
weather data.

Once again we
have the server
for the API.

And a very differentlooking description of
the resource we’re after.

http://api.openweathermap.org/data/2.5/weather?q=London

For this Web API
we start by telling
it we want data…
438  Chapter 10

…using version 2.5
of their API…

…and we’d like
the weather for
London.

web services

And one more example: how about looking up some
related artist information on Spotify?

Here we’re using version 1
of the API.

We’re asking for data for
a particular artist.

Here’s the ID of
the artist we’re
interested in.

http://api.spotify.com/v1/artists/43ZHCT0cAZBISjO8DG9PnE/related-artists

As always we have our
server name.

Any idea what
artist this
is? Would the
nickname “The
King” help?

And we’d like data
on related artists.

All of those
web addresses look
completely different, so how
are we supposed to know how
to make a request?

Consult the API documentation.
We’re serious. And it’s not as bad as it sounds. All Web APIs follow the
standard syntax for a web URL (and you can look that up if you’re
interested), but they expose very different APIs for very different types
of data. So, the way you request that data often looks quite different.
Look at our examples: the Open Notify URL is quite simple, while the
Spotify URL is more complex.
You’ll find that most Web APIs provide adequate documentation
for knowing how to form a request URL to access whatever type of
information you’re looking for. Hang tight, because we’re going to
explore Open Notify API in just a bit.

you are here 4  439

accessing web apis

I typed the
Spotify URL into my
browser and I got a “no
token” error.

Good point.
Making a Web API request through a browser is often
a great way to explore an API firsthand. That said,
many Web APIs require that you register (usually free)
as a developer before using their services. Registering
usually gets you an authorization token or an access key,
which you pass along when you’re making your requests.
Without the access token, many of these APIs will send
back an error, such as your “no token” error.
So, the best thing to do is to review the Web API’s
documentation for any access tokens it requires.
In our examples so far, the Spotify and Open Weather
Web APIs require access tokens. The Open Notify API,
however, doesn’t currently require a token.

Enter the following URL in your browser. What do you get?

http://api.open-notify.org/iss-now.json

Try retrieving this URL multiple times. You can also try entering the longitude and
latitude, like “-0.2609, 118.8982,” directly into Google Maps (latitude first).

440  Chapter 10

web services

Time for a quick upgrade
Before we start making requests to Web APIs, we’re going to
quickly add a new package to Python that will help us make those
requests. What’s a Python package? As we mentioned in Chapter
7, a package is just a formal name for a set of related Python
modules—you could also informally call it a library.
To add a new package we’re going to use the pip utility, which is
an acronym for “Pip Installs Packages.” Using pip, you can add
packages (or remove them later) to your local installation.

Can you say
“recursive acronym”?
Who might we be referring
to? Oh, just Twitter, Spotify,
Microsoft, Amazon, Lyft,
BuzzFeed, Reddit, the NSA, and
we could go on, and on, and on...

With pip, we’re going to add a handy package called requests,
which allows us to make requests to Web APIs. More specifically,
we’ll be using the requests module, which comes as part of the
requests package (not to be too confusing). Now we should tell
you that Python does have its own built-in module for making web
requests, but a lot of coders find requests easier to use and more
functional than the built-in module. Using the requests package
also gives us an excuse to gain some experience using pip.
Let’s get requests installed and then we’ll see what this package
is all about.

You can read more about the
requests package and module at:
http://docs.python-requests.org.

We often call packages
written by other developers
and organizations (outside of
Python’s core development team)
third-party packages.
you are here 4  441

installing the requests package

Doing the upgrade
The pip utility is built into Python, so get yourself into the
appropriate command line for your operating system and
let’s get the requests package installed.

[Note from editor:
Use your operating system’s command line like
love
we did in Chapter 9. If you’re using a Mac you’ll apparently, yokeu .]
want the terminal app, on a Windows machine use this Linux jo
the command prompt, and on Linux-oh, who are
we kidding? If you are using Linux, you live in the
command line.

Terminal
$ python3 -m pip install requests

Windows users
should use “python”
and not “python3”.

Go ahead and enter the command above to start the installation.
Note, you need to be connected to the internet for pip to
retrieve and install packages. Also note this is installing in
Python’s library directory, not your current working directory.
If you encounter any permission errors, make sure your user
account has sufficient privileges to install new packages.
Terminal
$ python3 -m pip install requests
Collecting requests
Downloading requests-2.18.1-py2.py3-none-any.whl (88kB)
100% |████████████████████████████████| 92kB 1.4MB/s
Collecting idna<2.6,>=2.5 (from requests)
Downloading idna-2.5-py2.py3-none-any.whl (55kB)
100% |████████████████████████████████| 61kB 2.5MB/s
Collecting urllib3<1.22,>=1.21.1 (from requests)
Downloading urllib3-1.21.1-py2.py3-none-any.whl (131kB)
100% |████████████████████████████████| 133kB 2.5MB/s
Collecting certifi>=2017.4.17 (from requests)
Downloading certifi-2017.4.17-py2.py3-none-any.whl (375kB)
100% |████████████████████████████████| 378kB 2.2MB/s
Collecting chardet<3.1.0,>=3.0.2 (from requests)
Downloading chardet-3.0.4-py2.py3-none-any.whl (133kB)
100% |████████████████████████████████| 143kB 4.1MB/s
Installing collected packages: idna, urllib3, certifi, chardet, requests
$

Here you can see pip retrieving the requests module, as well
as related packages it depends on. Note this is the macOS
X install, so your installation may look slightly different.
442  Chapter 10

Q:

Why do you prefer the requests module
over something built into Python? Wouldn’t
Python’s built-in module always be the way
to go?

A:

The creators of the requests package
created a library for making web requests that
is arguably more straightforward and provides
better functionality than what Python gives you
out of the box, so much so that it is in everyday
use by many online products and services. That
said, there is nothing wrong with Python’s built-in
requests module (which, by the way, is called
urllib, or is it urllib2? See, it’s already confusing!),
if you want to use that.

That’s one great thing about being able to
extend Python or most programming languages:
developers are free to create their own
extensions and share them with others.

Q:

That’s cool I can add new packages to
Python. How do I find out all the packages I
can add?

A:

You can search for new packages right
from the command line, like this:

python3 -m pip search hue

which will search for packages that match the
‘hue’ keyword. Another great way to find Python
packages is to search for them with Google using
search terms like “python3 request module” or
“python3 hue lighting,” as an example. Also, check
out http://pypi.org, a repository of software for
Python.

Q:

I’m running Python 3, but it doesn’t
seem to support pip.

A:

pip was added to Python with release 3.4;
double-check your version number and upgrade
to the latest version. Check with https://www.
python.org for the latest release.

web services

All we need now is a good Web API...
Now that you have the requests package installed, you’re finally ready
to make your first web request from Python. Of course to do that, you
need an interesting Web API to make a request of. If you’ll remember in
the beginning of this chapter we promised to take you to outer space and
back, and we’re going to do that using a Web API that provides the current
location of the International Space Station (ISS). You’ll find it at open-notify.
org. Let’s take a look at what Open Notify offers:

Visit open-notify.org
to check out more
documentation for the
ISS API. You’ll find the
highlights here.

Open Notify
supplies three
different services.

you are here 4  443

how web apis are documented

A closer look at the API
As you can see on open-notify.org’s front page, there are three “Open APIs from
space”: an API for getting the ISS’s current location, an API for making overpass
projections, and an API for getting the number of people in space. We’re going to
start by looking at the current location API, which you can see more detail on by
clicking on the “Current Location of...” link:

Here are the docs for the ISS
Current Location API.

If you were paying attention
earlier in the chapter, you’ve
already seen that this API
returns the current latitude and
longitude of the International
Space Station.
Here’s the request URL.

Here’s how to form the request URL to get
the current location.

And here’s an example of a
response from the API.
This looks like a set of key/value pairs.
Sound familiar? More on that in a bit.
We’re interested in the iss_position key, which
has a value that is another set of key/value pairs
holding the latitude and longitude.

The Open Space APIs were written
and are maintained by Nathan
Bergey, a rocketeer in Portland,
Oregon. You’ll find him on twitter
at @natronics.

444  Chapter 10

web services

The data returned from the Open Notify Web API should look a bit
familiar. Is there a Python data type it reminds you of?

Pronounced Jason, usually with
an emphasis on the “on.”

Web APIs provide data using JSON

Who’s JSON, you say? JSON is actually an object notation that most Web APIs use to
deliver their data. You’re going to find this notation familiar, because it is syntactically
quite similar to Python dictionaries.

There are other
formats for
exchanging data
over the web, but
you’ll find JSON is
the most popular.

Think of JSON as a set of key/value pairs specified as text. For example, say a weather
Web API wants to send you the current conditions for London. It would create JSON
that looks something like this and ship it your way:

The JSON
format wraps a
set of key/value
pairs in curly
braces.

{

All keys are
strings in JSON.

Note you can also
specify lists, or, what
JSON calls an array,
with JSON.

}

Each pair consists of
a string that acts as
a key, and a value...
"temperature": 67.2,
"precip_prob": "40%",
"location": {
"city": "London",
"country": "UK"
}

When used with a Web API, this entire
definition is placed into a text string
and sent back as a response.

...values can be
strings or numbers.
Values can even be
another, nested set
of key/value pairs.

st

e
requ

nse

{

o
resp

2,
": 67. %",
rature ": "40
"tempe p_prob {
"preci ion": ndon",
"Lo
"locatty": : "UK"
"ci
ry"
"count
}

}

JSON being
returned
you are here 4  445

more on json and dictionaries

Q:

That said, it is rather nice the two are so
much alike.

A:

So if I’m going to receive JSON as
just text, how can my code use that?

So are Python dictionaries and
JSON the same thing?

It is easy to confuse the two given
how much they look alike. Think of it like
this: in Python when you specify code for
a dictionary, at some point the interpreter
reads that code and translates it to an
internal data structure, which your code
makes use of. But with JSON, we’re
specifying a set of key/value pairs that we
can send over the network purely in a text
format to another program or service. So
JSON isn’t Python; rather, it’s meant to be
a general-purpose format than can be read
and interpreted by any language.

Q:

A:
Q:
A:

Hang tight—you’re asking the right
question, and we’ll get to that in just a bit.
Why is it called JSON?

Q:

I’m still confused. How do I know
what Web APIs are out there and how I
can use them?

A:

You’ll find clearinghouses for
web-based APIs on sites such as www.
programmableweb.com. Most companies
with APIs also provide documentation, like
dev.twitter.com, or developer.spotify.com, to
name a couple.

While JSON is meant to be a format
independent of any programming language,
it did grow out of one language in particular:
JavaScript. So JSON stands for JavaScript
Object Notation.

Take the JSON on the previous page and convert it, by hand, into a Python
dictionary in the space below. After you’ve done that, complete the code
below. You’ll need to look at the output to know what it does.

current =

______________________________________
______________________________________
______________________________________
______________________________________
______________________________________
______________________________________

Soon you’ll ask
your code to do
this for you.

______________________________________
loc = _________________________
print('In',

loc['city'] + ', ' + ____________,
'it is',

___________________,'degrees')

446  Chapter 10

Python 3.6.0 Shell

In London, UK it is 67.
2 degrees
>>>

web services

Now let’s look at that request module again
To make a web request we’re going to use the requests module’s get
function. Here’s how it works:

1

Use get to make the request.
First we call the requests module’s get function to make the
actual request to the remote web server.

url = 'http://api.open-notify.org/iss-now.json'

Here’s the URL we got from the
Open Notify documentation.

We pass the URL to the get
function, upon which the function
contacts the remote web server.

response = requests.get(url)

When a response is received from the
remote server, the function first
packages it up in a nice response object,
and then returns it.

When you call the get function,
a request is sent to the web
server at the supplied URL.

)
(url

get
est.

requ

text
status_code
headers

Response object

Your Python code

nse

o
resp

.2,
": 67 0%",
"4
ture
{
pera ob":
"tem cip_pr {
,
:
"pre ation" ondon"
"L
oc
"l ity":
"UK"
"c ntry":
"cou
}

The open-notify.org
web server receives the
request and sends back a
response in JSON format.

}

When the get function receives
the response it packages it up
into a response object. We’ll look
at this object next and see how
to use it.
you are here 4  447

the response object

2

Examine the response object.
As we said, after the get function receives a response from the
server it packages it up in a response object. Let’s take a look:

Here are some of the more
important attributes in
the response object. For
our purposes the text and
status_code attributes are
the ones to pay attention to.

The text property holds the real data of
the response, typically in JSON format.

text

The status code holds a diagnostic code
for the request. A code of 200 means the
request was fulfilled without errors.

status_code
headers

The headers are used to supply other
information about the response, such
as the content type of the response. For
instance, the JSON encoding is called
application/json.

Response object
response

3

Check the status_code and grab the text data.
When we get a response from the web server, we need to first check
its status code. A code of 200 means that the request was satifised. Of
course other codes are possible as well, such as error codes. Assuming
we get a 200 status code, we then use the text attribute to obtain the
data returned from the Web API. Here’s another little code fragment to
do that. Check it out and we’ll put it all together on the next page.

We’re using the response
object we got in 1 .

Let’s check the status code
and make sure we got a
code of 200, for success.

if (response.status_code == 200):
If the status code was 200, we should
print(response.text)
have some data in the text attribute.
else:
print("Houston, we have a problem:", response.status_code)

Otherwise, we've got an issue. For now we'll print
the error. Remember, you can find a complete list
of common status codes at https://en.wikipedia.org/
wiki/List_of_HTTP_status_codes.
448  Chapter 10

web services

Putting it all together: making a request to Open Notify
Alright, we truly have all we need to make a request, so let’s do that. You know how to call the get
function with a URL to make an actual request to a Web API, and we know from the open-notify.org page
what URL we need. You also know that the response object has a status_code you should check (and
make sure we get a code of 200, meaning success) before you use the response object’s text attribute.
So, let’s now put it all together and reach out and touch the Open Notify Web API:

First, import the
requests module.

Let’s set a variable url to
e URL
address of the ISS web serth
vice.

import requests

Here’s the
response
object.

url = 'http://api.open-notify.org/iss-now.json'
response = requests.get(url)

Then we use the get function, passing it
the URL we’d like to get data from.
Let's check the status code to make sure
everything was okay (that is, code 200).

if (response.status_code == 200):
print(response.text)
else:
print("Houston, we have a problem:", response.status_code)

If the response code wasn't 200,
we've got some kind of issue, so let's
print out what it is. You can find a
full list of status code on Wikipedia:
https://en.wikipedia.org/wiki/List_
of_HTTP_status_codes.

And then let’s print the
response to see what we got.
Remember, we’re printing
the text property of the
response, which holds the
data sent back to us from
the Web API.

...
Let’s do a quick sanity check before proceeding
g, healthy
At the time of writing, the ISS location service is a workin
so if for some
future;
the
in
tees
guaran
no
service. Of course there are
reading this
you’re
as
ing
operat
not
is
e
servic
n
reason the ISS locatio
page http://
the
book in 2036, we have a backup plan for you—check out
ng in
anythi
alter
to
wickedlysmart.com/hflearntocode to see if you need
check
to
hurts
never
it
but
the rest of this chapter; most likely you won’t,
first!

you are here 4  449

json and python

A Test Drive Flight

Get the code on the previous page into a file called iss.py, save your
code, and choose the Run > Run Module menu item to make sure
everything’s good.

You need a live internet
connection for this test drive!

Here's what we got.

Problems?
output,
If you aren't seeing this
is
e
ther
if
see
to
ck
che
first
200
a status code other than
it up to see what it
being reported (and look
us code? Paste
stat
a
ing
see
Not
).
signifies
make sure you
to
r
wse
bro
r
you
the URL into
ice is up and
serv
the
and
have connectivity
ck your code, and
running. Next, double-che
uest package really
double-check that the req
eptions reported to
exc
any
ck
Che
d.
alle
got inst
as if something is
s
ear
app
it
If
l.
wel
the shell as
ice, then revisit
serv
ify
Not
n
wrong with the Ope
e.
pag
s
viou
the Watch it! on the pre

Python 3.6.0 Shell
{"iss_position": {"longitude": "-146.2862", "latitude":
"-51.0667"}, "message": "success", "timestamp": 1507904011}
>>>

Any idea where this
location is?

How to use JSON in Python
So we’ve got our response from the ISS location service, and we’re seeing JSON in the response,
but right now it’s just a text string, which isn’t of much use, other than the fact we can print it. That’s
where Python’s json module comes in; the json module gives us a function, loads, which takes a
string containing JSON and translates it into a Python dictionary. How convenient is that? Here’s an
example of how the json module’s load function works:

import json

First we need to import
the json module.

And here’s a string containing JSON.
Note, to Python, this is just a string
with some text it.

json_string = '{"first": "Emmett", "last": "Brown", "prefix": "Dr."}'
name = json.loads(json_string)

So let’s call json.loads on the string, and assign the resulting
dictionary to the variable name.

print(name['prefix'], name['first'], name['last'])

And finally let’s use the
Python dictionary to
access the first and last
name attributes along with
the prefix attribute.
450  Chapter 10

Just a reminder: if you go
testing this code, do not call it
json.py (the same name as the
module json). We discussed that
issue in Chapter 7.

Python 3.6.0 Shell
Dr. Emmett Brown
>>>

web services

Using the JSON module on our ISS data
With the json module now on our toolbelt, let’s get it into our code and
get access to the space station’s latitude and longitude. Doing that is pretty
straightforward: we already have the JSON from the Open Notify servers
in our response object, so we just need to use the loads function from
the json module to convert that to a Python dictionary.

import requests, json

Add an import for the json module. We’re
separating module names by commas so we
can put them on the same line.

url = 'http://api.open-notify.org/iss-now.json'

Now let’s use json.loads to take the
JSON response, in string form, and
convert it to a Python dictionary.
(response.status_code == 200):
response_dictionary = json.loads(response.text)
Now let’s grab the
print(response.text)
iss_position value, which is
itself another dictionary.
position = response_dictionary['iss_position']

response = requests.get(url)
if

And then print
the lat and
print('International Space
long values from
position['latitude'] +
the position
dictionary. else:

Station at ' +
', ' + position['longitude'])

print("Houston, we have a problem:", response.status_code)

A Test Drive Flight

Make the updates above in your iss.py file, save your code, and give
it another test run.

Here's what we got. Looking nice!
It’s unlikely the ISS
What about this
will be in exactly the location? Any idea
same location when
where it is?
you run your code, so
your results will be
slightly different.

Python 3.6.0 Shell
International Space Station at -51.5770, -108.2028
>>>

you are here 4  451

adding a graphical display

Let’s add some graphics
Displaying the ISS location in text is, well,
just not acceptable. After all, this is Chapter
10, so we should be displaying the ISS’s
location on a nice map, or something. Well,
we have the technology—we’re going to dust
off our Python turtles and have a nice ISS
display in no time. We kid you not.

452  Chapter 10

“If I see one
more Python
Shell app,
you’re
fired!”

web services

Meet the screen object
Like we said, dust off those turtle skills. We’re actually going to
take those skills a little further by once again using the screen
object we saw in Chapter 7. Recall that the screen object lets you
change a few properties associated with the turtle window.
So let’s add the turtle module to this code, and see how we’re
going to use the screen object. You can go ahead and make these
additions to your iss.py file.

Add the turtle module.
import requests, json, turtle

We still say “window” would
have been a better name
than screen for this object,
but they didn’t really ask
for our opinion.

This first line just gets a reference
to the turtle’s screen object.
We’re going to increase the size of the window to 1000 by 500
pixels, which matches the size of the image we’re about to add.

screen = turtle.Screen()
screen.setup(1000,500)
screen.bgpic('earth.gif')
screen.setworldcoordinates(-180, -90, 180, 90)

This sets the entire window’s
background to an image.
Get the “earth.gif”
url = 'http://api.open-notify.org/iss-now.json'
image from the book’s
This resets the coordinate system source code ch10 folder.
response = requests.get(url)
of the turtle window; we talk
about this on the next page.

if (response.status_code == 200):
response_dictionary = json.loads(response.text)
position = response_dictionary['iss_position']
print('International Space Station at ' +
position['latitude'] + ', ' + position['longitude'])
else:
print("Houston, we have a problem:", response.status_code)
turtle.mainloop()

Just some housekeeping to
display the turtle window,
like in Chapter 7
You'll find this image in the
book's ch10 folde;, grab it and
place it in your own folder.
earth.gif
you are here 4  453

adjusting the coordinate system

Turtle Coordinates & Earth Coordinates
Let’s take a closer look at the call to the setworldcoordinates method, which you’ll see
plays a big role in determining how we locate the ISS’s position on our graphical map.

Center of turtle world
As you already know, the turtles live on a grid that
is centered on the x and y-coordinate 0,0. We’ve set
the grid to be 1000×500, so our grid ranges from
–500,–250 in the bottom-left corner to 500,250 in the
upper-right corner.

–500,250

500,250

0,0

–500,–250

500,–250

Prime Meridian
180,90

–180,90

The Earth has its own grid system with latititude
lines running west and east and longitude lines
running north and south. Longitude runs from
–180 to 180, with 0 being at the Prime Meridian.
For latitude, the equator is at 0, running to 90 at
the North Pole and –90 and at the south-most
point.

0,0

–180,–90

180,–90

We can easily reset the grid system of the turtles, and
have our turtle grid match the earth coordinates with
the statement:

–180,90

0,0

screen.setworldcoordinates(-180, -90, 180, 90)
which resets the turtle grid so the lower-left
coordinate is –180, –90 and the upper-right
coordinate is 180, 90. Our screen will retain its shape
and size, but all turtle movement will be mapped
using the new coordinate system.

This means we can now use common longitude and latitude
coordinates to position the turtles on the screen.
454  Chapter 10

180,90

–180,–90

180,–90

Using the setworldcoordinates method
we’ve set the coordinate system of the
turtles to match that of the Earth.

web services

Let’s add a turtle to represent the ISS
Have you figured out how we’re going to use a turtle to track the location of the ISS on the map? Well, let’s
add a turtle and find out. Make sure you add the code below to your iss.py file:
import requests, json, turtle
screen = turtle.Screen()
screen.setup(1000,500)
screen.bgpic('earth.gif')
screen.setworldcoordinates(-180, -90, 180, 90)
iss = turtle.Turtle()
iss.shape('circle')
iss.color('red')

Let's instantiate a turtle,
change its shape to a circle,
and make its color red.

url = 'http://api.open-notify.org/iss-now.json'

On the map, this circle is
going to represent (for
now) the location of the
ISS over the earth.

response = requests.get(url)
if (response.status_code == 200):
response_dictionary = json.loads(response.text)
position = response_dictionary['iss_position']
print('International Space Station at ' +
position['latitude'] + ', ' + position['longitude'])
else:
print("Houston, we have a problem:", response.status_code)
turtle.mainloop()

A Test Drive Flight

Oh, is
that all?

Let’s do a quick test before going further. Make the code additions
from the past few pages, and then make sure you have the image
earth.gif (again, you’ll find it in the ch10 folder in the book’s source
code) in the same directory as your code. Then give it a run.

If you see this, you’re in business.
All we’re doing so far is changing
the windows resolution, displaying
a background image, changing the
shape and color of the turtle, and
changing the coordinate system.

You should see a red circle, which is
really a turtle, positioned at its default
starting place at 0,0, only this time
that’s a longitude and latitude.

Python 3.6.0 Shell
International Space Station at -47.8777, -177.6666
>>>

you are here 4  455

using turtles to display the iss

Turtles can look like space stations too
Although it isn’t essential to building an app that uses a Web API, making our
turtle look like a space station on screen certainly adds to the realism and fun.
Here’s a quick piece of code to do just that:
import requests, json, turtle
screen = turtle.Screen()
screen.setup(1000,500)
screen.bgpic('earth.gif')
screen.setworldcoordinates(-180, -90, 180, 90)
iss = turtle.Turtle()
turtle.register_shape("iss.gif")
iss.shape("iss.gif")
iss.shape('circle')
iss.color('red')

You’ll find “iss.gif”
also in the ch10
source code folder.

Let the turtle module know we
want to use an image as a shape.
Then set the shape of the turtle
to be the “iss.gif" image.

url = 'http://api.open-notify.org/iss-now.json'
response = requests.get(url)

Note this step of
registering a shape
is an oddity of the
turtle module. In
other words, that’s
just the way it works.
Many image-based
libraries don't require
a registration step
like this.

if (response.status_code == 200):
response_dictionary = json.loads(response.text)
position = response_dictionary['iss_position']
print('International Space Station at ' +
position['latitude'] + ', ' + position['longitude'])
else:
print("Houston, we have a problem:", response.status_code)
turtle.mainloop()

A Test Drive Flight

Okay, make the code changes above and make sure the image iss.gif
is in the same folder as your code. You’ll find this image as well in the
book’s source code under the folder ch10.

Nice! We have a space
station image now!
Python 3.6.0 Shell
International Space Station at -47.8777, -177.6666
>>>

456  Chapter 10

web services

Forget the ISS—where are we?
We’re close! We’ve got everything in place and we’ve got a location back from
the Open Notify service, so all we need to do is put it all together. For starters
we need to actually get the latitude and longitude assigned to variables so we
can instruct our iss turtle where to move on the screen. One note, though:
did you notice that in Open Notify’s JSON those values are being represented
by strings, not numeric values? We’ll have to fix that. In fact, let’s look at how
to do that now:

Here's just a code fragment for now, the
part of the code that checks the response
and gets the latitude and longitude.
if (response.status_code == 200):
response_dictionary = json.loads(response.text)
position = response_dictionary['iss_position']
print('International Space Station at ' +
position['latitude'] + ', ' + position['longitude'])
lat = float(position['latitude'])
Let's get the latitude and longitude from
long = float(position['longitude'])
the position dictionary. Given they are both
strings, we'll convert them to floats first.
else:
print("Houston, we have a problem:", response.status_code)

You don't need to enter this code yet; we'll
make all the changes at once in a couple pages.
The next step is moving the ISS on the screen. You’re an old pro at this point
with turtles. All we need to do here is write a simple function—let’s call it
move_iss—that takes a numeric latitude and longitude and moves the turtle.
Let’s make sure and lift up the pen before we move.

Note we don't
technically have to put
the pen back down, but
we're just being a good
citizen given that is
Let's move to the right coordinate. Note that the default position
lat)
longitude is first, given the turtle takes its
of the pen. After all,
x-coordinate parameter first.
another developer may
be confused by this if
Now let's return the pen to its default down
they expect you to not
position and make the turtle visible.
change the pen in your
function.

def move_iss(lat, long):
global iss
iss.penup()
iss.goto(long,
iss.pendown()

The move_iss function takes a
latitude and longitude.
If we don't lift the pen, we're going to draw a
line across the display.

you are here 4  457

completing the implementation

Finishing off the ISS code
Let’s add the function move_iss to our code, as well as the code to get the latitude
and longitude from the position dictionary, and then we’ll call move_iss with
those values:
import requests, json, turtle
def move_iss(lat, long):
global iss
iss.penup()
iss.goto(long, lat)
iss.pendown()

We're just adding move_iss near
the top of the file.

screen = turtle.Screen()
screen.setup(1000,500)
screen.bgpic('earth.gif')
screen.setworldcoordinates(-180, -90, 180, 90)
iss = turtle.Turtle()
turtle.register_shape("iss.gif")
iss.shape("iss.gif")
url = 'http://api.open-notify.org/iss-now.json'
response = requests.get(url)

We no longer need a print
statement for the location—we've
got graphics now!

if (response.status_code == 200):
response_dictionary = json.loads(response.text)
position = response_dictionary['iss_position']
print('International Space Station at ' +
position['latitude'] + ', ' + position['longitude'])
lat = float(position['latitude'])
Let's take the latitude and longitude strings,
long = float(position['longitude'])
convert them to floats, and pass them to
move_iss(lat, long)
move_iss along with our iss turtle.
else:
print("Houston, we have a problem:", response.status_code)
turtle.mainloop()

458  Chapter 10

web services

A Test Drive Flight

Alright, make sure you’ve got your code all entered and cleaned up
and then you should be ready to finally see the ISS location on your
map. Give it a test run!

The ISS circles the earth every 92
minutes, so run your code a few times
to see the position change.

The ISS is also traveling at
4.75 miles per second.

Nice! When we ran our test
the ISS was somewhere over
the Indian Ocean.

Congrats! Mission
accomplished!

you are here 4  459

tracking the iss over time

Brain Building
Well, that was fanstastic: with a bit of code we’ve located the ISS and created a cool graphical display. But you
know what would make this code even cooler is if it tracked the ISS over time, rather than just tracking it one time.
Let’s do that, but before we do, this code could really use some cleanup. After all, we aren’t using all those great
abstraction muscles we’ve been building. Here’s our attempt to refactor this code a bit. See if it isn’t a lot more
readable (which you’re going to want on the next page):
import requests, json, turtle
iss = turtle.Turtle()
def setup(window):
global iss

Let's put all the window and
turtle setup in a setup function.

window.setup(1000,500)
window.bgpic('earth.gif')
window.setworldcoordinates(-180, -90, 180, 90)
turtle.register_shape("iss.gif")
iss.shape("iss.gif")
def move_iss(lat, long):
global iss
iss.penup()
iss.goto(long, lat)
iss.pendown()

We didn't make
any changes to the
move_iss function.

