Proceedings_of_the_Meeting_of_the_Eastern_Region_of_the_1620_Users_Group_196405 Proceedings Of The Meeting Eastern Region 1620 Users Group 196405
Proceedings_of_the_Meeting_of_the_Eastern_Region_of_the_1620_Users_Group_196405 Proceedings_of_the_Meeting_of_the_Eastern_Region_of_the_1620_Users_Group_196405
User Manual: Proceedings_of_the_Meeting_of_the_Eastern_Region_of_the_1620_Users_Group_196405
Open the PDF directly: View PDF 
.
Page Count: 481
| Download | |
| Open PDF In Browser | View PDF |
PROCEEDINGS OF THE
~TING
OF THE EASTERN REGION OF THE
1620'USERS GROUP
Marriott Motor Hotel
Arlington, Virginia
May
6, 7 aM 8, 1964
Hartin lnebel
Regional Secretal7
(',
PROCEEDIlfGS OF THE MEB'l'DG OF THI
EASTERN REGION OF THE 1620 USERS QROUP
MAY 6, 7 AIm 8,
1964,
TABLE OF OCtiTENTS
Sub
Section Section !!&!
SECTICJf 1 - GENERAL. " ., • • • • • • • • • • •
•
1.
o.
0
• •
• •
1.
2.
• •
1.
1.
1.
3.
1
1
1
• •
1.
4.
1
SECTION 2 - KEYNOTE ADDRESS. • • • • • • • • • •
Consolidation Through Systems Integration
by Dr. James R. OliTer • • • • • • • • • • •
2.
o.
0
2.
1.
1
• •
3.
o.
0
• •
• •
1.
2.
3.
• •
3.
3.
3.
3.
3.
1
1
1
1
1
• •
• •
3.
3.
6.
1
• •
• •
7.
8.
• •
3.
3.
3.
3.
9.
1
1
·..•• •
3.
3.
3.
9.
10.
1
li.
1
Progr_ Agenda. • .. • • • • • • • • • • •
Synopsis ot Papers •
• • • • • • • • • •
Ca.eat. ot the Program Chairman • • • • •
Some Hot•• on "SurYeY List ot Programaing
S7st••S n and Panel Discussions • • • • •
45
(~.
SECTION 3 - PROGRAMS TEAMS SESSIONS. • • • •
Statisti.. and Mathematiea • • • • • • •
Plotting • • .. ,. • • • • • • • • • • • •• •
Electronic •• • • • • • • • • • • • •
Education Institutions • • • • • •• • • •• ••
Structural and Cinl • • • • • • • • • • •
Abstracts "General Pipe Stress" by'
E. J. Orth, Jr. and J. Levis.
Electric Utility • • • • • • • • • • • • •
Abstracts "A Program tor ETaluating
Alternate Generating Plant
ExpanSion Pattern." by'
R. W. DaTiee. • •
CheJllical Engineering • • • • • •• •• •• • •• ••
Rate Engineering • • • • • • • • • • ..
•
Operations Research. • • • • • • • • •• •
Abstract: "Opt1aum Replacement ot
Capacit7-tdRited Equipaent"
by W. W. Pleines. • • • • • •
General Data Processing. • • • • • •
•
1710 Users Sub-Group • • • • • • • • •• • •
$
ill
C·
,',.1
• •
• •
•
•
•
•
4.
S.
S.
6.
3
7
1
4
1
1
I
· · ..
1
T
7["-r. ["8...... n"
\-
-±
TABLE OF OOHTENTS
(Continued)
Sub
Section Section
h.
4.
4.
4.
o.
o
1.
o
2.
o
h.
4.
S.
o
o
4.
6.
o
4.
7.
o
8.
o
9.
o
• •
4.
4.
4.
10.
o
• •
4.
11.
o
o
o
o
SECTIOI 4 - LIST OF PAPKRS • • • • • • • • • • •
Computation ot Coettic~t ot Correlation. • •
Croasing Point. - Rate Selection and
CCllpariaoD • • • • • • • • • • • • • • • • •
Electric Water Beatin, • • • • • • • • • • • •
MeetiDg AssumptioD8 ot Homogeneit7 of
Vari. .ea lIonliaear Data Tranatora.
and Bartlett'. Teat• • • • • • • • • • • • •
A General Panotion Subproll'_. • • • • • • • •
FORTRAN Programs tor CalpUting Elliptic
Integrals and FunctiCD8 • • • • • • • • • • •
The Us. ot Discrete ConvolutiOil in the
1n~si.
A
CCIIlpQter
Ana~8i8
(~\\
or Diffractions
• • • • • • • • • •
SpectrtDll
the Simultaneous Equations
Approach to O. .a
B7
Method • • • • • • • • • • • • • • • • • • •
ZIP - .l Multiple Line Keyword in
Context Program. • • • • • • • • • • • •
Eleotric Load Statistics Syatea for an
Ilectric Utility • • • • • • • • • • • -.
A Work Pertormance Monitoring &.retem
Designed tor an Electric utility Meter
ad _Installation Department. • • • • • •
A Multiple-Iteration Procedure tor
- Engineering Design • • • • • • • • • • •
• •
An
4.
3.
o
• •
4.
12.
A Distribution reeder and Substation Load
roree.sting
Procesaillg 0,,]) Surn,. Data on a Small
s.r.tea • • • • • • • • • • • • •
4.
13.
Computer • • • • • • • • • • • • • • • • • •
Plot Boutae tor BeE Load-and-Go FORTRAN • • •
Automatic Scheduling and Registration ill
4.
4.
14.
lS.
a Saall Colle ge • • • • • • • • • • • • • • •
AnT SPS with Simulated THF, DS and
h.
16.
o
MF Instructions. • • • • • • • • • • • •••
h.
17.
18.
19.
20.
21.
22.
o
lei High Speed SPS A8se.bler • • • • • • • • •
DOODLE - A Do-It-Your••lt Frobl.. Sol...r • • •
PDQ lORTRAR tor Paper Tape 1620. • • • • • • •
SI~' PORTRAH tor Paper Tape 1620. • • • • • • •
!he Art ot Debugging • • • • • • • • • • • • •
(SJMP) Seareh and M8Ilor.r PriIlt • • • • • • • •
S7abol Table Punch Prograu tor IS( UTO
o
!!I!
ad PDQ FCETRAN S7steae • • • • • • • • • • •
Monitor Superrlaor tor 1620-1311 or
4.
4.
4.
4.
4.
4.
23.
o
o
o
o
o
o
4.- --24.--
o
o -
4.
o
Sapermonitor • • • • • • • • • • • • • • • •
U.. of Computers in Design of Eleotronic
lCCllCLJlREIlt. • • • • • • • • • • • • • • • • •
o
26.
2
--------_..__...._._----------------_._...__._-_._---_._-_.-_._-.._.__ ..._-----_. -.... _..... _..__ ....... .......__................. _-...
_
------~~-~~-
()
TABLE OF OOHTENTS
c'
(Continued)
Sub
Section Section
SCTIOR
k
(Continued)
..t.work J.na17ei8 on the 1620 • • • • • • • • •
Ablltract s "All Experimental Personalized
Arra-r Tranelator System" by
H. Hellerman. • • • • • • • • • • •
Abstraots "Debugging in the POR n System"
by R. D. Burgess. • • • • • • • ••
SECTICII
S-
APPENDICES • • • • • • • • • • • • •
Registration List. • • • • • • • • • • • • • •
List ot Programs Teams Chairmen • • • • • • • •
Report of Meeting of Progr_s Team8Chairlllen ••
Surny of Type 1620 Progr8llDlling S7stema • • • •
(
''''''..""
~~/
l!I!
4•
27.
o
4.
4.
28.
o
29.
o
s.
s.
s.
s.
o.
o
1.
1
1
,.
2.
3.
4.
1
1
c
SECTION 1
GENERAL
Table of Content.
Program Agenda. • • • • • • • • • • • • • • • • • • • • 1.1.1
S1Dopsis of Papers. • • • • • • • • • • • • • • • • • • 1.2.1
Connents of- the Program Chairman • • • • • • • • ••• • 1.3.1
So.. Rotes on ·Sur.ey List of Programming
S78t... n and Panel Discussions • • • • • • • • • • ,1.4.1
1.0.0
~,
I'L,.,;;'
c
M'II!
'
'flM'!!Rf
run'N'N'l'" . ...,
"I'Ll'"
,
EASTERN REGION 1620 USERS GROUP MEETDlG
MARRIOTT MOTEL, WASHINGTON, D.C. - MAY 6-8, 1964
PROGRAM AGENDA
(D. D. Williams, Program Chairman)
Tuesdalz Mal ,th
6:00 - 8:00 P.M.
8:00
Registration
Sound-o.rr Session (Hardware, Programming Systems)
Wednesdazz Mal 6th
, 8roa A.M.
Registration
Note: WednesdaT was devoted exclusively to Programming
Workshops and special papers. There were four
(4) concurrent meetings, labeled A, B, C, and 0,
during certain time periods, labeled 1 through S.
This cross reference is utilized in the Synopsis
ot Papers.
(U~\
Summary or
A
9:00 - 10:30
(1)
FORTRAN P/S
Summary or
SPS
pis
B
D
Introd. to
Monitor
(Section I)
10:30 - 10:45
corree
10:4, - 12:15
(2)
DOODLE, A
FORTRAN
Language
High-Speed
SPS Assembler
12:1, - 1:4,
Luncheon (InfonMal)
1:45 - 3:1.5
(3)
Experimental
Translator
System
C
Introd. to
Monitor
(Section II)
Advanced
Monitor
(Section I)
Comparison of
SPS Systems
FORTRAN II
Workshop
(Section I)
PDQ FORTRAN -
0
Paper Tape
"Select External"
FORTRAN
1.1.1
5
Page 2
-
3 :15 - 3 :30
Advanoed
Monitor
(Seotion II)
(4)
(5)
D
Corree
3:30 - 5:00
8:00 - 10:00
-C
B
A
Aids to De bugging SAMP
Monitor Supervision for
Open Shop
Users
FORTRAN II
Workshop
(Section II)
Comparison ot
FORTRAN
Systems
Thursday, May 7th
8:00 A.M.
New Users Breakfast Meeting
Jim
Oliver
Registration
Note:
Thursday morning and Friday
afternoon comprise the
General Session. Program
Teams meet in the interim.
The Plotting and Eleotronics
Programs Teams were fonned
at this meeting.
9:00 - 10:30
General Session
...-""
/'''1'
Don Williams
Tom Wagner
Arnold Spitalny,
Marty Goldberg
Announoements
New Equipment Announcements
Systems Design in Eleotronic Industr.y
10:30 - 10.4,
Cofr,e
10
Comparilon in Depth of Naw 1I'ORTRlN Srateml
.4, -
12 .1;
1211, - 1.4,
Don Jardine,
Jan Lea
Luncheon
K~ynote
1:45 - 3 :15
\
Address t
Jim Oliver
Consolidation through
Systems ]ntegration
Program Team Workshops
Note I
4 - Lilt ot Papers tor papers presented or topics
treated.. '1'e. Chairaen are
See Seotion
l:Llted next to workshop.
A.
B.
C.
D.
E.
Frank Wella
N. Goldman
Eleotric utilitr
Eduoational Institutions
Rate Engineering
Ohemical Engineering
Plotting
K. W. Brady
r. Schroedel Aoting tor
T. H. Korall tl
TOM Soott
1.1.2
C.
6
C
. ~. "
I
ll"
I
HI!
'I!
VI"
I
'!I"
I
!'U'!!,!
wb##HH*¥#rt'
¥
re
Page 3
C
3:1$ - 3:30
Coftee
3:30 - ,:00
Continuation of Fir.t Four Workshops
E. Electronics
Arnold Spit.t11ny
,:)0 - 6:30
Social Hour
7:30
Movie on Type 360 System
IB1
8:00
Executive Board Meeting
Jim Davidson
Program Team Workshops
A. Electrie Utility
B. 1710 Users
C'. Structural and Civil Engineering
D. Education Institutions
Frank \ve lIt-;
J. J. Owen
Fr1dazz Mal 8th
9:00 - 10:30
C:\
10:30 - 10:h5
Coffee
10:u, - 12:1,
A.
B.
C.
12:15 - 1:4,
1:4, - 3:1,
Operations Researeh
Continuation of 9:00 - 10:)0
Continuation of 9:00 - 10:30
Tom Scott
N. Goldman
Jim Oliver
Luncheon
A'. . Electric Utility
B. General Data Processing
C.
Statistics and Mathematics
3:15 - 3:30
Coftee
3:30
General Session
Discussion ot Sound-ort Session
IBM Reports
Frank Wells
Robert Soucy Acting for
Arnold Canner
Frank Dickinson
Jim Ol:f_ver
Concluding Remarks
o
7
SYNOPSIS OF PAPERS
The sessions presented are summarized as follows: Cross Reference Number,
Title, Author (where applicable), Synopsis, and level of paper or workshop.
WednesdayzMal 6th
.1-1
Sunma of FORTRAN P Sand SPS P S - D. T. Nort
Camera and Instrument Company and E. Sinanian,
Fairchild
o rpo ration
A survey was presented by lIM and User personnel of the
major Il){ and user designed statement and symbolic language programming systems together with the additions and
modifications available from the librar,y.
Level: Intenn.ediate
B-1
Introduction to Mon!tor Programming - E. Sinanian and J. Grant I
lEM Corporation
A general discussion for new users of disk drives.
Section I di.scussed the Monitor in general and SPS
II-D in particular.
Level: Elementar,y
.1-2
DOODLE - A Do-It-Yourself Problem Solver - K. V. Farina a
General Electric Compagr
A FORTRAN programming system for 40K which interprets
and executes.
Level: Intermediate
AFIT SPS with Simulated TNFl'NS and MF - D. Olson, Newark College
or Engineerins
Level t' Intermediate
SPa High Steed Alsembler r,ar 20K - IC. Germann and G. Rumrill,
Rewarkljoi
age
ot Insi neering
Level:
B-2
Intermediate
Introduction to MoDi tor Programming - IEM COrporation
Continued from previous s'easlon.
in Section II.
Emphasis on FORTRAN II-D
()
Level t Advanoed
1.2.1
8
. p.o.'}",.
" Pfl",-j"bitbrlririt···dH
Page 2
Wednesdal, Mar 6th - Cont.
A-3
(cent.)
Q.
Level.
Boekhott
Intermediate
- P. G. Boekhott General American
A programming system based on PDQ FORTRAN embodying a
new I/O approach.
.
IAavell
B-3
Intermediate
Advanoed Monitor Programming - J. Grant, IB-t Corporation
A detailed discussion ot Monitor I tor the advanced
user.
Level:
C-3
Advanced
k Comparison ot SPS Systems - J. A. N. Lee and R. L. Pratt
evaluation and comparison in depth of the first
generation programming systems.
An
Level:
D-3
Elementary to Advanced
FORTRAN II Workshop - R. D. Burgess, Mechanical Technology, Inc.
A tutorial session reported to have been highlY
successful at MTI.
Level:~
B-4
Intermediate
Advanced Monitor Programming - IEM Corporation
Continued from previous session • Monitor II was
emphasized in Section II.
Level:
c-4
Advanced
The Art of Debugging - E. J. Orth, Jr., Southern Services, Inc.
Level:
Intermediate
Search and Memory Print (SAMP) - J. M. Wolte, Brooklyn College
Level:
Elementary
FORTRAN Symbol Table Punch for IBMi UTe and PDQ FORTRAN Systems -
0 ,/:
"
R. C.
Irons,
u.s.
Raval School orn.tion Medicine
Level:
Intermediate
9
Wednesday, May 6th - Cont.
C-4
~ont.)
3
Monitor Superrlsor for the 1620-1311 - E. E. Newman, Massachusetts
Institute
Technology
or
LeTel:
D-4
Page
Intermediate
FORTRAN II Workshop - R. D. Burgess, Mechanical Technology, Inc.
Continued from preTious session. Section II stressed
subroutine linkage and debugging.
A-S
Lee
and comparison in depth ot the first
generation programming systems, including IBM, AFIT,
An eTa1uation
UTO and PDQ FORTRAN.
Thursday, Mal 7th
9:00 - 10:)0
Welcome introduction and reports.
Comments on theme of meeting (systems) - D. D. Williams
Use of Computers In DeSign of Electronic Equipnent United lircraft corporation
A. SpftilDi;
Network Anagsis on the 1620 - M. Goldberg, United
Aircraft Corporation
Note: The preceding two papers support the
---- topic "Systems Design in Electronic
Industryll which is listed in the "Agenda".
10:45 - 12:15 An eTaluation and comparison in depth of the second generation
FORTRAN progrll1DUing systems including FORTRAN II and K1ngstran
FORTRAN by D. A. Jardine and J. !.N. Lee.
Papera clalled al .1ement~ presuPPole no background on the part at the
audienoe. Sellionl olassified as inter,mediate pre,uPPOle lome acquaintanoe
with the subjeot matter.
Sessions classified as advanced are intended tar persoDs with oonsiderable
experience in the field) questions ot an elementar.ynature will be discouraged at these sessiODs.
This agenda was supplemented between sessions by half-hour programmed
demonstrations on a type 1620 Model 2 computer with 40 K pOSitions of
memory, two disk units, a plotter and a card read-punch unit. The following
planned list ot demonstrations was altered slightly to accommodate available
hardware features:
Set AI
c·
Design Automation
Orbital TrajectolT Calculation
Similation of Analog Computer
1.2.3
10
I" ,inrH¥!!HH" 'FIW
ij'9
Page
Set BI
4
Monitor I
FORTRAN II
SPS III
Other Programs Available. COGO I
Circuit Analysi8
Electric Load Flow
Capital Investment.
FORTRAN with Format
Plotter Subrout1ne,
AnT
U.er des1r1n1 a speoifio QemOD,tration tram the
abow ,eleotion at times other thaD thole indioated,
or clelir:t.D1 to run &'lJ.7 ot hi, own prosram, va, asleed
to submit his request to tbl IBM representative at
thl demonltration oenter.
1I
Comments of the Program Chairman
The general response to your Program Committee's probe for
interesting material for the M'ay, 1964, Conference was quite gratifying.
Models of previous Proceedings, the wealth of papers offered
and the availability· and generous offers of speakers - expert in their
fields of interest, all helped to make the program planning easier.
The directive, for the best use ot the contributors' material
and talents, was taken from the Group's bylaws, and is repeated here:
"The primary object of the 1620 Users Group
is to advance the effectiveness of the
utilization of the IBM 1620 Data Processing
System • • • • • • • • "
In order to effectively advance this basic objective, the
Committee decided to build a program. blended with a theme which would
focus attention to the currently large stockpile of Programmirlg Systems
and Applications Programs. Thus, it was hoped to generate some constructive thought toward reduction of redundant programming efforts and
finally to atfect some integration of Application Programs into the
larger context of Data Processing Systems.
Hence, the them$, "Consolidation through Systems Integration"
is being highlighted.
The agenda vaa arranged 80 as to Mold the Conference into a
composite of two parts, each part ha.tng its own identity and flavored
by the theme.
Part I (Wednesday mornir.Lg to noon of Thursday) is mainly
identified with Programming SYltems. Wednesday' a tour sessions were
each of n1netr m1Dute duration, being held concurrent17 and being
three-deep. One ot these concurrent sessions i8 a panel discussion ot
SPS Systems which support the the•• relative to Part I.
Further support ot the theme, related to Part I, 1. i1'Yen in
two general session panel discussions cOTering the first and second '
gener'atioD of FORTRAN Syate.a.An enning session on Wednesday and one
on Thur8~ Morning campri.e the •• two 88ssiQD8.
Part II ('l'hUJ'1CIq morning to Friday atternoon) is m.aiDl,.
identified with AppUcationl Prograa8. The Programs Teams' activitie.,
are confined to this period. These meetings scheduled concurrentl1 are
four-d.eep on Thursdq afternoon and tl\ree-cteep during three •• ssions on
Friday. The ChaiNeD ot the various te_ bad been requested to limit
the scope ot their interest in programming to Applications Programs during
their teams' aessionl.
c
12
!lh1WijIP"UPP"i'iiIQ"IllTf3iri£ " "']i'HWiii'i"V
"
"n ..'"rril"t"
""J"""QT"WF"W¥'o/
The first aelsion at Thursd., morning - an overlap ot Part I
and Part II - il mentioned here aa an example at a tie in with the theme
by integrating aeveral A'pplicatioDs Programs into a Program STatem
(1.1. 'Spita1D7'1 paper, "System Dlsign in Electronic Industry").
The panel disculsion. previoull1 reterred tQ treat, in lame
depth, the lurvey lilt ot FORTRAN and SPS SYlteml whi~h are introduced
durinl the first leslion ot the Conference's opening. This lilt with
80me additions is reprinted in Section S of these Proceedingl.
Purther canment on the panels' aotirit7 will be tound in
the artiole following this one entitled "Some Notes on Sune;y List ot
Programming Systems and Panel Discussions".
.
The Procram Committee'a POlt-conferenoe activity ia confined to
the oompiling ot all the material in readinell tor prlntinl.
In oonclusion, the Committe. otterl lincere thank. to the m~
p.ople without Whol' alli.tanoe our million might not haTe been completed.
Donald D. Williaml
Program Chairman
John A. Rodgers
Program Secretary
1964 Spring Meeting
1620 Users Group
Eutem Region
1.3.2
J3
Some Notes on
"SUM'e), List ot Programming Systems·
and Panel Discussions
SurYel List of PrOgramming Systems
David T. Northrop devoted the entire opening session ot the
Conference to a review of a Sur.ey List of Programming Systems - both
IBM and User compiled - with a brief commentary relating to each listed
item. This session as planned served to introduce the subject matter
to be treated in depth during the panel discussions which were held
later.
SPS Panel Discussion
R. L. Pratt,· aided by Kurt Germann, otficiated at an
interesting panel discussion on SPS Systems. The Program Committee
is indebted to R. L. Pratt tor his post-editing of the SurY8y List
for reprinting in the Proceedings.
FORTRAN Panel Discussion
Report on Panel
User Written Compilers in Depth
D. A. Jardine
Brief mention was made ot IBM FORTRAN-Without-Format and
Format FORTRAN b7 J. A. N. Lee •
. R. L.Pratt described some ot the useful features· ot his
AFIT FORTRAN, such as compile time diagnostics and batch compiling.
AFIT FORTRAN, while an improvement OTer lIM systems, is somewhat slower
than other user-written compilers.
D. A. Jardine, in the absence of the authors, E. S. Lee and
J. A. Field J talked about UTa FORTRAN and the contributions made by this
system to improyed language and eaSier operation. UTO FORTRAN has few
compile t~e or object time error messages.
W. A. Burrows, in the absence of author F. H. Mask1ell,
described aca. of the teaturea of P.D.Q. FORTRAN. It 1s based on trl'O
FORTRAN with seTeral major improve.ente. It 1s the fastest running
program at object tillle tar mr 1620 not using hardware r.p. The 0011piler ia essentially the same as mo, but the object time subroutines
have been rewritten almost completelY.
c
1 I~
· R. L. Pratt then described' Auto-no at FORTRAN, deriT8d trom
AFIT FORTRAN, which oompiles hardware tloat-point add and subtract instructions in-line (not subroutined). It is probabl1 the tastest 20K
FORTRAN (at object time) currently available. Only limited FORMAT
capabilities are available.
C. H. Davidson talked about FORGO, the University ot Wisconsin's
load-and-go compiler tor teaching use. Extensive diagnostios at both
oompile and object time are available. 401 memoJ'l" is required.
Discussion trom the floor centered around details of operation,
use and construction or the various compilers.
o
1.4.2
--~~~--~-
..,.. ,"---.-,.,...... "." ...... -,"--.. -""",,,,,,,,
,~
i \
'~
SBC!I0I2
DDlO'l'B ADIRISS
!'able otConte.te
ConsolidatioD !hroqh S7ateu Iatesrat1ca
bT Dr. J. R. 011ftr • • • • • • • • • • • • • • • • • • 2.1.0
2.0,.0
1G
crnSOLIDATICN niROUGi SYSTEM) INTEGRATIGl
Dr. James R. Oliver, Dean
Graduate School and Director,
Computing Genter
University of Southwestern Louisiana,
Lafayette, Louis iana
I am pleased, as President of the Eastern Region 1620 Users Group,
to talk briefly with you concerning the theme of our meeting, "Consolidation
Throuah Systems Integration".
I trust that I will be able to leave you
with a thought or two today that might prove of benefit to you.
And
perhaps in so doing I will be able to help further the aims of our organization.
Few tems have as many definitions as the word "system". And I
say this even when talking to a group of people interested in
con~uting,
a field in which ordinary English words have frequently been given many
(~'"
new and sometimes surprising meanings I Isn't this a little strange when
we think of how precise our languages are for communicating with the
computer?
Al though "system" -has many meanings, including the people involved
in computing at an installation, I think we can profitably devote these
few minutes to hardware systems and software systems. Thus I will direct
my remarks along these areas of interest and will try to make three major
points which I believe to be of importance.
Many organizations have wide requirements as pertains to computer
usage- -business applications, management applications , scientific
applications, to mention a few--including many requirements which may not
yet be known.
Most of us would enjoy having our own private machines to work with,
o
but this is a luxury which few can afford.
I can speak fran the point
.---~--
..•..-.-~
..-....---,...-"".•..---..................
-2of view of the not-50-affluent Wliversity in saying that
~
certainly
cannot do so!
Perhaps you have heard of colleges and universities which have
many computers--even many large ones,
Unfortunately, these few cases are
overly publicized; the vast majority of educational institutions, however,'
are grasping and gasping for funds to even keep a computing center going.
As an example, which can be repeated over and over, in my own case I teach
two courses each semester, serve as Dean of the Graduate School, as
Director of all of our National Science Foundation activities on campus
and serve as Director of the Computing Center assisted by one full-time
unit record man and one half-time lady to help with computing activities.
Our main purpose for existence as educators in the computing field
is to serve the manufacturers and the users.
Relatively few of our
students are being educated for our own use.
But that is the subject of another talk.
Today we are here to
discuss a CODlJ1on problem--at least comon to most of us. We are faced with
a problem of maximum production for minimum cost.
Ih truth we very
practically are concerned with a problem in Operations Research - should
we suggest a computer solution to ,this problem?
Perhaps if we were able to put ourselves in the shoes of management
we would be better able to visualize the overall problem. But we have to
live with the fact that mOlt of us have to put up with a management which:
(1)
Cannot, or sometimes will not, comprehend computer concepts;
(2)
Often is in some way afraid of computer "take over", and
(3)
MUst be concerned with money justification to several levels-·
their superiors. shareholders, other members of organizatioo.
J8
C
-3-
Having acted and acting now on both sides of the proverbial fence,
(:.'!
I have seen sane of the problems faced by both sides. And I assure you
that the problems of the administrator are not trivial.
All too often we call upon the administrator to act in a capacity
for which he is not equipped. Can we expect a non-computer oriented
administrator to make decisions that we ourselves hesitate to make?
POINT 1: Are we making a sufficient effort in educating and in
keeping.wi4ft. our administrative staff infonned?
Nearly all of us sooner or later must discuss computer activities
with our colleges--if you haven't, don't despair, you willi
these colleagues are
non~canputer
Some of
oriented. Even worse, sane of them
are other-computer (particularly' other-concepts) oriented.
Here is one point that is certainly a most important consideration.
("
If the organization has a total of x dollars to spend on complters, and
one of our colleagues convinces management that y dollars should be
spent on a business use computer, it is sanewhat obvious that x - y
dollars will remain for our installation. Remember, because of the
non-scientific orientation of many in management, and because of the
strong arguments of persons in bUSiness applications, there will
frequently be sympathy for fragmentation.
Can't indeed two or more systems be consolidated into one? In
many cases, yes. 'We know that the 1620 is a very versatile machine.
If we inform our colleagues well of its capabilities, couldn't a Model II
1620, 40K, 2 disk drives and printer do most jobs as well as or better
than two systems costing more collectively?
o
We begin to see this idea more with the t«xiel 360. What are we
going to do about it? I can foresee all kinds of decisions to be made
19
-4-
with the 360. It is quite possible that we will be forced to resort to
c
the simulation of our computer installations to better detennine just how
they should be organized. I know of such a study made at Oak Ridge.
In the meantime, decisions with the 1620 still must be made.
Few
360's will be installed by most of us by two years from today. The 1620
is still a good machine and it
has much
wear left in it. Those who buy
or have bought 1620's have even a more serious problem. They will likely
keep in touch-with-the 1620 for many years to come.
POINT 2: Have we informed and communicated sufficiently well with
our colleagues who are, or are likely to be, interested with computer
activities?
Then we come to
sys~ems
as it applies to programming systems--mostly
referred to as software.
Unti 1 'now we have considered -two distinct divisions:
IBM and Users.
For some time now several people, including Jim Davidson, Charlie
Davidson, Don Jardine and others have been working to get better--no, to
get a--liaison between
~BM
and User-activities in this area of endeavors.
I would like to touch on two items:
(1)
Consultation with Users
when IBM decides to produce any kind of progranning package.
(2)
ConSUlta-
tion with IBM when Users decide to produce a significant programming
packaae.
Csn't IBM-use' our suggestions (1) because we know what we need and
(2) to utilize the ''brains'' of those in the Users Group? Also, why all the
duplicate effort? How many FORTRMs are needed? How many can we afford?
Shouldn't same of the effort, by Users in attempting to produce a FORTRAN
which will work as it should be devoted to experimenting with other
20
c
-5-
languages? Surely those Users who earlier today heard the presentation
4(;,
of the characteristics of an outstanding FORTRAN compiler produced by a
group of Users wonder why this has not already been available to us,
instead of some of the processors we have been forced to usel
We must not overlook the need for cooperation both ways. We
recognize that this brings up ideas not often
used--consultat~
IBM
when a User produces a significant proarammina packaae. At least three
reason. exist for this:
programminl effort,
or by IBM,
(3)
Perhaps IBM knowledge can help the
(1)
(2) We might avoid duplication of effort by Users
IBM could advise Users of impending changes to prevent
an effort being completely negated by new hardware or software being
developed. We owe our colleagues the production of worthwhile packages
if the Users' contributions to the Program Library is to mean anything.
POINT 3: Are we presently communicating among
ours~lvest
or with
IBM, sufficiently well to achieve consolidation through the integration
of programming systems?
will we fragment ourselves by writing an
Or
infinite number of compilers to the exclusion of better things to do?
Same of us are teaching--some are involved in computer applications.
A much smaller number is engaged in preparing software. Again I will
state that we are all concerned with getting the best production possible
fran us and our machines.
Aren't we charged with the obligation of doing best job possible?
Shouldn It we all-be concerned with the theme of our meeting--consolidation
through systems integration?
But I should
no~begin
winding up this talk.
o
21
.:!.1.5
-6-
A few years from now a uuch better talk than I can give will be
produced by a properly programmed computer.
c
The announcement will read,
"CUr program for this week will be, 'Consolidation Through Systems
Integration' by MJltivac 70809010 ••• ".
The canputer's speech will be really interesting.
coherence, and emphasis.
It will have Wlity.
It will meet all of the requirements of public
speaking as listeners like it.
The humor and jokes contained in the speech
will have a mathematical probability of at least 99\ of not falling flat.
Such a speech will not have any
te~s
in it that are strange to most of
the audience, so the audience will not get lost.
to all of Parkinson's Laws and MJrphy's Laws.
The talk will conform
At the same time it will
take into account what the audience already knows.
guilty of
unne~ssary
And it will not be
repetition of stupid cliches.
In fact, instead of a single speech to a whole audience, the canputer
will provide multiple simultaneous speeches for various segments of the
audience.
In froot of each member of the audience will be several small
dial switches.
(}le will be marked "Speed Control" with positions labeled
''You're going too fast" and ''You're going too slow".
Another switch will
be marked, "lbderstandina Control,", sayins "Yes, yes, I know all that.
pleu. come to the point" and "You're way over my head, please start over
and explain". A third Iwitch will be marked "Boredan Control" with
positions marked "You are beina mildly interesting--keep it up" and ''You
are borin. me to extinction·-please become lively".
I look forward to the time when a computer will be makina speeches
at meetings.
I think of all the poor speeches produced by human beings
that I have listened to and realiie that in the future. if I live long
22
-7-
enough, I will listen to 1Dliformly good, speeches produced by computers.
But 1Dltil that time you will be forced to listen to speakers like me.
I am able to recognize, however, that thOle of you who are still awake are
reaching for the switch .rked ''You've talked too ICl'lg--please shut up"l
So I'm going to pretend that you shut the power off.
o
23
J..I.1
_
- - - - - - - - - - ----------_.. ..__ ._-".. ,,_._.._..- - - - - - -
SECTION 3
PROGRAM TEAM SESSIONS
Table ot.Contents
Statistics and Mathematics Programs Te.
Attendance List. • • • • • • • • • • • • • • • • • 3.1.1
Report ot Activities • • • • • • • • • • • • • • • 3.1.2
Papers Presented • • • • • • • • • • • • • • • • • j .·1.2
Plotting ProE" Te8M
Attendance List. • • • • • • • • • • • • • • • •• 3.2.1
Report ot Activities • • • • • • • • • • • • • • • 3.2.2
Electronics PrOgrams Team
Attendance L1st. • • • • • • • • • • • • • • • • • 3.3.1
RePort ot Activities • • • • • • • • • • ••••• 3.3.2
Educ.ation Institutions Programs Team
Report ot ActiTities • • • • • • • • • • • • ••• 3.4.1
Papers Presented • ~ • • • • • • • • • • • • • • • 3.4.1
Structural' and Civil 'Insine.rins Proltas Te.
,
,
Attendanoe Lilt. • .'. • • • • •.• •• • • • • • • 3.;.1
Rlportot Aot,I't'1tll' • • • • • • • • • • • • • • ' '. 3.5.2
P.~r. Pr••• ft~cl • • • • • • • • • • • • • • • • • 3~S.2
.lttlnc!ano. Lilt. • • • • • • • • • • • • , • • • • 3.6.1
RepC)rt ot Acti"'1t1.~ • • • • • • • • • • • • • • • ,.~¥.2
P_petre Pre.~t.d··.'~ • • • • • • • • • • • • • • • ~:~,6 ..6
Cheaical BDline.rin, Pr~I~~! t~~
"J
~
" •.
~~~'.'~oration
ADJUS TMEN'!' OF SALES & REVENUES
FOR WEATHER AND TARIFlt' CHANGES
C
l', ,
"
..
......
PURPCBE OF THE PROGRAM
~.
In many cOJllpaniefJ the Rate Department is the group which provides
management with the theoretical figures that are reasonable answers to the "what
would happen if" quest.ions.
~here
is constant experinlent with ideas to provide
those answers trom the data Which 'are available.
A responsibility of our Rate &£ Economic Research Department is the
i
preparation of forecasts o~ customers, sales units and revenue dollars by months
tor seven different classes of 'business b.Y each type of gas (mixed and straight
natural).
The Rate Department, tor proper eValuation ot the forecasts I adjusts
the reported figures for prior years ,to the same standards which are being used
tor the forecast periods.
The 1964 li'orecast presented. special problems.
The
heating classes were
to be estimated, to the monthly requirements of a so-called "New Nonnal It
by the Weather Bureau.
&8
provided
The revenues were to be e~timated on the basis ot rate
schedules that were not yet in effect.
All of our
prio~
years data for normal
requirements were developed to meet the so-called "Old Normal".
The reverme data
had been produced under t~~ different sets of prior rate schedules.
There are various ways and means of determining theoretically What
revenues might have been if the weather had been 'normal, and there tlr,~ several
different 1r~ys ot calculating What re~enue might be on a proposed rate schedule.
Our Rate Department, la.te in 1963, began experimenting with
a mathematical method
ot developing these required answers monthly. Becaus e of the 'tremendous amount
of calculation required to test the formulas and to provide adjusted monthly- data
for several prior years for almost thirty different classes (b,y revenue code and
o
by rate code), and the need to provide t~ese statistics tor the forecast in a very,
3.8.7
Attachment #4
51
- 2 ,~
short period at tiJre J the problem ~as referred to the Data Processing Group.
job was completed by them trom start to finish in about a month's time.
~
The
The pro-
gram provided not only statistics for the 1964 Forecast but also provided data tor
the years 1960 to date on a truly comparable basis, that 18, all sales adjusted
to the requirements at the reNew Normal" and all revenues adjusted to theoretical
on the new rate schedules.
This program is now providing the Rate Department with
monthly sales and revenue figures for "normalized weather."
It is interesting to note here that for the first three months of this
year we have had, in total, almost normal heating requirements as related to our
billing cycles. With all the gyr:ations ot adjustments - plus and minus - the
deviation between the total revenues as reported for thdse three months and the
total revenues as developed for normal weather conditions is only 0.07% ot the
total.
This is certainly encouraging on the adjustnent tor the weather aspect
but we are still considering it to be an experiment.
In addition, the rate
sched.ule part must stand tho tests of time and statistical nethods.
'l'HEORY
There are two theories involved in this program.
adjustlDent of sales units tor weather.
One haa to do with
The other :Is concerned with adjustment
of revenues based on the revised sales units and/or a change in rate schedule.
c
At tachme nt l/4
3.8.8
52
OBJECTIVE OF 'lIFE PROORAH
The ol:>jective of this program is to accumulate the monthly
block analyses of the number of gas bills rendered and the associated consumption (in CCF) to an annual total within each block a.nd to a grand total for
all the blocks. A cumulative total is also calculated as each successive
block is summed. When the last bloc k i.s processed, the total number of bills
and total consumption is computed. F'rom the totals and the accumulated figures the consolidated factor is comf.?uted for each block. (See Appendix for
details.)
.
To facilitate checking, ·the total number of bills and consumptton is computed for each month R.nd visu.ally compared wi til the monthly
block a.nalysis figures already available.
T-he monthlY'data for bills and consumption is stored in
12 columns, one for each month. By summing the first figures 1n every
column, we obtain an annuAl total of bills or consumption for the first analysis. (See Appendix for' details.)
The prese~t monthly gas analysis involves eight tables
representing our two service clas:sifications, our space heating customers
and the irreguls.r bills rendered. Because part of Consolidated Edison's .
service area is subject to a 1% utiiity franchise tax, separate statistics
s.re maintained on each table for the system and the area for which the tax
is not applicable. For the sixteen required tables, whose number of bloc ks
vary from 47 to 57, the total running time on the l620I, exclusive of listing punched output, isapproxim.q.tely one hour. This compares with a manual
time of approximately 115 man-hours.
In the following appendix, the computer applicat.ion is
covered more technically.
o
Attachment #5
3.8.9
Attendants B.t Sessions of
Operations Research Programs Team
NW11ber Attending - 71'
A. H. Best
Jane Bonnette
Robert Burgess
William F. Bu.rl~igh
Harold Chancey
S. A. Clark
R. B. Davis
Dr. Kurt Eisemann
Jerry Eisenburg
Katherine Finnegan
Donald Flagg
Vincent Gangi
J. W. Garzane 11
Hector Ge ls1
E. H. Gerrish
G. S. Haralampu
F. R. Henderson
Tom Hoke
J. C. Hubbard
Samir S. Husson
Clayton Jensen
S. Kellman
John Kent
Alton Kindred
David Knight
Ted Krenzer
Thomas Latterner
Ralph Lee
Judith Liebman
Joyce C. Little
D. A. Lloyd
J. Y. Louis
Gene R. Lowrimore
Patricia Lussow
Thomas Morrison
Bennett, Nunberg
Dr. James R• Oliver - Team Chainnan
Brother Edward O'Neill
David Pixley
Pate Pointer
w. W. Pleines
C. H. Remilen
Theodore Sabine
John W. Sawyer
Joseph Schnecker
!1aynard Shore
Luke Sparvero
Arnold Spitalny
James Stansbury
R. L. Storrer
Carlis Taylor
William Thompson
John Tiers
B. Wingersky
Joseph Wingert
George Woodruff
Lawrence Wright
Charles Yacku1ics
K. V. Zajic
H. E. Blaicher
Reginald Harling
Len Karr
Denise Knight
Greta Larson
A. L. Lipson
Wayne Meriwether
Norman Ogilvie
John Owens
Joan Silverman
Welborn Smith
Marilyn Wingers~
r ti5"#ii5¥t*t¢
it
1,
rW"t'WM!
M'
t
tt
k t
!
h t
"
t 't
".,1"
t
t
t
d "'
8 \
t
tztH
It
t \
t
t
.'ht
j"t"
f rti t tt t
11
,I!
t ",
t"
t tt±
t" t
b
' .. ,' ~port, o~. , Activities of
.Oper.tions:. Re~earch Programs Team
The Operations Research Programs Team meeting met with
Dr. James R. Oliver serVing as Chaiman. There were approximately
75 persons in attendance.
Two papers were presented in their entirety: EMAGE - An
Economic Matrix Generator for a Distribution Transformer Management
Program - by W. W. Pleines, Baltimore Gas and Electric CompanY' (abst.ract
attached), and a Multiple-Iteration Procedure for Engineering Design by M. L. Baxter, Jr., Gleason Works (presented b.1 Theodore J. Krenzer of
Gleason Works).
Because of a time shortage, a third paper b.y Dr. James R. Oliver,
University of, Southwestern Louisiana, Lafayette, was presented in an
abbreviated torm. His paper was entitled, tilt Random Number Technique
tor the Study of Machine Breakdown and Repair". The technique used a
random nu~ber generator to produce a sequence of random numbers used
to determine whether or not breakdown would occur. During the study
the probability of breakdown was varied over a considerable range. If
breakdown occurred, a test was made to determine whether or not machines
were waiting to be repaired. If they were not, the machine undergoing
breakdown would begin to be serviqed. If other machines were waiting,
it was simply added to a line to be serviced as time became available.
If no breakdown oocurred, 8. oheck was made to determine whether machines
were waiting to be repaired. It so, work continued on them} if not, the
repairmen were then idle. In the case of machines waiting to be repaired,
a summation was made of the idle machine hours. When no maohines were
being repaired, a summation was made of idle man hours. The values re'corded by the program were the number ot time units used in the experiment,
number of breakdowns occurring during the time period, number of time
units the repair personnel were not engaged, and the numbe.r of time units
(accumulated) during which machines were not in operation.
Dr. Oliver pointed out an interesting but as yet undetermined
r@sult.. When the probability, in per cent, ot brea.kdO'taJ. multiplied by
the time units needed for repair equal approximately 100, a sudden discontinuity was observed in the curve representing the idle machine hours
versus t.iJne and the idle man hours, versus time. Also, a sudden growth
of the residual line occurred under similar conditions.
The speaker made it plain that the work reported was of a
prelilrlinary nature and that additional investigation was being continued.
In.terested persons were encouraged to write for further information concerning this study.
55
d
t r ...
There being no further papers to present or business to be
transacted, the Chairman declared the meeting adjourned.
Dr. 'Jsme sR. Oliver, Chairman
Operations Research Programs Team
Papers Prese'nted At This Session
Which Are Reprinted in Section 4
A Multiple~Iteration Prooedure for Engineering Design - M. L. Baxter, Jr.
dleason Works
..
3.9.3
c
-
••w
_
OPl'IMUM REPLACEMENT OF CAPACITY-LIMITED EQUIPMENT
w.
W. PLEINES
This paper explains the development ot a cost-based model for
transformer replacement which can be adapted for the replacement of
capacity-limited equipment in general.
The model, used as a component
of the Baltimore Gas and Electric Company's Transformer Load Mana,ement
System, is contained in a three-part program for the IBM 1620 computer.
Using fixed and demand-dependent costs tor all sizes and types of transformer in plac., it generates opportunity-loss matrices trom which it
determines, for each po.saible demand, whether a given unit should be
lett in place or changed to a larger or smaller unit of standard size.
c
At"tendants at Sessions of
General Ds.ta Processing Programs Team
Number Att.en.ding -
w.
24
F. Burleigh
S. Arnold
T. L. Drisooll
D. D. Williams
R. H. Davis
R. B. Riley
N. Ogilvie
R. E. Raver
G. M. Baber
D. H. Me.oLeod
K. R. Berger
J. P. Houlihan
E. M. Hamilton
R. A. Cooper
L. Schiffman
A. R. Kindred
Tfl. P. Wynn
M. N. Shore
/,-"'
"
l~
......:1
A. Spita.lny
J. Stansbury'
P. Lussow
E. H. Parker
S. W. Cr:tig
R. Souoy - Aoting Team Chairmarl
3.10.1
58
·-PiFMi ...... "q"'{" irllti!t!tlWH'''f-'!'''-'7''!!
Cj"
Jo
- 00- ..'····-
'-'-rp'U"'W""jR
Report of Activities of
General Data PrccesRing Programs Team
The Programs. Team Meeting fOl' General
to order by temporary Chairman, Robert Soucy •
D~ta
Processing
was
called
Following the theme of the May Meeting, three special session
program papers were delivered at the meeting. They were:
a.
"Electric Load Sta.tistic System" by E. H. Parker,
Baltimore Gas and Electric Company.
b.
II A Work Performance Monitoring System "by D. D. Williams,
Baltimore Gas and Electric Company.
Both of these papers clearly indicate how the integration of
data can be a source of valuable management reports and operating information.
A sample of the management reports produced by these programs
were distributed to those attending the meeting.
C'I
o.
"ZIP - A Multiple Line Keyword in Context Program" by
T. J. Scott, Sun Oil Company.
Mr. Scott's program is designed to perform a Keyword (or Key-phrase) in context analysis on a multiple line or card abstract, paragraph
or maga.zine article.
An outline of Mr. Scott's paper and a sample of "Keyword Phrases
in Context" output list was distributed to those attending the meeting.
R. Soucy, Acting Chairman
General Data Processing Programs Team
Papers Presented At This Session
Which Are Reprinted in Section 4
o
ZIP - A Multiple Line Keyword in Context PrOp-lint - T. J. Scott, Sun Oil
Company
Meter
3.10.2
59
An Eleotr.ic Load Sta.tist-ics S stem ror an Eleot'ric Utilit"\ - E. H. Parker,
Baltim'Jre G~.:4 &l~ ~. eotr 0 Company
c
3.10.3
60
Attendants at ·Sessions of
1710 Users Sub-Group
Number Attending - 17
J. J. Owen - Team Chairman
Patricia Lussow
G. F. Hamilton
A. J. Palmieri
A. C. Bailey
F.Wang
T. E. Morrison
s.
Jurnack
E. B. Svihla
Carol Braun
W. H. Beck
R. L. Storrer
J. W. Garzanelli
D. Pearce
S. Lopes
R. P. Paterniti
L. Stewart
o
3.11.1
61
Report of Activities of
1710 Users Sub-Group
Mr. J. J. Owen, Chainnan for the 1710 Users Sub-Group opened
the meeting with a welcome to all present, and a call for additional
papers at the next meeting.
There were approximately 30 people in the meeting at various
times although only 14 users and 3 IBM personnel signed the attendance
roster. A list of those signing is attached. It is assumed that these
are the people who are most interested in the activities of the Group.
It was announced that plans for the next meeting included a
discussion of 1710 Executive systems by IBM specialists.
The program tor this session consisted or two integrated presentations coyering aspects ot system planning. The first paper entitled
"Installation Planning tor an IBM 1710 System" by Mr. Lawson E. Stewart
ot the IBM Corporation, Baltimore, Maryland, covered System Planning and
1710 Installation Implementation. Mr. Stewart covered the general subject of project organization and planning. He began with a suggested
outline of broad goals for the project;
1.
Effective translation of Management plans into
a working system.
2. Education for future projects.
3. Meeting of project goals and deadlines with
least effort and cost.
With respect to the first point, Mr. Stewart emphasized -t.b&most
of the less than successtul computer control installations throughout the
industr,y were the result of improper definition and understanding of project goals. To prev~nt this occ~e, Mr. Stewart recommended direct
involvg nt of top management in the project via an initial report covering
goals and intended methods of achieving them, plus periodic reports ot
progress and plan modifications.
Education for future projects, reports Mr. Stewart, stems from
a conscious effort to man the project with this view and trom a comprehensive
system of documentation which covers not only programming but physical
planning and all other aspects of the project.
Meeting of goals and deadlines requires careful coordination
and control. The most important aspects ot this would be selection or
proper personnel tor each job function on the project team, assurance of
complete canmunication between team members, and use ot such planning
tools as critical path analysis. On this last point, Mr. Stewart dis-
c
3.11.2
62
( '"
~,,,
played a portion of a critical path schedule prepared using the IBM LESS
program. He emphasized the need to include all portions of the projects
in the t~~e estimates and job descriptions.
I
Following this portion of the discussion, Mr. Stewart presented
outlines of the functions to be considered in any computer control application:.
For Known Applications
1. Establish specific goals for this system, long range and short range.
What is to be accomplished, by when? What would be the ultimate
extension of this application?
2.
Organize the basic project team to include these functions or abilities Project leadership - a leader and administrator, not just a technical
specialist. Process knowledge - practical experience plus sound knowledge of fundamentals. Instrumentation - knowledge of measurement
techniques and signal treatment. Programming - math and system knowledge.
Installation - field installation experience and checkout techniques.
Documentation - probabl)" a fllll time job for one person, to maintain
documentation standards and assure the satisfactory preparation of
all reportsand final docUmentation.
3. Prepare critical path schedule for total project: Takes about three
("
;L"
weeks to complete the first time.
4.
,.
Establish a list of reports and calculations to be performed.
Generate a sensor list.
6. Analyze accuracy requirements for all measurements, so that high
precision instruments are used onl)" where needed.
7~
Follow a formal schedule of project meetings and benchmark progress
reports.
8. Check and plan completelY the ph)"sical installation.
9. Prepare program flow-special emphasis here on the use or Decision
Tables for both program logic planning tool and as program dooumentation.
10. Program checking by simUlation or use of a 1710 system.
11. Conduct electrical noise tests at the installation site.
12. Design the operator's console. This was considered a vital point
from the view that poor design would reduce operator cooperation and
system effeotiveness.
13. Operator training in use ot the console and computer program is a
must for suocessful operation and application growth.
3.11.3
6 ..
)
oj
v~.
Model Building Tests are required with just about any computer control application. This phase should be under the direction of an
expert in this particular work. This man norma1~ has as a backgroIDld, praotical eXperience with the operation of the speoific
process, knOW'ledge of statistical techniques and high-precision
measurement.
15.
Document all phases of project as you go, and prepare progress reports which serve to inform top management as well as continually
selling the advantages of the projec·t,. The greatest failing in
project teams is the lack of understanding that a project must be
contin~ally "sold" to top .management in order to keep interest and
support for the project.
For New Applications
List included only additional requirements above those for a
known application.
1.
Study the basic process. Accumulate all known information on theory
and practical operation.
2.
Run manual tests to establish important process parameters.
3. Establish, by data analysis, potential control areas.
4.
5.
Analyze measurement and control practicality.
Install system for data acquisition and reduction.
6. Develop math model of the proce s s and verify.
7.
Proceed as with known application.
Mr. Stewart concluded with a plea for more presentations in
this area as more 1710 systems go into operation.
The second paper of the session entitled "Some EarlY Experiences
in the Design of a System for an Electric Utility Power Plant by Mr. William H.
Beck of the Baltimore Gas and Electric Company introduced specific material
used in the documentation and planning of a 1710 System for the Herbert A.
Wagner Electrical Generating Plant of the Baltimore Gas and Electric Company.
Mr. Beck presented copies of the documentation and programming standards
adopted by his company for the 1710 installation project. In the discussion
of the attached standards, Mr. Beck emphasized the fact that a primar,y goal
of the 1710 installation was the acquisition of knowledge and education in
computer control of power plants. For this reason, additional care is going
into the development and following of standards for documentation of programs,
project planning, writing of reports and logging of data on effort and time
required far each aspect of the entire project.
As a specific example of the planning outlined in Mr. Stewart's
3.11.4
h
C"
#
ttdt
rt rt
;1
*
t
51
t
t
discussion, Mr. Beck's presentation was considered by" those present to
be extremelY valuable and exact~ the kind of specific, practical material
needed for other users about to embark upon an installation.
The meeting concluded atter a question and answer period which
covered the two presentations and additional material on measurements, etc.
J. J. Owen J Chaiman
1710 Users Sub-Group
-Note:
(
Formal copies of the papers by" Mr. Stewart
and Mr. Beck are not available for reprinting
within these "Proceedings".
\
,~
o
3.11.,
65
jl~
I
SECTION
4
LIST OF PAPERS
'C
Table ot Contents
"Computation of Coeffioient ot Correlation"
by R. C. Burrowes and J. E. Enterkin New England Power Sernc. COIIlP&n7 • • • • • • • • • • • • • •
"Crossing Points - Rate Selection and Capari8on"
by H. L. Fitch - Long Island Lighting Company •
"Electric Water Heating" by K. W. Brad7 Long Island Lighting Comp~ • • • • • • • • •
4.1.0
• • • • • • • 4.2.0
• • • • • • • • 4.3.0
"Meeting Assumptions ot Homogeneity of Variance:
Nonlinear Data Transtorms and Bartlett's Test"
by R. C. Irona .. U.S. Naval School ot Anation
Medicine. • • • • • • • • • • • • • • • • • • • • • • • • • •
"A General Function Subprogramn by M. E. Munroe -
UniTersity ot Hew Hampshire • • • • • • • • • • • • •
••
4.4.0
• • 4.,.0
"FORTRAN Prograu tor CC*lputing Elliptic Integrals
and Functions" b7 H. E. Fettis and J. C. Caalia Applied Mathematics Research LaboratOlT •• ' • • • • • • • • •
4.6.0
"The Use ot Di.orete CODYolution in the Analrsis
ot DiffractioDs n by' J. S. Lielxaan and
p. M. Johnson .. General Electric Comp~ • • • • • • • • • • • 4.7.0
"A Computer Approach to Gamma Spectrum AnalTSi.
By the Simultaneous Equations Method" b7
W. E. Carr - U.S. Department otBealth, Education
SUld We]Jr.ure • • • • • • • • • • • • • ~ • • • • •
••
• • • •
4.8.0
"ZIP - A Multiple Line Keyword in Context Progr8JI"
by T. J. Scott - Sun Oil Coap~ • • • • • • • • • • • • • • • 4.9.0
"An Electric Load Statistics System tor an Electrio
UtilitY'" by E. B. Parker - Baltimore Gas and
Electric Comp~. • • • • • • • • • • • • • • • • • • • • • • 4.10.0
"A Work Performance Monitoring System Designed tar
an Electric Utility Meter and Installation
Department" by D. D. Willi __ - Baltimore Gas
and Electric Company. • • • • • • • • • • • • • • • • • • • •
4.11.0
4.0.0
c
- ··Z·'-·· ..
T
'5"-
n
· _ f __
n.
)-:._'-_.
y."
"-e-'··uY"·rr··Y·'u
'm--
- ' .. ·dW· .. ·fij·W ...·i6wrif·
d··
Wtfrii·tfR¥ii"Ft·" 6···w·*
.J,
U!!!'.
pl.
IW!"" i"!! . , , '
I''tdd*tW··
Table at Content.
(Continued)
"1 Multiple-Iteration Procedure tor EngineeriDg
Design" by H. L. Barter - Gleason WorD • • • • • • • • • • • 4.12.0
"A Distribution Feeder· and Substation Lom
Forecasting Sy-st_" by J. C. Hubbard Baltimore Gaa and Electric Coap8D7. • • • • •
• • • • • • • • 4.13.0
"Processing O-D SurYey Data on a Small Computer"
by A. D. Stasi and M. B. Lipetz - Edwards and
Kelcey, Inc • • • • • • • • • • • • • • • • • • • • •
"Plot Routine tor NCE Load-and-Go PORTRANft
by' H. Sevard - Newark College of Engineering. • • •
"Automatic Scheduling and Regi8tration in a
Small College" by A. r. Jackson - Agricultural
and Technical College ot North Carolina • • • • • •
WAF!T SPS with Simulated ~. TIS and MF Instructions"
by D. Oleon - Newark College of EDginaeriDg • • • •
"NCE High Speed SPS Ass••bler"
G. Rumrill - Newark College
• • • • • 4.14.0
• • • • •
4.1S.0
• • • • • 4.16.0
• • • • • 4.17.0
b7 IC. GerJll8DD and
ot EngiDeering•• • • • • • • •
e
4.18.0
"DOODLE - A Do-It-Yourself Proble. Solver" b,r
H. V. Farina - General Electric Colllp8.JV' • •
• • • • • • • • • 4.19.0
"PDQ FORTRAN tor Paper Tape 1620" by J. W. Trant1Dl
and P. G. Boekhott - General hariem Transportation Corporation (MRD Division) • • • • • • • • • • • • • 4.20.0
"SEX FORTRAN tor Paper Tape 1620" by' P. Q. Boekhott _
General American Transportation Corporation
(MRD Division). • • • • • • • • • • • • • • • • • •
• • • • • 4.21.0
"'rile Art ot De bugging" by' E. J. Orth, Jr _ _
Southern SerYices, Inc eo • • • • • • • • • • • • • •
• • • • • 4.22.0
" (SAMP) Search and Memorr Print" by J. M. Wolte Brookl1n Colle,e. • • • • • • • • • • • • • ••
• • • • •
•
•
"S7Jllbo1 Table Punch Programs tor 11M UTO and PDQ
FORTRAN SY8teJu" b7 R. C. Irons - U. S. NaTal
School of Aviation Medicine • • • • • • • • • • • • • • • • •
~onltor
SuperYisor tor 1620-1311 or Supermonitor"
bY' E. E. Newman - Cirtl Engineering Systems
Laboratory at M.I.T •••••••••••••••
o
4.24.0
• • • • • • • 4.2;.0
4.0.1
67
r
!"\Ht"~
Table of Contents
(Contiriued)
/(Ir,
or Computers in Design of Electronic Equiplent"
by A. SpitalJl7 - United Aircraft Corporation ••• • • • •
"Use
"Network Anal1'sis on the 1620" by' M. Goldberg United Aircraft Corporation • • • • • • • • • • •
\0
• • 4.26.0
• • • • • • 4.21.0
"An Experimental Personalized Array Translator
System" by H. Hellerman - 11M Corporation Watson Research Center • • • • • • • • • • • • • • • • • • • • 4.28.0
"Debugging in the FOR II System" by' R. D. Burgess Mechanical Technology Incorporated. • • • • • • •
• • • • • • 4.29.0
4.0.2
68
COMPUTATION OF COEFFICIENT OF CORRELATION
(SLMPL'E, LINEAR CORRELATION) BY THE
SHORT-CUT CRUDE DATA METHOD,
STRAIGHT LINE REGRESSION, AND
STANDARD ERROR OF ESTIMATE
R. C. Burro'Wes and J. E. Enterkin
New England Power Service Company
6D
o
If, i. 0
1620
NEES PROGRAM NUMBER
15
I
TITLE a
Computation of Coefficient of Correlation (simple, linear
correlation) by the short-cut Crude Data Method, Straight Line
Regression, and Standard Error of Estimate.