Notice we used “window” as the
parameter name here (which is
technically going to be passed
as a screen object). This might
or might not have been a good
variable naming idea, but we
couldn’t help ourselves. We
should probably add a real code
comment to alert others.

We took the code
that talks to the
Web API and put
it in a function
called track_iss.

def track_iss():
url = 'http://api.open-notify.org/iss-now.json'
response = requests.get(url)
if (response.status_code == 200):
response_dictionary = json.loads(response.text)
position = response_dictionary['iss_position']
lat = float(position['latitude'])
long = float(position['longitude'])
move_iss(lat, long)
else:
print("Houston, we have a problem:", response.status_code)
def main():
global iss
screen = turtle.Screen()
setup(screen)
track_iss()
if __name__ == "__main__":
main()
turtle.mainloop()

460  Chapter 10

And finally we're being
good coders by providing
a main function.

web services

Alright, here’s the code to track the ISS in real time. With this code you’ll see the location of the ISS change
every 5 seconds. This being a Brain Building exercise, your job is to figure out how it works. Read the code
and see if you can guess, and then in the next chapter we’ll find out exactly how it does what it does.
import requests, json, turtle
iss = turtle.Turtle()
def setup(window):
global iss
window.setup(1000,500)
window.bgpic('earth.gif')
window.setworldcoordinates(-180, -90, 180, 90)
turtle.register_shape("iss.gif")
iss.shape("iss.gif")
def move_iss(lat, long):
global iss
iss.hideturtle()
iss.penup()
iss.goto(long, lat)
iss.pendown()
iss.showturtle()

We've added only two lines of code.

def track_iss():
url = 'http://api.open-notify.org/iss-now.json'
response = requests.get(url)
if (response.status_code == 200):
response_dictionary = json.loads(response.text)
position = response_dictionary['iss_position']
lat = float(position['latitude'])
With this code you'll see
long = float(position['longitude'])
the ISS location update
move_iss(lat, long)
every 5 seconds.
else:
print("Houston, we have a problem:", response.status_code)
widget = turtle.getcanvas()
widget.after(5000, track_iss)
def main():
global iss
screen = turtle.Screen()
setup(screen)
track_iss()
if __name__ == "__main__":
main()
turtle.mainloop()

you are here 4  461

review of web apis

 Using Python you can communicate
with Web APIs and incorporate their
data and services into your application.

 The requests package is freely available
and comes from an open source effort
to make web requests in Python easier.

 A Web API typically provides a
documented Application Programming
Interface (API) that describes the
data it can provide and services it has
available.

 A package is a collection of Python
modules.

 Often you’ll need to register and obtain
a key or authorization token to use a
Web API.

 You can install the requests package
using the pip utility.
 pip stands for “pip installs packages”
 Using the requests library we issued a
request using the get method.

 To use a Web API from Python, you
make a request to it using the web’s
HTTP protocol, just like your browser
does.

 The get method returns a response
object, which includes a status code,
the text of the response (often in
JSON), and the headers of the request.

 Web services respond to requests,
sending data, often using the JSON
notation.

 The status code is 200 when the
request was satisfied.

 JSON stands for JavaScript Object
Notation, and was created to provide
a standard way of data exchange
between any languages.
 JSON’s syntax is very similar to Python
dictionaries.

462  Chapter 10

 The built-in json module provides
methods to take a JSON string and
translate it to a Python dictionary or list.
 The screen object of the turtle library
allows us to set a background image as
well as reset the coordinate system of
the turtle grid.

web services

Coding Crossword
Welcome back to earth. Let’s give your right brain something
different to do.
As always, it’s your standard crossword, but all of the solution
words are from this chapter.

1

1
2

3

3

7

4

5

4

5

6

6

7

8
10

2

9

8

9

10
11

12

11

12

13

13
14
16

15

14

17

16

17

15

18

18

Across
2. WeAcross
had to register one of these with the

Down

Down
1. He wrote
the ISS location service.
turtle
station?
2. module.
We had to register one of these with the4. Which1.space
He wrote
the ISS location service.
3. What you
aremodule.
when you get a 200 status
5. Open 4.
source
package.
turtle
Which
space station?
code.
pip source
install? package.
3. What you are when you get a 200 status6. What does
5. Open
7. To access
a Web API you may need this.
9. Minutes
per orbit.
code.
6. What
does pip install?
8. Our 7.
code
neededathis.
used for per
weborbit.
requests.
To access
Web API you may need this.13. Protocol
9. Minutes
10. Format
for code
data exchange.
14. Turtles
on this.used for web requests.
8. Our
needed this.
13.live
Protocol
11. Run
north
and for
south.
17. Documented
functions
a Web API.
10.
Format
data exchange.
14. Turtles
live on of
this.
12. 200 is a good value.
11. Run north and south.
17. Documented functions of a Web API.
15. Data attribute of the response object.
12. 200 is a good value.
16. Language JSON was originally from.
15. Data attribute of the response object.
18. What's in a package?
16. Language JSON was originally from.
18. What's in a package?

you are here 4  463

exercise solution

Enter the following URL in your browser. What do you get?

http://api.open-notify.org/iss-now.json

Here's what
we got.

Very cool, we just got the current location of the space station.
Note it contains a few keys and that the iss_position key holds
another set of key/value pairs (the longitude and latitude).
{
pend on
"message": "success",
The formatting will deow
sers
your browser. Some bris int
"timestamp": 1500664795,
oa
may even download then open
"iss_position":
file, which you can th
{
as a text file.
}

}

"longitude": "-110.6066",
"latitude": "-50.4185"

Note that the longitude and latitude are being represented
as strings. We'll have to convert them to floats when the time
comes to use them.

Take the JSON on the previous page and convert it, by hand, into a Python
dictionary in the space below. After you’ve done that, complete the code
below. If you need a hint, look at the output.

current = {'temperature': 67.2,

'precip_prob': '40%',
'location': {

'city': 'London',

}

}

'country': 'UK'

loc = current['location']
print('In',

loc['city'] + ', ' + loc['country'],
'it is',

current['temperature'],'degrees')

464  Chapter 10

Python 3.6.0 Shell

In London, UK it is 67.
2 degrees

web services

1
2

3

Coding
Cross
Solution

S
7

A

T

A

U

4

J

12

S

T
I

S

F

I

E

D

T H O R

I

Z A

R E

F A

C

17

A
P
I

A
L

I

C O D E

E

A

Q

C

U

K

9

N G

E

A

I

S

G

S

C R

P

I

T
P

T U D E
S

S

Y

T E

I

I

T

H

15

G

P

A

E

13
14

6

R

O N

L O N G

T U S

N
V

I

11

A

O
J

T

T O R

S O N
T A

5

H

E

I

16

M A G E

N

8
10

I

N

X
18

T
W

M O D U

L

E

S

D

Across
2. We had to register one of these with the
turtle module. [IMAGE]
3. What you are when you get a 200 status
code. [SATISFIED]
7. To access a Web API you may need this.
[AUTHORIZATION]
8. Our code needed this. [REFACTORING]
10. Format for data exchange. [JSON]
11. Run north and south. [LONGITUDE]
12. 200 is a good value. [STATUSCODE]
15. Data attribute of the response object. [TEXT]

Down
1. He wrote the ISS location service. [NATHAN]
4. Which space station? [INTERNATIONAL]
5. Open source package. [REQUESTS]
6. What does pip install? [PACKAGES]
9. Minutes per orbit. [NINETYTWO]
13. Protocol used for web requests. [HTTP]
14. Turtles live on this. [GRID]
17. Documented functions of a Web API. [API]

16. Language JSON was originally from.
[JAVASCRIPT]
18. What's in a package? [MODULES]

you are here 4  465

11

widgets, events, and emergent behavior

Getting Interactive

This is a hardcore
programming chapter.
Work hard, get sleep,
drink water, have
patience, stick with
it—you’ll come out
the other side having el!
unlocked the next lev

Pronounce GUI
like “gooey.”

You’ve certainly written some graphical applications,
but you haven’t created a real user interface yet.That
is, you haven’t written anything that lets the user interact with a graphical user
interface (otherwise known as a GUI). To do that you need to adopt a new way of
thinking about how a program executes, one that is more reactive. Wait, did the
user just click on that button? Your code better know how to react and what to do
next. Coding for interfaces is quite different from the typical procedural method
we’ve been using, and it requires a different way of thinking about the problem.
In this chapter you’re going to write your first real GUI, and no, we’re not going
to write a simple to-do list manager or height/weight calculator, we’re going to do
something far more interesting. We’re going to write an artificial life simulator with
emergent behavior. What does that mean? Well, turn the page to find out.
this is a new chapter   467

the game of life

Enter the WONDERFUL WORLD of Amazing

ARTIFICIAL LIFE

Just ADD CODE—that’s ALL! By using our incredibly simple algorithm, with only four rules,
you’ll be viewing real artificial life with emergent behavior in no time. Here’s how it works:

Our artificial life consists of a set of
cells that live on a grid.
At each square of the grid a cell
can be alive or dead.
If a cell is alive, it is colored black.
We compute the next generation of cells
according to four simple rules.

THE RULES

These simple rules are actually known as the Game of Life. To play
the game, start by placing living cells on the grid and then start
computing generations by following the rules. While the rules
may seem quite simple, you’re going to see that the only way to
know how life is going to play out is to play the game. That’s right;
such interesting behavior emerges from these simple rules that
we can’t even tell you if the game will ever stop evolving without
actually running the computation—more on that deep topic in a
bit —but, for now, it sounds like if we’re going to play, we’re going
to have to write our own Game of Life simulator. Let’s do that!

468  Chapter 11

1

BIRTH: a new cell is born if it is
surrounded by exactly three live cells.

2

LIFE: a cell lives on as long as it is
surrounded by two or three live cells.

3

DEATH: a cell dies of loneliness if it
is surrounded by fewer than two live
cells.

4

A cell dies of overcrowding if it is
surrounded by four or more live cells.

was discovered by
The Game of Lifeician John Conway. More
British mathematkipedia.org/wiki/John_
at https://en.wi
Horton_Conway.

widgets, events, and emergent behavior

Ever heard of sea monkeys?
I think we’ve got a real business
opportunity here, even better than
Phrase-O-Matic!

A closer look at the Game of Life
Okay, you’ve seen the four simple rules, so let’s step through the game in a
little more detail so we understand how it works. You know that the Game
of Life happens on a grid, and at each location on the grid we can have a
cell that is either alive or dead. If it’s alive, we color it so we can see it. If it’s
dead, it’s just transparent.
Our Game of Life universe successively computes new generations of cells,
and to do that it applies the four rules to every location in the grid, after
which it updates all the cells at once, based on the outcome. It then does this
over and over and over, computing new generations. Let’s look at how the
rules apply to a few cases.

For all cells we count the
number of neighbors it has.

nter
Let focus on theplcee.
cell in each exam

Note the center cell
here is dead, unlike
the other examples.

3
4
neighbor count = 0 neighbor count = 1neighbor count = 2neighbor count = neighbor count = neighbor count = 3

By the rules,
zero neighbors
means our cell
dies of loneliness.

By the rules, one
neighbor means
our cell dies of
loneliness.

By the rules,
two neighbors
means our cell
lives on.

By the rules,
three neighbors
means our cell
lives on.

By the rules, four
neighbors means
overcrowding and
death. And so
on for five, six,
seven...

By the rules,
a dead cell
surrounded by
three live cells
means birth!

sadf

you are here 4  469

more on generative systems
Seriously? Four rules?
How interesting could
this really be?

Oh, you’d be surprised...
The Game of Life is a generative system. A
generative system has a predetermined, often simple,
set of rules, but generates behavior that you could not
have guessed, or would not have expected. Generative
systems have been used for art, music, as the basis of
philosophical arguments about the universe, and in
fields like machine learning. And there is no better
example of a generative system than Conway’s Game
of Life. From those four simple rules, here are a few
behaviors you might see in the simulator you’re about
to build...

that
Patterns between
oscillatetes
two sta

“Gliders” moving
through the grid

Stable structures that
live forever if they
aren’t disturbed

Chaotic states
evolving into
something else

How will it end?
Or will it ever
end? The only
way to know
is to let it
compute.
470  Chapter 11

widgets, events, and emergent behavior

Before you get started coding, work through a few generations of the Game of Life.
Use generation 1 as the starting point and then compute generations 2 through 6, by
applying the Game of Life rules. Here are the rules again:
• If a dead cell has three living neighbors, it becomes alive in the next generation.
• If a live cell has two or three living neighbors, it lives on in the next generation.

This is your starting point.

• If a live cell has fewer than two living neighbors, it dies in the next generation.
• If a live cell has more than three neighbors, it dies in the next generation.

Don’t forget to compute the
dead cells around the lives ones:
you may have a birth!

Generation 1

Generation 2

Apply the rules of the Game of Life, and
mark the results in generation 2.

Generation 4

Generation 3

plying the
And then continue on ap
d so on.
an
2
rules to generation

Generation 5

Generation 6
you are here 4  471

overview of the simulator

What we’re going to build
Well, as you can see from the graph paper exercise, it’s a little hard to get a feel for how the Game
of Life is a “generative system with emergent properties” without the help of a computer. So, what
we’re going to do now is build a Game of Life simulator. Our simulator will display the grid of
cells, let users interact by clicking in the grid to enter cells, and provide a few buttons we can use to
control the actions of the simulator. For starters we want to be able to start and stop the simulator.
We also might want to clear the grid and start over, or even load it up with a few pre-configured
patterns. So, we’re talking about building a full-fledged user interface.
Now, a great thing to do when creating a user interface is to, literally, sketch it out on a napkin.
We’re not kidding; it’s a great technique. Let’s make a sketch:

This is going to be
the main display
area, displaying e
the Game of Lif
as it evolves.

The Game of Life

You should be able
to click in this
area to turn on
and off cells.

At the bottom
Start/Pause
we’re going
to put three
buttons.
The start button starts the
computation and then morphs into a
pause button. The pause button stops
the computation and morphs back
into a start button.
472  Chapter 11

Choose a Pattern

Clear

This actually isn’t a button but
a set of options. Clicking this
button pops down a list of predefined patterns you can load the
game with. It also first stops the
computation if it is running.

The clear button stops
the computation and
turns all the cells to off
or dead.

widgets, events, and emergent behavior

Do we have the right design?
Interface design and testing is an entire subfield in itself—there are people who
spend their whole careers designing, testing, and perfecting interfaces. That
said, there are simple and effective techniques you can use to improve your own
interfaces, even if you don’t have a six-figure usability budget. One of those is called
paper prototyping. With paper prototyping, we mock up a user interface on paper
(we’ve already done that), and then have candidate users go through a set of use cases,
using the paper mockup as if it were a real interface. Doing so allows you to observe
real users and the mistakes and misunderstandings about your design.

Usability testing entails
testing products on
actual users to see how
well they work (or don’t).
Think of use cases as a set
of actions or scenarios the
typical user would go through.

Usability gurus claim you
can uncover about 85%
of usability problems
with this technique.

It’s time to do some paper protype testing. Photocopy this page, and the next one (or print a
copy from http://wickedlysmart.com/hflearntocode), and cut out the pieces where indicated.
Next, find a few friends, put the prototype down in front of them, and then ask them a few
questions (we’ll give you a little script of questions in two pages).

The Game of Life

Cut this out,
as it’s going to
be your paper
interface.

Start

Choose a Pattern

Clear

Main Interface

you are here 4  473

paper prototype testing

Random

Random Running

Glider Gun

Glider Gun Running

Choose a Pattern

Pause
Pause Button

Random
Glider
Glider Gun

Option Button

474  Chapter 11

Cut out each item separately; so you
should have six pieces of paper (seven
including the item on the previous page)
after you finish.

widgets, events, and emergent behavior

Wd

Here’s a script to take each user (your tester) through; ask your tester to say out loud what they are
thinking as they go through the exercise. Make notes about the mistakes they make along with any
interface misconceptions. Of course, you want to make note of the things that are working well too.
1.

Place the Main Interface in front of your tester.

2.

Give your tester a brief explanation of the Game of Life and what the simulator does.

3.

Tell the tester an easy way to start is to load a random pattern into the grid. Ask them if
they can figure out how to do that.
When they say they are clicking on “Choose a Pattern,” place the Option Button cutout
on top of the “Choose a Pattern” button.
When they say they are clicking on the “Random” option, place the Random cutout over
the grid, and remove the Option Button cutout.
Tell them they have now loaded the grid with a random selection of live cells.

4.

Ask the tester if they can figure out how to start computing new generations.
Wait for the tester to click on Start. If they don’t, give your tester some hints. When Start
is clicked, place the Random Running cutout over the grid; also place the Pause Button
cutout over the start button. Explain to them that they are watching an evolving scene of
living and dying cells.

5.

Ask the tester if they can now clear the grid. If they can’t find the clear button, then
give them hints until they do. When they click on the clear button, remove the Random
Running and the Pause Button cutouts.

6.

Ask the tester to directly enter some live cells. If necessary help the tester until they click
in the grid. Use a pen to drawn in points on the grid.

7.

Ask the tester to generate some new generations of cells. Again, the tester should click
the start button. When they do, place the Random Running cutout over the grid; also
place the Pause Button cutout over the start button. Explain to them that they are
watching an evolving scene of living and dying cells.

8.

Ask the tester to pause the cell generation. They should find the pause button; when they
do, remove the Pause Button cutout.

9.

Ask them to now load the glider gun pattern. When they click “Choose a Pattern,” place
the Option Button cutout on top of the “Choose a Pattern” button. When they click on
“glider gun,” place the Glider Gun cutout on the grid and remove the Option Button
cutout.

10. Ask them to start the simulator again. When they click the start button, place the Glider
Gun Running cutout over the grid; also place the Pause Button cutout over the start
button. Explain they are watching the simulator generate an endless stream of gliders
crawling down the grid.
11. Tell the tester that concludes the test and thank them for participating.

NOTE: If you find major problems in your testing, well,
we’d like to stop and address them, but this is a book in
print, so we unfortunately don’t have that luxury. But
keep those issues in mind as we proceed through the
chapter and think about how you’d do things differently.

It never hurts
to have a few
cookies on hand
as a thank you.

you are here 4  475

how to approach the implementation

How we’re going to build the simulator
Now that we’ve got the interface designed on paper and we’ve also done a bit of paper prototype testing, we can
feel pretty good about moving forward on the implementation. We do need some idea of how we’re going to
tackle implementing all this, though.
To do that we’re going to use a tried-and-true design that is used across the industry for building user interfaces,
and it involves thinking about our code in three conceptual pieces: the underlying data model, the view, and the
controller, otherwise known as MVC.
Now, separating things into a discrete model, view, and controller typically requires that we use object-oriented
techniques—something we’re not even going to talk about until the next chapter, but we are going to follow this
design conceptually. In other words, we’re not going to go crazy following the MVC pattern, but we are going to
take some inspiration from it, while keeping things as simple as we can.
Here’s how we’re going to think about the design:
The View

The view code is
concerned with what is
displayed on the interface.
It does not care about the
specifics of what is being
computed other than
knowing it has to display
data on a grid.

The Controller

The controller takes
any user requests
coming through the
interface (like to start
the game) and maps
them to requests that
need to be made of the
model or the view.

The Model

The model code only cares
about the data for the
grid and cells, and how to
compute new generations.
It doesn’t care or know
about the view.

You’ll often hear this
design referred to as the
MVC design pattern, or as
model-view-controller.

An entire subfield of study
called “software design
patterns” concerns how
these designs emerged over
time and how they work.

So we have three distinct pieces to code: a view that only worries about how to display things; a model
that only cares about computing a grid of cells and doesn’t want to know anything about how it is being
displayed; and a controller, which manages user interaction and relays commands to the view and model
as necessary.
So why all the trouble? Why do we need MVC? Well, history has taught developers that apps built with
user interfaces can all too quickly become a pile of unmanageable spaghetti code (yes, that is a technical
term; Google it). MVC allows us to avoid that by keeping each piece focused on one responsibility
(among other reasons).
At this point you don’t need to fully understand MVC, but at least now you’ve had some exposure to the
idea, and as you’ll see, it’s going to give us a good way to approach building our simulator. So, with that,
let’s get started!
476  Chapter 11

widgets, events, and emergent behavior

Building the data model
Even though we just did all that interface design work, we’re actually going to set it aside
for a bit and work on the data model for the simulator. As we’ve discussed, by the model we
mean that we need a way to represent all the cells on the grid, and then a way to compute
each generation of the game on that grid. Let’s dive right in.

Representing the grid
To represent the simulator grid we’re going to store a grid of integer cell values, with a
value of 0 meaning a dead cell and a value of 1 meaning a live one.
Now you’ve written plenty of code that uses a one-dimensional list of items, but a grid is
2D—it has a width and a height. So how do we create a 2D list? By using a technique of
lists within lists. Here’s how it works: say you want a grid that is three items high and four
items wide. You can do it like this:
my_grid = [[0, 1, 2, 3],
[4, 5, 6, 7],
[8, 9, 10, 11]]

We can model a two-dimensional
grid of values with a list of lists.

This represents a twodimensional grid three
high and four wide.
Python makes this even easier by giving us a syntax for
accessing two-dimensioned lists:

Python 3.6.0 Shell
11
>>>

Remember the indices
are still zero-based!
Get the value of the location at
Study this code and convince
height = 2 and a width = 3.
yourself that it actually creates a
list of lists representing a grid that
So let’s create a list to hold the cells of our grid:
is height-by-width items.
height = 100
These two globals are going to hold
width = 100
the width and height of the grid.
You might have assumed we'd
Remember list multiplication from Chapter have used a Boolean for our grid
values (because they represent
grid_model = [0] * height
4? This creates a list that has height
live/dead well). While that would
number of zeros.
have been a reasonable choice,
for i in range(height):
And now let’s iterate through
representing cells as integers will
grid_model[i] = [0] * width
each item in the grid_model list
reduce our code complexity when
and replace it with a list that has
we
count a cell's live neighbors.
width number of zeros.
value = my_grid[2][3]
print(value)

you are here 4  477

how to compute a generation

Computing a generation of the Game of Life
Now that we have a place to store our cells, we need some code to compute each
generation of the Game of Life as it evolves. To do that we’re going to have to
examine every single cell and to figure out whether it lives, dies, or is born in the
next round. Let’s first just check out how to iterate through each cell in the list:

Let’s create a function
called next_gen to do this.
def next_gen():
global grid_model
for i in range(0, height):
for j in range(0, width):

…and figure out if
grid_model[i][j]
lives, dies, or is born
in the next round.

We’ll iterate through every single
cell in our two-dimensional list…
We’ll compute this
for every single cell.

Remember this is a nested for
loop; for each value of i we
iterate through every value of j.

Computing each cell’s fate
Each cell’s fate is controlled by the Game of Life rules. Let’s take another look:
• If a dead cell has three living neighbors, it becomes alive in the next generation.
• If a live cell has two or three living neighbors, it lives on in the next generation.
• If a live cell has fewer than two living neighbors, it dies in the next generation.
• If a live cell has more than three neighbors, it dies in the next generation.

So for each cell we iterate over we’re going to have to apply these rules, and follow our
logic here. If the cell is dead, then we need to check if it has three neighbors; if it does,
it is born. If the cell is already alive, then it has to have either two or three neighbors to
continue living. In all other cases, the cell is dead in the next round. So even though we
have four rules, it really comes down to those two conditions to determine if the cell is
alive or dead. We’ll apply that logic in a bit. First, though, there’s one crucial piece of
information we need before we can begin computing any of this: the number of live
neighbors each cell has. As you might imagine, to do that we’re going to have to look at all
the adjacent cells around any given cell to know how many neighors it has living.
Let’s see how to do that.

478  Chapter 11

widgets, events, and emergent behavior

Here’s the count_neighbors function; it
takes a grid and a row and column in
the grid and returns the number of live
neighbors for that location.

Ready
Bake

def count_neighbors(grid, row, col):

This is Ready Bake code. All you need to do is
type it in (or grab it out of the book’s source code
folder). While we think you’re fully up to the task
of writing this code (at this point, we know you
have your conditionals down), going through this
code in detail is not the real focus of this chapter.

count = 0
if row-1 >= 0:
count = count + grid[row-1][col]
if (row-1 >= 0) and (col-1 >= 0):

That said, it is great exercise for your brain to
go through this code and understand it. The
difficulty of this code isn’t so much that it has
to check every neighbor, but it also has to take
into account the edge conditions of cells near
the edge of the grid. So at a minimum, take the
time to understand this code. And if you want to
take it further, when you’ve finished the chapter,
come back and write a new count_neighbors
function from scatch.

count = count + grid[row-1][col-1]
if (row-1 >= 0) and (col+1 < width):
count = count + grid[row-1][col+1]
if col-1 >= 0:
count = count + grid[row][col-1]
if col + 1 < width:
count = count + grid[row][col+1]

So here’s our Ready Bake code just for you.
Enjoy.

if row + 1 < height:
count = count + grid[row+1][col]
if (row + 1 < height) and (col-1 >= 0):
count = count + grid[row+1][col-1]
if (row + 1 < height) and (col+1 < width):
count = count + grid[row+1][col+1]
return count

Conceptually, this code, for a
given cell, adds the number of
live cells to the neighbor count.
The code becomes more complex,
though, because we have to check
for edge conditions of the cells
on the edge of the grid.

Here’s how you access each
neighbor around a cell.
grid_model[row][col-1]

grid_model[row+1][col-1]
grid_model[row-1][col-1]

grid_model[row-1][col]

grid_model[row+1][col]

grid_model[row-1][ col+1]
grid_model[row+1][col+1]
grid_model[row][col+1]

you are here 4  479

completing the next_gen code

Code Magnets

Lucky you. While you were reading about Ready Bake code,
we went ahead and wrote the next_gen function and put it
on the refrigerator. But as is always the case in the Head First
books, someone came along and messed it up. Can you help
us get it back together? Careful, there are extra magnets you
don’t need.
Here’s the

rules again.

We put these
up for you
already.

def next_gen():

THE RULES
BIRTH: a new cell is born if it is
surrounded by three live cells.
LIFE: a cell lives on as long as it is
surrounded by two or three live cells.
DEATH: a cell dies of loneliness if
it is surrounded by fewer than two
live cells.
A cell dies of overcrowding if it is
surrounded by four or more live cells.

global grid_model
for i in range(0, height):
for j in range(0, width):

cell = 0

cell = 1

and

else:

cell = 1
cell = 0

480  Chapter 11

or

if count == 3:

if count == 2 or count == 3:
elif grid_model[i][j] == 1:

if grid_model[i][j] == 0:

count < 2:

if count > 4

id_model, i, j)
count = count_neighbors(gr

widgets, events, and emergent behavior

A Test Drive

Before we get too far along, let’s get this code organized and give it a test drive just
to make sure we’re on the same page. Copy the code below into a file called model.py.
Make sure you include the Ready Bake code as well. Although this code won’t do much
(yet), go ahead and give it a test drive and fix any errors.

height = 100
width = 100

Let's pull together the code
from the last few pages.

grid_model = [0] * height
for i in range(height):
grid_model[i] = [0] * width

Here's what we got—
not much yet, but
it's a start.

def next_gen():
global grid_model

Python 3.6.0 Shell

for i in range(0, height):
for j in range(0, width):
cell = 0
print('Checking cell', i, j)
count = count_neighbors(grid_model, i, j)

Don't miss this line
that we put in just
for testing.

if grid_model[i][j] == 0:
if count == 3:
cell = 1
elif grid_model[i][j] == 1:
if count == 2 or count == 3:
cell = 1

def count_neighbors(grid, row, col):

Put the entire function
body here from the Ready
Bake code two pages back.

Ready
Bake

if __name__ == '__main__':
next_gen()

If printing takes a
while on your machine,
you can always choose
the Shell > Interrupt
Execution menu item
to stop the program.

You should see all the cells being checked
if things are running correctly.

Checking cell 99 76
Checking cell 99 77
Checking cell 99 78
Checking cell 99 79
Checking cell 99 80
Checking cell 99 81
Checking cell 99 82
Checking cell 99 83
Checking cell 99 84
Checking cell 99 85
Checking cell 99 86
Checking cell 99 87
Checking cell 99 88
Checking cell 99 89
Checking cell 99 90
Checking cell 99 91
Checking cell 99 92
Checking cell 99 93
Checking cell 99 94
Checking cell 99 95
Checking cell 99 96
Checking cell 99 97
Checking cell 99 98
Checking cell 99 99
>>>

you are here 4  481

finishing the model code

Completing the model code
We’re not quite there yet. Oh, the logic of our code is just fine, but right now we’re just
computing the value of a cell in the next generation; we aren’t actually doing anything with
it. Here’s the problem, though, if we were to store the next-generation cell values in the
current grid (as we’re computing them), we’d throw off all the count_neighbor calculations
because we’d be computing with a mix of current and next-generation values at the same
time. So, to solve this problem we’re going to need two grids, one to hold the current values
and one to hold the next generation. Then, when we’ve completely computed the next
generation, we’ll have to make it the current generation. Here’s how we do that:
grid_model = [0] * height
next_grid_model = [0] * height
for i in range(height):
grid_model[i] = [0] * width
next_grid_model[i] = [0] * width

Add a global
declaration.

def next_gen():
global grid_model, next_grid_model

for i in range(0, height):
for j in range(0, width):
cell = 0
print('Checking cell', i, j)
count = count_neighbors(grid_model, i, j)
if grid_model[i][j] == 0:
if count == 3:
cell = 1
elif grid_model[i][j] == 1:
if count == 2 or count == 3:
cell = 1
next_grid_model[i][j] = cell
temp = grid_model
grid_model = next_grid_model
next_grid_model = temp

Once next_grid_model is completely
computed, we need to make it the
grid_model. To do that we swap the
two so that grid_model now points to
the data in next_grid_model’s grid,
and vice versa.

482  Chapter 11

Rather than show the
entire file, we're only
showing the code with
changes; go ahead and
make these additions to
your code.

Let’s create a
second grid, called
next_grid_model.

After we’ve computed
a cell, we’ll store it in
the correct position in
next_grid_model.

Q:

Why do we have to swap
grid_model and next_grid_model?
Isn’t it enough to assign next_grid_
model to grid_model?

A:

That will work fine until you start
computing another generation, and
then you’ll have grid_model and next_
grid_model assigned to the same list,
which means you’ll be counting cell
neighbors and changing them in the
same list. Not good. So, we go ahead
and swap the two lists so that when
we do compute the next generation,
grid_model will be assigned to the
current generation and next_grid_
model will have the next-generation
values written to it.

widgets, events, and emergent behavior

We’re not really exercising the next_gen code if we’re testing with a grid full of
zeros. Write a function, randomize, which takes a grid, a width, and a height and
places random ones and zeros at each cell location:
import random
def randomize(grid, width, height):

Your code here

A Test Drive

Let’s do another test run. While we have no way to view our model yet—that’s the whole
point of building a user interface—this is another good time to test things out. Just to be
clear, here’s all the code.

import random

We'll be using the
random module.

height = 100
width = 100
def randomize(grid, width, height):
for i in range(0, height):
for j in range(0, width):
grid[i][j] = random.randint(0,1)
grid_model = [0] * height
next_grid_model = [0] * height
for i in range(height):
grid_model[i] = [0] * width
next_grid_model[i] = [0] * width
randomize(grid_model, width, height)

Add your randomize
function at the top…

…and call it once the
grid is created.

Continued on
next page...

you are here 4  483

testing the model

def next_gen():
global grid_model, next_grid_model

If printing takes a while on
your machine, you can always
choose the Shell > Interrupt
Execution menu item to stop
the program.

for i in range(0, height):
for j in range(0, width):
cell = 0
print('Checking cell', i, j)
count = count_neighbors(grid_model, i, j)

if grid_model[i][j] == 0:
if count == 3:
cell = 1
elif grid_model[i][j] == 1:
if count == 2 or count == 3:
cell = 1
next_grid_model[i][j] = cell
print('New value is', next_grid_model[i][j])
temp = grid_model
grid_model = next_grid_model
next_grid_model = temp
Add another
def count_neighbors(grid, row, col):
if __name__ == '__main__':
next_gen()

print statement
just for testing.

Ready
Bake

Here's what we got;
your results will differ
of course because the
cell values are random.
Python 3.6.0 Shell

New value is 1
Checking cell 99 91
New value is 0
Checking cell 99 92
New value is 0
Checking cell 99 93
New value is 0
Checking cell 99 94
New value is 1
Checking cell 99 95
New value is 0
Checking cell 99 96
New value is 0
Checking cell 99 97
New value is 1
Checking cell 99 98
New value is 1
Checking cell 99 99
New value is 1
>>>

Where are we?
Well, that was a nice chunk of code. For the most part, we’ve now
completed the code representing the model of the simulator. And,
while we can’t visualize the grid in any way yet, we’ve built a way
to store and compute each generation of our generative game.
Now we’re going to treat model.py like a module and move on to
code the user interface, which is going to visualize and control the
model.
Let’s get started.

484  Chapter 11

Before we move on, make sure
you remove the two print
statements from the next_gen
function, as you’ll no longer be
needing them.

Make sure
you do this!

widgets, events, and emergent behavior

Building the view
Ready to actually see something on the screen? Us too. This is where we start building
the view, the onscreen representation of our simulator. To build the view we’re going
to use the built-in Python Tkinter module, which allows us to create GUIs using
many of the common components you typically see in user interfaces, like buttons,
text entry boxes, menus, and a canvas you can draw on programmatically. In Tkinter
we call those components widgets. Here are some of the widgets we’re going to use:

The outer window is a
widget that holds all
the other widgets.

Start is a
button widget.

While there doesn't
seem to be a definitive
pronunciation of Tkinter,
most folks call it “Tea
Kay inter” or “Tea Kin
Ter.”

The simulator grid is a
canvas widget, which allows
you to draw geometric
shapes (we’ll be drawing
lots of little squares to
represent live cells).

Clear is a
button widget.
This is known as an option menu
widget; it allows you to choose
from a set of options.

That’s just a small sampling of the widgets available in Tkinter; if interested, you’ll
find a lot more information at https://en.wikipedia.org/wiki/Tkinter.

you are here 4  485

using widgets

Serious Coding

Creating your first widget
You have some idea of what a widget is going to look
like on your screen, but what does a widget look like in
code, and how does it relate to Python? From a coding
perspective, widgets are just objects. Typically we
instantiate a window widget first (which will appear on
your screen as soon as you instantiate it), and then we
instantiate and add other widgets to the window. So let’s
start by creating a new window widget and then we’ll
add a start button to it. One note, in Tkinter, the class
that represents a window is called Tk.

Let’s start by importing the tkinter module, only
we’ll do it in a slightly different way.

There’s another way to import a
module you haven’t seen yet. When
we use the from keyword, like:

from tkinter import *

We no longer have to prepend every
function, variable, and class name with the
module name. So, for instance, we can use
Tk() rather than tkinter.Tk().

Don’t forget to pay attention to case sensitivity;
we have lowercase tkinter, Tk (uppercase T,
lowercase k), and so on.
And then let’s instantiate a new window; remember
we do that with the Tk class.

It’s customary to call
the top-level window
the “root.” So we named
our variable root.

from tkinter import *

Notice that when we
create a widget, we pass
it the root window it’s
going to be part of.

start_button = Button(root, text='Start', width=12)

root = Tk()
root.title('The Game of Life')

And let’s set the title attribute, which
will appear at the top of the window.

Next we instantiate a Button object,
giving it a couple of arguments to control
text on the button and the width of
the
mainloop()
the button (in characters) on the screen.
This line tells the Tkinter module to place the
button in the window wherever it can. This is called a
And finally, just like we did with
layout
manager, and we’ll talk more about it in a bit.
our turtles, we need to turn over
Basically
, if you have a bunch of widgets in a window,
control to Tkinter and allow it to
you
need
a layout manager to help you position them.
monitor the window for clicks and
similar events.
This is a function from the Tkinter module, and again, we don’t have to prepend the
module name because of the way we imported Tkinter.

A Test Drive

start_button.pack()

Place the code above in a file named view.py and give it a test drive.

You should see a window on the screen similar to this one.
However, depending on your operating system, operating system
version, and so on, you may see something slightly different.
Click the button; what happens?

486  Chapter 11

If you don't see the full title, you can
manually resize the window a bit.

widgets, events, and emergent behavior

Adding the rest of the widgets

The Game of Life

Let’s add the rest of the widgets we need for the simulator. We’ve got a start
button, but we still need a canvas, a clear button, and an option button (otherwise
known as an OptionMenu). We’re going to go ahead and add a few other things
too and put it all in a setup function. Feel free to re-enter all this code from
scratch in view.py if it’s easier than editing the previous version.

el module (the Start/Pause
Clear
The view is going to need access to the mod
one we wrote), so let’s go ahead and import it.
You’ll see what this is used for in a bit. Our
cell_size = 5
cells on the screen are going to be bigger
Here’s our setup function. Let’s get
than one pixel, so we need to adjust the
the global variables we need out of
screen size to accommodate this.
the way and get to the widgets.
def setup():
And notice we’re
global root, grid_view, cell_size, start_button, clear_button, choice
using keyword
arguments. The
root = Tk()
tkinter module is
Here’s the Tk top-level window as before.
root.title('The Game of Life')
full of them.

from tkinter import *
import model

grid_view = Canvas(root, width=model.width*cell_size,
height=model.height*cell_size,
borderwidth=0,
highlightthickness=0,
bg='white')

First up we need a canvas to
draw all those cells on. You’ll
note we’re supplying a fair
number of arguments, like the
width, height, border thickness,
and background color.

start_button = Button(root, text='Start', width=12)
clear_button = Button(root, text='Clear', width=12)

Here’s our start button from before.
We need a clear button too.
This works with the option menu;
we'll come back to this shortly.

choice = StringVar(root)
choice.set('Choose a Pattern')
option = OptionMenu(root, choice, 'Choose a Pattern', 'glider', 'glider gun', 'random')
option.config(width=20)
grid_view.pack()
start_button.pack()
option.pack()
clear_button.pack()
if __name__ == '__main__':
setup()
mainloop()

Remember we need a layout
manager to place the widgets
in the window. So let’s call
pack on each widget to make
that happen.

And this is the option menu widget, which
we had in our original design. You’ll see we
instaniate an Option Menu object just like
the other widgets, but this widget has a
few other things we need to discuss as well.
We’ll come back to this.

And let’s not forget
to call setup!
you are here 4  487

testing the layout

A Test Drive

Get all that new code on the previous page in view.py and give it a test drive.

Here’s our output. Remember, yours
may look different depending on your
environment.
For comparison, here’s our design.
The Game of Life

Start/Pause

Do you see this? The layout manager isn’t
doing such a great job. We don’t want these
buttons to stack on top of each other.

Clear

We want our buttons like this. Note, we’re
not going to worry about the border
around the canvas, as you’ll see it looks fine
without the border in the actual interface.

Correcting the layout
The Tkinter layout manager, which packs our widgets into the window as
best it can, isn’t giving us the layout we want. We could spend some time
tweaking the pack layout manager, but, as it turns out, Tkinter has a few
layout managers, and we’re going to use another manager that is better
suited for our layout: the grid layout manager. The name is coincidental, in
that it has nothing to do with the simulator grid. Rather, the grid layout
manager allows us to place our widgets into a grid structure within the
main window. Doing that is pretty straightforward if we know where we
want things placed.

488  Chapter 11

Choose a Pattern

Vertically stacked is
not what we want.

widgets, events, and emergent behavior

Placing the widgets into a grid layout
Let’s take our design and think about it as if it were being placed into a
grid. Here’s what it looks like:

The Game of Life

Think of this as row 0,
which has a widget, the
canvas widget, spanning
three columns.
Do we even need to
mention that when it
comes to computer science,
and coding everything
starts with 0?

Start/Pause

The start
button is in row
1, column 0.

Choose a Pattern

Clear

Think of this as row
1, with three columns,
each holding a widget.

The option menu The clear
widget is in row button is in row
1, column 2.
1, column 1.

Translating the grid layout to code
Now that we know where each widget goes, here’s how we tell the grid
manager to place the widgets into the grid:
grid_view.grid(row=0, columnspan=3, padx=20, pady=20)

Put the grid_view canvas in a grid
at row 0, spanning three columns.
We also added some visual padding
to make it look nicer.

start_button.grid(row=1, column=0, sticky=W, padx=20, pady=20)
option.grid(row=1, column=1, padx=20)

clear_button.grid(row=1, column=2, sticky=E, padx=20, pady=20)

Note the sticky parameter tells the layout manager to keep the buttons, essentially,
stuck to the west (left) and east (right) sides, instead of being centered. This just
helps the buttons stay in a good position for different window sizes.

Put the buttons and the option
menu widget in row 1, in their
respective columns, also with a
little padding.

you are here 4  489

using a grid layout

A Test Drive
grid_view.pack()
start_button.pack()
clear_button.pack()
option.pack()

Open view.py and replace the existing layout manager code
with our new grid-based code. After that give it a test drive and
see if things look better. Also, if you haven’t already, remove
both print statements from your model.py file.

Find this code in view.py, and
replace it with this code.

grid_view.grid(row=0, columnspan=3, padx=20, pady=20)

start_button.grid(row=1, column=0, sticky=W, padx=20, pady=20)
option.grid(row=1, column=1, padx=20)

clear_button.grid(row=1, column=2, sticky=E, padx=20, pady=20)
The Game of Life

Much better! Short of
displaying cells, we’ve
pretty much got the view
looking like we want.

Start/Pause

Q:
A:

Why is it customary to call the top-level window the
“root”? Like the root of a tree?
It’s a computer science term, and you’re not too far off
on the tree analogy. Think of a root system with the big thick
root at the top branching off into smaller roots. In the same
way, the big top window branches off into smaller components.
For instance, an OptionMenu widget has a pop-up menu, and
in that you’ll find a number of options to choose, and each of

490  Chapter 11

Choose a Pattern

Clear

those has a string it displays, and so on, smaller and smaller,
like roots. You’ll find tree and root analogies all over computer
science: the root of a file directory, the root of a complex data
structure, and so on.

Q:
A:

How come our buttons don’t do anything when we
click them?
Because we haven’t told them what to do yet. Hang on!

widgets, events, and emergent behavior

Relax

Moving on to the controller
We’re getting there. We’ve got our model all ready to hold cells and
compute new generations, and we’ve got a view that looks like our
paper prototype. Now our job is to wire all this up and get this interface
actually doing something—in other words, we need to start implementing
the controller aspects of our simulator. To do that we’re going to take your
brain on another mind-bending trip into a slightly different way of thinking
about computation—don’t worry, this is nothing like the detour we took on
recursive functions; it’s much more straightforward, yet different than what
you’re used to.
First, though, let’s quickly connect the view to the model. Doing that is our
first real step toward having a functional simulator. After we’ve connected
the two, we’ll start writing our controller code and start incrementally
implementing the functionality behind each component in our interface.

Adding an update function
To get the view and model connected, we’re going to write a function,
update, that we’ll use over and over. The update function will be
responsible for calling the model’s next_gen function and then using
the view to draw the model’s cells on the screen, or more precisely, on the
canvas we created in the interface.
Let’s write a little pseudocode to understand the update function:

First, compute a new
generation.
CALL next_gen()
Then iterate
through every
FOR variable i in range(0, height):
cell in the model.

DEF update():

FOR variable j in range(0, width):

cell = model.grid_model[i][j]
IF cell == 1:

This is a long chapter
with a lot of code and
concepts. Take breaks,
get sleep, give your brain
time to absorb the material. You
might want to take a break now.

Q:

If I use tkinter import * won't it clash
with my own variable, function, and class
names?

A:

The from/import statement imports
a module such that all its variable, function,
and class names are defined at your top
level, meaning you don’t have to prepend
their module name when you use them. The
advantage is that this saves you having to
always use a module name. The disadvantage
is that your own variable, function, and class
names may accidentally clash with those in
the imported module. So why risk it? Well, if
you think it is unlikely that you’re going to
define names in your own code that are going
to clash with the module (in the case of Tkinter
that would mean clashing with names like
Button, or Tk, or other widget names), then it
can often make your code more readable.

If the cell is alive, then
we’ll color in a pixel at that
location in the view.

DRAW black cell at location i, j on grid_view

The only difference between this pseudocode and the code we’re going to
write is that on the canvas we’ll draw small rectangles rather than individual
pixels. Why? Small pixels are very hard to see, so we’ll draw a 5×5 pixel
square on the canvas instead. You might remember the global variable
cell_size, which controls this. You’ll see how this works in just a bit. Oh,
one more thing: we’re only going to draw cells that are living. So, to get rid of
any cells that have died since the last generation, we’ll need to erase the entire
canvas first before redrawing any cells. Let’s look at the code...

Why do you think we’re only
drawing live cells instead of
live and dead cells?

you are here 4  491

completing the update function

def update():
global grid_view
grid_view.delete(ALL)
model.next_gen()

Delete anything drawn on the canvas using
the canvas object’s delete method.
Compute the next generation
Following the
of cells in the model.
pseudocode...

for i in range(0, model.height):
for j in range(0, model.width):
if model.grid_model[i][j] == 1:
draw_cell(i, j, 'black')

If a cell at i, j is live, we then
draw a small, black rectangle.

This is our function, not Tkinter’s, so we
will still need to write this function.
That implements all the logic of our update function, but we left one thing
unimplemented: the function to draw a rectangle, draw_cell. Let’s do that now:

The draw_cell function
takes a row and column
position and a color.
def draw_cell(row, col, color):
global grid_view, cell_size
if color == 'black':
outline = 'grey'
else:
outline = 'white'

Note, this first bit of code is
a purely aesthetic addition.
This code picks an outline color for rectangles. For
black rectangles it chooses grey outlines, which gives
the rectangles a nice look. You'll see.
Upper-left coordinate
of rectangle

grid_view.create_rectangle(row*cell_size,
col*cell_size,
row*cell_size+cell_size,
col*cell_size+cell_size,
And it draws a
fill=color, outline=outline)
small rectangle to

represent the cell.

Then we have the fill color of the
rectangle as well as the outline color.

Think of the draw_cell function as just a helper function that uses Tkinter’s
canvas method create_rectangle to draw a rectangle. This function
“helps” by handling the arithmetic of figuring out the top-left and bottom-right
coordinates of the rectangle we need given the value of cell_size.
492  Chapter 11

Bottom-right
coordinate of
rectangle

widgets, events, and emergent behavior

A Test Drive

Things should start to get interesting. Open view.py and add
the update and draw_cell functions. Also add a call to
update to test things out. Give it a whirl.

def update():
global grid_view

Place update below your
setup function, just above
the check for __main__.

grid_view.delete(ALL)

model.next_gen()
for i in range(0, model.height):
for j in range(0, model.width):
if model.grid_model[i][j] == 1:
draw_cell(i, j, 'black')
def draw_cell(row, col, color):
global grid_view, cell_size
if color == 'black':
outline = 'grey'
else:
outline = 'white'
grid_view.create_rectangle(row*cell_size,
col*cell_size,
row*cell_size+cell_size,
col*cell_size+cell_size,
fill=color, outline=outline)
if __name__ == '__main__':
setup()
update()
mainloop()

Let’s put in our call to update.

Here's what we got. Now we're
getting somewhere!

you are here 4  493

event-driven programming

Ready for another new style of computation?
If you think about all the code you’ve written, you’re in the driver’s seat in that the
computation is always proceeding according to your direction. At every point in the
computation, you’ve got code that says what to do next. For a lot of code, though, this isn’t
the case. A lot of code follows a more reactive style of computation.
To understand what that means, imagine your start button just sitting there doing nothing.
Suddenly a user comes along and clicks on it, and then what? Well, some piece of code is
going to have to wake up and start doing something. In other words, you need code that
reacts to events that occur in your app. Often those events will be a user clicking on a button,
choosing a menu item, typing into a text box, and so on. But they could be other things
too, like, say, a timer going off, data arriving over a network, and so on. Often we call this
event-based or event-driven programming.

Someone clicked on me, now
what? Oh right, I have this
function I’m supposed to call.

def start_handler():
print("Yup, you clicked on the start button alright.")
print("Thanks for stopping by.")

The code that handles an event, like a button click, has different names in different
languages: event handlers, observers, and callbacks, to name a few. We’ll stick with
event handlers because we’re writing code that handles an event that occurs in our
code. No matter what you call these bits of code, the way you construct them is
similar: we tell the object that generates the event, like our button, the function
we want it to call when the event occurs. One other thing to know: often event
handlers are also passed a special event object when they are called. More on that
in a bit. For now, let’s see if we can get an event handler written and working.

494  Chapter 11

widgets, events, and emergent behavior

Let's add a click handler
Our controller code isn’t going to require writing all that much code, so we’re just
going to put the controller code in with the view.py file (instead of creating a new
controller.py file). Go ahead and open view.py again. Add this line just below the line
that instantiates the start button object:
start_button = Button(root, text='Start', width=12)
start_button.bind('', start_handler)

You can call bind on any
widget to associate an event
with a function to call when
that event occurs.

And we want to call the function
start_handler whenever the
button is clicked.

The event we're interested in is a
click from the left button of a mouse
(which works on Mac or Windows
machines, or just about any machine
with a mouse, for that matter).

Now below your setup function and above the update function, add a new
function, start_handler.

start
This is the function we told theked
on.
clic
button to call when it gets

Handlers get passed an event object that contains
information about the event, like which button was clicked.
Here we don't really need any of that information, but we'll
be making use of it later in the chapter.

def start_handler(event):
print("Yup, you clicked on the start button alright.")
print("Thanks for stopping by.")

A Test Drive

Make those two updates to view.py and take the code for a test drive. Click on the
start button and then check your Python Shell for any output.

We clicked the start
button a few times.
Notice how each time you click the function
start_hander is immediately called.

Python 3.6.0 Shell

Yup, you clicked on the
start button
alright.
Thanks for stopping by.
Yup, you clicked on the
start button
alright.
Thanks for stopping by.

you are here 4  495

more thoughts on event-driven programming

So in this style of
programming, my code is sitting
around waiting to be called as different
actions or events happen in the user
interface?

That’s right. Whether the user clicks on a button,

chooses an option from a menu, clicks on the canvas to
perhaps add a live cell (yes, we’re going to do that), or
some other event occurs, a common style of coding is to
register functions that will be called (that’s where the name
callback came from) when the event occurs. As you get used
to this style of coding, you’ll find it is a natural way to
program.

You might be wondering, though, if none of my code is
executing and I’m just waiting for events, how come the
program doesn’t just end? And is there any code that
is being executed as my code waits? That’s where the
mainloop function fits in. You’ll notice in this code and in
the turtle examples in the book we always call mainloop
as the last thing our code does. When you do that, the
code in mainloop takes over monitoring everything that
is happening in the interface, and when it sees a user
interaction it then calls out to your code. So, there is always
code running, and in this case it is the code in mainloop.

496  Chapter 11

widgets, events, and emergent behavior

Here’s some code you’ve never seen before. What do you think it does?
Hint: this code also calls a handler function when an event occurs.

Yes, this is the same root
assigned to the Tk object.

This is a time in millis
(1 sec = 1,000 millisececononds
ds).

root.after(100,update)
This is the update function
we wrote already.

This is a method in the Tk object.

You might be having a déjà
vu moment after that ISS
code we looked at in the
previous chapter.

How the start/pause button is going to work
The start button is an important button because, when clicked, it needs to tell the simulator to
start computing new generations. If you remember the paper prototype, it also needs to morph
into a pause button. Let’s work up a little state diagram showing how the start button works before
we jump into code. In this diagram we’re going to use new global variable, is_running, which is
True if the start button has been clicked and the simulator is generating new generations. If the
game hasn’t started or is paused, then is_running is False.

If the user clicks on the
start button, then we set
is_running to True and we’re
going to call update(). We’re
also going to change the text
on the button to “Pause.”

You can do this over and
over, toggling the state
between the start and
the pause button.

set is_running to True

s
ck
cli
r
e
us

ca
ll u
pd
at
e(
)

ks
lic
c
er
us
set is_running to False

If the user clicks on
pause button, then wethsee
is_running to False. We’ret
going to change the text also
the button to “Start.” on
you are here 4  497

starting and stopping in the simulator

Implementing the start/pause button

set is_running to True

First we need a global is_running variable. Let’s add that to the
top of the view.py file:

s
ck
cli
er
s
u

from tkinter import *
import model
cell_size = 5
is_running = False

ca
ll u
pd
at
e(
)

s
ck
cli
er
us

set is_running to False

Let’s add this global,
initially set to False.

Now we just need to follow the state diagram. Find the start_handler
function you just wrote, and we’re going to make some changes:
def start_handler(event):
print("Yup, you clicked on the start button alright.")
print("Thanks for stopping by.")
global is_running, start_button
if is_running:
is_running = False
start_button.configure(text='Start')
else:
is_running = True
start_button.configure(text='Pause')
update()

A Test Drive

Let’s get rid of this
old code.
If we’re already running (meaning the set
button is in the paused state), then we on
is_running to False and change the butt
name to “Start.”

Otherwise, we set is_running to True, set
the button text to “Pause,” and call update
to compute a generation.

Make those two updates to view.py and take the code for a test drive. Try clicking
on the start and pause buttons over and over. Do you see generations being
computed?

Make sure the start
button toggles between
Start and Pause.
Another example of a
small bit of code making
a huge difference, but
there’s an even bigger
moment coming.
498  Chapter 11

When you click the start/pause
button repeatedly, you should see
generations being computed!

widgets, events, and emergent behavior

Another kind of event
If you click fast enough, you can actually see generations of cells
being computed by your next_gen function. That’s not too bad,
but your fingers are going to get tired, and these being computers,
you’d probably like to compute them a little faster than you are
capable of clicking. To make that happen, we’re going to use
another kind of event, one not based on what the user does (like
clicking a button), but one based on time.
The Python Tk object provides an interesting method named
after. Let’s take a look at how it works:

You’ll find similiar
functionality in most
programming languages.
There are 1,000
The first argument is a
milliseconds in a second.
You can call the after time in milliseconds.
The second argument is
method on the root
a function to call after
Tk object.
that time has elapsed.

root.after(100, update)
Okay, but what does it do exactly? Let’s drop in on some code
invoking the after method.
Hey, root object, I’ve got
a function I need run after
100 milliseconds.
def update():
global grid_view
grid_view.delete(ALL)
model.next_gen()
for i in range(0, model.height):
for j in range(0, model.width):
if model.grid_model[i][j] == 1:
draw_cell(i, j, 'black')

Got it—in 100 milliseconds
I’ll let the update function know
it is time to run. Don’t you worry,
it will happen.

Your code

The root Tk object’s
after function
you are here 4  499

the after method

Hey, update
function, it’s been 100
milliseconds, so you’re up:
get running.

Perfect, it worked
just as advertised. When 100
milliseconds had elapsed, my
update function ran. Just what I
wanted. Thanks, root!

Trace through this code; can you tell what it outputs? How does it work? Check
your answer at the end of the chapter.
from tkinter import *
root = Tk()
count = 10

Brain Building
Do you think this
qualifies as a recursive
function?

def countdown():
global root, count
if count > 0:
print(count)
count = count - 1
root.after(1000, countdown)
else:
print('Blastoff')
countdown()
mainloop()

500  Chapter 11

widgets, events, and emergent behavior

If only we had a way to
call the update method over
and over so that we’d see new
generations of cells computed
without all that clicking. That
would be dreamy. But I know
it’s just a fantasy...

We have the technology: the after method
The after method is just what we need to get our simulator computing at
regular intervals. In fact, we hope you spent a little time understanding that
last Sharpen exercise because we’re going to use a similar technique here in
our update method:

def update():
global grid_view, root, is_running

We’ll need to add the root
and is_running globals, as
we’ll be using them.

grid_view.delete(ALL)
model.next_gen()
for i in range(0, model.height):
for j in range(0, model.width):
if model.grid_model[i][j] == 1:
draw_cell(i, j, 'black')
if (is_running):
root.after(100,update)

Okay, now, when update is called, if is_running is
True, then this call will schedule another call to
update in 100 milliseconds (1/10 of a second).
you are here 4  501

testing the after method

A Test Drive

There are those times when one line of code makes all the difference. Guess
what? This is one of those times. Get those last changes into view.py and get
ready to see your app transformed into a Game of Life simulator.

(1) Click the start button and
watch the simulator run.
(2) Click the pause button to
temporarily suspend it.
(3) Repeat as often as you like
and restart the simulator to get a
new round of random starting cells.
This is starting
to look like more
than just a random
selection of cells.
Did you see any
oscillating patterns?
Gliders going across the
screen? Areas of chaos
that come and go?

Are we there yet?
Well, we’ve certainly come a long way. In fact, we have all the big
stuff working. We just need to finish up a few interface details. Up
first, we have the clear button to implement—it’s going to clear
the screen and set all the cells to dead. With a clear screen we’ll
want the ability to click and add our own live cells to the canvas.
And finally, we have our option menu, which is going to let us
preload some patterns into the grid. We’re going to tackle these
one at a time, starting with the clear button.
502  Chapter 11

widgets, events, and emergent behavior

How about the clear button? It needs to set is_running to False and then set
each cell’s value to 0. It also needs to set the start button’s text to “Start” again.
And, it should call update to update the display before ending (with every cell in
the model set to 0, the screen should clear).
Use the start button as an example and write the code for the clear button:
start_button.bind('', start_handler)
def start_handler(event):
global is_running, start_button
if is_running:
is_running = False
start_button.configure(text='Start')
else:
is_running = True
start_button.configure(text='Pause')
update()

You’ll need to let the clear
button know about your
handler, just like we did with
the start button.