II
~~
R.C. Burrowes
III
SCOPEt
In common usage, the term "correlation" is used to mean the c(.\us~l
effect or co-relationship which exists between variables. The
degree to which variables are co-related or influence one another
is represented by a computed measure called the "coefficient of
correlation." Values of this coefficient range from zero (no
correlation at all) to approaching +1 or -1 (approaching perfect
positive or perfect negative correlation~.
an~
J.E. Enterkin, New England Power Service Co.
Positive correlation results when the movement of two series of
variables is closely related in such a way that they move together. An increase in one series accompanies an increase in the
other,. or a decrease in one series accompanies a decrease in the
other. (For example: the related movement of degree days and
heating fuel consumption) •.
Negati~e
correlation results when one variable increase as the
other decreases, such as KWH used and average cents per Ia'-lH.
This p~ogram will concern itself only with simple linear correlation;
the correlation of two variables (one of which is independent and
one dependent) to be described by a straight line regression.
Suffice to say that there are more elaborate methods Wlich exist
to determine the degree of correlation among more than two variables,
and for computing curved regression lines.
The method outlined in this program is applicable, subject to core
limitations, for any two series (one dependent and one independent)
that may logically be assumed to influence one another.
IV
TEST
-
DATA:
The following data (taken from a sample of New England Utilities~
will be utilized to illustrate· the 1620 Fortran program designed to
c~lcula~e the coe.rfi.cient of'cQrrelation, the "Least Squares"
s~~a.igh~ line·regression-that best.describes the relationshiPI and
1;,h.~ s:t.~Ildarderror of estimate which indicates the amount of
dispersion in t!le dependent variaQle, which we have failed to account
for by.the regression line.
x • independent variable
Ult.Customers per pole mile
12.
19.
21.
.36.
43.
49.
Y • dependent variable
KVA per line transformer
9.3
11.1
10.9
16.4
21.1
20.5
5s
60.
17.8
28.1
70.
44.5
t
65.
23.6
/r---.\
~'
ImtT
!·""FX·""iMthiH ·+"**i*r+Hwt"fiitfWFf"'"j"""""
1620
NEES PROGRAM NUMBER
30.1
105.
120.
183.
(:'
V
-2-
15
43.4
50.0
REFERENCES:
I..
'Formulas
1. Coefficient of correlation
where:
• coeff. of correlation
• sum of
• sum of the individual
deviations from their
• sum of the squares of
mean of X
. • sum of the squares of
mean of Y
The short-cut method utilizes the
lined in appendix D ) to arrive
of the deviations of XJ Y and the
deviations X times the deviations
products of each x and y
respective means •
the x deviations from the
the y deviations from the
"centering process" (as outat the sum of the squares
sum of the products of the
Y.
Linear !agression Equation
where::
T
~
~~
X
y
X.
Y
•
•
•
•
•
•
•
As a further
~y
computed trend of Y variable
same as defined above
same as defined above
arithmetic mean of X series
an thmetic mean of Y series
original values of independent variable
original values of dependent variable
short-cut~
•
•
l::XY
the two normal equations
Na +·b~X
a '2x + b ~X2
can be reduced. to
o
a
•
b
•
Y - bY (origin at X •
~
%
XI
0)
71
--~---"'"---"'-~-"""-'''''''''''''--"-'''.--
1620
NEES PROGRAM NUMBER
15
where:
a • the computed or trend value of Y when X • 0
b • the amount and direction of slope at the point
where X • 0
B.
BIBLIOGRAPHY
Croxton, F.E., and D.J. Cowden, Applied General Statistics,
Prentice-Hall Inc., New York, 1939
Davies, G.R., and" D. Yoder, Business Statistics,
John Wiley & Sons~ Inc., New-rork, 1949
Mills, F.C., Statistical Methods, Henr,y Holt & Co.,New York, 1938
VI
PROGRAl-l FORMAT:
A.
Input Data
The X and Y series may be of any length up to 366 pairs, but there must
be one X value for each Y value.
Two F 10.4 formats permit use of X and Y data of six whole numbers and
four decimals. However, machine storage capacity and output formats may
restrict the use of long series of large digit numbers.
The first entry card must be an .tItem" card showing the number of pairs
of X and Y variables in the series to follow. This item number must be
punched with its units position in Col. 3. (see appendix A).
X, Y data cards follow, and are entered in the following sequence:
X 1
X 2
Y 1
Y 2
Card 1
Card 2
If
'ttl"
X n
Y n
Card n
with the X 1 value punched inools. 1 through 10 (the c01s.1,8,9,10
representing four pl~ces to the right of the dec~ma1 point) and the
corresponding Y 1 value punched in cols. 11 through 20 (the co1s.17,18,
19,20 representing four places to the r~ght of the deoimal point).
Each card contains an X value and its corresponding Y value.
72
~.l3
-'%.""" ... -
1620
NEES PROGRAM NUMBER 1,
.
-
.. -
-
T···Y w
···!
*-
·=·-1 .. ·...
-4-
An attempt has been made to make the 1'ormats as broad as possible J
but on occasion we have experienced an overtlow in printing the
sum of the squares of the y deviations from the mean of Y (symbolically ~2)
coded as SOYSQ and printed in output as SUM DEV Y SQ. When this overtlow
occured J the number was printed 10 "floatlog po1nt." Th1a "r-B overflow
error" did not 1nval1date the computations, it mere~ lb1tted the mode
ot output tor this particular value trom tlxad point to floatine po1nt.
B.
Symbols Used in the Prosram are,
(See Append.1x B)
ITEM • the number of pairs of X and Y items to be correlated
.' 2:x • total of X aerie a
• 2!. Y • total of Y series
• ~ X2 • sum ot sqUires of each X value
• %y2 • sum of squares of each Y value
• ~XY • sum of the product of each X value multiplied by
i ts'Cotr.esponding Y value
AVOX.
X • arithmetic mean of the X series
AVOY.. Y • arithmetic mean of the Y series
SOXSQ • ~x2 • sum of the squares of the deviations of each X value
SUMX
SUM!
SMSQX
SMSQY
SMPRO
I
from the mean of the X series.
SOYSQ • ~y2 • sum of the squares of the deViations of each Y value
from the mean of the Y series.
SOXY • ~xy • sum of the cross products'of the deviations of each X
value from the nean of X times the deviation of each Y
value from the mean o·f Y.
I'll
R. the coefficient ot correlation
T. the computed trend values of Y
~"
A. the Y intercept • Y- B X
B. the amount and direction of s lope at the point where
X • 0 ; or Zxy -;- ~x2
.
D. dev~ations of actual Y values from the computed trend values 01' Y•
• SMDSQ. ~D • sum of the squares 01' the deviations of actual Y values
from the' (xnuputed trend values of Y.
SUMT - sum of the trend values of Y
STDEE-
J::£
D2 - standard. error ot e stim te
COUNT
SQRTF - square root
READ PUNCH PRINTCOUNT C.
o
read via punched card
punch output on card
print output on typewriter
ITEM
OUTPUT DATA
(See Appendix C)
During the course of the program execution, the value of the coefficient
of correlation R, the trend equation constants A and B, and the count of
the number of pairs of X and Y data are printed on the typewriter as soon
as they are computed to provide immediate answers.
73
itr'M&bwdii6d'
"....,._
---_._-_._---
.•__....
1620
NEES PROGRAM NUMBER
15
-5-
c
Complete statistics are punched on cards for separate listing, and
these include the followingt
.
R • coef. of correlation
A • regression constant} Equation, Y • A + EX
B • regression constant
Count • no~ of pairs of I and Y data
List of X, Y, Trend Values, Deviations from Trend
Std. Error of Estimate
Mean of X series
Mean ot Y series
Sum of'Dev. ot X, squared
Sum ot Dev. of Y, squared
Sum of Dev. ot XI
Sum of Dev. trom Trend, squared
NOTEs'
----The standard error of esttm,te is a measure, in absolute terms, of the
variation of the actual values of the dependent variable (Y) tromtheir
estimated values. 'It would be expected that 68% ottha actual Y values
would fall within. 1 std. error est. of the regression line, 95% within
+ 2 std. error est: and 99.7% + 3 std. error of est.
-
-
ANALYSIS OF VARIANCE
c
The std. error of estimate, when squared, represents the amount of variance
unexplairiedb1 the regression line.
The SllDl DEV Y SQ divided b1 Count or
(tt
( N
) represents total variance.
)
Total variance minus unexplained variance ,. explained variance.
The explained and unexplained variance can be turned into %of total variance for
quantitative analysis ot the success of predicting from the regression equation.
The square of the coett. of correlation also equals the % of explained variance.
Although the analysis ot variance baa not been programmed, the Mcessar)"
components are read1lr ava11able and the hand calculation is 10 minor that it wa.
thought to b, worth :1nclu.41nI to round' out the, correlat1on analYI:La.
JD/mb
3-11-64
c
4.I.d
!!Ill"
..
APPENDIX A
ILLUSTRATION OF INPUT CARDS
sooooo
/
DATA DECK -
Y13
LA~T
CARD
00000011110000011111000000000000000000000000000000000000000000000000000000000000
I 1 l • , , , • , I' " U U •• 15 '1 II ••• , 21 ,. n II .. IS • " ]• .It 10 JI Jll3 l4 JllUI lilt ...1 .2tu••, ... , ..., 5UI N ill M " M " 5UI .. II U ., 54 .~ " fl WI ., 11 71 I: I] a I) ;6 1/ " 1, a,:
I I 111 I I I I 1 1 I I I I 1 1 11 I 1 1 1 1 I 11 1 I I 1 1 I I' I I 1 1 1 I 1 1 1 1 1 1 I 1 I 1 .1 1 I 1 1 1 1 1 1 1 I I 1 1 1 1 1 1 1 1 1 1 1 1 1 I 1 I 1 1
2222222222222222222222222222222222222222222222222222222222222222222222222222222~
33333133333333333333333333333333333333333333333333333333333333333333333333333333
,
.
444444444444°444444444444444444444444~4444444444444444444444444444444444444444444
5 5 5 5 5 5 5 5 5 5 5,5 5 515 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5·5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 55 5 5 5
6666666666666666&6666666&66666666&66666666666666666666666666666666666666666666&'
77777777777777777777777777117717111117111111717111111111171111771177777777777777
888aI888'1'8888'8a88'18888a88888'888'8~8888888888888a888888888.88888888888888888
99999999999999999999999999999999999999999999999999999999999999999999999999999999
I
'
Y1
,.
"I I ,
1/ I
DATA DECK - FIRST CARD
00000011110000000111000000000000000000000100000000000000000000000000000000000000
t I J 4 S I J • • II II II II I....,,' II It 21 II II II .. 'UU' 2Ilt • il JUU4 lUUJ atlt 40 •• 424144 4S ...7 .. 41101i IUIIUIII Sf .. H ...IIUl54 IS .. II .. H JO II " I) J. " II " '. I. 10 .
1111111111111111111111111111111111111111' 1111 t 1111111111111111 t 11111111.111111111
22222122222222222222222222222222222222222222222222222~ 22222222i22~22222222222222
333333333333333313333333'333333333,33333333333333333 3 33333333333333333333333333333
.44444444444444444444444,444444444444444444444444444444444444444444444444444444444
55555555555555555555555555555555155555~5555555555555555555555555555555555555555~
6666666666666.&661&&&666666&66&6666666~6666666666&66&666666666666666666666666666
7 7 1 7 11111 7 1 7 7 7 7 1 7 7 1 7 7 7 7 1 7 1111 7 1 7 1 7 7 1 7 7 1 7 7 7 7 1 7 7 7 7 7 1 7·1 7 11.11 7 1 7 1 7 111 7 11 7 7 7 7 7 7 7 7 7 7 1 7
88'888888888888888888888.8888888888888888888888888888.888888888888888888888888888
999999999999999199J99999999999"999999999999999999999899999996999999999~9999999999
4'
• 10 " I' 12 .. 1,1' If' II" 'I 22.n .,.
,
, 2' II 1031 l' U)4 25 21 3J JI 'U'U 41 U44 4~ 41 ., •• 41 "'51 N U ~ "'UI ShU.. II"U4 IS " " U "'0 II " 1.\ " I) )' " I. 1110
ITEM CARD. Showing that there are 13 pairs of X and Y
values inth1s input, ~Q tollow.
00000000000000000000000000000000000000000000000000000000000000000000000001,000000
I I J 4 S I , I I 10 II " 13 14 " I' " II IUUI n 'HUll'" lilt 10 21 » l3 JUI» II at • to 41 4UH441 .. 47 .. 4UIII U U M 'UI" II .. co IIIUI .. IS " " IU. 10 ,. " 'J I. I, " 11 II ,. 10
11111 t t 1111·' 111111111 t 111111111111111111·11111111' 11111111111111111111111111111 t 11
.
..
,
222212222222~}2222l2222222222222222222222222222222222222222222222222222222222222
i31333333333~3333333J3333333313~3~333333333333~3333333 33333333333333333i33333333
4,4444444444444444444444444444444444444444444444444444444444444444444444444444444
5$555555'55555555555555555555555555555555555555555555555555555555555555555555555S
6066~666&'&66i&I'I'&&I"""'II'I'&'&I&I&III'I"'&'I"."'16&&6666666666666666
711171777771.177771771117717171777717177777777777 7117 777777777777711777'1711711117
8888888888888.888 ••••••• 88 •••••••••• 8 ••••••••• 8 •••••• 1 ••••••• 1888888888888888888
9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 , •• I 9 • S • I • I 9 IIII 9 I t 9 9 t • 9 9 II • 19 9 9'i I 9 119 ••• ,1 t 9 9 9 9 9 9 9 I 9 I 9 9 9 9 9 9 9 9 9 .
I' I 4 I I ' " •• ".' .... 'illltJ ..... 'IIIl.,.,..l121I1.'I.'\'Jta.p'.....I.,.' ........... ,U•• luu....r»I11 ...'uU ....." ..... ttlll'J.""",." •
..ac
508'
.
" a~
"""'
1
·"Mti'_. i'"
.---.---~
"
APPENDn B
Ie
COEF. OF CORRELATION, STRAIGHT LINE REGRESSION AND STD.ERROR .,.
0'8300 C
R•BURRO~JES AND J. ENTERK IN
JANUARY 1964
...
08300 C
DIMENSION X(366), Y(366)
08300
10 SUMX a o.
08300
08324
SUMY - o.
08348
S~1SQX =0.
SMSQY CI~O.
08372 .
SMPRO' -0 •.
08396
READ 1 , I TE~1
08420
1 FORMAT (13)
08444
08466
. COUNT - ITEM
00100 ' 1=1 ,I TEM
58502
08514 100 READ 2, X(I), Y(I)
08634 .. 2 FORMAT (2Fl0.4)
08662
'. 00101 l.l~JTEM
SUMX .. - SUMX + X( I )
08674
'SUMY - SUMY + Y( I)
08734
SMSQX ... SMSQX + X(1}**2
08794
SMSQY = SMSQY + Y(I)**2
08866
101 SMPRO .. SMPRO + X(I)*Y(I)
08938
.. _.. AVGX = SUMX/COUNT
09070
AVGY .. SUMY/COUNT
Q9106
SOXSQ =SMSQX - AVGX * SUMX
09142
SOYSQ =SMSQY - AVGY * SUMY
59202
S~Xy
=SMPRO - AVGX * SUMY
09262
R = SOXY/(SQRTF(SOXSQ * SOYSQ»
09322
B • SOXY/SOXSQ
.
09382
i)9418
A = AVGY -B * AVGX
PR INT145 ,R, ITEM
09478
PUNCHl4S,R ITEM
0'9514
'., 19/)
_,F14.4/1X14HCOUNT
145 FORMAT(1.X14HR
Q9550
... PRINT147,A,B
09676
PUNCH1.47,A,B
097.12
147 FORMA:r.(..lX14HEQUATION, Y -,F17.7,3H + ,F17.7,4H * XI)
Q9748
.... PUNCH148".
.... ..
:....
09852
148 FORMAT. (1 OX1 HX, 16Xl HY , 14X5HTRENO, 13X4HDEV • I)
09876
... SUMT -0.... ..
...,.. . ,.. .. .
.. ...
' ..
T0082
SMDSQ .0.
T0106
SMSQX·O.
TOl30
OlS4
00102 1.1"ITEM
T fI4 .. A + Boo X( I )
01.66
D. Y( 1)- T .
0238
f
0298
"0334
r0382
T0418
;0562
;0608
I0656
tg~·~~
-0888
0924
+
;1044
SM·SQX.SMSQX+O
SMOSQ • SMOSQ+O**2
SUMT • SUMT + T
102 PUNCH 48, X( I'), vel ),~T),;,
.. 48 FORMAT(1X4F17.7)
. STOEE ... SQRTF..('SMOSQ./CQU;N2j)
PUNCH 48.! SUMX., SUMY, i&UIr1r,.;SM$Q~
I
PUNCH471~TDEE
AVGX
.
.
47 FORMAT.(.l111X14HSTD ERR.
• ,:Fl1 •..7ltX14HMEANX
., F17 .7)
.. PUNCHSO ,AVGY .• SOXSQ
.... , .....
50 FOR~1A T. (1X14HMEAN Y . ,·F17 .7/1,X14HSUM OEV,X <,SQ. •• ,F17. 7)
PUNCH52.S0YSQ,SOXY
7:.6
f!..J·7
c
Dvwr'u'@twmm
·_j .. Wp··
__ .. W .. ,· .... ·.. ·t" ........ "fo·Hij"· - --
APPENDIX B (contd.)
11080
11200
1'1224
1290
':1298
0
52
'.
54
..
FORMAT(lX14HSUM DEV Y SQ.-,F17.7/1X14HSUM
PUNCH54.,SMOSQ
. .... ..
.
FORMAI(lX14HSUM DEV 0 SQ.~,F17.7)
GO TO 10 ...
END
DEV
xv
.,F17.7)
PROG SWl OFFFOR SYMBOL TABLE, PUSH START
19999 SIN
19989 S.INF
19979 COS
T9969 COSF
19959 ATAN
·19949 ATANF
T9939 EXP'
T9929 EXPF
T9919 LOG
19909 LOGF
T9899 SQRT
T9889 SQRTF.
T9879 ASS
19869 ABSF
19859 X
T6202
_
T6199 Y
12549
12539 0'010
f2529 SUMX
.. ~ 12519 0000000099 .. ,.,
.
T2509 SUMY
12499 SMSQX
12489 SMSQY...
12479 SMPRO
12469 0001
T2459 0001
r2449 ITEM
T2439 COUNT
12429 0100
T2419 I .'
T2409 ~002
T2399 0002
T2389 01.01
T2379 000
f2369 OOO~
T23S9 AVGX
T2349 AVGY
T2339 SOXSQ
T2329 SOYSQ
12319 S~Xy
T2309 R
f2299 QOl
T2289 B
T.227.9 A
12269 ~145
f·2259 ~14S
T2249 0147
12239 0147
12229 0.1.48
. '221.9 0.1.48
T2209 SUMT.
12199 SMCSQ
T2189 0.1.02
.. 217.9 if"
T2169 0
T2.1.S9 0:048
T2.1.49 Qo48
T2.1.39 §.TOE.E
T2129 ~047
T2.1.1.9 Q047
T2.109 OOSO
r2Q99 0.050
T2089 Q052
T2079 0052
. 12069"0054 ..
r20S9 OOS~
SW 1 OFF.tO IGNORE.SUBROUtLNES,PUSH START
PROCESSING COMPLETE
o
,
.
o
77
I
•
"t'tth
"~
....
~.---..;-~-
--
..---------------..
-.----..•-.__._
...................... .••.•..•.............
.•.•
_-_..
~-~
..........,. ...•....-...-_.._-_.-
APPENDIX C
R
,.
COUNT
M.
.8836
13
y'"
EQUATION.
8.8943910 (,
12.0000000
19.0000000
21.0000000
16.0000000·
43.0000000
49.0000000
~9.0000000
totc..L'\'V '-+
60.0000000
6t.O.OOOOOO
70,0000000
101.0000000
120.0000000
18·1.0000000
842.000000.0
5TO ERR.
,.
9.3000000
11.1000000
10.9000000
16.4000000 .
21.1000000
20.'000000
17.8000000
28.1000000
.23.6000000
44~'OOOOOO
30.1000000
43.4000000.
50.00000QO
326.8000000
II
6.0621261
64.7.92300
MEAN V
,
SUM DIY X SQ ••
,.,0,21.1184610
•• 1010000
. MEAN X
1110.9110000
.,'0.1160000
.
SUM DIV V SQ,' '
SUM DIY XV
,
SUM DIV. 0 SQ..
'
• .".10."00
.2~O7991
*
TREND
Y
X
C ~
C
,
11.903.981 0 .
13.6~9'7S0
14.1611730
17.9231610
19.6787560
21.183"10
23.6915430
23.9423420
21_1963380
. 26.4503340
31.2283050
,8.99.02930
1'4.7906410
326.79·99700
X
DEV.
-2.6039810
-2.5595750
-3.2611730
-1.5231610
1.4212440
-.683~~10
-~.8915430
4.1516580
'-1.5963380
18.0496660
-~.1283050
4.4091070
-4.7906410
.0000070
(
if"
..,Jv
I
Appendix D
1.
COMPUTATION 0' COEFFICIENT OF CORRELATION
C:\
(Short-cut.crude data method)
. (1)
(2)
X
INDEPENDENT
Y
DEPENDENT
VARIABLI
KiA PIB
VARIABLE
X2
9.3
11.1
10.9
.16.4
21.1
20.5
11.8
28.1
23.6 .
44.5 .
30.1
43.4
2°. 0
18~
326.8
CbTALS
842
MEANS
r ··64.77
(5)
y2
(6)
(7)
REGRESSIOH
·Ye VAWES
ULT.• CUSTOMER '
PER POLE MILl LINE TRANSI.
12
19
21
36
43
49
· 59
60
65
70
105
120
(4)
XY
(3)
OOMPUTED .
SQJAlES " PRomOTS
111.6
210.9
228.9
l44
361
441
111296
1,849
2,401
.3,481
3,600
4,225
4,900
11,025
14,400
~~1489
81,612
590.4
86.49
123.21
118.81
268.96
44.$.21
.. 420.25
316.84
789.61
'556.96
1,980.25
906.01 .
1,883.56
2,,00.00
907.3
1,004.5
1,050.2
1,686.0
1,534.0
. 3,115.0
3,160.5
5,208.0
211 0•0
a7,951.3 10,396.16
z
. DEVIATIONS OF
ACTUAL Y VALUES
FROM REGRESSION
OR COMPUTED y.
VALUES
11.9
13.7
14.2
17.9
19.7
21.2
23.1
23.9
25.2
26.5
'35.2
39.0
- 2.6
- 2.6
- 3.3
- 1.5
+ 1.4
.7
- 5.9
+ 4.2
.. 1.6
+18.0
- 5.1
+ 4.4
- 4.8
-
2~·8
~ + deviations-28. 0
. ~- dev1at1ons a 28.1
. 326.9
!- 25.14
54,;36 21,166.8 8,215.75
CORRECTIONS. .
~, 790.5
. 27 ,076
CENTERED
* Corrections
l:x2'. 1:X7
.
2,180.41
i -
I:~
64.77 x 842;8 54,$36
64.77 x 326.8 - 21,166.~
25.14 x 326.8 • 8,215.·75
.
r-
•J
ari thm:3 ti c
.me~
of X
'! •
ari t~~tis:
r.
~Ul'l
.mean """ I
ot
.'
it~.o
6,790.5
59,0)6,781.16
6,79O.S
•
6,790.5 .
1,683.S4
•
• +0.8838
Coefficient
ot Correlation
79
"
'"
.~---~~-~-~------- ...
-... -.... -" ...."-
'Appendix D cont' d.
2.
h' ..
•
.
c
.
.........-.............-
rrl
(:O:1a'L i'A r j\)tJ Oft' it&.Ut~dSl ON 1,1
,..-...
L;x.v
_ ....._..
6:;! '190. C;
E.x2·
27,076
.2S00·
a ~ Y-bY • 25.14 - (.2$08)(64.77)
• 2S.14 - 16.24.
'
.. 8.90
Tl'er£d Lino or Yo .' a+bX
• 8.9Q+.2S08 X
Oriain at I • 0
. Equation tor straight line Yc· a+bI, where.,
I !II independent variable
.
Yo· computed I value or trend value of,
. the dependent variable
a • the Ic value when X-O
b • ,the amount and direction of slope
at the point where 1-0
Substituting the' original value. ot I in the above ,eQ.uation
Yc· a+bX, the'e.timated 10 value. ot the regre8sion l1n~are computed.
For example.
Yo • a+bI
Yo • a+bI
, • 8.9 + (.2$08)(183)
• 8.9 + (.2508)(12)
• S4.8
,(S.e Column 6
• U.9
,G
At tn. point where ar1linal X valu•. - 12 the oomputed or regression'
val"e ot y. - 11.9, and where orilinal I value • 183 the oomputed regression
value ot Y • S4.8. Connectinc the •• two computed I valu•• bya 8traight
line sives the rear.ssion line.
of
Not. t~at t he 8U~
tne r~srel8ion or oOJlllNted Yo values approximates, U.e IWI ot the liven Y value. (See .oolwm 6)..Al.lo, the sum ot the .
claviationa ot the clot.•. above the rep•• aion 11_ approx1ute. the ,SUM ot the
clan.tion. of the dot. below tile 'regre. 81 on line (See column 7)., Thus, the
relr.asion .line calculated br the "method .ot 1...
18 ·termed the
line ot best tit.
"
.
t-.cpU'....
o
The coefficient ot correlation haa been computed and found to be
+.8838. Ualna tba knowledge that +1.00 18 perteat pOlitiv. correlation and
-1.00 "epre••nte perteot negative correlation, it IIllht be interpreted that ,
+.8838 repre.ent,a talrl,. hieh po.itive corr.lat:l.OD between ult1ute au.tomers
per pel. aile ancllCVA per line t-ranatoraaer. H~er, there reu1ns' atinal
computation to telt, the .ilnit1aar.Q8 of the aoetftG\6nt, of oo~.l.tion datermned trOll t,hese two ••ri..and appra1ae :I.t,a yal.ue -- whet.her :I. t be real and
81p1tioant 00lT.1&t1on or Ml'e1, •• to obuaoe.
8n
'-I. I. II
C
Appendix D cont'd.
The teat ot, aigniticance 'ia baaed upon the "null hypothesis" which
aaawre. tha~ no correlation eXista between the aeriea ana17l8d., Chance
values ot "1'' the coetticient ot correlation or related measures (F) are
calculated and evaluated in a special table ot r values. From this table
(tound in Daviea and Yoder, !3USINESS STATISTICS, pp. $86-7, or Croxton and
Cowden, APPLIED GENERAL STATIStics, pp. 678-9 -- 8ee reterence sources) the
value ot , computed trom tii8 coettic1ent of oorrelation derived trom any
correlation .nal~i. can be compared with r values expected due to "chancen
(chance operatinl S ti•• out ot 100) and d,. to ".ilft1ticant co-relation"
(chance operatins 1 or lei. tim out ot 100). It 'our cOlI;)uted , value dieprowl the "null hJpotble1I", then we have real correlation. '
The , table liats value. tor .0; or chance level, and .01 or s1g-
niticant'lavel.
Note that the , values are in 1nwrs. relationship to the
probabili ty ot chance or the chance le.,.l. The greater the opportunity ot
chance the naller the , value and vice veraa. The table can be interpreted
•• tollow••
. 1)
'~','
C
,J
It the computed r val,. tor a partiCular problem is
, lee. than the .0; val,. or " then the correlation
can be said to be due to c~ce.
2)
It the computed r val,. tor a particular problem is
greater than the' .0; value of r but les~ than the .01
\,&1,., than the correlation -7 be tel'D8d 118ign1ti~ant".
3)
U the computed r Tal,. tor a particular problem is
greater than tha .01 Talue ot r, than. the correlation
mq be interpreted as being lth1shq all1liticantll •
'rEST or SmlDICWlCE
Computation and Interpretation ot', vallIS tor our problema
_01 explanation
',.
.
Appendix
4.
D cont' d.
c
,
Frona " table ~h'.!N columns are em-I) and lines are (N-m), we find
the F VAluEts tor Ollr prOblem' in· Column III (2-1) and Line Nll (1)-2)
Each
lind tlUa,tbor hAS .two values of P, the.OS level and the .01 level (1'1\ DaViAS
and fodor, pp"
In Croxton and Cowcien (Appendix 0-2) the .COl
f
S86-sa".
level 1••hown in acti1tion to the
.OS and .01' r .81\188.
., VALUE .
FOR 2 SERIES OF
13 PAIRS OF ITEMS
CHAHCI I&VEL
.OS
(DAVIES AND YODER)
(ch:mce oparatilJg S t1_~ out ot 100)
or ohanee la ... l or r
=4.84
.01 (chonce ,operating 1 t1. out at 1(0)
or 81pd.t1oaDt level ot r
Since ourcQIlpUt8d value of .,
:I
39.2S and 18 a value greater than
9.6S, \he ., "alue ot the .01 or 81gn1ticant level, chance i8 operating
less
, than 1 t1Jl8 out ot'lOO. Therefore, it JIlUSt be concluded that the coefficient
of oonala14on ot +.8838 detendD8d trom 13 pair ot i tams ot Ultimate
Custo.r8 per Pole X:Lle andlV! par Line Tranator.r taken trom the 19~7
'1' " D Stu41 is hilhl{ significant.
In o1;her worda, the items oorrelated
def1n1telT pro.. tiii a reil cause and ettect Nlati~nsh1p exists.
The
iacreaa. in XVA per J.d.ne ~anstor_r is influenced by the increase in the
number of O11'4ut. C.tcmm'8 per Pole M1le.
1he r teat further indicates
t.hat. our ••p1e -ot lJ pair ot itt., while pemapi a,lmall '&l1q)1e, val lars.
enousb to pnft oonalwd.velT tha~ real correlation enats between t~se two
, aerie8e
(if"'\
. ".,
On the other hand, it the oomputed value ot ., had been smaller than
the .0S level of " thell it would have indicated that there was no real correlation present, and either the coefficient of correlation would have had to
ha... been larger than .88)8 in 'arder, to be significant or that our sample of
13 'pair o,t it.~ vas ,not large enough to prove that Nal correlation existed.
The toreloing humen a brier attempt to 'acquaint the reader With
ot the ItaU.t1oal tel'l8 &880clated vlth oorrelation analysis, to illustrat, • method of ooaPut:l.DI thl o08tt:1c1ent ot correlation and regression
~o t,••·' ,tM ....rd.tloIftCl ot tb. Nlu1t1, IDe! to point out lome of the
1t.1Il,.t.i 011 in ""111 0111 'I inbI,rpNtat4on ot thl U,NI ot oorrelation 10111,
QO" va. 0019'1'-.
of "Z'I (\hI oOlt'ti~i•• of o.relat.ion).
80.'
a_,
I
,,11,.
I
,
c
CROSSING POINTS
1A'1'I SILIC'lIctI AND CCllPARISOH
R. L. Pitch
LaDI I.laDd L1lht1nl Ccapaxv-
--- .. ---.--.-.-...
-~"-~.'."
- ,-....
-,.,' ............. '"-,.. ....
,~-.----
'--
'.
CROSSING; lOINTS
Rat. SeleetioD _cl·CoaaparisoD
,·
C
I
H. L. Pi¥b • LILCO
',~,
.,.II'
Rat. dest.p ael simplification 1s oc business. We ..e constatly reviewing
our strucCur•••~cht.Da for improved and reasonable rates, not onJy-to reduc. our wort; loael aDd e&pallle, but a180 for clarity aDd underst_dinl by
our CODs. .
LILCO il DO exceptioD.
-
r..
1 would like Co refer to the N. Y. S. Public Service Commission'. memoraodua
adopted J . .~ 31, 1955 quote - tI--After tboroup consideration of the
v8&'ioua .thode of brlDliq about the desired rata uciformity and tariff
8'-Pl1ficaC'oo, it is tbe judgement of the Commis8ion that, with the two exceptioas of ••nice froa \Ulderaround linea aDd aealonal service, there should
be ODe t~iff schedule for the entire territory for each of the major cat..
aori.. of '."ice t resideDtial. seUl'al or coaaerc.ial, and industrial."
SiDce that t:lae ou .p.emene baa baa en_avoriol t9 attain this 80a1.
Coadd,olaa or cacel1lai ratea, because of the a8ve...8 effect on cu.tomer' 8
billa,
be doDe ooly'after • complete review of each account .ffected
aDd wttb sc.p. of jU8ti~lcation.
mu.,
'aced with chia probl. . . . have for tbap.at five years apent endiesl day.
of cl .... lcal computationl .xam1nlna our coof1ictlna ... at. with the ulti.ate
loal of el11Dlnatina it froa the acbedule.Gradu.allythl'oupc:..eful rate
s81ectloDacd reductioaa, utilizinl fully our 1BK 1620 dilital computer, va
are O~ th. doo~ atep ofattainin& complete rate uniformity aDd :tariff
.~plificat1oQ.
.
,,-,
. if"'"
thl. ,ropt.wttceD 18 fOlUtAT rCBTRAN for card inp"t into an IBM 1620
dillt.l computer, will •••ble you to compare rates by produciul cros.inl
polat. aDd bill ~abl.. , if de.ired, 18 . . .tter of minute••
Bxhibit I (Cwo , ....) 11 a s.lf-explaDatory flow dialr...
Ixh1ltltl1 (tH•• , ....) 1a the pl'ol... a writtea. in FORMAT FORTRAN
Exhibit 111 18 a .ap1. of the input data card aDd a gloalar:y of teru aDd
'witch.. u ••d.
IMblblt IV'.....pl. of tbe heade... c.rdl.
_.lblt V,• • ' 1&.1. oltb. ~UCp"t
'a'a lellal'ded .b)'th. ,&-01&-••
BaibiC VI ,. ID . . .,1. of ,r:LIlt.d output,' data ill .cap.: .of.501W Damud
BalbiC VII, i. • lik. ....,1. of pl'lDta. out,uC da'. 1D .tep. of 1 KW ne.lDd
labibit VItI 1. • ....1. of • btll tabl. aeDerated by the prol~" if desired.
'rbI, p~oar- "... .. lDput elat., ."ea1e cud ...laoeel by )'ou cOlltaiDiDI •.
deuDd ,UDOIa aDd • Clae ,\lOch. Toa.""1' v1tb tbe utllizatioD of awitch.s
tbls offe~. UDlt.iteclfl.l1bt~iC, .. t:~ .t.rtlDI polDt aDd output de.ired.
I{. 2.1
8 "tJ
· . . rww" ""( "("mOO' r ¥#*5£'"" E'if" N" "w±
- 2 -
Crollins Points (cont'd)
Nat.rally the complexity of the rates to be compared will determine the iDput scope. In explainiDI our Crossing Points program to you, allow me to
refer to the specific examples exhibited. Pl.s8 f •• l free to ask any questionl.
Prolr,! Utili ••tion
With raf.r.Dce to exhibit II. tlli. pcolrlll "a. iDittated to compar. S.C. No
2 with S.C. No.4 by calcul.tiDI tha differ.Dc. in the billiol of a.pacific
d••aad io atape of hour. ue.. IncorporatiDI .witch 1 for Primary 01' SeCODdary servica. .witch 2 to incremeDt the demamd .fter tha 730 hour cycl& ,and
.witch 3 88 aa. option to obtain a b:Lll tabl,e. we _ra abl. to produce
Cro.a1.DI points to ,I'aphlcally show the .dYamc.... of our rata 2 VI rata 4.
Actually we are compariDI four rataa:
aata 2 secondary vs., Rata 4 scoDdary
Rate 2 primary
VS.
Rate 4 primary
complete cycle of 730 hours (avg. hours in avg. month), selecting each,
time the proper rate formula ( 9 - rate 2, 4 - rate 4), and generating output cards on either Crossing Points or bill table, requires approAimately
eight minutes machine time. A review of both Exhibits I (flow diagram)
together with Exhibit II (the program) will clearly detail the sequence of
,8 Single cycle.
A
'
C
,~,
The output data, printed examples (VI" VII, VIII) was geoerated from one
input card (Exhibit III), maoually operating the switches at will, in
approximately 40 minutes.
Let us examine this output. -
(Blackboard Exhibit)
Conclusion:
This demonstration, however brief, indicates the advantage of l620digital
computer application in rate comparison or rate aelection work. Possibly
aome of you are presently confronted with this problem. Although this
program was designed to aatfafy our clemand, only slight modification would
beneces.ery to solve innumerable comparison problems. The general FORMAT
1. here for 8ene~al use br any member.
o
"'j
01""l1li1.
D'#(~)
c
#fIIJ/CH
,
.5I~t:W~II~Y
;I'II"'AI~
-Pull/eN
.....--~7
1;#81.1'
1I1"'''IN~
,..
c;
t:1I'cit.
II
MINIMUM
,..
,~~
HININU/rtf
tlNM~.
411111¥1
~(//I
,
.I
'-0
~INr~"l
IIIrl
"""/11111." 'os
4"ll.lar
MN'EII,
21
If
eli'"
~lIi"
c
W"UfN'W"'n"r"
T
--"'if Fa!" E"JP-"-Cr
-"F
"" r -.j@""""WW"-PJ""]T
.It,
ON
+
~
-]-
7"0
-~~-~
#VaJCH
2,
~,.,,..Jp
T,I<4C
CD,J'f)H~ • .J
.------.!?_,,,,,J6)
;
.
C
C
C
C
',:,'.
C
CROSSING POINTS S.C.NO.2 YS S.C.NO.4 EFF 6/1/64
H.L.FITCH 3/64
SWITCH 1 ON FOR PRIMARY, OFF FOR SECONDARY
SWITCH 2 ON FOR INCREMENTS OF 50KW, OFF FOR INCREMENTS OF lKW
SW'~i~E~SY~N-D~~~f21ABLJ,- OfE__ B_I-!_AS_S_
-- --DIFF\ 1).0.0
OIFF\2)-O.0
IF(SENS~ SWITCH 1)906,907
906 PUNCH 921
:'
i
_________GQ_ TO 9..Q_~_ __ _ ________.'___ ... _ _ ___ .__ .. __ _
907 PUNCH 920
904 PUNCH 1
104
PUNCH 4
READ 3,O,T
c.o.o
- -- --- - GR=O.O --.. - -- _. -.--.. -.- -- -.-- -------.. ---.- - --- - C-D*T
IF(SENSE SWITCH 1)5,6
6 IF(Q-l00.)7.7.8
7 IFCC--D*.30. )9,9,10
·-~··~~~to~~gOD._10 IF(C-D*120.)11,l',12
11 PRzO.8S1*O+O.0233*C
620 IF\PR-l.800*O)9,9,14
PR~O.(j·6~_"
.--.. -_. -.. ----.. - . ---.'
,
GO TO 500
- '-"--'--
i
15 PR=2.027*D+O.of3S*C
!
16 PRu;:2.027*D+(L.008S*C+360.000
GO TO 620
~--.-.-
.
.... "
GOrO 620
--- ~~~jOg~D) 1_8~~8._~__ . --- ----...-----.. ---"'-"-"
....- .- . - -..
GO 1'0 500
- 19 iF ( C-0" 1 20 ~ ) 20 , 20 , 21
20 PR-O.8o1*O+O.0233*C+5.000
27
IF(PR~1@800*D)18~f8.22
22 PR&;PR
-.. ----.. - ..... --... - .-- ----- _._._._- ._-- -..
21 IFfC-72000.)23,23 24
24 I FfO-200.) 2S ,'2-5. 2bJ
---Gc)-TCf~6(r
23
P~-1.977*D-tO.0135.*C+5.000
GO TO
0
"
\"'-.,,;
! - --·T2·-'-t\·~'~noooo J1S-:-l5-~15----'
:
(£--1
14 PRs:PR
21
...-. ~~+69 H-*I):~ .!.Q.O~2~C~1.6..h00.9_____ ----.._..
26 IF(C-360.*O)23 23 2
29 Prt.3. 777*0+0. 0685'C~5. not' , -. ..0
GO TO 27
500 IF(C..gO.)501,501,;02
501 GR-l.500*O+O.039*C-2.588
-------lm-rn-Tomr···------·------------·---··-.. - -.- .... -.
502 IF f C-210.)503 503 504
503 GR~1.500*D+o.031*t-l .868- .. - _ ..
GO TO 1000
504 I Fe C-l 500.) 505, srrs-;)06" .-. --_ . _._.- - ...
S05GR-l.500*O+O.02S*C-O.608
87
I
C"'
.
GO TO 1000
506 IF(C-12000.)S07,507 l S08
507 GR=1.5CO*O+O.Ol·9*C+ts.392
GO TO 1000
508 IF(C-36000.)S09,509 ;12
509 GR~1.S00*D+O.013*C+~O.392
GO TO 1000
512 IF(D-100.)S13,513,514
513 GR~1.500*D+O.OOR*C+26n.392
GO TO 1000
514 IF'C-360.*O)509 509 515
39 2....... - ..... -.
515 GR-3.
GO TO 3no*o+O.
, 000 "'-' oo~":c+Ao.
_..... _._
'
5
600
IF~D-100.)6no1600,601
IF·C-O*30.)6nl,6~2,6~3
602 ?R; 1 • 8n~';'~O
GO TO
70~
603 IF'C-o*120.)604 604,605
60~ plr~O.551*D+o.nrl3~t·-- ---- -_.-_.._.- ... 609 IF'PR-l.800*O)6~2,6~2,606
606 PR=PR
GO TO 700
605 IF(C-72u00.)6C7,607,608
607
PR~t.727*D+O.0135*C
- -'-' . -. -- GO
TO- bUg' - .
-- -_.- . --- .---- _.- ... - --603 PR=t.727*o+o.on85*C+36n.ooo
GO TO 609
601 IFC-O*O.)61',611,612
611 PR= 1 .800*0
GO TO 700
- - --- -sT1"lF - C~D i~ -zo ';)"6-11, ol.T;-6n;.-- -.. " -.- -. --_.. --613 PR=O.501*D+O.0233*C+3.000
610 IF(PR-l.800*O)61',6Tl,6TS
615 PR=PR
GO TO 700
614 IF~C-72000.)616 616 617
----·-oTi-1FTU~wu:i6Ttr;ota-;6t9 .--- ------.--. - -- -- ---.,. ... -616
PR=1.677*O+O.0135~"C+5.0()O
GO TO 610
61~ PR=1.677*D+o.oo85*C+365.000
GO TO 610
619 IF:C-360.*O)616 616 622
.-- - 612-P1r=3 ~477wn+O".Oo-A"5W:C-~"5~otm - -- -._--_. --" -..,GO TO 610
,no IF·O-100.)701,701,702
701 IF.C-90.)703,703 704
703 GR=f.350*O+O.039 t C+4.912
GO TO 1000
.. - -jOq- rF.....-C~2"TO. )70-5,'703":700---- --'--~-'''- -'705 GR=1.350*O+O.031*t+S.632
GO TO tooa
706 IF(C- 150 0 .)707 t 70 7,708
707 GR=1.3S0*O+0.Oz5*C+6~892
TO 1000
'70B GO
lFTc~ 11000 ~ 1709;709' 7 rc; --_._--._.-.. -._--._... 709 GR=1.350*O+O.019*C+fs.892
GO to 1000
710 IF(C-36000.)7111711A714
~O···
..
-
711
714
GR=1.350*O+O.Olj*C~7.892
GO TO 1000
GR·;-Y·~·JSO*D+o-.oog-~C+"I6T.89-2-··-·--·-
- --
GO TO 1000
70211=":C-90.)803,803,804- - 803 GR=1.200*D+O.039*C+19.912
GO TO 1000'
- . -.------.-.. --804 IF.C-210.)80S,80s,806
-
.88 __ .
805 GR=1.200*O+O.031*C+20.632
GO TO 1000
806 I-F( C-1500.) 807 807 808
807 GR=1.200*O+O.0~5*C+21.892
GO TO 1000
808 IF{C-12000.)809,809,,810
809 GR=1.200*O+O.019*C+JO.892
GO TO 1000
810 IF(C-36000.)811,811 814
811 GR=1.200*O+O.013*C+f02.892
GO TO 1000
814 IF(O-oO.)815,815 816
815 GR=1.200*O+O.OOa l C+282.892
GO TO 1000
816 IF(C-360.*O)811 811,817
817 GR=3.000*o+O.ooA*c+102.892
1000 DIFF'1)=GR-PR
IF'SENSE SWITCH 3)2006 1500
1 son - I FTD1FF'· 1» 2000,2000,2601
2000 IF OIFF 2»2005,2005,2006
2006 PUNCH 101 1 D,T"GR CPR,OIFF: 1)
IF(SENSE ~w TCH 3J903,2500
2500 CC=O*'T-l.0)
TT=T-l.0 '
-
P. 3 col' J
- PUNCH, ~005,O,TT,CC,DIFt'2)
2005 DIFF(2 =OIFF 1)
DIFF(l -0.0
GO TO 903
2001 I Fe 0 IFF· 2» 5000,2005,2005
5000 PUNCH 101 1 0,T,C,GR,PR,OIFF 1)
- 'GO TO 250u
903 T=T+l.0
iF,T-730.)104,104105
105 IF\SENSE SWITCH 2'41,42
, 42 0-0+1.0
GO TO 43
4 T 1)=D+50-.0
43 T=1.0
IF(SENSE SWITCH 3)104,1501
1501 DIFF~ 1)-0.0
0IFF(2)-O.O
GO TO 104
--'1
101
F-ORHAT(41H
KW-
HRS
KWH
S.C.NO.2
S.C.NO.4
,10HDIFFERENCt)
FORMAT(F5.0,FS.O,FS.0,Fl0.2,Fl0.2,FS.2)
3 FORMAT(FS.O,FS.O
920 FORMAT( 18H SECONDARY SERVICE)
.
921 FORMAT(16H PRIMARY SERVICE)
.. FORMAT(lX)
)005 rGItMAT( F5.0 ,FS.D ,Fa.o, 20XF8~ 2'
END
8H
.L.L
"1
T,.;;t.,,·t
Wtf
t
C. RO~5JAJ"
C'
s.~. AJ~. 2 ~~.
.pOI"JT~
4-
GLQ5SARy'
'\
"
4.
b.
I
c
C.
D
d.
T
,...I,
.s.
.cec
J"
TT
e.
D +~e'e"t! f!..
PIF'F
Con~" ",p+-lt:1n
- :P. ~ Itr1 d
I i rn t. ( J,~H. ~)
PR,
/i),Jv ~ic,n u"f..
~"a/ ~e.~C47I4(....
-. P"'.VIQU~ C'fJns,.""",p-l-,t:lf?
-
7i ,.,., '-
du f'1.n?, ~ IJY
en c. /"ok,"-
~wl-'-(!J.lES
l
ON
.pr'lrt:I~y
I
aFF
~e~ol7ky
2
aN
2,
1)':':-
~
0,.)
3
o~F
I
( ..
,
.;-
~NPU""
70.
ale
I?a;k.~
litd-e:J
~,,~ ""n?~r7h G./'_ 50 KW
~r1C "e"".,-/-:1 o.P I K",
?J,I/ 7a61L
'8y - 'ass
-
J)ATA
1•
~aOOOOCOD~OCOCOQa0030COOGOoonoooo~oaJOOtionooooaOG~aoo.o.aOO~Qoco~nn~oaJuti~UG
',.': -
',~!lf'
:,1!4,~""'·'.M~I:Z;';'.~l'i·'~_~~:
.;)'IM~J~J'IIJ'j
... '~;4:~4"4i.,C.4~'~)I~'·'!"~"'S-·~!I9ce.I~P"Q."."."nT.I;.~"'I.).IS"
l1iII111:!1111111:i t :llilli:lI11Jl"1111111111111111111111111111"11,.'
,,11Il11
, : 1 2? 212 2 2 2! Z 2 Z, 2 =~ 2 L 2: 12 Z; 2 22? 2 2 ~ 1. 2 2 ~ 2; ~ 21.12 J.12l2 2 2 Z 2 2 L 22 i 2 2 222 t 2 211:' 2 22222222 Z
lZ3.i 3333
44 f ~ , " 4 ~
3llll~J33313l3SJJ3jJ1Jjj3:J~J33J~~3j31Jl~lj31jJ3l33333333333333333333l
4 ~ .. 4 • 4 4 .. 4 4 .t 4 ;. , -1 4 ~ 4 4 4 ~ • , ~ 4 " 4 4 4 4 4 4 4 " C4 4 4 4 4 • 4 4 4 4 4 4 4 ., 4 , 4 4 " " 4 4 " 4 4 4 " 4 4 ~ ,~ 4 " 4 4 " 4
S 5 ) S :) 5 t; 5 J ~ ) ')
,
~
: t.
!J "
5 ~ 3 ~ 5 5 5 5 ~ 5 ~. 5 ) 5 ~ iSS S 1 5 5 ~ 5 5 5 5 5 Sal
6 ~ 5 f: & ; 6 & i 6 6 6 ,
£,
s ~ 1I b 6 &6 S 6 Ii 6 , 6 i
~ j ~ ';
S5 ~
~
5 5 S5 SS ~ 5 5 5 5 5 5 5 ~ 55 5 S5 5 5 5 5 5 5
6 & 6 ~ & 6 6 & 6 I) 6 ti 6 6 G6- , 6 , 5 ~ 6 Ii G6 i i • 6 , , 6 6 i GI 6 5 5 Ii 6 6 6 & , &, 0
'71177/111711'1777'771i11j111111?111117171711177171111/1111111111117111171111111
o
~A8~
~.a.
~8t!8Ia88S~~ia~i8R~~aa8a8a6iil'8aa8188a81"8!'8a8'888!,~a~8]aS8~~aaI8'
9Y]~~~9~99~~9999~=9Q9'99999!~9~99q9~~~q99q999951~99q99l~9Y9993~9S~;~'9ng9~9~q999
I • :; ••
T:"J!
..... __~ " ., -! lJ I. :~ ;! ..
~O.,
Ito -,
lC :'1 12 :J.4 1'\ oS 21 2111
)II
3. 102 II .. ~ "1;
)I
:os 48 t,
C! 4, ..
t$ .. 'f
41 •• ~
,t
~ '1 M ~." • • 111 ';(.16,a . . . . ,,,., "_I -271 Ie ,~ "
11
II " ..
1f.~.
E.)fI4161T
i
.J1L
..................
--......----- ........ -.-.-.-..------.....-
... ~~~.
4 .•.
Sf.".
• GD081DU·O .... aOe'OQIOCOOaUI000800CDDftOOUOOOOOU01900eeeeIOOI.OeIOI •• )0060000 'G
a
, t ' • 5 • } • • '8 \I 11 :3 14 ~5 II 11 " N 10 ~I :7 :U. '" A "
2t r: ~l " II ,. l'·
» 31 •
.Ill • •' 41 4J 44 "; ...., . ~'1» ~I >l 4l !I& ~,.
51 R lUI D • ., II a,. J1 12 Jl14 J!I "
n"
lI.
t 1 I 1 1 l I t 1 1 1 1 1 Ul t 1 1 , 1 1 1 J • I t 1 , 1 1 1 tIl I 1 I 1 1 1 , , , 1 1 1 1 1 1 lit 1 1 1 1 1 1 , 1 1 1 1 1 1 , I 1 1 1 1 11 1 1 11
Z
~~i22l222
33~
222222222222Z222222~22222Z222Z2Z2tZ2222222222~22Z2222222222l2222222Z
333333331331 3333333333333l3Jl333lJ3333l3l33lJ3S3333333333313333J31 133333133
4444~. '4.C444C4444444444444444444C4444~~'4444444444'4444.44.44444444444444444444
5S 51 '55555 S 55
'~6&
5555555555~S5555~5555555555~555555~555555S55555555555~55555555
6Se&~&66~86666&&I'&6Sli'6&1~8i66'i&1666661'Sil&6i6'111 ••• 116.' •• "16&."1.1
1 1 1 7 i 1 7 1 11 111 7 11111 7 7 1 7 1111.1 7 1 7 7 111 , 7 1111 7 111 7 1 7 J 7 7 111 7 11 7 J 1 J 7 1 J 111 7 7 11111 7 7 1111
..
.\
888968881 .Sa888 •• I' •• 18.81 •• aI88Iaaa,allaatall.18'.laIIS.II •••• 88IS"" ••• I'."
9-,!9!999999tI9"'9999'9!9't~~9'18tl'91"999'9'1'9999199StSIS!9"!'
991999!9 9S!9
, l J • , • , • \ .~ II 17 " 14 '; " " illS:: ~ Z2 Z' 24
"'I,!!,~,.' 5 0 . I
a 21 %l • a" .;. U'l S
~. l'
•
a
JS." ., 4.1 .. G ••, • • • II»." ~ a 5J ,. . . . II IUI . . . . .I • • II )' .,. n M n " 'I la 11.
....
,.
r::..,1~
l .. ' .
1' . . .
,
910
,nO~O
~ ~
• , II 7 • S
~OOO.
8aOOOoneOODO'~OOOeOIC~UDOD'OaOOO
O!IO 1001000000
••• ,o'Oe"OOI ;!O
i,.
n 24 JJ:t :! :'I ~ .'1 , • .l3 :;411 • V .» ••1 42 U 44 " • . , . 4.llIII ~~ 52 ,UU,,. sr tt.8 .111 U ...., .,,1 8 • • ,. 12 !Z M n 1111 "
1111111111111f111111111111111111111tllI111t"11'111111tl111111111111111111111111
•
!I)
11
l'
'J ,. I!I II " It" 211"
!Z
.
.
22 "2222 2221' 22222 222222 222 22222221222222222222222222222222212222222222212
333331333333333331311
33 3J333
33
1~IS33333'
444444 ••• 444.'.4444444444444.4'44.4444 4.
~'55555555~S'55&55'55'SS
5'55!~1'15
333333331333331.31131133333131
444444'44"""'4'~44444444"4'44'444.
115515555 I
lSI 66'1"68S6S II.IS&I •••• , ••••••• III1I1
I
15155ISS'l.55555'!155i55~'5'15
11811.1111118111118111&11111&1.161.
11171"7111171777777717777117777771J777J17777777777771 '7'77lll"'J'J""')'1)77'
'8118~1
~g'999~9
181811111 IS •• I I •••• 1. , •• , •••• 1 •••••• 1111' ••• 11111111.' ••••••••••••
, 2 3 • , • • v~
".:.:~
!1q'99'9~1'11998'999!t!"'9'!"19
'II
I
I'
acel
n ,e
"tt
tf II • '" 11%" lJ 'M It It l? • •
"'1
U JUI " . ) ' •
'9'
'.""SI""'!lltt!S"9'~'99'~!"
» ••, ~ .......... lUI au .... •, • a .. ,; WD It ... aua •
III ~
.'-1-
}II) " ,. I( ..
~.
EXHIIJIT
'I " .
r
$.
(J
t
~
i
i
I
I
I
C~O:'!..:;/"'6
____ - 9 UT P'''!T_
-_---..--- .pQ"JT~
_---..-..-- .......