And here’s the start_handler
again, which should serve as a
good example.
Your code here!
If you’re stumped on zeroing out
the cells in the model, refresh
your memory by looking at how
the update function iterates
over the cells.

When you’ve written the code, double-check it with ours at the end of the chapter and then
get it in your view.py file. Give it a try.

you are here 4  503

manually editing cells

How to enter and edit cells directly
Right now the game works by randomly choosing a bunch of live
cells and then when we click the start button it begins computing
new generations. What would be nice, though, is to be able to click
and draw our own live cells on the grid before we click Start. To
do that, we’re going to use a technique similar to the one we used
to handle button clicks. That is, when the user clicks on the canvas,
we’ll use an event handler (in other words, a function) that takes that
click and translates it into a live cell onscreen and in the model.

...which is going to generate
an event and send along an
event object with the xand y-position of the click.
x: 315
y: 134
event

def grid_handler(event):
global grid_view, cell_size
x = int(event.x / cell_size)
y = int(event.y / cell_size)
if (model.grid_model[x][y] == 1):
model.grid_model[x][y] = 0
draw_cell(x, y, 'white')
else:
model.grid_model[x][y] = 1
draw_cell(x, y, 'black')

The user is going to
click on the canvas...

We’ll begin building this code by binding a left button click to the
canvas just as we did with the start and clear buttons.
grid_view.bind('', grid_handler)

When the user clicks the left button
on the grid_view canvas, call the
grid_handler function.
504  Chapter 11

event_handler

And that event object is
going to get passed to our
click handler function.

widgets, events, and emergent behavior

Writing the grid view handler
Let’s think through how we want the grid_handler function to work. How about this: if the user
clicks on a white (dead) cell in the grid, then we make it a live cell and turn it black. And, if the cell is
black already, we’ll change its state to dead and turn it back to white. Of course, the cell also needs to
change visually in the view, and we need to update its value in the model.
Let’s look at how to do this in code:

Remember the grid handler function
takes an event as a argument.
def grid_handler(event):

global grid_view, cell_size

We can get the x- and y-position of
the click from the event object. You’ll
find there is an attribute for each.

x = int(event.x / cell_size)
y = int(event.y / cell_size)
if (model.grid_model[x][y] == 1):
model.grid_model[x][y] = 0
draw_cell(x, y, 'white')

else:

model.grid_model[x][y] = 1
draw_cell(x, y, 'black')

A Test Drive

Remember our grid is scaled by the cell size, so we need to
find out the true x and y (row and column, if you like) in
the grid model. To do this we divide by the cell size. We
use the int function to make sure the result is an int and
not a floating-point number.
If the current cell in the model is 1,
we set it to 0 and call our draw_cell
function to color in the cell white.
Otherwise, we set the model to 1 and
call our draw_cell function to color in
the cell black.

Add the grid_handler function above to view.py, just below the setup function.
And also remember to put the call to the bind method in your code (see below).
Finally, it’s time to delete that call to randomize in model.py; that way you’ll start with a
clear grid.
Don’t miss this step.
grid_view.grid(row=0, columnspan=3, padx=20, pady=20)
grid_view.bind('', grid_handler)

Place the call to bind just below where
you set the grid for the grid_view.
Now, when you run the simulator you
could see a nice blank screen. Click
in the screen to add live cells before
clicking the start button.

We recommend drawing a
glider pattern and clicking
Start. If you mess up, just
click again to erase the cell.

If clicks aren’t working, make
sure you’re clicking in the
middle of the window.
you are here 4  505

adding pre-built patterns

Now it’s time to add some patterns
We’re going to put one more bit of polish on this app by providing an option
menu to allow users to choose from a list of preexisting patterns. Our design
supplies three patterns, but you should feel free to add your own.
Remember that you’ve already instantiated an option menu in your code, but
we purposely put off discussing that code because the option menu widget
works a little differently than the button and canvas widgets. Let’s look at the
code we have so far to create the option menu:

The Tkinter module provides an object that stores values.
Here we’re creating an object to hold a string and assigning
it to the variable choice. You'll see how this is used in a sec.

We then set the value of the
choice object to ‘Choose a
Pattern,’ which is the initial
choice we want selected in
the widget.

choice = StringVar(root)
choice.set('Choose a Pattern')
option = OptionMenu(root, choice, "Choose a Pattern",
"glider",
"glider gun",
"random")
option.config(width=20)

To make things look better,
we also widened the widget
by giving it a width of 20.

We then instantiate the option menu and
pass it the root window (as is always
the case with widgets), our variable
for holding a string value, and a set of
options that appear in the menu.

As you can see in the interface this all looks great, but it doesn’t do anything.
As with the buttons and the canvas, we need to add an event binding to the
option menu, but the way we do that is a little different than we’re used to.
Here’s how we add a binding for the option menu:
choice = StringVar(root)
choice.set('Choose a Pattern')
option = OptionMenu(root, choice, "Choose a Pattern",
"glider",
"glider gun",
"random",
command=option_handler)
option.config(width=20)

506  Chapter 11

We add one more argument to the
OptionMenu constructor, which is the
command (think “handler”) to invoke
when a user chooses an option.

Add this small
bit of code to
your view.py file
before we forget
to do it!

widgets, events, and emergent behavior

Writing a handler for the OptionMenu
Writing an option menu handler is a bit different as well, because we aren’t
just reacting to a simple button click. Instead the user has chosen an option in
the menu, and we need to figure out what that option is and then act on it.
As you might have figured out, the first item in our menu, “Choose a Pattern,”
is an instruction to the user; it isn’t meant to do anything. For the rest of the
options, let’s write some code to see how we’re going to handle them:

This handler gets an event too, although
we won’t need to make use of it.
def option_handler(event):

We’re going to be using the Tkinter
StringVar object we saw earlier.

global is_running, start_button, choice

Let's stop the simulator
from running and reset
start_button.configure(text='Start') the start button.
The choice variable holds a StringVar
selection = choice.get()
object with the value the user selected in
the option menu. A StringVar provides a
if selection == 'glider': get method to get the value it holds.
is_running = False

model.load_pattern(model.glider_pattern, 10, 10)

elif selection == 'glider gun':

model.load_pattern(model.glider_gun_pattern, 10, 10)

Once we have the user's
selection we can test
to see which option
it was, and either
load a pattern or call
randomize.
Note that we need to write
the load_pattern function
as well as define what the
patterns look like.

elif selection == 'random':

model.randomize(model.grid_model, model.width, model.height)

update()

After we change the
model, we need to
update the display for
the user.

If the user chooses
random, we conveniently
have already written a
randomize function, so
we just need to call it!

Go ahead and place this code
into your view.py file above
your start_handler function.
We can’t run it yet, because
we haven’t written the model’s
load_pattern method or
defined the patterns. We’ll do
that next, and then come back
and give it all a test drive.

you are here 4  507

the stringvar object

Okay, you threw me
with that StringVar object. So,
it’s an object that holds the choice we made
with the option menu? Is it like a special object
that acts like a variable? Given we have variables
already, why do we need it?

It is a little confusing. After all, we’re happily coding
along, using everyday Python variables, and out of nowhere

Tkinter throws in its own variables in the form of an object. Why?
There are a couple reasons we need to use Tkinter variables here.
The first is that the Tk graphics library (which Tkinter is built

on) is actually a cross-platform library (meaning it works with lots of
languages, not just Python). As you gain experience with Tkinter,
you’ll see a few things that don’t feel Python-like because Tkinter
was not designed specifically for Python.

The other reason is that the StringVar class allows you to do a bit
more than just store and retrieve values; using a StringVar you
can track the changes to a variable’s value. For instance, say you’re
building a weather monitor and you want to update your display
anytime the temperature changes. Well, with a StringVar you can
use its trace method to be notified anytime its value changes, like
this:
temperature = StringVar()
temperature.trace("w", my_handler)

Anytime temperature is changed (written to in
Tk terminology), our my_handler gets called.
We won’t be making use of this additional functionality of
StringVar in our code, but it’s handy to know about for future

reference, and is a great example of another form of event-based
programming.

508  Chapter 11

widgets, events, and emergent behavior

How to define your patterns
Let’s define our patterns in the form of a two-dimensional list, like our grid
model. So a pattern for a glider would look like this:
glider_pattern = [[0,
[0,
[0,
[0,
[0,

0,
0,
0,
1,
0,

0,
1,
0,
1,
0,

0,
0,
1,
1,
0,

0],
0],
0],
0],
0]]

Can you see the resemblance?

And the glider gun is a little more complicated:
glider_gun_pattern =

[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0],
[0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]

Can you see the
resemblance here too?

You don’t need to type these in. Just look in your Chapter 11 source code and you’ll find
glider.py and glider_gun.py. Open them and copy these two assignment statements to
the bottom of your model.py file (just below the count_neighbors function). No need to
test yet, other than making sure you don’t have any syntax errors.

you are here 4  509

writing the pattern code

Writing the pattern loader
Now we need to write the code to load a pattern into the grid mode. All the pattern
loader needs to do is take a list and copy its ones and zeros into the grid model. Actually,
before it does that, it should clear out the grid model and write zeros into all its cells.

We’re going
to load our
patterns by
copying them
into the
grid_model.

We’ll also support
an optional x- and
y-offset for the
pattern.

We should also allow the pattern to be offset on the grid, so it can be placed, say, in the
middle of the grid. We’ll do that by taking an x-offset and a y-offset as arguments, and
rather than copying the pattern starting at the top of the grid (location 0, 0), we’ll place
the pattern at the offset location. Here’s the code:

Our function takes a
pattern (like glider or
glider_gun) in the form of
a two-dimensional list.

It also takes two other arguments, an xand y-position to locate the top-left
cell of the pattern on the grid.

def load_pattern(pattern, x_offset=0, y_offset=0):
global grid_model
for i in range(0, height):
for j in range(0, width):
grid_model[i][j] = 0
j = y_offset

And zero out
all the cells.

Alert: there's duplicated code
here (look at the clear_handler);
there's an opportunity for
refactoring this code later (a
homework assignment).

Let's set i and j to the
offsets.
Now we just iterate through each value
in the pattern and assign it to the
corresponding point on the grid_model.

for row in pattern:
i = x_offset
for value in row:
grid_model[i][j] = value
i = i + 1
j = j + 1

510  Chapter 11

You’ll remember of course
parameter defaults from
Chapter 5.

widgets, events, and emergent behavior

A Test Drive

Add the load_pattern function from the previous page to your model.py file,
just below your patterns, and you should be all ready for an exhaustive test of
the Game of Life simulator! Given all the code and code changes in this chapter,
you’ll find the complete source code for the simulator over the next several pages.

model.py
import random
height = 100
width = 100
def randomize(grid, width, height):
for i in range(0, height):
for j in range(0, width):
grid[i][j] = random.randint(0,1)
grid_model = [0] * height
next_grid_model = [0] * height
for i in range(height):
grid_model[i] = [0] * width
next_grid_model[i] = [1] * width
def next_gen():
global grid_model, next_grid_model
for i in range(0, height):
for j in range(0, width):
cell = 0
count = count_neighbors(grid_model, i, j)
if grid_model[i][j] == 0:
if count == 3:
cell = 1
elif grid_model[i][j] == 1:
if count == 2 or count == 3:
cell = 1
next_grid_model[i][j] = cell
temp = grid_model
grid_model = next_grid_model
next_grid_model = temp

you are here 4  511

the complete code and test

def count_neighbors(grid, row, col):
count = 0
if row-1 >= 0:
count = count + grid[row-1][col]
if (row-1 >= 0) and (col-1 >= 0):
count = count + grid[row-1][col-1]
if (row-1 >= 0) and (col+1 < width):
count = count + grid[row-1][col+1]
if col-1 >= 0:
count = count + grid[row][col-1]
if col + 1 < width:
count = count + grid[row][col+1]
if row + 1 < height:
count = count + grid[row+1][col]
if (row + 1 < height) and (col-1 >= 0):
count = count + grid[row+1][col-1]
if (row + 1 < height) and (col+1 < width):
count = count + grid[row+1][col+1]
return count
glider_pattern = [[0,
[0,
[0,
[0,
[0,
glider_gun_pattern =
0, 0, 0, 0, 0, 0, 0,
[0, 0,
0, 0, 0, 0, 0, 0, 0,
[0, 0,
0, 0, 0, 0, 0, 0, 0,
[0, 0,
0, 0, 0, 0, 0, 0, 0,
[0, 0,
0, 0, 0, 0, 0, 0, 0,
[0, 1,
0, 0, 0, 0, 0, 0, 0,
[0, 1,
0, 0, 0, 0, 0, 0, 0,
[0, 0,
0, 0, 0, 0, 0, 0, 0,
[0, 0,
0, 0, 0, 0, 0, 0, 0,
[0, 0,
0, 0, 0, 0, 0, 0, 0,
[0, 0,
0, 0, 0, 0, 0, 0, 0,

512  Chapter 11

0,
0,
0,
1,
0,

0,
1,
0,
1,
0,

0,
0,
1,
1,
0,

0],
0],
0],
0],
0]]

[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0],
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,
0, 0, 0],
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0,
0, 0, 0],
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0,
1, 1, 0],
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0,
1, 1, 0],
1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0,
0, 0, 0],
1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 1, 0, 1, 0, 0,
0, 0, 0],
0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,
0, 0, 0],
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0],
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0],
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0]]

widgets, events, and emergent behavior

def load_pattern(pattern, x_offset=0, y_offset=0):
global grid_model
for i in range(0, height):
for j in range(0, width):
grid_model[i][j] = 0
j = y_offset
for row in pattern:
i = x_offset
for value in row:
grid_model[i][j] = value
i = i + 1
j = j + 1
if __name__ == '__main__':
next_gen()

view.py
from tkinter import *
import model
cell_size = 5
is_running = False

def setup():
global root, grid_view, cell_size, start_button, clear_button, choice
root = Tk()
root.title('The Game of Life')
grid_view = Canvas(root, width=model.width*cell_size,
height=model.height*cell_size,
borderwidth=0,
highlightthickness=0,
bg='white')
start_button = Button(root, text='Start', width=12)
clear_button = Button(root, text='Clear', width=12)
choice = StringVar(root)
choice.set('Choose a Pattern')
option = OptionMenu(root, choice, 'Choose a Pattern', 'glider', 'glider gun', 'random',
command=option_handler)
option.config(width=20)

you are here 4  513

the complete code and test, continued
grid_view.grid(row=0, columnspan=3, padx=20, pady=20)
grid_view.bind('', grid_handler)
start_button.grid(row=1, column=0, sticky=W,padx=20, pady=20)
start_button.bind('', start_handler)
option.grid(row=1, column=1, padx=20)
clear_button.grid(row=1, column=2, sticky=E, padx=20, pady=20)
clear_button.bind('', clear_handler)
def option_handler(event):
global is_running, start_button, choice
is_running = False
start_button.configure(text='Start')
selection = choice.get()
if selection == 'glider':
model.load_pattern(model.glider_pattern, 10, 10)
elif selection == 'glider gun':
model.load_pattern(model.glider_gun_pattern, 10, 10)
elif selection == 'random':
model.randomize(model.grid_model, model.width, model.height)
update()
def start_handler(event):
global is_running, start_button
if is_running:
is_running = False
start_button.configure(text='Start')
else:
is_running = True
start_button.configure(text='Pause')
update()
def clear_handler(event):
global is_running, start_button
is_running = False
for i in range(0, model.height):
for j in range(0, model.width):
model.grid_model[i][j] = 0
start_button.configure(text='Start')
update()
def grid_handler(event):
global grid_view, cell_size
x = int(event.x / cell_size)
y = int(event.y / cell_size)
if (model.grid_model[x][y] == 1):
model.grid_model[x][y] = 0
draw_cell(x, y, 'white')
else:
model.grid_model[x][y] = 1
draw_cell(x, y, 'black')

514  Chapter 11

widgets, events, and emergent behavior

def update():
global grid_view, root, is_running
grid_view.delete(ALL)
model.next_gen()
for i in range(0, model.height):
for j in range(0, model.width):
if model.grid_model[i][j] == 1:
draw_cell(i, j, 'black')
if (is_running):
root.after(100,update)
def draw_cell(row, col, color):
global grid_view, cell_size
if color == 'black':
outline = 'grey'
else:
outline = 'white'
grid_view.create_rectangle(row*cell_size,
col*cell_size,
row*cell_size+cell_size,
col*cell_size+cell_size,
fill=color, outline=outline)
if __name__ == '__main__':
setup()
update()
mainloop()

Nice! Here’s what we got
playing with the everfascinating glider gun.

To replicate this, choose the “glider gun”
option and click the start button.

you are here 4  515

review of widgets, events, and emergent behavior

Nice job! The Game of Life
simulator is spectacular! Let me know
when you want to discuss marketing this.
I’m thinking comic book classified ads.

 Generative code produces
output you could not have
guessed by looking at the code.

 A reactive or event-based style
of computation is often used
when creating user interfaces.

 The Game of Life was
discovered by mathematician
John Conway.

 In this model we supply handlers
in the form of functions that
are called when certain events
occur.

 GUIs are graphical user
interfaces.
 Paper prototyping is a method of
testing a GUI before coding it.
 We designed our app
by separating code into
model, view, and controller
responsibilities.
 Tkinter is a Python module for
building user interfaces.
 Tkinter provides widgets that are
presented in code as objects
and onscreen as common user
interface components.
 The Tk object represents a main
window.
 Tkinter provides several layout
managers, which organize
widgets within a window.
 We used the grid layout
manager to place our widgets.

516  Chapter 11

 We use the bind method on
Tkinter widgets to register a
function to handle an event.
 Most handlers are passed an
event object, which contains
more information about an event.
 Most programming languages
also provide a method of
registering a handler to be called
after a duration of time has
elapsed.
 Tkinter provides the after
method on the Tk object to
schedule code to be executed at
a later time.
 The OptionMenu keeps the
user’s selection in a StringVar
object.
 The Tkinter mainloop monitors
the interface for user interaction.

widgets, events, and emergent behavior

Even Further
Further!

Taking your Game of Life Simulator

You’ve built a great little simulator! And it’s just the beginning because there’s so much
more you can do with it. In lieu of a crossword, here are some ideas to consider.

LEARN MORE
nford.
 Start by learning more: http://web.sta
edu/~cdebs/GameOfLife/
e,”
 Google “Game of Life” along with “Maz
,”
erse
“Rev
or
”
City,
lled
“Wa
ay,”
“Night/D
g
to see some other examples of interestin
alternative rules.
out
 Research Cellular Automata to find
its
and
Life
more about the Game of
mathematical basis: https://en.wikipedia.
org/wiki/Cellular_automaton

CODE MORE
 Tweak the rules. He
re’s another set of
rules:
1. If a cell is alive, it sta
ys alive in the
next generation.
2. If a dead cell has two
live cells
adjacent to it, the dead cel
l changes
state to live.
 Write code to save an
d load your patterns
from a file.
 Implement a torus: ou
r implementation
uses a rectangle that ends
at the borders.
Change your code so tha
t the lefthand
side wraps around to the
right, and the
top to the bottom so the
entire grid is one
continuous surface (it’s no
t that hard with
a little thought).
 Add color: how abou
t basing the color on
how long a cell lives?
 Add “ghosting”: if a cel
l has been alive,
give it a light grey color. Ha
ve it fade over
time.
 Find ways to optimize
the code and make
it super fast.

you are here 4  517

exercise solution

Before you get started coding, work through a few generations of the Game of Life.
Use generation 1 as the starting point and then compute generations 2 through 6, by
applying the Game of Life rules. Here are the rules again:
• If a dead cell has three living neighbors, it becomes alive in the next generation.
• If a live cell has two or three living neighbors, it lives on in the next generation.
• If a live cell has fewer than two living neighbors, it dies in the next generation.
• If a live cell has more than three neighbors, it dies in the next generation.

Generation 1

Generation 2

What do you notice
about these two?

Generation 4
518  Chapter 11

Generation 5

Generation 3

In four generations it repeated
itself and moved one square down
and across. If we keep computing
this, it will move across the grid!

Generation 6

widgets, events, and emergent behavior

Code Magnets Solution

THE RULES

Lucky you. While you were reading about Ready Bake code,
we went ahead and wrote the next_gen function and put it
on the refrigerator. But as is always the case in the Head First
books, someone came along and messed it up. Can you help
us get it back together? Careful, there are extra magnets you
don’t need.

BIRTH: a new cell is born if it is
surrounded by three live cells.
LIFE: a cell lives on as long as it is
surrounded by two or three live cells.
DEATH: a cell dies of loneliness if
it is surrounded by fewer than two
live cells.
A cell dies of overcrowding if it is
surrounded by four or more live cells.

def next_gen():
global grid_model
for i in range(0, height):
for j in range(0, width):
cell = 0

Most cases result in the cell
being dead in the next gen,
so we'll just set it to that.

count = count_neighbors(grid_model, i, j)
if grid_model[i][j] == 0:
if count == 3:

This implements the
BIRTH rule.

cell = 1

elif grid_model[i][j] == 1:
if count == 2 or count == 3:
cell = 1

This implements the
LIFE rule.
We didn’t need all
these.
if count > 4
and
cell = 0
count < 2: else:
or

you are here 4  519

exercise solution

We’re not really exercising this code if we’re testing with a grid of all zeros. Write a
function, randomize, which takes a grid, a width, and a height and places random
ones and zeros at each cell location:
import random
def randomize(grid, width, height):
for i in range(0, height):
for j in range(0, width):
grid[i][j] = random.randint(0,1)

Iterate over the grid,
and assign a random
int to each location.

Trace through this code; can you tell what it outputs? How does it work?
from tkinter import *

Brain Building
Do you think this
qualifies as a recursive
function?

Technically the function
countdown never calls itself;
instead, it is asking the root
Tk object to call it at some
point in the future (1 second
in this case). But it certainly
has a recursive feel!

root = Tk()
count = 10
def countdown():
global root, count

We explicitly call countdown the first time, which
schedules the first “after” call of countdown.
Each time countdown is then called, it keeps
scheduling more invocations of countdown 1 second
later—that is, until count == 0, when the
function ends without scheduling another call.

if count > 0:
print(count)
count = count - 1
root.after(1000, countdown)
else:
print('Blastoff')
countdown()
mainloop()

Python 3.6.0 Shell

10
9
8
7
6
5
4
3
2
1
Blastoff

Because you’re instantiating a Tk widget, you’ll
see a new window pop up when you run this code.
520  Chapter 11

widgets, events, and emergent behavior

How about the clear button? It needs to set is_running to False and then set
each cell’s value to 0. It also needs to set the start button’s text to “Start” again.
And it should call update to update the display before ending (with every cell in
the model set to 0, the screen should clear).
Use the start button as an example and write the code for the clear button:
start_button.bind('', start_handler)
def start_handler(event):
global is_running, start_button
if is_running:
is_running = False
start_button.configure(text='Start')
else:
is_running = True
start_button.configure(text='Pause')
update()

clear_button.bind('', clear_handler)
def clear_handler(event):
global is_running, start_button
is_running = False
for i in range(0, model.height):
for j in range(0, model.width):
model.grid_model[i][j] = 0
start_button.configure(text='Start')
update()

You’ll need to let the clear
button know about your
handler, just like we did with
the start button.

And here’s the start_handler
again, which should serve as a
good example.
Here’s the code to add the
clear_handler.

First we set is_running to False.
Then we zero out the cells in the
model.
Now reset the button text to ‘Start’.

Finally, update the display.

When you’ve written the code, double-check it with ours at the end of the chapter and then
get it in your view.py file. Give it a try.

Clear’s working great! But,
boy, we really need a way to
click and add our own cells.

you are here 4  521

12

object-oriented programming

A Trip to Objectville
We’re
leaving this dusty
old procedural town
behind. We’ll send you
a postcard!

In this book you’ve used functions to abstract your code.
And you’ve approached coding in a procedural manner using simple statements,
conditionals, and for/while loops with functions—none of this is exactly objectoriented. In fact, it’s not object-oriented at all! We have looked at objects and how to use
them in our code, but you haven’t created any objects of your own yet, and you haven’t
really approached designing your code in an object-oriented way. So, the time has come
to leave this boring procedural town behind. In this chapter, you’re going to find out why
using objects is going to make your life so much better—well, better in a programming
sense (we can’t really help you with other areas of your life and your coding skills, all
in one book). Just a warning: once you’ve discovered objects you’ll never want to come
back. Send us a postcard when you get there.

this is a new chapter   523

object-oriented programming

Breaking it down, a different way
Remember back in Chapter 1 when we said there were two
skills that you needed to learn to code? The first was breaking a
problem into a small set of actions, and the second was learning
a programming language so that you can describe those actions
to a computer. At this point, you’ve learned a good deal of both
those skills.
And that’s good, because those skills are the foundation of any
coding you (or anyone) will ever do. However, there is another
perspective on breaking problems down, one that almost every
modern language encourages and one that most professional
programmers favor: it’s known as object-oriented programming, with
which we had our first encounter in Chapter 7.
With object-oriented programming, rather than writing an
algorithm using the techniques we’ve learned for abstraction,
conditional logic, and so on, we instead model a set of
objects and how they interact. In some ways object-oriented
programming is an advanced technique that is full of its own
jargon, techniques, and best practices. But in other ways it’s
often an intuitive way to break problems down, as you’ll soon
see.

Remember classes
from Chapter 7?

524  Chapter 12

Now object-oriented programming (otherwise known as OOP),
is a topic that could (and has) filled many books. In this chapter
we’re going to try to get the gist of OOP down so you’re in a
good position to read and understand code written in objectoriented style. We’re also going to get you to the point where
you can use object-oriented techniques in your own code, which
includes creating your own classes. And, after this chapter,
you’ll also be in a great position to continue your objectoriented studies.

object-oriented programming

What’s the point of object-oriented
programming, anyway?

Toast
kind
eat()

OOP allows you to design your code at a higher level. It allows
you to focus on the big picture.
You’ve heard that before—remember when you learned how to
take a piece of code and abstract it into a function? That freed
you to start thinking of your code as a set of functions that could
be called to solve your problem rather than having to mentally
track the spaghetti code of low-level ifs, elifs, for/ins, and
assignments. OOP takes all that to the next level—with OOP
you model real (or virtual) objects, including their state and
behavior, and we let them interact to solve problems.

Toaster

State
Behavior

temperature
elapsed_time
button
load()
toast_it()

For example, take the difference between having to:
1.

Create a heating coil out of wire,

2.

Hook it to electricity,

3.

Turn the electricity on,

4.

Get a piece of bread,

5.

Hold it 2 cm from the coil,

6.

Keep holding bread until done,

7.

Remove bread,

8.

Turn off electricity.

and just using some objects:
1.

Place toast in toaster.

2.

Push toast_button on toaster.

3.

Remove toast when toaster finished.

Button

Procedural way to think
about toasting bread

Object-oriented way to
think about toasting bread

is_pressed
push()

Remember from
Chapter 7 all objects
are instantiated
from classes and
have attributes
(state) and methods
(behavior).

The first way is procedural, while the second way is objectoriented: you have a set of objects (bread, a toaster, and a toaster
button), so you’re thinking at the level of the problem (my bread’s
in the toaster, so now I just need to push the toast button), not
down in the nitty-gritty of every step that needs to occur (while
the heating coil is at 240 degrees and toast is 2 cm away and the
toasting time hasn’t elapsed).

you are here 4  525

more on object-oriented programming

Say you were implementing a classic
ping-pong–style video arcade game. What
would you choose as objects? What state
and behavior do you think they’d have?

What do you like about
object-oriented languages?
“It’s easier to understand how to use
an object than a random bunch of
functions and data in a module.”
—Joy, 27, software architect

“I like that the data and the
functions that operate on that data
are together in one object.”

Pong!

—Brad, 19, coder

“It helps me code in a more natural
way. My coding feels closer to the
real problem.”
—Chris, 39, project manager

“I can’t believe Chris just said that.
He hasn’t written a line of code in
five years.”
—Daryl, 44, works for Chris

“Other than turtles?”

Paddle object

Player object

Ball object

Paddle object

526  Chapter 12

Player object

—Avary, 7, junior coder

object-oriented programming

Designing your first class
In Chapter 7 you saw how to use classes to instantiate objects, and since that
chapter you’ve seen quite a few objects: the built-in types (string, floats, and
so on), turtles, widgets, and the HTTP requests object, to name a few. But
you’ve never created a class of your own. We’re going to do that now.
Like using pseudocode to plan out your procedural code, it helps to plan out
your objects (or more specifically the classes that create them) before jumping
into code. Let’s start simple and plan out a Dog class:

A class has a name, Dog…
…and some state in the form
of attributes…

Dog
name
age
weight

…and some behavior, a
method called bark.

bark()

Before we write the code for this class, let’s think about how it is going to
be used. Here’s some code that uses our yet-to-be-implemented Dog class:
codie = Dog('Codie', 12, 38)

jackson = Dog('Jackson', 9, 12)
codie.bark()

jackson.bark()

Then we make use of
the bark method to
have each dog make a
sound.*

First we
instantiate
two dogs with
different
attributes.