...------.....
1 ',.
... t:
~:
.:,
~IJ l~acOOr~~DOO
G·!P'~~
OOO'Dpnou a
~
~ 1~ .~ ,~ :~
'~2'
~ ~I
:~ J~
I •
,
" :!
; •
i 'I! 'I:
.4
,'".
I'"
" -, '::
! '11111'
?6
'4
•• :"1 oK:
:tI .' :IJ '4
:Ii
II
ODDltOo oO 'loaoa O Boaoe£oaOOloueO;DI' •• 08
~, 40.,
4;' I ••! U • • / 48 4t' :It
~. W ~ ~ ~ s. ~ ••• ~ ~
)..1
1
. . . . &Ii . " . ill It 7!
•
n n ,. 11 " " "
JI.
i l l l l l l l l l l l l i l l illllllllllllllllllllll111111111111111111111
2'."" 2' 2 2 2 21 i 2 2' 12: 211 212 22212272 7122112 2 2 U t 2 222121122222 2122221H2 2222222 22221
JlIJ3Jl 3J3J13333 3333:33 111133333l1 3 1 133 331l3333333l33333333
~'11414'4411111'11:4'14'lll 11411141111141444 44444444444' •• 4 •• 4.4 ••• 4.44444444
),-.
11)J;:3
'11
~:,;~,! ~ ~,\, ~ 5 • j ': \55 51 5 5 5) \;" \ 5 51 H 5S; \ 5 5 5555 5n 55S >5 55 55B5 5 55 5555 555 55 555 5$ $.
~'H',' ;S,HH, 5', ;~'!iSH HHbF66&66&666i& UIS&UU6651I1IHU.UIIIUII"."
,;
1
C"
1'111) 1 I 111' I 1111//11717111 1111,1711111711777711717177717717771717117717
II/} 111
up.; HH
-e~i-i! S9!!BI~, ~n"38!1
~
99S"'!9!9"
'
:
'"
~:.~
;, 1 ~...;.,'
3
"<;-.:".'" I
"'3' iSt!t
9l"'9.'llt99S9~Sf19999'99IS99SS9119SI99'9'99
.~ ~. :~ ~ g~j~9S999
l: , ~ -=
'i;~ ~
~~,~ ~~ :.c~' ~ I' II .. a .. " Iii.
JJ~' ~ ,. ~
'"
' '.',
".. :
! ,
~ \ J,
I ! 1 , ! I:
2!;
n? n
!;:. JI J'J Jb
4J .. :.. •• : .... !III
'" I.
j.
'>.I
"
51!
..
V
1111 )J
/l
".
- ..-.;.
OCGC'"ijo~(aGoooaOO'OCCbOG801
;";.:~:"~'
0000U08'0010008.0000100100010
i1:.4'~ ~I ~
•
'J".s "" &Ii", :" $' U
_
V ...U10 . . . . .1 11.:'1" )113 J. JI JI
01n _ " .
I I I , , I , I , , , , , I I 1 I I I , I I , I 1 ',I 1 I ! I , I I , , , , , , I , 1 I I , 1 1 , , , 1 1 , I 1 , 1 1 1 I , 1 1 lit
t ". :' 11 '. I'
I, 1 , ,
,~,~
2. ::. ... :' i1
II, .t,
~
,au '80r,ca~oo~
• : .:
BeUBa! HllliSi!I&II.UII.1t811 .. S8UIU8!1
'S "'11'
p "
,'I
.... :"J
"C ;Ili l;).I 1'.)1. ,.: , "'" t' .-
211'1;: 1 n2l,,1111: 11 11 n l l l
3:11 ,')3 Ill!ll)
n
212 11112221) 111
iJ3IZ!33!13131113Iil"1~1
n
11112122 U222! 222222222122
113333333333333333333333333333333333
•
44t,.II!'414l141414444'.441.444444444.444444I'144444.,414444444444.444444.444444
,'.' '", '; E , \ 5 \ \ 5 , l , ) •
': " , ; ; -,; is
I
~ ~ ~ , S\ H
&", ,; ~ ~ l r;, r,,;
t
5 S5 5 5 ~
~\H
5,
HH 55 ••• In. &.
~ \ , \ " 5 , 5 5 5 5 J \ 55 $5 U
e,! 6
H H 55 H 5 5 5 51 58 55 5 55
5US&6H 5 05"6& 6 666 66&& S""UiUIi
,':; 71,; r i ! 1: l' I I I ' 11)) 171 1;, i 1'; '17//11177/1'1' 1171177711177711117117171777777
288l8iU~IH0B8B9n81idi UH38I8.USlI .... UU . . . . . . . . III . . .
~'3999~~!!9~Si9999§!999.9'9~t9999'!.~!9'99a"'999~!"9.199~99'111"9.A'I.'111
9 2_
.•:.r.......;,.,:,
~",.,,,.,,.,,
u . . . . . ~ .. » ""''''" ••• ,' .............. h. h" •• " n.
'an
o
HIS
&!Ha&8
j ••
I,
'
" , ' , •• " " ',',,,., ... ',"
.... " II"."
, ,," ', •• ' " " ..
Ll.d..IO
, C)(/.II~' T3L
------------- ------------ '----' --
:
I
SECONDARY SERVicE
KW
HRS
50.
50.
1~.
100.
1 •
100.
o.
,~.
l'l·
1
110.
200. 14.
200. 13.
200. 10~.
200. 10~.
2t:'0. 1 •
250.
•
('I~
2S0.
2'jO.
300 1
30e.
....3,90.
300.
250.
350.
.3 50.
620 •
1 •
100. ~Z8.
100. 377.
lS0. lJ.
150. 12.
l~O.
KWH
"".) 1
97.
1~1
1 ~.
93.
92.
'~.•
1
c.~.!'
600.
'~OO.
1 00.
3Z800.
37700.
'~20 •
1 00.
11~~~O.
D.
2300.
2600.
21000 1
20800.
o.
330 •
24500.
24250.
,I
S.C.NO.2
S.C.NO.4
~0.64
~o.OO
181 .8S'
180.00
Z'2.Z~
11J.00
210.44
2Z0.CO
~~~'24
~~4.06
361.5;'
360.00
65~ '2~
6~4.;O
4 Sl~ .6 1+
L~'O
1131 82
1Z6.10
.00
• nc
DIFFERENCE
.64
-.60
1.SB
-.6
-.20
• 3L~
.l.l~
-2.40
1 .39
-2.20
-1 • 1C
nr
.'
·
--.------
-- ---..•. -.----_ _-_...
-
-------
.)
4.Gl~
• '','"'
-2.20
",I'
... --_._------_._-_.. -
---
.J\.)
5L.3. 8~)
~!}o
2,g-0O.
2 00.
GCl3.09
(19- ",1.,.'2·j "
-2.2!
r'
~250.
6~3. 11.
630.00
2 50 •
"
1010'24
1011.14
3.1!.
-3.:;0
-.80
900.
.,.
--.------ .
2
-.8
1. 2
4500.
4200.
C
"')
.2
'"'<:,
!--'.to
-1.80
·
'-'
------
~.'I',':
\~
_--3!:C:..:..::!I(a...:!(}:....::S=-=$..:...:/AJta~L--...a:;~-JII"'I&I,~A1lL.rL-KS____~IIIIE.L.JTEe~:sIIIL--JQ~e:..-.....J5c-..::BQ~K::L..:kI:...1It..-_____________ ._
E 5+1, BIT
=vr:.
------------------------------------------ ..
----------------------1c·
Z}.-·;r.7;----(" '1
9·J!:
'
TZGitZij"u.m....
c)
SECONDARY SERVICE
K~'J
1\\'1:·1
'HRS
S.C.NO.2
;1. 21~: 10~~g~:
4914.
351. 14.
351. n~. 31~~j.
351. U·J. 3Qou' •
~~O.
~
..
j.;2.
l~.-.
~~~:
~'9.
nn
352.
~~~:
0..;.
1~.
1014.01
-3.5'.
-1.02
-,.57
- .23
~2~5.
638.49
635.40
640.28
037.20
Bn:
1020.97
102.2.04-
355. 14.
355. 89.
4970.
31595.
1023.62
j~t:
14.
2.38
-J:~1
to.LB.l1..----.l 01~..u...- -.. ~1..45. .
4956.
~li06.
1 52.
354.
rU"_'_"%''''''''"11"''''''' ..... - ...
2.SS'
. 633 • .6!l--·- . J .• 11
·1016.89
9 .2.
-J.~~ .. .Ag,.. -i 1.4.17 ••
88. 31l'G!••
5310.
3 • 1·5.
3U·
1012.99
1015.65
1 •
r,ur ..
-98.13
3.12
631.GO
31j~~:
30976.
.
DIFFERENCE
S.C.NO.4
63'... 92
636..7-1
5280.
....
., 1'"
.:..
b
3.00
-3.64
-1.6~
1.9
- 3-55.--~r---·~.--- .642.06 . ... -o..w~ao..-.- -.~.--.-.
-3.67
-1.89
102S.51
C...··,,-,I
'
-_._--- ---- ----_ _-----.-.-.----_ ...----_
..
..
---- -
-_._ _...
'---'"
--
_ .. _ ........... __ ., - ____ - ___ ...___ ...___ ...._ ..__ .u_. _____ -. _.' _____. - -
IKMl
. C~~~~ . . '?oIAl.r~---··~ .:'TEP:S
_ - :__ XHI!JIT
:vn: . __ ...______.___ . .-_-.___
..
o
---_
....
_
...
_--
~--
..-----~ . .,..--------~----- .....
-----.. -. - ... -- --.-- -- ---.
-_. ----
.....
--.~
- _._..
--
__
_. --- -_.
..
-'
- ----
._ . .··t'OO]'!t""·W
···"!' '12'1"1"\ TOO ...
w··· .....
c
SECONDARY SERVICE
KW
HRS
KWH
S.C.NO.2
48300.
1233.~9
12)7.~
48'50.
1242.39
350. 140. 49000.
1246 .94
350. 141. 49350.
1251.49
350. 142. 49700.
1256.04
350. 143. 50050.
50400.
1260.59
350. 144.
--35G.l4;.· S01!iO. -- 126s..1~-1269.69
350. 146. 51100.
1274.24
350. 147. 51450 •
1278.79
350. 148. 51300.
1233 .34
350. 149. 52150~
1237.80
350. 150. 52500.
350. 138.
350. 139.
1
.~e-. 151.·
350.
350.
350.
350.
35('.
-- 39h350.
.3 50.
350.
35Q.
350.
152.
153.
52850 •
53200.
53550.
53900.
54250.
156. 51.600.
-1-5-1. - 5495fi. -154.
155.
153.
55300.
162.
56350.
56700.
350 •. 164.
350. 165.
350. 166.
57400.
57750.
159.
160.
161.
55630 •
56000.
. ,3-5&.- -+6-3 •.- 5-/~5G.
-e, L.L.
581'1::'.
S.C.NO.4
DIFFERENCE
1349.00
1353.72
1358.45
1363. 17
1367.90
1372.62
1377.35
-115.70
1336.80
1391.52
1396.25
1400.97
1-405.70
-117.10
-117.28
-117.45
-117.63
-117.80
1419.37
1424.60
1429 •.32
1434.05
-113.33
-113.50
-11 C.68
-113.85
"ll~.~-
-116.05
.. t 16 .23
-116.40
-116.58
-116.75
138.2.07 -~116.93
1292.lf1+. --. 141-U~ ~ 117. 9S .
·1 296.99
1415. 1 5 -110. 15
1301.54
1306.n9
1310.64
1315.19
t- 3- t 9 • 74
1324.29
1323.n4
1333.39
1337.94
13l~2.49
--1433 .. 77 .., 1 1.9-.0 J.
1443.50 -119.20
1448.22 -119.38
14j2.95 -119.55
1457.67 -119.73
1462.40 -119.90
-t 347.04 - -1467 .12-.
1351.59
1356.14
136c.69
1471 85
1476.57
0
~l2.0..DB
-120.2~
-120~43
1481.30 -120.60
.AeL.c
.- _. __ "'XfJJ~.tr_ .
YIII
9 r"
- -d
c'
"iiwe8tbeirtWSrldwi"i - #*itt ---f-'
(
-'--bwd'iJid&,6:d" t·
itbbt6WairiM8Wf &n
--
Pf
_'_gn
c-
r-
[[W
.
'f
-
f*'IiiiHdt*¥Wb#,
.,
"'.'
1
I
f
I
C"
"
ELECTRIC WATSR SEATING
nCREMENTAL APPLIANCE USAGE
K. W. Brady
toDi .Ielallci £~p\lft1eO#lPf.1l1'
0""
~I 4
I I *'
7 ....~.
I'
t.
C,~
rpI""BW"-
It\MM!t1lt
·"W'"
ELECTRIC WArER HEATING
Incremental Appliance Usage
K. W. Brady - LILCO
This program is written in FORMAT FORTRAN for card input into
an IBM 1620 digital computer.
Tbe initial phase of this program began with a mail survey of
our electric .residential customera. LILCO baa better than 600,000 sucb customers; all were not surveyed. Some 250,000 of the.e accounts were selected;
I believe this vas accomplished throulb the use of random numbers. Included
in this survey were J!! our electric water heating customers, approximately
29,000. We were going to attempt to determine the usage of an electric water
heater a8 a 80le major appliance a8 well as in conjunction with various other
major appliances.
The program was scheduled to run for one year at which time
results would be tabulated to deter~De the necessary as well as the reliability of the data. This final phase has not been completed as yet, ho~
ever, the 1620 section has, and that is what 1 would like to demonstrate to
you today.
On page one (1) of tbe exhibits there is a copy of the original
survey card which was mailed to the customer. Many departments in the co~
pany had their fingers in this survey so that there is some information on the
card which is superfluous. This is the reason we designed a "condensed card",
also shown on page 1, which contained information pertinent to the water
heating study. This card was key punched directly from the original survey
card and verified. This deck of condensed cards, one for'each customer who
replied to the survey, then became our master control deck. InCidentally,
with regard to returns we received approximately a 52% return from the water
heating customers.
This master deck was th4D broken down by types of major appliances in conjunction with electric water heating to determine the population of each appliance group. Where it was felt that the size of the group
was too 8mall to give significant answers, the group was dropped aDd put ioto
an overall non-conforming group. Page 2 shows the various appliance groups
we retained as well as tthe code~81:van to:.each, group. You will Dote the groups
are split between yes and 00. We have some honest customers, who, though receiving the benefit of the water heating reduction in our electric rate, indicated that they no longer had an electric water heater.
The next step in the program was to match detail billing cards
with this master deck by appliance groups. Page 3 oftha exhibits shows a
flov diagram which illustrates the method used for this match. LlLCO has
both monthly and bimonthly billing 80 it was necessary to run four passes
through a collator to obtain the proper match. These passe8 were:
~a.J
98
- 2 -
1. Y.s group vs. 30
2. Y.s group vs. 60
3. No group vs. 30
4. No group va. 60
day
day
day
day
billing.
billing.
billing.
billing.
When all the detail bill cards were collated with the master
deck, those that matched had their proper appliance code number punched into
them;thos8 that did not match had the non-conforming group code punched into
them. This gave us a code number for all of the customers who replied to
the survey. For those who did not reply. a code, (30), waa given and- they
were handled as a separate item in the program.
The detail bill cards containing the code number -were then -,
sorted by appliance codes into a seriea of predetermined frequency ateps.
these steps were actually an assumed rlass interval of 50KWH. We merely
counted the number of items for each class interval and produced a summary
card for the interval or frequency step. We set up 51 steps; an illustration
of this is on pages 12 and 13 in the rear of the exhibits.
- Getting back to the flow diagram, once the detail bill cards
were summarized into frequency steps, they were retained in account number
order as we wanted to obtain a print-out of each customer's account for the
teet year. The frequency decks, 14 for the 'yes' group (the flow diagram
ahows 13, one was added-at the last munute). 10 for the 'no' group, and 1
each for the non-conforming groups were then processed on the 1620. These
decks of carda were white; the two non-conforming groups were merged with
their respective yes or no group and an overall summary group produced (for
yes and no). These groups were coded 10 for 'yes' and 20 for' no'; the color
of the cards, pink. We now had detail by appliance groups as well as su~
mary, both 'yes' :land • no' for all the cu. tamers who rep lied to the survey.
'~
A pink deck was also produced for the no-reply customers in
the same frequency steps. We had all the detail bill cards for our electric
water heating rate, 80 it was decided to include the non-replies as a single
group. The three pink decks, summary decks, were then merged into a single
overall summary deck (blue). This deck gave us a check on the number of
customers billed for each month to make certain we had all of the detail
cards. It WOuld also tell us if we hac to many.
The next pbase was the 1620 _program. I am not a statistician
by aDy mean., 80 tbat 1n produciDS the prosram I worked with the statistician
in the d.p.rtment. ae .uppli.d the formul •• n.c ••••ry to obtain the anawara,
and b. inc:Ldentally 1. the on. who is,pre.eutly 8na1Ylioi the reaults. My
',o1. fUDctioD wal to produce a machine m.thod which would provide answers,
thereby rell.vinl extended clerical calculationa,
The statistician .et down the formulae io atatistical form.
which I cODverted to my own langu8,e, These are shown on page 4 of the exbtbits; one look at some of the symbols and you wl11 understand why 1 used
my oWQ 1aquage •
<
Page 5 contains a flow diagram of the program. 6 and 7 a
i)'t!'1:tit-out and 8 a gloasa:ry of the terms used in the program. I believe they
a. aelf-explanatory and suffice it to aay the proaram.utl1izing the fre99
quency distribution data set up in a class interval 0 f SOKWH, an 88aumed mean
6'£ 225, provides the arithmetic mean, Itanclard deviation, standard error of
me8l1, and co-efficient of variation for the various appliance groupings.
~3 . .:L
c
C··,'"
'
..
Page a-A illustrates a sample output of the program. Column 1 is the frequency step. (class interval); colu1D11 2, the deviation from the assumed mean
(225); column 3, the deviation squaredj column 4, the frequency (number
of items by step); columo S, the frequency times the deviation and column
6 the frequency times the deviation squared. Columns 5 and 6 are summarized
by the prolram as they are used in the formula calculations.
Page 9 is an illustration of the four beader cards used in the
A set of these cards is necessary before each code being processed
by the computer.
program~
Pages 10 and 11 are card layout forms illustrating the fields used
in the input data cards and the output. data cards.
I am presenting this program to you to demonstrate one of the
many ways rate men can make use of the 1620 computer. This program was
originally intended to be a one-shot set-up. however, with certain
modifications,' I believe it could be written in a way to be of general use
to all of us.
1 0 ()
.. ...-,,.-----,-,,,-------.,
,~~-,-
,-,,,
I
C.:Ae [)
PLEASE ANSWER QUESTIONS BELOW BY MAKING A CHECK IN THE BOX MlO FILLING IN NUMBERS ON THE LlNFS
YOU CAN HELP US
kITCHEN APPLIANCES
HOME HEATING-CENT!!AL SYSTEM
IERYI YOU BETTER!
..!!M!U..
.nw..
~.
?;
..rI SlttGL£ - r 1 rwo
~ fIIIIILY ·"L.J .....Y
ELECTRIC
00[]
l1LCO GAS
!lID BOTTLED GAS
410
BOTTLED GAS
!l20
!I.G
540
IUFRIGERATOR
4 { ] ELECT"IC
LILCO GAS
610
620
OIL
440
':i~
4e[] SEiIM"Tl fREEZER
!Ie(]
"
t?lj./"H'~.C /"'n')-lIe.1 f (~nJLtlHtll"};l
S1CI71
.
r7/nr/
C!/Q.1S
,vUhliH'
:t:::L
,
j:'"
7P I-oI?-1
(!ktss
~
'e!Ue..,.,
,4SSU"., c~l hJ~c. ~/.
J)ey,d 1 ~)ot
X
-(';)01
,.",J.,.s.
.J)ev;C/.
-
-
S?
,~1rj..
/J?
~'>.7 tN' r . c(
/
.' "Ct,/
( :/"" J S
-lS
5
111
~
~
~:rZ/l1
/ It/t·,·~/<(.
~I
t/n/t'~(;
1
"-/1 I
.~
,i,'-"! J~{ <,
(
-t..--H
"re,·,
S
-I,
el '.'
Lie'
ft )
J
C
r
y
I
,
"
0::
3.
,. ~
~~~,(.')
IV
,0'
f7i1
- (U; ~)
i,
l-
D
,
=,
=
CO
I
I
-
~FM
=
I
I
V :;
4·
I
~
~
X
!
iV"~.e
!
!
i
!
-
i
i
I
I
I
,L}SI'}1
B.s
I
=
,~)
FPS
..c-
..;;:;;-
T
X'
Av
~.sA4 ~
..
(to
:f"])
.sf)
SE'M
~,7
!
•
. -. ..l)
ma. (
:1-
I
, ...
.-ff=
~1i~
WHV~J'
j-I~
ST) p
C]
X
d ')L
(el
's
)1 re- t;,//:
!11.- f'
/8/.
£
fUSS
I
~.
riJlt/: ~J~J-tt
/
C·
-~
~/.i'"1J
"ld~'~
« L~
.1'
IN
r=
Ie
C
I
''- <-.7' f'2i 5""::'
S{,
,5~( ;'/1/«(.. ~. ~J
)(= x'.,. t,:-.t
2.
1:"
S'Int IJ,('
!
I.
V-YC j ~ (i--Wl
'
"../"/t':-'/'/ , "1//'1
A
~.
5" ...,., <"",/
hi I/.:'J.' ~
1-,
/-,
- ---
__ 0
~"'hlb ~,- cl' (:
T~/r. i' 1//1111)
/~.$
!1 Je4.11
Sf"~~( 'un:! j ~"M~n
'$ a.. ... ~ C( ,-c( .. I'-f'1~"'Z'"r
o (ef?"t t>,euf
v'
I
-Frzu~ ~c.:t X -i' ~tS""" t< I, lot
+~lW6 'r<~.£, )( l ,,",y,/.f/,,">, s
...,-;:~/ e:Y (1
SCf/j'; ~rJ'
(11')105
"I ~J.f~S
~~~
t!> ,,t'
"
St"H.
7840.
9251.
7200.
14415.
4096.
9801.
5780.
11025.
3888.
1369.
4332.
6084.
6400.
3362.
•
•
•
2.
86.
3698.
•
•
•
180.
16020.
••
·C
••
••
••
••
•
14.
42.
84.
164.
150.
234.
273.
440.
360.
490.
473.
444.
624.
756.
810.
816.
714.
648.
684.
880.
693.
59-4.
667.
528.
550.
416.
405.
280.
319.
240.
465.
128.
297.
170 .•
315.
108.
37.
114.
156.
160.
82.
4.
1113.
••
128.
27.
32.
13.
*.
*.
*.
*.
••
••*.
*.
*.
*.
*.
••
*.
*
*
*
*
*
•
:c
*
*
••*
••
*
*
*
*
*
*
•
*
*
*
**
*
*
•*
8100.
323538 •
*
:C
lH;~,
.' 'f ,; .
~;;./~.
*
14,
C
1ll
·,··,·'i
'I'
.: 1 U II 0 0 I) COO 0 0 0 • 0 0 0 0 0 0 0 0 U 0 0 0 0 0 0 0 0 0 0 0 0 0 0 U 0 0 0 0 0 U 0 ~
I
•••
'
••
II II I~ ') It I'
I ., .. I' "
'1
'~:1 • I~ II 1 ,' ....
I
1 II
,,~
I
'. . . . . ,
~ :",
n non rt a n 'I
""""'IIIII'!tlll1l"'111111111!111'11I1:111:'"
'I
? 1 1 7 7 2 2 2 2 7 2 7 2 1 i ? 2 2 2 2 2 2 ? 1 2 2 2 1. 2 2 1 ~ ] , ; 1 1 'l :' 1 ~ , 7 2 2 2 2 :
11J J
1J113l 1l 131Jl 1l 3l 133J313J33J33333
~
331
j
I
J
1
)
U1r , n~
'I
,
. . . ' .' . . . . . '.
1
','1<1'
2 'J 7 : : -: i;' .
11 33 331 J ? 333
j .1
I
',i"
7 'I 1 ; '; ,
,
13333
~
»
j
? '} .' : }
.
., .
33333}3}333~
)
"I
1
,
.,
..
,
..,
,
J3)
'.4 •• 4444 •••• '444444.444.4.444 •• C444.4'.44.C~.4J.\4444.4.'"4"'~'4'444!4""~~j
S~5;~555SSS55S~5555SS~~55S~SJ5~))5)5)S5S~)S5~S55~5SSS)5"~~~5SSS~~5~S5SS~S5~J~~~~
ci666'11'111116&'&'i'&I'I&&6&6&16C6ii6666b&6~66666666~66&666~66~~;t6tfi66b6b666665b
.
a , 0 a C,a ~ 0 COO 0 0 0 0 0 0 0 0 0 0 8 0 3 0 0 0 0 n0 a 0 0 nco 0 0 0 0 0 0 J ~
'
,
., 4(.
4
•
0 C r DOC ~ 0 0 0 0 0 0 0 0 0 0 0 a ~ ~ 1 n U.~ :~
"C
'"
,
,,, F •
I
..
"
'.
"
'If ., 1:10
t
,.~.
I
'1 ) '1
l'
ceo 0 n 0
1
'
(4
: 1 ! 1 I l I t ! 1 11 t 1111111 11111 t t t l 1 I 11 t I l l I I 1 I , t I l I 1 1 I t t t I t 1 t 1 ; 1 1111 1 1 I 111: 11 1
22?~2272222Z2221i22222222~22122]~2222222']222222222221 222212212227?
1
1I , 1
':~:, ~::!')
3 333 3J J 33 3 13 3 J J 33 J 3333 333 33 J J J J 33 33 333J 33313 3 3J 33 33J 333 3 33 33 3 333 3 3J : 3 J
1
~
., :.
J 31 3" ;
4 4 ~ 4 '\ 4 4 4 4 4 4 ~ 4 4 4 4 4 4 4 4 .l 4 4 4 4 4 4 4 4 ~ 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 • 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 ~ ,~ 1 f 4 4 4 4 4 4 4 4 ~ t
5~5~555~:5~5555555S555555S55~5555555S5S55S5555~55555555S5S555S555555~555S55S555
e',
5E606666fE€b66666''6666666666666f66666i'''6666666666f6 66666£66666666~t.66666666;J
'l.
A SSJt-1ED
100' 00000000001010100800000000000.8000.00000000000000000000000000000000000000000
• 1 J • ~ • ) • • " " I; I) '. I~ II" II 11.71 r. n,. ~ 1117 11 n )G l: l, Jl )4 1':> II J' • :II • • , 41.1 .. " t6 ~8~·:-!f~)~) ~I 2/ 1' 23 3 'i 3:, 35 ~/6 . .: :
3~9 39 3040 : : ~_ 3: 4 ~~ 4'35" 3:' ~~
7
5
! .."'-' . '..(....
;::(.~;.~...,.,/;...
,~/:,,~. >! . 'c '/
:.. ..~
,:,/..0.:/"'"
67
'I "
5°31 51
J~2
70
7.
5..30
£)';/,,# .::.'.,4£~
--
,,- - ._=Y',L_
~_ '-",'
' -~ !
"
~" ~'
~
I :-. " ,./ . .;
..v,,->.
,',
'
~ i ~:~
~)
•
,--'; 2)
•i
.. 1;';)- ,
•
~ ~~
000000:00000000 000 0 0 0 0 0 0 0 0 0 0 0.01 0 0 0 0 0 0 0 000 010 0 0 0 0 0 000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00000008 I I-I
4 5 6, I • • , , " " '''4''18 ,,11819 20 '12' 23 24 25 1121 28:291>1 3t "33" 35 36 3138 39,".' " ........ " .... SO 51 " " 54" iii "58" 10"12 ...... " 12" ",. II 12 n 14 II II Jt.
i
~.r!..!.!:,
; ,...'-...
~. -/
'.. '
:•
i '" e
'I
,.
~
222222122222222222 2/2222 22222 2222 22 2 2 2222'222 222 222222 2222222222222222 2 ~ 222122121221222
~
33 333313 3,33.33 3 3 3 3 3 3i3 3 33 3 3 3 3 3 3 3 33 3 3 3 3 3 3 333 33 3 3 3 333 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3333 3'3333313313133133'3
~
44 4 H 4144 4, 4 4 4.4 4 4 4 4 414 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4.4 4 4 4 4 4 4 4 4 4. 4, 4t'l.4 4144.4
~
5 5 5 5. 5515 .5-~ 555 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5'5 555 5 5 5 5 5·5 5 5 5 5.55551
IJ
6666 6 616 6 6'6 6,6 6 6 6 6 6 666 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 666 6 6 6 6 6 6 6 6 6 6 6 6 6 66 6 6 6 6 6 6 6 66 6 6 6611111". II.fl I.
I
'
I
,
111111111111111111111177 7 777 777 71717 7 7 7 1 1 7 1 111111 1 1 1 1 11 1 7 7 7 7 7 7 7 1 7 7 7 7 7 7 77 7 7 7 1 1 771/7 7 ~ 7.7t~ 17
I I. I I 81118 I I 8 8.8. I 81818 8·8 8 8 8 8 88 81818 8 8 8 888 8 88181888,88 88888888 88888888888 I 8 8 I I II I I I I . I~"
t
9 9 9 9 9 91919 9 9 9.1 9 9 9 9 919i9 9 9 9 9 9 9 9 9 91919 9 9 9 9 9 9 9 991919 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 99 9 9 9
II."
1 2 3 4 5 8171 • • 1811121314151817 ,.:,.28 2122 23 24 25 26 27 28 lis 30 31 32 33 34 35 3& 37 38 39 40~1.42 43 44 45 46 47 48 49 50 51 52 53 54 55 5& 57 51 59 6811 &2 63 5415.11 • • 711112 n 741Js1Ja nlJa 11.
'1'1' '1'1'
~ c::r:::>CI:>c:::!::>c:I:J c:::!::)D::)C::Z:::>, ~CL) CJ2::) CJ:D c:::!!::>OD c::::!D CE:>CE::>c:::!DCJDCE::>~C!Dc::i!:)c!c)CE::>c::u::::>(' .~
MAU SIIISI...
I lsI
I I
PI".
.
,
2. I
PIS".
COMPLETE THESE SECTIONS FOR
. .AIICH Off ••• __ _
nANSMI •• _. __ •
CUSTOMUNO.
::::107 YES
aIS~IIAMII
0 1100
CVST_ • _ _ _ _ DATI
fIOOn TO
~
,"
• , ' ........
{
'.
.'
~,.:',I
,. • <
J.:;.~
" 1 1_
/",.
LAYOUT .WE'Al5O IT
_
I
Is thi~ a. rIYision of. or replacement for.
YES 0
an ulstlnllBM cad form?
If (YES). indicate
present
IBM card form no.
••
.,
May the Illstlnl pnntlnl plate be scrapped? YES 0
If (YES).indicate authority for scrappinl:
1_
.LL! _ _ _
u~_!
DAn
POIII'''' fOIl '} I) OITI.MIIIISHOIIT CAIO COLUMN NUM.ElS IT UfEIIING TO GUIDI ... ND COLUMNS
S_T CA.D
'.INlIO, .now. USAIl( COLUMN lANGE IS ESTA.USNED IT THI C... , ... CITY LOC ... "ON
LAYOUTS
OIlIACN 5101 O. GUIDI. E. G" ON ... 51-COLUMN CA.D, USUA.U COLUMNS .... E 15-.S.
_.
•
___
AUTHOIITY
...... __
NO 0
NO D
DATE
i
___ .................
~
6
.A .....
CORIIERS TO IE CUT
IIOILE
0
UPPER
Q
0
LI
"In
LOWEI
0 0
i
J
.
.... _______ ._......
II
M-t.' ...
COMPLETE T H E S E _ ,
SECTIONS WHEN APPLICABLE
_
. . . . . ._ _ _ .._
.... _
• • _____
===
::-..=.
. INTERPRETATION
R£GUII£.£IITS
.!!I.!!!I
lilliE
0
0 SeORllII
0
0
45 :";'I~~'I
0 ~~~~CI 0 PIEII.I£lIIIIO
0 c.--.----,
ILPU·
INDICATE ON THI LAYOUT THI LOCATlOII
60 .IIIIlL
552 D
557 0 (Of IACN fEATU.I S.IClfIED nOVI., )
:~~:·:.tl.~~~:::~.::.::: IILL FEED _ _
IIC'O STIli D PAID e••TlILHIS 0
CII.. "IIIIU.".I1.14 Ell ..11I1I1IIsnl.0 SPEc.O If can ISTlIE PIIIUI}
1"10 c:=>
call "III '11 11I"'n Win
II IITI SillS 1IIICaTE
"IILII "PI FEU'. (C·3 II
.IUllllln.1
SHORT CAID COLUMN GUIDI:
C.... D c ... , ... CITY (COLUMNS)
3 4
2 3 4
II
ALL LAYOUTS
~:~~TJ::
TYrE.lnn
IPAC'"
0
lin
L1ln
"
STYLE.
.
T•• IU
SE~. CI!'. --.!I..!!_
----
-
_-"~I=-·""""~··I
c 4::> ~
~
L1.II"
2) CUT OUT THE snIP Of COLUMN 'NUMIIIS.
3) ''''STE THE STRIP ON THE L...YOUT fO.M IN DESIIED LOC ..."ON fOI THE SHOIT CA.D.
4) .ULE ENDS Of SHO.T C.... D 2 1/2 COLUMNS InOND THE snIP Of COLUMN NUMIIIS.
.
C;UT-OUT5 fOI
• ASTIN.
I
I
fIJ
COlUMII.UIOI
__
12
I,1 1,1•1 1 111 1 1 1 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 il11 1 1 1 1 1 1 1 1 111/1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 }1 1 1 1 1 1 1 11 1 1
00
Hm
-
16
'-~~.
I
~
15
",
iO.
FORM X74-4049-12
IBM CARD LAYOUT FORM 1
0
4
c4=>
5) DUEIMINI MAIIt SENSE 'OSITIONS. If ANT. IY ASSOCIATING STI" IIUMIEU WITH CAID
L...YOUT .OSITIONS. I. G .• ON A 51- COLUMN CAID. M. S.• OSITIONS .-22 AWE USAlLE.
.
~
'-----"
4
II
SCALE: A,.IOX. DOUlLE
SIZI.
,
ACTUAL CAID SlZI:
3 1/4"):( 7 :! /1"
1I 1
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 6i 64 65 66 67 &8 69 70 71 72 73 7( 15 78 1111 79 •
9 10 11 12 13 14 15 1& 17 18 1920 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 4243 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 68 61 62 63 64 65 66 67 II 69 7071 72 73 74 75 7& 1111 79 •
e
¢
r?
J/.3.;l
o
IBJ4
n
~
IBM CARD LAYOUT FORM z.
~~:~:I~~
I
TL.!.2
•
3
..
7
8
9
10111213 Jot 15
91011
8
16 17 18 19
12
13
...
"0
!.--
:::r·/:""'~: _. '.#.~
i
il5145
!
~I
.
-
I
;
EtV/h7 c/v
~
I
~
I
'c£E~:"E.A./.-/'
~.-:... J ..'='#"'p" ~,:.r-.I
.
[)EV/H//C-"
~
; I
i
..
FRF~~r"/cy
;
I
I
---.
rH ..~: .. _~r C,,:, .....r~· -- _!~.C.::..,
I --.
./...
c
y
,-££~l/.... / o / C ·
~.e£";:·:'/£.-I,
..c'yY
-e
.,,p.~.,~/,.
'Z
:
;
;>~ 0 0 0 ~ 0 0 0 o'i~ 0 0 0 0;' 0 0 ;i.1~;; ~ 0 0 ~ io 0~~foo)o 0 0 0 ; io 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 / :
1m
'
20212223,2.1 25 26127 28'29;30 3132133'34'35.36:373839.40 41"421.t].44 45 4' 47 48 49 50 5152,535.55.565758596061626] ' 4 6 5 6 6 6 7 ' I 6' 70 7 1 7 2 7 3 ' 4 7 5 , , 7 7 7 . 7 . 1 0
15
16
17-181'92021
2324
3031323334,353637383940414243
....
4514,.7
.. ,
49
50
51
52
53
54
S5
56
57
51
5.
60
12
25
26
28
33
35
36
"S
I
,$/cp
It-
FORM X74-4049-12
i
1 2 3 4 5 6 7 i 8 9 1011121314151617 18 1920 2122 23 24 25 26 27 28 29;30 313233 34 35 36 37 38 39 4041424344 45 46 4748 49 50 51 /52 53 54 5556
.. 5758 59 &0 61 2!&3 64 65 66 6768 69 70 111213141511771111.
ill
1111 11 111 11 11 11 111,1 1 111111 1 111 :11 1 1 1 1 1 1 11 1 1 1 1 1 11 1 1 1 1 1,11 111111111 ':11 1 1 1 11 11111111111
121
222222'2222222222212 12222222222 2!2 22222 222212,2222222222222222222222'222222222222222222
~
3 3 3 3 3 3 3 3 3 3 3 3 3 3 33 3 3 3 3 3 3 3 3 3 3 3 3 3~3 3 3 3 3 3 3 3 33 3 3 3 3 3 3 3 3 3 3 3:3 3 3 3 3 3 3 3 3 33 3133 3 3 3 3 3 3 3 3 3 3 333 333
iii
j44 444444444444 44 4 4 4 4 4 4 4 4 4 4 4 4 4i44 4 4 44 4444 4 4 4 44444444:444444444444144 4 4 4 4 44 44 4444 44 44
~
15 5 5 5 5 5 5 5 5 5 5 5 5 5 55 5 5 5 5 5 5 5 5 5 5 5 5 515 5 5 5 5 5 5 555 5 5 5 5 5 5 5 5 5 5 51515 5 5 5 5 5 5 5 5 51515 5 5 5 5 5 5 5 5 5 5 5 5 55 5 5 5
no
!6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 &61 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 66 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 &6 6 6 6 6 6 6 &&&6 6 &&&
~
1777 7 7 717 7 7 7 7777 77 7 7 77 77 77 7 7777 77 7 7 77 7 7 7717 7 77 7 77 7 7 7 7 7 77 7 7 7 7 7 7777 77 7 7 7 7 7 7 7 7 7 7 7177 17
~
go
18 8 8 8 8 818 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 88 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 a8 8 8 8 8 8 8 8 8 I 8 8 8 8 8 8 88 ••• 8
~
I
I
.
[j]L19 9 9 9 9 91 9 9 9 9 9 9 9 9 9·9 9i9!9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9919999999999999999999-9999999999999999999999
1 2 3 4 5 .,7 8 9 10111213141516 n18:19 20 2122 23 24 25 26 27 28 29 30 3132 33 34 3536 37 38 3940414243 44 45 46 47 48 49 50,51,52 53 54 55 56 57 58 59 &0 61
&3 64 65 66 67.69 70 7t
n 13 74151177 7111.
JIil
COLUMN GUIDE ~c::r::::Jc::J::)CI:>CI:>C::OCL:>CI:)CL:>C::OCE:>C:::U::::>c:::!C) C!E::) CE::>ODCE:JODCJ!:)CE:>CE:>CE:)CE:>CE:>OD~C!!:)C!f:)
FOR MAli SENSING
'0._"
0_0 ,
_
1111";111,111'
nsiilils I
COMPLETE THESE SECTIONS FOR
IIANCH OFFICE NAME
YlANSMITTAL NO_
CUSTOMEI HAME
CUSTOMER NO.
CUSTOMER IEfElENCE AND DATE
::g~r:ED7 YES 0 NO 0
MAIL
'IOOFS TO
-,2', ;,u I~(t •. ,",4
LAYOUT 'U'ARED
'OINTERS FOI }
jjI\.
Is this a revision of, or replacement for,
an existinl IBM card form?
NO
0
NO 0
:ji.. JL:-~
• 'DAh
AUTHOIITY
DATE
CORIIERS TO IE CUT
1I0ilE
0
UPPEI
Q '~I
LOIER
0 0
INTERPRETAtiON
1I0ilE
0
0
45 :";,,~lill 0
~:~~~r 0
60 ~~~~:aL
552 0
557 0
:~;'~:·:.:·,~::,m.:·::: IILL FEEO _ _ uln TI 11110
::::'.::C:E.::E:.~~:r~I:I:: ENO PRIIiTlIIG nlt·D sm.O
:~~~:~.m:.,fm
•.
IC-I /I
m'c'l::
nnr.'JTEI
2' CUI OUT THE snIP OF COLUMN NUMIUS.
'ASTE IHE STRIP ON THE LAYOUT FORM IN DESIIED LOCATION FOR IHE SHORT CARD.
4' IULE ENDS OF SHOIT CARD 2 112 COLUMNS InOND THE STRIP OF COLUMN NUMIEIS.
3'
16~1~!161.15~1.5~15~1~!15~14~1.4~14~14!1"~139
3~13!1~!13~12~12~12~12!1
66 64 62 60 S8 56 54 S2 50 48 46 44 4240 38
36 34 32 30 28 26 24 22
3 4
1 2 3 4
0
May the existinl printinl plate be scraPlled? YES 0
If (YES),indicate authority for scrlppinl:
11 DETERMINE SHOIT CAID COLUMN NUMIUS IY IEUIIING TO GUIDE AND COLUMNS
'I/NiED Inow. USAlLE COLUMN lANGE IS ESiAIUSHED IV THE CA'ACITY LOCATION
ON EACH SIDE OF GUIDE. E. G., ON A 51. COLUMN CAID, USUAlLE COLUMNS ARE~r-65.
CUT-OUTS FOR
PASTING
YES
If (YES),indicate present IBII card form no.
'.
COMPLETE TliESE
IDI~''':''I
SECTIONS- WHEN APPLICABLE
ALL LAYOUTS
SHORT
CARD
COLUMN
GUID~
CARD
CAPACITY
(COLUMNS)
~~":ICI 0
REQUIREIIENTS
SCIRIIiS
0
G
PREI••• ERIIiS 0
::::::'~I ·'0
:::::m::t. .
(:FD~~~!: :::I~~~ ~~~~I~:E:H!I~;~TiON)
STUI
IF
0
can
PAO 0
C.ITlIUIUS
II TI IE '1I1lEI )
lilli, I •• ICI!E
::,~:TI
TI••
. . .1
U
.!~!!!1~!~Cl!S.:.!!!A!EJ!!!t.
., ~-
C:::;·;;4?
0
0 .c::::>
0
.
.1
22 24 26 a8 3D 32 34 36 3S 40 42 44 46 48 50 52 54 56 58 60 62 604 66
.... .
~
c4::::J c::::>
4 ~
5' D"UMINE MAU SENSE 'OSITIONS. If ANY, IY ASSOCIATING SlIIP NUMIUS WIIH CAID
LAYOUT 'OSITIONS. E. G .• ON A 51· COLUMN CAID. M. S. 'OSITIONS .-22 AlE USAlLE.
~ IZ~12:1 z~I~:13; 1~~13:13~13:14 ;1~~1~:14~14:15~15~15~1 !5~ 15:1,6~1,6~1~:1
......
I
SCALE, " " 0•.
SIZE.
ACTUAL CARD SI;
31/4" X
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
7 8 9 1011121314151617 181920 2122 23 24 25 26 27 28 29 30 31323334 35 36 37 38 39 4041424344 45 46 47 48 49 50 51525354 55 56 57 58 59 60616263 64 65 66 6768 69 70 71727374757677 78 79 80
, /'
~.E,./J
MEETING ASSUMPTIONS OF,HOMOGENEITY OF VARIANCE:
NONLINEAR DATA TRANSFORMS AND BARTLETT1S TEST
By
Richard C. Irons
UoS. NAVAL SCHOOL OF AVIATION MEDICINE
UoSo NAVAL AVIATION MEDICAL CENTER
PENSACOLA, FLORIDA
USERS CODE: 1159
. 113
,,' !f.D
C
INTRODUCTION
When working with various statis~ical routines, an investigator may have reason
to suspect that there are substantial differences in variance among the sets of scores or
observations
0
Under these conditions he may wish to perform Bartlett's Test for
Homogeneity of Variance.
This test is usual Iy performed on the original data, and if the variance is found
to be heterogeneous, then a nonlinear transformation on the data can be accomplished,
and Bartlett's test done a second time. The process is usually repeated unti I a transform
is found which will produce variance in the sets of data which does not depart significantly from homogeneity.
In programming the above operations on the IBM 1620 computer the slowest
operation is in reading the data into the computer. Consequently, in the present work
it has been found advantageous to read in the original data, per.form several nonlinear
c~'
transformations, and to punch out a card containing the original score and the transformed scores. In addition, summary cards which contain the number of cases, the sums
of the scores, and the sums of the squared scores are punched out for the original data
and for each transformation of each set of scores. The four transforms used are discussed
by Edwards (1). They are: 1)
X + 05, 2)
X+
X + 1, 3) Log of (X), and
4) the reciprocal of (X) where (X) is the original score.
The input for the Bartlett test is the summary cards for the sets of scores, and the
test is performed for the original data and each transformation. If the variance of sets
of scores is found to be homogeneous for the original data, the investigator need not
concern himself with the tests for the four transformations. If, however I Bartlett's test
1 1 It
JI. '-/. /
on the original data shows the variance of. the,10346
ld0488
AI0644
AI0778
AI093C
A10974
AI10S4
A11106
Al1162
Al1208
Al1256
Al1318
Al1358
t.,11422
A,11546
411546
00162
7
12 n
~.~.1
TRANSFORMATION PROGRAM SYMBOL TABLE
i~~
,~Dl)RESSES
19999
19989
19979
19969
199?9
19949
19939
19929
19919
19909
19899
19889
19879
19869
19859
19849
19839
19829
19819
19809
19799
19789
19779
19769
19759
19749
19739
19729
19719
19709
19699
19689
19679
19669
19659
19649
19639
1')629
19619
19609
19599
19589
19579·
19569
19559'
19549
19539
19529
19519
19509
19499
194'89
19479
19469
NA~~E
D I ~"iENS I ON STATE LOC
FORti:
SIN
Loe
FIX
FLOATING
A
~
A
...
SINF
COS
COSF
ATAN
ATA,NF
EXP
EXPF
LOG
LOGF
SQRT
SQRTF
M
A
A
A
A
A
A
A
A
A
A
SIGN
A
SIGNF
ABS
ABSF
RDNM
RDNMF
A
A
A
A
A
A
0909 11422
CON
43429448 00
SUMX
00000000 99SUMX2
SfVlXA 2
Sr"1XA
5iviX B
SiviXB2
Sf·1XC
srviXC2
SMXD
SMXD2
A
A
A
A
A
A
i~
A
'~J'
A
A
A
A
A
A
A
0022 07346
0001 '09898
A
NOGRP
KOUNT
A
A
A
0001-
0023 07394
A
OOOZ 09920
A
NutviGP
NOCAS
LAST
A
A
A
A
00000021 07454
A
A
A
A
A
0003 09948
XRA\-J
KASNO
000
001
A
A
XA2
500COOOO 00
XA
XB
A
A
A
10000000 01
8
1. 2 1
A
¥/I. ?
(~"
.,;
.,.. ....r..w
un
n
:·
..
_····VT·_JWiWj-
ttibiittH&iffbkWwHti6rt;;*t'hfwrlHrWHidWiirifttt"tdd d\H"uH titiHribz"""·".tTj
t»
bit
Hridf··riHr»tftt6···rit
"HI
tt t···. tttttt 'M
"ME ·wHlP. u!I!lHb'¥t1l
i !!W,' .,tIfll"¥t'f"w"W?
t :"' lIfi·'<'·',ltt·"MW±W*"+t'tur"
,hI,
TRANSFORMATION PROGRAM SYMBOL TABLE (Cont'd)
(i~,
.~
C·"
"",,"
19'459
19449
19439
19429
19419
19409
19399
19389
19379
19369
19359
19349
19339
19329
19319
19309
19299
19289
19279
19269
19259
19249
19239
19229
19219
XB2
XC
X.C2
XD
A
~
t::.,
/I
M
0006 C8106
A
A
A
XD2
NUMTR
0013 08690
0007 08390
A
A
A
A
A
/4,
A
A.
0009 09976
0002-
0041 10346
OOC3OC42 10488
0004-
l:'
0043 10644
A
0805-
/->..
0044 10778
A
A
A
A
A
A
0015 08890
0014 08794
0026
0027
0028
0029
0031
OC32
0033
0016
0017
19'2~9
19199
19189
19179
19169
19159
19149
19139
19129
19119
19109
19099
A
0004 08034
0005 07954
CASES
11054
11106
11162
11208
11\256
11318
11358
10116
10250
f:.,
A
,4.
A
A
.A
0019 09102
A
A
0018 10302
0045 1093C
A
A
A
A
0024 09866
0025 10974
OC092
o
9
1 ?... ?-
t
,~..
I
TRANSFORMATION PROGRAM INPUT DATA
~
2
1
6
24.
41 •
37.
40.
37.
·32.
6
2
40.
36.
29.
44.
38.
35 •
3 17
35.
38.
39.
49.
58.
35.
36.
54.
52.
40.
43.
38.
42.
38.
44.
46.
37.
4
227
231
233
204
206
236
209
222
230
237
239
240
205
207
211
212
213
215
216
217
218
220
221
224
225
232
234
238
251
10
30.
34.
44.
36.
43.
48.
25.
42.
44.
32.
4
5
29.
31 •
36.
28.
6 12
31 •
33.
46.
39.
40.
33.
37.
44.
46.
45.
35.
25.
5
7
39.
45.
53.
37.
48.
8
203
208
210
214
219
223
226
228
229
235
9
10
247
256
261
262
246
249
258
253
254
255
259
260
263
265
266
243
242
244
245
291
248
10
11
\
12
5
47.
35.
24.
50.
42.
5
33.
37.
37.
46.
35.
11
36.
28.
37.
38.
32.
44.
18.
33.
46.
35.
35.
5
41 •
38.
38.
36.
27.
3
41.
48.
·26.
241
252
257
258
264
.~.
~v
270
274
275
288
290
267
269
271
272
278
281
282
284
285
286
289
268
273
276
280
283
A".....·,
I
".",
277
279
287
123
c
J./-.4/0 ,
f -
TiKU.IIT&HritH u *#'W.· rib
riri'# ribt
w#ifrittRb±'F H' if 'WW"±rlitritriftntt'
*
j
t tt ttt
t t
Hd t $ "t,' '2
I
MI'U""
,ti \
:
j't
'tt" t'lt tr'j:
'U"
W!'WWi:T:!!IWt:Il:W*=Mi:trIe\".• .e" ....'+ei.='·+Mtb"WIh....'''mb\lfi\W,''fdM'ritWt'N¥W'.'"m'+m'f'lJ'i'I.'~1fI}1.,,'$wru '!!i(l'·@,J~~;..,
TRANSFORMED SCORES TABLE
C;'
A=GROUP NUMBER
B=SUBJECT NUMBER
C=RAW SCORE
D=SQRT(X+.5)
E=SQRT(X)+SQRT(X+1)
F=LOG(X)
G=RECIPROCAL OF X
A
C'
';
C\
".,1
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
4
·4
4
4
4
4
4
4
4
4
5
5
5
B
227
231
233
204
206
236
209.
222
230
237
239
240
205
207
211
212
213
215
216
217
218
220
221
224
225
232
234
238
251
203
208
210
214
. 219
223
' 226
228
229
235
247
256
261
5
262
6 246
6 249
C
24.00
41.00
37.00
40.00
37.00
32.00
40.00
36.00
29.00·
44.00
38.00
35.00
35.00
38.00
39.00
49.00
50.00
35.00
36.00
54.00
52.00
40.00
43.00
38.00
42.00
38.00
44.00
46.00
37.00
30.00
34.00
44.00
36.00
43.00
48.00
25.00
42.00
44.00
32.00
29.00
31.00
36.00
28.00
31.00
33.00
D
4.9497
6.4420
6.1237
6.3639
6.1237
5.7008
6.3639
6.0415
5.4313
6.6708
6.2048
5.9581
5.9581
6.2048
6.2849
7.0356
7.1063
5.9581
6.0415
7.3'824
7.2456
6.3639
6.5954
6.2048
6.5192
6.2048
6.6708
6.8190
6.1237
5.5226
5.8736
6.6708
6.0415
6.5954
6.9641
5.0497
6.5192
6.6708
5.7008
5.4313
5.6124
6.0415
5.3385
5.6124
5.7879
E
F
9.8989
12.8838
12.2471
12.7276
12.2471
11.4014
12.7276
12.0827
10.8623
13.3414
12.4094
11.9160
11.9160
12.4094
12.5695
14.0710
14.2124
11.9160
12.0827
14.7646
14.4912
12.7276
13.1906
12.4094
13.0381
12.4094
13.3414
13.6379
12.2471
11.0449
11.7470
13.3414
12.0827
13.1906
13.9282
10.0990
13.0381
13.3414
11.4014
10.8623
11.2246
12.0827
10.6766
11.2246
11.5755
1.38021
1.61278
1.56820
1.60206
1.56820
1.50515
1.60206
1.55630
1.46239
1.64345
1.57978
1.54406
1.54406
1.57978
1.59106
1.69019
"
11
G
1.69~97
1.54406
1.55630
1.73239
1.71600
1.60206
1.63346
1.57978
1.62324
1.57978
1.64345
1.66275
1.56820
1.47712
1.53147
1.64345
1.55630
1.63346
1.68124
1.39794
1.62324
1.64345
1.50515
1.46239
1.49136
1.55630
1.44715
1.49136
1.51851
t 2 lJ
.041666
.024390
.027027
.025000
.027027
.031250
.025000
.027777
.034482
.022727
.026315
.028571
.028571
.026315
.025641
.020408
.020000
.028571
.027777
.018518
.019230
.025000
.023.2 55
.026315
.02380~
.026315
.022727
.021739
.027027
.033333
.029411
.022727
.027777
.023255
.020833
.040000
.023809
.022727
.031250
.034482
.032258
.027777
.035714
.032258
.030303
I./: Jf. / /
TRANSFORMED SCORES TABLE (Cont'd)
250
253
254
255
6
259
6, 260
6
263
6
265
6
266
6 243
7 242
7 244
7 245
7 291
7 248
8
241
8 252
8
257
8
258
8
264
9 270
9 274
9 275
9 288
9 290
10 267
10 269
10 271
10 272
10 278
10 281
, 10 282
10 284
10 285
10 286
10 289
268
11
11
273
11
276
280
11
283
11
12 277
12 279
12- 287
6
6
6
6
46.00
39.00
40.00
33.00
37.00
44.00
46.00
45.00
35.00
25.00
39.00
45.00
53.00
37.00
48.00
47.00
35.00
24.00
50.00
42.00
33.00
37.00
37.00
46.00
35.00
36.00
20.00
37.00
38.00
32.00
44.00
18.0U
33.00
46.00
35.00
35.00
41.00
38.00
38.00
36,.00
27.00
41.00
48.00
26.00
6.8190
6.2849
6.3639
5.7879
6.1237
6.6708
6,.8190
6.74'53
5.9581
5.0497
6.2849
6.7453
7.3143
6.1237
6.9641
6.8920
5.9581
4.9497
7.1063
6.5192
5.7879
6.1237
6.1237
6.8190
5.9581
6.0415
4.5276
6.1237
6.2048
5.7008
6.6708
4.3011
5.7879
6.8190
5.9581
5.9581
6.4420
6.2048
6.2048
6.0415
5.2440
6.4420
6.9641
5.1478
13.6379
12.5695
12.7276
11.5755
12.2471
13.3414
13.6379
13.4905
11.9160
10.0990
12.5695
13.4905
14.6285
12.2471
13.9282
13.7838
11.9160
9.8989
14.2124
13.0381
11.5755
12.2471
12.2471
13.6379
11.9160
12.0827
9.0547
12.2471
12.4094
11.4014
13.3414
8.6015
11.5755
13.6379
11.9160
11.9160
12.8838
12.4094
12.4094
12.0827
10.4876
12.8838
13.9282
10.2951
1.66275
1.59106
1.60206
1.51851
1.56820
1.64345
1.66275
1.65321
1.54406
1.39794
1.59106
1.6.5 3 21
1.72427
1.56820
1.68124
1.67209
1.54406
1.38021
1.69897
1.62324
1.51851
1.56820
1.56820
1.66275
1.54406
1.55630
1.3p103
1.56820
1.57978
1.50515
1.64345
1.25527
1.51851
1.66275
1.54406
1.54406
1.61278
1.57978
1.57978
1.55630
1.43136
1.61278
1.68124
1.41497
.021739
.025641
.025000
.030303
.027027
.022727
.021739
.022222
.028571
.040000
.025641
.022222
.018867
.027027
.020833
.021276
.028571
.041666
.020000
.023809
.030303
.027027
.027027
.021739
.028571
.027777
.050000
.027027
.026315
.031250
.022727
.055555
.03,0303
.021739
.0285-71
.028571
.024390
.026315
.026315
.027777
.037037
.024390
.020833
.038461
(10'
~
,1""-"",
~~#J
c
12
J .,....t
~
S%ff,,·,...ffttbHj-· in
j· ..
···'jfifijiftiv - r ..!ff["Tffil")" .. ·'"ffft
.... r·N ... t "r -ff·"
• ••
•
r·
.
II"
_... r""I!"-··-lflrttH.r
··5izftf··.. ··,'t... irit'¥tlt*t.. tFri·6tW...
··biritfrihtri"frif¥rl4 tH±" ht ttzrittZttf""Yi tHtt"i±H.·.· dt-·,#It.
INPUT DATA FOR BARTLETT PROGRAM FROM TRANSFORMATION PROGRAM
t t
j.
IIHUtt
----_._............ ,_... _..... .......
_
... _._.........._-
_
INPUT DATA FOR BARTLETT PROGRAM FROM TRANSFORMATION PROGRAM (Cont1d)
12
N
~,
• (j 0
6 .·00
17.00
10.00
4.00
12.00
5.-00
5.00
5.00
11.00
5.00
3.00
SUMX,
.17
.16
.41
.27
.13
.32
.11
.13
.13
.34
.14
.08
SUM X sa
X=RECIPROCAL OF X
.005393
.004611
.010120
.007899
.004276
.009261
.002671
.003972
.003668
.·012297
.00412i
.002508
5
5
5
(
,-"
."
5
5
5
5
5
5
5
5
5
5
f'--".
'...._-,./
1>4
- --ljijMwil"ijj']WIl-r- ·"00""]" "2"C-·
f
!"-WrY"" -N . -
bi#I±i.iti. . . ··•... *rli'#±h*btthHti:#tt.t
'it b.· ttt
:H
THE BARTLETT TEST FOR HOMOGENEITY OF VARIANCE PROGRAM
c
This program computes the Bartlett Chi-Square Test for homogeneity of variance,
degree of freedom, and the level of significance!)
In addition to these measures the
program provides a table conta ining many of the computational steps in performing the
Bartlett Test and gives tables values of chi square for the appropriate degrees of freedom
for the following levels of significonce: 0.10, 0.05, 0~02, 0.01, and 0.001
It
FORMAT FOR INPUT CARDS
Columns
EDIT
Example
Number of groups
0012
Transform Type
1,2,3,4, or 5
Number of cases
000000013.
1- 4
65
CASE
Informati on
1-10
Blank columns may be used in place of leading zeros in the datd fields for the
(:"
original data, and for each set of transformed data the .jnput consists of an ED IT card
and a CASE card for each group. These are followed by the cards containing the values
for the sums of (X) and sums of (X) squared. The groups of data can follow One another
directly.
OPERATING THE PROGRAM
1. Place the program in the card reader
0
2. Place the five decks of data cards produced by the transformation program
on top of the program deck.
30 Press the RESET button on the console, the LOAD button on the card
reader, and the START button on the card punch.
15
t ",MWMtthil "
....
__.._ __
...
..
--- -----
4. When LOAD DATA is typed, the program will automatically read the data
and process the problem.
50 When END OF JOB message and instTuctions for initiating anew run are
typed, NON PROCESS RUNOUT for the CARD PUNCH must be pressed
to obta in all the output cards
0
6. The output cards may be listed directly with an 80-80 board on a 407
machine.
PROCESSING TIME
The approximate computer time necessary to accomplish the Bartlett Program is
given by the following equation:
Execution time = 120 + 6K seconds
where K
= number of data cards per transform group.
1 2 !J
16
Lf. Lt.
I~-
(~
"'
IU.'.
ill I
~06
iOR
j 10
II
I~
II
11
I,
I·
FLOW DIAGRAM FOR BARTLETT PROGRAM
"I
0:
READ
CHI-SQUARE TABLES
(11
ZEROIZE SUMS AREAS
Ob
07
READ NUMBER OF
GROUPS AND FORM
(lR
DETERMINE
WHICH TRANSFORM
SOLVE FOR
CHI SQUARE
!
i
~-()
i
(II
READ VALUES FOR N,
SUM X AND X SQUARE
I
I.~
cl
III
I
i
I
PUNCH CHI SQUARE
AND
DEGREES OF FREEDOM
CONSTRUCT
BARTLETT TABLE AND
SUM VALUES
I'
FIND CHI-SQUARE TABLE ENTRY
NO
If>
PUNCH CARDS FOR LEVEL OF
SIGNIFICANCE OF CHI SQUARE
AND THE CHI-SQUARE TABLE ENTRY
FOR THE DEGREES OF FREEDOM
17
I
!~
i
PUNCH HEADER CARDS
!
19
2_~
~ __21
PUNCH
BART LEn TABLE
I
;
C~
J
,.-
.'1
j
,
23
,I
7,1
NO
i
~~
17
'BARTLETT PROGRAM LISTING
cr
DEFINITION OF SYMBGLS FOLLOW.
AC7C70
;",C7070
AQ7D70
A0707.']
AC7070
A0707C
A0707G
A07070
'-
c
c
c
c
MATRX= STORAGE LEVELS FOR CHI SQUARE TABLE.
S U H uF = S U i/, ; ,1 A T ION U F D • F •
SIGN = SUMMATION OF N.
SMRCP= SUMMATION CF THE RECIPROCALS OF D.F.
SUMS2= SUMMATION OF S·SQUARE.
SMSD2= SUMMATION OF THE SU~ OF X SQUARE.
SMDFL= SUMMATION OF D.F. TIMES TH~ LOG OF ~ S~UARE.
NOGRP= NUMBER OF GROUPS.
NU~TR= TRANSFORMED DAfA TYPE.
CHIDF= LEVEL OF SIGNIFICANCE.
KCUNT= A COUNTER FUNCTION.
LEVEL= LEVEL OF .::iIGNIFICANCE.
CASES= NUMBER OF CASES IN GROUP.
S U ~"1 X = SUM 0 F X.
SUMX2= SUM OF X SQUARE.
DF
= N-l, WHERE N IS NUMBER OF SU~JECTS.
RECDF= RECIPROCAL OF D.F.
S2
S SQU,ARE.
TLOG = LOG OF S SQUARE.
DFLOG= D.F. Tl~ESLOG' OF S SQUARE.
ONE
= FIRST EQUATION IN SOLVING FOR CHI SQUARE.
TWO = SECOND EQUATION IN SOLVING FOR CHI SQUARE.
Trl R E E = THI RI) E QUA T ION INS 0 LV I N G FOR CHI SQUA RE •
(HI
= CHI SQUARE.
LAST =COUNTER TO FIND END OF ROUTINE.
r
'-
c
c
c
c
C
,......
c
(
c
c
c
C
C
C
c
c
c
c
.t..0787 tJ
AC7070
A07:~ 7C
,£4.07070
AC7C70
A07070
A07070
A07J7C
,6.0707C
;",07070
A.07070
;",07070
A07078
A07070
A07070
A,07070
A07070
A 0 7:) 7 0
A0 7 0 7 0
A07070
,A.,0707C
r
PROGRAM BEGINS.
'-
c
DIMENSION T(5,30)
TR X = 1
CHI SQUARE TABLE IS READ IN AT THIS POINT.
READ 2,T(1,MATRX),T(2,MATRX),T(3,MATRX),T(4,MATRX),T(5,MATRXl
IF(MATRX-30)3,4,4
!\1A TRX=I,1A TRX+ 1
GO TO 1
CN=.43429448
PROGRAM HEADER (IDENTIFICATION) CARDS TO BE PUNCH NOW.
PUNCH 5
PUNCH 6
PUNCH 7
ZEROING SUMMATION AREA NEXT TO BE DONE.
LAST=l
SUMDF=O.
BIGN=J.
SMRCP=iJ.
SU~I1S2 =0.
SMSD2=u.
Sr"iDF L = U •
READ VALUES FOR NUMBER OF GROUPS AND TYPE TRANSFORM USED~
READ 9,NOGRP,NUMTR
ROUTINE FOR.IOENTIFICATION OF TRANSFORM TYPE.
IF(NUMTR-l)ll,l2,ll
~'1A
r
.....
1
3
4
C
C
8
C
C
~,--",
. '\".,"",,/
18
131
A0707G
A0707C
A07070
A07070
A07094
A07094
A07406
A07474
A075l0
AC75l8
A0754;:"
/>,07542
A07554
A07566
A07578
A07578
A07602
A07626
A07650
A07674
A07698
A07722
A07746
A07746
A07762
A07782
c
(
BARTLETT PROGRAM LISTING (Cont'd)
",",\'
rJ',
12 PUNCH 18
A'J785C
/:.,07862
GO Te 13
11
If- (NUf·1rr~-2)
14,1':"lL+
A0787C
A'J793B
A079r;,C,
15 PUNCH 19
(JC
TO 13
14
IF(NUMTR-3)16,17,16
17
PUi'~CH
A 0 7'-) ':) 0
21
A08026
GO TO 13
A08(:Jb
16 IFCNUMTR-4)22,23,22
23 PUNCH 24
GO TO 13
22 PUNCH 25
1.3
C
ACJd046
;'.(;8114
ACB126
A.08134
A08146
(HIDF=I\JC(lRP~l
'KOUNT=l
LEVEL=CHIDF
26 CASES=U.
SU~. 1X =~:.
SUi'
A
A
A
A
A
A
00000000 99-
A
A
A
A
A
A
A
A
A·
A
A
A
A
A
A
A
A
A
A
0002-
0003-
I
j j
JI. J/.. ;;. ~ .
C
X':'&,'"1-'
Itt
f"
-
-
-
fl
.
---J
"'2"-
-!
IWjij:d-""-
r--r··-Hij""'1¥[""f¥"Y'''-
ern -
---r-mmq"f"'"RUf"'TW"EW"f"
BARTLETT PROGRAM SYMBOL TABLE (Cont1d)
,I.,
e;\1
1796 q
0017 08026
~7959
17949
(~"\i
!,
C'\
0022 08134
0023 08114
A
A
A
0024 13354
0025 13456
A
CHIOF
KOUNT
LEVEL
I:..
A
A
A
A,
0026 08254
CASES
SUI"1X
sur"1X2
A
~.
0027 10856
DF
A
A
10000000 01
RECDF
SUMD2
A
f~
A
t,
A
002
S2
TLOGS
A
A
A
A
A
A
A
A
001
DFLOG
0028 08934
0029 08910
17709
17699
17689
17679
17669
l7659
l7649
l.7639
l7629
17619
l.76(;9
1,:.7599
1;7589
17579
l7569
17559
17549
17539
17529
17519
17509
17499
17489
17479
17469
17459
17449
17439
17429
A
A
0004-
17739
1" -.,-,
." ,',
I ';I L ';/
l7Y19
l7909
l7899
l7889
17879
17869
17859
17849
17839
17829
17819
17809
1 (799
17789
17779
17769
17759
17749
17739
17729
17719
A
0021 13230
0031 11146
\
0032 11302
0033 11476
0034 09154
0035 09110
A
A.
ONE
TWO
A
,A.
23026000 01
THREE
30000000 01
005
CHI
0036 11534
0037 11616
A
A
A
A
A
A
A
A
A
A
A
A,
A
0038 09718
0039 09698
0041 09678
0042 11886
0043 10354
0044 11968
0045 09874
0046 09834
0047 09854
A
A
A
A
0048 12042
0049 12116
.A.
0051 10030
0052 10010
0053 09990
A
A
A
0054 12218
23
A
1 ~'" f»
.If~. ~,--3
BARTLETT PROGRAM SYMBOL TABLE (Cont'd)
174\:~;9
17399
17389
17379
17369
17359
1.7349
17339
17329
17319
'
1 7309
.
1 7299
\
17289
17279
17269
17259
17249
A.
OC55 12320
17419
A
A
A
0056 10186
0057 10166
0058 10146
0059 12394
0061 12496
A
A
A
0062 10342
'0063 10322
0064 10302
0065
0066
0067
0068
0069
A
A
A
12570
12672
12746
10902
11048
A
A
A
A
0005-
0071 10746
0072 10790
0073 13546
A
A
A
A
00128
\
24
CHI-SQUARE TABLE
2 • 706 3 .84 1 5 .41 2 6 .635 10 .83
4.605 5.991 7.824 9.210 13.82
6.251 7.815 9.837 11.341 16.27
7.779 9.488 11.668 13.277 18.46
9.236 11.070 13.388 15.086 20.52
10.645 12.592 15.033 16.812 22.46
12.017 14.067 16.622 18.475 24.32
13.362 15.507 18.168 20.090 26.12
14.684 16.919 19.679 21.666 27.88
15.987 18.307 21.161 23.209 29.59
17.275 19.675 22.618 24.725 31.26
18.549 21.026 24.054 26.217 32.91
19.812 22.362 25.472 27.688'34.53
21.064 23.685 26.873 29.141 36.12
22.307 24.996 28.259 30.578 37.70
23.542 26.296 29.633 32.000 39.25
24.769 27.587 30.995 33.409 40.79
25.989 28.869 32.346 34.805 42.31
27.204 30.144 33.687 36.191 43.82
28.412 31.410 35.020 37.566 45.32
29.615 32.671 36.343 38.932 46.80
30.813 33.924 37.659 40.289 48.27
32.007 35.172 38.968 41.638 49.73
33.196 36.415 40.270 42.980 51.18
34.382 37.652 41.566 44.314 52.62
35.563 38.885 42.856 45.642 54.05
36.741 40.113 44.140 46.963 55.48
37.916 41.337 45.419 48.278 56.89
39.087 42.557 46.693 49.588 58.30
40.256 43.773 47.962 50.892 59.70
o
25
BARTLETT PROGRAM OUTPUT
c
BARTLETT TEST FOR HOMOGENEITY OF VARIANCE
i-<~FERENCE
EDVJhF~DS
REV.
t~D.,
PROGRAMMED BY R.C. IRONS
P.125-129
SCHAVMED PENSACOLA,FLA
DATA FROM RAW SCORES USED
SUM XII}
CASE
NO
1
2
3
4
5
6
7
8
9
1U
11
12
D.F.
N
S( I )
SQ
SO
I/D.F.
6.
6.
17.
10.
4.
12.
5.
5.
5.
5.
5.
16.
9.
3.
11.
4.
.20000
198.83330
~g.76666
.20000
~?R.OOOOu
.06250
• 1111 1.
.33333
2s.6000G
.~ -l .;, b 0 -::l.1
501.60000
38.00000
12.66666
.090 :;l.'1
4~5.6670G
4S.060~~
.25 CiU
171.20000
4.
4.
.25 (,; ()C
433.20000
11.
732.00000
5.
3.
10.
4.
2.
.250UO
.100(,0
42.800(0
108.30000
24.80000
73.20000
89.
77.
.25(;00
t.)
97 • 7650 G
5~.71l33
99.20000
28.50000
.50 f)OO
114.00000
252.66670
126.33335
2.59785
3762.13200
620.12096
CHI SQUARE;:
LOG S( I)
(N-l )lOG S( I)
SQ
SO
1.59951
1.40824
1.57241
1.74611
1.10266
1 .~r)-::l.70
1 • r,'~ } 44
2. • (: '346::'
). ,; 2 r:;, 77
d • l.~ ~, ~~ 1
.5 7 7SC
1•
~h) (l L.·;
!
• .s4i)~1
H4
~.819?-7
2.101c-,1
4.2030~
1•4
~ /~
DEGREES OF FREEDOM=
7.958
7.99759
7.04120
25.15856
15.71503
3.30798
]8.19177
]26.32178
1]
NOT SIGNIFICANT AT .100
LEVEL OF SIGNIFICANCE
.100
.050
.020
.010
.001
CHI SQUARE TABLE ENTRY 17.275 19.675 22.618 24.725 31.260
139
26
BARTLETT PROGRAM OUTPUT (Cont1d)
C.'·'
I"
DATA FROM SQRT(X+.5)
CASE
NO
1
2
3
4
5
6
7
8
9
10
11
12
SUMS
TRANSFORMATION USEb
1/D.F.
N
6.
6.
17.
10.
4.
12.
5.
5.
5.
11.
5.
3.
5.
5.
16.
9.
3.
11.
4.
4.
4.
10.
4.
2.
.20000
.20000
.06250
.11111
.33333
.09090 .
.25000
.25000
, .25000
.10000
.25000
.50000
89.
77.
2'.59785
CHI SQUARE=
SUM X ( I )
SQ
S(I)
LOG S(I)
SQ
SQ
1.58500
.88518
3.51742
3.54400
.33590
3.41997
.98702
3.05672
.64878
6.08836
.93662
1.79917
.31700
.17703
.21983
.39377
.11196
.31090
.24675
.76418
.16219
.60883
.23415
.8Q958
-.49894
-.75193
-.65789
-.40474
-.95091
-.50737
-.60773
-.11680
-.78996
-.21549
26.80414
4.44623
(N-1)LOG S( I)
SQ
-~63049
..... 04595
DEGREES OF FREEDOM=
9.386
-2.49470
-3.75969
-10.52633
-3.64273
-2.85273
-5.58107
-2.43093
-.46721
-3.15985
-2.15499
-2.52198
-.09191
-39.68417
11
NOT SIGNIFICANT AT .100
LEVEL OF SIGNIFICANCE
.100
.050
.020
.010
.001
SQUARE TABLE ENTRY 17.275 19.675 22.618 24.725 31.260
C~I
o
1~ 0
27
"""."
..
""-""-~~~,,-"--- ..
------..."-
... ~--."- ..
-- .. "''''
..•..... ".".".,,",,.,,......... ,,"" .. '"
.,.""'."--""""
8A~TLETTPROGRAM
OUTPUT (Cont'd)
DATA FROM SQRT(X)+$QRT(X+1) TRANSFORMATION USED
CASE
NO
1
2
3
4
5
6
7
8
9
10
11
12
SUMS
SUM' X( I )
N
I/D.F.
SQ
S(I)
SQ
6.
6.
17.
10.
4.
12.
5.
5.
5.
11.
5.
3.
5.
5.
16.
9.
3.
11.
4.
4.
4.
10.
4.
2.
.20000
.20000
.06250
.11111
.33333
.09090
• 25000
.• 25000
.25000
.10000
.25000
.50000
6.29000
3.73518
. 13.69920
13.85960
1.30360
13.58990
3.91808
12.18688
2.55512
24.26350
3.46542
7.16667
1.25800
.74703
.85620
1.53995
.43453
1.23544
.97952
3.0.4672
.63878
2.42635
.86635
3.58333
89.
77.
2.59785
106.03315
17.61223
CHI SQUARE=
LOG S(I)
SQ
(N-1)LOG S·(Il
SQ
.09968
-.12665
-.06742
.18750
. -.36197
.09182
-.00898
.48383
-.19464
.38495
-.06230
.55428
DEGREES OF FREEDOM=
9.548
NOT SIG~IFICANT AT .100
LEVEL OF SIGNIFICANCE
.100
.050
.020
.010
.001
CHI SQUARE TABLE ENTRY 17.275 19.675 22.618 24.725 31.260
\
28
.49840
-.63329
-1.07879
1.68757
-1.·0;,,8593
1.01005
-.0359 •
1.93533
-.7715'
3.84.~'
-.24921
1.10857
6.22769
11
T·tf··W·· .. Nzr·
BARTLETT PROGRAM OUTPUT (Cont'd)
C'
DA.TA FROM LOG(X) TRAN S FOR r·1 AT ION USED
N
NO
1
2
3
4
5
6
7
8
9
10
11
12
SUMS
S(I )
SO
SUM X( I )
CASE
D.F.
l/D.F.
SQ"
LOG S ( I ) (N-l)LOG S ( I )
SD
.05118
.03593
.07520
.08239
.01937
.07979
.03918
.08638
.01408
.18737
.01648
.05666
.01023
.00718
.00470
.00915
.0064:)
.00725
.00979
.02159
-1.98983
-2.14346
-2.32790
-2.03836
-2.18987
-2.13943
-2.00899
-1.66564
.OO35~
-?.4534 rj
2.
.20000
.20000
.06250
.11111
.33333
.09090
.25000
.25000
.25000
.10000
.25000
.50000
.01873
.00412
.02833
-1.72729
-2.38?10
-1.54770
77.
2.59785
.74403
.13109
6.
6.
17.
10.
4.
12.
5.
5.
5.
11.
5.
3.
5.
5.
16.
9.
3.
11.
4.
4.
4.
10.
89.
4.
CHI SQUARE:;:;
DEGREES OF FREE DO~v1:;:;
12.034
SQ
-9.94917
-10.71732
-37.24643
-18.34531
-6.56963
-23.53376
-8.03598
-6.66258
-9.81382
~17.27293
-9.54041
-3.095 /+0
-160.78279
11 .
NOT SIGNIFICANT AT .100
C
1
.,
.001
.050
.020
.010
.100
LEVEL OF SIGNIFICANCE
SQUARE TABLE ENTRY 17.275 19.675 22.61d 24.725 31.260
CHI
o
1 42
29
.. _.
·"j"··j"·.,'!W.tt,-1
BARTLETT PROGRAM OUTPUT (Cont'd)
(f~'
'V'
D~TA
FROM RECIPROCAL OF X USED
Cl\SE
NO
1
2
3
4
5
6
7
8
9
10
11
12
SUMS
N
D.F.
I/O.F.
6.
6.
17.
10.
4.
12.
5.
5.
5.
11.
5.
3.
5.
5.
16.
9.
3.
11.
4.
4.
4.
10.
4.
2.
.20000
.20000
.06250
.11111
.33333
.09090
.25000
.25000
.25000
.10000
.25000
.50000
89.
77.
2.59785
CHI SQUARE:::
sut". X( I )
SQ
S(I )
.00057
.00034
.00023
.00060
.00005
.00072.
.00025
.00059
.00028
.00178
.00020
.00037
.00011
.00006
.00001
.00006
.00001
.00006
.00006
.00014
.00007
.00017
.00005
.00018
.00603
.00104
22.221
LOG S ( I ) (N-l)LOG S ( I )
SO
SO
SCJ
-3.93829
-4.16199
-4.83907
-19.69148
-20.80995
-Lt.1696c~
-37-.52662
-14.30865
-45.91406
-16.80954
-15.31895
-16.57067
-37.47654
-17.17825
-7.45476
-4.76955
-4.17946
-4.20238
-3.82973
-4.14266
-3.74765
-4.29456
-3.72738
DEGREES OF F REEDO~'1:::
-77.4251~
-326.54467
11
SIGNIFICANT BETWEEN .050 AND .020
.001
.020
.010
.050
.100
LEVEL OF SIGNIFICANCE
CHI SQUARE TABLE ENTRY 17.275 19.675 22.618 24.725 31.260
00
t
';ttmtt \ tit",' It; \ kktt\>
,Hwt#t"t,
th I'
t"5
t t
REFERENCE
1 ~ Edwards, A. L., £xperimental ,Qesign
.!!! PsycholC?gical
Research. New York:
Holt, Rinehart and Winston, 1960.
o
J ,,, "
31
-=----
....=..-
~-~-'--=---.-~--
=~~=~=
- ,-', . _-,---------------
',1' . •
A. Gl.:NERAL, FUNCTION SlJBPltOOlWl
M. E. Munroe
Uhiversity of New Hampshire
~,
C
"
~
5. :, ;: '
A General Function Subprogram
C:'
by M.R. MUnroe, University of New Hampshire
This is a Fortran II function subprogram that accepts
program and returns
F(X).
of any desired function.
X fran the main
This single subprogram can be used to yield values
The function to be computed is controlled by data
read into the main program.
The code for describing a function by numerical data is as follows.
A
block of four integers will give instructions to combine to functions in a
specified way and store the result.
The first integer in each block of four
specifies the combining operation to be performed.
The next two specify the
basic functions (or previously stored results) that are to be combined in this
way.
The fourth integer specifies a label for storage.
The specific code
chosen is this:
Functions
Identity
Function
1. addition
11. log
[!(x) •
2. subtraction
12. sin
3. multiplication
13. cos
L.
division
14. exp
'5. composition
15. sqrt
(!(g(x)
16. atan
Library
Operations
D
6. integer
exponentia tim
Constants
~
30
Storage
101 - 115
21
17. aba
Every time a 30 appears there must be an auxiliary entry to indicate what
constant.
0)
Thus, the data consists of a vector of 4k fixed point numbers plus
another vector of floating point numbers - one for each 30 in the first vector.
/
J./. &.1
Illustration:
Suppose we want the subprogram to compute
,~
j
l\l.~7
sinnx + 3x5
•
The vector of integers would be as follows, and we indicate at each storage
stage what is being stored.
sinnx
30
)
21
12
101
101
3x'
3
)0
103
102
6
21
J sinnx + 3XS
Sinnx+3x'
1
104
102
104
105·
5.
3.
5
15
105
1<1>
The vector of constants would be
3.1415926
Note that the exponent
5 is read as a floating point number. The subprogram
will convert it t·o an integer before exponentiating.
A listing of the subprogram follows.
bers as branching controls
The idea is to use these code num-
qy means of IF statements and computed GO TO state-
mente.
1 ~7
¥.5.~
C
n
C
c
C
3
C
C
1
C
4
C
C
2
C
41
(~.
C
-,)
C
42
C
43
44
45
46
47
48
49
GENERAL FUNCTION SUBPROGRAM
FUNCTION FCX)
o I MENS ION I (200) ,C ( 100) • FF C15)
COMMON I.C.NI
INITIALIZE CONSTANT COUNTER
KON=l
DO 1 00 J= 1 • N I ,4
IJ=I (J)
PREPARE FPR FIRST PASS (THROUGH SECOND FUNCTION)
L=2
S=X
M=J+L
IM=I(M)
BRANCH FOR FUNCTION. CONSTANT, STORED RESULT
IF ( I M-30 ) 1 .41 ,2
BRANCH FOR LIBRARY, IDENTITY FUNCTION
IF(IM-20)4,4,42
LIBRARY FUNCTION
K=IM-lO
BRANCH TO SPECIFIC LIBRARY FUNCTION
GO TO (43.44.45,46.47.48,49).K
STORED RESULT
K=lM-IOO
R=FFcK)
GO TO 65
CONSTANT
R=CCKON)
RESET CONSTANT COUNTER
KON=KON+l
GO TO 65
IDENTITY FUNCTION
R=S
GO TO 65
SPECIFIC LIBRARY FUNCTIONS
R=LOGFCS)
GO TO 65
R=SINFCS)
GO TO 65
R=COSF(S)
GO TO 65
R=EXPFCS)
GO TO 65
R=SQRTF(S>
GO TO 65
R=ATANFCS)
GO TO 65
R=ABSFCS)
\~
C
65
C
67
C
C
C
11
C
C
66
C
31
32
33
34
35
36
C
70
C
C
C
100
BRANCH ACCORDING AS THIS IS FIRST OR SECOND PASS
GO TO 166.67,.L
STORE RESULT OF FIRST PASS ASZ
Z=R
PREPARE FOR SECOND PASS (THROUGH FIRST FUNCTION)
L=1
START SECOND pAss UNLESS OPERATION IS COMPOSITION
IF ( I J-5 ) 3. 1 1 • 3
RESET S FOR COMPOSITION
S=Z
START SECOND PASS
GO TO 3
AFTER SECOND PASS BRANCH TO OPERATION CALLED FOR
GO TO <31.32.33.34.35.36>.IJ
PERFORM OPERATION AND RECORD RESULT AS F
F=R+Z
GO TO 70
F=R-Z
GO TO 70
F=R*Z
GO TO 70
F=R/Z
GO TO 70
F=R
GO TO 70
lZ=Z
F=R**IZ
NOTE STORAGE INSTRUCTIONS
K=IeJ+3)-100
STORE RESULT AS BOTH FF(K) AND AS F
IF DO LOOP CONTINUES IT WILL BE CALLED AS FF(K)
IF RETURN TO MAIN PROGRAM IT WILL BE CALLED AS F
FFCK)=F
RETURN
END
rf-~
\-l7'
u. T)snH'
"'0!!iX·
sewn- .we: .... ~
.
MhliMwfijfflf ... ·Mi¥8Mu· .. T f
T
·-n
·f'·'"[
'·-z"""
w·· ..
W···TTT .. T·····hiiiw'ii'HP .. ······· .. ·-timVe·W··"?r··T· nb
ft#riiHriitF£W6*dUhtttrtHbrl*!Hidbh'n.,
c'
FORTRAN PROGRAMS FOR COvIPUTING ELLI PTI C INTEGRALS
AND FUNCTIONS
llenry E. Fettis
James C. Caslin
May 1964
Applied Mathematics Research Laboratory
Aerospace Research Laboratories
Office of Aerospace Research
Wright-Patterson Air Force Base, Ohio
o
150
FORTIW'J
1.
Scope
pnlC
PROGI~\1C;
FOR
CO~n'JTING
ELLIPTIC INfEGHALS AND FUNCTIONS
.2£..:h:. Programs
programs prcsentcJ jn this paper provide subroutines for di rect
computation of the three types of elliptic integrals in so-called "Legendre
~orma1
Fom".
They are
Integrals of the first kind:
11
a.
r(~'Ii)il_::$"'''~;'
it
a:=it
<-I J·
~
Inter-ra1s of the second kind:
b.
E f4
c.
-!f)-I: - -t.sil1~
if
dIJ j
Integrals of the third kind:
2r(~«~
d~
)_(8
) ,1i
l
(/-CX:~ih~)VI-~~ih
I;
t
k
eX = 1
+ seeS)
11
f
=J-~X 1,,(t.1I''''5eC;(;)-fll1I5:::;::~
=
/_~.1.11Jt"t6l"~$'~c;e)+/~Jtd;rta/5';h~J 1
if271.{ta7? 1)~5'
=
2
O('~~ 0
ftC
~
8) + ici71. 615'~~ e} J
1 5 'J
• -
CX'~ /
~f"··~
•
3.
~'1athematical
Fonnulation
TIle prQgram..~ presenteu here are based on a transformation, known as
Gauss' transformation or sometimes Lrulden's second transformation.
It
penni ts each of the three nonnal integrals to be expressed in tenns of
elementary fW1ctions and similar ones with a lesser modulus.
Since the same
transfonnation may be applied to the ne\" set, and the process continued until
a mouulus of sufficiently sraall magnitude is obtained,
a point is ultimately
reached where the integrals reuuce to element functions as outlined in
section 2.
It is then a simple matter to go backwards and find the desired
values of the original integrals.
a.
The transfonnation goes as [0110\0,15:
A new modulus kl is computed from the formula
l-kt
kl = l-k'
where k t =
v'i"=k! is
known as the complementary modulus.
/r'~
"-\;.c.."".i
b.
A new amplitude Q1 is computed from
sinQl
c.
=
(1+k')sin9
1+ (1-k 2sin2 9
A new parameter ((, is computed from
0(. Of:: 0(
I
where p '=
¥ /t.."jiJ..
(/+1t'
~1-k2 /o(~
TIle original integrals, functions of k,Q, oc.~ may now be expressed in terns
of ones \Vi th argt..1Irents k 1 ,
F(Q,k)
9,,0~~,
and it is easily shown
that if these conditions ate satisfied by the original values of k and ~~
it \...i 11 always be satisfied by the subsequent values of these quanti ties.
TIl is leaves one remaining case, namely
o .c:: tX.~<~;l..
,-:hcre the transformation wi 11 not
\~ork
di rectly.
FortLDlately, there cxis ts
the following relationship:
II(Q,tX~~2)
+
II(QJ~~,k2) =
5.1
.J/+tJ../
/-u
where
(l-rx~)(lI~~/) t 1, TI(Q,k2/~,k2) can be calculatcd by the previously
dt'scribed method, and I I (9, "'~ k 2)
6.
SOr1C ~
:~ur.lerous
L2!.
from the above fonnula.
.!!!£. Ntunetica 1 Accuracy
sources of inaccuracy may be i.nherent in the previously
described fonnulac.
a.
Improving
fOlUlc.l
111ese will be briefly discussed be low.
Subtraction of large but nearly equal quantities.
This occurs both in the calculation' of E and II.
In the fOrnlulac
for E, Tn and Sn increase in magnitude but their difference remains finite;
5j
mi larly, in the expression for II, the quantity Yn Gn - Zn approaches a
finite limit although both Yn (;n and Zn incre'Ls,c.
111e inaccuracy whidl
mi ght result due to loss of significant figures may be avoided by refonnulnting
the exprcssions eli rectly ill tenn..'i of the differences.
b.
Inaccuracies.!!!. Computing
~
from SinQN.
:,tost computers find inverse trigonometric functions by means of
(til
inv(;~rse tangl'tlt suhroutine.
cllculatcd.
\',"j
TIlis requires that both sin9 N antI cos9~~ be
T[ sin9N is close to zero or one, inaccurate values of cos9~
11 result if the latter is computeJ from the relation:
coSQ N :: '/1-sin 2QN'
to avoiJ this I.lifficulty. c.osQn+l should be computed directly
[roJ1\~inQn
and cosQn' s i.Jlc{'~ the ini tial of those quanti ties may be computed wi th
reasonable precision, and the final results "'ill accordingly be r:lore aCCLlratc.
A suitable fomula for accomplishing this is the follO\'Jing:
7
157
•.-...,.•,..-..-
... - ...
'filW . p
Unreasonably large numbers may also occur if the ini tial value nf
small, since this quantity· appears ill the denominator in 13q. 3.1.
1J
is
For this
reason it may be preferable to avoid uSing the direct method when C(t4 is
ncar to k 2 , and this may be accomplished' by' employing Eq. 3.1 only when 't'~ 0
oro(",.1f... and using the transfonnation to the parruneter
o
8
~~J.
15S
othcl1dse.
;("~,
\,
1:,lj:~1
I:JULY i-(SSK,Tll)IS crH,:iPUTED ,
1;,JD==2 F(SSK,TH) Ai"jD E(SS!","nl)
/\RE~~jjIIPi.JTED ~j
~'J"')01'J"'-I'''IE'
ELL 1P (coS"
(""11''\, r"''')
.;)
I, () hi,,) l I
i' ~,
;J
I'" ..I",..", I i:''JD' ,I:".)
C. i \
P 1=1 • S/079632Gj'91~G/ 5
Tl'll =F> l"(TH/90. 0
C" S I,
1/
('l
r"l
. · 1 ....0') \,J,
";J, n
f: ( ,),
9 I F( TH- 90 .0) 12, 13, 12
1 2 S;,1=$ I i,IF( 'j"H 1)'
,
f
0,,)
A=(1.0~SN)/(1.0-SN)
JK~O. !),,\L(3G I:" (/\)
£1(=51,1
G!~ TO 30
13 G(1 TrJ (1 J}, 1:~ ) , I i'·J D
1:;~J!(==9 .9999999E+9D
11~
G
6
5
'/
£1-<=1 .0
GCj TrJ 30
PR I iff 100, SSK
;,H(=9.9999999E+93
Gt1 TO 30
IF{'rH-90.)S,b,5
$ I',J:: 1 .0
C;',!=O.O
GO TV:) j'
S;~ol=S 'I i·lF (TH 1)
C;"1=C(1S F ( TH 1 )
S1<:::5 SK
H=l .0
T=:l _0
Q=O .0
S=1.0
D::SQRTF( 1.-51<~'\SK~\S~",11"St"J)
3 SI,2=S;('i($K
SKP=SQRTF(1.0~SK2)
GC1 T(~ (11,10), li'JD
10 Q==Q + ( "1",( Si',ji C~",l~'( S1<2! ( D+1 • 0 ) )
11 SK=(l.-SKP)/{l.+SKP)
~(:.=( 1 • O·~SKP) 1(1 .0+0)
l
(
SrJ=X"kS i'l
O=SQrbihirtiWttiiiiMi" \ ..... ··..rff .....·..w··jj""·-
1"]""'
-
.
flt"tf'J""X,.··YIl"f""Nj··T'W5·· ... Y .. · -
... _.. ·C"·W'F"" .. '··-.. H'P ... ·g'···
APPENDIX B
C F0RTRAN II PR0GRAM THAT CALCULATES ELLIPTIC INTEGRALS 0F THE
C THIRD KIND.
INPUT:
SK=M0DULUS, (7J=AMPLITUDE IN RADIANS
PARAMETER (AL)=ALPHA SQUARED.
PI_/~ . '" (/19
C AND
C
-
(1-/1(
f'inllt')VI- ~;tSl~.;l&
1
3
4
43
44
45
9
10
13
11
14
15
5
2
o
16
ACCEPT 100,SK,0,AL,E
Pl=1.570796325794875
ALN=AL
SKN=SQRTF(SK)
SN=SINF(0)
CN=C0SF«(7J)
R=1.0
S=1.0
T=O.O
Q=O.O
TS1=AL-SKN
TS2=ALN-1.0
TS3=SN-CN
IF(AL)35,40,3
IF(TS1)4,4,35
ALN=SK/AL
TS2=ALN-1 .0
T3=1 • -ALN
T4=1 • -SKI ALN
IF(TS2)43,45,44
U=SQRTF«T3*T4)/(1.-SK*SN*SN»
GO T0 45
U=SQRTF«(ALN-1.)*(1.-(SK/ALN»)/(1.-SK*SN*SN»
IF(TS2)9,14,15
C=SQRTF(T3)
IF(TS3)10,10,11
Tl=ATANF(U*(SN/CN»
T2=ATANF(C*(SN/CN»
C0RR=(Tl-T2)/SQRTF(T3*T4)
G0 T0 5
Tl=Pl-ATANF(CN/(U*SN»
T2=Pl-ATANF(CN/(C*SN»
G0 T0 13
T1=SQRTF{1.-SK)
T2=SQRTF{1.-SK*SN*SN)
C0RR={-SK*SN*CN)/«T1*T2)*(Tl+T2»
G0 T0 5
Tl=(U*SN+CN)/{SQRTF(ALN-l.0)*SN+CN)
T2=(1.-SK*SN*SN)/(1.-(SK/ALN)*(SN*SN»
C0RR=(.S*L0GF(Tl*Tl*T2»/(SQRTF«ALN-l.)*(1.-SK/ALN»)
P2=SQRTF(1.-(SKN*SKN)/ALN)
SK2=SKN*SKN
P=P2
SKP=SQRTF(1.-SK2)
16U
AL1=ALN
IF(TS3)16,16,17
SAL2=SN/CN
G0 T0 18
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
35
36
37
38
39
40
41
42
SAL2.CN!SN
SN-«1.+SKP)*SN)/(1.+SQRTF(1.-SK2*SN*SN»
SKN-(l.-SKP)!(l.+SKP)
TO-S1 RTF (1.-SKN*SKN*SN*SN)
eN-( 1.+SKN*SN*SN)/TO)*CN
R-R* 1. +SKN)
P2-SQRTF(1.-SKN*«1.-P )/(l.+P »)
ALN-ALN*«l.+P )1(1.+SKP»)**2
IF(AL1-l.0)19.22,23
Tl-SQRTF(1.-AL1)
IF(TS3)20,20,21
.
HN.(2.0!Tl)*{ATANF(Tl*SAL2)-ATANF(SQRTF(1.-ALN)*(SN/CN»)
GfD T0 24
C-SQRTF(l.-ALN)
HN.(2.!Tl)*«Pl-ATANF(SAL2/Tl»-(P1-ATANF(CN/(C*SN»»
GfD T0 24
T2.SKN*SKN
T3·SQRTF(1.-T2)
T4.SQRTF(1.-T2*SN*SN)
,
HN.(-2.*T2*(1.+SKN)*SN*CN)/«T3*T4)*(T3+T4»
G0 T0 24
Tl-(CN+SQRTF(ALN-l.)*SN)**2*(1.-«SKN*SKN)!ALN)*(SN*SN»
T2-(Ct~*SQRTF( 1.-SKN*SKN*SN*SN)+P2*SQRTF(ALN-l. )*SN)**2
HN.L~GF(f2/Tl)/SQRTF(AL1-1.)
.
Tl·1.+SKN
T-T+(tiN/P )*5
~:T~l;~~i~~jp-J*~/P )*s)
IF SKN-E)25,2S,2
IF SN-CN)26,26 t 27
0N-ATANF(SN/CNJ
G" T0 28
0N.Pl-ATANF(CN/SN)
PfBAR-(lN*Q-T
IF(AL)31,40,29
IF(TS1)30 30 31
PI-0N*R-p1sAR+C0RR
G0 T0 32
PI-PIBAR+(G0/RH')
PRINT 101,PI
G0 T0 1
RH0.SQRTF(1.-SK/AL)
IF(AL-l.0)36,39,42
C.SQRTF( 1.-AL)
IF(SN-CN)37,37,38
G0-ATANF(C*(SN/CN»/C
G0 T0 5
G0-(Pl-ATANF(CN/(C*SN»)/C
G0 T0 5
IF(CN-E)40,40,41
PRINT 103,ALN,0
G0 T0 1
G0-SN/CN
G0 T0 5
C-SQRTF(AL-l.}
Tl-(CN+C*SN)/(CN-C*SN)
~~:t~~tE~A~(Tl»/C
100
103
101
11::\ '
G0 T0 5
1 G!
F~Rt1A T ( E17 • 10)
F0RMAT(30HPR0BLEM N0T DEFINED F0R AHPHA=F16.10 10HAND THETA=F16.9)
t~~t/1AT (3HP I =- F15. 10)
ij.~ , //
o
tttit#
t t
T~USE
OF DISCRETE CONVOLUTION IN THE ANALYSIS OF
DIFFRACTION PATTERNS
J. S. LIEBMAN
P. M. JOHNSON
GENERAL ELECTRIC COMT)!lNY
. ADVANCED ELECTRONICS CENTER
IT:AA.CA, NEW YORK
o
162
Jf. 7. D
1
This paper is about a program for convolution of two complex
discrete functions.
We will discuss uses of complex convolution, de-
scribe the experimental problem which led us to develop the" program,
and then will present the technique itself.
The convolution process is used in the analysis of several engineering problems.
In electrical signal processing,by networks, the
circuit output equals the complex convolution of the input time function
with the circuit impulse response function.
In schlieren optics, light
passini through transparent media undergoes phase diffraction instead
of amplitude diffraction, and may be analyzed by complex convolution.
Our experimental system, a schlieren system, is a special case of
a coherent optical processing system, and it uses the phenomenon of
ri'~
\~LR
phase diffraction for a rather unique purpose.
Figure 1 is a diagram of the experimental setup.
light is passed through a
~ooved
If monochromatic
transparent film and is partially
blocked in the diffraction plane, it will then project a pattern on the
image screen.
The image intensity varies (in a manner depending upon
the type of operation at the diffraction plane) with the groove depth
and slope.
The variations of light intensity on the screen present the
information that is stored on the film.
If no blocking of the light
takes place in the diffraction plane, then there will be no variation of
light intensity on the image screen.
We desire to find out what kind of
blocking at the diffraction plane produces the best results.
16 J
'I. ZI
J' ·ET." },} ...-.... t 'dt. Ht#si'bw .. "±ti· tid
***#*#1* ."t -. h· riririe¥itiirirl iztM+tdw·w
·
N*fH_
_
•• J
.(".:-
·-f .. ·..
n
:lI TT"If"Til!IfW .. l"'!-ij'm·!!t""p·· ... WT
..
Wf"JjWIIIT
T
2
It happens that Fourier Transform relations exist between the
object~
diffraction, and image planes.
It also happens that the light
amplitude at the object plane can be represented as a multiple product.
If we use convolution theorem that the Fourier Transform of-a product
is equal to the convolution of the Fourier transforms of the individual
functions, and if we know the individual transforms of the factors in
the product, we can easily compute the diffraction pattern.
Hence, a
major part of our computer simulation of the optical display system 'Was
the development of an efficient technique for the convolution of two or
more complex discrete functions.
SOURCE
PLANE
OBJECT
PLANE
(THERMOPLASTIC
RECORD)
1-----.".- -
--
x
FIGURE 1
o
IMAGE PLANE
(SCREEN)
Xl
X"
3
Because the functions we convolved have real and imaginary parts,
the following series of equations develop the model used for our complex
convolution.
If
;= ~.;=~
I
are complex functions of x, the distance from the center
of the diffraction plane.
~ (X) =
c(
(z) E:' -~ '0< (x)
~ 6) -= if (z)e-"(6U)
Let
Since
,.5't11':;:'<'" C'6-Q
(0
•
I-
ceo oJ. 51il iG
I
165
tI.73
C.
II' . .·· or'cm'T
t t
'j
t
tth tit
.
'hj .. $j
'bi&fH,-' ···[· ..-· ... f·j'·-.,·
{rite> fow tirtwrt_**
.... j ... · j
_.
- '"!i'¥iitii'HrW •.•.
"]
r- .. n n"'f""""'%!PW ..
·i.,.-n.
. . . . i¥W··!s",w··
4
Thus, complex convolution may be expressed in ter.ms of four real
convolutIons.
The following operations represent :part of oursimu-
lation and, for various
~easons,
they were carried out by several
At the end of this paper is a Fortran
puter programs instead of One.
program for the simple convolution
(a)
c:'
~erfor.med
in part b.
Prepare four tables.
(1) X
(2) X
(3) x
( 4 ))(
(b )
COID-
vs
a CC)sc>(
VS Q SIn e><...
~ b-- CCAS
V5 tr S' t'n {3
G
Convolve
(1)*(3) : R (ll)
(2) *(4) =- R~ (,x)
(1)*(4)-=- ~I (x)
(2)*(3) ·=~~(X)
(c)
Combine into real and :imaginary parts~
li(x) =~, (x),- R~()()
cVCx) .= I, 6V+ J;. (x)
(d)
The amplitude of the roth point of the diffraction pattern is
(e)
The phase of the roth point of the diffracti~n pattern is
eo< ()11) ~
o
Ta'l
-/ Lri(11)
f\
)
("tl) .
166
Lf.71
5
In addition to possible: uses in y;our own computing group, this
technique offers interesting oPPC?rtunities for University physics and
electrical engineering labs.
More and more universities and colleges
are teaching their students how to program and are giving their students
machine time.
Complicated diffraction patterns could be predicted
before being observed in the physics labs, as well as schlieren optics.
Computer calculations could replace the electrical engineering students'
frequent use of the graphical technique for complex convolution.
In closing, we present the Fortran program for the basic Convolution of two discrete functions that is the fundamental unit of our
complex convolution.
It is offered as an example of a technique to be
used when the variable (m in our case) can be restricted to integer
values.
Then each F(m) can be stored in location m for efficiency of
storage.
1 G7'
c
r·r-t·st ....!i·en·····
c
CONVOLUTION OF
C
J.S.LIEBMAN.~ENERAL ELECT~IC.ITHACA.N.V.
C:
T~O
DISCRETE FUNCTIONS
DIMENSION Ml(201).M2(201).Al(2Ql).A2t201).AlA2(402)
READ 95
PUNCH 95
READ 90.Nl
DO 10 l=l.Nl
READ
90.
M1 ( I ) • A 1 ( I )
j! 1 0 M 1 ( I ) =M 1 ( 1 ) + 100
READ 90.N2
DO 20 1=I.N2
READ 90. M2(I).A2(I)
~
20
M2 ( I ) =M2 ( I ) + 1 00
DO 30
30
K=1.402
AIA2(K)=0.
DO 40 1=1.201
DO 40 J=I.201
K=Ml (I )+M2(
40
( ""~:
J)
AIA2(K)=AIA2(K)+Al(I)*A2(J)
DO 60 K=1.402
.
"
IF (AIA2(K»
50
50.60.50
J=K-200
PUNCH 90.J.AIA2(K)
60
CONTINUE
PAUSE
GO TO
90
FORMAT (14X.I6.14X,F6.4)
95
FORMAT (49H
END
* THESE STEPS WERE MERELY MANIPULATIONS TO AVOtb
NEGATlVE!NDICES~
~.'i".1
~.
J 68
·rr···..''fjtl'!]]i'n
---~---"
-,------
-----,---,------,-,,-,----,-----------,----,---""-,-
---
A COMPUTER APPR.OACH TO THE ANALYSIS OF GAMMA SPECTRA
BY THE ~THOD-OF CHARACTERISTIC ENERGY REGIONS
by
,
to
f
~
o't
i
I
:L.1-+
. 1-,
~D
::z. b t,L1l : I
Sf.! 81 J {Il t£
II 9lS'
OPERATOR
240O-0eoo HRS.
II
oaO()"1600
•
i_~J
£.;t;t:.;
PD Veeder
1 Unit
p __ _
__ JJ£_
&
<:;
L
~ CP~ ~
(pi
_e_J._.
n ~'.-(;
II -
;
,_I~:
G-Id
,7
Ih ~~111 ,5:'03
tsa ') 112Y?:~lJ..q
_ :tfi 1_
tot/-
PRES·Ir!5I~F.
~1 LJ 011_
91)-,-:;' 7 l'Rilll S-
33974 "IN"
33981 "IN"
II
__
- '_fi~
~~
Jo I
.L~tLb~ ~l 3/7_
f;.
33980 "XN"
I:J,------~--~--------L---------L-----~
II
119 3J-1 j
JI ~aJ t;
3009525
... Unit
•
I:. 7,
}f!~
~:
7~
~ It.
'l.3{,
~71 1/ 0
~~
J)lIJ;"'1
2.ea
33973 "IN"
3009622
Turbine 30084604
1 Unit
96
'ji;.r
r/ .J/_~
~/,·SjZ,_-
60
911 ~ll")t..? G'31 ~ II S :),1 · 9 I
~ s:L/_:A.l_ _t{
g' ~
[lY)7lI7- 0 i~~_-__ ~L.--====1lj-o~iH.rZ:-2-.ii d-hYlj. L '1·31 J.Ij·o7111o-·'j l1-:--;- :
3 ~,~:-B--t--¢ ~ I !i 2=l!
~ .j-,Ib.. ~ ", J rr-=8f;l3-~-j.1ci.·~-T~WR=7-=¥1 ; ! !
1134
I a
33980 "OUT" 3009446
PD
~
,_J-~
.1
:(p!z.
2206173
b.l1l
6' 7J7J
/6 ')
~
I G,. I
3398 i "OUT" 3009440
2~""
..
1201905
/a
113
JS.J
;4)
!, . __ _
Ij~'
:/I~' / ;:36,
/' r
Ilfi.
/1 ,
,:;.
?
/" ij ttll~ ~~4ll 2. Z ~II ~"o b~(1-'&-~ ~
/ (} z
1500
1600
1700
S'-/71
I ;'
~3
.1
b
], ~.3
!4Hr.TotBl"~ 810
1 __
~
7l7
,2Hr. Total/ll
'_
'~~
l·~
2300
2400
,":
!
I
--&,,-"
2100
2200
!
;,
1400
_
READING
;-
~o
•
PLANT
LOAD
N
2106125
GAS
~ 7.\ 8 IFIRM
ENGINES URBINES SALESTO
.
25 N
25'"
STEEL
·,r~
1800
1900
2000
Ag:h:aud-
1101857
NET
NET
251\1
2S a, 60 IV
TIES TO
TIES
STEEL TO STEEL
IJ r
COOE~I 3209471
() , I
DATE!JeAf
I " ..5
!2Hr. Total II}
'l:.
Bar
I
Ibbb I b Ibbblbbbi bb
It'!
1 00
~
I () r
1100
• __
@
Enltineerln& StaU.tlca
,:l}.l
J 3 Lj •..•
1200
1.'
t
LOA
& E. Co.
MWHR
1112993
SON
TIES
FROM
STEEL
D.F. CODE
c.
n
........
-..
r
ITEM If'ENTIFICATION
1~~~1
AOORfSS
~r
DATE
MO. DAY
CONST.
:1
YR.
•
';iig ~...
24 'HOUR
HOURLY DATA
fTJ7 ~~
TOTAL
0100
0200
0300
0400
66 6) 68 6~ )0 71 7Z 73 74 75 Ib 11 )~ )9 8D
,! 1
.~ ~ i,11 2, 2 211 2 2 2 12 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 22 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 2 2 2 2 2 1 2
.~ ~I 13 i3 313313 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 33 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3' 3 3
~ i~ I,} i4 414 414 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4T 4' 44 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4
~ ~ 1:Is 5' 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 55 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5
1
+
1
1
l~j'~!~!S SiC 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 666666666666666666666 S 6 6 6666666 G6 66666666666666666666666666
"ilh III I 111111111111111111111111111111111111111111111111111111111111111111111111111
~
!G!a 8 8 08 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8888888888888888888888 8 8 8 8 8 8888 8 8 88888888888888888888 8 8 8 888
iqi g 9 9 999 9 9 9 9 9 9 9 999999999999 9999 9 9 999 9 9 9 9 :l 9 9 9 9 9 999 9 9 999 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 99 9 9 9 9 9 9 9 9 9 91
i liz
J 14 516 ) 8 9 10 II 12 13 14 IS 16 11 18 112021 22 23 24 25 26 27 28 2. 30 31 32 33 34 35 3ti J7 38 39 40 41 42 43 44 45 46 4) 48 4950 51 $253 54
_.. _.... __I.~9JJ9.1~ w
t
~
MIse
ITEM
! I
6
SuB
t ...
I
MASTER
IOENTI'-ICATIQN.
ITEM IDENTIFICATION
i
·#1 .•" . ,t
DATE
'·r.
t Kn"
en .rt
. ., **
~ 56 57 ~8 ~9 60 6162 6l 6·1 63 6i 81 68}~ 70 11 7273 74 75 16 ))
w', ? .
Mr.,' ' t
f
:;!4HOUR
Ji 19
i~
~
!:~ ~..:
HOURL.Y
OATA
FOR
HOUR
ENDING IA.M. OR
P,M.t
Poii~Co;o OS'~~O t~~~J<~ ~y~ ;°0 ~A~ ;~ ~ OTOOOTOA~ O:~ ~~ ;~oOoO ;~~~ ;~~~ ;~~~I;~~~ ;~~~ ;~~oO ~~~oO ;:~~ ~ ~~oO ~ ~ ~oO ~;~~
11
III
~
"".,~
)
g
I sys·1
__
•
w, • •
Z 3 4 516 7 8 9 10 II 12 13 14 15 16 11 18 19 2C 21 22 13 24 25 26 27 28129 30 31 32 33 34 3536 37 38 39 40 41 42 4344 45 46 41 41 495051 5 53 54 5558 51 58 59 6~ 61 82 63 64 656667 68 69 70 71 72 13 14 15 76 77 78 79 80
1 t 11':1 t 11111 t 111111111 t 11111:1"'11111'1'11111111111111111111111 t 1111111111111111
:
~
g.
~ <{ 22 2 2 212 2 222222 2 2 2 2 2 2 2 2 2 2 2 2 2 2 212 2 2222 l 2 2 2 2 22222222222222222222222222222222 222222222 f'I
.~ ~
NOTES: (a) Voltoge code (b) Milceilaneoul equipment code.
~
~ 0 3 3 3 3 313 3 333 3 3 3 3 3 3 3 3 3 3 3 333 333 3 3 3 3333333333 333 3 3 333 3 3 3 3 3 333 3 333 333 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
C)
~
Z
a:
56
IJ I
z
C)
Z
jW
I
(cl Duration of meosurement code.'
Cd) Unit of mealurement code.
4 4 4 4 414 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4
I
(el Number of places to right of decimal point.
UI
~
0
f'I
5 5 5 5515 5 5 5 5 5 5 5 5 5 55555 5 5 5 5 5555155555555555555555555 555555555555 55555555555 5 5555555 5
I
I
6 6 6 66166 6 6 6 6 6 6 6 6 6 6 6 6 6 66666666166666666666666666666666 6 66666666 66666666666 6666666 6 6
~
n
~
I
n
7 7 7 7 717 7 7 7 7 7 7 7 7 7 7 7 7 7 7 777 7 7 7 1117 7 7 7 7 7 7 7 7 7 7 7 77 11 7 777 7 7 7 7 7 7 7 7 7 777 7 7 7 77 777 7 7 7 7 7 7 777 777 g
I
I
I
8 8 8 8 BI8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 818 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8
I
I
.
9 9 9 9 91919 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 919 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 99999999 999 999999 999
,
I
1 1 Z 3 4 516 7 8 9 10 11 12 13 14 15 16 17 18 1921 21 22232425 2& 2J 2s129 30 31 323334 3536 31 38 3U( 41424344 4546 41 4 495051 5 5354555657 585IGC &1626364 6""HUt 10 /I 12 13 14
.... 000505
1~ 76 11 787980
92
I
I
"1*
/ ""
:S '- I .P t::'
J \
?'¥
4B_G. & E.
(M-1679)
co.
BALTIMORE - COMP. 8
MD. P.R.R. INCL. BFCL - COMP. 9
STEEL PLANT - COMP. 10
BALTIMORE GROUP - COMP. 11
WASHINGTON - COMP. -12
BALTO. WASH. GROUP - COMP. 13
BALTO. PLUS WASH.
BALTO; AND STEEL PLANT
BALTO. AND STEEL SALES
BALTO. AND STEEL SALES AND MD. P.R.R.
1
2
3
4
5
6
7,
8
9
10
~~
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20,
211
22
23
24
fOTAl
.
511
484
473
467
476
502
573
757
911
989
1029
1031
1012
1039
1022
1007
995
971
999
1020,1
.