Remember this is called
calling the constructor.

There’s no need to
type in or run this
code yet.

Here's the output.
Python 3.6.0 Shell

Now you’re going
to objectify me?

Codie says "WOOF WOOF"
Jackson says "woof woof"
>>>

* Did you notice each dog has a
different bark? Hmm, how'd that happen?
Based on this code example, we need to have a constructor in the Dog
class that takes a name, age, and weight and then creates a new dog
object for us. We also have a bark method that differs in its output based
on some aspect of the dog—if we had to guess, we’d say big dogs say
“WOOF WOOF” and smaller dogs say “woof woof.” Sound familiar?
Let’s write some code to do this.

Remember Codie?
you are here 4  527

how to create a class

Writing your first class
We’re going to start with what we’ve been referring to as the constructor;
in other words, we’re going to write the code that initializes our dog
objects. After that we’ll implement the bark method.
So study this code, and then let’s talk about it:

We’re creating a class
named Dog.
class Dog:

In Python the constructor is a function named __init__.
Actually we say it is a method, not a function, because it is
defined within a class.

def __init__(self, name, age, weight):
self.name = name
self.age = age

self.weight = weight

In the body of the constructor we’re
setting up each of the object’s
attributes; name, age, and weight.

Okay, but inquiring minds want to know
why is there a “self” in the parameter
list and why are the attributes
prepended with “self”?

How the constructor works
The key to understanding how the constructor (and other methods) works
is understanding the role of the self parameter. Let’s walk through what
happens when the constructor is called. Pay close attention; if you blink
you’ll miss it.

codie = Dog('Codie', 12, 38)
1

When the constructor is called, which happens
when we use the class name along with parentheses
and some arguments, the first thing Python does is
create a new, empty Dog object.

Here’s our brand new Dog object, only
right now it doesn’t have any attributes.
528  Chapter 12

Dog object

object-oriented programming

2

Next, your arguments are passed along to the __init__
function. Something else happens too: Python takes the
newly created object and passes it as the first argument,
which we named self.

...and the newly created
object is passed as the
first argument.
After step 2 the newly created
Dog object is assigned to the
parameter self.

Next, the class’s
constructor is invoked
with the supplied
arguments, and one
addition...

Dog object

__init__(

, 'Codie', 12, 38)

Dog object

self

def __init__(self, name, age, weight):

3

After step 3, all the
arguments we passed to
the constructor have been
assigned to attributes in
the Dog object.

The body of the constructor is then executed.
We assign each parameter (name, age, and
weight) to an attribute of the same name
in the instance of the Dog object using dot
notation.

Again, our brand new Dog object is
passed as the first argument.

name = ‘Codie’
age = 12
weight = 38

def __init__(self, name, age, weight):
self.name = name
Because self is the new Dog
object,
we’re assigning the
self.age = age
values name, age, and weight
self.weight = weight
to attributes in the Dog

Dog object

self

object.

4

When the constructor completes, Python takes the
Dog object and returns it as a result of calling the
constructor. In this case, when the Dog object is
returned, it is assigned to the variable codie.

ne

o
dt

ig
Ass

name = ‘Codie’
age = 12
weight = 38

Dog object

After step 4, the new Dog
object, complete with the
attributes name, age and
weight, is assigned to the
variable codie.

Ret

urns

codie = Dog('Codie', 12, 38)

name = ‘Codie’
age = 12
weight = 38

Dog object

codie

you are here 4  529

more on self and __init__

Q:

How does __init__ get away without
returning a value, yet one is returned
from calling the constructor?

A:

There are all kinds of things happening
behind the scenes with constructors. When
you call the constructor it first creates a new
object for you, then it takes care of passing
that object as the first argument to __init__,
and finally it handles returning that object for
you. In other words, it’s all built into the way
Python works.

Q:
A:

Is there something special about
the name “self”?

No, anytime a constructor is called
it passes a copy of the new object as the
first argument to the __init__ method. By
convention we name this parameter self. But
we don’t have to. That said, this is such a
widely adopted convention that if you don’t
call the first parameter self you’ll get strange
looks from your fellow developers.

You’ll find many object-oriented
languages have a concept like self with a
different name, like “this” for instance.

A Test Drive

Here's our
new class…
…and a
function for
printing dogs.

Let's create
two instances
of Dog objects
and pass them to
print_dog.

530  Chapter 12

Also avoid using self as the name of a local
or global variable, as it will cause mass
confusion.

Q:

So object attributes are just
variables that hold normal Python
values?

A:

Pretty much. An attribute can be
assigned any valid Python value, just like
a variable. And, before you ask, methods
are just like Python functions, only they are
defined within an object instead of globally.
There actually is one other difference too:
methods also have a self parameter, which
we’ll look at shortly.

We haven’t written the bark method yet, but let’s test what we have. Copy the code
below into dog.py and then give it a test drive.
class Dog:
def __init__(self, name, age, weight):
self.name = name
self.age = age
self.weight = weight

This isn't a very
object-oriented way
to print our dogs, but
we'll see how to improve
it in a few pages.

def print_dog(dog):
print(dog.name + "'s", 'age is', dog.age,
'and weight is', dog.weight)
codie = Dog('Codie', 12, 38)
jackson = Dog('Jackson', 9, 12)
print_dog(codie)
print_dog(jackson)

Notice as long as we pass our
print_dog function a dog,
things work as expected!

Python 3.6.0 Shell
Codie's age is 12 and weight is
38
Jackson's age is 9 and weight
is 12
>>>

object-oriented programming

Writing the bark method
Remember when we call
a method, we often say
we “invoke” it.

Before we actually write the bark method, let’s talk about the difference
between a method and a function. Sure, methods are defined in a class, but
there’s more to it than that. Think about how a method is called: it is always
invoked with respect to an object. Like this:
codie.bark()

Objects

Or:

jackson.bark()
And, as you’ll see, methods typically operate on the attributes of that
specific object. Given that, methods are always passed the object they are
being invoked on as their first argument.

Similar to how __init__ is passed the
object as its first argument.

Let’s write the bark method now and you’ll see how this works:

class Dog:

def __init__(self, name, age, weight):

We defined a method
with the same syntax
as a function.

self.name = name
self.age = age

self.weight = weight

def bark(self):

Notice how
we always use
self to access
the object's
attributes.

bark is passed the object it is being
invoked on as its first argument (it
doesn’t have any other parameters).

if self.weight > 29:

print(self.name, 'says "WOOF WOOF"')

else:

print(self.name, 'says "woof woof"')

Because we’re always passed
the instance the method
was invoked on as the first
argument, we can access its
attributes through the self
parameter.

Here we check the
weight of the dog,
and if it is over
29 we print WOOF
WOOF; otherwise, we
print woof woof.
We also print out
the dog’s name in
each case.

you are here 4  531

understanding methods
name = ‘Codie’

How methods work

age = 12
weight = 38
def bark()

Let’s walk through how the method invocation works to make sure we
understand it.

codie.bark()
1

Let’s see how the bark
method works when it is
invoked on the Codie object.

When we call a method on an object, in this case
the codie object, Python takes that object and
passes it to the method as the first argument,
along with any other arguments you supply (bark
obviously doesn’t have any others).
name = ‘Codie’
age = 12

def bark(self):

Dog object

weight = 38

codie

When the bark method is
invoked on the Codie object,
Python passes that object
as the first argument,
which gets bound to the
parameter self.

def bark()

codie

Dog object

if self.weight > 29:

name = ‘Codie’
age = 12
weight = 38
def bark()

Dog object

self

print(self.name, 'says "WOOF WOOF"')

else:

print(self.name, 'says "woof woof"')

2

Next we evaluate the body of the method. The first
line compares self.weight to 29. In this case
the object assigned to self is Codie’s Dog object,
and the value of self.weight is 38, so this
condition is True and the first clause is executed.

print(self.name, 'says "WOOF WOOF"')

3

4

We execute the print statement, which first prints
the value of self.name. Again, self is set to the
object that the bark method was invoked on, so
that’s Codie’s object, which has the name Codie,
so we print Codie says WOOF WOOF.

With that, the method completes. As we coded it
the bark method doesn’t return a value, although,
like a function, it could have.

532  Chapter 12

self is set to Codie’s
object, so self.weight
is 38 and greater
than 29. So we invoke
the first clause.

The print statement first
prints the name attribute of
the object assigned to self.

Python 3.6.0 Shell
Codie says WOOF WOOF
>>>

object-oriented programming

In Chapter 2 we wrote some code to compute a dog’s age in
human years. Add a method to our Dog’s class to compute its
age in human years. You can call the method human_years; it
takes no arguments and returns the result as an integer.

your dog's name? ")
dog_name = input("What is
r dog's age? ")
dog_age = input("What is you
* 7
human_age = int(dog_age)
print('Your dog',
dog_name,
'is',
human_age,

Here's our code from way
back in Chapter 2. Oh, the
memories...

)
'years old in human years'

Here's our code so far. Add
a method, human_years,
that returns the dog's age
in human years.

class Dog:
def __init__(self, name, age, weight):
self.name = name
self.age = age
self.weight = weight
def bark(self):
if self.weight > 29:
print(self.name, 'says "WOOF WOOF"')
else:
print(self.name, 'says "woof woof"')

Add the new method here.

def print_dog(dog):
print(dog.name + "'s", 'age is', dog.age,
'and weight is', dog.weight)

codie = Dog('Codie', 12, 38)
jackson = Dog('Jackson', 9, 12)
print(codie.name + "'s age in human years is ", codie.human_years())
print(jackson.name + "'s age in human years is ", jackson.human_years())

you are here 4  533

understanding inheritance

Adding some inheritance
Let’s say we have the need for a new kind of dog in
our code: a service dog, those trusty companions
who have been well trained to help those needing
assistance. They’re dogs, but they’re dogs with skills.
So, do we need to start over and define a totally new
class ServiceDog from scratch? That would be a
shame because we’ve put a fair bit of time into our
existing Dog class (work with us here), and it would be
great if we could reuse that work. Well, we can.
You’ll find almost every modern programming
language provides the ability for classes to inherit
attributes and behavior from another class. We call
this capablity inheritance and it is a cornerstone of
object-oriented programming.

This is how we draw a class diagram
showing ServiceDog inheriting from
the Dog class.
Dog
name
age
weight
bark()

Returning to our service dog, we can define the
ServiceDog class so that it inherits the attributes

(a name, age, and weight) and the ability to bark
from our original Dog class. But going further, our
ServiceDog class can add new attributes, like a
handler (the person the dog is assisting, not to be
confused with a event handler in the last chapter) and,
say, a walk method to help their handler with walking.

ServiceDog
handler
walk()

Let’s see how to define the ServiceDog class...

What other dogs can you imagine inheriting from the Dog class? What new attributes might
they have? What new methods?

534  Chapter 12

object-oriented programming

Implementing the ServiceDog class
Let’s implement the ServiceDog class and walk through its syntax and
semantics:

This syntax says we're declaring a new class
ServiceDog that inherits from the Dog class.
The class we’re
Our new
class name inheriting from
class ServiceDog(Dog):
def __init__(self, name,

Remember, syntax is all
about how we can type it
in, and semantics is all about
what it means.

Here’s the ServiceDog
constructor.

The ServiceDog
Notice it has the same parameters as the Dog constructor also expects
class, but also has a new, extra parameter.
to receive an extra
handler argument.
age, weight, handler):

Dog.__init__(self, name, age, weight)
self.handler = handler

This adds a new attribute,
handler, to self.

This line calls the constructor
of the Dog class and passes
it all the arguments it needs,
including self.

def walk(self):
print(self.name,'is helping its handler', self.handler, 'walk')

We also have a new
method, walk.

The walk method uses attributes from the Dog and the ServiceDog classes.

Alright, that’s a mouthful, so let’s see how this code works.
rody = ServiceDog('Rody', 8, 38, 'Joseph')
print("This dog's name is", rody.name)

print("This dog's handler is", rody.handler)
print_dog(rody)

rody.bark()
rody.walk()

…or methods just
We can also call
inherited methods, ServiceDogs can do,
like walk…
like bark.
Uh oh, service
dogs should
not be barking
loudly...

Let's create a ServiceDog object
with a handler named ‘Joseph’.
We can access inherited attributes, like name…
…or attributes from ServiceDog, like handler.

A ServiceDog is a type of Dog, so we can
still call print_dog and have it work fine.

Rody, a service dog

Python 3.6.0 Shell
This dog's name is Rody
This dog's handler is Joseph
Rody's age is 8 and weight is 38
Rody says "WOOF WOOF"
Rody is helping its handler Joseph walk
>>>

you are here 4  535

subclassing another class

A closer look at subclassing
When we create a class that inherits from another class, we say that class is a subclass of the other. We
also commonly use subclass in the form of a verb and say that we’re subclassing another class. In this
case when we defined the ServiceDog class, we were subclassing the Dog class.
But let’s take a closer look at the syntax and semantics we used creating the ServiceDog subclass,
starting with the class statement:

More terminology: often Dog is called
the base class, because it is the base of
all the classes that derive from it (right
now we just have one, ServiceDog, but
we'll make some more).

When you define a new class you can either
define it from scratch, or you can supply a
class in the parentheses that you want to
subclass.

It's also common to call this class
the superclass. So here Dog is the
superclass of ServiceDog.

class ServiceDog(Dog):

Now let’s look at the constructor method, __init___:

We told you OOP had
a lot of terminology!

The declaration of the parameters is nothing out
of the ordinary, just notice that we've added
additional parameters (well, one in this case,
handler, but you could add as many as you want).

def __init__(self, name, age, weight, handler):
Dog.__init__(self, name, age, weight)
self.handler = handler

And the last line is assigning
the handler parameter to the
attribute of the same name.
Note that only ServiceDogs
will have this attribute
because only ServiceDogs will
execute this _init__ method.
Finally we have the new method, walk:
def walk(self):

Typically this should be
you
the very first thing wh
en
or
ct
ru
do in a const
subclassing.

This next line may look a little
different. Here we're calling the
constructor of the Dog class. This
sets up all attributes that are
common to all dogs. If we didn't do
this, then the name, age, and weight
would not be set up in the object
we're instantiating.

In a subclass you can define a method just as you would in
a base class. Notice that this method is available only in
objects instantiated from the ServiceDog, not any old Dog.

print(self.name,'is helping its handler', self.handler, 'walk')
536  Chapter 12

object-oriented programming

A ServiceDog IS-A Dog
When we have an inheritance relationship between two classes we say we have an IS-A
relationship—a ServiceDog IS-A Dog. The concept doesn’t just apply to direct descendants: for
example, below you’ll see that SeeingEyeDog inherits from ServiceDog, which, again, inherits from
Dog, so a SeeingEyeDog IS-A ServiceDog, but a SeeingEyeDog IS-A Dog as well. On the other
hand, while ServiceDog IS-A Dog, it’s not a SeeingEyeDog (because SeeingEyeDog inherits from
ServiceDog and can do things that ServiceDog doesn’t know how to do).

Dog is our
base class.

A ServiceDog
inherits from Dog,
so a ServiceDog
IS-A Dog.

Dog

HotDog

name
age
weight

calories
cook()
eat()

bark()

ServiceDog

A HotDog
IS-NOT-A
Dog.
FrisbeeDog

ShowDog

handler

frisbee

handler

walk()

catch()

trot()

A SeeingEyeDog
IS-A ServiceDog.
SeeingEyeDog

A FrisbeeDog
IS-A Dog
too.

handler

jump()

And even a ShowDog
IS-A Dog
We can have any number
of subclasses of Dog.

listen()

A SeeingEyeDog
is also a Dog.

And we don't have to stop
there, we can subclass our
subclasses, for instance a
SeeingEyeDog here inherits
from ServiceDog.

you are here 4  537

the isinstance function

Testing IS-A in code
If we have an arbitary object, can we tell if it IS-A certain class? For
instance, let’s say someone instantiates this object:

mystery_dog = ServiceDog('Mystery', 5, 13, 'Helen')
They then hand you the object. Is it a Dog? ServiceDog? Something
else? How can we tell?
That’s where the isinstance built-in function comes in. Here’s how
it works:

isinstance takes an
object and a class as arguments.
if isinstance(mystery_dog, ServiceDog):
print("Yup, it's a ServiceDog")
else:
print('That is no ServiceDog')

Let’s try another:

isinstance is True if the object is
the same class, or inherits from
the class—in other words, if the
object IS-A class.

if isinstance(mystery_dog, Dog):
print("Yup, it's a Dog")
Now isinstance is True if
else:
the object IS-A Dog.
print('That is no Dog')
And one more:

Now isinstance is True if the
object IS-A SeeingEyeDog.

if isinstance(mystery_dog, SeeingEyeDog):
print("Yup, it's a SeeingEyeDog")
else:
print('That is no SeeingEyeDog')

538  Chapter 12

In this example isinstance evaluates
to True because mystery_dog is a
ServiceDog.
Python 3.6.0 Shell
Yup, it's a ServiceDog
>>>

In this example isinstance is
still True because mystery_dog
inherits from Dog.
Python 3.6.0 Shell
Yup, it's a Dog
>>>

Not this time: isinstance evaluates to
False because our ServiceDog is not a
SeeingEyeDog.
Python 3.6.0 Shell
That is no SeeingEyeDog
>>>

object-oriented programming

Given the class diagram on the left, fill in the values that isinstance
evaluates to on the right. Remember isinstance always evaluates to
True or False. We did the first one for you.

Your answer,
True or False
simple_cake = Cake()
chocolate_cake = FrostedCake()
bills_birthday_cake = BirthdayCake()

Cake
flavor
bake()
cut()

False

eat()

isinstance(simple_cake, BirthdayCake)
isinstance(simple_cake, FrostedCake)
isinstance(simple_cake, Cake)

FrostedCake
frosting

isinstance(chocolate_cake, Cake)
frost()

isinstance(chocolate_cake, FrostedCake)
isinstance(chocolate_cake, BirthdayCake)

BirthdayCake
name_on_cake
add_name()
add_candles()

isinstance(bills_birthday_cake, FrostedCake)
isinstance(bills_birthday_cake, Cake)
isinstance(bills_birthday_cake, BirthdayCake)

you are here 4  539

how to subclass str

Subclasses aren’t just for your own classes; you can subclass Python’s built-in classes too. How
about we subclass the Python string class (otherwise known as the str class)? We’re going to
create a subclass called PalindromeString, which includes the method is_palindrome.
Study the code below and then give it a try. Can you think of another method you’d like to add to
str? There’s no solution at the end of the chapter; everything is on this page:

We're going to create a new class
PalindromeString, which is a subclass of Python's
built-in str class (otherwise known as string).
class PalindromeString(str):

We're just going to add a new method
to our subclass, so we don't need to
implement a constructor.
def is_palindrome(self):
i = 0
If you don't supply a constructor,
j = len(self) - 1
then when this class is instantiated
while i < j:
the superclass's constructor is used
Here's our new
method. Remember
if self[i] != self[j]: (which is str's constructor).
the iterative version
return False
of is_palindrome?
i = i + 1
Notice the use of self throughout.
j = j - 1
In this case, self is the str object
return True
itself, so we can do anything with it
we'd do with a string.
Now let's put our subclass to the test. Remember a
PalindromeString IS-A str, so we can do anything with it we do
with a string, and it inherits a lot of functionality from str.
word = PalindromeString('radar')

word2 = PalindromeString('rader')

print(word, 'length is', len(word), 'and uppercase is', word.upper())
print(word, word.is_palindrome())

print(word2, 'length is', len(word2), 'and uppercase is', word2.upper())
print(word2, word2.is_palindrome())

Here's the output. Can you think of other ways you
could use your new ability to subclass the str class?
By the way, do you remember in Chapter 9 we were looking
for a way to match Verb, VERB, VeRB, verb, and so on? You
can use the upper (or lower) method to convert it to all
uppercase (or lowercase) before the comparison.
540  Chapter 12

Python 3.6.0 Shell
radar
radar
rader
rader
>>>

length is 5 and uppercase is RADAR
True
length is 5 and uppercase is RADER
False

object-oriented programming

How would you describe yourself?
How would you describe yourself ? Using the __str__ method, of
course! It’s time to replace that print_dog function with something
more object-oriented. Here’s a quick Python convention: if you add
a method to your class named __str__ and in that method return a
description in the form of a string, then when you print out any object
from that class, print will use your description.
class Dog:
def __init__(self, name, age, weight):
self.name = name
self.age = age
self.weight = weight
def bark(self):
if self.weight > 29:
print(self.name, 'says "WOOF WOOF"')
else:
print(self.name, 'says "woof woof"')
def human_years(self):
human_age = self.age * 7
return human_age
def __str__(self):
return "I'm a dog named " + self.name

Q:

So when I have a class like
ServiceDog and I create a ServiceDog
object, behind the scenes are there
actually two objects, one with the Dog
attributes and methods and one with the
ServiceDog attributes and methods?

A:

No, there’s just one object with all
the attributes. For methods, the objects
refer to their definitions in the classes, so
they don’t really live in the objects. But
conceptually, just think of an object you
instantiate as a single ServiceDog object.

Q:
A:

I heard Python has something
called multiple inheritance?
Yes, that’s where you don’t inherit
from just one class, but more than one.
Like a flying car might inherit from both
an automobile and an airplane class. It’s
good you know multiple inheritance exists,
but not all experts agree that it’s a good
idea. Some languages consider multiple
inheritance such a horrible, error-prone
practice they forbid it. Put it on your list of
things to look into as you become more
experienced in OOP, but remember, there
are often better ways to approach your
object-oriented designs.

Add a __str__ method to the Dog
class to craft a custom string for
print to use.
Let's test this out by calling
print on a few objects.
Python 3.6.0 Shell
codie = Dog('Codie', 12, 38)
jackson = Dog('Jackson', 9, 12)
rody = ServiceDog('Rody', 8, 38, 'Joseph')
print(codie)
print(jackson)
Note we're in no way changing
print(rody)

how print works; we're just
changing how any instance of
the Dog class prints.

I'm a dog named Codie
I'm a dog named Jackson
I'm a dog named Rody
>>>

Wow, it even worked for Rody,
who is a ServiceDog.

you are here 4  541

overriding methods

Overriding and extending behavior
A few pages back, did you notice that Rody barked a big “WOOF WOOF”? That behavior
would probably be frowned upon while Rody is working as a service dog. If you look at the Dog
class, that’s how the behavior is implemented for dogs over 29 lbs. So, are service dogs stuck
with that behavior for good? Actually, no, we can always override and extend the behavior of
classes we inherit from. Here’s how:
class Dog:
def __init__(self, name, age, weight):
self.name = name
self.age = age
self.weight = weight
def bark(self):
if self.weight > 29:
print(self.name, 'says "WOOF WOOF"')
else:
print(self.name, 'says "woof woof"')
def human_years(self):
human_age = self.age * 7
return human_age
def __str__(self):
return "I'm a dog named " + self.name

Note that you can add
any attributes you want
to a class in the __init__
method. The __init__
method is your chance
to set up any attributes
needed to model the
internal stage of your
object. Attributes don't
have to mirror the
parameters of the method.

class ServiceDog(Dog):
def __init__(self, name, age, weight, handler):
Dog.__init__(self, name, age, weight)
Let's add a new attribute to the
self.handler = handler
ServiceDog,
a Boolean attribute named
self.is_working = False

is_working that’s initially set to False.

def walk(self):
print(self.name,'is helping its handler', self.handler, 'walk')

Notice in this string we need both double and
quotes.

def bark(self):
single quotes. So we escape the single
if self.is_working:
print(self.name, 'says, "I can\'t bark, I\'m working"')
else:
Dog.bark(self)

Now we're going to redefine the bark method in ServiceDog.
Whenever we call bark on a dog of type ServiceDog, this
method will be executed rather than the one in Dog. We say
we’re overriding the bark method.
542  Chapter 12

If is_working is True, then the dog says that it is working
and it can't bark. Otherwise, we call the bark method in
the Dog class, passing it self. In this latter case, the bark
method will end up doing what it normally does (barking).

object-oriented programming

A Test Drive

Let’s give this code a try. Replace the contents of your file
dog.py with the code on the previous page. Then add the
code below at the bottom to test it out.

Create our dog Rody.
rody = ServiceDog('Rody', 8, 38, 'Joseph')
rody.bark()
rody.is_working = True
rody.bark()

We haven't seen a lot
of this, but remember
to change an object's
attribute value you just
assign it as you would a
variable to another value.
The ServiceDog has now
extended the behavior of
Dog, yet when is_working
is False it still has the
same behavior as Dog.

Have him bark
(remember is_working
is initialized to False).

Now set is_working to
True and try again.

Python 3.6.0 Shell
Rody says "WOOF WOOF"
Rody says "I can't bark, I'm working"
>>>

you are here 4  543

review of object-oriented terms

Welcome to Jargon City
Remember we mentioned OOP has a lot of jargon?
Let’s cover some of that jargon before we get too far
down the road. Welcome to Jargon City.

Classes can have
attributes and
methods.

Dog
name
age
weight

We say that Dog is
the superclass of
ServiceDog. We also
often call Dog a base

bark()

class.

Both Dog and Service
Dog are classes.

We say that
ServiceDog
inherits from Dog.

ServiceDog
handler
walk()
bark()

We can instantiate Dog or ServiceDog
objects by calling their constructors.

We say that ServiceDog
is the subclass of Dog.
We also call it a derived
class, because it is based
on Dog.
We say that ServiceDog
overrides the bark
method in Dog.

True or false? All ServiceDogs are Dogs but not all Dogs are ServiceDogs.

Answer: True.

544  Chapter 12

object-oriented programming

class Car():
def __init__(self):
self.speed = 0
self.running = False

Take a look at the class definitions on the left, which contain a few
overridden methods. Execute the code below (in your head) and write
down the output here. As always, you’ll find the solution at the end of
the chapter.

def start(self):
self.running = True
def drive(self):
if self.running:
print('Car is moving')
else:
print('Start the car first')

Trace through this
code in your head,
writing the output you
expect below.

class Taxi(Car):
def __init__(self):
Car.__init__(self)
self.passenger = None
self.balance = 0.0
def drive(self):
print('Honk honk, out of the way')
Car.drive(self)

car = Car()
taxi = Taxi()
limo = Limo()
car.start()
car.drive()

taxi.start()
taxi.hire('Kim')
taxi.drive()
taxi.pay(5.0)

limo.start()
limo.hire('Jenn')
taxi.drive()
limo.pour_drink()
limo.pay(10.0, 5.0)

Your output here

def hire(self, passenger):
print('Hired by', passenger)
self.passenger = passenger

Careful,
this one
is a little
tricky.

Python 3.6.0 Shell

def pay(self, amount):
print('Paid', amount)
self.balance = self.balance + amount
self.passenger = None
class Limo(Taxi):
def __init__(self):
Taxi.__init__(self)
self.sunroof = 'closed'
def drive(self):
print('Limo driving in luxury')
Car.drive(self)
def pay(self, amount, big_tip):
print('Paid', amount, 'Tip', big_tip)
Taxi.pay(self, amount + big_tip)
def pour_drink(self):
print('Pouring drink')
def open_sunroof(self):
print('Opening sunroof')
self.sunroof = 'open'
def close_sunroof(self):
print('Closing sunroof')
self.sunroof = 'closed'
you are here 4  545

understanding the has-a relationship

Object can HAS-Another object
Object attributes aren’t limited to just simple types like
numbers and strings, they can be assigned to lists and
dictionaries as well. They can even be assigned to other
objects. When an attribute is set to another object, we say
there is a HAS-A relationship. If we have a class named
House and its objects have an attribute set to a Kitchen
object, for example, we say the House HAS-A Kitchen.
When there is a HAS-A relationship between objects, we
say that one object is composed with another object (a
House is composed with a Kitchen). Object-oriented
experts get even more precise as to whether both objects
can stand on their own or not (a kitchen doesn’t make
much sense without a house, but a person, who owns
a house, makes sense without the house) and they have
specific terms for those. We’ll let you explore these
relationships as you gain experience.
But you’re probably wondering, what’s the big deal? So
what? An object can have another object as an attribute—
that’s sort of what I’d expect, so why the big production?
You’ve seen that an object can inherit behavior—for
instance, a ServiceDog gets its human_years behavior,
as well as some of its bark behavior, from the Dog class.
Well, another common way to give an object additional
behavior is to compose it with another object. Think
about that House object—when you use composition, or
compose the House object with a Kitchen object, all of
a sudden you have the new ability to cook.
Let’s look at how to compose objects and then we’ll see
how we can add new behavior (and even delegate some
responsibility) to another object.

How many $5 words can
you find in this chapter?