~
15
21
18
19
19
20
21
30
17
14
11
12
12
12
21
23
24
28
24
24
218
233
239
222
224
238
237
272
239
236
241
241
241
262
265
251
253
245
262
-
1285
1280
1278
1220
1079
- I
236 ,
744
738
730
708
719
760
831
1059
1167
1239
1281
1284
. 1265
1313
1308
1281
1272
l2~A
-
968
881
749
2246
2101
1828
LS21
1986
2022
1974
1809
1564
131q
1±_5.27_4
3R9S4
928
815
'14
19
261
278
245
?()
/~?
Q--'i.8
hl~
44q
SR71
26()LJ. ~
l_q2~~
iq7?~
.
2282
-
~
1232
1190
1165
1142
1152
1214
1378
1790
2061
2231
2313
2326
2284
2347
2337
2311
2275
2226.
2272
11
7()f)
-
488
452
435
434
433
454
547
731
894
992
1032
1042
1019
1034
1029
1030
1003
5182
987
1002
1006
-:t
999
9'36
908
901
909
956
1120
1488
1805
1981
2061
2073
2031
2073
2051
2037
1998
25 CYCLE
BALTIMORE
BALTIMORE PLUS STEEL SALES
BALTIMORE GROUP
14-15
14
15
60 CYCLE
r
BALTIMORE
BALTIMORE PLUS STEEL SALES
--o
634
616
609
584
578
621
681
896
1024
1077
1100
1117
1117
1155
1135
1168
1158
1115
1157
1154.
1157
1070
937
649
637
627
603
597
641
702
926
1041
1091
1111
112_9
1129
1167
1156
1191
1182
1143
1181
1178,
1168
1084
956
q~R
R71
/t;SQ4
??hAI
729
717
712
689
700
740
810
1029
1150
1225
1270
1272
1253
1301
1287
1258
1248
1216
1261
1256
1267
1206
1060
0#
1~53
11-13
11
12
13
. 1.
. 2-
22,
14
..
13
.~
18
18
18
17
19
17
18
21
19
17
21
18
21
18
20
25
25
33
34
27
26
29
27
35
31
23 ,
28
37
28
39
23 --31
24
32
24
34
22
33
33,
30
42
43
43
47
47
46
45
57
62
61
66
493
467
456
449
459
484
555
737
886
962
1003
_D5
10~4
63
67
989
1011
bb
~94
62
63
65
59
61
984
971
947
977
998.
59
64
55
905
795
A41
21
23
20
21
33
26
27
52
nas
?~1':\()
f;)?:l
f,7Q
l~f,()
19l~
985'
,...
~
"~
-
ENGINEERING STATISTICS
COLUMN HEADINGS
COlUMN HEADINGS
~
1.l.lA
(-;
c,.,:;
SYSTEM LOAD COMPONENTS
MWHR - MAIN UNIT BASIS
~~
"
;/
MON APR 20 1964
~~~
\1 )
15
--
616
598
590
563
559
600
660
871
991
1043
1071
1082
1086
1118
1096
1137
1126
1081
1124
1121
1127
10371
911
1
7q4
/?()()?
o
5LI
Pt.;:- # J B.G. & E.
(M-1679)
co.
SAFE HARBOR NET SEND OUT
230 KV
138 KV
69 KV
TOTAL - SEE NOTE
2/3
1/3
1 .- 6
1
2
3
4
5
6
1.
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
102
102
108
110
109
109
104
94
104
107
110
110
112
109
111
110
107
104
107
20
1071
21
22
23
110
109
106
108
24
TOTAl
~
~
~
2-
4.
3
~
5
52
51
51
46
52
51
52
58
59
61
62
64
64
64
64
64
61
65
61
20
18
16
17
14
20
23
36
48
52
57
57
52
55
53
52
54
49
50
174
171
175
173
175
180
179
188
211
220
229
231
66,
50
2~3,
64
58
57
48
43
35
28
221
216
199
105
- -63
-
£~8
228
228
226
222
218
218
116
114
117
115
117
120
119
125
141
147
153
154
152
152
152
151
148
145
145
149
147
144
133
129
1410
2569
947
4928
3285
INCL.RAMP ENERGY OF 0 MWHR
~ND P.P.& L STA.SER.OF 2 M~vHR
~
":':1.1.25
ENGiNEERING STAHSTICS
COLUMN HEADINGS
COLUMN HEADINGS
HR -
n
~
' MANOR-BRUNNER IS. TO BRUNNER IS.
7
NET TO BALTO.-WASH. GROUP FROM OTHERS
9
NET INTERCHANGE TO B.-W. GRP.
NET INTERCHANGE TO BALTO. GRP.
11
NET INTER. FROM WASH. TO BALTO. GRP.
12
S.H. 230 KV NET TO BALTO.-WASH. GRP.
8
10
6-
.7
~3
73
74
74
72
66
~
150
147
132
103
79
66
1643
8
21
10
-11
-14
-5
35
80
133
179
190
192
179
158
165
159
144
159
153
147
58
57
58
58
58
60
60
63
70
73
76
77
76
76
76
75
74
222
.. 249
233
228
190
166
51
-41
-173
-213
-209
-164
-53
-102
-79
-24
-73
-62
-28
-45,
-33
0
98
121
9,
10
106
135
116
113
73
46
-68
-166
-314
-360
-362
-318
-205
-254
-231
-175
-221
-207
-173
-194
-180
-144
-35
200
214
205
191
149
97
-21
-90
-227
-217
-224
-183
-126
-158
-144
-106
-141
-156
-131
.~
-8
-163,
-151
-125
-10
64
11
..
94
79
89
78
76
51
47
76
87
143
138
135
79 .
96
87
69
80
51
42
31
29
19
25
72
12
.-
259 -3026 -1253
2685
1773
-116
TO •• 1558 •••• 589 ••• 1120 ••• 1773 •••• 660
FROM.1299 ••• 3615 ••• 2373 •••••• 0 •••• 776
S.H. ENTITLEMENT &c INTERCHANGE DATA
MWHR
MON APR 2
1964
13
.
~
14
..
81
92
119
124
114
74
24
-39
-75
-83
-82
-69
-46
-56
-48
-34
-52
-49
-40
-43
-37
-23
3
20
15
i
I
I
i
.:.
( )
c~!
B. G. & E. co.
(M-llll)
ENGINEERING STATISTICS
COLUMN HEADINGS:W-IND. MINUS
OOLUMN HEADINGS
NET BEN. 138KV FROM WASH.
NET BEN. 115KVFROM WASH.
1
2
.°
.....
NET
1,
4 ,'
I .
2'
,.
31 3
40. 3
3
5
c;]
5
i 1
II 15
. II 13
11 11
M
IS·
12 54
3
iN 16
11 1
~ 5
16
12
14
;20 16
: 11
Ie
, .1
.9
tz3
1
RK PERFORKANCE MONITORING SYSTEM
DESI (WED FOR AN ELECTRI C UTILITY
METER AND IHSTALLATICI{ DEPARTMENT
D. D. W111i_a
Bngiaeer1Jlg Computer .lpplieatioDs Group
Baltimore Gas and Electric
Comp~
Baltimore, MU7land 2120)
o
t/. 11.0
D ~ D v T~iJ~l:L ::~~',:~~
:;~·ti:'.. v~~o C·3.~ ~~:. ::"ct.;;~~:-':!.o C::.,.~:;~):;;~.::~~r
.it
ir.i)l"OVO:-;l~;::~~
irt budc:;o·::'i:.:..C
to
19
/.f,II./
:r
P~'''''
2
c
fer -:::-':'0 purposcs of this report, all
fo:-cc~, ~~cl
ccnt:;,~ctor3
regardless
as ccmprisi:ng District 9.
of
produced
I 1Jill not
t~bul~tic~s contaL~~d
~":,ter.;.pt
in the
fo~
the dictricts as
to discuss here tile contents of all of tL3
rePCY~.
Hc,,;,;eV0l."', from a general data pro-
As
control d'3vice, the rC90rt
~c:Oi.:~ld
bo virtually useless 'Hi thout the ba.ses
In
r;l~~to!'
end
tr~v81
tiffie as
L~dicetcd
:l r;:~t.::;:·,
by
Division has been divided
b:"o~d.ly
~nG v~~ious
::l1d a job io
~.;.:J
are
~C:dition
to
fuz:ction cards uhich arc i::put, to the system, and VJhich pro7icio
or rc:::ovi..i-:2
o
Do
in:~:>
tabulations. By definition,
c. unit of 110rk llhich cover.c a
isdicr.tcd above,
'~h~ ~rork
singl~
of the In.stallation
approximately a thoUSD.l.""!d functions 'Which
clt!sc.if'ied as product.ive or non-productive.
198
/f.ll r2.
- page,
3
c
re~pect
to ouch
it~::ls CoS
the cedition and
system consists of tevt;;re.l
and district.
pl"O[;;T~,li;;i,
sub"~~action
the firs·t of
l~hich
compiles
~1d
The sor'c,il1g takes botueen oight tJ:ld tan hom"s to complete
The
headings ruld interp... 0tively
t~..l.bu.la:tions
of tabulaticus to
prog::"~m."Ced
in the report, o;i:cept
half to tt-Jo hours to turn out the
th~
no:~t t~~o prOgl"'or~ls
contai."l the
steps for compiling all of the
one contained
cc:;~:plote
-hl
sheet
typ3
9.
(c\,
report on a type '1620 Hodel 1
l"'cpo:."'t cO!'lsists of appro:::irr. \!tely 60 pagos, r-ll of uhich arc turneci o,ut
It ta!-tes ab-Dut si:: hours to prin"c 0 co:nplete l"oport.
iiS
experience iD gained in the intel:"pretatiol'l a:.'ld usc of the
infOl".:~a.ticn COlYC:iC.i.!Z,lJd i~1
be
exte~d~d
to
C.OVCl·
this report, it is
e:~cctcd
that tho cy3tem 'Ilill
the opert!tions of ether divisionS and dop.:ll"t,ments in
It is hoped that its flc:::ibility ",;ill
rna!~c
it easily edaptable
to their 11cods.
199
'thWNJ:t=leH!'MMWrt't
"l!'1!U
i
"ttrttiWI't
tl'
Milt'*'
u'u'"
!
"
,'tMtftJtetf"
t
!,
'Mt' 'f wtWt'ltlf
I
" '
tt'
1-0
JANUARY 1964
INSTALLATION DIVISION
MONTHLY OPERATIONS REPORT
DIVISION SUMMARY
DISTRIBUTION OF WORK TIME BY FUNCTIONAL CATEGORY (1)
CATEGORY
CURRENT
MONTH
LAST
MONTH
SAME MO
LAST YR
I-MO.'
AVG.
PRODUCTIVE WORK
JOBS RECEIVED
TOTAL
RATE (NO/MAN-DAY)
JOBS COMPLETED
TOTAL
COMPANY
CONTRACTOR
RATE (NO/MAN-DAY)
TOTAL MEASURED
ACTUAL JOB TIME (HRS/MAN-DAY)
EXPECTED JOB TIME (HRS/MAN-DAY)
JOB PERFORMANCE INDEX
TOTAL UNMEASURED
ACTUAL JOB TIME (HRS/MAN-DAY)
SUPER., INSPECT., CLER.,(HRS/MAN-DAY)
3148
3148
1.2
1.2
4002
3273
729
1.5
4002
3273
729
1.5
3475
2.85
2.13
1.34
527
3475
2.85
2.13
1.34
527
.32
.32
2.49
2.49
.49
.27
1.22
1.09
1.09
.49
.27
1.22
1.09
.48
.48
NON-PRODUCTIVE WORK
JOB-RELATED
HEADQUARTERS (HRS/MAN-DAY)
FIELD (HRS/MAN-DAY)
WORK PERFORMANCE INDEX (MEASD FUNC)
TRAVEL (HRS/MAN-DAY) (2)
TRAVEL PERFORMANCE INDEX
OTHER (HRS/MAN-DAY) (3)
1.09
DISTRIBUTION OF WORK TIME BY 'TIME CATEGORY (1)
(HRS/MAN-DAY)
CATEGORY
S TRAI GHT TI ME
OVERTIME
o
NOTES.
CURRENT
MONTH
LAST
MONTH
SAME MO.
LAST YR.
7.88
.12
l-MO.
AVG.
7.88
.12
(l~TIMES GIVEN IN TERMS OF EQUIVALENT MAN-DAYS.
(2 EXCLUDING TRAVEL FOR SUPERVISION, INSPECTION, CLERICAL.
(3 MEETINGS, TRAINING, AND ABSENCES COUNTED AS WORK TIME.
200
tlHE " t~
I
- - _... _- .-.... -~-.:.:. -.: --.':-.:. ~:':"':" •.:....... -.•. ~. ..•
•....-." _.-....,::._*';; -.:- ::.:.: ..----........:.:...:.~:.:..:... .....:....:.:•..:.---:.:..:..:..-..:.:..::....:: :".. -:..,~,,:... --:.~ ..- .,,-
~'"
-"
-----~----------
-----
--~.- .•.
---.--------------.---.--------- ...
1-2
JANUARY 1964
INSTALLATION DIVISION
MONTHLY OPERATIONS REPORT
c
ANNAPOLIS SUMMARY
DISTRIBUTION OF WORK TIME BY FUNCTIONAL CATEGORY (1)
CURRENT
MONTH
CATEGORY
LAST
MONTH
SAME MO
LAST YR
I-MO.
AVG.
PRODUCTIVE WORK
JOBS RECEIVED
TOTAL
RATE (NO/MAN-DAY
JOBS COMPLETED
TOTAL
COMPANY
CONTRACTOR
RATE (NO/MAN-DAY)
TOTAL MEASURED
ACTUAL JOB TIME (HRS/MAN-DAY)
EXPECTED JOB TIME (HRS/MAN-DAV)
JOB PERFORMANCE INDEX
TOTAL UNMEASURED
ACTUAL JOB TIME (HRS/MAN-DAY)
SUPER., INSPECT., CLER.,(HRS/MAN-DAY)
209
209
.8
478
478
478
478
1.8
1.8
410
410
2.98
2.33
1.28
68
2.98
2 .• 33
1.28
68
.25
.25
2.10
2.10
.50
.21
1.07
.50
.21
NON-PRODUCTIVE WORK
JOB-RELATED
HEADQUARTERS (HRS/MAN-DAY)
FIELD (HRS/MAN-DAY)
,
WORK PERFORMANCE INDEX (MEASDFUNC)
TRAVEL (HRS/MAN-DAY) (2)
TRAVEL PERFORMANCE INDEX
OTHER (HRS/MAN-DAY) (3)
,
1.54
1.07
1.54
1.10
1.10
.42
.42
DISTRIBUTION OF WORK TIME BY TIME CATEGORY (1)
(HRS/MAN-DAY)
CATEGORY
STRAIGHT TIME
OVERTIME
NOTES.
~
CURRENT
MONTH
LAST
MONTH
SAME MO
LAST YR
7.94
.06
I-MO.
AVG.
7.94
.06
l~TIMES GIVEN IN TERMS OF EQUIVALENT MAN-DAYS.
.
2 EXCLUDING TRAVEL FOR SUPERVISION, INSPECTION, CLERICAL.
3 MEETINGS, TRAINING, AND ABSENCES COUNTED AS WORK TIME.
201
c
f"fffll"
- " TrrOI" "" -FtMW.B!*h'" fHd Uf&#tri
"±tLiriri"b±Mriiftf "rt. 'h ""Erb" " "'
.. b&+rif"".. b#ih:rt t" iif. """\ " " t t tt t t i l t
c'
2-0
JANUARY 1964'
INSTALLATION DIVISION
MONTHLY OPERATIONS REPORT
JOB PERFORMANCE INDICES
(MEASURED JOBS)
CURRENT
MONTH
(~.1
SAME MO.
LAST YR.
I-MONTH
AVG.
DIVISION
1.34
1.34
MONUMENT STREET
1.58
1.58
ANNAPOLIS
1.28
1.28
ESSEX
1.16
1.16
GLEN BURNI E
1.20
1.20
HOWARD
1.12
1.12
COCKEYSVILLE
1.43
1.43
097
.97
WESTMINSTER
..
1008
1.08
CONTRACTORS
1.57
1.57
iY
BEL AIR
o
LAST
MONTH
202
W'
3-0
JANUARY 1964
r~
~I
INSTALLATION DIVISION
MONTHLY OPERATIONS REPORT
TRAVEL INDICES (MEASURED TRAVEL)
. PERSONNEL PERFORMANCE
LAST
MONTH
(1 )
(2)
SAME MO.
LAST YR.
(1)
(2)
I-MONTH
AVG.
(1 )
(2 )
1.09 42.84
DIVISION
CURRENT
MONTH
(2 )
(1 )
1.09 42.84
MONUMENT STREET
1.04
34.51
1.04
34.51
ANNAPOLIS
1.10
51.86
1.10
51.86
ESSEX
1.63
45.48
1.63
45.48
GLEN BURN! E
1.68
50.64
1.68
50.64
HOWARD
1.05
38.92
1.05
38.92
COCKEYSVILLE
.89
44.30
.89
44.30
BEL AIR
.78
51.92
.78
51.92
WESTMINSTER
...
1.00
64.61
1.00
64.61
CONTRACTORS
1.03
34.54
1.03
34.54
NOTES. ~1~ ACTUAL TIME/NORMAL TIME.
2 TOTAL TRAVEL MINUTES/COMPLETED PRODUCTIVE JOBS.
203
((~\
~
'ift"'';
w'tt'
un 'ilL
I
»M"
e'L" e
i¥iN''!!
"tHY
-'"'uI'
II
kIItMHtMI" 'J
I
H
t
11M
H ""it""" #{tt,#"
"If
c
""tf "whitt t t t
4-0
JANUARY 1964
INSTALLATION DIVISION
MONTHLY OPERATIONS REPORT
TRAVEL INDICES (MEASURED TRAVEL)
DISPATCH PERFORMANCE
CURRENT
MONTH
CI
DIVISION
(1)
5.6
(2)
10 8
MONUMENT STREET
3.1
5 9
ANNAPOLIS
9.6
ESSEX
4.9
GLEN BURNIE
6.4
HOWARD
5.7
COCKEYSVILLE
6.8
BEL AIR
8.'5
WES TM INS TER.
CONTRACTORS
10.1
3.2
LAST
MONTH
(1)
0
0
17.8
8.6
11.4
12.1
14.2
16.4
21.0
6.2
(2)
SAME MO.
LAST YR.
(1)
(2)
I-MONTH
AVG.
(1 )
5.6
(2)
10.8
3.1
5.9
9.6
17.8
4.9
8.6
6.4
11.4
5.7
12.1
6.8
14.2
8.5
16.4
10.1
2 1 00
3.2
6.2
NOTES. (21 ) VEHICLE-MILES/FIELD JOB o
( ) MAN-MILES/FIELD JOB.
o
~//. "~
"j
5-0
JANUARY 1964
INSTALLATION DIVISION
MONTHLY OPERATIONS REPORT
DIVISION DETAIL
DISTRIBUTION OF·WORK TIME BY PERSONNEL CLASSIFICATION
(EQUIVALENT MAN-DAYS)
CURRENT
MONTH
(1)
(2)
LAST
MONTH
(1)
(2)
SAME MO.
LAST YR.
(1)
(2)
I-MONTH
AVG.
(1)
(2)
1609'
652
CLASSIFICATION
M.+To
1609
652
32
22
32
22
196
88
196
88
1838
761
1838
761
TRANSFERRED
BORROWED
CONTRACTOR
SUB-TOTAL
TOTAL
2599
/r'
\,4...J1;I
2599
NOTES. (l)PRODUCTIVE WORK (2)NON-PRODUCTIVE WORK
205
1/./1
I
C,.'
" ""
rn"G"" " If"
",."")""""
"f""Y&:ffiwfl"t!kH±fE"'"jf""6tririrtbHt "w"""Hd"bih 'b 6Hbt
c'
II
t
t
j
ttt
t
11"",,,,,,,**,,,»," II 'M '"
5-2
JANUARY 1964
INSTALLATION DIVISION
MONTHLY OPERATrONS REPORT
ANNAPOLIS DETAIL
DISTRIBUTION OF WORK TIME BY PERSONNEL CLASSIFICATION
(EQUIVALENT MAN-DAYS)
. CURRENT
MONTH
(1)
,(2)
LAST
MONTH
(1)
(2)
SAME MO.
LAST YR.
(I)
(2)
I-MONTH
AVG.
(1)
(2)
CLASSIFICATION
M.-!I.
C:
170
86
2
1
173
87
170
86
2
1
173
87
TRANSFERRED
BORROWED
CONTRACTOR
SUB-TOTAL
TOTAL
260
260
NOTES. {l)PRODUCTIVE WORK (2)NON-PRODUCTIVE WORK
'-'J
#"'t.!'i
206
u_ /./ ..
rr-,
l/'"
/c
6-0
JANUARY 1964
INSTALLATION DIVISION
MONTHLY OPERATIONS REPORT
DIVISION SUMMARY
DISTRIBUTION OF PRODUCTIVE TIME BY FUNCTIONAL CLASSIFICATION
(HRS/MAN-DAY)
CLASSIFICATION
SUPERVISION
I NSPECTI ON
CLERICAL
INS TALLATI ON
SELF-CONTAINED
TRANSFORMER
MISCELLANEOUS
TOTAL
STORM TROUBLE
'>'JTAL
CURRENT
MONTH
LAST
MONTH
SAME MO
LAST YR
I-MO.
AVG o
.62
1013
.14
.62
1013
.14
2.51
.18
.48
3.11
2.51
.18
.48
3.11
5 066
5.66
OISTRIBUTION OF JOB-RELATED NON-PRODUCTIVE TIME BY FUNCTIONAL CLASS.
(HRS/MAN-DAY)
,,,.("',,
CLASSIFICATION
HEADQUARTERS
PREPARG & COMPLTG DAYS WORK
MISCELLANEOUS
TOTAL
CURRENT
MONTH
LAST
MONTH
SAME MO
LAST YR
I-MO.
AVG o
.30
.18
.48
.30
.18
.48
.06
.08
.13
.21
'.15
6
.08
.13
.21
.15
FIELD
CGI
TURN-BACKS & TURN-DOWNS
MlSCELLANIOUS
TOTAL
TOTAL
DISTRIBUTION OF OTHER NON-PRODUCTIVE TIME BY FUNCTIONAL CLASS.
(HRS/MAN-DAY)
CURRENT LAST
SAME MO
I-MO.
MONTH
MONTH LAST YR
AVG.
CLASSIFICATION
MEETINGS
TRAINING
ON-THE-JOB
OFF-THE-JOB
ABSENCES COUNTED AS WORK TIME
TOTAL
.05
.05
.10
.30
.03
.48
.10
.30
.03
.48
2 u7
J-f-. I /~ i/
\~
- -- r,m"i:!ft"HS£wirif5iMTtt#'ii'j"diri'Ht#H "#t##rttrihY""
t"
-."".
tt t H
Itlf"P'wq
6-2
JANUARY 1964
INSTALLATION DIVISION
c
MONTHLY OPERATIONS REPORT
ANNAPOLIS SUMMARY
01 S TR I BUTI ON
OF PRODUCTI VE TI ME BY FUN' TIONAL CLASS I F I CATION
(HRS/MAN-DAY)
CURRENT
MONTH
CLASSIFICATION
SUPERVISION
I NSPECTI ON
CLERICAL
INS TALLA TI ON
SELF-CONTAINED
TRANSFORMER
.MI SCELLANEOUS
TOTAL
STORM TROUBLE
TOTAL
LAST
MONTH
SAME MO
LAST YR
I-MO.
AVG.
.70
.68
.73
.70
.68
.73
2.41
.21
.60
3.22
2.41
.21
.60
3.22
5.32
5.32
DISTRIBUTION OF JOB-RELATED NON-PRODUCTIVE TIME BY FUNCTIONAL CLASS.
(HRS/MAN-DAV)
CURRENT
r·10~n
CLASSIFICATION
LAST
HMO[\.! rq
SAME MO
I'-MO.
lAST YR
AV(j ..
HEADQUARTERS
PREPARG + COMPLTG DAYS WORK
MISCELLANEOUS
TOTAL
.30
.19
.49
.30
.19
.49
FIELD
CGI
TURN-BACKS + TURN-DOWNS
MISCELLANEOUS
TOTAL
TOTAL
.02
.11
.08
.21
.70
.02
.11
.08
.21
.70
DISTRIBUTION OF OTHER NON-PRODUCTIVE TIME BY FUNCTIONAL CLASS.
(HRS/MAN-DAY)
CURRENT
MONTH
CLASSIFICATION
MEETINGS
TRAINING
ON-THE-JOB
OFF-THE-JOB
ABSENCES COUNTED AS WORK TIME
TOTAL
LAST
MONTH
SAME MO
LAST YR
.03
.03
035
.03
.42
'.35
.03
.42
208
..w_
..
-t,u"-aW_igU¥~·.·~4~_
____
~,"(_~;~@",_£A_,nt_,AM_M;~_._~a
__________________
*~a
~
~
1-MO e
AVG.
__________________________________
II
/1· i.{:
........
_-.
__
._---
~--
7.. 0
JANUARY 19;6.4
INSTALLATION DIVISION
MONTHLVOPERATIONS REPORT
If~
~,,#
DI5TRI8UT10N OF JOB-RELATED TRAVEL DISTANCE BY WORK UNIT (1)
(MILES AND PER CENT)
CURRENT
MONTH
( 2 ) ( 3)
LAST
MONTH
(2)
(3)
SAME ·MO.
LAST YR.
(2)
(3)
I-MONTH
AVG.
( 2 ) ( 3)
:W·ORK UNI T
DlVISION
DIS TRICT
MONUMENT ST.
ANNAPOLIS
ESSEX
GLEN BURNIE
HOWARD
COCKEYS VI LLE
BEL AIR
WESTMINSTER
CONTRACTORS
27317
6039
5512
2499
2342
2119
4036
2348
2422
2837
27317
22.1
6039
5512
2499
2342
2119
4036
2348
.242,2
2837
20.2
9.1
8.6
7.8
14.8
8.6
8.9
10.4
22.1
20.2
9.1
8.6
7.8
14.8
8.6
8.9
10.4
if ''\
\lccY
DISTRIBUTION OF JOB-RELATED TRAVEL TIME BY WORK UNI~ (1)
. (VEHICLE HOURS AND PER CENT)
CURRENT
MONTH
(2)
(3)
LAST
MONTH
(2 )
, (3)
SAME MO.
LAST YR.
(2 ) ( 3 )
I-MONTH
AVG.
(2)
(3)
WORK UNIT
'DIVISION
OISTRfCT
MONUMENT ST.
ANNAPOLIS
esse'x
':SEi'EN BURNI E
HOWARD
COCKEYSVILLE
's-tiLAIR
W:!STMI NS TER
CttNTRACTORS
1396.0
1396.0
478.0
208.2
173.1
118.3
92.0
152.1
73.1
101.4
213.9
34.2
·14.9
12.4
8.5
6.6
10.9
5.2
7.3
15.3
478.0. '
2Q8.2
173.1
118.3
92.0
,152.1
r"l3.1
101.4,
213.9
34.2
14.9
12.4
8.5
6.6
10.'9
'5.2
7.3
15.3
c
EXCLUDING TRAVEL FOR SUPERVISION, INSPECTION, + CLERICAL.
TOTAL DISiANCEOR TIME.
PER' CENT OF DIVt SION TOTAL.
2ftH
I
8-0
JANUARY 1964
INSTALLATION DIVISION
MONTHLY OPERATIONS REPORT
DISTRIBUTION OFJOB RELATED CREW MEMBER TRAVEL TIME BY WORK UNIT
(MAN HOURS AND PER CENT)
CURRENT
MONTH
(1 )
(2 )
LAST
MONTH
(1)
(2)
SAME MO.
LAST YR.
(1 )
WORK UNIT
DIVISION
01 STRICT
MONUMENT ST.
ANNAPOLI S
ESSEX
GLEN BURNI E
HOWARD
COCKEYSVILLE
BEL AIR
WES TMI NS TER
CONTRACTORS
I-MONTH
AVG.
(2)
(1)
(2)
1440.6
1440.6
442.3
192.0
133.9
132.4
102.0
227.4
92.2
118.3
205.8
30.7
13.3
9.3
9.2
7.1
15.8
6.4
8.2
14.3
442.3
192.0
133.9
132.4
102.0
227.4
92.2
118.3
205.8
30.7
13.3
9.3
9.2
7.1
15.8
6.4
8.2
14.3
NOTES. (1) TOTAL MAN-HOURS MINUS TOTAL VEHICLE-HOURS.
(2) PER CENT OF DIVISION DIFFERENCE.
DISTRIBUTION OF JOB RELATED EXCESS CREW MEMBER TRAVEL TIME BY WORK UNIT
(MAN HOURS AND PER CENT)
CURRENT
MONTH
(1)
(2)
LAST
MONTH
(1)
(2)
SAME MO.
LAST YR.
(1)
(2).
I-MONTH
AVG.
(1)
(2)
WORK UNIT
01 VI SION
01 S TR I CT
o
MONUMENT ST.
ANNAPOLIS
ESSEX
GLEN BURNI E
'HOWARD
COCKEYSVILLE
BEL AIR
WES TMI NS TER
CONTRAC TORS
238.0
97.5
25.1
26.2
12.7
14.5
36.2
8.0
18.1
41.9
238.0
41.0
10.5
11.0
5.3
6.1
15.2
3".4
7.6
17.6
97.5
25.1
26 2
12.7
14.5
36.2
8.0
0
18.1
41.9
41.0
10.5
11.0
5.3
6.1
15 2
3.4
7.6
17.6
0
NOTES. (1) TOTAL TIME DUE TO CREW SIZE LARGER THAN NECESSARY.
(2) PER CENT OF DIVISION EXCESS.
21ft
'tIl /,./~I
9-0
JANUARY 1964
••..•.
___
.u
--
--
•.•.
me.
0
d.~
91
··rUW·T-
00
9-0
,
37
38
C'39
140
41
42
43
44
45
46
47
48
49
50
51,
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
-
940
978
843
496
476
483
919
933
946
841
938
475
972
910
942
480
939
945
944
743
909
846
957
811
921
745
773
917
479
489
474
853
488
494
407
821
020
908
922
713
492
907
918
844
935
765
755
983
953
911
865
984
803
863
03()
84.3
78.3
77.9
77.7
76.3
76.0
74.8
70.9
64.2
54.0
53.9
50.8
50.6
50.3
50.1
50.0
49.4
48.4
47.9
45.9
45.9
43.9
43.9
39.2
38.8
36.6
34.9
31.2
31.0
30.0
29.2
23.3
21.9
21.9
19.0
19.0
18.3
17.9
16.3
15.5
14.8
14.3
14.1
13 0 1
13.0
12.6
12.4
11.7
9.5
9.4
9.2
9.1
9.0
8.3
8.2
65
13
7
71
17
38
62
5
34
5
51
8
86
15
46
35
13
14
52
11
11
2
5
11
54
7
9
36
,17
88
17
2
47
30
.26
.73
.79
.53
1.30
6.02
11.13
1.09
4.49
2.00
1.21
14.18
1.89
10.80
1.06
6.35
.59
3.35
1.09
1.43
3.80
3.46
.92
4.17
4.17
21.95
8.78
3.56
.72
5.23
3.88
.87
1.82
.34
1.72
11.65
.47
.73
18
35
10
1
2
28
7
11
.83
.03
.44
.33
.58
.68
.40
.66
1
7
2
13
9
3
2
15
3
2
10
1.17
.17
.37
.16
.09
.88
1.04
.29
.02
.33
.67
.60
1.40
.43
.01
.58
.29
.42
1.07
.68
.59
.93
2.02
.29
.67
.56
.66
.62
.77
.17
.22
1.37
.34
PAGE 2
1.80
1.80
1.12
1.12
7.13
1.28
1.00
.96
7.13
1.28
1.00
.96
1.61
1.05
1.37
1.61
1.05
1.37
2.31
3.12
1 26
3.50
1.34
1.48
1.53
1.16
2.31
3.12
1.26
3.50
1.34
1.48
1.53
1.16
3.59
1.91
1.58
3.59
1.91
1.58
.75
.75
1.10
1.10
1.06
.52
1.79
16.30
7.75
.53
2.04
1.28
1.14
1.20
.84
1.63
11.65
1.36
.63
.94
1.14
1.20
.84
1.63
11.65
1.36
.63
.94
13.00
1.80
6.20
.90
1.06
3.13
4.60
.61
3.00
4.15
.82
1.30
1.67
2.57
.83
1.06
1.04
1.62
1.82
1.47
1.56
1.36
1.30
1.67
2.57
.83
1.06
1.04
1.62
1.82
1.47
1.56
1.36
0
2]2
~I:
/ /, /6
9-0
92
93
- 94
95
96
97
98
99
100
101
102
103
104105
106
107
108
'109
110
111
112
113
11. t~,
115
116
,117
11~
119
120
121
,122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
TOTAL
793
498
950
763
943
080
486
200
491
851
963
916
783
435
752
126
725
090
485
733
602
969
715
813
971
601
120
9'/9
434
923
449
723
958
603
110
964
432
861
439
408
422
013
116
418
437
023
117
428
431
562
440
609
442
559
7.1
7.0
6.6
6.5
6.3
6.2
6.0
5.9
5.8
5.6
4.8
4 06
4.5
4.3
4.0
3.8
3.6
3.3
2.8
2.5
2.4
2.4
2.3
2.3
2.3
2.2
2.0
2.0
1.9
1.9
1.8
1.7
1.7
1.6
1.5
1.5
1.4
104
1.2
1.0
1.0
.8
.8
.8
08
.7
.7
.6
.6
.6
.5
.5
.4
.2
17,957.6
5
9
3
1
5
16
7
16
10
2
1
3
1
10
1
4
2
6
1
2
10
2
1
1
4
1
2
1
4
1
7
1
1
3
2
.27
.46
.67
.42
.01
.31
.62
.45
.60
.30
.27
.89
.22
1.05
.21
.67
.SO
.41
.73
.42
.26
.14
1
4
1
.24
3
.90
4
1
1
6
3
1
1
3
2
1
1
1
2
1
.33
.33
7,242
.45
.42
.21
.08
.42
.23
.16
.17
1.42
.78
2.20
6.50
1.26
.39
.86
.37
.58
2.80
4.80
1.53
4.50
.43
4.00
.95
1.80
.55
2080
1.25
.24
1.20
2.30
2.30
.58
2.20
1.00
2.00
.48
1.90
.26
1.70
1.70
.53
.75
1.50
.35
1.40
.40
.25
.80
.80
.13
.27
.70
.70
.20
.30
.60
.50
.50
.20
.20
PAGE 3
.92
.92
2.27
8.62
.99
.96
3.03
.82
1.48
.20
036
.99
1.29
1.13
2.29
1.15
2.00
1.37
2.27
8.62
.99
.96
3.03
.82
1.48
.20
.36
.99
1.29
1.13
2.29
1.15
2.00
1.37
2.86
2.86
2.92
2.55
.82
1.50
2.92
2.55
.82
1.50
1.44
1.44
.91
.16
.95
2.22
1.32
.91
.16
.95
2.22
1.32
1.09
.38
1.60
.53
.68
1.09
.38
1.60
.53
.68
.56
1.02
.94
.561.02
.94
.91
076
.91
.76
1.31
.67
1.00
1.31
.67
1.00
.69
•.69
21:]
;i~~
II. /7
C
C
C
9-2
JANUARY 1964
('\
INSTALLATION DIVISION
~J
MONTHLY OPERATIONS REPORT
ANNAPOLIS SUMMARY
DISTRIBUTION OF WORK TIME (*)(EXCLUDING DEFINED TRAVEL TIME) BY FUNCTION
PERFORMANCE
INDEX
CURRENT
MONTH
LINE
NO.
FUNC.
NO.
TOTAL
TIME
(MHRS
C:,
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
771
902
900
901
781
477
481
941
915
845
947
493
751
937
741
933
482
407
484
020
791
499
976
801
977
487
944
945
913
946
984
910
030
721
496
921
~*)
199.8
189~1
181.3
175.8
124.5
90.8
58.6
50.2
40.8
34.9
30.6
29.6
24.3
21.6
20.4
20.2
19.1
19.0
18.8
18.3
17.6
16 0
15.4
12.6
11.5
11.3
10.6
10.5
10.1
9.6
9.1
8.3
8.2
7.8
7.2
6.6
0
NO.
JOBS
AVERAGE TIME
TRAVEL FUNC.
(MHRS) (MHRS)
PERF.
INDEX
LAST
MO.
SAME MO.
LAST YR.
12-MO.
AVG.
71
4
.95
2.81
47.27
1.36
1.36
34
.90
3.66
1.18
1.18
105
19 .
1.
3
18
·39
7
8
9
1
107
.92
.91
1.36
.72
.80
.90
1.05
.47
1.40
.85
.55
2.64
3.70
11.63
1.70
.75
3.47
2.70
2.26
20.20
.17
1.56
1.37
.74
.83
1.13
1.23
2.67
1.16
2.09
10.10
.54
1.56
1.37
.74
.83
1.13
1,,23
2.67
1.16
2.09
10.10
.54
1.20
1.57
1.20
1.57
3.49
3.49
1.52
·1.58
2.03
1.28
1.82
1.45
1.36
3.16
1.52
1.58
2.03
1.28
1.82
1.45
1.36
3.16
3
35
12
27
22
2
6
2
7
5
3
5
15
2
10
4
18
9
.03
.51
.03
1.08
.66
.67
.89
.43
.94
.87
.34
.50
.20
.72
6.26
.52
1.46
.59
.70,
6.30
1.91
5.65
1.51
2.10
3.36
1.92
.60
4.15
.82
1.95
.40
.73
2I
!~
/,
-.-
..-...... ,,-
... ...... ...... "'.-.-•...
~
~
______.... _M. _ _ _
~
....................·.. '""···,'·'·" ..
~·,,···,·_~·'·
. , ..
,--""-~---,,,,-,-,...:;,.:~..::.':"
.. ';- .. :..
::::.:.:.~;~;
..; .•.:..,-.-.-:::....:...•
..
:~;::.:.:;.;.;....::.::;.~ ~:. ".~
.
o.
~"
•• " ••••• _._~_. __
.~
.... _ _ _ ..... , ...... " '
9-2
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
TOTAL
480
495
917
080
811
200
476
821
975
435
483
490
773
126
474
951
904
090
939
475
731
479
489
602
942
120
434
983
449
711
603
918
110
488
432
938
940
408
422
013
116
418
437
439
023
117
492
428
431
562
440
486
609
761
442
559
6.4
6.4
6.4
6.2
6.1
5.9
5.8
5.6
5.0
4.3
4.2
4.2
4.0
3.8
3.8
3.8
3.4
3.3
3.1
3.0
2.9
2.5
2.4
2.4
2.1
2.0
1.9
1.9
1.8
1.8
1.6
1.6
1.5
1.5
104
1.3
1.3
1.0
1.0
.8
.8
.8
.8
.8
.7
.7
.7
.6
.6
.6
.5
.5
.5
.5
.4
.2
1,679.2
6
6
15
16
2
16
2
4
4
10
2
13
1
4
1
1
2
6
2
2
1
23
10
2
2
4
2
7
1
3
2
2
7
4
4
1
4
1
1
6
3
2
1
1
3
3
2
1
1
1
1
1
2
1
.42
.32
.01
.68
.31
.81
1.18
.27
.13
.78
1.00
.77
.14
.22
.63
1.85
.21
.33
.98
.26
1.65
.14
.80
.76
.24
.11
.33
.33
.45
.26
.48
.42
.21
.08
.42
.23
.16.17
1.06
1.06
.42
.38
3.05
.36
2.90
1.40
1.25
.43
2.10
.32
4.00
.95
3.80
3.80
1.70
.55
1.55
PAGE 2
.64
.96
1.19
.82
1.19
.82
1.55
1.55
1.13
1.13
1.04
1.71
1.15
1.04
1.71
1.15
5.63
2.86
1.37
1.31
5.63
2.86
1.37
1.31
3.48
3.48
1.26
1.44
.91
.85
.95
3.00
1.26
1.44
.91
• 85 (t'-~\
• 95 \",
3.00
.67
1.09
067
1.09
1.60
.31
1.11
1.60
• 31
1.11
.56
1.02
.94
.56
1.02
.94
.64
.91
.76
.90
.64
.91
.76
.90
1.31
.67
1.00
3.00
1.31
.67
1.00
3.00
.50
.75
.20
.20
.69
.75
.69
1.45
2.50
.10
.24
1.05
1.00
.47
.95
.25
1080
.53
.80
.75
.21
.35
.32
1.30
.25
.80
.80
.13
.26
.40
.70
.70
.23
.20
.30
.60
.50
.50
.50
.644(~""
.96~0'
215
848
-
_
...
-
- ... -..... -
..
~/.//'
--
._.
- ..
'
19.
C'
10-0
JANUARY 1964
METER AND INSTALLATION DEPARTMENT
INSTALLATION DIVISION
MONTHLY OPERATIONS REPORT
CONTRACTOR DETAIL
DISTR1BUTION OF WORK TIME (INCLUDING DEFINED TRAVEL TIME)
BY BUDGET CATEGORY
(MAN- HOURS)
CATEGORY
SPONSORED
BY M.&I.
OPERATION
(~II\'
449.8
27.5
119.0
1139.0
482.2
1621.2
81.9
1757.5
TOTAL
DEFINED TRAVEL TIME
DEFFERENCE
TOTAL
91.5
MISCELLANEOUS
SUB-TOTAL
SPONSORED
BY OTHERS
4.6
445.2
MAINTENANCE
CONSTRUCTION
SPONSORED
BY E.D.D.
81.9
509.7
2271.8
4.6
2271.8
419.7
1852.1
21 6
11-0
JANUARY 1964
METER AND INSTALLATION DEPARTMENT
",". #
INSTALLATION DIVISION
MONTHLY OPERATIONS REPORT
SERVICES INSTALLED AND CHANGED IN DWELLING UNITS
UNDER·
100
100
AMPERES
- OVER
AMPERES
TOTAL
RESIDENCES
NEW
INSTALLED
26
452
418
OLD
INSTALLED
CHANGED
26
161
26
311
52
412
NSTALLED
220
229
449
OLD
INS TALLED
CHANGED
48
52
APARTMENTS
NEW
'I
___ TOTAL
.. _-- .....
~--
..
~".-
.. - .-_ .
'.__ ¥_'.--'--'."_ .... ,._. -
1144
\'.Y
65
96
-
NUMBER
PER CENT
533
33.1
1019
66.9
1612
100.0
c
,
%11
....... ---~-... --.---
(-r~
I
----.------ -
-'- -,- ..
-
.. .,..
~
-
- _ ... -
----r
----.-.-.~-----.---.---
.. - ..--.--.---- .. -
.....
,~.-
.-.--.---~
----.. ---.--- .. ---- '-'-"-",.- ----- .. - . - - _.-
A MULTIPLE-ITERATION PROCEDURE FOR ENGINEERING DESIGN
Meriwether L. Baxter, Jr.
Chief, r'lathematical Research
Gleason Works
Rochester, N.Y.
c::
Prepared for presentation -at the 1620 Users Group
Eastern Region Meeting
vJashington, D. C.
May 6-8, 1964
c'~
2 1~
A MULTIPLE-ITERATION PROCEDURE FOR ENGINEERING DESIGN
Introduction
The process of design is essentially one of synthesis; but
in most real engineering design problems direct synthesis 1s
out of the question due to the' c0mplex interrelation of variables.
Methods of
analysis~
on the other
hand~
have vastly
improved~
and
there is hardly any structure or mechanism that cannot be analyzed
(\,lith the aid of computers) to determine quantitatively any
property we ·may wish to s'tudy.
Fo~tunately~
of
iteration~
we can convert analysis into synthesis by means
and obtain thereby a powerful design tool.
The
purpose of this paper is to describe one such procedure that has
been found effective at the Gleason Works.
Statement of the Problem
The elements of the problem are the following:
1) A set of n variable design parameters
(a~ b~
c ---)j
2) A ca1:cuation procedure of any complexity based on
the parameters;
3) A set of n results·" functionally dependent on the
n
parameters~
whose desired values are kno't'm.
The problem is to establish a procedure for determining the
values of the n parameters that will yield the desired n results.
2 1U
Obviously, some restrictions are necessary for such a
procedure to work, and some discussion of them will come later
in this paper.
However at this
~tage
it is sufficient to
require only that
1) a solution to the problem probably exists (a
matter of engineering judgment or experience)
2) a set of reasonable starting values for the
n parameters can be chosen.
Outline of the Method
It is not particularly difficult to handle multiple
iterat·ion when the relating functions are simple and short
enough to make differentiation easy.
problems,howeve~
In many engineering
this is not the case and we need a method
which does not require any analysis of the functions involved"
but merely the ability to calculate them r'apidly.
Let us say that if VIe are given values of three parameters
a, b, c, we can
on a, b", c.
calculcd;~
three quantities" x" y, z which depend
It is not, however, possible to calculate a, b, c
from given values of
:1..,
y, z.
Nevertheless we must find
what values of a, b, c to choose to make x = y
=
someho~·,r
z = 0, and ,'!e
know as a practical fact that such a solution exists.
22ft
-2-
-----------.. . . . .---zw_. . __________
..
,!III".!,.!11144.,..""·4,..~_iiI!!!iIi!iIIIlM
• .iiiijl'IJIl'Ilj;:z1illillliR4A1ii&_A!Ifii1!ij,.4-~,i\i~,¥.. -,£'JJ.!!!IIi'l.'m,i¥l-,.'iIf#!'!!!!,.i!F"1"!'-h\4"'"l1l.4f,~i~A.Y!!I'I\!.·
.. M-.,..MM.i\!m
... ,,,,,,".,,."""'4=.~""""·.:·q_.'"""I.'_
..I."""-_v,::..'
WIl\lir.lIQi,,!'i'.:M:;;m¥\i.UaX"i¢ftml'l"iG~,,*,:·,m,,;il"!lll'(4fT1l,~,,+~hg~,~·
~iI!!I!2.I.t
If we make a small change in a and determine its effect on
x,
y
and z, we have obtained approximations of three partial
derivatives
Similarly by stepping band c small amounts we can
com~lete
the
matrix of· 9 partial derivatives, or rates, and vlith these we can
calculate changes in a J b, c
linear equations.
~erely
solving
by
t~ree
sirr~lta~eous
Repeated application of this procedure will
produce convergence, if
a) there 1s a unique ans't':er
b) the starting values of the parameters are reasonable
c) high accuracy is maintained in COrT.putc..tic:'1."
particularly in the solution of the
li~~ar
equations.
Figure I is a computation flovJ-CD.2rt i'0:.:
handling such an iterative procedure; not all
·:';':-.0
: .. ;:~~:
o~'"
~ecessary
details
are included, but only enough to indicate the plan.
He begin by initializinz the star'cing valuc::s of . . . " oJ c and
specifying the small steps
the first pass.
~aJ ~b, ~c
to be used to get ratos in
We also set switch I for the first pass.
The computational block calculates x, y and z, and the
test block determines whether the errors are sufficiently
If so, we exit from the iteration procram.
fail and control will pass to switch 1.
-3-
s~all.
Usually the t8st will
Branch (1) of switch 1 resets itself to branch (2)~ proceeds
to set s'Vlitch
2~
computation.
On subsequent passes a net-I[ set of rates is figured
step a, store the last a.."YlSVlerS, and return to
following each change in a, b, or c.
\fnen three sets of rates have been
t:.a,
~b~
determined~
new values of
/J.c are calculated '\'lhich are first approximations to a
solution.
If we had reason to believe the functions \'lere
lineal~
it would be desirable to subs';;itute all of these changes at once;
however it is generally more practical to apply the changes one
at a time as was done
wit~
the original steps, obtaining new
rates at each stage in preparation for subsequent passes.
As a
result, we only expect that the test may be passed every third
time through the loop.
This procedure, for
two, three or more variables, has
one~
been of tremendous value in our engineering co:r:putz.. tieD; ho:-.rever
it must of course be applied intelligently.
troublesome condition is unsuspected
functional relationships; if
SUCrl
O1':.e
part::'c~larly
disconti~~ity
in the
ciiscont1nuities exis-c they
may be located in advance, and the program modified to prevent
stepping across this region.
Multiple solutions have not been
troubleso~e
since their
existence can usually be predicted; sufficiently good starting
values will force the
solu~ion
we are after.
-4-
~hen
the iteration does not converge, and the program has
been carefully written to avoid round-off errors, we usually
find that there is a physical meaning corresponding to the
non-convergence.
For example, in certain gear. analysis programs,
failure to converge indicates either
a) line contact instead of point contact (infinite
'number of solutions), or
b) contact at or near the cusp of a generated surface
(dual solutions very close together).
Use as a Suborogram (FORTF.AN)
Iteration of the type described has sufficiently varied
usefulness in our work to justify preparing standard
for various numbers of variables.
subprogra~s
The subprogram approach is
also desirable for us since the iteration seldom constitutes the
\-Ihole problem, but is more likely to represer. . c.
r.:e:~ely
a detail
in the main-line computation.
It is desirable to leave as many controls as possible in the
hands of the progr8.mmerj we therefore set up the
:~-v,:.)::!rosra:n
-'co
include only the necessary mechanics of iteration, leaving to the
main-line prograEl such
i ~-;ems
as
a) initializing the variable parameters
b) testing the results
c) takinG action on failure of iteration.
We also permit the
progr~~er
to specify in his CALL
state~ent
the values of the initial steps of the parameters.
-5-
._._-------.
------
._..- -
- -
223
c
$ "ittritteWt±ttd""fif m
. dillN'If"Jij
It is recommended that the iteration loop be made within
the range of a DO
statement~
to provide easy initializing of
the iteration and an automatic count for stopping iteration
after a preset number of tries.
Figure"2
INTP3A.
ShO'\'lS a
program employing the iteration subprogram
This program Vias designed as a short test of the
subroutine and includes its own data.
Naturally a real pr'ogram
would be much more complex.
Starting values of the parameters
A~
B, C are set at +25°,
-25°, and +25°, and the f.unctional relations are
x=
sin A+ sin B = 0
y = sin (A + C)
-
(.ll.. + B = 0)
sin 50° = 0
Z = sin (B + C) + sin 10°
::;:;
0
(A + C
:::;::;
~,"",O)
';IV
(B + C =-10° )
It will be seen that the c0rrect final values of A, B, C fulfilling
Notice that the main calculation and the test lie
the range of the DO loop.
~ithin
Tne limit of 100 for t::e inc.8X I
permits about 33 tries before
exiti~g;
this is
prob~bly ~or2
than will generally be needed.
Failure of the test sends contr'ol to th2
1'lhich
changes the value of ei ther A:; BJ or C,
2.S
t·he dashed line in the Figure; control is then
CONTINUE statement so as to restart calculation
of the DO loop.
indicated
r.etu~ned ~~·o
a~
the
b~r
t:le
be;ir.~ing
Passing of the test transfers out of t:1.8 loop.
o
-6-
~_",==GMWU===i"'=;;
;:;,,,,,;;;,,;,; h,
4ii'.#M " . . .
)i~
II'
.
I~. I
Tt.10
refinements that we generally employ are not shown in
this example.
One of these is to follow the CONTINUE statement
1'lith a PRINT statement indicating that convergence has failed
within the specified count, followed by other appropriate action.
Tne second is a sense-sw1tch-controlled PRINT statement, just
before the test, causing print-out of A, B, C, X, Y, Z, I
each time through the loop for tracing purposes when convergence
is in question.
Structure of the Subprogram
Since the logic of the subprogram is essentially represented
by the right-hand part of Figure 1, it is not
flow' chart.
used.
Figure 3 sho:1s the FORTR.!:..N l::'s tine
Switch 1 of
Figul~e
and switch 2 by the index
1 is
~~epr'es8nted
to show a
neces~ary
by
2.S
no:'! being
"the DO index I,
K~
One addition has been xade to cover the
sometime during the iteration the c2.1culated
poss~bility
incl~eme~t
that
of oY).e of
the parameters ,may be zero" rcsultin3 in division by zer'o 1.-::-'2:'--':'
new rates are being computed.
In reviewing this prozram at this time I 'Delieve ·
several ways in 't.'!nich it might be improved.
use appropriate subscripting (if you have
S3G
One of these: is to
~ co~pilcr
that can
handle subscripting efficiently); another possibility is to call
a separate subroutine to solve the set of
simalta~eous
linear
equations.
22~
-7-
C
Extension of this method to handle other numbers of
variables should present no difficulty; one can also consider
the possibility of
1~riting
a general procedure for handling
n variables.
Co~~ents
and Conclusions
Since we have established this method without proof
of existence or convergence, it is essential that it
applied with care and common sense.
here that
\~ill
There is no
oe
~agic
produce ans'V'ers Hher'e nor-J.e exist.
Within these limitations there are
~~~y e~zi~eo~~ng
design problems w"here this f:1ultlple iteratio:1 pl'"'ocea.ure is
extremely valuable.
226
-8-
. r
44.. .
4,\ AAga4
INITIALIZE
a, b, c
60,
6b,
6C
SW1~1
-.
STORE LAST X,Y,l
'
1
COMPUTE
~+~b
X(a, b, c)
0+60
Y (0, b, c)
C +6C
Z(o, b, c)
-
SW1<-2
.
,.
SW2-<-1
SW2~2
S~V2~3
j~
~
f
i
TEST
I
X=y=Z=:O ?
NO
,.
~
COMPUTE
SVv'1
60,
1
3~2
y
~
l'....
,
l'~
6C
2
---lYES-ON
6b,
2
COMPUTE
RATES
'---------
::.J
N
~
~
~
'\ /
!:.,
~. )
,
c~\
INTP3A TEST
C
P. == ADEG/ 57 . 29 58
~------ - - - - - - - j
I"
B==-.A.
DO 1
1==1
J
1 00, 1
I
I
---"
-<
j, \
i
X=SINF(A)+SINF(B)
Y=S I !~F( A+C )-0.76604
Z==SINFlB+C)+O.17365
IF(ABSF(X)+ABSF(Y)+ABSF(Z)-.00002)2,2,3
3 CALL INTP3A(A,B,C,X, Y,Z, I, .01, .01, .01.)
C,
,j"'"
CO~~T
2
I t~U E
ADEG=A'i'~57.
BDEG=Bi'~57
2958
. 2958
C0 EG = Ci'~ 5"1 • 29 58
PRINT 4,ADEG,BDEG,CDEG, I
4 F 0 Ri'1 A T ( 4 X , 3 ( F 7 . 4 , 2 X) , I 3 )
STOP 11111
END
rig.2
228
t -',
H- /. . .
1}
__
__ii1t1nw:mQ,&l'IfW ,.. ¥,f,?+".",ttt.f\.~"
T
AT
•
't)
SUB R0UTI NE I NTP3f.\ ( A, B, C, X, Y, Z, I ,A STEP, 8 STEP, CST EP )
IFl!-1)1,1,2
DA::;;ASTEP
DB=BSTEP
C::;C~DC
DC=CSTEP
GO TO 4
3 K= 1
11
t,=A+DA
4
1~(DC}12~13,12
12 DXDC=(X-XO)/DC
xo=x
DYDC=(Y-YO)/OC
YQ=Y
DZDC=(Z-ZO)/OC
ZQ=Z
13 Dl=DYDS*OZDC-OYDC*OZDB
RETURN
02=DXDS*OZOC-DXDC*OZDB
2 GO TO(5,8,11),K
D3=OXDB*OYDC-OXDC*OYDB
5 IFCOA)6,7,6
OEN=DXDA*Dl-DYOA*D2+DZDA*D3
'" DXDA=(X-XQ)/DA
OA=(Y*02-X*Ol-Z*03)/DEN
0
DYDA=(Y-YO)/OA
DS=(OXDA*(Z*OYDC-Y*OZDC)-JYDA*(Z*DXDC-X*OZDC)
OZDA=(Z-ZO)/OA
+DZDA*(Y*DXDC-X*DYDC))/DE~
7 K=2
B=B+DB
+DZOA*(X*DYJB-Y*OXD8))/~E~
.\.
GO TO 4
GO TO 3
8 IF(OB)9,10,9
9 OX08=(X-XO)/OB
OYOB=(Y-YO)/DB
DZDB=(Z-ZO)/OB
Fig.3
)-/ i~.
II
(,,
',
",,,,
~'
"A Distribution Feeder and Substation Load Fore casting System"
by
J. C. Hubbard
Baltimore Gas and Electric C
..... ,.,
rf''\
'-.,Ii!
-4For each item a complete set or input data can be entered on from two to
tour cards, according to whether and what type of adjustments are applied. The
first card tor each ita. includes a field tor the variable indicating how IUI~
cards are to be used. Input data are assembled by mastersJ within the master,
by stations. Items are processed one at a time. 1s each card is read, checks
are made on the correctness or its order and on whether it is the rinal data
card tor the item. All identification fields on each card are cheeked, to
prevent the mixing of data.
Each item proceeds by its appropriate path to its forecast peak loads.
all distribution substations and CU,.,tomer stations set a program
evitch to positive values, and cumulate their peak load components. The switch
. is set to zero betore the next itenl's data are read. Individual reeders of a
multi-reeder station cumulate their _ peak loads) the switch is lett to have
the value "zero".
In the case of a multi-teeder station, the forecast is computed first
from reference and growth rate. Later the program switch initiates the computation of certain of its forecast peak loads by' means of the summation of its
feeder ~aks and its diversity factor. The results are compared. Where the
difference is outside of' a preset allowable percentage difference, the fact is
so noted in output; the note serves as an indication to the forecaster that
further study of the station and its feeders is called tor.
For a maeter substation, the forecast is computed by use of its
diversity factors and the summation ot the peaks of the stations which it
supplies. The program switch is set negative.
The initialisation procedure preceding each !roup of input cards 1s
determined by the value (positive, lero, negative) ot the program switch.
On the
(
""
~v
W87,
Output for each i tam includes the actual load and the forecast load in
mva and components tor all five years. The llYar of capacitors considered to
be in service during the period of the forecast appears also. The mvar component
of the forecast is that which the reeder or station must carry. This output,
by means of an SPS fonnat modification program, becomes the five-;y:ear foreeaet
for feeders and substations.
Program
4:
The revised forecast
In practice, the initial forecasts are made strictlJ on the basis ot
change wi thin a given reeder or station load area. They are released for study
in connection wi th system plaming and developnent. The studies result in a
schedule ot load transfers and capacitor instillations prior to each seaeon.
The planning schedule is inserted into • modified forecasting program.
This hae all of the inputs ot the earlier program; provision is made also tor
including the capac itor changes and the IlW and mar components ot load traIlS tere
for each ot the first three torecast years.
The output ot this program is identical in format with that ot the
initial torecaet. The new figures nveal the etfects of the planning schedule,
and assist in planning tor the intermediate-term. Full annotation describes
how the revised torecasts difter trom the original.
C
,·I.".
,I
J. C. Hubbard
Bal timore Gas and Elec tric Company
May S, 1964
J-j. ,'!} ;j
LIST OF J'I0UItIS
ProF.
1
now
Maar..
Input-Output Lists
Sample 'ag. of Input Data
Saple Ol1tput
ProS!'rlpre S
6
7
nov
2
D1.~..
Designation of Variabl••
Saple Qltpl t
Prop. 3
11_. 8
9
10
Iuput-Output Lilt.
Sample Page of IDput Data
nov Maar-
'll'1r. 11
Sample Pa.. of Foree ••t
235
~ ..I I~
,'t.
. , ;;1.
/
.:1
FEEDE./ff AND TRANSFORMER CIILCS.
CRW + JCH, ENitNEEIf,". STAT.
5-31-"'2.., (MOD. "-/¥-fJ2~DIAI&.REV.
LFP7
MJ
LiP=O ...-..-...t' Nfl r=O
2
READ)
100
j
...._.. \
--,
I
\
i NN=I
J
\
If
MM-!
MM-O
5'1 CALC
>-----4ii--.... YL 11
60 .,0 2.
pr:'.1.
2, Y,X
PF~
S
CALC
LFP=LF
LTP=LT
CALC
Xl') (;
YL)PF
PF::PFZ
0
YL;YLZ
-
PF-PFI
'--------4
YL =Y L 1
o
F16. I
i&&I&i&WMWIJIIlJ.lMi(t\•. ;
::ct,:; ,,- ... \#,¥I
-
m
.
"iiG\¥.iii\i\~'i'.iiMiW'I!""" .k.,n.,Q::;Iifl?!;
.f .; . . 1.. ·...... Ji¥Iid+..""'¥h.. ,·,;:'iw;::r., .•';
#,~iii4.L..,.,., •. t.."' •.. __ ._ .. :.. __ ".. .'._~ .....•,:_
: .• ,.. ~_ .. ,.'
'T.
H
I~
1:1
'I
~~~~io~~
~-o~~
n
• •
em
t
~u
~Ti
o~
L~
~lf)
vi
V
~r
~
V
V'
V
V
V
I
N
J...y
XA
XP
~·t~
~~
~
(
o.JjdinwHt
~~ ~ )-Jfffi11
V
,,,-r-~
YC
hN/t1Jd ~.
vi
V
A
~)Jff
v'
~)
V
V
V
~
V
V
/YI1JJ'rJJJyJ
V
,/
-#WwJJd:v
V
,/
:z.
X
Y
PF
YL
X])
r;
J)X
])y
cwwftTj; ~
~~~
~mtwKkJ
I.
\',J!
V
y
~NiJj)
V
~~~
V
~~~
Flo,2
V
237
Jf.;g.7
(.;
n
o
~
'.
~(OE~ AND TQ.~NSI="OQ..t.1E1l DATA $l.IEET·
@
=IF~/q
.
~e:OCIt
~
TIt,,~
5
.
..
(
Yl~,·SI~
~
ttWM~
\~ 15
.!t1O
~
510
I
. ~1"l9
I
'.3.
I
'3d
'.6 ()
AW"STMUIT
20
-;
l>~cv. T•• ""'"
tl\WMIl
'\1
25
J
~
.L/-; 7.Q ~
)1 2.0
1
!
(J
!1/-5
.~3.1
\
~
,10 I
.7iAtJ
.L.311
!()
bl
?J..7.QQ
1. 3. I
.7~.~
.'-311
I
11t-.s
1 (J
I
.~
i ()
.lJ..1.'1.2
.7.J.t. .2- £..3.1
! S?
.7.~.()
.7.'1 c
3.0
·.7 t!J c
l'~
R.j ~
~
j,
.'4-.7.1 I
.<1
.L
.(,.3. J
.f.. 3.1
.Ll-.7.1
1
~
.tJ..7/.f
,
i;
i
QJ
.11-: 11 q
I
~~L
g~.,
,
. ,i7.g
.I
.
:
2l5~
.2L;.1
I
i
.
i
J
12:9
. 12l
J L~.R
i
.~
~"Q
I
'J J
17.~
I
25. 1
I
i
.t./..~1q
i
I
·
J.J5.3
210
.2:5~
.352
J; I
5?
I
r.
.L.3. J
I
G
.t..3./
. t, 3.1
2 .t,3./
' c
. I (;
I"
'
'J
1('
:7. ~
jCJ3
t;~
J!~ ~
3.0
. i,.:;
170
FIG,3
25~
-
I~
21,5.~
1
g~
!
.~.~
i
·
LJ.l 7 .~
I
25<
-
.~.Lf
~_'i
.7.00
J,Q.4
~
;
'I
t;
j
.K 1
;
2.LJ.!'
.
i
.~
'.
2:5-"
l
"
'
t;
I
'is~
.R,!J:
JL3
'3.1
.~.3./
.l. 31 I
j"lJ
1
..
7C,
5. I
.8.'
.i.3
,1tJ.CJ
'J.
!
~I 5?
/ .3.9
i
~'.Q
2:5
I·
i
f70
~~.a
g"
-117
.
.
5.0
t.
.~3g
,
i
!
):5.3
} .t. 3.1
JI~3
}I~l
K.'
I
I
I
:
I
i
I
ITt] 4
g~
. !J~
I: 1. £::
~
l
!
I
113.!
11)
.
-1/5
,
j
3.3.'
701- 72
""~m
.
65£('
:
i
./.J-;3.'
PMW'
~1l"Q.
12 lJ.~
I
!7!i
'" 1.("
.t2 if.tt
· II, g
I
'!1/.1
. i;J)
()
zJ2[J
i
t'Alft::
't1ACT.e....... UM"!.!Q
~
II
,
}'3.g
gil..
gJ.~
.J.)
~-q
~
55
. ~ CJ
I
~
1',9 .
,\WNtt
e,].,
50
~5
FDJ:J )
~CT"'EClI4M11""
2 .~.3.1
.
i
.LI-.7_' 2
I
.
!,'.11
l3.~
;
I
; ,C
0..7
I
~j_'1J
"1VAMCl,
;l5CJ
7.'.tJ
i
?t1.t.
" .. PEqq
- ' ..0-
£j
J ;2J).
t..3.1
I(V
....
35.
I
lJ
1
...
l
I
i
.7.~.tJ
.
,ka
i
.LJ..7J5
.4.1. I.~
"\t-_
v
"UIP"'!ENT WM(
Jj ~ (]
.,.~, J
J
t1
~UIt.~EMT'
J
i
.t. j .li
.-.
"O&'TA~£
~
1
(,~~,I
C"PS."T "~K 11..,
J-';!-
gj
;
,
I
' } 'D.?
-1.4-
II!
~
I
Fdr
Sta,
No. No.
0
No.
of
~s
Yr.,
Seas
8 1 632
Fdr. ar sta. Peak
- Mva
Mw
1.15
1.15
Mvar
Load
PF
-.05 87.
%
D. F.
Mw Growth Mw Mvar
Adj.
Mvar
.64
1.15
1.7 100. 100.
(if'
~J/
o
136 1 632
1.19
1.19
-.15 93.
.45
1.19
4.3 100. 100.
6~2
1.44
1.44
-.09 87.
.81
1.69
3.0 100. 100.
o
1076 1 632
.89
.81
• 37 90.
.37
.81
2.5 100. 100 •
o
2076 1 632
• 81
-.31 94.
.26
.75 -3.8 100. 100 •
1.58
.61 83.
1.06
1.58
8.2
.97
7.4
o
4780
56 1
196
a
632
1.69
•
•
Q
•
0
•
•
•
.~
4783
196 0 632
1.03
• 97
-.34 87 •
.55
4784
196 0 632
.69
.69
.01 83.
.46
1.30 14.5
4786
196 0 632
1.46
1.46
-.06 83.
.98
1.46
3.2
0
196 4 632
4.40
4.40
-.10 83.
2.95
7.99
6.5 107. 104.
4796
232 0 632
.34
.32
.14 91.
.14
.32 14.2
•
•
4797
232 0 632
• 86
.59
-.63 90 •
.27
.59
1.7
•
•
o
232 2 632
1.00
• 88
-.49 90 •
.41
.88
4.7 103. 100.
o
300 1 632
.88
• 85
.26 86 •
• 49
.85
3.6 100. 100 •
o
302 1 632
1.34
1.06
-.82 84.
.68
1.06 -6.1 100. 100.
o
1304 1 632
1.35
1.34
.23 89.
.68
1.34 13.5 100. 100.
o 2304 1 632
4.34
4.30
-.60 94.
1.50
4.30 16.2 100. 100.
(4~",
I
'~p
'l ,)
~
~.,;
• ,
.,~'
l-/
I
I}
cJ .
f>~'tJaJYN *334-
1-pwJ;
o/amn,Uffw
kd.J .wr,·'Jt ~.vx
W i:bit. .=:-.,.~~'!~"!:~~ _~. ~f.!. .'.
- - - -..
------.-.------~'
3
.
(READ I
\WEEKLY Pk'S.
I
\" WKS /-/0
I
,
/0' /
>---~.....\
\..
-
!
- --- '"
._. .... ..
PRINi '\
107»
!--~
- - - - -
lor /J
--.."
"...
.(, PUNCJ.I
10'/-
I
I
...
I
15
CALC.
iN
WStlA!, TS(JM
WT$vNl
c!
\
'
I····
I
I
II
nSUM
I
WPEAH
L__________________
..J
I
IXN=- N
)
IY-2000
'~.----.....t rY:;
F16.5
240
ilefv r/rJ'jJ;n~
/liN =/
T
r
»n~ ..«"Uj.~
fu ~JW~ #:fJJ,)
(I): f1u. Jip)if'
w (r);
"
~ IY =.IY- 2.000
mtW' h~
r
j,~I{ ~ ld"aJ,
:;b
1/
/ - a-o
/
-
{I= /iIJJ 20)~ ; 2/- ,,"0 d>J3
20
)
IXI'/= XN :: 1Zo·1.A.~'Jnj~,
W S ut-r1::
T6(jtrJ '"
WTSU"1::
f
k~ mw J.u~r!.J
~ "JJmy. ~
~ WI1 Tvt
11:1
20
<
TZSUNJ :
~TI1-rY1
n:/
WF£~K
-ff: Y=
= ~ ~ haw
A +BT
J-1--- c~}~ ~~,
(=:;t~~,
TN~RM -:.:i:fm}J~M1I.n~.~~
YN ¢ R11/ :: A1U-y' ~ at *IN ?RN)
flArr¢=~~~;'YN?~0
y/oo =;;rwr~ rJj; T:: 100
Y fNS = WSf)tv! /XN
Flfi. ~
J/., ·?./I
rl/"",
'\o\".;.lIi.i
o
~
~
*3~1
________________
VARIATION OF PEAK LOAD WI TH WEATHER
fRons 'YR.
TI~LOAD
REGRESSION
PAGE
TEMPERATURE
NO.
SEA.
__ _l_~~ _
(NOR"'. ,NO.
~
B
___ ~72039E +03 .29598 7E-00
AVG. NORM.
AVG.
NORM
TI
Tl
100i A TIo WKS,
TI
RATIOU@e)
188.2
183 0 8
201.6
1.082
5
.249677E~00
45.8
40.
111.6
110.1
125.1
1.116
8
1389
632 .341257E+03 .136376E-;-01
45.2
40.
402.9
395.8
477.6
1.162 10
2628
632 .191679£-:03 .272586E-00
45.2
40.
204.0
202.5
218.9
1.076 10
632
.383324£-00
45.2
40.
255.3
253.3
276.3
1.065 10
759 _ 632 .287007E+03 .610460E-00
45.2
40.
314.6
31J.4
348.0
1.098 10.
45.2
40.
24.5
24.2
26.8
1.071 10
3874
~
A
INDEX
40.
676
"-
EQUATION CONSTANTS
54.6
296
~-
PK.TO
LOAD AT
~su~
__
"~ __ ANQ ____
1
.§~.?
632 .100171E+03
.237973E~03
632 .225404E+02 .433526E-Ol
!=' JFi. 7
N
_
.......
I'~
1
:'I,',::1
I
!
----~------.,.-~-----,------
____y. . 8. . . . _ _ _/1_ _~.ILJ<.J<_J<_
- - - CE
¢A
--~X..£...f-A..
+ + _ _ - - -..- .....- - - - - - - - - - - -
_ _ _ _
l",¥""J.&t1 in'ltil~ f:ua~~ ·
--.-CL..--
aiM,at anva.1fKaFt;-
-.--.u-
.,.,
,"_._~L_~ rbmf', W{Z1L
-_.XA--..--.--------------~-.----- . -.--.---------.--_ _
---Yli------'-'.---~-" . -.----~-.EL
'---CI1E ,c,rJf1 Mu!kJIDl~ ~ fe!.
-----rdle----
---JLt.4---'li1n1k-;;r~ --J)N 12 ~-+=:
ft.
_----A..-lN~·
H- ____
!
u
-
- - - - - - - . - . - - -
-
--_.- - - - -
JiM.
N/2
-•
.!:I..I:I----.------..
-------------
---.JiJL tfn. 'callfi.tjl1JJfJidL ------- ----.{;!4.,K:;&S.-~~1fwdi r.Qj~
--------- ~J_. ___ ~w1dl1l114_-- . -------.. . ---..-------f;<4,K~s -2lk.~5k("~ g:1 ----~
------.Il(1'f--~~~ft:~---¢i4ryf{::L~ $_ ftfuwtL.·~-. ~(Z:~ ~~
--s15)
II
_---.!!~.!.!.-v_~.-
______ .___.__.______________ .____ . ____ ._._. ___ . . __
TJX(;) 11k dJ;~4pto) 1&Af'.
-------_. - - - - - -
II
VK)
K
=, Ii
10 CfYWa.tu
II
II
"
"
1/
1/)
/I
ff,.""-'
~?
Fli, (
./3./3
I
o
~
" ,.
~
". ,1
EDE; ~:il®l'.lrJn« Statlat1ca
PEAK LOAD FORECAST DATA
01-1791 )-2B-20
GLJ""POWDE"IZ. (j)
~~
YEAR
•
5
10
I~ / .TaE.
F
15
.MAlf?lS tl-l. • •
IT' ,i.8.'j
~9.1.1
1M
t
20
•
•
•
Ilv
•. $.'/.~ ,
T
~
•
2S
•
•
,
,
"
tM1
5
..ti.'.I.S
J
1'1./,01 ,
,
-'---.1
.I.
L, ,
L
L'
IT
.4".1
. ,Id
.
,tI. 1.01
.~
•••
,
•
I
•
I
1.
----_I
F
T
'-'-
--t I...........................
~
....L......L
,
V
.$
.~:I
I
,
1
L LI
•
I
IRe WAR
I
.1
I
:
ICE
,
L
~.
"
-,
2S
.. ,. L~.8
IG1
L' ,
L
I
I
IReWAR
,.,
IG5
20
I
I
I
,
J
, .t:..3: I
I
I.
15
1
J
,
I
Mt
J
20
JY • ,(;,..3:/
•
I
IOM1
1.1..
I
:
UM4
:
IR4
.:
U,
I
:,
'I
:,
'
DRS
•
50
,
I....J
I
J
I
I
.,
I
fR3
1
1
I .1. ,
:
IA-MVAR
I
1_.
,
,I.,
~R-MVAR
I
I
:
"
i
ICE
:~'LL..L
I
I
I
i-r TI
IOM1
I
2
I
:
I
•.
L
1
,
RA-MVA
,,
1
:
IA-MW
,I
I OMS
M3
IR3
:
,.
I/!/,
IDRt
J..
I
t
Il_L.-l
I
.,
30
"
;
M4
..l.
1
•
i
I-'-----'---.L_'
L
,:2:/.0
I.
L-,--.
40
,
ICE
i.L ••
"'
•
"
IA-WA
I
45
50
,
•
l
80
~
,.
•
I
•
,
••
•
:
,
~
,
J
I....J
•
L--'--
I
•
IAeCAP
2, 11 • •
, U
I
.. , , ,.
,
'I
~
•
,
• •
70
..1---.1..
~.D
_..L.
,
,
I
L2
I
I
.3
,
L...l.
11M
,
•• •
·
10
__....L...~---'_ _~
L~
.10. IH.
•
IN.'2JI1 1
,
,
I
I
II
,2
•••
I
I
•
U
.3
,
L...L.
70
65
"""
IA-CAP
:2~Lj
•
11M
...L.-'----'---1..L---'----l.. .
..
:
IR4
!A-MW
1\ M5
:
10
,.
IH
10
UN
..311 ,1
~1,----,---.1
:
"., Il.,O{, ('f'/~,a ,X.l..3
.l..
60
•
IA-MVAR
,
,
i
65
•
,
lA-CAP
~/,00 .•. .2.1,0 . . . ~71L_l .• !4.2 ..
2
r--........._1.
rf. o,tI~.".1"SilJ)
IRS
L.
,III
1
55
n
A--"'T"'"",
;
I'.L..JL 1_-'_
..D.........t.4
75
L!6.0 I' . . 1\
,1\."
.l..4
75
,
R5
:,
,
..
:
:
I
i
,
I
. . 1 ' 1---.L !_...L
IIMS
""
I
35
.3
['~O "'J/~O ~
I
65
I
,
~AeWAR
• :.
'I:
,
12
R5
60
LL,. I .. I.. ,II, i . I. L LLi. I .
Li
1
---r IDR5
I
...L.....LLLl....t
..
I
!6.DU •• , ';:2,2 ,
1
1"0:0,0 , , .. ;.
DM2
L
55
L_L...1.
I
,1.1".-...-.........11--L..-...L
so
4S
J
10 IH liN
:'.DIIOM ~ ,.,
. .'2II_L 1
... I
In,
]D~51
~4
"
I
JII..J.. 1
70
,
IA~~P
60
, , :/.11 , ,
M4
U
IL~
:
•
11M5
•
•
•
I
•
I
• .,
IDR1
1
,
IH liN
10
:.3.51.L..L~,5~A1"C{"t~f't+JJi4
I
1. ,
55
,
"I"l.~
I."
I
L .. ~'~
'JAe WAR
,
UAeMVA " I . IA-MW
I I , , ~f.J
1
,
:-
I
LII., .1.,',II.,i,
1. _L_'
40
.L....l_L...1
./,fI!O,O , ,
i
IR1
:
• .• ,2.1
M'
'I
TT~M5---""
I
I
35
··:srR~MW--n ~ ~;r~~AR--:
G'
I
.,
•••
65
!Q.4l_L..L ,- _L_!ll.lo
lOR 1
Ll!lO
60
,
I
•
II.'
i
I
--L
IR5
••
I
45
'-
I
JA-MW
R3
I
1
I : 1\
1:3.0 ,,:3,00
,
tt1.011
•
.1.11.
25
S.U,B.S,T.AT,I.oh.
I.~Jol L_J-3~q1
,
;-!J.51 .. -:2.s1
L....J
1
:
I
11----'--:
55
,
I
:
80
7S
..L....1._~.
IAeCAP
11M!
,
•
..L.....L....1.
ICE:
"I'
30
G1T-rIG5
I
I
.~
IR4
,
i'
,
I
I
,
••
1/14.0
•
I
...L_...L
_1
I
I
Ii
I
•••
'5 ReMW
i~1
,
I
I
~2.SI
I
:
:.
.L_.L
I:
,
,
'IDR5
I
I
•
• , ,
I
,
1 1
1 1 ~._J L
I :9t41 , •
OMS
IM3-
I'
50
,
I I
•
IIM4
45
A-MVA
Li
I
,/~
I
:2.51 .. -.
I
40
L
·
1
'JM3 L,·
.
i~J , I L' • I. 11. I,
25
,
II
I
I
:, ,I
•
R3
,
..:,
L....L...J.... 1
It,
!&lO
35
...l..--'
,
.-]Ae MVAR
I
-'---.1..---'---1.
1._.I...-L
• ,t,,3ll
--
I
I
T
.tl ~ .3.9.
:,
30
,:
V
J
1
,
40
I.
IReMW
,
to
~-
•
IOR1
1
1M3
35
1 1
I
,
,E.L1.N.
OMS
, ,(,.3: I I . J.O:OJ) , , /.0:0,f) . . • , :,
M2
•
IR2
. . I.-. . L. .-&..--L---'ILM_'.L,. . . .--L_J:~ ~LL._L_
I'----'----l.
L':'-'
I,
L.:· 1
30
.L!5.~.
...L.....L.....1
ILl'!
I
:-rw --:
70
6S
60
55
"III'I.'.".'I'.'I.j
A-MVA
'...L~O
:
50
45
,
'.:1
tR2
I
20
1..1.
.1,t../.71 . .}u.ol.
I. /
5
'W.I.Ji./.E.R.S.
,
:S[ReMW
15
)..",./.7IT 1 .1.1,01M I
~ .4.-i.l.71--~- I~ 1,01
e
.
.
10
'I
I
25
I
,
•. 1_...1.....l.
5
F
I
~L.L...1.----'--.L.....l.
~
:0.0
j
8M2 I
L-,---K 1 Ll
....I.-....L...-&..~---'~.L-~~~.J
,'1./.0
I
...L.....l.
1.-'---1
,
IOM1
:
....l
1
.'
,
~...3!/L.L' , l'lJL..L' 1 ~51 I..L'
'I TIG5
IOM1
,~~!I 1, ,J,Dio.o , , ,/..0:0.0 , •
-:-IR11M2 r- IR2-
15
M
ICE
,
I
I
10
I
.4.,,/,/0,
1
I
5
,
F
•
•
I ' f'
I :rT
'1 •• 1:
, . , ! I,. ,I., . I.. . I, .
ILLI.
I
I
,
• •
I
-
:
40
•
,1/,J .. lJIJ . , k,J , , 12o,
20
IV
I
, IG5
·lR1
I
.L L-,--,
1M
3S
I
I/W .. , ./:J,o ..
•
':
.',3:1....L....L!
I
jT-"
I
I
510----
.J."/,5 , ,II.
/,0 _Li
-,
I·
.,
i .JJReMVAR
[G 1
~,~,7./I ,LU311 L3.4,A
''',~/l,
--------r
Mil .N.r.E.p.5~LE~~£.t~ _
F
30
•
~~1_L..L L~D
---.~
I
,
,
is IReMW
1",3:1 • 1
,
,3.tf.
,
I
.
--r--)./1,7.1 . •1,',3 13.'1.
.4If.7./1 . ,1.8,31
•
SEASON
...L
i--'._L .L.-l.
70
.: . 11M
75
...
,
.4
80
I
~_L ,~~,~~.JN.3O 1
--_·-_-·-_T_~_II
-L
I
I
l--__
._.1
--_._--- -
INITIALIZE: '
MM.LTP,
XEI. XE5acj
I'NIT/~LIZE: :
l
-1
II
I
READ
STATION
NAME
, CAR
I,
7
_"
,CARDN,.4)
r; L- ,
l
j
1I
.,'K)
-----r-
3
I
)
,
111
•.,
(LT)
'
'
fIJ,X(I)
0
,"
i ,"
r
0
,±!
'-....0
"
70-1
~K-Z
'=.J
"
,0
I )+
•
'
-
'
I
i'
0,
!!."
0
);0 " iI
I tfUNCH - , " - / .
--- OJ
"'[
"I
'7/ ./
_J,
/(MM)
/.'.
Llf"-L..I
I
0
"
'
MM=I
I
I
-
'
S3jLM.P,-;LM,,
!
_,
~::
- ,--
5%'
",-v
t
---
. ,-->-=-
"J/:f"
5
Mv:""SR
,I
l
r
"
I;.____
.L_
LF=O
PUNeH5
MVAHfI
t
'00CIf~
STA. NAME'
-.--~
CA =0.0
~~~';;;"-;;;J;~~ "Ul{J.AJy.~r';) 3/13/'3
I
..
I'
,
-,0
_I;
......
..---"
+
----.---L.---.. ----__.
L __ _
"-
'!!!. CRROR M£S5.\
~"""
,PVNCH
\
$'-0.00
' ',: ERROR MESS.
L
PEAK LOAD FORECAST/Nfi
ENS.S'AT.
11-~2
~
/PUNCH
@11:f(
+
-~i--
I
i
LL=I
0'
'"
'
I
,
"'.,
",,";R.
"
Z\
.., 7'
"No. Z '.----...<'
'-'./_
_
.
-
._~rLTP=LT· ,,,,'7
"'t.-.~" ~ l----*---=x
1: 4~~-L~"
~,
~>f,!L..,...J
7
(},READ 3 "
,/1...
_
'
1"
._9'"
/
;;X(K) I
L---,..-J
,e:,
I
,READ
,
t-
~
, REAl> 3
\CARD'/O.,)
I 7)
-'~LT~
'
'KK=O-I
30, \l'~
~-3)"'~'- t-~ /~'~ 6~C;' ~
o '-.. lO( /
-.------
_"
KK
Yo
,,. _ .,_,"" N-'C? ~/"" ~O
""
I
xM. vlQ»
,W-~..,
""I"
r " 'E:W.'H
,..,. ME"" 'l'o
J'
1'1O®7- r
(LF)
"- 6 0
" ,
/' 0
!t-M)
102.
:C_
'
-_.'.
J./t~o
.~ -' f.)
'
ICf1
~-l
_
Fie. 10
. ,..
"-
JC~
N
f13
I: .., ..
.;;:...
~,
:.rr
\..
~
\)
1-'\
\,
I
,,-,_/
f .\
\, j
ACTUAL AND FORECAST ONE-HOUR PEAK LOADS
ADJUSTED FOR PLANNED FEEDER AND CAPACITOR CHANGES
60 CYCLE FEEDERS AND SUBSTATIONS
SUMMER, 1963 TO 1968
MVAHR, MWHR & MVARHR
S TATI ON,
FEEDER
NUMBER
CAPS
MVARHR
SION HILL
4939
.48
.48
ACTUAL
1963
TIME
LOAD
7/18
22
VAN BIBBER FEEDERS
7053
1.50
2.40
7054
("',
1964
1965
FORECAST
1966
1968
1967
1.26
1.22
.31
1.06
1.06
-.03
1.17
1.17
.03
1.29
1.29
.12
1.43
1.42
.20
1.59
1.56
.30
3.17
3.17
-.10
4.09
4.05
-.55
4.53
4.52
-.35
5.01
5.01
-.12
5.54
5.54
.12
6.17
6.16
.39
2.91
2.77
-.89
3.15
3.05
-.77
3.42
3.35
-.64
1.20
2.10
VAN BIBBER SUBS TATI ON
7/18
2.70
22
4.50
1.94
1.90
-.40
2.54 .
2.71
2.29
2.52
-1.10
-1.00
4.74
4 .. 70
-.60
6.19
5.93
-1.77
6.75
6.59
-1.48
7.38
7.29
-1.15
8.09
8.05
-.80
8.92
8.91
-.41
WEST ABERDEEN
4738
1.53
1.53
7/18
22
2.22
2.22
-.15
2.45
2.45
-.00
2.72
2.71
.16
3.01
3.00
.33
3.35
3.31
.53
3.73
3.66
.74
7/19
15
40.00
37.20
14.70
42.96
41.66
10.50
46.50
45.10
11.35
50.42
48.42
14.03
54.71
52.02
16.93
59.44
55.94
20.07
"
HARFORD MASTER
19.64
0'·"
"
ARSENAL 4913 - 600 KVAR CAPACITORS ADDED-IN 1964.
BELCAMP INCLUDES A LOAD INCREASE IN 1964.
CHESAPEAKE CARDENS 4918 - 600 KVAR CAPACITORS ADDED IN 1964.
CHURCHVILLE 7051 - 300+J140 FROM SION HILL 4939 IN 1964.
CHURCHVILLE 7052 IS A NEW ITEM AND WAS PLACED IN SERVICE AFTER THE SUMMER
PEAK OCCURRED.ITS LOAD CONSISTS OF ALL CHURCHVILLE 4KV AND PORTIONS OF
FOREST HILL 4912 AND WINTERS RUN 4617.FORECAST INCLUDES NEW LOAD FOR
HARFORD JUNIOR COLLEGE IN 1964, 1800 KVAR CAPACITORS ADDED IN 1964.
FOREST HILL WAS RELIEVED BY CHURCHVILLE 7052 AFTER THE SUMMER PEAK OCCURRED
300 KVAR CAPACITORS ADDED IN 1965.
HAMILTON COURT 4927 - 300 KVAR CAPACITORS ADDED IN 1964 AND 300 KVAR ADDED
IN 1965. ~
HAVRE DE GRACE 4619 - INCLUDES LOAD INCREASE FOR HARFORD MEMORIAL HOSPITAL
IN 1964, 300 KVAR CAPACITORS TRANSFERRED TO 4625 IN 1964 AND 750 KVAR
ADDED IN 1965.
.
HAVRE DE GRACE 4625 - EXCLUDES LOAD INCREASE FOR HARFORD MEMORIAL HOSPITAL
IN 1964, 300 KVAR CAPACITORS TRANSFERRED FROM 4619 IN 1.964 AND 750
KVAR ADDED IN 1965.
SION HILL 4939 - 300 KVAR CAPACITORS ADDED AND 300+J140 TRANSFERRED TO
() , ."\
F/i. II
~
;iAGAiIM;;Ai;i:;;;'n;"
~
,; ,.,:..' . Ii·
i :t
»." .,iMfflhi$.i$ 9f4AIP..,
#
. T,¥.-, 44 - A4 ..¥ t, .
_
W#
4. ;tD:;; W!&4i!l .....,iWIi,tp,; .. -
PROOESSING Q-D SURVEY DATA
ON A SMALL COIIPU1ER
A. D. Stasi and M. B. Lipetz
idwards and Kelcey, Inc.
b·t· "frlitrW·""
irk
MM
rlit'Hi+iMiH.i'i=b;h!Mt!6W ... ·&WB£rij·,···Y·W·
lfr)'"f)·
PROCESSING O-D SURVEY DATA ON
A SMALL COMPUTER
By
A. D. Stasi and M. B. Lipetz
The title of this paper is not intended tq lead anyone to believe
that Edwards and Kelcey has found a viay to process a complex
comprehensive transportation study on its bare minimum Model I,
paper tape, 20K IBM 1620.
Although there are programs available
to do this on the 1620, they all require 60K and are limited in the
number of traffic zones they can handle.
Rather
I
'W·ould like to
describe to you some ,of the things 'W"e have done in the way of preparing
the data for the larger computers such as the 7074 and 7094.
Most
of us do not have ready nor easy access to these larger computers
which are used with the Bureau of Public Road's battery of Traffic
Forecasting and Assignment Programs for comprehensive traffic studies.
Consequently, when we do get on them, we must use them most
efficiently.
Being certain that our data is as correct as it can possibly
be and in the proper form for processing is, therefore, of prime
importance.
Editing for correctness and re -formatting of the original
data is well within the capabi1ities of the small computer, i. e.
~-:\ 2~)·· S,l.!C
IBM 1620. We have used all configurations of 1620's by the way our
'e
I
J./. /J././
_ _ _ _ _ _ _ _. ._ _ _ _ _ _ _ _ _ _ _......IIIOWIo!IJIIH!IIIMlIIiti'Jliii,\hiiliWii",...Iiii'i\iiiiiIi1!iliWGiihtIll&iMIlt:r...L4!&i'li .,::uHiiHI!iG'O:;,\bt..~"lt3lI't.ffi'lifuFw!tE_.tIitIF'fr,WF··'·
u
~.~
..
-2'f~,
\~iiP"
own Model I - 20K with paper tape I/O; a Model I - 20K with a slow
card reader I/O and a souped-up Model I 60K with paper tape, card
reader and 6 magnetic tape drives.
This last configuration was
available at Rutgers University until about a month ago -- we
w'ere
sorry to see it go -- it cost us only $50/hr! !
To get down to specifics, w'e have written editing programs,
volume factoring programs, zone conversion programs and
re-formatting programs.
The remainder of this paper will be devoted to an explanation
of the purpose and some of the mechanics of these programs.
I. Ed it ing Pr 0 gr am
O-D surveys are generally based on a sampling of the
travelling public which is later expanded to represent the entire area
being studied.
It is, therefore, essential that the maximum number
of sur vey trip records be maintained in the file.
Towards this end,
it is absolutely necessary that all trip records be edited for validity.
The specific requirements for the editing program are dictated by the
particular trip record being edited (Roadside Interview, Home, Truck,
Taxi, Tourist, Non-Tourist, etc.).
Our program simply checks various items of data, both single
characters and fields, on each card read in.
Several methods are used
J..I./~.1.
.. r--
-3 -
to accomplish the checking.
between the item of data
The simplest is a direct comparison
and its parameters for each item.
However"
utilization of this method will sloV\!" the program down considerably.
A faster method involves the use of binary decision tables.
Here the
item of data itself is used to compute a location 'W·ithin a corresponding
table.
The value of the digit {zero or one} found at this location
indicates whether the data is acceptable or not.
Another check incorporated into the program is the logic or cross
data check.
This is done to insure that several interrelated items
do not contradict each other and wreak havoc in subsequent programs.
(~'.'
For example, a trip entering the survey area at its 'W·estern boundary
.j."
'W·ith an indicated travel direction of West, contains an obvious inconsistency.
A zone of origin that does not lie within the county of origin
also indicates an inconsistency.
The only option available to the
programmer, and hence to the program, is to designate both items
of data as errors and leave the decision as to which is right and which
is wrong to the traffic engineer.
Error indication posed a problem.
How could the
program clearly
indicate which items of data are wrong and yet not destroy the original
250
o
1-/. 11f...1
-------------
-- ....-.----.- ..-.---.-~-~--..
---.-.--.- ..."---.-- .. -- ..
-4-
value of the data?
The answer was to punch two error cards.
first card is simply a duplicate of the input card.
is also a
duplicate~
data fields.
The
The second card
however" it contains asterisks in all the incorrect
Printing the error cards on a 407 produces a.listing from
w'hich corrections may easily be made.
The two types of error cards
are tied together by a sequence number assigned by the program
each input card.
to
This number permits each output record to be
uniquely identifiable in all future processing.
It also provides us with
a relatively simple way to insert corrected data cards back into their
original position in the file.
To aid in the determination of the necessity for error correction"
the program categorizes and summarizes the types of errors.
Should
the decision be made to correct the errors, these summaries indicate
which types of errors may economically be corrected by the program
and which can only be done manually.
The program also keeps a count
of the total number of cards read in and the total in error.
Ideally, the program should completely separate the acceptable
data from unacceptable data by punching or writing two files.
This
eliminates the problem of separating the data at some future time by
makeshift means, facilitates the application of corrections by a program
and maintains the order of the original input file.
On computers
c
-5'!.'I~"
(
essentially equipped with only one output device~ the 1620 typewriter
being much too slow to be seriously considered for high-volume
applications,
the production of
t\~lO
output files becomes a problem.
An easy solution is an identifying punch in an available card column
and subsequent separation on a sorter.
The best solution is to
utilize a 1620 with both punched card and magnetic tape input/output.
The acceptable data file could then be written on the tape and the
error file punched on cards.
This has the added advantage of placing
the acceptable data on the proper medium for input to the larger
computer at relatively little extra cost.
A program for editi ng 70, 000 trip records has been satisfactorily
run on a 20K 1620 with card reader,
and one for editing 500,000 trip
records on the souped-up 1620 previously described.
As is evident from. this description of our Editing Program,
quite a bit w'as packed into it.
however, such as
There is need for more information,
what are the numbers of correct and error records
at each interview station.
This informatim is valuable and often
will aid in pinning down the location of large blocks of errors.
.1./. /tIf-..0
-6-
A station volume accumulation program was, therefore,
written to summarize, by interview station, the output of the
editing program.
So used, it produces an up-to-the-run tabulation
of the number of acceptable trips through each interview station.
By using the error cards as input, a tabulation of unacceptable
trips can also be produced.
These two figures for a particular
station provide an indication of how many, if any, unacceptable
trip records must be corrected to increase the sample to a
s atis factory level.
Part one of the
c
program reads and accumulates the number
of trips for each station.
The address of the proper accumulator
is computed directly from. the station number, thereby eliminating
the need to search a table for each record.
The sec ond part tes ts
each accumulator for zero and, if not zero, punches a card containing
the corresponding station number and the number of trips through
that station.
After all the non-zero volume stations have been
punched, the program types the total number of input records and
the number of input records which did not have a logical station number •
.) .)
r'
..."
3
~
I
-7-
After both the
editing and summary runs are complete, a
decision must be made to either discard or attempt to correct
the unacceptable trip records.
It has been our experience that
these records usually comprise a significant portion, between 5 and
10 per cent, of the total number of trip records.
The usual decision,
therefore, is to correct all the records that are correctable by
a program and discard the remainder.
in groups.
Many errors seem to occur
One intervievv"er at a particular station, for example,
consistently entered the wrong hour period on his interview forms.
All the trips recorded by this man at that station on that day contained
the wrong time.
This can be corrected by scanning the other records
for that station and determining during what part of that day the
station was being operated.
Errors such as this lend themselves
well to correction by a program.
These corrections normally
increase the acceptable portion of the interview sample to approximately 98 per cent.
The correction or FUDGE program,
known, is a conglomerate.
as it has come to be
It contains the appropriate routines from
the editing program that originally discovered the particular errors
1\
-8-
to be corrected.
program.
In a sense, the errors are re-discoverE::!d by the
This tiIne, however, instead of simply indicating the
errors the program applies corrections and pun.ches the· corrected
card.
A card is punched for each error card read whether corrections
are made or not.
After the unacceptable trip file has been essentially corrected,
it is run back through the editing program.
This is necessary to
provide a final check and to weed out the relatively small group of
records that contain items of data \XI"hich
are"
therefore,
still unacceptable.
the final unacceptable trip file.
v~rere
not corrected and
These records now become
The acceptable trip records are
combined and tabulated by the station volume accumulation progl;am.
A tabulation is also made of the unacceptable records.
of these records is now complete.
Proces sing
They are deleted from the survey
da ta at thi s point.
As previously mentioned, there is presently available from
the BPR a package of programs related to O-D surveys.
These
programs are all written fOl a definite purpose and have
d~£inite
requirements relative to their use.
This is not to say, however,
that we cannot fool the computer into thinking it is doing one thing
c
-9-
while we are actually making it do something else.
For example,
a program written to prepare tables of trips taking place between
various zones based on various trip purposes has been used to prepare
tables of trips between specific zones based on the intervie'w station
rather than the trip purpose.
This was accomplished by placing a
"dummy code" in a specific field and defining this field to the program.
as the purpose field.
Not really difficult and· not really fancy, but
a simple 1620 program did the job for us and a simple 1620 placed
the factor.
A not so simple 7094 produced the tables which would
have had to be produced manually othervv-ise.
Another problem frequently encountered in O-D surveys has to
do with the numbering system used to designate the survey zone.
These survey zones frequently require in excess of five digits to
uniquely identify a trip end.
These survey zones are later com.bined
and must be reduced to a 5-digit maximum code in order to be
acceptable to the BPR programs.
Again this is a job for a small
1620 -- we have done this for 70, 000 records on a 20K card 1620.
The 500, 000 record job vv-ill be done on a 7074 although we might
have tried it on the souped-up 1620 if changing zones were the only
changes required and it had not been moved out.
c'
256
I
-10-
These examples show how Vie have made use of our computer
for preparing data from an O-D survey.
There are many more uses
which we 'Iv-ill undoubtedly uncover in the future. We feel our 1620
has a definite place in conjunction with the larger computers and
intend to make maximum use of it.
C,,~,"
I'
'
\"
J
I
I
,
•
I
,
..
'
."
./ G
CI
THE PLOT ROUTINE AS AN
EXAMPl~
OF METHODS OF ADDING
SUBROUTINOS TO THE NeE LOAD AND GO FORTRAN
Hubbard Seward
""'~''''''_~d'''''''''''''''---'''' ____ ' _ _ _ _ '''.''''_' __ '_'_'_'''' __ ''''_'_' ___ "'_~_ _ _ _ _ ' _ _ _ _ _ ""'_' ___ "._.""
I.
The Plot Routine as an Example of Nethods of Adding
Subroutin~s
to the NeE Load and Go Fortran
by
Hubbard Seward
Newark College of Engineering
Presented at the May meeting of the IBM 1620 Users' Group
in Washington, D.C.
NCE Load and Go Fortran was written to provide a 20K Load and
Go system with greater arithmetic
system.
abi~ities
than the IBM Gotran
In order to conserve space it was necessary to omit or
simplify some features of Fortran I.
Whenever possible features
were provided to replace those which had to be eliminated.
These
substitutions make it necessary for the programmer to compile in
the Fortran source program some things which are handled
auton~tic-
ally in other aystems, however in many cases additional usefulness
was created.
The original description of the sYstem states that
subroutines cannot be added, however this is not exactly true.
Because of space limitations, no facilities for adding subroutines helve been :i.ncluderi, but any user may add such subroutines
as will increase the processors utility.
Of course he must do
so with the realization that, at least for those restricted qy a
20K machine, space is at a premium.
'_~~,
Recon~ilation
will be necessary to add subroutines, but an
unlimited nUIllber of functions and up to four new nonari t.hmetic
pseudo op codes can be added.
Functions can be added by merely
including their names and the associated addresses ill the function
table symbolic location FTBL.
The table format is an eight digit
constant representing the alphameric function name followed by a
five digit constant for the address of the execution routine.
address must have a flag over the low order digit.
The
Note that
functions can be made removalble by locating them after the I/O area
and properly relocating the record mark placed by the initialization routine and also changing the address used to initialize
several fields.
(Chi, output, wk, and associated constants should
remain in their present relationship but may be moved to any location.
between 10K and 20K.
on the listing.
See for example the routine at location 15824
For maximum space the EXP and LOg routines can be
eliminated but when this is done the instruction at 01380 should
be made to branch to an error message, since the
**
operation
requires these subroutines.
To add new features to the system, additional nonarithmetic
type pseudo instructions can be added.
Because of the way this
part of the interpreter is set up there is a limit of four more
which can be added.
ions is variable.
The length of the nonarithmetic pseudo instructThe op code must have two digits and the
instruction must be terminated b.1 a record mark.
c
All operands are
two, three, or four digits long and are of four types:
260
- - - - - - - - - - - - - · · - " " " ' ,•."."'""',',,',0."....... """"""""ll>A\M,w"/lth%",,,,
"",;
:,1
3.
(1)
Variables and constants use the four-digit s,ymbol table pseudo
address which must be decoded by SUBER; (2)
statement numbers are
represented by their three-digit symbol table address;
(3) USE
and
do end pseudo ops which may have to refer to statements which have
no statement number use the four low order digits of the location
of the pseudo ops for that statement;
(4)
IF (sense Slrltoh)
pseudo ops contain a two-digit indicator as the first operand.
Currently the maxinru.m size of pseudo instructions is limited to
26 digits by the size of the area WK.
The plot pseudo instruction is complied by recognizing the
word PLOT on the source statement.
The control is passed to the
I/O compiler routine which assembles the operands in groups of
five.
If more operands are present, a new pseudo op is produced.
/«""0'\
'~"~'
The execution routine increments the card number and then
enters a loop which puts a character in the output area for each
variable.
\ihen there are no variables left in the list, the card
is punched and control returns to the interpreter to execute the
next pseudo instruction.
If more than five variables are needed
on a plot or if axes are desired this routine can be easily modified
to provide these features.
More variables could be
accommodate~
by
changing the plot routine to examine the next pseudo op before
returning control, and returning control only if it were not
another plot.
Note that the card numbers in columns 79 and 80 also
provide a means of s orting out plot cards from simultaneously
punched data output.
261
c
h.
C'
I
NON -
ARIT~IC
FORI' RAN
Pseudo
Equivalent op code
PSEUDO INSTRUCTION TABLE
Description
Length
Operands
END
+00
Branch to compiler
5
X
*FXFLT
- 00
Change mode of FAC
3
X
*RSGN
+ 01
FAC
3
X
*DO END
- 01
7
C
STOP
+ 02
Increment Index; check
for Index Limit
Branch to. Intr~prete~
Initialization
Set Index Equal to Initial
Value
Halt
3
X
19
A,A,A,A
-r4"DO
STARr - 02
PAUSE
+
GO TO
O~~
=-
FAG
- 03
3
X
6
B
3-23
X or From
READ
('~'
ACCEPI'
1 to 5
TYPE A
PFa:NT
- 05
PUNCH
)
+ 06
Tests FAC For -,0,+
12
B,B,B
IF(s~nre - 06
Tests Indicator
11
D,B,B
USE
stores LOC in Table
6
B
Returns to LOG in Table
3
X
*IF(
RETURN
*
- 07
These routines are not exact equivalents of FORTRAN Statements.
Operand types
A
Pseudo Address - The three center digits of a symbol table
address the first and last being always land 9 respectively
pluB a fourth digit to indicate subscript.
c
J.I../& ..J
5.
B
sa.
as:! without subscript digit.
Refers to .ta tement number (, loc.a tion in symbol table.
c
4 low order digits of location of DO st.art pseudo
D
2
d1g1~s
OPt
for machine indicator number
/
/
X
No operands
Note
On I/O instructions flagged subscript digit indicates
t~edpoint
vari$ble.
263
d"
t . .to
t .r1 h "" ;;rtrtft n,;f'±tf_i11
"#
******
***
***
'".'2
DSA
PLOT ROUTINE FOR LOAD AND GO FORTRAN
COMPILATION ROUTINE AND CONSTANTS
PLACE AFTER CARD NO. 24
INPLOT
PLACE AFTER CARD
PLT.CHI+6
COPLOT
155
C
BE
PLACE AFTER CARD
4,PLOT
178
***
***
PLT
" - [[318" TtY" ..
OAC
OS
0006
PLACE AFTER CARD 526
\MK+l,80.9R
B
COREAO+12
DORG *-3
***
COPLOT TFM
***
BNE
IF PLOT ARRAY
ER 1 1 -120
STATEMENTS ARE DESIRED. REPLACE CARD 392 WITH
EXECUTION ROUTINE
IF THE PLOT ROUTINE IS DESIRED TO BE REMOVABLE, ( SW 1 OPTION
PLACE THIS AFTER CARD 1930,
IF NOT PLACE AFTER CARD 1810
NOTE THAT NUMBERS MUST BE FIXED POINT AND IN THE RANGE 0-77
0007
AM
LOC-l,4,10
TR
20214 31 14625 14629 12-0031
\AlK.IAIK+4
AM
P4-1.2.10
20226 11 20321 000-2 12-0032
MM
FAC,2.10
20238 13 00079 000-2 12-0033
BN
20250 47 20382 01300 12-0034
ER66
AM
20262 11 00099 ~4459 12-0035
99.0UT
CM
20274 14 00099 ~4613 12-0036
99.0UT+154
BP
ER66
20286 46 20382 01100 12-0037
20298 24 20333 0009R 12-0038
P4+11.-99
C
TF
20310 26 0009R K0415 12-0039
-99.CHAR.7
20322 46 20346 012-0 12-0040
P4
BE
*+24 •• 10
20334 16 0009R OOOKI 12-0041
TFM -99.21.10
20000 11 09139 000-3 12-0009
INPLOT AM
LOC-l, 3. 10
20012 16 20321 K0413 12-0010
TFM P4-1.CHAR-2
20024 27 08482 08735 12-0011
SVJEEP.OUTX
BT
20036 11 20145 000-1 12-0012
AM
CDNO.I.IO
20048 43 20096 20145 12-0013
SO
*+48.CDNO
20060 11 20144 000-9 12-0014
AM
CDNO-l.9.10
20072 43 20096 20143 12-0015
BD
*+24.CDNO-2
20084 16 20143 OP170 12-0016
TFM CDNO-2,7170.8
20096 26 14617 20145 12-0017
TF
OUT+158,CDNO
PI
8NR P3.\MK+l
0019
******
EXIT HERE IF FINISHED
20120 39 14459 00400 12-0020
WACO OUT
B
LOC
***
***
***
******
CDNO
P3
4.7070.*+2
*+2
SUBER.WK+4
EM-...,.-SUBI-23
*+20.FAC
ER71
B
DORG *-3
BNF ER74.FAC-3
DC
DORG
BT
TF
BNR
20142 PO 70
20145
20158
20170
20182
20190
20190
20145 04-0022
0_023
26 00099 1135J 12-0025
45 20190 00079 12-0026
49 09060 00000 12-0027
0028
44 20370 00076 12-0029
J-I.I5.l·
I
I
"
!)A
6-c
7
I,,,,
~
a
P1
ERRORS 74 FLOATING POINT VARIABLE,
ERROR,77400.5Rl1
ERROR. 67600.7R
OUT+154,4559.8
B
PI
DORG *-3
LIST OF CHARACTERS FOR PLOTTING
******
CHAR
DAC 1 • +
DAC 1. *
DAC 1.X
DAC 1.0
DAC 1 , •
******
aTM
ER74
BTM
ER66
TFM
EXIT
66 OUTSIDE RANGE
20.370 17 027-4 7740R
20382 17 02704 0760Z
20394 16 14613 OM559
20406 49 20108 00000
20414
20414
20416
20418
20420
20422
,)0
,)4
07
PO
-3
20415
20417
20419
20421
20423
IN ORDER TO MAKE THE PLOT REMOVABLE ADD THIS AFTER CARD 1947
PLT
AM
CM
BNE
AM
IF MORE THAN FIVE VARIABLES ARE TO BE PLOTTED. REPLACE
THE
BLOC IN THE ABOVE ROUTINE WITH THE FOLLOWING
AND MAKK SUITABLE EXPANSION TO THE LIST OF CHARACTERS.
LOC-l. 2. 10
l-LOC.4.10
EXIT
LOC-l.l.10
B
PI
SM
LOC-l.2.10
LOC
B
0044
12-0045
12-0046
12-0047
12-0048
()()49
0050
SF
***
***
***
***
20358 49 20108 00000 12-0043
02-0051
OZ-0052
002-0053
02-0054
02-0055
4'-",
,,~,
THE CARD NUMBERS IN THIS LISTING REFER TO THE CARDS IN THE LISTING OF THE
UNCOMPRESSED DECK AS DISTRIBUTED BV THE PROGRAM LIBRARV.
ALL MACHINE
LANGUAGE INSTRUCTIONS AND ADDRESSES ON THIS LISTING SHOULD BE IGNORED.
if.l&"
7
tirtNeteri:fiittbririi +·
Tm!!· ri8Wu!ihii*··
.') A 6-
c=
'ii
A=O.O
PI=3.1415926536/20.
DO 8 1=1,41
DIMENSION ISIN(41)
ISIN(I>=SIN(A)*35.+3S.S
8 A=A+PI
PLOT. ISIN(I ),,1=1.41
(~\
,-
END
01
02
03
04
05
06
07
08
09
10
+
+
+
+
+
+
+
+
+
+
+
+
11
12
13
14
15
16
17
18
19
+
+
+
+
+
+
+
(,."
20
+
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
40
-.
41
42
26ti
C
"I
""'1
1l1l'CMATIC SCHEDULING AND REGISTRATION
IN A SYALL COLLEGE
Arthur F. Jackson
Director ot Guidance and Computer Center
Greensboro, North Carolina
~.
If"'.
I
'.
AUTOMATIC SCHEDULING AND REGISTRATION IN A SMALL COLLEGE
by
Arthur F. Jackson
Director of Guidance and Computer Center
The Agricultural and Technical College of North Carolina is a
small land grant college located in Greensboro, North Carolinaan enrollment
of approximately 3,000 students.
graduate schools:
Agriculture, Education and
Engineering, Nursing, and Technical Institute.
with
It has five underGeneral Studies,
Each student registers
in one of sixty-five different major fields.
The college has been using unit record equipment in the registration process for several years beginning with a Remington Rand
(:"
installation of Punch and Sorter and developing into a complete
Remington installation.
In 1962, the college changed to an IBM Series
50 including the Sorter, Interpreter, 402 Accounting Machine.
With
this equipment, registration proceeded as follows.
The Deans of the various schools requested that a specific
number of class cards be prepared for a certain quarter.
This informa-
tion including class code number, class description, meeting place,
time, and credit hours, was punched into a Master Course Card.
These
cards were then duplicated on the reproducer in the numbers indicated
by the Deans and then sent to the Registrar.
At pre-registration
time, the students completed a course schedule on a four-part form
at a pre-registration station.
o
These forms were completed and sent
26>
J,I./G. I
-_.
__ _.•...,._--,
....
-2-
to the Registrar's office where clerical help manually matched each
course on the schedule form with a pre-punched class card.
Each
student's class cards were assembled in an envelope and returned to
the Data Processing Center where student's name and number were
gang punched into his class cards.
From these cards, class rosters
and individual student schedule forms were printed on the 402
Accounting Machine.
In cases where cards or courses had all been
distributed before a student had received all his class cards, his
schedule was sent back to his advisor for revision.
In June, 1963, the college accepted delivery of an IBM 1620
Computer, a card system with 20 K Memory positions and without
automatic divide.
It was used primarily for teaching purposes ex/,~
cept for occasional use in the computation of statistics and compilation of certain registration data.
'",/
During the 1964 Spring Quarter,
the 1620 played a prominent part in the registration.
This was
possible because of the development of three programs:
STUDENT, a
program devised by IBM, and two post processor programs designed by
the author.
STUDENT, an abbreviation of "Scheduling Technique Using Defined
Increments of Time", was designed for use in scheduling students into
a section where the number of sections does not exceed 1000.
It
1S
designed to store a Master Schedule (not to exceed 1000 sections),
accept student requests (not to exceed 8 per student), and to place
students in sections.
Students are allowed to designate
sections desired and also to request free time.
s~ecific
Students who request
26D
c'
-3-
(~:"\
specific sections are scheduled until requested sections fill;
students who do not specify sections are scheduled as long as any
section is open.
When a student is not scheduled, a report 1S
written on the typewriter indicating the reason for failure to
schedule.
This program is controlled by course code number, section
number, and load limit.
This means that each student request is
checked to deterJrine whether or not the number of spaces provided
for that section has been reached.
On the basis of these two checks,
the student is scheduled or not scheduled.
The card format for the
student request allows for twenty-eight columns of information
about the student while the rest of the request card is allotted to
(
H,
"'..
the eight courses.
In order to test the feasibility of STUDENT for use in our
system, the decision was made to use the class cards and Master cards
from the 1963 Fall Quarter registration in a trial run.
A computer
program was, therefore, written to condense the several individual
class cards for each student into the single request card format
required by the program STUDENT.
in the STUDENT format.
The Master Cards were also transformed
The process was then reversed, the l1asters
were loaded into memory, and the students' requests fed into the program.
Every possible conflict situation was simulated in the test.
The test resulted in the "go ahead" signal for the Spring Quarter.
Once this decision was made, there remained two problems.
The
Dunched output from STUDENT could not be used effectively as the class
.) 1'1
ow "
fI
I
I
i
I
-4'~7
card which is used as basis for class rosters, admission to class,
and finally as a grade report.
The second problem was that of
determining the spaces remaining in sections which were still open
as soon as the scheduling had been completed.