546  Chapter 12

How would you define a set of
classes to build a object-oriented
house?

object-oriented programming

Your turn to create a new class. How about a frisbee-catching dog? We’ve
already created a Frisbee class for you:
class Frisbee:
def __init__(self, color):
self.color = color

A Frisbee can't do much;
it just has a color and
a __str__ method so it
prints well.

def __str__(self):
return "I'm a " + self.color + ' frisbee'
Your job is to help us finish the FrisbeeDog. It needs to catch a frisbee
and give it back. It should have a __str__ method as well. Oh, and the
dog can’t bark if it has a frisbee in its mouth, so you’d better override the
bark method.
class FrisbeeDog(Dog):
def __init__(self, name, age, weight):
Dog.__init__(self, name, age, weight)
self.frisbee = None
def bark(self):

def catch(self, frisbee):

def give(self):

def __str__(self):

This is a challenging exercise.
Give yourself the time to work
through it. Use the solution at
the end of the chapter to the
extent you need to. Be patient
with yourself.
You’ll want the
FrisbeeDog to bark
like other dogs, unless
it has a frisbee in its
mouth, in which case
say “I can’t bark, I
have a frisbee in my
mouth”.
If catch gets called,
take the frisbee
passed to you and
store it in the
frisbee attribute.
When give is called,
set the attribute to
None and return the
frisbee.
If the dog has a
frisbee, then return
the string “I’m a dog
named 
and I have a frisbee”;
otherwise, return
what all Dogs return.
you are here 4  547

testing the frisbee dog

A Test Drive

Double-checking your answer to the previous exercise with our solution, add your
new code to the dog.py file. Remove the previous test code and add the test code
below to the bottom of your file.

def test_code():
dude = FrisbeeDog('Dude', 5, 20)
blue_frisbee = Frisbee('blue')

Notice how this code
is looking a lot more
object-oriented than
procedural?

print(dude)
dude.bark()
dude.catch(blue_frisbee)
dude.bark()
print(dude)
frisbee = dude.give()
print(frisbee)
print(dude)
test_code()

548  Chapter 12

Print the dog and make it
bark, and then have it catch
the frisbee again.
Then have it bark with the frisbee in its mouth.
Print the dog (now that it
has the frisbee) and then
have it give back the frisbee.

And then print the frisbee
that was given back by the
dog, and print the dog again
(now that it's given up the
frisbee).

Here's the output we got.

Dude, a
frisbeecatching dog.

Create a frisbeeDog
and a Frisbee.

Python 3.6.0 Shell
I'm a dog named Dude
Dude says "woof woof"
Dude caught a blue frisbee
Dude says, "I can't bark, I have a frisbee
in my mouth"
I'm a dog named Dude and I have a frisbee
Dude gives back blue frisbee
I'm a blue frisbee
I'm a dog named Dude
>>>

object-oriented programming

Designing a Doggie Hotel
As you know we’re never one to pass up a
good business opportunity, and we hear dog
boarding is a real growth industry. All we
need to do is keep a few dogs, allow them to
be checked in and checked out, and of course
give them all a chance to bark every once in a
while. Let’s create some code to do that. Let’s
do some planning for the hotel by creating a
class diagram for it.

Every hotel has a name attribute, to
hold a name like “Doggie Hotel” or
“Doggie Ranch” or “Puppy Playground”.
Hotel

A Hotel also has two
attributes that are parallel
lists. The lists hold the names
of the dogs along with a
corresponding dog object.
And we have two methods...

name
names
dogs
check_in()
check_out()

The check_in method
takes a name and a dog,
and checks the dog into
the hotel.

When it comes to writing some code for
the Hotel class, there are a few things to
consider. First, we should only check objects
into the hotel if they are truly dogs. To be a
dog an object should either be an instance of
the Dog class, or a subclass of Dog. And we
can only check dogs out of the hotel if they
were checked in, so we need to check that
condition as well.

Dog
name
age
weight
bark()

The check_out method
takes a name and returns
the dog if the dog is
staying at the hotel.

We can say our Hotel
is composed with a set
of Dog objects, the
ones stored in the dogs
attribute list.

Can you think of a better way to
store dogs in the hotel than using
two parallel lists?

you are here 4  549

implementing the hotel class

Implementing the Doggie Hotel
Now that we’ve created a diagram for the Hotel class, let’s work through
implementing it. We’ll start by implementing the constructor and then work
through implementing the two methods, check_in and check_out.

When we instantiate the
Hotel class, we give it a name,
like “Doggie Hotel".

class Hotel:

def __init__(self, name):
self.name = name

self.kennel_names = []
self.kennel_dogs = []

And we'll use two lists, one to hold the
dog names and another to hold the
corresponding Dog objects.

We've got a method to check dogs into the hotel.
The method takes a Dog object.
First make sure this is
def check_in(self, dog):
a Dog; no Cats or other
objects allowed.
if isinstance(dog, Dog):
self.kennel_names.append(dog.name)
self.kennel_dogs.append(dog)

print(dog.name, 'is checked into', self.name)

else:

When we check in, add the
dog's name and the Dog
object to the respective lists…
…and just a little
output for testing.

If the value passed
to check_in isn't a
Dog, we don't check
it in. Sorry, cats.
to make sure
check
First
the kennel...
in
is
dog
the
...and if it is, get the Dog object
from the list and then delete
the name and object from the
kennel lists.

print('Sorry only Dogs are allowed in', self.name)

def check_out(self, name):

To check a dog out of the hotel,
we just need to supply its name.

for i in range(0, len(self.kennel_names)):
if name == self.kennel_names[i]:
dog = self.kennel_dogs[i]
del self.kennel_names[i]
del self.kennel_dogs[i]

print(dog.name, 'is checked out of', self.name)
return dog

We also return the Dog object at the end of checkout.
After all, you should get your dog back, right?

print('Sorry,', name, 'is not boarding at', self.name)
return None

550  Chapter 12

If the dog isn't in the kennel, we
let the user know this is the wrong
kennel and return None.

object-oriented programming

A Test Drive

Add the Hotel class on the previous page to your dog.py file. Remove the
previous test code and add the test code below to the bottom of your file. Oh, and
don’t forget the Cat class too.

class Cat():
def __init__(self, name):
self.name = name

Check out our
new Cat class!

def meow(self):
print(self.name, 'Says, "Meow"')
def test_code():
Let's create a bunch of Dogs, and dogs of
different types, like ServiceDogs and FrisbeeDogs.
codie = Dog('Codie', 12, 38)
jackson = Dog('Jackson', 9, 12)
sparky = Dog('Sparky', 2, 14)
rody = ServiceDog('Rody', 8, 38, 'Joseph')
dude = FrisbeeDog('Dude', 5, 20)
We’ll try a Cat too—you never know!
kitty = Cat('Kitty')
hotel = Hotel('Doggie Hotel')
hotel.check_in(codie)
hotel.check_in(jackson)
hotel.check_in(rody)
hotel.check_in(dude)
hotel.check_in(kitty)

Create a hotel and get all
these pets checked in.

dog = hotel.check_out(codie.name)
print('Checked out', dog.name, 'who
dog = hotel.check_out(jackson.name)
print('Checked out', dog.name, 'who
dog = hotel.check_out(rody.name)
print('Checked out', dog.name, 'who
dog = hotel.check_out(dude.name)
print('Checked out', dog.name, 'who
dog = hotel.check_out(sparky.name)
test_code()

Now let's check them all out
and make sure they're giving
us back the right dog.
is', dog.age, 'and', dog.weight, 'lbs')
is', dog.age, 'and', dog.weight, 'lbs')
is', dog.age, 'and', dog.weight, 'lbs')
is', dog.age, 'and', dog.weight, 'lbs')

And let's keep the Doggie Hotel on
their toes—we never checked in Sparky.

Output on the
next page.

you are here 4  551

understanding encapsulation

Here's the output!
Fine, be that way.

Everyone got
checked in...
...except the cat.
And the Doggie
Hotel gave us all the
right dogs back.
And they figured
out we didn't drop
off Sparky.

Python 3.6.0 Shell
Codie is checked into Doggie Hotel
Jackson is checked into Doggie Hotel
Rody is checked into Doggie Hotel
Dude is checked into Doggie Hotel
Sorry only Dogs are allowed in Doggie Hotel
Codie is checked out of Doggie Hotel
Checked out Codie who is 12 and 38 lbs
Jackson is checked out of Doggie Hotel
Checked out Jackson who is 9 and 12 lbs
Rody is checked out of Doggie Hotel
Checked out Rody who is 8 and 38 lbs
Dude is checked out of Doggie Hotel
Checked out Dude who is 5 and 20 lbs
Sorry, Sparky is not boarding at Doggie Hotel
>>>

Using lists seems very clunky in
the hotel. After all that work understanding
dictionaries, that would seem a better way to
implement the kennel, right?

Good call. You beat us to it—we were thinking
the same thing. Of course a natural reservation is
that once we start changing lists to dictionaries, those
changes are going to propagate into the rest of the
code, and then next thing you know we’ll be changing
everything.
Ah, but not so! After all, this is object-oriented
programming and one of the benefits is
encapsulation. Think about encapsulation like this:
our objects are keeping internal state and behavior
together, and as long as from the outside everything
works as expected, the method we use to implement
the inside of that object is our business. It shouldn’t
have any effect on the code using our object.
Let’s fix the hotel and you’ll see what we mean.

552  Chapter 12

object-oriented programming

Renovating the Doggie Hotel
Get out the jackhammer; it’s time to rework this
hotel. Let’s just rework check_in and check_out
from scratch without showing each change; in this
case, it’s easy to look at:

When we instantiate the
Hotel class, we give it a name,
like “Doggie Hotel".

class Hotel:
def __init__(self, name):
self.name = name
self.kennel_names = []
self.kennel_dogs = []
self.kennel = {}

And we're now going to use a
dictionary to hold the dogs;
let's call it kennel.

We've still got a method to check dogs into the
First make sure this is
hotel. The method takes a Dog object.
a Dog; no Cats or other
def check_in(self, dog):
objects allowed.
if isinstance(dog, Dog):

When we check in, we add the
dog to the dictionary, using
the dog's name as the key…

self.kennel[dog.name] = dog
…and just a
print(dog.name, 'is checked into', self.name)
output for
else:
print('Sorry only Dogs are allowed in', self.name)

def

To check a dog out of the hotel,
we just need to supply its name.
First check to make sure
check_out(self, name):
the dog is in the kennel...
if name in self.kennel:

little
testing.

...and if it is, get the
Dog object from the
dictionary and then
delete the dog from the
dictionary.

dog = self.kennel[name]
print(dog.name, 'is checked out of', self.name)
del self.kennel[dog.name]
We also return the Dog object at the end of checkout.
return dog
all, you should get your dog back, right?
After
else:
print('Sorry,', name, 'is not boarding at', self.name)
return None

Nice! Much clearer code!

If the dog isn't in the
kennel, we let the user
know this is the wrong
kennel.

Remember from Chapter 8
that dictionaries are more
efficient than lists when we
need to find an item.

you are here 4  553

testing and expanding the hotel

A Test Drive

Replace your check_in and check_out methods in dog.py file. And give it
another test drive.

Notice we didn’t change any other code! And
we totally changed the way our dogs are
stored and accessed. That’s encapsulation
at work; your other code doesn’t need to
know or care how you implemented the dogs,
as long as you’re providing the same object
interface (in other words, the methods you
can call on a Dog).

Python 3.6.0 Shell
Codie is checked into Doggie Hotel
Jackson is checked into Doggie Hotel
Rody is checked into Doggie Hotel
Dude is checked into Doggie Hotel
Sorry only Dogs are allowed in Doggie Hotel
Codie is checked out of Doggie Hotel
Checked out Codie who is 12 and 38 lbs
Jackson is checked out of Doggie Hotel
Checked out Jackson who is 9 and 12 lbs
Rody is checked out of Doggie Hotel
Checked out Rody who is 8 and 38 lbs
Dude is checked out of Doggie Hotel
Checked out Dude who is 5 and 20 lbs
Sorry, Sparky is not boarding at Doggie Hotel
>>>

Adding some hotel activities
Any guess what the favorite dog activity is at a dog hotel? Barking, of course. Let’s add a barktime
method to the hotel and let all the dogs get a chance to bark.
def barktime(self):
for dog_name in self.kennel:
dog = self.kennel[dog_name]
dog.bark()

Go through each dog name in the
kennel and use it as a key to grab
the Dog object.

And let the dog bark.

A Test Drive

Add the barktime method to your dog.py file. Remove the previous test code and
add the test code below to the bottom of your file. Give it a spin.

def test_code():
codie = Dog('Codie', 12, 38)
jackson = Dog('Jackson', 9, 12)
rody = ServiceDog('Rody', 8, 38, 'Joseph')
frisbee = Frisbee('red')
dude = FrisbeeDog('Dude', 5, 20)
dude.catch(frisbee)
hotel = Hotel('Doggie Hotel')
hotel.check_in(codie)
hotel.check_in(jackson)
hotel.check_in(rody)
hotel.check_in(dude)
hotel.barktime()
test_code()

554  Chapter 12

Looks like all the dogs are barking.
Python 3.6.0 Shell
Dude caught a red frisbee
Codie is checked into Doggie Hotel
Jackson is checked into Doggie Hotel
Rody is checked into Doggie Hotel
Dude is checked into Doggie Hotel
Codie says "WOOF WOOF"
Jackson says "woof woof"
Rody says "WOOF WOOF"
Dude says, "I can't bark, I have a frisbee in my mouth"
>>>

object-oriented programming

Relax
I can do anything you can, or Polymorphism
Let’s take a look at the output from our barktime test again:
Codie says "WOOF WOOF"

Jackson says "woof woof"
Rody says "WOOF WOOF"

Codie is a Dog.
Jackson is a Dog too.
Rody is a ServiceDog.

This is a
chapter with a
lot of new concepts. Take
breaks, get sleep, give
your brain time to absorb
the material. You might
want to take a break now.

Dude says, "I can't bark, I have a frisbee in my mouth"

And Dude is a FrisbeeDog.
Technically, all the dogs were different types (with the exception of Codie and Jackson, who
are both Dogs). Yet, we were able to write code that iterated over all of them, treat them
uniformly, and call the bark method.
for dog_name in self.kennel:
dog = self.kennel[dog_name]
dog.bark()

No matter what kind of dog it is,
we can call bark, guaranteed.

What if ServiceDog or FrisbeeDog doesn’t have a bark method? Ah, we know they
are guaranteed to have a bark method because they are subsclasses of Dog (said another
way, they inherit from the Dog class), and the Dog class has a bark method. Now, in
our case, ServiceDogs and FrisbeeDogs override the bark method, but it doesn’t
matter—one way or another, they have a bark method.
Why is this a big deal? Because we can count on this quality in our objects, we can write
code that operates without worrying about how those objects do what they do, even if
down the road we change how those objects work, or we create totally new dog types
we never anticipated (like a ShowDog or a PoliceDog). In fact, all Dog types, current
and future, can participate in the Doggie Hotel’s barktime without any changes to the
Doggie Hotel code.
This quality we’re discussing has a technical name: polymorphism. It’s another one
of those $5 words from the object-oriented crowd. Polymorphism means that you can
have different objects with different underlying implementations (like a FrisbeeDog
and a ServiceDog) that present the same programming interface (that is, they
both have a bark method). This quality plays out in several ways in object-oriented
programming and is a deep topic to explore that is related to inheritance. For now, just
remember it is possible to write code that can be applied to many different objects as
long as they have the set of methods you’re expecting.

The more experienced you
become with coding, the
more you’re going to see
this last bit as a huge
advantage.

you are here 4  555

using polymorphism

It's about time we teach the other dogs to walk
The only dogs in our implementation that know how to walk are service dogs. That’s not right; all
dogs walk. We’ve got the main Dog class and we have the FrisbeeDog class, which inherits from
Dog, and both need to be able to walk. Do we need to add a walk method to both? Well, as we saw
with the __str__ method, given that FrisbeeDog inherits from Dog, if we add a walk method to
Dog, then FrisbeeDogs will inherit that behavior. Given that, let’s do a little reworking of the code
to take advantage of that. We can even improve the ServiceDog in the process. Let’s see how.
class Dog:
def __init__(self, name, age, weight):
self.name = name
self.age = age
self.weight = weight

We’re showing only the
classes that changed.

def bark(self):
if self.weight > 29:
print(self.name, 'says "WOOF WOOF"')
else:
print(self.name, 'says "woof woof"')
def human_years(self):
human_age = self.age * 7
return human_age
def walk(self):
print(self.name, 'is walking')

Let's add a simple
walk method to Dog.

def __str__(self):
return "I'm a dog named " + self.name
class ServiceDog(Dog):
def __init__(self, name, age, weight, handler):
Dog.__init__(self, name, age, weight)
self.handler = handler
self.is_working = False

And for the ServiceDog, if it's
working, we’ll show a special
message. Otherwise, we'll just do
what all other dogs do.

def walk(self):
if self.is_working:
print(self.name,'is helping its handler',
self.handler, 'walk')
else:
Dog.walk(self)
def bark(self):
if self.is_working:
print(self.name, 'says, "I can\'t bark, I\'m working"')
else:
Dog.bark(self)
556  Chapter 12

object-oriented programming

A Test Drive

Add the walk code additions and changes from the previous page to your dog.py file.
Remove the previous test code and add the test code below to the bottom of your file.
Then take this code for a walk.

def test_code():
codie = Dog('Codie', 12, 38)
jackson = Dog('Jackson', 9, 12)
rody = ServiceDog('Rody', 8, 38, 'Joseph')
frisbee = Frisbee('red')
dude = FrisbeeDog('Dude', 5, 20)
dude.catch(frisbee)
codie.walk()
jackson.walk()
rody.walk()
dude.walk()
test_code()

Looks like all the dogs are walking.
Python 3.6.0 Shell
Dude caught a red frisbee
Codie is walking
Jackson is walking
Rody is walking
Dude is walking
>>>

Override the FrisbeeDog’s walk method so that if a dog has a frisbee he says, “I can’t
walk, I’m playing Frisbee!” Otherwise FrisbeeDog acts like a normal Dog. Go ahead and
add the code to your dog.py file. Use the same test code as the last Test Drive.
class FrisbeeDog(Dog):
def __init__(self, name, age, weight):
Dog.__init__(self, name, age, weight)
self.frisbee = None
def bark(self):
if self.frisbee != None:
print(self.name,
'says, "I can\'t bark, I have a frisbee in my mouth"')
else:
Dog.bark(self)
def walk():

.
.
.

Override walk here.

def catch(self, frisbee):
self.frisbee = frisbee
print(self.name, 'caught a', frisbee.color, 'frisbee')

The rest of the FrisbeeDog class
goes here.

you are here 4  557

using inheritance

and responsibility

The power of inheritance
All we had to do is add a new method, walk, to the Dog base class
and magically our dogs (the ServiceDog and the FrisbeeDog)
that didn’t walk before began walking. That’s the power of
inheritance. We can add, change, and extend the behavior of a
whole set of classes by simply changing a class they inherit from.
Now, in general that is a powerful tool in your hands, but as with
any tool, it can be misused if you aren’t careful. With inheritance
we can also cause unintended consequences when we add new
behaviors—what if we added a chase_squirrel method to
the Dog class without thinking about the consequences for our
ServiceDog?
You can also overuse inheritance to extend the functionality of
your classes. There are other ways too, and we’ve mentioned one
of them: composition. By putting classes together—or composing
them—we often come up with object-oriented designs that are
more flexible than using inheritance alone.
Using inheritence properly and not overusing it requires good
object-oriented design and analysis skills (something you’ll learn as
you continue to develop your OOP skills). Now we’re talking about
some subtle aspects of OOP and object-oriented design, but it’s
good to be aware of these concepts in the beginning—too many
coders get far into their coding careers before they become aware
of topics like the power of composing classes.

You know how to draw
inheritance with a class diagram.
Here a Kitchen is inheriting
from a Room class, or, as we
say, a Kitchen IS-A Room.

Room

entrance
width
height
enter()
leave()

House
kitchen
lock()
unlock()
sell()

Kitchen
description
cook()
clean()

Here’s how you draw composition.
Here a House HAS-A Kitchen.
HAS-A means composition.

So, again, these are subtleties you’ll learn as you gain experience
with OOP. For now let’s take our hotel just a little further, and
maybe we’ll use composition in the process.

There is still an issue with our new dictionary-based implementation of the Hotel class. For
example, what if two dogs are checked in with the same name? What happens? How would you fix
this issue?

558  Chapter 12

object-oriented programming

Creating a Doggie Hotel walking service
The Doggie Hotel is looking for extra revenue opportunities, so what about
providing a dog walking service to the dogs staying in the kennel? Sounds like a
winner to us. Let’s implement that:
class Hotel:
def __init__(self, name):
self.name = name
self.kennel = {}
def check_in(self, dog):
if isinstance(dog, Dog):
self.kennel[dog.name] = dog
print(dog.name, 'is checked into', self.name)
else:
print('Sorry only Dogs are allowed in', self.name)
def check_out(self, name):
if name in self.kennel:
dog = self.kennel[name]
print(dog.name, 'is checked out of', self.name)
del self.kennel[dog.name]
return dog
else:
print('Sorry,', name, 'is not boarding at', self.name)
return None
def barktime(self):
for dog_name in self.kennel:
dog = self.kennel[dog_name]
dog.bark()
def walking_service(self):
for dog_name in self.kennel:
dog = self.kennel[dog_name]
dog.walk()

Adding a walking service is pretty
straightforward; in fact, it looks
almost exactly like our hotel
barktime method, except the dogs
walk instead of bark.
Iterate through every dog in the kennel
and call each dog's walk method.

Okay, that was easy enough, but let’s say we don’t have time to walk
a zillion dogs; after all, we’re busy running this hotel! What we’d love
to do is delegate the responsibility to someone else.

you are here 4  559

adding a person object

How are we going to hire a
person to walk our dogs if we
don’t even have people objects?

Serious Coding
In OOP, when an object
asks another object to
perform a task, it’s called
delegation. Delegation
is a way to give an object
extra behavior without
having to inherit it or
implement it directly.

Good point. Let’s fix that. How about we create a
simple Person class, and then create a subclass called
DogWalker who knows how to walk dogs?

Otherwise known
as “subclassing”

class Person:
def __init__(self, name):
self.name = name

Just a simple Person class with a name.
We could also add more later.

def __str__(self):
return "I'm a person and my name is " + self.name
class DogWalker(Person):
def __init__(self, name):
Person.__init__(self, name)
def walk_the_dogs(self, dogs):
for dog_name in dogs:
dogs[dog_name].walk()

A DogWalker is just a Person, but it
has a walk_the_dogs method.
This method just iterates through the
dogs and calls each dog’s walk method.

Great, now let’s rework the Hotel class so that we can hire a
DogWalker and then delegate the dog walking through the
walking_service method:

These methods go in the Hotel class.
In our hire_walker method we make
and
as an

sure the object is a DogWalker,
def hire_walker(self, walker):
if so hire them by adding them
if isinstance(walker, DogWalker):
attribute.
self.walker = walker
else:
print('Sorry,', walker.name, ' is not a Dog Walker')
def walking_service(self):
if self.walker != None:
self.walker.walk_the_dogs(self.kennel)

560  Chapter 12

Now in walking_service, if there is
a walker attribute, then we ask the
walker to walk the dogs.

object-oriented programming

If you look at the
DogWalker class, it
knows that the Dogs are stored in
a dictionary. If we change the code
back to lists, then DogWalker will break.
With encapsulation I thought we weren’t
supposed to know how things in an object
were implemented?

Good question. It is true that the walker knows

the dogs are being stored in a dictionary, or rather a set of
nested dictionaries. That said, a dictionary is a common
data structure and provides a perfectly reasonable way of
passing the walker a set of dogs. Hypothetically, the hotel
could use a complex internal data structure to store the
dogs, and when it comes time to pass them to the walker,
it could package them up in a nice dictionary.

You are right, though: if we do change our existing
internal implementation of the hotel, we will have
to keep in mind that walkers are expecting to get a
dictionary of dogs. So we haven’t fully encapsulated this
aspect of the hotel.
All that said, if you really wanted to encapsulate this
aspect of the hotel, and further separate the hotel and
the walker (so that the walker doesn’t have or rely on
knowledge of the hotel’s implementation), you could
improve the implementation with the iterable pattern
that we mentioned much earlier in the book. This pattern
allows you to iterate over sequences of values without
knowing anything about how they are implemented.
Doing so is a bit beyond the scope of this book, but,
again, it’s a great question and something to continue to
exploring in your own designs.

you are here 4  561

testing the dog walker

A Test Drive

At this point you want to add the Person and DogWalker classes from two
pages back, as well as add the hire_walker and walking_service
methods to the Hotel class. Then use the test code below (replace your existing
test code):
def test_code():
codie = Dog('Codie', 12, 38)
jackson = Dog('Jackson', 9, 12)
sparky = Dog('Sparky', 2, 14)
rody = ServiceDog('Rody', 8, 38, 'Joseph')
rody.is_working = True
Rody wouldn't
dude = FrisbeeDog('Dude', 5, 20)
hotel = Hotel('Doggie Hotel')
hotel.check_in(codie)
hotel.check_in(jackson)
hotel.check_in(rody)
hotel.check_in(dude)
joe = DogWalker('joe')
hotel.hire_walker(joe)
hotel.walking_service()
test_code()

be
working at the hotel,
but let's try it out
anyway.

Let's create and hire
our walker…

…and delegate the
responsibility to Joe.
Python 3.6.0 Shell

Looks like delegating
the dog walking is off
to a flying start!

562  Chapter 12

Codie is checked into Doggie Hotel
Jackson is checked into Doggie Hotel
Rody is checked into Doggie Hotel
Dude is checked into Doggie Hotel
Codie is walking
Jackson is walking
Rody is helping its handler Joseph walk
Dude is walking
>>>

object-oriented programming

Meanwhile, back at the turtle races…
Remember back in Chapter 7 when something went awry with the green
turtle? If you’ll recall the green turtle was mysteriously winning by large
margins. The police are still stumped. With your new OOP knowledge, can
you take another look and see what is going on?
import turtle
import random

turtles = []

class SuperTurtle(turtle.Turtle):
def forward(self, distance):
cheat_distance = distance + 5
turtle.Turtle.forward(self, cheat_distance)
def setup():
global turtles
startline = -620
screen = turtle.Screen()
screen.setup(1290,720)
screen.bgpic('pavement.gif')

turtle_ycor = [-40, -20, 0, 20, 40]
turtle_color = ['blue', 'red', 'purple', 'brown', 'green']
for i in range(0, len(turtle_ycor)):
if i == 4:
new_turtle = SuperTurtle()
else:
new_turtle = turtle.Turtle()
new_turtle.shape('turtle')
new_turtle.penup()
new_turtle.setpos(startline, turtle_ycor[i])
new_turtle.color(turtle_color[i])
new_turtle.pendown()
turtles.append(new_turtle)

Study the code again
carefully; it definitely looks
like someone has hacked
the program and added
some new code. What does
the new code do? What
object-oriented concepts
does the hack depend on?

You’ll find the
solution in two pages!

def race():
global turtles
winner = False
finishline = 560

while not winner:
for current_turtle in turtles:
move = random.randint(0,2)
current_turtle.forward(move)

setup()
race()

xcor = current_turtle.xcor()
if (xcor >= finishline):
winner = True
winner_color = current_turtle.color()
print('The winner is', winner_color[0])

turtle.mainloop()

you are here 4  563

a guide for using objects

Welcome to Objectville
A guide for better living through objects.

handy guide with tips &
Welcome, new Objectville resident! Please accept our
your stay.
tricks for better living in Objectville. We hope you enjoy
objects (and classes)
)) Get yourself acquainted. Take the time to explore thecode to get a feel for how it’s
around you. Look at other object-oriented Python
been constructed and how objects are being used.
es, which you can do just
)) Don’t be shy about extending Python’s built-in class
as with your own classes.
lle is a life-long pursuit.
))Know up front that learning your way around Objectvi
the experience. The only
)) Continue learning. You know the basics; now you neederately, otherwise known
way to gain experience is to study and to practice delib
as “build stuff!”

ritence and making
))Spend more time on the object-oriented basics, like inhe
use of polymorphism in your designs.
way up to complex object
))Keep your objects simple and focused, and work your n skyscrapers.
designs; practice building small houses before you desig

using delegation in your
))Learn more about building with composition and
code. You’ll find this may make your design more flexible.
learning.
)) Don’t stop. You’ve got the momentum going, so keep

564  Chapter 12

object-oriented programming

The turtle races case SOLVED
Nice job!
You got him!

Did you figure out what went awry? It looks like our sneaky hacker has used his
subclassing skills to create a subclass of Turtle called SuperTurtle. Then in
SuperTurtle he overrode the forward method so that 5 units get added onto the
distance parameter before the call is made to the base class’s (Turtle, in this case)
forward method. The hacker obviously has a good knowledge of polymorphism
because he knew that the game’s race method would call forward on any kind of
object, as long as it IS-A Turtle. See, a little OOP knowledge CAN be a dangerous
thing!
import turtle
import random

turtles = []

A subclass of Turtle is defined!

class SuperTurtle(turtle.Turtle):
def forward(self, distance):
cheat_distance = distance + 5
turtle.Turtle.forward(self, cheat_distance)
def setup():
global turtles
startline = -620
screen = turtle.Screen()
screen.setup(1290,720)
screen.bgpic('pavement.gif')

turtle_ycor = [-40, -20, 0, 20, 40]
turtle_color = ['blue', 'red', 'purple', 'brown', 'green']

for i in range(0, len(turtle_ycor)):
A SuperTurtle object
if i == 4:
is being instantiated
new_turtle = SuperTurtle()
else:
every
time for turtle
new_turtle = turtle.Turtle()
new_turtle.shape('turtle')
#4, the green turtle.
new_turtle.penup()
new_turtle.setpos(startline, turtle_ycor[i])
new_turtle.color(turtle_color[i])
new_turtle.pendown()
turtles.append(new_turtle)

The hacker has created
a subclass of Turtle that
overrides the forward
method and adds 5 more
units anytime forward is
called.

def race():
global turtles
winner = False
finishline = 560

while not winner:
for current_turtle in turtles:
move = random.randint(0,2)
current_turtle.forward(move)

setup()
race()

Polymorphism at
work: this code calls
forward on any
object that IS-A
Turtle, even if it is a
SuperTurtle.

xcor = current_turtle.xcor()
if (xcor >= finishline):
winner = True
winner_color = current_turtle.color()
print('The winner is', winner_color[0])

turtle.mainloop()

you are here 4  565

congratulations
Got through the
entire book? Including the
OOP part? Well, I think
we may have found our
candidate.

Have you considered a career in coding?
If you’re reading this page, assuming you didn’t just skip to the end,
congratulations! And you should really ask yourself how far you could take this!
In case you haven’t noticed, you’ve covered a huge distance in this book. The
exciting and daunting thing is that this book just scratches the surface of software
development. If nothing else, we hope you’ll consider our suggestion, as you
clearly have the right stuff to take this even further.
Here are a few suggestions for what’s next:

Start on either path.
Jump in head first to
OOP and an industrystandard language,
Java.

Follow it with
learning high-level
object design with
design patterns and
principles.
566  Chapter 12

Learn the languages
that provide the
structure and style
of the web…

Learn browser-based web
development by reading
these books in parallel.
…and the
language that
provides the
behavior

Follow it up
with even more
sophisticated
browser coding.

object-oriented programming

 Object-oriented programming (OOP for
short) focuses on solving problems by
modeling real (or virtual) objects.
 Objects have state and behavior.
 Objects are created from classes, which
act as blueprints.
 When we create a new object, we say we
instantiate it.
 A constructor is a method that initializes a
class.
 In Python the constructor is named
__init__.
 The constructor is passed the object being
created as the first argument.
 By convention we call the first parameter of
a constructor self.
 Attributes are like variables within an object
and can be assigned to any Python value.
 Methods are like functions, except they are
passed the self argument.
 We can inherit the attributes and methods
of another class by subclassing.
 The class we’re subclassing is often called
the superclass or supertype.

 We say we have an IS-A relationship when
we subclass.
 We can test for IS-A relationships with the
isinstance function in Python.
 The isinstance method returns True if an
object is an instance of a class (or any of
its superclasses).
 Override the str method to return the string
you’d like displayed when your object is
printed by print.
 When we write code that works for objects
that expose the same interface we’re
taking advantage of polymorphism.
 We think of an object’s interface as the
methods we can call on an object.
 We compose objects by assigning another
object to the attribute of an object.
 Composition is a common technique to
extend a class’s behavior.
 When we rely on another class to do work,
we’re delegating work to that class.
 Multiple inheritance occurs when you
inherit behavior and state from more than
one class.

 We can override the methods of a
superclass by defining methods of the
same name in the subclass.

you are here 4  567

coding crossword

Coding Crossword
This is it, the final crossword, and it’s object
oriented. Well, at least all the words are.
1
2
3
4

5

6
7
8

9

10

11

12

13
14

15

16

17

18

568  Chapter 12

Across

Down

4. Creates an object and initializes it.
6. Built-in function to check IS-A.
10. Dog that helps humans.
12. Code that works over objects with the same
interface.
14. Convention for the first parameter of a
method.
17. Holds state of object.
18. Has a lot of jargon.

1. Not allowed in the hotel.
2. Type of inheritence that inherits from two
or more classes.
3. Class from which others are subclassed or
derived.
5. IS-A relationship.
7. Hiding implementation details.
8. Redefining the behavior of a method in a
subclass.
9. HAS-A relationship.
11. Asking another object to do the work.

object-oriented programming

In Chapter 2 we wrote some code to compute a dog’s age in
human years. Add a method to our Dog’s class to compute its
age in human years. You can call the method human_years; it
takes no arguments and returns the result as an integer.

class Dog:
def __init__(self, name, age, weight):
self.name = name
self.age = age
self.weight = weight

Here's our code so far. Add
a method, human_years,
that returns the dog's age
in human years.

def print_dog(dog):
print(dog.name + "'s", 'age is', dog.age,
'and weight is', dog.weight)
def bark(self):
if self.weight > 29:
print(self.name, 'says "WOOF WOOF"')
else:
print(self.name, 'says "woof woof"')
def human_years(self):
years = self.age * 7
return years

To compute the dog's age
in human years we use the
dog's age attribute, multiply
it by 7, and return it.

codie = Dog('Codie', 12, 38)
jackson = Dog('Jackson', 9, 12)
print(codie.name + "'s age in human years is ", codie.human_years())
print(jackson.name + "'s age in human years is ", jackson.human_years())

Python 3.6.0 Shell
Here's the output.

Codie's age in human years is 84
Jackson's age in human years is 63
>>>

you are here 4  569

exercise solution

Given the class diagram on the left, fill in the values that isinstance
evaluates to on the right. Remember isinstance always evaluates to
True or False.

Your answer,
True or False
simple_cake = Cake()
chocolate_cake = FrostedCake()
bills_birthday_cake = BirthdayCake()

Cake
flavor
bake()
cut()

False

eat()

False
True

isinstance(simple_cake, BirthdayCake)
isinstance(simple_cake, FrostedCake)
isinstance(simple_cake, Cake)

FrostedCake
frosting

True
True

frost()

False

BirthdayCake
name_on_cake
add_name()
add_candles()

570  Chapter 12

True
True
True

isinstance(chocolate_cake, Cake)
isinstance(chocolate_cake, FrostedCake)
isinstance(chocolate_cake, BirthdayCake)

isinstance(bills_birthday_cake, FrostedCake)
isinstance(bills_birthday_cake, Cake)
isinstance(bills_birthday_cake, BirthdayCake)

object-oriented programming

class Car():
def __init__(self):
self.speed = 0
self.running = False

Take a look at the class definitions on the left, which contain a few
overridden methods. Execute the code below (in your head) and write
the output here.

def start(self):
self.running = True

Trace through this
code in your head,
writing the output you
expect below.

def drive(self):
if self.running:
print('Car is moving')
else:
print('Start the car first')

car.start()
car.drive()

taxi.start()
taxi.hire('Kim')
taxi.drive()
taxi.pay(5.0)

limo.start()
limo.hire('Jenn') Careful,
taxi.drive()
this one
limo.pour_drink() is a little
limo.pay(10.0, 5.0)

class Taxi(Car):
def __init__(self):
Car.__init__(self)
self.passenger = None
self.balance = 0.0
def drive(self):
print('Honk honk, out of the way')
Car.drive(self)

car = Car()
taxi = Taxi()
limo = Limo()

Your output here

def hire(self, passenger):
print('Hired by', passenger)
self.passenger = passenger
def pay(self, amount):
print('Paid', amount)
self.balance = self.balance + amount
self.passenger = None
class Limo(Taxi):
def __init__(self):
Taxi.__init__(self)
self.sunroof = 'closed'
def drive(self):
print('Limo driving in luxury')
Car.drive(self)

tricky.

Python 3.6.0 Shell

Car is moving
Hired by Kim
Honk honk, out of the way
Car is moving
Paid 5.0
Hired by Jenn
Honk honk, out of the way
Car is moving
Pouring drink
Paid 10.0 Tip 5.0
Paid 15.0
>>>

def pay(self, amount, big_tip):
print('Paid', amount, 'Tip', big_tip)
Taxi.pay(self, amount + big_tip)
def pour_drink(self):
print('Pouring drink')
def open_sunroof(self):
print('Opening sunroof')
self.sunroof = 'open'
def close_sunroof(self):
print('Closing sunroof')
self.sunroof = 'closed'
you are here 4  571

exercise solution

Your turn to create a new class. How about a frisbee-catching dog? We’ve
already created a Frisbee class for you:
class Frisbee:
def __init__(self, color):
self.color = color

A Frisbee can't do much;
it just has a color and
a __str__ method so it
prints well.

def __str__(self):
return "I'm a " + self.color + ' frisbee'
Your job is to help us finish the FrisbeeDog. It needs to catch a frisbee
and give it back. It should have a __str__ method as well.

class FrisbeeDog(Dog):
def __init__(self, name, age, weight):
Dog.__init__(self, name, age, weight)
self.frisbee = None

We have a simple
constructor; it just
sets a frisbee attribute.
Notice frisbee here is
another object, so some
composition going on.

def bark(self):
if self.frisbee != None:
print(self.name,
'says, "I can\'t bark, I have a frisbee in my mouth"')
else:
We're overriding the bark method. If the frisbee dog
Dog.bark(self)

currently has a frisbee, then it can't bark; otherwise, it

def catch(self, frisbee):
barks like any other dog.
self.frisbee = frisbee
print(self.name, 'caught a', frisbee.color, 'frisbee')

The catch method takes a frisbee and assigns
def give(self):
it to the object’s frisbee attribute.
if self.frisbee != None:
frisbee = self.frisbee
self.frisbee = None
print(self.name, 'gives back', frisbee.color, 'frisbee')
return frisbee
else:
The give method sets
print(self.name, "doesn't have a frisbee")
the frisbee attribute
return None
to None, and then
def __str__(self):
str = "I'm a dog named " + self.name
if self.frisbee != None:
str = str + ' and I have a frisbee'
return str

572  Chapter 12

returns the frisbee.

And here's the str method, which
conditionally prints based on whether
the dog has a frisbee or not.

object-oriented programming

Override the FrisbeeDog’s walk method so that if a dog has a frisbee he says, “I can’t
walk, I’m playing Frisbee!” Otherwise, FrisbeeDog acts like a normal Dog. Go ahead and
add the code to your dog.py file. Use the same test code as the last Test Drive.
class FrisbeeDog(Dog):
def __init__(self, name, age, weight):
Dog.__init__(self, name, age, weight)
self.frisbee = None
def bark(self):
if self.frisbee != None:
print(self.name,
'says, "I can\'t bark, I have a frisbee in my mouth"')
else:
Dog.bark(self)
Dog has a frisbee if self.frisbee

is not equal to None.
def walk(self):
if self.frisbee != None:
print(self.name, 'says, "I can\'t walk, I'm playing Frisbee!"')
else:
If the dog has a frisbee, then output that the
Dog.walk(self)
dog is playing. Otherwise, do what all Dogs do by

def catch(self, frisbee):
calling the superclass's walk method.
self.frisbee = frisbee
print(self.name, 'caught a', frisbee.color, 'frisbee')

def give(self):
if self.frisbee != None:
frisbee = self.frisbee
self.frisbee = None
print(self.name, 'gives back', frisbee.color, 'frisbee')
return frisbee
else:
print(self.name, "doesn't have a frisbee")
return None
def __str__(self):
str = "I'm a dog named " + self.name
if self.frisbee != None:
str = str + ' and I have a frisbee'
return str

Python 3.6.0 Shell
Dude caught a red frisbee
Codie is walking
Jackson is walking
Rody is walking
Dude says, "I can't walk, I'm playing Frisbee!"
>>>

you are here 4  573

crossword solution

1

C

2

A
T

4

5

N

Coding
Cross
Solution

10

S

6

I

H

7

I

S

I

O

E

R V

R

E

I

R

T

P

I

N S

N C E

P

T

A

R U C

C

11

L

C

E

D O G

O

L

E

M

A

P O

L

Y M O R P

R

S

E

A

I

U

G

N

D

L

A

C

E

12

17

A
T

T

T

I

E

I

O

16

R

H

T O R

R

A

E

C E

U

T

9

N
I

S

C O N S

E

8

M

3

13

S M

14

S

E

L

15

F

S

S

T

U

I

H

I

O

S

D

B

B

U

C

T

E

I

18

R

I

L

O O P

E

O

A

N

S

N

E
D

S

O
G

Across

574  Chapter 12

4. Creates an object and initializes it.
[CONSTRUCTOR]
6. Built-in function to check IS-A. [ISINSTANCE]
10. Dog that helps humans. [SERVICEDOG]
12. Code that works over objects with the same
interface. [POLYMORPHISM]
14. Convention for the first parameter of a
method. [SELF]
17. Holds state of object. [ATTRIBUTE]
18. Has a lot of jargon. [OOP]

Down
1. Not allowed in the hotel. [CATS]
2. Type of inheritence that inherits from two
or more classes. [MUTIPLE]
3. Class from which others are subclassed or
derived. [SUPERCLASS]
5. IS-A relationship. [INHERITANCE]
7. Hiding implementation details.
[ENCAPSULATION]
8. Redefining the behavior of a method in a
subclass. [OVERRIDE]
9. HAS-A relationship. [COMPOSITION]

appendix: leftovers

The Top Ten Topics
(We Didn’t Cover)

We’ve covered a lot of ground, and you’re almost
finished with this book.We’ll miss you, but before we let you go,
we wouldn’t feel right about sending you out into the world without a little
more preparation. We can’t possibly fit everything you’ll need to know into this
relatively small chapter. Actually, we did originally include everything you need to
know about Python programming (not already covered by the other chapters), by
reducing the type point size to .00004. It all fit, but nobody could read it. So we
threw most of it away, and kept the best bits for this Top Ten appendix.
This really is the end of the book. Except for the index, of course (a must-read!).
this is a new chapter   575

list comprehensions

#1 List comprehensions
You’ve seen how we can create a list of numbers using the range function
in Python, but there is an even more powerful way to construct lists that is
similar to the way mathematicians construct sets of numbers. We call them
list comprehensions and they can construct lists of any type. Let’s first look at an
example with numbers:
[x + x for x in range(10)]

Doubles every number in the range
from 0 to 9

Python 3.6.0 Shell
[0, 2, 4, 6, 8, 10, 12, 14, 16, 18]
>>>

Or, how about an example with strings:
lyric = ['I', 'saw', 'heard', 'on', 'you', 'the', 'wireless', 'back', 'in', '52']
[s[0] for s in lyric]

Python 3.6.0 Shell

Grabs the first letter of each
word in the list lyric

['I', 's', 'h', 'o', 'y', 't', 'w', 'b', 'i', '5']
>>>

Okay, but how does this actually work? Well, essentially a list comprehension
creates a list from another list. To see how this works, let’s look at the format of a
list comprehension:

Revisit the examples above with this
formula in mind.

The first part is an expression that
uses a variable that represents each
We then have a for expression with
item in the existing list.
the variable and the old list.
And finally, we have
a conditional; we
haven’t
seen one of
[expression for item in list if conditional ]
these yet
Here’s how we can add a conditional to our example:
[s[0] for s in lyric if s[0] > 'm']

Python 3.6.0 Shell

Only add the item
to the list if it is a
character greater
than ‘m'.

As with many things in this appendix, this will take some study to master, but you
can see it is a powerful way to create new lists.

576  Appendix

['s', 'o', 'y', 't', 'w']
>>>

leftovers

#2 Dates and times
Dates and times are an important part of many computations. You can
import the Python datetime module like so:
import datetime

Then you can create a date object by instantiating it with any date up until
the year 9999 (or going to back to year 1):
my_date = datetime.date(2015, 10, 21)

Year, then month,
then day

Or you can create a time object by instantiating it with any time, providing
hours, minutes, and seconds:

Hour, minutes, and seconds

my_time = datetime.time(7, 28, 1)

Or combine them together with the datetime object:

Or combine them all together.

my_datetime = datetime.datetime(2015, 10, 21, 7, 28, 1)

Let’s print them all:
print(my_date)

We can print each date and time
object to see what it holds.

print(my_time)

print(my_datetime)

print(my_date.year, my_date.month, my_date.day)

Python 3.6.0 Shell
2015-10-21
07:28:01
2015-10-21 07:28:01
2015 10 21
7 28 1
>>>

print(my_time.hour, my_time.minute, my_time.second)

You can also get the current time with:
now = datetime.datetime.today()
print(now)

Notice you can get the
current time down to
microseconds.

Python 3.6.0 Shell
2017-07-27 19:12:07.785931

Or use datetime’s formatting facilities:
output = '{:%A, %B %d, %Y}'

print(output.format(my_date))

The date object supports a
rich formatting language.

Python 3.6.0 Shell
Wednesday, October 21, 2015

That’s just a start—in any language there’s a lot to know about handling dates
and times; check out the datetime and related modules in Python to dig
deeper into this topic.
you are here 4  577

regular expressions

'ac*\dc?'

#3 Regular expressions

Remember all the trouble we had matching text that contained words and punctuation?
We could have solved that problem using regular expressions. A regular expression
is, formally, a grammar for describing patterns in text. For instance, using a regular
expression (regex for short), you can write an expression that matches all text that starts
with t and ends with e, with at least one a and no more than two us in between.
Regular expressions can get complex fast. In fact, they can seem like an alien language
when you first encounter them. But you can get started with simple regular expressions
fairly quickly, and if you like them, become a true expert.
You’ll find regular expressions supported by most modern languages. Python is no
exception. Here’s how you use Python regular expressions:

import re

Import the regular expression module, re
Let's test a few strings...

for term in ['I heard you on the wireless back in 52',
'I heard you on the Wireless back in 52',
'I heard you on the WIRELESS back in 52']:
result = re.search('[wW]ire', term)
if result:
loc = result.span()
print('found a match between:',
else:
print('No match found')

found a match between: (19, 23)
found a match between: (19, 23)
No match found>>>

Search using a regular expression.
Here's the regular expression part; it
matches ‘wire' or ‘Wire'.
If there's a result, there's a match and the span
where the match is in the string.
loc) method tells us
Let's print out where the match is…

…or if there wasn't one.

In this example we used a very simple regular expression, but regular
expressions are capable of matching sophisticated patterns—for example, valid
usernames, passwords, and URLs. So, the next time you’re about to write code
to verify a valid username or something similar, pull out the regular expressions
instead—a couple lines of regex could save you a lot of code.
Search is just one method provided by Python for matching strings with an
expression. Python also has a sophisticated regex compiler in the re module
you’ll want to check out for intensive pattern matching applications.
The real key to understanding and using regular expressions is learning how
to read and create patterns. To do that you’ll want to study the general subject
of regular expressions, and more specifically, the notation that Python uses for
regular expressions.

578  Appendix

Python 3.6.0 Shell

a very brief intro to
jquery
leftovers

#4 Other data types: tuples
Python lists have a sister data type we never told you about: tuples. Syntactically the two
are almost identical. Check this out:
my_list = ['Back to the Future', 'TRON', 'Buckaroo Bonzai']
my_tuple = ('Back to the Future', 'TRON', 'Buckaroo Bonzai')

So, what's the difference? Well,
syntactically one uses square brackets
and one use parentheses.
You also already know how to use them, for instance, in iteration:
for movie in my_list:
print(movie)

for movie in my_tuple:

Same output!

print(movie)

A Python list of movies—well,
of strings, actually
A Python tuple of strings

Python 3.6.0 Shell
Back to the Future
TRON
Buckaroo Bonzai
Back to the Future
TRON
Buckaroo Bonzai

And of course you can refer to the items in a tuple with an index too, like my_
tuple[2], which evaluates to 'Buckaroo Bonzai'. Tuples even support most of
the same methods as lists, so what’s the difference? Tuples are immutable—meaning you
can’t change them like you can lists. Once you make a tuple, that’s it, you’re done—
you can access its items, but you can’t change them.
So why would you use them? Why do they exist? Because of what we call time and
space; in other words, tuples take up less memory and you can operate on them faster
than you can with lists. If you’ve got collections of items that are very large and/or
you are computing lots of operations on them, then you may want to consider tuples,
which can significantly improve both memory use as well as execution time.

It can also be
computationally safer to
work on data structures
that can't change!

You’ll also want to take a look at a few things you can do syntactically with tuples you
can’t do with lists:
x, y, z = (1, 2, 3)
apples = 'mac', 'red', 'green'
single = (3.14159265,)

Extra comma

This is called unpacking; Python assigns each
variable to the corresponding item in the tuple.
Even without parentheses, comma-separated
items become a tuple.

Careful, a quirk: because of unpacking, if you
want to assign a single-item tuple to a variable,
you need to add an extra comma.
you are here 4  579

the set data type

#5 Other data types: sets
There’s another common data type we haven’t discussed: sets. Do you
remember sets from algebra class? You might recall sets have only
one of each element value, and they have no order. You might also
remember that common operations on sets include taking the union
and intersection of one set with another. Well, Python has sets too; let’s
see what Python provides:

Sets are represented by comma-separated
values enclosed in curly braces.
set = {1, 3.14159264, False, 77}

Dictionaries have key/value pairs within
curly braces, while sets have just values.

Sets can hold practically any Python
type, but the values need to be unique.
Like lists, and unlike tuples, sets are mutable, meaning we can change
them and add values to them:
set.add(99)

Or remove them:
set.remove(1)

And of course they provide more interesting methods we can use:
even = {2, 4, 6, 8, 10}
odd = {1, 3, 5, 7, 9}
prime = {1, 3, 5, 7}
even_and_prime = even.intersection(prime)
print(even_and_prime)
odd_and_prime = odd.intersection(prime)
print(odd_and_prime)

Python represents the empty list
(having no items) as set().

Python 3.6.0 Shell
set()
{1, 3, 5, 7}
{1, 2, 3, 4, 5, 6, 7, 8, 10}
>>>

even_or_prime = even.union(prime)
print(even_or_prime)

You can also check to see if a value is in a set with the familiar if x in set syntax.
Check out the Python set documentation for even more interesting methods like set
differences, symmetric differences, and supersets.
580  Appendix

the arguments
object
leftovers

#6 Server-side coding
Many applications run as web services, providing services like
Google search, social neworks, and ecommerce. These applications
are powered by many different languages, including Python.
Learning server-side coding often requires familiarity with
technologies such as the hypertext transport procotol (HTTP),
HTML (a page markup language), and JSON (as we’ve seen,
for data exchange), to name just a few. In addition, depending
on the language you’re working with, you’ll most likely want
to use a web framework or package. These packages provide
functionality that takes care of many of the low-level details
of serving web pages and data.
If you are working with Python, two of the more popular
frameworks are Flask and Django.

@app.route("/")
def hello():

return "Hello, Web!"

Server-side code
executes on a server on
the internet.
r

t

es

u
eq

Client-side code executes
on the client—that is, on
your computer.

Here's a quick Flask example. With Flask we can
route web traffic from the “root” (think home
page) of a website to the function hello. Hello
simply outputs a simple string.

Typically you’d output
HTML or JSON here.
Flask is a small, more minimal framework aimed at smaller
projects you want to get up and running quickly. Django is a
much heavier-weight framework aimed at larger projects. As
such, Django takes more time to learn but will do a lot more for
you. Django provides page templating, forms, authentication, and
a means to adminster databases. Because we’d guess you’re just
getting started in web development, so we’d recommend Flask as
a place to start, and then upgrading to Django as your skills and
needs increase.

Learning HTML and CSS is also a vital part
of creating web pages. After Head First
Learn to Code you’re more than qualified
to take on this book, and perhaps even Head
First JavaScript Programming at the same
time—JavaScript is the de facto standard
for programming in the browser.
you are here 4  581

computing with lazy evaluation

#7 Lazy evaluation
Say you wanted to write some code like this:

This code returns the nth
prime number...
…from a list of primes.

def nth_prime(n):
count = 0
for prime in list_of_primes():
count = count + 1
if count == n:
return prime

Clearly we can't really do this,
because creating an infinite list of
primes takes, literally, forever.

def list_of_primes():
primes = []
next = 1
while True:
next_prime = get_prime(next)
next = next + 1
primes.append(next_prime)
return primes

We could generate an upper
limit, say the first 1,000,
but that could be very
inefficient as well.

We'll leave this function as a
homework assignment (or a
quick Google search).

We can fix this code using a technique called lazy evaluation or calculation on demand, and
in Python this is supported by something called a generator. Here’s how we alter our
code to create a generator:

need our primes

We no longer
def list_of_primes():
so we've removed it.
array,
next = 1
while True:
next_prime = get_prime(next)
next = next + 1
Now each time we generate a new
yield next_prime
prime, we use the yield statement.
So what does yield do? A bit of magic. Think about it like this: in the nth_prime
function we iterating over all the primes. When the for statement first calls list_
of_primes and the yield statement is used, a generator is created (otherwise known
as an iterator in many languages). The generator is an object that has one method,
next, which can be called (behind the scenes in the case of the for statement) to
get the next value. Each time next is called, the list_of_primes magically picks
up computation where it last left off (computing the next prime) and as soon as the
yield statement is invoked, it returns another value. This repeats as long as there are
values being generated.
Lazy evaluation is a fascinating and powerful form of computation worth looking
into more deeply.
582  Appendix

a very brief intro leftovers
to JSON

#8 Decorators
Decorators come from an object-oriented design practice
called design patterns. Python loosely implements the
decorator pattern by allowing you to “decorate” functions with
other functions. For instance, if you have a function that just
returns some text, you could create a set of decorators that
add things to that text, like HTML formatting. Say we wanted
to easily add HTML paragraph tags to a piece of text. We
could use decorators like this:
def paragraph(func):

def add_markup():

return '' +

return add_markup
@paragraph

def get_text():

return 'hello head

To make text a
paragraph in HTML,
we just start it with
a 
 and end it with
a .

We need to use our higher-order
function knowledge here. First we’re
going to be passed a function, func,
func() + ''
and then we’ll create and return
another function that surrounds
whatever func returns, when called,
with  and .
Then we use the @ syntax
to decorate another
first reader'
function, like get_text.

print(get_text())

Now when get_text is called, it will be
called within the decorator code, which
causes the  and  tags to be added
to whatever the get_text function returns.

As you can see, this can be a little mind-altering and probably
gives you some indication that higher-order functions are worth
studying.
Decorators are a powerful feature and can be used far beyond
decorating text. In fact, you can add one or more decorators to
any type of function that you’d like to give additional behavior
to, without altering the original function.

Python 3.6.0 Shell
hello head first reader

you are here 4  583

first-class functions

#9 Higher-order and first-class functions
You’ve come a long way in your understanding and use of functions. We’ve talked
about functions as a way to abstract code, but there is so much further you can take
them. In fact, you can do anything with a function that you can with other data
types: you can assign a function to a variable, pass a function to another function,
and even return a function from a function. When we can treat a function like any
other object or data we call it first class, and when we either pass or return a function
from another function, we call that a higher-order function. But what does it even
mean to pass or return a function from a function?
Understanding higher-order functions is a topic worthy of some lengthy study; for
now, whet your appetite with some code that demonstrates using functions in a
higher-order manner.

def pluralize(str):

def helper(word):
return word +

Did you know you can declare a
function within another function?
The helper function is only available
within the pluralize function; no
"s"
other code can call it.

return helper(str)

Python 3.6.0 Shell
girls
>>>

The pluralize function uses helper to
add an “s” onto a string.

val = pluralize('girl')
print(val)

Okay, that’s new, but let’s make this truly higher order. Hang on to your hat:
def addition_maker(n):
def maker(x):

return n + x

return maker

The function addition_maker
takes a number n...
…and defines a function that takes Python 3.6.0 Shell
another number, x, and adds n to it.
3
And then, it returns that function as
>>>
the result of calling addition_maker.

add_two = addition_maker(2)
val = add_two(1)
print(val)

So if we pass addition_maker the
number 2, it returns a function, which
here we assign to the variable add_two...
…that, when called, adds two to
whatever it is passed, in this case the
number 1.

If that’s a little hard to wrap your head around, welcome to the club. But, like
recursion, with a little study and work, thinking in higher-order functions
becomes second nature.
584  Appendix
  

recursive functions
leftovers

#10 A lot of libraries
We’ve seen a few Python libraries throughout this book, but there are many more built-in
and third-party Python packages waiting for you. Here are a few to take a look at:

requests
Flask, Django

sched
logging

Pygame
Beautiful Soup

Pillow

This third-party package is worth mentioning again. If you
need to make HTTP requests from your application, this is the
package for you.
You’ve seen these frameworks in passing as well. If you need
to build server-side Python code, be sure to check out both.
Flask tends to be for small-scale projects and Django for large
applications, but both should serve you well.
Need to execute code at specific times or on a schedule? The

sched module is a standard Python module that allows you to

create schedules and invoke code on that schedule.