This information
would eventually be obtained from temporary class rosters, but
administrative officials felt the need to have this information
earlier than possible by this method.
This information was vital to
the opening of new sections or the deletions of courses in which too
few persons were registered.
Two programs, both storage and search
programs, were written to achieve these ends.
The program to punch class cards was designed to use as input
the condensed output from STUDENT.
This 'was a single card with name,
,,,,r~,
I.D. number, and as many as eight course code numbers and sections.
'"' ':.#
The program first stored up to 225 Master Course Cards in memory.
It was limited to 225 because of the alphabetic content of our Master
cards.
The program then attempted to match each course on the student
request card with one of the masters in storage.
When there was a
match, a duplicate of the Master Card was moved to output, the
student's name and number were added from the request card, and the
card was punched.
The search was sequential, from the first master stored to the
last.
There was, therefore, no need to sort the master course cards
before storing in memory.
If no match occurred during a search, the
computer simply moved to the next course on the cdrd or the next cdrd
to begin the search over again.
.
.) r"1
.....
1
J-f. I~.
1;'
-5The inability to store all of the Master Course Cards
necessitated four passes with the request cards.
The Masters, how-
ever, were loaded by school groups which made possible the printing
of class rosters for one school while the computer was punching
course cards for another school.
This process was not timed with
stop watches, but it was possible to prepare all of the class cards
within one working day.
In addition, the preparation was much more
complete than ever before.
The card count program was similar in nature to the program
for punching class cards.
The course code number and section of
each Master Card was stored in memory together with a two position
counter.
Each unit was thirteen storage positions as follows:
a
four digit course code number, a blank; a two digit section, a blank;
a two digit number representing the number of spaces allotted to that
section, a blank; and finally, a two digit counter.
Storage was
searched sequentially, and when a match occurred, the number "one"
was added to the count position of the section matched.
At the end
of the run, the typewriter printed out the contents of memory where
the course codes had been stored.
This was a report of the number
of spaces allotted to each section and the number used at this point.
Pre-registration was conducted as previously with the exception that each person listed, in addition to the courses, the course
code number and section for each course on a special card.
These
cards were sent to the Data Processing Center where, as soon as
received, they were punched.
No sorting was required, but the cards
o
J../.. I~. . .,
-
..---.----..-.- ...
---"~,,
--...
..................-- .. ''' ....
~.--~--
."
.............. ".. ".._.......... "" .... ""....-......
".~---.--------.---~
-6-
,
."'-'"
were sorted by school for ease of handling of individuals who could
not be scheduled.
These were then run into the 1620 with the program
switches set for no punched output.
checking such things as missing
the like.
This afforded an opportunity for
master cards, wrong sections, and
Errors in course code numbers were expected since this was
the first time the student body had been required to use them.
All
errors thus discovered were corrected and the request cards were then
run again to discover schedule conflicts.
Most of these resulted
from the choice of favorite hours for classes.
Where possible, these
were resolved by reproducing the request cards, omitting the section,
or by punChing in sections which could not create conflicts.
No
punched output had been obtained on either run.
With conflicts reduced to a minimum, the request cards were
once more fed into the 1620, but this time the program switches were
set to obtain punched output in condensed form.
Those individuals
(their request cards) who were not scheduled on this run were sent
to the department in which they were enrolled for a resolution of
this conflict.
Now, the program for punChing class cards was used.
Class
cards were made for each class section for all of those who had been
scheduled successfully and a count made of the number of seats taken
in every section.
Class rosters were then printed for every class
and section on stock paper.
Schedules were printed on the 402
Accounting Machine for the individual students who had been scheduled.
In the interim between the final run on the computer and
J-f. It;,. (,.
.....
\'
-7-
C\
registration, the various administrative officials used the data
provided by the program described to open new sections, expand limits
in sections, et cetera.
This made possible the scheduling of most of
those individuals who had not been scheduled prior to this time.
At registration time, the student picked up his schedule and
other things such as meal books, et cetera.
The time required for
this was about fifteen minutes as opposed to several hours under the
old system.
Lines moved very rapidly so no long lines were formed.
Problems in registration were reduced to a minimum but the
problems in changing courses was not solved.
In our system, a student
must re-take a course he has failed in the next quarter it is offered.
This created most of the requests for changes.
The information gained in this effort suggests pre-registration
in May for the next Fall Quarter, registration of the entering
freshmen with test scores taken into account, and eventually preregistration for an entire year together with automatic scheduling.
o
,
--"-"--"""--'''''''-'''-''''
""",,""----
I
Iii
I~
I~
!FIT
SPS WITH SIJrIDLATED TNF, TNS AND MF
INSTRUCTIONS
David Olson, Camput1n& Center
Newark Colleee or Enaineerine
#.17.
0
AFIT SPS with Simulated TNF, TNS and lvlF Instructions
David Olson
Computing Center
Newark College of Engineering
Introduction
AFIT
SPS by Mr. Pratt has just recently become widely
available through the Program Library.
The system has many advantages and improvements over SPS II.
However the system requires a machine with indirect addressing,
auto-divide, 40K and the TNF, TNS and MF instructions.
Description of Simulation Method
We have removed the requirement of the TNF, TNS and MF
ins tructions.
TNS
The TNS instruction is used only twice in the processor,
therefore TD instructions were used to simulate this instruction.
TNF
The TNF instruction is used several times therefore it was
felt that a subroutine would be the most efficient method of
simulating it.
This was done as follows:
Replace the 73 with a 41, then follow the TNF with BTM
TNF, *-1.
The TNF subroutine looks like this:
o
~/7./
2.
DS
TNF
SF
TNF-l
TF
A999,CONl
TF
304,1-TNF
A
A999,CON2
TF
310,RSTOR
AM
TNF-l,5
TF
1-TNF,A999
BB
CONl
DC
10,6070707070
CON2
DC
10,00(1)1020304
A999
DS
10
RSTOR
DC
12,801234.567891
This subroutine is essentially the one described by Mr.
Pratt in the May 1963 issue of the 1620 User Group Newsletter.
MF
The MF instruction is also used quite frequently, therefore
it was also simulated b,y a subroutine.
This was accomplished as follows:
Repla ce the 71 with a 41 then follow the MF with a BTM MF, 1-*
The MF subroutine is as shown below:
QADD
DS
5
MF
TF
PADD, QADD
AM
PADD,5,10
CF
-PADD
DS
.5,*
BNF
*+36,-QADD
CF
-QADD
SF
-PADD
PADD
<) ,.., "'"
.... • I
BB
.t./.17 <-~
C
3.
or
course there was some additional rewriting of the processor
to allow use or the BTM instruction, since the routines in the
processor which used TNF or MF were quite often entered by another
BT or BTM instruction.
Mr. Pratt is to be congratulated on. his excellent contribution
to the growing list of Symbolic Programming Systems.
We have
round no significant errors in his system.
In our opinion, it is the best 40K SPS available.
(
,0
""",.
o
276
JI..I7-3
,0
\""",
NCE "HI OH SPEED ",SPS ASSEl4BLER
I
,r-',
1;1,,,,,.,1
Kurt Gerrr.ann -and O8orie Rumrill
Newark 0011eaeo-t Enaineerini
c
C:
NCE HIGH SPEED SPS ASSID1BIER
by
Kurt Germann and George Rumrill
Newark College of Engineering
Presented at the May meeting of the IBM 1620 Users' Group
in Washington, D.C.
This assembler is an attempt to provide Users' Group members
with an assembler that has the language capabilities of AFIT SPS,
and a symbol table size of the order of LAMP (File 1.OO.0a,l in a
20K system.
Although some compromises had to be made, the resulting
system, provides space for 450 labels, and has a language capability
simi1iar to IBr-l 1620/1710 SPS.
The need for an assembler with a large label capacity in 20K
developed when the Computer Center Staff of Newark College of
Engineering was writing NeE Load and Go Fortran (File #2.0.029).
During compilation of the Load and Go System, the IBM SPS Symbol
table overflowed, so another version of SPS was looked for in the
1620 LibrarJe It was found that OSAP (File #1.01.013) best suited
the need, but to be useful to us, several modifications and
corrections had to be made.
These initial modifications were the
work of Joseph Petersack, an undergraduate student at NCE. After
the 20K Load and Go processor was completed, the authors decided
to follow through on these initial modifications, and to rewrite
tl7e
completely OSAP processor.
"J-
L/.I1, I
2.
The NCE assembler is approximately two to three times as
(l('~
''L.)l'"
fast as IBH 1620/1710 SPS, exclusive of input-output.
THis means
that on a Model I 1620 with a 500/250 cpm reader-punch, the assembler
will run at reader speed on Pass 1, and at punch speed while punching
an uncompressed object deck on Pass 2.
The assembler requires card I/O and indirect addressing.
Automatic
divide is required only if division is performed in address adjustment in the source statements.
The assembler has been written so that the input format can
be of
~
version.
types.
It can be standard SPS, or a slightly modified
The modified format is easier to keypunch, and also easier
to read on a 80 - 80 printout.
Colunms
The format is as follows:
1 - 6
Label
7
Blank
8 - 11
12
Op
Code
Blank
13 - 50
Operands and Remarks
51 - 80
Arbitrary
This format does not allow page and line counts; if the.y are desired
the standard SPS format can be used by setting a console switch.
The output format of an uncondensed object deck is in most
cases a single card per statement, and can be printed on a standard
80 - 80 board.
The uncondensed format is:
281
3.
¢olumns
1 - 50
52 -
80
Original Source statement
Machine Language and Sequence Numbers
This format enables using the object deck as a source deck, making
the original source obsolete.
Besides being more convenient J this
format produces object decks which are about one half as large as
. those of SPS II.
In addition, a compressor is built in, Which
produces compressed decks smaller than those of SPS II.
Also
provided, at the operator's option, is a symbol table punch-out,
five s,ymbo1s to a card, in alphabetic order, at the end of pass 1.
The number of labels that can be stored by a 20K machine
varies.
If, upon initialization, the compressor, its loader, and
the uncompressed loader are eliminated, approximately
digi t J.. be1s can te stored.
450 six-
If the compressor and loaders are not
eliminated, about 350 labels can be stored.
If a larger model 1620
is used the assembler expands automatically, increasing the s.ymbol
table accordingly.
Depending on the size of the machine, the size of the program,
and the number of labels, it is possible to store the source
statements of a program in core during pass 1.
If the program fits in cofe, it can be run in a single pass,
if it does not fit, however, the source program must be reentered
for pass 2 as usual.
In a 20K machine, again depending on the
program, 60 - 100 source statements can be stored and processed.
A 60K machine on the other hand can store 1000 - 1300 source
statements.
Besides reducing card handling, this will also, of
course, save machine time.
-.) S-.)
H.18.
:~
In addition to the above, a different error message format, has
been incorporated.
<~,
I,~,~",
Each error message consists of a typeout of the
actual source statement in error, the type of error, followed by
the sequence number of the card in the uncompressed or compressed
object deck that contains the error.
column
51
To aid in t rouble shooting,
of the uncompressed object card in error contains the
letter E.
As far as the language is concerned, all SPS II op-codes for
the 1620 card system a re llalid, both mnemonic and numeric, with
the exception of HEAD, TRA, TeD, and SEND.
In addition, many of
the op-codes used in AFIT SPS have been incorporated, as well as
several interesting programming procedures devised by Richard L.
Pratt.
A variable and fixed length subroutine package similar to
that of AFIT SFS is also included.
,.-11-----,
\ ....,"iIV'
c
5.
Blanks in Source statements
Blanks are ignored in the operand field of all
sourc~
5tatements,
includinc- the nag operand of instructions.
In a DAC statement, the right.most character of the constant
doe:; not have to be followed immediately by a conmia.
If no address
Examples:
C,
A
DS
, 1001*10-9
B
DS
J
20002
TF
A
J
TR
A:
DAC
5,
DAC
2, A
i~
B,
4 ;;
,
i6 10001
20002
31 10001 20002
00 00000 000
@
,
,
tJ. 0*
spaces between the second and third conWAS will not
cause the Qonstan~ to be loaded ~t zero
The
,
DC
TLr;
B , 123
,
r~·.~~::."~'1;;'. !., >;}:L c~~n
0000*
not be greater than 35 because of the
restriction on the length of the source statement (columns 1 to 50).
DSAC StateJrent
The Define Speoial Alphameric Constant is the same as the DAC
except no flag is placed over the high order digit.
This allows long
alphameric constants to be constructed by one DAC followed by DSAC's
and this long constant can be moved by a transmit field instruction.
6.
DC aod.DSC statements
A)
.
The processor
~an
~
.
.
I"·
supply artY number of lending zeros, p.p to
the maxilTlUl'll constant length of 99999
Example:
5600,@
DC
X23
This will cause 5599
~eros
followed bya record mark to
be loaded into care, the label will refer to the location
of the record mark and the high order zero will l::x:l
flagged.
B)
The
@
sign which indicates a record rnarkmay appear anywhere in
a constant.
Example:
DSC
6,-@123@@
Will cause
C)
i
12~' to be stored.
Any character of the constant may be flagged· by punching a
minus sign over that character in the source statement.
Example:
DC
7"
1~I4*P8
..0044i
- - -18
t1il1 cause
D)
to be stored
will produce
,*.
1,-@
will
:;
DC
1,*
"
DSC
1,-@
n .
DC
I:;
DC
~'1
"
"
"
*
:#=::;
Constants may be preceeded by a pltis.sign if desired.
c
7.
DFC statements
(.,:
"\
Floating point constants, e.s rcr;u.ircd
l)~r
the subroutines, may
be specified as follows:
second
If
Third
"
mantissa length, Which may be symbolic,
(must be less than 50)
mAntissa (trailing zeros automatically
supplied. )
exponent (may be symbolic)
fourth
"
address (if desired)
first operand
The label, if present, will refer to the units position of the exponent.
Example:
DFC
10 ,-2,1
will produce
--
2000000000 01
Operands
The previous
IB}~
restriction on - 0 is removed and this will
now be assembled correctly.
Example:
TDfJi
15 10001 00000
A , - 0
Division is now allowed in address arithmetic (if the auto divide
feature is available). Division can be used in conjunction with
subtraction and multiplication to produce truncation.
E&le:
B/A
*
1000, B/A +213
11 02000 00215
Record Marks in the QR position of instructions are frequently
required.
This usually has required the two instructions:
TFM
A
" 10
However on this system an R in the flag operand will cause a
record mark to be placed in the Qll position of the instruction:
C
,
"'J
'
286
8.
Example:
TFH
A
TFN
A
TDH
A
,,
lOR
16
10001
0000*
R
II
"
16
10001
-0000*
llR
11
n
15
10001
oooo~
"
"
results in
.,
rf'~
\\.
DSA statements
The number of operands in a DSA or DSSA is no longer limit.ed
to ten, but may be any number of address operands up to column
50.
DSSA statements
This is the same as DSA, except that no flags are set over the
high order digit of the DSSA's, and the label refers to the high
order digit of the first DSSA.
TABLE
DSSA
A,B,
10001
20002
00000
DNB statements
If·~
\4..,,7
The length in the DNB and DSNB statements must be no greater
than 99999.
DSNB statements
Define Special Numeric Blank acts the same as DNB except that
the label, if present, is equated to the high order position of
the numeric blanks.
INC statement
mnput Converwion Subroutine allows input data in the free
format style of AFIT Fortran to be oonverted to internal floatine
point form:
2S7
c
9.
Ci
first operand
address where exponent of number is to be stored.
second operand
alphameric record address of data to be converted.
Data in the input area may be separated by blanks or commas and
although the input area must contain an alphameric record mark, no
flags are required.
Library No. 1.6.053.
OU'l'C statements
Output Conversion Subroutine allows internal floating point
numbers to be converted to external Fortran style constants under
forrrat control.
first operand
second operand
third operand
alphameric record address where external form is to
be stored
address of exponent of floating point number to be
converted.
five digits specifying the format:
In this five digit constant) the first digit represents the
type of format desired:
o
for E
1
for F
2
for
G
The next two digits are the length of the field, and the last two
the number of spaces after the decimal point.
E 14.8
01408
7.2
10702
G 16.10
21610
F
The E and F conversion work exactly the same as 1620 Fortran.
The G w.d conversion gives F type output if as many as d significant
figures can fitted in the space provided.
E
w.d will
bE;
used.
Otherwise, a format of
Library File # 1.6.053
1
DOODLE
Do-It-Yourself Problem SolTer
M. V. Farina
General EleQtric Compal11'
DOODLE
Do-It-Yourself Problem Solver
M. V. Fa ri Yl q
~. E. COvY\pd. ny
An engineer busily at work in the solution of a mathematical problem
reminds me of a trombone player. He can manipulate a slide rule with
all the skill of a musician.
Unfortunately, slide rules are slow and
limited in accuracy.
Suggest to him the possibility of programming his problem for
solution on a computer, however, and chances are hetll give you a hundred
reasons why he can tt. He doesn rt know anything about programming;
Fortran has too many rules to remember; his problem is too small for
a computer, and he can go on and on with excuses.
An accountant industriously engaged at making debits equal credits
reminds me of a pianist. He can finger the keyboard of a desk calculator
easily as well as Van Clyburn.
But the typical accountant, like the engineer, believes that he
himself cannot program his problem for a computer.
Perhaps it does take some skill to program an engineering or
business problem.
But have you noticed -- it takes no skill to doodle?
We all do it -- while on the telephone, during a meeting, while waiting
for the doc tor.
If programming could be made as simple as doodling, then maybe
engineers and accountants would program.
In an effort to simplify pro-
gramming we devised a system for those people to use and called it
DOODLE.
C
'I
,;.,/
1
29{l
J.!. /9. /
My discussion of DOODLE is in three parts: First, what DOODLE
is; second, how we are inducing engineers and accountants to use it;
and third, some ideas of what the second generation of DOODLE might
produce.
DOODLE is a system which enables the user to write out a problem
in a very simple way. There are few rules to know and a knowledge of
programming is not required. Wefll show you some examples in a few
moments.
DOODLE is an interpretive scheme. It works directly with the
userfs program as he has devised it.
The system interprets each
instruction as it encounters the instruction and executes it.
This means
that no assembly or compilation process need be performed before the
problem is given to the computer.
Computer time is thus conserved.
Very often it takes longer to compile or assemble a program than it does
to run it.
This is particularly expensive on one-shot jobs. DOODLE by-
passes this expense.
Mechanically, DOODLE is a Fortran II program written for an
IBM 1620 Computer with indirect addressing, auto-divide, floating-point
hardware and card 1-0. Our Fortran compiler in Johnson City has been
changed to origin at 11K instead of the standard 12K, as specified in the
Fortran II Operatorfs Guide, Page 25.
(aro J' W'\(A.c~i ~e)
2
29 1
(
~'. \
{,
•.•.
\
.
~iO"
c
I can best continue showing you what DOODLE is by telling you how
we are inducing our engineers and others to use DOODLE.
The words
Utricking them It seem to be applicable here since the write-up of DOODLE
does not mention the words "program" or "programming." The write-up
pre-supposes no programming knowledge and speaks only in terms of:
"you have a problem to solve -- a computer can solve it for you--herers
how you do it. "
The DOODLER publication exists as a 2-volume work. The first
booklet is the ttPrimertt and the second booklet is called ttFor The Advanced
Doodler." Pd like to show you a few of the pages from the Primer.
(SLIDE 1 - PAGE 1 OF DOODLE WRITE- UP)
The key word here is "EASY." On this first page we are hoping to
spark the readerrs interest.
The second page tells the reader that DOODLE is a method of telling
the computer what his problem is; that he writes out the problem on a
simple form; that cards are punched from the form and delivered to the
computer; and that he gets answers in return.
(SLIDE 2 - PAGE 3 OF DOODLE WRITE-UP)
This slide shows how the user tells the computer what numbers his
problem deals with. Note the illustration. We feel the illustrations, of
which there are many, are very effective.
A card will later be punched for every line shown on this coding
sheet. Notice that item numbers have already been filled in.
o
_" ..£.", ........ _......... 4.4IY4.?¥4i4#.Lf!L+.\4:F. .. _._ ...
3
2U2
(SLIDE 3 - PAGE 5 OF DOODLE WRITE- UP)
We have told the reader how he feeds numbers into the computer.
He will now be told how to tell the computer what to do with those numbers.
(SLIDE 4 - PAGE 6 OF DOODLE WRITE-UP)
This slide shows how the trops· are used.
Page 7 of the DOODLE write-up reads as follows:
tryou told the computer to divide item 101 by item 102. The result
became item 105.
Then you told the computer to multiply item 103 by
item 104. The result became item 106.
Finally you told the computer
to add items 105 and 106. The result, item 107, is the answer to your
problem.
If
Finally the reader is instructed to put the op IIEND" at the end of
his solution procedure, and he is shown how the computed coding form
wi 11 look at that time.
(SLIDE 5 - PAGE 8 OF DOODLE WRITE- UP)
At item 107, note that a 111 n has been placed under the column
headed "P. II This causes the computer to print out the result of the
execution of the instruction at item 107. ThiS, of course, is the answer
to the example problem.
(SLIDE 6 - PAGE 15 OF DOODLE· WRITE- UP)
Skipping ahead in the DOODLE write-up, page 15 shows alist of the
ops that are available to the DOODLER.
4
2 U3
c
C:
At the bottom of the slide, notice that the publication "For The
Advanced Doodler rt is mentioned for the first time. It is not necessary
for a DOODLER to ever read the advanced volume.
It Dcludes such
exotic features as looping, address modification and use of subroutines.
We recognize that these features are such that only professional programmers would be interested in them or DOODLER enthusiasts.
The DOODLE write-up continues in this light vein.
The reader is
let step by step to the point where he learns how to get repetitive solutions
from problems, and how to make use of smybolic nomenclature.
He is told how to do this in a chapter called UThat Mysterious
Column C.
If
Here is part of that chapter:
(SLIDE 7 - PAGE 17 OF DOODLE WRITE- UP)
In the last part of the booklet, we give an example of how the
quadratic formula could be written up.
(SLIDE 8 - PAGE 21 OF DOODLE (MODIFIED)
Notice that the values for A, B, and C are given to the computer
at item numbers 101, 102, and 103.
The constants used by the problem, a zero, a four and a two are
given next at items 104, 105, and 106.
The actual programming of the problem start at item 107. We need
a minus B, so we subtract B from zero and store it in MB, which stands
for rtminus B.
0
"1
n
5
,,'
.~.If· j -
We need a B square, so we multiply B times B and put the result
in B2.
The names where results are stored can be made up by the
DOODLER as he goes along. In the example, we have used names that
remind us of what has been calculated. Notice that when A is multiplied
times C, we have chosen as the result name oAC,
It
and when AC is
multiplied times four, we have used the result name "4AC.
n
There are two answers to this problem, and they are shown at
items 116 and 117. These are the positive and negative roots respectively.
Notice the ttl U under the column headed np.
If
This causes the computer
to print out the results at items 116 and 117.
When the computer comes to item 118, and spots the op KENO,
n
it goes back to the card reader looking for more cards. If it finds more
cards, it wi 11 load what it finds on top of the program being run starting
at the beginning of the program and continuing to load until it comes to
another end card.
In this example, new values of A, B, and C will be loaded on top
of the old values of A, B, and C.
The program will then be executed
with the new values, new answers will be printed out, and the computer
will again look for new input in the card reader.
It is possible for DOODLE to set up branch conditions, to make
tests and to branch about within the program depending upon the value
of test conditions.
It is even possible to do address modification and
branching to subroutines. We donrt recommend that the DOODLE user
get involved with these more sophisticated features of DOODLE. It is
possible, for example, to get into endless loops.
c
6
2U5
C'"
'I
Besides, a person who was this sophisticated would probably be
better off learning and using Fortran.
It is easy to find mistakes when using DOODLE. If sense switch
one is turned on before the program is run, the res ult of every instruction
is printed out.
Before c losing I'll touch upon some ideas which have come to mind
about how DOODLE can be used in some novel ways.
Suppose an engineer has a problem he wants to solve. Rather than pick
up a coding sheet, writing down the instructions and then having them key
punched, he can simply reach into his desk, pull out some pre-punched
cards, put them together, and take them to the computer direct, thus bypassing the key-punching step.
Of course the situation isn tt really as simple as this.
(""
The user
is required to maintain in his desk several hundred cards made up of
about 100 standard instructions.
The next slide shows what some these
standard instructions look like:
(SLIDE 9 - SPECIAL)
The secret in using this idea is that every instruction revolves
around ARG1 and ARG2. If one wants to add two plus two, for example,
the ltdesk programmer tf would have to put a two into ARG 1, then a two
into ARG2 and finally cause the addition of ARG1 and ARG2.
The result
would go to ARG1 and the programmer would then usually have to put
ARG1 somewhere in order to use its results later.
r~
~I
7
. ~ ~J 6
Jf.19·
7
This sounds very wasteful and slow.
this idea has any practicality at all.
It is! We tre not suggesting
Or does it?
Suppose every standard statement had a unique 2-digit number
associated with it.
Then, the programmer could call for a standard
statement by number rather than reaching for it in his desk drawer.
Suppose he punched up all the standard statements he required for a
particular problem on cards starting in column 1 and continuing to
column 80. He could get forty 2-digit numbers per card, which could
represent a fairly elaborate sequence of events.
For example, suppose
the programmer wants to compute the following:
X
= SIN
(e
2
- 1)
He reads in his data so that X is represented by ftstandard A II and
e by
ttstandard B.
If
The solution of this equation could then be as follows:
(SLIDE 10 - SPECIAL)
We have arbitrarily assigned 2-digit numbers to each of the standard
instructions shown here. Note that the value for 1 is standard. So would
other numerical values selected such that the programmer could build
a number as he needed it.
For example, the number 100 could be built
by multiplying standard 10 by standard 10.
We need not stop here. All sorts of possibilities exist. Suppose
the function X;: SIN (e
2
- 1) is to be used often in a program.
The
sequence of operations which compute this function (010214..... 37) could
be summarized by another 2-digit number (say 99) which will cause the
proper sequence to be set up.
In this way the programmer can generate
his own pseudo macro instructions.
.,
'~."\
{
8
297
"
c
"n
""]""Z'P
But now I feel Pm treading in·the area of Fortran programming or
of computer design and will close. ltd like to leave you with this thought:
If you do see an engineer playing tunes on a slide rule or see an accountant's
fingers pecking swiftly at the keys of a desk calculator, please hand him a
penci I and ask him to start doodling.
o
9
298
1-/. /9
r
c
,---
.
,..-.
c'
3,
/I
:';'./9. / i
J
w
A
~
:~
~
\
:t
DO YOU HAVE PROBLEMS ?
do you do RE PETITIVE WORK
on a desk calculator?
do your problems involve NUMBERS
and
WHA T HAS TO BE DONE with those numbers?
why not have a COMPUTER
f/.!~
...
solve your problems for you?
~
...
~
t.;;
its EASY
$
.........
:()
\.,'"
"'"
.sl'P(
1
-
--
-
--- --
----
--_.-........---
first, tell the computer what NUMBERS
191.'25
1654.0
2.0
.3
you are going to use.
suppose they are:
on the· form shown below, yourll see a column headed OP
and a
columnhe~ded
NUMBERS
writeLOD and 191. 25 on the first line under these columns.
do thesa:me for the other numbers.
h.ere'~$how
°'
1.. 0;D·
~--
~t£.~ ~
11 ~p~ 1 1
~r"-'2
~f"iQ
......I-
~O-A
~
.,.,
..:....;,
f8-~
,. 1..
11t
A
a
IIO~-
.'
1.:0 V _
the form wi 11 look after you have done this:
I:'
l1.l &
.
·LQ,D
~
R~ :
,;.,,,
,
.'.
L
I
•
..
!
I
j
i -1
0'
'"
".~.
".'
a·' ...•
.j
I
.L
i
~
I
~
--L.;~L..
I~
I
,
j
I
,
I
I
I
i
-'
. • 3 ~.
r'- . --....-l..
t.t.1.
.cr
J.
I
.1
Nu~eERS
I • ,2., S.... - 1 - -. ~~
~f t. .6,5 ..'t~
I
.
...l.
I
La
,II'>(
.
,
ry
I
.
...l.
I
.
.f
j
I
j
a
~R.
L-J-- t-- --4--'---+_ I - ~
i
-~~.-
c..
II
I
-_ ..
i
t
I
I
1- I
,
.~'
>t . '
"I
-.- ----
NT .s
i
i.
I
(.0 N\ N\.~
.
.----- ..
,
_ _.
I. I,
~J-!-----.--
-'
I
,
I
I'
.,
.~
~O
-.-...
Sl , 01 1..
~
\
/
t{
f \
~"
)
M
m
~I
~I
o
~
~
now tell the computerWHA T TO DO
to do this, you'll need some OP'so
you've already used one.
it was LOD.
some other Opr's look like this:
ADD
SUB
MPY
DIV
~
~" ~
"-
you cal1 guess \vhat these 111ean .
~
........-.
t~
~
.stl"E .J
..
the opt s you'll need are:
suppose your problen1 is this:
\
+-
(1)
191.25
1654
(2)
2 x .3
MPY
(3)
add the t\VO results together.
ADD
DIV
. here's ho\v you \vould \vrite out the problen1
0' '- 1
l
~t.M 1t- I I. .
.01 i L.,O ,:? ~
-u'2
•
!
-~
:8-:~
~o.
,&;
.0.-7
..l O~
__ 1
...\ 0
....
• J.
-', -0
~~.
I
1',
.... I
L(,')i
.-'--
I y',. Ul. I
"'1
~ i 0 3
: !A.D Df
;
; '0.5
""
.
():E1
I
.
,
'
'
--"-
,
.
'- -t>
.--.- --
~-~
I
•
i
I
-.
--~
I
t
I
',:n :lJ-,
,
,
I
.
I
NU""DEI'tS
1M
~_~
I
l(;~'II
'.' ~ -
_1-2-
1
E'>
I
t
I
I
I
.
.
.
'--'"
~ .~
... C.
.
.~
J.:a
•
.
•
.
.
I.'
t
.
I
.
.
,
.-L-J.--..
.
•
I
•
I
--
•
I
•
-~---
NT 5
£,
L
I
I
I
<:"OMN\~
•.
• ,3.
~~
1.1-
t ..
I.q
~
•
','O.L.
I
~
. .
.
J
~~
C.
i..
&
.
I
'J
~
..,. I",
bC.~~:
:o:s_
i fL p'
_
, i): Y Y
I
!
I)~
~
A
itt
-----..
---.-
.--------.
I
I
•
.I
""
L"It,·~"...
\~
--
~
SLIDE ,.
.
. r.._
C.·1
~
f
-..
'.
~,
)
(-J
o
~J
~
you'll have to tell the computer to
print out item 107.
place a "1" in the column headed P
place END as the last OP
here is how the form looks when you have written out your problem completely:
JliT~M ~
1:'.1.. 0' _LJ1.
A
~lf·l",
&
I.
c.
lill [11
II:
~UM B£~S
J:
c.oN\~i NT 5--
-r I [F=lifr9~ .• .}.;..s. -" . .
ff'~:3' Rt~:~LI : : : I,' I : ~ ~ I I : : : I I m~ -?: q~.
..-...-~+-III f - - - - - "4 I IL~ o.D t ;.1 _~ __
I· I : I I 1
I I I I ~ I>f~- -:-.3. I . I
.' 1
J
1LO.t> I
T~~n
-- rr I .-- .--.--r-I .
~.
I
-I
-+-::........
:
I
-
I
I
:
I
I
I
!J:Q:§: 11£~I'U:~:1: UJ:gtll : : : II iTEtI · : ~ : : : : :u: II
.u:.L
'~:~:gi I':O:~: II'~O:(): II : : : 1111'b~~ : :_::_::: : :-H-------1~.IOR
rO~9.
.. t\
~
~
,~~
rn .g."
1
1
_1 __ .1
1
....L
"-
,~
~.,~
c.',J
~J
5L'''f S'
LIST OF opts
OP
LOD
ADD
SUB
MPY
DIV
SIN
COS
EXP
QRT
ATN
ABS
LOG
END
*MUV
*RET
*BCH
*SUM
WHAT IT DOES
load a nunlber into the SI)e<-ified itenl
itenl A ... itenl B
itenl A - itenl B
itenl A x itenl B
itenl A f itenl B
sine of item A (radians)
cosine of itenl A (radians)
itenl A raised to the itenl B l)o\\·er
square root of itenl A
arctangent of itenl A (radians)
absolute value of itenl A
natural log of item A
last card of problenl solution· or of ne\v set of numbers
move item A
last OP in a subroutine
branch to item A, B, or C
sum from item A to item B
*For these OP's see chapter called "For the Advanced Doodler."
~
"-
.~
"
'~:..,:.
@
-=..
5LII>E
,
"-I,
~
\, )
I' '\
\1 j
o
n
~
" _3.'
It
suppose you need the value of" ." " in your problem.'
you can tell the computer what the VALUE of " .". "is
and
at the same time, write down a NAME for the ." item.
you will then be able to refer to that item by the
~NAME
you have given it, as well as its NUMBER.
here's how to tell the computer to calculate the
area of a circle whose radius is 8.1623
the equation is A = .". r2
-- - ... _.... - ..-.:_: : I ~sf;l; II nI:;r~:~:J;6~~~ :: II
~ a£i:II~J;6I'r:~:; II~:~:
:II ::·
II
I
I~r
Y1
•
•
•
•
•
•
•
•
•
~'(
,~
I
'<.'
~
...........
0')
~'"
j
l. 10' 7
I
I
I I
I
....
!",
,
I
I
I
I ,,
!
•
...'"
!
I
t
I,
!
i
'Z
It
tal
I
f
f
I
~
0
t
,
. _.
.
•r
.
....
,•
.-
.- r-- .-
\J
.-
.. .. . .. · . - . .
..
.. . ·
.
.
.. . . .. ·
- .. . . . .. -
:
-
.1-,
_.
J
...
- .
. ·
·
.
..
.:
'"
.. · ..
·
~
III
ID
• ",..=~
t
.
.,-
.·
..
·
..
·r
'" -, 1 .
W
rl..
tJ-Tf
,.;. 1.. . _if-~
:J
"2
.1
I
I
~
~
-
...
1...
.- -t- ... .. - ~.!S
-l_ -.. ..... - ,J~ -- .. - ."';;
10.-
..
\J
f-
r.
~
·
.,..-
-'
.
~.
4
.
,-
r-
- -.- -
10-
-- .-..-
.- t-- •..
- 1--"
·
..
..
· . · .. ·
·
-
... .-
·
..
<
._-
~
- .....
~ lc ~ G.\ 10
~ ~ "2 . .
.. ~ ..J I'~ uJ
.
0
-
.
· .
~ I- I•
~
·
~\l
-!I ~ '.-i"'
~
~.
i---
-- ~
..
~
...-
_._
.....
I·
I .
•
• .d~t''S::
. . c...,_
~
~,
I
I'"
!£,~ ..
I
ldC"" C!
±k d~
Ll/9.~f'~ (I
-r.
309
~.
.~
~
OP
A
LOD
LOD
LOD
MUV
MUV
MUV
A
B
C
B
c
OP
A
B
c
A
ADD
B
C
SUB
MPY
ARGl
ARGl
ARGl
ARG2
ARG2
ARG2
ARGl
ARGI
ARGI
ARGI
MUV
MUV
MUV
ARGl
ARGl
ARGl
ARG2
A
B
MUV
ARG2
ARGI
ARGI
ARGI
~
,
~
}~:~
-"-,
\~.~
,,-c:
jL '''' ,
OPNO.
A
OP
B
.C
..
01
02
14
25
25
40
5-2
34
;
3·7
A
LOD
LOD
LOD
MUV
MUV
MPY'
ARCl
MUV
ON-E
SUB
SIN.
ARGl
ARGl
B
ONE
ARGl
ARG2
ARGl
ARG2
ARGl
B
B :.
ARG2
ARG2
1
0102142540523437
I
I
1
I
I
I I
I
I
I
I
I
I·
:t".~
...}.i~'·
. . ..
-..
".;'
~
\, )
..sLa.E 10
.rOi't
r
~-.~
\. /
.~
'\ ;I
... ......-..-.
-......-.......-
PDQ FORTRAN
For Paper Tape 1620
Edited by
J .. W. Trantum
P. G. Boekhoff
-3193
MRD DIVISION
General American Transportation Corporation
7501 North Natchez Avenue
Niles, Illinois
Phone:
312
NI7-7000
Ext. 233
April 1964
We wish to acknowledge the generously given help of Frank H. Maskiell,
and all associated with the development of P.D.Q. Fortran.
3J2
TABLE OF CON'I'ENTS
SECTION
I
II
III
.ABSTRACT
If
•
•
•
..
•
II
..
0
•
•
..
•
•
•
Lan.guage Specifieations
0
..
•
0
e o . ..
•
•
It
•
•
•
•
•
0
•.•
•
•
•
•
•
•
8
....................................
8
•
•
8
•
0
0
4f
•
•
•
1
.
2
List Of Permissible Statements ••
2
Additional Specificationsv ••••••
Relocatable Subroutines •••••••••••••••••••
3
9
Error Messages During Processing •.••••••••
11
Operating Instructions •.• v•••••.••
Consol
13
Operating and Set Up.
Error' Procedure •.•••.••••••••• " •••••••••••
Debug and Protect Feature ................... .
Object .. Progra.rn. Errors ..
IV
Physical Description ........
0
•••••
0
00008.888
e e. . . . . . . . . . . . . . . . . . . . . .
•••
00
........
8
••••••
"'
0
•••
•••••
13
13
13
14
15
/1r-··~"
",~,/
Modification of the System ••.•••••••••.•••
15
15
16
Object Time Error
17
Tape Layout s .......... v••••••.••••••••••••••••••••.••••
Adding Re10catable Subroutines............
V
. •••••
~odification.
Error in the System ••••••••
o. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •
i
18
I
1620 USERS GROUP LIBRARY PROGRAM ABSTRACT
Title:
PDQ Fortran for Paper Tape
Author; Organization:
Date:
Apri.l. 20, 1964
Direct Inquiries to:
J. Wo Trantum, P. G. Boekhoff
MRD-GATC
Users Group Membership Code:
3193
J.,W. Trantum, P. G. Boekhoff
MRD/GATC
7501 N. Natchez Avenue, Niles, Ill.
Phone: NI 7-7000, Ext. 233
Description/Purpose:
PDQ Paper tape is a modification of PDQ card system
which generates shorter, faster object programs than any other system available
for machines without Floating Point Hardware.
Specifications:
A.
Storage:
I.
Processor requires 17250 allowing
2.
Class A subroutines require 6400
275
symbols
B.
Equipment Required by System:
Paper tape, Auto Divide, 20K.
Program will adjust for additional memory.
Additional Remarks:
Any program written in FO-004 or U.T.O. may be compiled
in PDQ. Additional facilities in PDQ include:
COMMON statements, batch
compilation, continuation of r/Q and FORMAT statements, expanded format (A and
D specifications) and multiple specifications (nFw.d), "Free Format" input,
improved listing of source program, PROCEDURE statements, TRACE statements.,
MOVE statements, IF statement with continue feature, additional relocatable
subroutines, Fixed Point ADD and SUBTRACT in line - No Fixed Constants i.n the
Symbol Table.
o
3 13
1
· - - -_ _ _ _ _ _ _OOM~~:&lIlIJIIiiIiji_iWIiliiU1'idiii$i!_i••liitiiii&I"J .. 11Lm,.t_:J\Ull~~M"...·_l ...·-"'-I'.[
-
-----
II
_________
~
__
~.
_ _
~~~_
..... _
......... ___ .....
_~
'OM.
Language Specifications
Any proper statement in the FO-oo4 or UTa languages is allowed.
Permissible Statements.
A.
Arithmetic Statements:
A
B.
List of
=B
op
C etc.
Control Statements:
1.
GO TO n
2.
GO 'ro (n , n ,
2
l
3·
IF (a) nl , n 2 , n3
4.
IF (SENSE SWITCH i) n l , n 2
(any of the n. in 3 and 4 above may be replaced by the letter c)
n ), i
m
l
5.
DO n i
= m1 ,
DO n
m or
2
= ml ,
i
m2 , m3
6..
BEGIN PROCEDURE
7.
RETURN
8.
END PROCEDURE
9.
EXECUTE PROCEDURE n
/il" '",
n
\,~,.",
n
10.
CONTINUE
11.
CONTROL
12.
PAUSE
n
m
13·
. STOP
14.
END
15.
BEGIN TRACE
16.
END TRACE
17.
MOVE
m
n, A, B or
MOVE
n, A, B, m
3 1 !~
2
c
A procedure is defined as a group of FORTRAN statements which are to be
executed as a unit more than once in any larger program. Such a group of
statements is preceded by a statement
BEGIN PROCEDURE n
where n is a procedure number. (Note - This number must not appear anywhere
in the program as a statement number, and cannot be used in more than one
procedure.)
The procedure statements must be followed by the statement
END PROCEDURE n
where n is the number of the procedure.
The procedure n may be executed at any time by the statement
EXECUTE PROCEDURE n
which transfers control to the statement immediately following the BEGIN
PROCEDURE n statement a.nd' obeys the 'statements of procedure n until
3j
3
~
a)
the END PROCEDURE n statement is executed, which transfers control
to the statement immediately following the EXECUTE PROCEDURE n
statement that entered the procedure.
b)
a RETURN n statement is executed. This return$ control to the
statement immediately following the EXECUTE PROCEDURE n statement
that entered the procedure. A RETURN statement is valid only
within the procedure from which it returns control.
Any statement may be included within a procedure, including EXECUTE
PROCEDURE statements referring to other procedures, except:
B.
a)
BEGIN PROCEDURE, RETURN, or END PROCEDURE statements referring
to other procedures.
b)
EXECUTE PROCEDURE referring to the procedure within which it is
placed.
CONTROL m
This statement controls the typewriter carriage
C.
m
=
101
spaces typewriter
m
=
102
returns typewriter carriage
m
=
108
tabulates
PAUSE
The program execution is halted by a 48 op code in the main line program.
Depression of the "START" key will cause the co~puter to continue to the next
instruction.
D.
STOP m
The program execution is halted after the carriage is returned and STOP m
is typed. If m is omitted, STOP 0000 will be typed. Depressing "STARr" will
cause the computer to continue to the next program instruction. m, if present,
must be an Unsigned constant, a < m < 9999.
E.
END
The program execution is halted after the .carriage is returned and END is
typed. Depressing "START" will cause the carriage to return and END to be typed.
316
~;l(). ;/
C
•
F.
TRACE
The results of all arithmetic statements may be t~aced without the generation
of any additional in-line instructions. Replacement type statements (A == B,
C = D(K), I == J) will not be traced since no arithmetic operation is required and,
hence, FAC is not used. If the replacement statement includes a subscripted
variable on each side of the equality, the statement may be traced. The tracing
of arithmetic statements at object time is by the use of SWITCH 4.
BEGIN TRACE will "tiur;n on the trace feature. Thi.s will replace the
normal FMFAC instruction (26 ADDR, FAC) with a combination "FMFACTRACE'· instruction (17 FMI'R, ADDR). Each arithmetic statement
thereafter will permit tracing. Tracing will be terminated when
an END TRACE statement is encountered. Succeeding statements will
not be traced until another BEGIN TRACE is read.
G.
-PRINT
PRINT n, list will cause the variables in the list to be printed according
to the format specification.n after a carriage return has been executed. This
conforms with the PRINT 'statement of l,i'0-o04. No indexing is permissible within
the list.
(
""'"
H.
TYPE
/
TYPE n, list performs the same function as PRINT n, list with the exception
that no carriage return will be executed prior to the output of data. This will
permi t the printing of cO.lumns of data utilizing the typewriter tab setti,ngs and
the CONTROL 108 statement ...
I.
INPUT/OUTPUT LISTS
Jf the length of the list in an input/output statement requires more
characters than are availabl.e before colurrm 78 ,thelistmay be. continued in
the next statement. ' Terminate the list preceding a continuation wi.th a comma.
Continuation statements must have a digit (a continuation number of character)
preceding the remainder of the list.
J.
COMMON
'The COMMON statement 1,s followed by·a'list of variables whose object tlme
addresses will b.eg.in in the first symbol location after the function symbols.
This will enable the' assignment of. addresses to symbols cOmmon in sever91 programs.
Dimensioned variables which are listed in COMMON statements will have the necessary
space reserved and must riot be again listed i.n a DIMENSION statement. CAUTION:
The COMMON staterp.ent(s) must be the first compiled statement in the program.
EXAMPLE:
COMMON A,B, C,D(2,4),E
31 7
5
." .. ,........ ,.. ......... ,...."........:..r...".................,""'-,.....,-"_
. ..
..,......"'-'-_
~
K.
MOVE
MOVE
n, A, B
or
n, A, B, m
,1f""'-
The purpose of the MOVE statement is to transmit a specified set of fields
to specified locations. m (if present) and n must be unsigned constants. A
and/or B may be subscripted; the contents or the locations to be transmitted
may be fixed, ,floating, or note If m or n are specified ,to be more than 2 digits)
they will be treated modulo 100Q
\",.$'
If m is omitted or specified to bel, n fields will be transmitted from A
and (n-l) successively lower locations to' Band (n·-l) successi,vely lower locations.
If m is specified .to be >1" n fields. 'will be moved from every mth location
beginning with A to every mth location beginning with B,
The MOVE statement generates 3 instructions in the object program (36 core
positions).
Example:
MOVE 3, x(l), y(2), 4
(36 positions)
is the real way to say
100
IX)
100
I
= 1, 9,4 .
y(I+l)
=
x(I)
(120 positions in line + 20 in
symbol table for I and 100)
/~ °gJ
evaluates signum (x)
' 0.0,
1.0,
Method -
obvious
Length -
52 digits
Speed -
2.5 msec.
x
x
x
Accuracy - Perfect
9.
,
( "'"''
,
ONE -
evaluates the Heaviside unit step function l(x)
Method -
obvious
Length -
76 digits
Speed -
2.2 msec.
==
r10.5,
0.0,
x <
x ==
(1.0, x >
~oJ'
~.'~
AccuraC!y - Perfect.
Error Messages During Processing
PDQ FORTRAN has,a limited ability to detect errors in the grammar and
syntax of FORTRAN. This capability is not adequate and is included mainly to
ensure that the processor does not destroy itself.
The PDQ FORTRAN error messages are:
Error No.
Condition
1
Incorrectly formed statement
2
Subscripted variable for which no DIMENSION statement
has previously appeared in the program, dimensioned
variable 'used without subscripts, ,variable in DIMENSION
statement has already appeared in the source program.
o
~ ,;l{).
11
II
Error No.
Condition
3
Floating point number not in allowable range of values,
or fixed point number contains more than four digits.
4
Symbol table full.
5
Mixed mode expression.
6
Variable name in an expression contains more than
five characters.
7
Switch number has been omitted in an IF (SENSE SWITCH n)
statement.
8
A comma follows the statement number in a DO statement.
9
A DIMENSION statement ends with a comma, more than two
dimensions have been specified in a DIMENSION statement.
10
Unnumbered FORMAT statement.
11
Incorrect representation in a FORMAT statement in one
of the following ways:
*
+ . ,) in numerical
a.
Special character ( = @ field specification.
b.
Alphabetic character other than D, E, F, I, or A
in a numerical field specification.
c.
Decimal pOint missing in an E or F-type numerical
field specification.
d.
The number of positions to the right of the decimal
point has not been given in an E or F-type numerical
field specification.
e.
A record mark appears in a numerical field specifica-
tion or an alphanumeric field.
f.
The first character following the word FORMAT is not
a (
g.
The last character is not, or ) .
12
The total record width specified in a FORMAT statement is
greater than 99 characters.
13
A
FORMAT statement number has been omitted in an input/output
statement.
12
III
Operating Instructions
1.
Clear Core
2.
Load PDQ Processor
3.
Switch Settings
36 00000 003
Parity and I/O t~ Stop, OFLOW to program
Program Switch
(''''''
Note:
On
Off
1
List ONLY object time addresses
of source statements
List source statements and
symbol table
2
Do not list names and locations
of relocatable subroutines
(functions)
List complete symbol table
3
Do not compile subroutines on
object tape
Compile subroutines on
object ta~e
4
Source program is on tape
Enter source program on
typewriter
If switch 1 is on at the end of the compilation, the symbol table
will NOT be typed, but the highest address below the object-programsymbol-table will be typed.
Error Procedure
If a source statement error is encountered the processor will type an
error message and halt. Pressing START will transfer control to the typewriter
so that the statement may be correct'ly typed in. Sense Switch 4 serves its
usual role with rega'rd to typing errors and must be turned off before releasestart is pressed. In some circumstances a source program error may result in
a MAR CHECK. In this event one may enter a branch to the error routine (4911~B't)
and proceed as above.
Debug
and Protect Feature
The symbol-table locations for which entries are not defined at compile
time will be initialized to flag-record marks in low address (high-order address)
of each symbol location at the time subroutines are loaded. This will protect
the subroutines and the remainder of the symbol table from the disastrous effects
of referring to undefined symbols (fixed variables excepted). Further, the
record mark will generate a MAR CHECK if an undefined symbol is encountered by
the subroutines.
c
325
Jl02(). )3
13
Object Program Errors
Error conditions detected by the subroutines generate the following error
messages and results:
Condition
Message
Contents of FAC
Operation
Overflow in fixed
or floating
arithmetic
OFLO
NINES
continued
Underflow in
fixed or floating arithmetic
UNFLO
ZERO
continued
Overflow in EXP
Underflow in EXP
EX OFLO
NINES
ZERO
continued
continued
EX UF10
Negative argument
in SQRT
I AIN'T REAL
Argument out of
range in SIN-COS
Function Undefined Beyond Allowable
Range (FUBAR)
""argumentt'
argument
continued
terminated
(If-"
l,eJiI.....,pJ
Zero argument in
LOG
LOG 00
MINUS NINES
continued
Negative argument
in LOG
I CAN'T TAKE
YOUR LOG
log \argumentl
continued
Input out of
range
(none)
NINES
continued
Fixed Point Arithmetic
The operators +, ~ and * in the fixed point mode give results mod 10000.
Hence, the addition of 9998 and 4 gives 2 with no error detection due to the
overflow. Should division by 0 be attempted, an error overflow will be indicated.
326
14
C", "
IV
Physical Description
~.
Processor tape consists of the following records
1.
Loader and identifier (numeric)
2.
Tables and program (numeric)
3.
Subroutine; names
dump of locations 00100 through 17850.
(alpha)
Subroutine tape consists of the following records (all numeric)
1.
Loader and identifier
2.
Tables
3.
Relocator (dump or locations 00500 through 02553)
4.a
Relocatable subroutines (many card images)
4.b
End-of-relocatables record
-5.
Class A subroutines (dump or locations 00100 through 06367)
6.
Loader
7.
Symbol-table initializer
8.
I/O buffer initializer
9.
Type message, halt, and branch to 06400
A duplicator program is provided with the system.
both the processor and subroutines.
It will duplicate
Adding Relocatable Subroutines
The name of the subroutine must be added to the last record of the processor.
Its order number in that record is the subroutine #+ 1., Subroutines must be
written and processed in an SPS which produces card-image output e.g., 1620-1710
SPS. Subroutines must begin with DORG 5000. Flags on the zero and one positions
of each instruction will cause the P and Q addresses respectively to be relocated.
No other flags may be used in addresses' to be relocated except for positions 6
and 11. Flags in those locations will gener~te proper indirect addresses.
327
C
"',I
,','
15
-
--------------------
-,._---.-.. ...- ' - - ,
To assemble a relocatable subroutine, delete all records from the SPS
output except those containing information generated by program instructionsj
i.e., delete the loaders and tables. Then add header and trailer records as
follows:
Header Record
XX (Subroutine No.)
62 blanks of:
(Subroutine No. must be less than 23)
Trailer Records
1.
Memory-capacity digit records
64 blanks Xxxxxf: (location of any digit which must be adjusted
for memory size.
2.
Subroutine length reoord
xxxxx (length of subroutine - must be even) 57 blanks
aooof:
•
Modifications of the System
The following modifications to the system may be incorporated at the
user's discretion.
10
Compilation Errors
If the user prefers not to correct source program errors at compile
time he may (1) ignore the erroneous statement by changing the digit in
location 13241 of the processor to 9, or (2) terminate compilation by changing
the contents of locations 13241-6 to 909860.
2.
Trace Format
To elimi.nate printing of the 4-digit address of the traced variable,
change subroutine locations 02694-5, 02736-'7 to 41.
To modify the Format for Fixed and Floating variable change the fields
at 2327 and 2359 respectively to the desired specifications.
3.
Plus Signs in Output
Plus signs are normally omi.tted in both tape and typewriter output.
To obtain plus signs in output change the digit in location 03196 of the subroutines to 1 for floating output, change 04694 to 1 for fixed output.
32b
16
· tttt
rtMH
c
d bHtiU·xtiH+ritin+·\
*"¥+++witt.l'··ii+,.r· ... ··'··J·
u·iljGiiiAlI,Hi.-··T ... ·'trrn · zvu·P"2
.-
Object Time Error Modifi.cation
Any error condition message may be deleted by changing the carriage return
and output instruction op-codes to 41. The system now has a Nap instruction
following each of the instructions which write error messageso These may be
changed either to Halt instructions (48) or 'to branch to END (·\490262~) to
terminate the program.
329
17
V
Error in the System
There is one known error in PDQ FORTRAN inherited from FORTRAN with FORMAT.
The statements
A ::; - B
**
C or
will compile as if they were
A
=
(-B)
**
C or
A = (-B) ** I
To obtain the desired result, it is necessary to write
* * C)
A - - (B * * I)
A = - (B
or
/("",
(",,..lV'
c
18
SEX FORTRAN
For Paper-Tape 1620
Edited by
P. G. Boekhoff - 3193
MRD DIVISION
General American Transportation Corporation
7501 North Natchez Avenue
Niles, Illinois
Phone:
312
NI 7-7000
Ext. 233
(""~"
,.'
May 1964
The author wishes to acknowledge the assistance of
John W. Trantum, Frank H. Maskiell, John W. Holmes,
Messrs. Forbisher and Laroche, and all others whose
work has made possible the development of this system.
o
331
I
THE CONCEPT OF SEX
SEX FORTRAN is a system based on, and largely identical to, PDQ
FORTRAN for Paper Tape. To the best of the author's knowledge, no one
presently disputes the claim that PDQ is the most efficient and powerful
FORTRAN in existence for a poor man's 1620 (20K, auto divide); this is a
sufficient (though not entirely the actual) reason for taking PDQ as a
starting point.
One reason for the writing of SEX FORTRAN is that any FORTRAN must
be less than maximally efficient, and that many progrffiTh~ers are sophisticated
enough to recognize and circumvent (e.g., by judicious patching) the inherent
inefficiencies. In most systems, this requires the programmer to operate
outside the language. SEX is designed to give this type of programmer a
head start toward program efficiency by providing several new, frightfully
nonstandard statements in the language, and by streamlining the treatment of
parts of the standard FORTRAN language.
A second reason for SEX is that, in several instances at the author's
installation, programs have appeared which have overlapped by several thousand
locations - too much to permit a simple tightening of the source program, but
small enough to suggest that there should be a reasonable way to squeeze them
into core.
SEX provides an answer in two ways. First, for the programmer who does
not wish to stray from conventional FORTRAN, any program in aIde FORTRAN,
. FO-004, UTO, or PDQ may simply be recompiled in SEX from the original source.*
This will save from 3500 to perhaps 3000 locations on a large program, depending upon the system against which a comparison may be made. For example, any
program requiring 25,000 locations in UTa, or about 23,500 in PDQ, will almost
certainly fit a 20K machine in SEX. The SEX object program begins in location
03100, and both the object program and the symbol table are generally shorter
(never longer) than those generated by any of the above systems.
-<''''\
\~,.,J
Second, for the benefit of the programmer who has outgrown the idea that
the 1620 is a mysterious black box, the more powerful language of SEX will
permit corners to be cut in the source program.
The other side of the question of efficiency is time. Exclusive of I/O,
running time in SEX is slightly but not significantly faster than PDQ. The
I/O'subroutines process faster in SEX; actual time spent on the I/O operations
depends in other systems on the format, which is rigidly fixed in SEX. Thus no
direct comparison is possible. The author claims, however, that SEX in general
does yield a faster as well as a shorter object program than any other system
for this simple-minded machine configuration.
*
A few statements are not accepted by the SEX processor:
SAY and OUT
332
,Jf. ;2/. /
/.
()
II
THE TECHNIQUE OF SEX
The best way to permit a significantly longer object program is, of
course, to remove unnecessary instructions from the Class A subroutines.
Since the author did not wish to cripple any of the PDQ arithmetic subroutines, the input-output routines (some 3500 locations) were discarded
and replaced by a set of rather austere I/O subroutines requiring only
350 locations. This is the greatest space-saving innovation in SEX, and
it is from the two new statements corresponding to these routines (SELECT
and EXTERNAL) that the system derives its name.