An upgrade from putting simple print statements in your code,
the logging module is a built-in module that gives you a way
to output informational warning or error messages. Even better,
you can configure the type of messages that are output based
on the type of execution (say, testing versus a production run of
your code).
Pygame is a library for writing video games. Pygame provides
modules for dealing with game graphics as well as sound. If
you’re just starting to write games, this can be a great place to
start.
Data on the web isn’t always available in JSON, and often
you’ll just need to grab a web page and sort through the
HTML to find what you need. Unfortunately, many web
pages contain poorly written HTML. The Beautiful Soup
library tries to make all that easier and gives you a nice—no,
beautiful—interface to web pages from Python.
Pillow is an image library, with everything you need to read,
write, and process many common image file formats. Pillow also
provides support for displaying files and converting images to
other formats.
You’ll find a lot more at https://wiki.python.org/moin/UsefulModules
you are here 4  585

I can’t believe the book is almost over.
Before you go, you really should read the
index. It’s great stuff. And after that you’ve
always got the website. So I’m sure we’ll see
you again soon...

Don’t worry, this
isn’t goodbye.
Nor is it the end. Now that you’ve got an
amazingly solid foundation in learning to code,
it’s time to become a master. Point your browser
to http://wickedlysmart.com to explore what’s
next!

What’s next? So much more!
Start on either path.
Jump head first in to
OOP and an industrystandard language,
Java.

Follow it with
learning high-level
object design with
design patterns and
principles.

586  Appendix

Learn the languages
that provide the
structure and style
of the web…

Learn browser-based web
development by reading
these books in parallel.
…and the
language that
provides the
behavior.

Follow it up
with even more
sophisticated
browser coding.

Index
Symbols

- (negation operator) 47–48

\ (backslash) 400

() (parentheses)
Boolean expressions and 102
in functions 60, 183
grouping operations 45
operator precedence and 48
subclassing 536

: (colon)
file paths 401
indented statements and 82, 105, 183
key/value pairs 364
slice syntax and 272–273
, (comma) 255, 364, 579
{} (curly brackets) 445, 580
. (dot notation)
attributes and 321
classes and modules 318
docstrings 300–302
examples of 77, 294, 308, 529
methods and 321
variables and 321

!= (not equal operator) 81, 101–102

+ (plus sign)
addition operator 47–48, 158
concatenation and 25–26
>>>> prompt 40
“” (quotation marks)
declaring lists 25–26
strings and 42, 88, 250–251
[] (square brackets) 24–25
- (subtraction operator) 47–48

__ (double underscore) 51

@ syntax 583

== (equality operator) 81–82, 93

_ (underscore) 51

= (equal sign)
assignment operator 38, 82
keywords arguments 212

A

** (exponentiation operator) 47–48
/ (forward slash)
as division operator 47–48
as file path separator 400–401

ABC programming language 15
absolute paths 401

< (less than operator) 93

abstracting code
about , 179–182, xiv–xvii
avatar example 195–199
object-oriented programming. See OOP
organizing modules for reuse 292–293, 299–304
turning into functions 183–190, 525
Web APIs and. See Web APIs

% (modulus operator) 47–49

access tokens (Web API) 440

* (multiplication operator) 45, 47–48

addition operator (+) 47–48, 158

> (greater than operator) 80, 93
>= (greater than or equal to operator) 81
# (hash character) 24, 99

This is the index

587

the index
after method 499–502

Boole, George 81

and Boolean operator 92–93, 101–102

break statement 405–406

append method 156, 158, 320

bubble sort
about 227–229
computing numbers 236
implementing in Python 234–235
pseudocode 231–232

arguments in constructors 320
arguments in functions
about 61, 183, 186
evaluating 204
keywords in 212, 487
order of 192
parameters comparison 192, 204, 211
argv attribute 423
assignment operator (=) 38, 82

bugs (software) 54, 414
Button object 486

C
calculation on demand 582

assignment statements 41

call stacks 351–354

associative maps 367

case sensitivity
class names 318, 320
from keyword and 486
Python programming language 51
strings 103

asterisk (*) 45
attributes
about 317, 319, 541, 544, 546
accessing through methods 319, 531
dictionary items 368, 370–372, 450
dot notation and 321
response object 448
state in form of 315–317, 525, 527
turtle example 306, 308, 316

B
backslash (\) 400
base class 536, 544
\b (backspace) 408
Beautiful Soup library 585
behavior of objects
about 306, 308, 315–317, 527
OOP and 525, 542–544
overriding and extending 542–544

choice function (random) 25–26, 86
classes
about 316, 320–321, 544
behaviors and 316–317
built-in 320–321
case sensitivity 318, 320
designing 527, 549, 560
diagramming 316, 525, 527, 549, 558
dot notation and 318
implementing 535, 550
inheritance and 534–537, 541, 544, 558
objects and. See objects
terminology 536
using 318
writing 528–530
client-side coding 581

Bergey, Nathan 444

close method 402

blocks of code. See code block

closing files 402, 420

Boolean data type 80–82, 87, 477

code block 105, 183–188, 418

Boolean expressions 81–82, 87, 92

colon (:)
file paths 401

Boolean operators 92–93, 101–102
588

index

the index
indented statements and 82, 105, 183
key/value pairs 364
slice syntax and 272–273
comma (,) 255, 364, 579
command-line execution 423–424, 442
comments 24, 98–99, 300–301
communication protocol 438
composition 546, 558
computational thinking , 1–4, x–xiii, 14, xiv
concatenating
list items 24–25
strings 46, 96, 139
conditional expressions
about 86
choice function 25–26, 86
if statement 82–85, 92
infinite loops 113
while statement 105–110
constructors
about 318, 544
arguments in 320
functionality 528–529
subclassing and 536
controller (MVC pattern) 491–510
converting
numbers to strings 140
strings to numbers 59–60
Conway, John 468
C programming language 9
Crazy Libs game
about 394–395
creating template 396
handling exceptions 413–415, 417–420
processing template text 411–412
prompting for user input 397
reading template 397, 399
signifying placeholders in template 396–397
storing 421
updating code 420–424
writing completed file 397

CSS (Cascading Style Sheets) 581
Ctrl+C keyboard combination 113
curly brackets 445, 580
current directory 400
current time 577

D
data science 246
data types
about 57–58, 316
Boolean 80–82, 87, 477
classes 316–321
dictionaries 362–372
lists. See lists and list items
None 214, 420–421
numbers. See numbers
primitive , xiii
sets 580
tuples 579
date object 577
datetime module 305, 577
debugging
about 54–56
suggestions for 19, 22, 60
decorators 583
default values
assigning in functions 195, 197–198
parameters with 210
def keyword 183
delegation 546, 559, 560–564
deleting
dictionary keys 363
list items 157–158
del function 157–158, 363
delimiters
commas 255
space separators 24–25, 61, 96, 139
whitespace 253–256
derived class 536, 544
you are here�

589

the index
design patterns 409, 583

even method 580

dictionaries
about 362, 367
accessing items 366
adding items to 362, 366
attributes of items 368, 370–372, 450
JSON format and 446
key/value pairs and 362–364, 580
list comparison 367–368, 553, 561
memoization and 376–377

event-based programming 494–498

division operator (/) 47–48
Django framework 581, 585
docstrings 300–302
Doctorow, Cory 292–293, 299–300, 304
documentation 98–99, 439–441
dot notation (.)
attributes and 321
classes and modules 318
docstrings 300–302
examples of 77, 294, 308, 529
methods and 321
variables and 321
double underscore (__) 296

E

exception handling
about 418
Crazy Libs game 413–415, 417–420
debugging process 54–56
detecting invalid input 100–103
infinite loops 113
int function 59–60
lists 134, 139–140
numbers 53–55
parameters 211
return values 56
strings 53–55, 140
suggestions for 19, 22, 60
syntax 22, 60, 102
variables 207
Web API requests 440
except keyword 418
executing code
about 6, 13
from command line 423–424
from IDLE 18–21
from Python Shell 40–41, 188
exponentiation operator (**) 47–48

elif keyword 83–85
else keyword 82–85
empty lists
about 134
checking for 344
creating 156, 159, 321
empty strings 88, 198, 349, 409
encapsulation 319, 552, 561
EOL error 22, 43
equality operator (==) 81–82, 93
equal sign (=)
assignment operator 38, 82
keywords arguments 212
escape sequences 407–408
590

event handlers 494–498

index

expressions
applying operator precedence in 48
assigning values to variables 45–46, 80
Boolean 81–82, 87, 92
conditional 24–25, 82–87, 92, 105–110
regular 578
extend function 157–158

F
False value 80–83, 87, 92, 106
Fibonacci sequence 357–359, 373
file management
closing files 402, 420
file modes 399, 421
naming considerations 309

the index
opening files 399–402, 418, 420
reading files into code 402–409
updating code 420–424
writing files 421
File menu (IDLE editor)
New File item 18, 26
Save menu item 20, 26
file object
assigning to variable 399
close method 402
readline method 407, 409
read method 403
write method 421
file paths 400–402
first-class functions 584
Flask framework 581, 585
Flesch-Kincaid readability formula 247–248, 278, 282,
301–303. See also readability formula
Flesch, Rudolph 247
float function 59
floating point numbers 58–59, 320
flowcharts 76
for statement
about , 105, xiii
break statement and 406
bubble sort example 231–232, 234–235
in dictionaries 364
in operator and 364, 409
iteration and 105, 142–148, 151–152, 409
in lists 142–148
sequences and 78, 145, 152
while loop comparison 142, 151–152
forward slash (/)
as division operator 47–48
as file path separator 400–401
fractals, generating 379–382
frames 351–354
from keyword 486, 491
function calls
call stacks 351–354

how to perform 184–188
memoization 376–377
recursive functions and 350–354
to other functions 192
functions
about 37, 77, 183, 192, 315
abstracting code. See abstracting code
accessing within modules 252, 294
arguments in. See arguments in functions
defining 183–185, 192
executing 183–188
first-class 584
hash 367
helper 410
higher-order 583–584
methods and 317, 531
naming 183, 192
parameters in. See parameters in functions
passing functions to 192
process overview 184–188
recursive 343–359
return value 39, 56, 85, 183, 193
turning blocks of code into 183–188
variables and 194, 200–209

G
Game of Life
about 470, 490
adding design patterns 506
building data model 477
building screen view 485
building simulator 472–476
building the controller 491–510
building the data model 477–484
building the view 485–490
Clear button 472, 485, 502–503
complete source code 511–515
computing a generation of 478–484
computing at regular intervals 499–502
correcting layout 488
creating widgets 486–488
defining design patterns 509
you are here�

591

the index
editing cells directly 504
Option Menu 472, 474–475, 485, 506–510
Pattern Loader 472, 474–475, 510
Pause Button 472, 474–475, 494–498
placing widgets into grid layout 489
reactive computation for 494–497
rules in 468–471, 478
sketch of 472
Start button 472, 475, 485, 497–498
translating grid layout to code 489–490
writing grid view handler 505
writing option menu handler 507–508
writing pattern loader 510
generative systems 470
generators 582
get function (requests) 447–448
global keyword 207
global variables
about 194
local variable comparison 208–209
__name__ 296–298
scope of 200–203
using in functions 207
Golden Ratio 357
graphical user interface (GUI). See Game of Life

higher-order functions 583–584
history of Python 15
HTML (Hypertext Markup Language) 581, 583
HTTP (Hyptertext Transfer Protocol) 438, 448–450, 581

I
IDEs (Integrated Development Environments) 13, 18
IDLE editor
about 13
installing xxxiv
Python sample programs 18–21, 24–27
if statement
conditional logic and 82–85, 92
in operator and 261, 269, 363
imaginary numbers 58
import statement 24–25, 77–79, 294, 491
indenting statements 82, 105, 183
Index Error 134
indices
dictionary comparison 363
for lists 132–133, 138, 157, 235–236, 477
for strings 272–276
for tuples 579

greater than operator (>) 80, 93

infinite loop 113

greater than or equal to operator (>=) 81

inheritance 534–537, 541, 544, 558

GUI (graphical user interface). See Game of Life

in operator
for statement and 364, 409
if statement and 261, 269, 363
sets and 580

H
HAS-A relationship 546
hash character (#) 24, 99
hash functions 367
hash maps 367
help documentation 99, 300–304
helper functions 410
help function 300
heuristics 264–273, 276–277
592

index

input
defined 22
detecting invalid 100–103
file-based. See file object
prompting for. See prompting for input
input function
functionality 37–39
quotation marks and 42
return values 39, 56, 85
syntax 37

the index
insert function 158
insertion sort 239

K

installing
IDLE editor xxxiv
packages 441–442
Python xxxiv–xxxv, 17

KeyError 363

instances. See objects
instantiating objects 318, 525, 544

key/value pairs
dictionaries and 362–364, 580
JSON format and 445–446

integer numbers 58–59, 140

keyword argument 212, 487

Integrated Development Environments (IDEs) 13, 18

keywords 19, 51. See also specific keywords

International Space Station 443–444, 449–461

koch function 379–382

interpreting code 6, 13. See also Python interpreter

Koch, Niels Fabian Helge von 380

int function 59–60, 140

Koch snowflake 380–382

instance variables. See attributes

keys
dictionary 362–364, 367
in JSON 445

IS-A relationship 537–539

L

isinstance function 538–539

lazy evaluation 582

iteration
about , xiii
breaking out of 405–406
dictionaries and 364
for loop. See for statement
in Game of Life 477–479
lines of files 409
lists and 138, 142–148
recursion comparison 355–356
tuples and 579
while loop. See while statement

len function 131, 158, 254, 268

iterations
infinite loop 113

list function 156

invalid syntax error code 22, 60

iterators 409

J
Java programming language 8
JavaScript programming language 9, 581
JSON format 445–448, 450–451, 581
json module 450–451

less than operator (<) 93
library. See modules
line feed (\n) 403, 407–408
Linux environment
file paths in 400–401
installing Python 17
installing requests package 442
LISP programming language 9
list comprehensions 576
list multiplication 156, 477
lists and list items
about , 125–128, 134, xiii–xvi, 146
accessing 129, 134
accessing last item 132
adding to another 157–158
altering 263
creating 128, 156
declaring 25–26
deleting 157–158
you are here�

593

the index
determining number of items 131
dictionary comparison 367–368, 553, 561
empty lists 134, 156, 159, 321, 344
indices 132–133, 138, 157, 235–236, 477
inserting items 158
iteration in 138, 142–148
lists as classes 320
ordering 134
output considerations 139–140
parallel lists 162, 235, 368
range of numbers 145–146, 148
slicing 273, 344
tuples and 579
updating values in 129
zero index value 132, 138, 157, 477
lists within lists 477
literal notation 364
loads function 450–451
local variables 194, 200–203, 207–209
logging module 585

M
macOS environment
installing Python xxxiv
MacOS environment
command 442
command-line execution 424
file paths in 400–401
installing Python 17
Python sample program 18
mainloop function 309, 496
Massachusetts Institute of Technology 307
math module 305
memoization 376–377
merge sort 239
metacognition xxix
methods
about 317, 319, 527, 541, 544
accessing attributes 319, 531
594

index

constructor 318
dot notation and 321
functions and 317, 531
invocation process 532
tuples and 579
writing 531–533, 554, 559
model (MVC pattern) 477–484
modules
about 22, 77, 252, 294–295
accessing functions within 252, 294
accessing variables within 252, 294
datetime 305, 577
dot notation and 318
importing 486
json 450–451
logging 585
math 305
naming 309
organizing for reuse 292–293, 299–304
popular Python 305
random. See random module
request. See requests module
sched 585
sys 423
TKinter. See TKinter module
turtle. See turtle module
module view controller pattern. See MVC (module view
controller) pattern
modulus operator (%) 47–49
multiple inheritance 541
multiplication operator (*) 45, 47–48
MVC (module view controller) pattern
about 476
building data model 477–484
building the controller 491–510
building the view 485–490

N
__name__ global variable 296–298
naming considerations

the index
Boolean data type 81
class names 318
files 309
functions 183, 192
modules 309
variables 51
National Research Institute for Mathematics and Computer Science (Netherlands) 15
negation operator (-) 47–48
negative indices
for lists 132–133
for strings 273
nested loops 231–232, 234–235
New File menu item (File menu) 20, 26
newline (\n) 403, 407–408
next method 582
\n (newline) 403, 407–408
None value and type 214, 420–421
not Boolean operator 92–93
not equal operator (!=) 81, 101–102
numbers
about 58
converting strings to 59–60
converting to strings 140
data types supported 57
error handling 53–55
Fibonacci sequence 357–359, 373
floating point 58–59, 320
imaginary 58
integer 58–59, 140
pseudo-random 78
Python interpreter and 57
sequences of 78, 145–146

O

behavior of. See behavior of objects
classes and. See classes
constructors and 318, 320, 528–529, 544
HAS-A relationship 546
instantiating 318, 525, 538–539, 544
IS-A relationship 537–539
methods and. See methods
state of 315–317, 525, 527
using 318
object variables. See attributes
odd method 580
OOP (object-oriented programming)
about 436, 524–526
composition 546, 558
delegation 546, 559, 560–564
designing classes 527, 549, 560
encapsulation 319, 552, 561
HAS-A relationship 546
implementing classes 535, 550
inheritance 534–537, 541, 544, 558
IS-A relationship 537–539
overriding and extending behavior 542–544
polymorphism 555–556
subclassing. See subclasses
superclasses 536, 544
terminology 536, 544
writing classes 528–530
writing methods 531–533, 554, 559
open function 399–401
opening files 399–402, 418, 420
Open Notify API 439–440, 443–444, 449
open-notify.org website 443–444, 447
Open Weather Map organization 438
operator (expressions) 45. See also specific operators
operator precedence 47–49, 93
or Boolean operator 92–93

Objective-C programming language 8

organizing source code xxxvi, 20, 22

objects. See also OOP
about 315–316, 321
attributes of. See attributes

output
defined 22
error handling 53
you are here�

595

the index
file output 420–422
list considerations 139–140
print function 22–23, 61–63, 96, 139

plus sign (+)
addition operator and 47–48
concatenation and 25–26

overriding behavior 542–544

polymorphism 555–556

P

pop method 363

packages
about 294, 441, 585
installing 441–442

prime method 580

positional arguments. See arguments in functions

palindromes 171, 347–354
paper prototyping 473
Papert, Seymour 307
parallel lists 162, 235, 368
parameters in functions
about 183, 192, 201
abstracting code example 197–198
arguments comparison 192, 204, 211
default values 210
keywords in 212, 487
listing required first 211
Python interpreter processing 185–186
recursive functions and 350
variables comparison 204
variable scope and 200, 202
parentheses ()
Boolean expressions and 102
in functions 60, 183
grouping operations 45
operator precedence and 48
subclassing 536
path separator (/) 400–401
PEMDAS mnemonic 48
period (.). See dot notation
Perl programming language 9
persistent storage. See file management
PHP programming language 8
Pillow image library 585
pip utility 441–442
pixels 310
596

index

primitive data types , xiii
print function
about 22–23
fixing output glitches 139
space separators 96, 139
user-friendly output 61–63
problems, breaking down 2–4
programming languages 6–10, 14
prompting for input
continually 104–105
Crazy Libs game 397
handling errors 56
step-by-step process 37–39
storing user responses in variables 85
while loop logic 110
pseudocode
about 4, 34, 37
anti-social network example 371
attributes of items 368
bubble sort 231–232
computing readability 249
crafting 6
creating 362, 364
deleting keys 363
flowcharts and 76
Game of Life simulator 491
index comparison 363
iterating through keys 364
lists comparison 367–368
order considerations 364
retrieving values by keys 362–363, 367
testing for keys 363
turtle races example 323
writing code from 36

the index
pseudo-random numbers 78
punctuation, removing for readability 272–276
.py file extension 20, 252
Pygame library 585
Python interpreter 13, 57, 183–188
Python library. See modules
Python NameError 22
Python programming language
about 10, 14, 23
brief history of 15
case sensitivity 51
decision-making in 80
installing xxxiv–xxxv, 17
sample programs 18–21, 24–27
versions of 15–16, 442
writing and running code 13
Python Shell window
closing 113
entering sample programs 18–19, 24–26, 40–41
viewing program output 21

Q
quicksort algorithm 239
quotation marks (“”)
declaring lists 24–25
strings and 42, 88, 250–251, 301

R
randint function (random) 77–79, 158
random module
choice function 24–25, 86
importing 24–25, 77–79, 294
randint function 77–79
random number generation 77–79
range function 145–146, 148, 576
readability formula
about 247–248, 278, 282, 301–303
analyzing text for 250–251
computing sentences for 247–248, 257–263

computing syllables for 247–248, 264–273, 276–277
computing words for 247–248, 253–256
customizing for reuse 292–293
dealing with extraneous characters 256–257
implementing 278–283
pseudocode for 249
removing punctuation 272–276
setting up function for 252–253
reading files into code 402–409
readline method 407, 409
read method 403–404
recipe. See pseudocode
recursive functions
about 343–346
determining end of 350
Fibonacci sequence 357–359, 373
generating fractals 379–382
how recursion works 350–354
iteration comparison 355–356
palindrome example 347–354
refactoring code , 179–182, 190, xiv–xvii
regular expressions 578
relational operators 80, 93
relative paths 400–401
requests module
about 305, 441–442
get function 447–448
requests package 441, 585
requests (Web APIs)
about 437
access tokens and 440
error handling 440
forming URLs 444
request module process 447
responses (Web APIs)
about 437
ISS example 444
JSON format and 445
request module process 447–448
retrieving stored variable values 44
you are here�

597

the index
return statement 192

skeleton code 258

return value (functions)
about 183, 193
assigning to variables 39, 85
handling errors with 56

slicing
lists 273, 344
strings 272–276

reverse method 320

sorting data
about , xv–xvi
bubble sort 227–239

Rock, Paper, Scissors game 74–76
root window 486

sort function 239

Run menu (IDLE editor) 21–22, 27

space separators
concatenated items and 25–26
multiple arguments and 61
print function 96, 139

Run Module menu item (Run menu) 21–22, 27

spaghetti code 476

runtime errors
about 54, 417
examples 59, 363

split function 253–256

Rossum, Guido van 15
r (read mode) 399

S

Spotify website 439
square brackets [ ] 25–26

sched module 585

statements
about 2–3
assignment 41
indenting 82, 105, 183

Scheme programming language 9

state of objects 315–317, 525, 527

scientific notation 58

status codes (HTTP) 448–450

screen object 327, 453–456, 460

steps, breaking down problems into 2–3

search method 578

storing variable values 38, 44, 85

semantic errors 54, 417

str function 140, 183

semantics of programming languages 7, 38, 535

strings
about 38, 42, 57–58, 262
accessing last character in 263
adding multi-line text into code 250–251, 301
analyzing for readability formula 250–251
calculating terminal characters for sentences 257–261
case sensitivity 103
as classes 320
concatenating 46, 96, 139
converting numeric values to 140
converting to numeric values 59–60
empty 88, 198, 349, 409
error handling 53–55, 140
immutable nature of 263
indices for 272–276

Save menu item (File menu) 20, 26

sentences, computing for readability formula 247–248,
257–263
sequences
of characters 58, 257, 262–263, 268
for loops and 78, 145, 152
len function and 268
of numbers 78, 145–146
of values 142, 260
server-side coding 581
sets 580
shadowing variables 202
shoushiling (game) 74–76
598

index

the index
keys in JSON as 445
matching with expressions 578
Python interpreter and 57
quotation marks and 42, 88, 250–251
slicing 272–276
splitting apart 253–256
stripping characters off 413–414
substrings. See substrings
string (str) class 540
strip method 413–414
subclasses 536–537, 540, 544, 560
substrings 253, 263, 272–276
subtraction operator (-) 47–48
sum function 342
superclasses 536, 544
Swift programming language 8
syllables, computing for readability formula 247–248,
264–273, 276–277
syntax 7, 37, 535, 583
syntax errors
about 54, 417
examples 22, 43, 60, 102, 211
sys module 423

T
templates 396–397
terminal characters, calculating for sentences 257–261
terminating programs 113
testing code 40, 165, 473
text editors 13
text files 396, 399
text strings. See strings
thinking computationally , 1–4, x–xiii, xiv
third-party packages 441
time object 577
Timsort algorithm 239
Tkinter module 305, 485, 506

Traceback error message 417
triple quotation marks 250–251, 301
True value 80–83, 87, 92, 106
try/except block 417–418, 420
\t (tab) 408
tuples 579
turtle module
about 305–307
adding turtles 311–313
creating turtles 308–310, 314
generating fractals 379–382
OOP rework 563–565
purpose of attributes in 319
purpose of classes in 317
purpose of methods in 319
purpose of objects in 315
screen object and 327, 453–456, 460
turtle races example 322–333
using classes in 318
using objects in 318

U
UnboundLocalError 207
underscore (_) 51
unpacking 579
upper method 320
URLs (Uniform Resource Locators) 438–440
usability testing 473–475
use cases 473, 475
user interface widgets 485–489

V
variables
about 38, 44, 46
accessing within modules 252, 294
assigning default values to 195
assigning objects to 399
assigning return values to 39, 85
assigning values to 41, 44, 58
you are here�

599

the index
computing expressions with 45–46, 80
creating 88
dot notation and 321
functions and 194, 200–209
global 194, 200–203, 207–209, 296–298
instance. See attributes
local 194, 200–203, 207–209
naming 51
parameter comparison 204
shadowing 202
storing values in 38, 44, 85
in while loops 105–110
versions of Python 15–16, 442
view (MVC pattern) 485–490
vowels, computing for syllables in words 266–273,
276–277
\v (vertical tab) 408

W
web addresses 438–440
Web APIs
about 436–437, 446
making web requests 443–444, 447–449
using JSON 445
web addresses and 438–440
while statement
break statement and 406, 409
bubble sort example 231–232, 234–235
fireside chat 151–152
for loop comparison 142, 151–152
introduction 105–110
whitespace 253–256
widgets 485–489
Windows environment
command-line execution 424, 442
file paths in 400–401
installing Python xxxv, 17
Python sample program 18

600

index

words, computing for readability formula 247–248,
253–256
write method 421
w (write mode) 399, 421

Y
yield statement 582

Z
zero index value
for lists 132, 138, 157, 477
for strings 274

Don’t know about the website?
We’ve got updates, video,
projects, author blogs, and more!

Bring your brain over to
wickedlysmart.com

Source Exif Data:

File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.6
Linearized                      : No
Language                        : Ÿ§[S1
Page Mode                       : UseOutlines
XMP Toolkit                     : Adobe XMP Core 5.2-c001 63.139439, 2010/09/27-13:37:26
Create Date                     : 2017:12:13 10:51:52-05:00
Modify Date                     : 2017:12:14 11:56:59-05:00
Metadata Date                   : 2017:12:14 11:56:59-05:00
Creator Tool                    : Adobe InDesign CC 13.0 (Macintosh)
Instance ID                     : uuid:ed5be413-454f-ef42-a29b-2dfff34eac94
Original Document ID            : adobe:docid:indd:4cd9de28-cdcb-11dd-a607-cffab7c9696f
Document ID                     : xmp.id:ecf725bc-0ddf-4dc4-8807-f9b31a828f78
Rendition Class                 : proof:pdf
Derived From Instance ID        : xmp.iid:1d730af4-8762-4a4d-8e91-858c14307c9d
Derived From Document ID        : adobe:docid:indd:4cd9de28-cdcb-11dd-a607-cffab7c9696f
Derived From Version ID         : 1
Derived From Original Document ID: adobe:docid:indd:4cd9de28-cdcb-11dd-a607-cffab7c9696f
History Action                  : converted
History Parameters              : from application/x-indesign to application/pdf
History Software Agent          : Adobe InDesign CC 13.0 (Macintosh)
History Changed                 : /
History When                    : 2017:12:13 10:51:52-05:00
Format                          : application/pdf
Title                           : Head First Learn to Code
Creator                         : Eric Freeman
Producer                        : Adobe PDF Library 15.0
Trapped                         : False
Tagged PDF                      : No
Page Layout                     : SinglePage
Page Count                      : 641
Has XFA                         : No
Author                          : Eric Freeman
EBX PUBLISHER                   : O'Reilly Media

EXIF Metadata provided by EXIF.tools

Head First Learn To Code A Learner's Guide Coding And Computational Thinking

Head%20First%20Learn%20to%20Code%20-%20A%20Learner's%20Guide%20to%20Coding%20and%20Computational%20Thinking

Head%20First%20Learn%20to%20Code%20-%20A%20Learner's%20Guide%20to%20Coding%20and%20Computational%20Thinking

Head%20First%20Learn%20to%20Code%20-%20A%20Learner's%20Guide%20to%20Coding%20and%20Computational%20Thinking

Head%20First%20Learn%20to%20Code%20-%20A%20Learner's%20Guide%20to%20Coding%20and%20Computational%20Thinking

Head%20First%20Learn%20to%20Code%20-%20A%20Learner's%20Guide%20to%20Coding%20and%20Computational%20Thinking

Navigation menu

Versions of this User Manual:

Views

Navigation