Accompanying the adoption of this new set of I/O routines there are
of course restrictions on data format. All numeric data must appear in
excess-fifty or in 10-digit fixed-point representation (DIO format), one
value per record. Alphameric data, in A5 format, may also be input or
output if desired. Since all format (except typewriter control) is strictly
defined,there are no FORMAT statements in the system (they will be accepted
but ignored by the processor).
The arithmetic subroutines for fixed-point addition and subtraction
have been eliminated. The add and subtract instructions now appear in line.
This results in a saving of 30 locations in the Class A subroutines, at the
expense of one instruction (12 locations) in the relocatable LOG subroutine.
Further, smaller savings of space were introduced by a slight change
in the STOP routine and by the author's willingness to accept a somewhat
questionable format for "TRACE output (these are described in Section IV
'below) .
The symbol table has been shortened in SEX by the elimination of fixedpoint constants (these now appear in line) and by ignoring FORMAT statement
numbers. The adoption of the PDQ "continue" feature for IF statements makes
possible the elimination of many statement numbers from the source program
(and hence from the symbol table).
The object program itself is for the most part the same as that generated
by PDQ (i.e., shorter than in any other system), with two exceptions: FORMAT
statem~nts are ignored, and fixed-point arithmetic statements of the form
I = I - J (where J is any fixed-point expression) are compiled correctly by
SEX, which saves one or two in-line instructions depending on the form of J.
In many programs, logic is determined by defining variables which function
only as binary switches to be tested by the program. To eliminate the obvious
waste of reserving 10 digits in the symbol table for each such creature, SEX
provides a bank of ten single-digit SENSE LIGHTS, which are described in
Section V.
For the sophisticated programmer, it is possible ln SEX to save further
space in cases which are beyond the limited psychic capabilities of the SEX
processor. Provision is made for the programmer to patch his program at
compile time. The HANG PATCH statement is discussed in Section V. .
III
THE LANGUAGE OF SEX
The following is a list of permissible SEX statements. For the most
part, these are the usual statements found in any 1620 FORTRAN, from FO-oo4
on up. Those which were introduced (or altered in meaning and/or treatment)
by PDQ have been marked with a single asterisk. Statements marked with a
double asterisk are peculiar to SEX.
A.
Arithmetic and Data Transmission
A
=B
op C etc.
* MOVE n, A, B or
* MOVE n, A,
B.
B, m
Input/Output and Format
**
SELECT, wxyz
**
EXTERNAL, List
ACCEPT n, List
P:IRINT.n,List
PUNCH n, List
READ n, List
*TYPE n, List
FORMAT (
)
CONTROL n
C.
Program Control and Logic
CONTINUE
GO TO n
GO TO (nl , .•. , n j
DO n i
= ~,
DO n i
), i
m2 or
::: ~, m , mg
2
33'1
.If. ~/. 3
C
**
LIGHT i
**
LIGHTS
ON
OFF
ON
OFF
IF (Expression) n , n , n3
l
2
IF (SENSE SWITCH i) n , n
2
l
IF (SENSE LIGHT i)nl , n 2
PAUSE
*
*
**
STOP or
STOP
n
EXECUTE . PROCEDURE n
BEGIN PROCEDURE n
RETURN n
END PROCEDURE
n.
END
D.
Declarative and Miscellaneous
**
HANG PATCH
DIMENSION
* COMMON
* BEGIN TRACE
* END TRACE
All statements are interpreted and compiled as they are in PDQ FORTRAN
except as herein noted.
335
IV
NEW TREATMENT OF STATEMENTS
~,
Most statements are handled by SEX as they are by PDQ.
A.
The exceptions:
',,,v
ACCEPT, PRINT, PUNCH, READ, TYPE - each of these I/O sta.tements
is compiled in the "standard" manner, subject to the format
restrictions of SEX; i.e., each generates a SELECT instruction
(q.v.) defining the format that I have decreed to be the most
"natural tf, followed by an EXTERNAL instruction for each list
variable. AN I/O statement with no list (SUCh as PRINT 1) will
generate (a) a SELECT instruction in the object program, and
(b) a compilation error. The error routine allows the operator
to leave the SELECT instruction in the object program or to
overlay it with the next statement.
I/O continuation statements (as in PDQ) are accepted by the system.
B.
FORMAT - the SEX processor will accept but ignore FORMAT statements
and continuations (their statement numbers will not appear in the
symbol table). All I/O defined by the statements listed in paragraph
A must be numeric, in DIO format. Alphameric I/O (in A5 format)
is also allowed but must be specified by a SELECT statement.
C.
STOP or STOP n - The only innovation here is the abbreviation of
the printed message to S NNNN.
D.
TRACE - The format for trace output has been modified to
AA.A.AAVvvvvvvvvv for floating and AA.A.AAOOGGGGVvvv for fixed
variables, where
A is the address of the variable,
V is the value of the variable, and
G is whatever garbage may be in. the accumulator.
(f--",
'c",.",,/
w. 'M'HI t
r
h"
t \\
bIt ttrte
Ifi
V
~.
TREATMENT OF NEW STATEMENTS
The following is a description of the new statements mentioned in
Section I.
A.
SELECT, wxyz
This is the statement which defines format and controls all input
and output. The arguments are as follows:
w =
x
('~"
y
..."".
=
=
(
(
(
(
0
for numeric input
1
for alpha input
2
for numeric output
~
3
for alpha output
(
(
(
(
1
for typewriter
2
for paper-tape punch
3
for paper-type reader
(
0
for no control before first list item
1
for space before first list item
2
for carriage return before first list item
3
for tab before first list item
~
(
(
(
z
=
same as y; applies to all items in list except the first.
The polypotent SELECT statement generates one instruction in line which sets up
the I/O subroutines to handle the list variables in succeeding EXTERNAL statements. The arguments corresponding to standard" r/o statements are:
TI
ACCEPr ... 0123
C
PRINT
- 2123
PUNCH
- 2200
READ
- 0300
TYPE
- 2133
337
~------
B.
-.-".
..
EXTERNAL, List
This statement generates, for each list variable, an instruction to
input or output the variable as specified by the last SELECT statement encountered.
As an example of the use of SELECT and EXTERNAL statements, consider the following
two programs:
C
PROGRAM I (PDQ)
READ 1506*, A, B, C
PRINT 1506, A, B, C
1506
FORMAT (D10, 5X, D10, 5X, D10)
c
PROGRAM II (SEX)
SELECT, 0300
EXTERNAL, A, B, C
SELECT, 2123
EXTERNAL, A, B, C
/t:iIIf~,
<",./IV'
END
The operation of these programs is identical (in fact, the SEX object programs
generated would be identical). In Program II, the first SELECT statement tells
the I/O subroutines that data are to be read numerically (w = 0) from paper
tape (x = 3), with no typewriter carriage control (y = z = 0). The first EXTERNAL
statement causes the values of 3 variables to be read and stored. The second
SELECT tells the subroutines to write numeric (w = 2) on the typewriter (x = 1)
with a carriage return preceding the first item on each EXTERNAL list (y = 2)
and a tab preceding each succeeding item (z = 3). The final EXTERNAL statement
causes the values of the list variables to be read from storage and output as
specified.
c.
LIGHT i ON (or OFF)
The function of this statement is obvious except for the definition
of the symbol i, which can be either a single digit (0 through 9) or the name
* cf.
Forbisher-Laroche
338
7
G
· . . mn . Jrn""5nHfre
.
of a vari~ble presumably (though not necessarily) defined elsewhere. If a
variable, the light affected will be that corresponding to the low-order digit
of the value of the variable at the time this statement is executed. If the
variable is fixed-point, this will be the units digit. Use of a floating point variable in this statement is permissible, sometimes useful, and usually
dangerous. The sign of the variables will have no effect unless the program is
being run with IA turned on, in which case the author disclaims responsibility.
Example:
The sequence
Q
= 0.0
LIGHT Q ON
DO I IDIOT = 6903, 6907, 2
I
LIGHT IDIOT ON
LIGHT 9 ON
will cause lights 0, 3, 5, 7, and 9 to be turned on, in that order.
Loading the subroutines to start the program. will reset all lights OFF.
If any light is turned on, it may be turned off again only by means of an OFF
statement; interrogating a light does NOT turn it off.
D.
LIGHTS ON (or OFF)
This statement simultaneously sets (or resets) all 10 lights.
IF (SENSE LIGHT i) n , n 2
l
As in the ON and OFF statements, i may be the name of a variable;
as in other IF statements, one (or both) of the statement numbers may be omitted.
If the light referenced is on, the program branches to ~; if off, to n 2 •
E.
If both statement numbers are omitted AND i is the name of a variable, the
object program may not run properly.
The continuation feature for all IF statements has been modified from PDQ;
the letter C for the continuation may be replaced by anything (or nothing)
neither beginning with a numeric digit nor including a comma. This should be
of great value to those of us who delight in writing fudge recipes and/or bad
poetry in our programs.
Ci
., (
3. II ~J
F.
HANG PATCH
This statement allows a machine-language patch to be entered at
compilation. When a HANG PATCH statement is encountered, the·SEXprocessor
transfers control to the typewriter to accept the patch, the idea being that
the map listing being produced will make the necessary addresses available.
Two of the more immediately obvious uses of this statement are for alpha
output (such as column headings) and for direct branches within the object
program to avoid using symbol-table space for statement numbers.
A few words of caution are in order:
(1)
normally, no record mark should appear in the patch. In
this case, the last digit entered at the typewriter will
be the last digit inserted into the object program;
(2)
if a record mark with no flag appears in the patch, it will
go into the object program as the last digit of the patch,
and anything following the record.markwill be ignored;
(3)
if a flagged record mark appears in. the patch, the preceding
digit will be assumed to be the last digit of the patch;
(4)
the processor will accept unflinchingly a patch with an odd
number of digits. Care is.advised.
34 II
9
pH
l1{h·
"··b {rttHlHWfhri
·db.drtri¥Ri&*WMiHW.....([ TIi ''''!Ii"'t ..
VI
r"
r
III
,r
FI II I". P
DESCRIPrION AND OPERATION OF THE SYSTEM
The processor occupies locations 00100-17480, permitting 252 entries
in the symbol table. The Class A subroutines end in location 03099.
Relocatable subroutines are stolen directly from PDQ (q.v.).
Errors detected at compile time are for the most part the same as
those of PDQ, UTO, etc. The routine for handling them is a bit different.
When an error is detected by the processor, a message is typed and the
machine halts. If Switch 2 is turned on and the START key depressed, the
carriage will return and a corrected statement may be entered at the typewriter. If Switch 2 is off when START is depressed, the erroneous statement
will be ignored and the next statement will be read in.
This error routine and the existence of the HANG PATCH statement make
it advisable for the programmer to be available at compile time. Simple
patches are provided with the system to permit changing these procedures for
closed-shop operation.
Except as noted here, operation of SEX is the same as with PDQ.
and instructions are provided with the system.
10
Details
THE ART OF DEBUGGING
('If">-'
"~""jOII"
E. J. Orth, Jr.
Southern S,M:1ces, Inc.
~'" Ii ..
q
({~'
,I-i
------------------------~----~--~--~--------- ~----
The Art of Debugging
E, J. Orth, Jr.
Southern Services, Inc.
600 North lath Street
Birmingham, Alabama. 35202
This talk is based on experiences' in writing and debugging large FORTRAN II
programs at the Computer Center at Southern Services, Inc., in Birmingham,
Alabama. • As such, it is the sum total of' the experiences ot several' people.
Although specific reference will be to FORTRAN II oriented debugging, the techniques may be applied to any system. No claims are made tor new ideas. Bather
it is hoped that this representation of old ideas in what is hope tully a new
framework will be useful to all concerned.
FORTRAN II provides two types of' trace: An equals trace, and a logical
trace of sorts. The equals trace prints the result ot each arithm~tic statement;
the logical trac'e prints the results of oODlPUtations specified within an IF
statement. The normal output mode i~ the typewriter; however, an' object time
modification described in The 1620 Users Group Newslet1;er, Vol. 1, No.2, allows
conversion to punch mode. The normal FatTRAN trace will be called the micro
trace in the rest of' this discussion.
(:
A long program involving many computations would require reams ot micro
trace to reach a bug oceuring late in the program. The tollowing techniques mal"
be used to simplify the problem:
1. The subprogram concept
2. The PAUSE N statement
3. A special debugging subprogram.
The Subprogram Concept
Subprograms can materially-cut the volume of a one tor one trace. Atter a
subprogram is checked out, it may be compiled in the no trace mode, thus elimi...,.
nating all trace output from that subprogram at mainline debug time.
Subprograms need not be restricted to repetitive routines.
mainline logic may be checked out and CALLed as a subprogram.
Sections of the
The PAUSE N Statement
Use of this statement at strategic points in the program, each with a
unique N, will allow the programmer to step through his program observing N in
OR2. When a trouble spot is reached, trace may be initiated. Thus the PAUSE N
statement allows an awkward type of macro trace.
A Special Debugging Subprogram
A more useful type of macro traoe may be provided by a single argument
subprogram whose function is to type and punch "STNO In (where I is the' argument)
and pause. Type/punchot "STNO I" would indicate that the program is about to
execute statement number I. Such a subprogram may be CALLed at the beginning ot
each main logical block. The argument would be the statement number or that
logical block. Let us call this subprogram STNO.
3l
~
-1
v
- - - --~------.. --.-~---~------------,
The· utility of' STNO may be expanded by addition of' a dump routine such as
No. 1620-01.6.093. This dump routine is in the f'orm ot a library' function, and·
is relocated trom the subroutine deck by J = DUMP ( J) • J, the argument, is an
. integer ot tour digits, and specif'ies the section of' core to be dumped. For
instance i f J is 1214, all core f'rom l2000thru 13999 will be dumped in 5/75
torma.t. With the proper argument, the library function DUMP mal" be used to take·
snapshots otselected portion ot core such as the symbol table or the common
storage a"a, on up to all ot core.
c
As a tinal ref'inement, both the STHO subprogram and the· use ot the DUMP
routine within that program may be placed under console switch control. Thus,
the entire subprogram .1' be bypassed, or only the DUMP routine bypassed. See
now chart in appendix B • The STNO debugging tool may now· be used as tollows:
1.A macro traoe of the logic through the punch/type "STNO I"
2. Selected core dumps
·3. Micro traces as desired, selected through CS4 at the proper
"STNO I" typeo
.
If punched micro traoe is used, the punched "STNO I" will allow the programmer
to readily find his place. Punched micro trace is most definitely recommended
to save computer time.
It an on-line printer is available, read print for punch in the above.
Theusetulness ot select!ve core dumps cannot be overemphasized._ A prime
example concerns matric manipulation. A snapshot ot the 1«)rk area before and
after completion of a mat ric operation maybe compared to the micro trace ot a
FORTRAN statement. Indeed, a subprogram such as STNO incorporating some sort ot
dump routine is absolutely necessary' in .programs involving matrix algebra. S'l'NO
and DUMP are to matrix algebra as the FORTRAN micro trace is to arithmetic
statements.
Another Use of DUMP
The following sequence may be placed at the very' beginning or a program
during debug time to allow variable length core snapshots at will.
1 FCBMAT (14)
JF (SENSE SWITCH 4) 2,.3
2 ACCEPTl,I
I=DUMP(I)
PAUSE.lllll
:3 CONTINUE
FloW Charting and Coding
Certain simple documentation p:ractises will speed debugging. For instance,
each page of the flow chart. should b~ numbered. It is most useful if' each numbered statement concerned with the logiC on a g:i.ven page uS.e the flow chart page
number as a hundreds digit.
Each logic block on the flow chart should be cross-indexed to the FORTRAN
sOttrce program by statement n;umbero That is, the number of the first statement
in a block should appear above that block on the now chart.
The obvious utility of' comment cards need
not .be
3·4 I{
belabored.· here.
2
o
itid.f·ti·""'··
(
""
J'"
t ··
f ±b±ilii,Hftd+·dHdiH6i8i±r'idri·,ti,tt&&HiW···*wg .
-3
•.•.••.••
n
i"ff·_··
.
r·r··
nrT
·!··
..-
Rim· ....-
"·ij'_·!·[W···h-
The FCRTRAN statement STOP N is most useful as an error trap. Each N should
be unique, and a table of N, statement number, and type of error should be prepared
STOP N should be used to trap unexpected, illogical conditions. For instance in testing a variable in an IF statement: If the variable may be positive
or zero but never negative, use STOP N on negativeo Programmed switches and the
computed GO TO are very convenient
When testing a set of switches, STOP N may
be used to trap theoretically impossible conditions. The only cost is the statement number, and the STOP statement o The STOP N statements may be removed when
the program is debugged and no errors will occur.
0
0
Conclusion
Some sort of disciplined debugging procedure must be adopted in order to
assure optimm utiliza.tion of programmers, operators, and equipment. If we are .
to grow in our profession, we must adopt and submit to this discipline. The
days of bit chasing and typewriter core dumps are fast disappearing. We must be
ready to program and debug without touching or seeing a computer. A painful
thought? Yes
But have you heard of System 360?
II
APPENDIX A
When a card punched as below indicated is LOADED after a FORTRAN II object
deck is loaded, all micro trace will be in the punched mode o The output may be
listed 80-80 on the 407, with a card taped over printwheels N thru SO to block
printing of garbage. N=ff+3 or kk+l, whichever is greater. ff is floating
precision; kk is fixed precisiono
Procedure: Reset, LOAD this card. Typewriter will print "PCH TRACE MODE".
Then branch to beginning of object program
0
16 06701 00041
15 06733 00004
15 06753 00004
39 00051 00100
48 57434 80063
59 41434 50054
56 44450 I
Replace RCTY with NOP
Replace WNTY with WNCD
Replace WNTY with WNCD
Print "PCH TRACE MODEtt
In most cases it is desirable to permanently modify the processor to punch
trace.
o
345
3
APPENDIX B
One pos,siblecontigux-ation of the STNO subprogram is shown.
CSl"controls type/pmch of "STNO I" and CS 2 controls DUMP.
In th1sversion,
ON
Pause
OFF
ON
EXAMPIE:
* These may be
removed after debug time.
(i
.. pew')
···f·····
'-p
"H"""P"iI"W'
I
_.
-
rqnrn"l
.
c'
SAKP
Sea reh and Jlemot"1' Print
Jack II. Wolte
BrooklYn Colleee
Brooklyn, New York
(81M P)
Jack M. Wolte
SIlRCH
1Nl)
MDl>RY PRINT
Brooklyn College, Brooklyn 10, R.Y.
(Ueer 11026)
DESCRIPTION
This utility program is a .... ilabl. from the 1620 Gen.ral Program
Library und.r I 1.6.100. It CLn be used as an aid in debuggiDg
machine language programs in which the successive instruction
10eatlo~ are st.epped up uniformly be 12 and in which the
iastructions are not interspers.d with record areas. If these
condi tioll8 do not prew-il for the entire progr. . , however, this
utility program can be applied. separately to each individual
portion of the program tor which these conditions are satisfied.
The .-.orr print part ot thi8 program will write out on the
typ.wri ter &JV' select.ed portion ot the memory in either ins truction format. or da.ta tor..t. In instruction tormat the typeout
con'taina the location, the OP code, the P-e.dclres8 and the Qaddress, appropriately spaced tor ease of reading. Plags and
record. marks are typed out. Bach instruction is typed on a new
line.
In cia ta tonat each lino contains the 10eation and the contents
of the ten Jl.8JIIOZ7 pos1 tiona starting at that address. Becaus.
the error in the program will otten become apparent before the
eompl.tion ot the memory print, the typeout, whether in instruction or cia ta. for...t, can be terminated order 11' by program swl tch
control at ~ ttm. in the processing.
The _.reh part ot this program _1' be particularly usetul when
the prog~ being run stops on an invalid operation code due to
a branch to a location that 18 actually not the beginning ot an
.instruction. BT diaplaY'ing Illl and subtracting 12 from that
adelre,., ve know the address of the instruction wi t.h the invalid
OP code. The BElRCH can then be used to search the instructions
part of the program being debugged tor all branching instructiou
whose P-part. i8 the address of the instruction with the inn.lid
OP coele. The SEARCH will typo out the locations and the instructions ot all conditional and unconditional branches to that
adelr.... It take. indirect addr.ssing into account to one level
when aD in4irect addr... appears aa the P-part ot a branchins
iM true tion.
,.. a by.~proclu.et ot the .earch, the· prolram types out all la.dlate instruetioM tound. without a 1'1ag in the tirst tour poll tioD'
ot the Q-part of the instruction, except, ot course, tor the OP
code 15. The locations of the typed instructions are also typed
out.
o
, t
,,-.
(S AMP) contlnued
(~-
Sometimes lt will be found that the search is complete4 .itboU\
finding aDY branching 1n8truction that could bav. caused the
program to branch to the addr.ss being sought.
be clue
to the tact tJ.w.t a branch aigbt haft -,' been _de to a 100&tioD.
that 1s 12 poaitioM 10." tan the ..delre.s "bere the program.
bung up. It the 12 digit. begiDD1ns at that addre•• oonstituted
an executable 1ruatructloD, the coaput.. would. Dot have hlmI up
until the 1'ollowing instruction was reached. !bus 1t 18 desirable
te pertorm a • .-017 print startiDg • .,.eral instructlon legtha
betore the aclclress where the prop-a huDg up. Thi. vill reT.l
whether a branching might haTe taken place to aD acldr ... prececliDg
the addreee 01' the 11'1....11d OP oocle. "1th thi. intoratioD, the
SEARCH can be p1a.rmed more ettectl.,ely. Quit. ott., lIor...,.. ,
the m8l1017 print i taeit will re.,.l the actual error and the SBlRCH
will then be _4e onl1 when it 18 apparent that the 'prograa could
baTe r_ch.d the iJmll1cl OP cod. only hy branchUIg to It.or to aD
iD8truct1on prece41Dg it.
Th1e.,
STORAGE
Thi. utilit, program itselt oocupie8 ...or" positioD8
~400-'9999.
The following chang.. 11'1 the source progr. . , which is written in
SM, will enable the program to be relocated. tor a 20]( . . .ry.
Source Card
01010
ol4Oeo
oliOS,
ChangeaDORG
(or 8.!l¥ other address at which
.2600 cona.cutive positions
are &ftl1able)
Change column -16 fraa , to 1
01,088
04100
04110
04118
041,0
04150
04160
17400
•
•
•
•
•
•
•
•
Por modltlca tion tor 201, p081tloD8 19995-19999 BlUSt be a_llable.
Indirect a44resslag.
Without auto.atl0 41ylde.
o
It autGlat1c 41ricle harcl~e 1• .!re.ent, cSDce the second. operancl
01 eource card 02270 fro. 18 to '2"0. It the present object deok 1.
aftilable, oSDce ooluma ,,..~ 1'1'_ .18 to 20 on,card 1,.
349
(8 A K p) continUed
OPBRlfIlfG IlSftlJOTIOIS
'or the _st part the operatiDg instructions are cODyeyed by
t,ypwzeittcm. ....age.. '!'h. SAMP object d.eck .., b. loaded in1tially
.... ';, bruched to "hen n ••ded. It, howeyer, an error' in the proFto be clebugged' clest.roye the utili t7 program in
i tou be
rel_elect r_dil7J 1 t conta1Da 70 carda.
."17,
,
,
The source program, write-up, and. illustrative output appear ln the
.taral a_llable :trom. the 1620 General Program L1brarr a8 1.6.100.
35U
..
il
",.,
"
. 3'"
~
SYMBOL TABLE PUNCH PROGRAMS
FOR
I,. B.M., U. T.OD AND P.O.Q. F,ORTRANSYSTEMS
, By ,
Richard C. Irons
u.s. NAVAL SCHOOL OF AVIATIQN MEDICIN~
u.s. NAVAL AVIA-TION MEDICAL CENTeR
PENSACOLA, FLORIDA
'
USERS CODE: 1159
351
TABLE OF CONTENTS
Page
INTRODUCTION. • • • • •
1
I.B.M., U. T.0. SYMBOL TABLE PUNCH DESCRIPTION • • • • • • • ••
2
P.D.Q • SYMBOL TABLE PUNCH DESCRIPTION •• • • • • • • • • • •
2
OPERATING INSTRUCTIONS • • • • • • •
3
ADAPTAT10N TO OTHER THAN 20K MACHINES. • • • • • • • • • • ••
,4
GENERAL FLOW DIAGRAM. • •
5
I.B.Mo, U.T.O. SYMBOL TABLE PUNCH PROGRAM LISTING.
.6
LISTING FOR SAMPLE PROBLEM "NUMBER 'O'NE'.
~ ~ ~
'0'.
•
•
•
•
9
SYMBOL TABLE LISTING FOR SAMPLE NUMBER ONE • • • • • • • •
13
P.D.Q. SYMBOL TABLE PUNCH PROGRAM LISTING • '. • • • • • • ••
15
LISTING FOR SAMPLE PROBLEM NUMBER TWO • • • • • • • • • • • ••
18
SYMBOL TABLE LlSTIN,G FOR' SAMPLE NUMBER TWO '. • .' • ~ • ~~ ~ ~.
21
0
•
•
0
•
•
•
•
•
•
•
•
•
•
•
•
0
•
•
••
0
0
0
•
•
•
•
••
•
•
•
•
0'
••••
•
•
•
•
0
•
.'
•
•
•
•
•
••
•
•
•
•••
0
•
•
•
•
. .'
352
ii
c
INTRODUCTION
Fortran (FORmula TRANslation) is a popular computer language because it
may be used without knowing the internal operations of the computer. A mai or
drawbock of the Fortran language, however, is the difficulty in finding program errors
when they occur.
In processing ,a Fortran language program a table containing the storage
addresses for calcu lated or constant values is produced. This table is known as a symbol
table. The table can be a good aid in locating errors in a program because the values
in question can be looked at and inferences mQdeas to the causes of the difficulties.
Since the time required to type out the entire symbol table is relatively long,
it is desirable to have the symbol table punched onto cards and then listed, perhaps, on
an I.B.M. 407 machine. This makes the thne required to obtain the table negligible,
and the operation, therefore, more desi rabl e •
Programs for punching the table have been mentioned in the 1620 USERS
GROUP NEWSLETTER but these programs do not include all of the possible information
that could be punched. In addition to the information punched out by these programs
it is possible to provide the program execution phase addresses for any numbered fortran
statement, or format type statement. With this additional information, statements can
be referenced without necessari Iy referring to a program listing for the information.
To accomplish this, two programs have been developed, one of which is
compatible with the I. B. M. I U. T •
0', ' ,
'''I
o.
Fortran systems, and the other with the P.o. Q'.
Fortran system. These programs will provide the symbol table on cards, at the computerls
..
card punch speed, in a format that is easy to read and.th,at co~~ains t;he maximum amount
of information available.
DESCRIPTION OF PROGRAMS
I.B.M., U. T.0. SYMBOL TABLE PUNCH
The symbol table, described above,
is generated during the
processing of the
.
.
~
,
,~
Fortran source program and is not complete unti I the entire source program has been
...
processed. This makes it necessary to load the symbol table punch program after the
processing of the· source program has been completed. In general it is advisable to use
;
.
the Fortran Subroutine Deck during the compilation phase rather than during the object
deck execution phase. This is mandatory for the U. T .0. version, optional for the
I.B.M. versions. The U.T.O. processor will automatically clear the symbol table
If"\
storage area if the subroutines are not used during the compi lation of the source program.
P.D.Q. SYMBOL TABLE PUNCH
This program, like the program for the U. T.O. version, requires that the
subroutines be used at the source deck compi lation phase, since the symbol table storage
area is also cleared if the subroutines are not used during this phase.
The original P. D. Q. system was developed with a symbol table punch r()utine
incorporated within the CLT2 processor. The routine that was used will punch the table
.
'
J
.
.
ina numer ic form I wh i ch I besides be i ng very d iffi cu It to read I destroys the area for .
•
j,,'
•
•
•
' .
.
.'
the program discussed here. The routi ne used by the P.o. Q. processor is governed by
switch two on the 1620 console. If switch two is off then the source program listing and
3 5 I~
2
\~,¥
the symbol table wi" be punched on cards.
C,"'_ }I.'
It is desirable to have the program listing
on cards because of the time saved, but it is also desirable to eliminate the symbol
table punch routine used in the original P 00. Q. system. If the IIbranch control switch
two" instruction in the symbol table routine is changed to an "unconditional branch"
then it will be possible to retain the program listing on cards and eliminate the symbol
table punch routine.
The above change would be necessary in the P.D.Q.-CLT2 processor only.
In the P.D.Q. documentation from the 1620 USERS GROUP LIBRARY, page 193, line
number 7552 is the instruction which reads tlBC2 EXIT." This instruction shou Id be
changed to read "B EX~T • II
OPERATING THE PROGRAMS
1. Compi Ie source program.
2. Place symbol table punch program in card reader and press LOAD
button on the reader.
3. Press START on the card punch.
4. Press START on the 1620 console when machine returns to a manual mode.
5. When punching is completed, remove cards from the card punch and list
on an I. B. M. 407 machine with an 80-80 board.
PROCESS ING TIME
The approximate computer time necessary to accomplish the symbol table
punch is given by the following equation:
Execution time = (A x B) + 20 seconds
where A = number of symbols
and B = card punch speed of the computer
355
0,"',
.
-"
3
ADAPTATION TO OTHER THAN 20 K MACHINES
It is necessary to change the instruction labeled uDOIT" in both programs
for machines having memory packages larger than 29K.
The instruction reads: IIDOIT TFM INFAD, 19999. II For a 40K machine
the Q portion must be changed to read "39999 11 , for 60K it would be "59999 11 ,
et cetera.
35 t»
4
c
."
I ..
GENERAL FLOW DIAGRAM
c'
SET ADDRESSES NOW
CLEAR
CARD AREA
HEADER
CARD
CHECK FOR NEGATIVE TYPE
MANTISSA AND CHARACTERISTIC. MOVE VALues TO
THE OUTPUT CARD AREA.
MOVE THE NEGATIVE SIGN
AND VALUE TO OUTPUT CARD
SUBTRACT SIX TO
MODIFY OBJECT
ADDRESS. MOVE
NO FORMAT NUMBER
AND 0 BJECT TO
OUTPUT CARD
AREA.
MOVE NAME OF
VARIABLE TO THE
OUTPUT CARD
AREA.
o
I
MOVE STATEMENT NUMBER
AND OBJECT DECK ADDRESS
TO OUTPUT CARD AREA.
~~--<..
CALCULATE THE DIMENSION
SIZES AND MOVE THEM TO
THE OUTPUT CARD AREA.
5
I.B.M., U.T.O. SYMBOL TABLE PUNCH PROGRAM LISTING
10
00020
00030
00040
50
6 J C;O
6000
6012
6,,)24
6() 36
6048
6060
6072
6J84
6096
S 1 ;..;8
I 612(;
6132
6144
6156
,.) 168
6180
6192
6200
62UO
6212
6224
16
26
26
15
14
46
11
49
16
44
26
26
26
26
39
~':<
JJ
S ,2 96
() 3 \i 8
33
32 c)
6 332
16
73
49
/'
-....... ;VL.:>O
S26(,
() 272
~ 20 i,
i~
:'·!t
~9999
07179 07174,
07184 C7174
87186
06042
06;)96
06042
06036
06:142
"062CO
07217
07251
07281
07317
07187
06843
49 G6036
73
12
44
44
33
16
73
12
4/+
62/, d
07174
90OOO
67345
01200
00081
aoooo
B7186
06043
07081
:':17111
07139
07169
00400
00000
00000
07195
07184
06352
06272
''07179
0717~
00000
OGOrO
07321
G7319
07184
G6332
0718 ..
07315
07313
\.:,6636
00061
0718 ...
0717 ..
0717 ...
00091
C718 ..
GeOCO
OOOrO
0718 ..
oU0 c) CJ
')
',~
E.,
,,
(J ..-
,
it
,
-
00060
00070
00080
90
100
00110
00120
00130
00140
150
00160
00170
180
00190
00200
00210
00220
00230
00240
00250
00260
00270
00280
00290
00300
00310
00320
00330
00340
00350
00360
00370
00380
00390
00400
00410
00420
00430
00440
004:'0
()0460
Ci0470
()O480
004')0
')
,.~
~~ ~~
*
*
~--it-
1f- ,'{-
SYHbOL TABLE PU~ICH PROGRAi'~ Fcr' THE 1~- ?r
I • B. ~·1. AND U.T.O. FORTf~AN PRO C:: S .5 I t~ G-:: ~~
-;(- r,
SYSTErv1S.
"><
~*'
;C' ji"
?(
-~t..1~
..if
*{{-
DEVeLOPED bY R.C. IRONS, u.s. i.,;'::' \/,~ L
SCHOOL OF ;',VIATI00i tv1EDICII~[, I'~.,"'.S.
PENSACOLA, FLORIDA.
"-,,
->*-
?r
{~- -)~
-l~
~~
** TO CHANGE THIS PROGRAf;; F"OR LAY,GcY, i,- ;\
f,* SIZE fviEtviOR YS IT IS ONLY i'~ECESSAf\Y
?~~~
-,..--,,TO CHANGe THE FIRST DIGIT OF-- F"iL:
Q FIELD
r
INSTRUCTIOi~ LAGLED i)U I T -;.,-,,**
*
*
~~- ~j~-
-l("
~-
/,"" r-
\j
,<-
~f-*
*1r
**
ir
DOLT
EL 11·1
LOCATICN
THE USER.
DOI~G
f'/;,t.,
Y DE CriA t'1(; EI) TO
5~IT., 00500 0
31 ooooa 01940
49 00000 0
34000700070136000700070234000560070136000560070249075000
1 08600 200 20000 1 08400 199 00100 Z
37 0250100100
LOG FILE WILL BE PUNCHED OUT.Z
37 02561 (J010e
RE'MOVF LOG CARDS, PUSH START Z
OOOOOQOOOOOOOOOOOOQO
1 05229 001 02718
WHERE TO START NEXT LOG ENTRY
1 05228 001 02900
UPDATED FIELD FOR ABOVE
1 05229 170 03000
LOAD AND UNLOAD LOG FILE
381
7
••
iI'
•••
•
.'
PLOT
B&G8nr..n·
LOAD
SIJB$£T
SSGQ,£.D I T"GO
~~!-;I' /~n'l T
. OlD OF .
JOB
SSGO
EDIT
fitOt1l TOR 1
•
.
"
.
3400032007013600032007024900500010520002900100
SUPERMONITOR COLD START
o 0:~~.~ a o···s 0 :" 0 0 0 ~ 'Ii ~i; 0 0 ~ ~2: 0 '2: 0 0 0 '2; '2i 0 3(;' ~ 0 ; 0 0 ~;; 0 0 ~:. 0:': 00 !! 0 BOO 0 0 0 ~ 00000000 UUODD 000.00000
12)'5'7 •• "nnUMR.u ••• ~nn~»anua~~n~~ •• n»u"~uu~U.U~U.~NUMUH~WHN~~U"""v."n~nnu~"n~~.
1 1 1 1 lIt 1 1 1 1.. 1 1 1 1 1 1 1 1 1 1 1 1 1'1 1 1 1 1 1 1· 1 1 1 1 1 1 1 1 1 1 .~ tIl tIl 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 lIT 1 1 1 1 1 1 1 1 J: 1 1
22222 2 ~'12 2 2 2 2 2 2222" 2 2 2 2 2 2 2 2 2 2 22222· 2 2 2 ~ 2222222 2 ~ 22222222222222222 '12 2 2 2 2 2 22222'22
:.: 3 3 3 3" 3 3 3 3 3 3 . 3 l 3 3:; 3 3 33 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 J 33 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 ~ 3 3 3 J: 3 ;r 3 3 3 ~: 3 3 r 3, 3 3 3 3 3 3-33
4~4444444444444444444444
444444444444444444444444~444~4444444444~:4444444444444 44
5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 S.: 5 5 5 5 5 " 5 5 5 5 5 5 5 5 5 5 5 55 5 5 5 ,: 5 5 $ J: 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 ~ 5 5 5 5 5
6 6 6 6 i &6 I" II i &' 6 6 &6 6 6 &66 &&&6 6 6 6 6 6 6 6 6 6 6 6 i &6 6 6·6 &6 6 s' 6 666 6 6': 6 &6 ~] 6 6 6 i~ 6 6 6 6 i 6 , 6 & &, 6 6 , I &•
i.
7 7 7 7 7 7 .,. 7 7: . 7 7 7 7 7 7 7 7 7 7 1 . 7 1 7 7 1 7 7 7 1 7 7 7 7 7 7 7 7 1 7 7 7 7 7 7 7 7 7 7 '~7 7 7 7 77 7 1 7 7 1 7 7 7 7 7 7 7 7 7 1 7 7 7 7 7 7 7 7
t.
88aala.118111888888aa8181.18888.ia88888888aS8888888888888888888888181.1 •• 1811111
1
.I' 9 9 9 9 9'99 9 9 9 9 I 9 9'99 9 9 9 9 9·9
-_.__ _"".
99 9 ~ 9 9 9 9'9 9 9 ••
9 9 9 9 9 9 ;; 9 9 99 9 9 9 9 9 9 9 :~ 9 9 9 :~ 99 :'; 9 9 9 9 9 9 99 9 -'9 9 ,';9 9 9 S S 9 9
I;
123'5171'~"U~M.M»MM~Dnn~uana~.~u~~
.
--
n.H~~UU~U~~.UH~N»M"~~~H.AaPM._v.".n.»M~NnnN"
....
.-
.;;w-~"
.....
I,
Fig. 2. SYSTEMS CARDS
.~ ·"-'--rD-r·iE~~t1Atf,-EDWARD-t:----·-,--·-S-SYSTE~1S' DE'~E[OPMENT:----~-"-"---"
r
~
C
c c :~
~
~
c:
M
,...
,..
...
r"'
~
c ::
r
...
eooOOOOOODO:OOOCODOCOOODOOODOOOOOO~OO=:::OO:DOOJOOOOOOO~OOOODOOOODOOODOODOOOOODO
'234"'I."nnQM"au.~~~nnu»unH~.~U~M~.n.Uq~~u~U.~ •• H~UUMUH~.Q.~ap"u"v""~nnnH~"nnnQ
1 1 1 1 1 1 1 1 1 1 J 1 Ie 1-1 1 11 Ie 11 I 1 1 1 1 1 1 11 1 1 1 I 1 1 1 1 1 1 1 1 1 1 1 1 11 11 1 1 1 1 1 1 1 I 1 1 1 '1 1 1 1 1 1 1 1 1 1 1 1 1 1 I 1 1
2222222Z222222.2222222222i222221222:22:2J222:222222222 2222222222222~2122222222222
3333333a3333333:33JJ33333~33333333333333:33333S33::333 33J333333333333333333333333
[444444C4444:44444[444:~444444444.444444444:44['44444J44~444444444444444444444444'
555551555=:'II::55~555555=55555555S555555S~55~IJ==5115JJ555~555555555555555555555
I , , , 1111'11:: '1111 , I =1111. , • , 11.1 , I , III , liS' I'S I 1.111 JIG &, && && 6 &6 6 6 6 &, 6 &, I , 6. 6 i 6 ':6 ,
177}}11J}}}J}1}1}17117}7}7J}7"}}7171}}117117)11}17:7777771777777177717771771777
•••••• " •• , •••••••••••••••
CI"I""""I'=IIIIIII"~:""".""'~'I""""'II' ,
99 S 9 S t : , I 9 • ,', I • 9 9 I • It: ~ I t 9 ,'•• S 9 I 9 9 ••• I • • II , • 9 t • t t • t I ,'11,' 9 • t •• IS. 9 • III. S • , S I. t s S I •
'z
3 4'1 II
.lonaaQM.1117 ••••• 23 ...... ZI ••• ~M ....... 4141G ............................ ., ••••••••• 1711 • •
_IO~I.....
Fls- 3_ ·'··OSlm .. m--WD·--·..-----····~--'-·-··. · . .· - -'--_.'-"
L
...
"
C
8
4· *
5·
Data Card (when required)
End Job Card
The FORMAT for the User ID card is as shown in Fig. 3.
C
The 1620 does not have a clock so a system to keep
track of the 'time was devised. To do this a Clock card containing the time and date are read in and entered into' the log
in the same manner as the ID card. The Format is identical to
that of an ID card except the user's name 1s replaced by PowerOn, Clock, 'or Power-Orr and the time and date replace the
use. See Fig. 4. One clock card for each hour of the day is
pre-punched and entered into the log, on the hour, while the
machine is in use,. This is accomplished by stacking them in
the read hopper along with the jobs being processed. When the
log is punched and. listed it ~ontains the time power was
turned on for the particular day and the names and application
of all users to the next hour. Tq.e .same information wi th
hourly time is repeated until the power is turned off. As was
stated previously,' the logta~le is automatically punche.d . out
when full. A Log card provides the capability to punch out
the log whenever it is desired.
The results of this log were amazing; on two different ,days 344 starts were logged. One such log is shown in
Fig. 5.
-EDIT
Edit 1s a modified version of IBM's FORTRAN with
FOR~~T PRECOMPILER.
It will detect many of the common
..
errors, particularly language, key punching and specifica-'
tion mistakes but it is not exhaustive.
Its suggested use is 'in conjunction with the SSGO
system which does not require Format. But it may be used
C'')
1',/
38J
9
.If''''
"",y'
firIf
••
If
•
•
•
•
•
oQ 00000000 0 ~ 0 000000000000000 O~' O~· 0 {) O~· 0 0 00' 000000000000000000000000000000'0000000000
~234$17 •• a"n»M~~Udg~~nn~~auan~~n»~»~v3U"~~U~U"~~UR~M~MHHnMU~~"UM •• VQ"ronn»H»nnnn"
1-11 1 1 1 1 1 1111 1 1 1.1 1111 1 11 1 1 11 tIll 1111 1 11 ': 1 . : 11 11 111 t 1_111-1 1111 111 111111111111 1 11"-11
222222222222222222222222222222222222222222222-22222222222222222222222222222222222
3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 : ' 3 3 : '3 -
3 3 .- 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
;: 4 4 4 4 4 4 ;~ 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 , ' 4 4 4 4 4 4 4 4 4 4 4 . 4 44 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4 4
555555555555:555.:555555555555555555555555555555555555555555555555555555555555555
6666666666::66.~66666666666666666,:66&66666
:i666666666666666666666666666666666&66
77711771707177777177777777777777771717711777717777777777777771777777177717777177
l'
~
::: 8 8 8 8 8 8 8 B8 8 8 8 8 aa8 8 8 8 8 8 8 aaa8 8 8 8 : : 8 . , a8 8 _~ 8 8 ~.: 8 8 8 8 aa8 88 a8 aa8 8 8 aaB8 8 8 8 8 8 8 aaaI 8 •• I 8 8 8 aaI
9lZJ451).,a"nqM.RU
~ 9 9 9 9~ 9 99999 C99 9 ••
99 '9 9999 9-i 9999999999999 9 9~9 f999 9 99 9 9 9 9 9 999999999999999 99I. 999' 999
HanD~nava~~~unu»~»»~~~UU~~U~QURRU~MH~~NUM~U~Mu~~."~nnnH~~nnnq
t~~. . . . ~ __.~,,'"-:- ___I!M~~_:. _ _~_...~._....
>'",
."''aj'~~CO': .'~"""""""''''''''''__"",,::'''~~.''- ~...,..~~~.. "
\-
Fig e' 4 e
CLOCK CARDS
384
-- --- --
-- ------_
....
C
DEPARTMENT OF CIVIL ENGTNEERING
COMPUTER LABORATORY
RUN lOG
HOURLY ENTRIES FOR THU.MAR.26,1964
TIME
PROJ.
TOTAL ENTRIES
-- ..
('"
C
----~-
12.01 AM -
1.00 AM
0
15
1.00 AM -
2.00 AM
0
14
2.00 AM
3.00 AM
0
15
3.00 AM -
4.00 AM
0
17
4.00 AM -
5.30 AM
0
21
8.30 AM - 10.00 AM
2
16
10.00 AM -
1.00 PM
0
12
1.00 PM -
2.00 PM
6
9
8.30 'AM -
10.00 AM
2
18
2.00 PM -
3.00 PM
20
21
3.00 PM -
4.00 PM
33
34
4.00 PM -
5.00 PM
28
32
5.00 PM -
6.00 PM
32
35
6.00 PM -
7.00 PM
12
15
,1.00 PM -
8.00 PM
30
30
8.00 PM -
9.00 PM
14
16
9.00 PM -
10.00 PM
22
25
10.00 PM - 12.45 AM
4
15
-----~-
TOTAL NUMBER OF ENTRIES
203
'Y=\ c::... S
--.....
-~-
...
...
----..-
385
344
¥-~~/O
. ...... .. ,................ .... ....
,
~.~
• • • _"
~
. _ •••••
~
••• " •••• " ........... M... " " " " .... ~~~". _ _ ~ •••
"'
1'./i
DEPARTMENT OF CIVIL ENG I NEER I ,,<~
COMPUTER LABORATORY
RUN lOG
;11(""-'"
POWER-ON
SUHRBIER. JOHN
GARDNER.D.
GARDNER.D.
GARDNER.D.
GARDNER.D.
GARDNER.D.
GARDNER.D.
GARDNER,D.
GARDNE'R,D.
GARDNER,D.
GARDNER,D.
GARDNER.D.
GARDNER.D.
GARDNFR,D.
GARDNER.D.
THU.MAR.26,1964 12.01
1
S
S
S
S
S
S
S
S
S
S
S
S
CLOCK
GARDNER.D.
GARDNER,D.
SUHRBtER. JOHN
GARDNER.D.
GARDNER,D.
GARDNER.D.
GARDNER.D.
SUHRBIER. JOHN
JESSIMAN
JESSIMAN
GARDNER.D.
SUHRBIER. JOHN
GARDNER.D.
SUHRBIER, JOHN
S
S
1
1
1
1
1
CLOCK
167~F
1673F
1673F
1
9407
G
G
9430
9430
1613F
9407
1673F
9407
1673F
1673F
1613F
1673F
S
5
S
S
S
S
S
S
S
S
S
S
S
S
S
1
"1""
I
\'·j~,.7
9407
THU.MAR.26,1964
1
1.00 AM
1673F
1673F
S
S
5
S
S
5
S
S
S
S
"'"---","
A.M
9407
1673F
1673F
1673F
1673F
1673F
1673F
1673F
1673F
1673F
1673F
1673F
THU.MAR.26.1964
CLOCK
SUHR8tER, JOHN
GARDNER.D.
GARDNER,D.
GARDNER.D.
GARDNER.D.
GARDNER.D.
GARDNER.D.
GARDNER.D.
GARDNER.D.
GARDNER,D.
GARDNER.D.
GARDNER.D.
GARDNER.D.
GARDNER.D.
GARDNER.D.
MURRAy.W.E.
S
S
S
(
2.00 AM
9407
1613F
1673F
1613F
1673F
1613F
1673F
1673F
1673F
1673F
1673F
1673F
1613F
1613F
1673F
"
386
S
THU.MAR.26.1964
C
3.00 AM
~.:(JS: II
..."
10(1..)
SUHRBIER, JOHN
GARONERtD.
GARDNFRtD.
GARDNF'RtD.
GARDNER,D.
GARONERtD.
GARONERtD.
MURRAY,W.E.
GARONER,O.
MURRAV.W.E.
GARDNER,D.
MURRAV,W.E.
GARONfR,D.
GARDNER,D.
GARDNER,D.
GARDNER,D.
GARDNERtD.
("
(~"I.
,""
C'
'
"'/
1
S
9407
1673F
1673F
1673F
1673F
1673F
1673F
1
S
S
S
S
S
S
S
S
1673F
1
S
S
1673F
1
S
S
S
S
S
1673F
1673F
1673F
1673F
1673F
S
CLOCK
THU.MAR.26.1964
GARDNER.D.
GARDNER,D.
GARONER,O.
GARDNER.O.
GARDNER.D.
GARDNER,D.
GA.RDNER .0.
GARDNER,D.
GARDNER.D.
GARDNER,O.
GARONER,D.
GARDNER,D.
MURRAV,W.E.
GARDNER.D.
GARDNER.O.
GARDNER.o.
GARDNER,O.
GARDNER,O.
GARDNER,D.
GARDNER.D.
GARDNER,O.
GARONFR,O.
S
S
S
S
S
S
S
S
S
S
S
S
1673F
1673F
1673F
1673F
1673F
1673F
1673F
1673F
1673F
1673F
1673F
1673F
1
S
S
S
S
S
S
S
S
S
S
4.00 AM
1673F
1673F
1673F
. 1673F
1673F
1673F
1673F
1673F
1673F
POWER-OFF
THU.MAR.26.1964
5.30 AM
POWER-ON
THU.MAR.26.1964
8.30 AM
KtM,JIN HWAN
KIM,JIN HWAN
KyM,JtN HWAN
NEWMAN. EDWARD E.
JOHNSON,WILLIAM F
JOHNSON.WILLIAM F
JOHNSON,WILLIAM F
JOHNSON.WILLIAM F
JOHNSON,WILLIAM F
JOHNSON,WILLIAM F
1
1
1
1
1
1
1
1
1
G
G
G
S
S
S
S
S
S
5
DSR 8790
DSR 8790
DSR 8790
DSR NO 9107
DSR
DSR
OSR
DSR
OSR
NO
NO
NO
NO
NO
9107
9107
9107
9107
9107
387
~.~S:/~
10(4)
CLOCK
THU.MAR.26,1964
FAIRBURN, DON
MUMFORD,J.
TELSON,M.
MUMFORD,J.
ALAM
TELSON,M.
BATOR RICHARD
KINGSNORTH, R C
GOLKA, JOHN J.
LEARY D F
LEET KENNETH
TAYLOR, WILLIAM A.
'.:
WANEK. GLENN
PROKOPY, JOHN C
WANEK, GLENN
NAINIS
LAZAR .. JAY L.
MOORER, JAMES A.
MOORER, JAMES A.
MOORER, JAMES A.
MOORER, JAMES A.
KINGSNORTH. R C
LEARY 0 F
BERNHARDT,L.J.
BERNHARDT,L.J.
GOLKA, JOHN J.
BATOR RICHARD
MOORER, JAMES A.
MOORER, JAMES A.
MOORER. JAMES A.
MOORER, JAMES A.
PROSEK, J. R.
SNYDER, M. B.
LAZAR, JAY L.
'~
1
1
18
1~O9
1
1
1
1.15
1
I
1~15
G
1~15
1
1.15
1.15
1.15
1.15
1.15
l.15
1.15
3
G
":':::1 C
2
I
I
1
PM
1'.15
1
1
1
1
',; 1" :' '.' ,,1.1-.5 .,' .:
"
",
.~.
~"'
'
".
':i
1.15
1.15
1.15
4
1
1
2 '. ,', ') ,."1".1'5,
1
1.15
1.1.5
1. .1.5
1. 1.5,
1
1
1
1
1
1
1
1
18
1.1.5
1.15
1".15'
1,.15
1.:1$
3
1."15:',
1.15
1
1
1
1
1
1.15'
1.' 15~
1.'15
1.15,
1."15'
1.15
3
4
2
1
CLOCK
THU. MAR .26.1964
KAVALLIEROS ANTONY 1
KAVALLIEROS ANTONY 1
NEWMAN, EDWARD E.
FAIRBURN DON
1
TENNISON. THOMAS J.
SIKES,P.G.
I
NAINIS
I
ERNST, D.
21
DEDRICK D F
16
CHAPMAN, D G
MATTES.D.A.
HUIE,J.L.
KITTREDGE
18
TENNISON. THOMAS':J ."';, ";
JOSEPH,GUNTER
LEARY 0 F
PHILLIPS, C.R.
VI
WANEK, GLENN
I
KINGSNORTH, R C
G
G
l-
C
G
'3~OO
4.00 PM '
r-~;2:7;
1'~:2,1,'
S
1.2.7
1 .. 15-
G
1
1
1
L.15,
1,.1-5'
1',.1,5
3
2
1
1
1
2
,'-,
,t;:, ,!.
G
1
4
1
1
1:.15
L~15:
11..15 :
1,.15 '
} ,~
1.15
:1 ~.1 ~"" .,', ,
1.15
1.15
1'.'15,"
1.15
1.15
389
------.- ...
-.-.---.-.-------~-----
."'---"-._--
"'
... ...
_
~~.~--.----~.~-"'-~"'~
......
~-~.~~~~
....
...........,
~
...........
,,~~
10(5)
KUBO. A. S.
SEXAUER,W.L.
SEXAUER.W.L.
SEXAUER,W.L •.
SEXAUER.W.L.
SEXAUER,W.L.
SEXAUER,W.L.
ALAM
MABRY R. E.
PHILLIPS, C.R.
DEDRICK 0 F
BERNHARDT,L.J.
BERNHARDT,L.J.
tA
G
1.1~
1
.1
1
1
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1
1
1
G
I
G
4
2
VI
16
1
1
CLOCK
THU.MAR.26.1964
NEWMAN. EDWARD E.
NEWMAN, EDWARD t.
COOPERBERG
NEWMAN, EDWARD E.
SEXAUER,W.L •.
MATTES,D.A.
JOSEPH,GUNTER·
WANEK. GLENN
SHWtMER, JOEL
GOLDMAN.F.W.
SEXAUER,W.L.
LEET KENNETH
HUIE.J.L.
SEXAUER.W.L.
SEXAlJER,W.L.
BERNHARDT.L.J.
JOSEPH,GUNTER
SEXAUER,W.L.
DEDRICK D F
DEDRICK JD F
GOlDMAN.F.W.
MATTES,D.A.
WANEK. GLENN
BERNHARDT, L.J.·
JOSEPH,GUNTER
JOSEPH,GUNTER
JOSEPH,GUNTER
GOLDMAN.F.W.
CHAPMAN', t') G
GRAASS J H
BERNHARDT,l.J.
CHAPMAN, D G
BERNHARDT,l.J.
CHAPMAN, D G
CHAPMAN. 0 G
5
,~
5.00 PM
S
7
S
1
1
I
1
G
1
1
1
1
G
1
1
1
1
G
16
16
I
1
2
2
1
1
1
1
G
14
OPERATOR
1 •. 15
1.15
1.15.
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
.1.15
1.15
1.15
1.15
1.'15
1.15
1.15
1.15
1.15
G
G
1.15
1
1.15
1.15
1.15
1.15
1.15
,1.15
1.15
1.15
1
4
1
1
1
1
1
1.~5
CLOCK
THU.MAR.26.1964
NEWMAN,E.
ROBINSON.J.V.
NEWMAN.E.
BERNHARDT.L.J.
BERNHARDT,l.J.
S
S
S
,1
1
1""
~i·l;-.J,",.l
6.00 PM
CI
9407
1.15
1.15
390
JI..2S:/5'
,otto)
C
CHAPMAN, 0 G
CHAPMAN. 0 G
CHAPMAN, 0 G
LEARY 0 F
LEARY 0 F
CHAPMAN, D G
CHAPMAN, 0 G.
CH~.PMAN,
D G
CHAPMAN. 0 G
('' 1
-.-J'
;
1
1
1
1
1
1
1
.
i
1.~5
1~15
l.iS
1.15
,
~
~ .1~
1~1~
..
LEARY 0 F
1
1
CLOCK
THU.MAR.26,1964
ENGER,TOM
ENGER.TOM
ENGER,TOM
LEARY 0 F
LEARY 0 F
CHAPMAN, D G
CHAPMAN, 0 G
CHAPMAN. 0 G
LEARY 0 F
LEARY 0 F
MOORER. JAMES A.
MOORER. JAMES A.
MOORER, JAMES A.
MOORER. JAMES A.
LEARY 0 F
BERNHARDT,l.J.
LEARY 0 F
CHAPMAN, o G
LEARY D F
CHAPMAN, o G
GOLKA. J. J.
LEARY 0 F
WILL I AM S ,IR E
WILLtAMS.RE
BERNHARDT.L.J.
CHAPMAN, 0 G
DESMOND,BARBARA A.
DESMOND,BARBARA A.
HAUSSLING H J
MOORER. JAMES
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I
SKODOH, JOHN G.
SKODOH, JOHN G.
ENGER.TOM
ENGER.TOM
ENGER ,TOM
DESMOND,BARBARA A.
CHAPMAN. D G
DESMOND.BARBARA A.
BERNHAROT.l.J.
GOLKA, J. J.
DESMONO.BARBARA A.
WIlLIAMS.RE
GOLKA. J. J.
1
1
1
1
CLOCK
0
1.15·
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
l.15
1. ~.5
1.15
l.15
1.15
1.,.15
l.15,
1.15
1.15
1.15
1.15
1.15
1.15'
1.15,
1.15
1.15
1.15
1.1.5
THU.MAR.26.1964
I
I
S
5
1
1
1.15
~,
1
1.15
1
1
1.15
1.15
1.15
1.15
1.15
1
1
1
8.00 PM
5011
5.01,1
1,.15
1.15.
1,.15
1.15
1
7.00 PM
391
~.~s:/"
-----_.._-.
__._--_.-.. _._.__...__.__. _ - . _ _._-_. __.._._--_ _-_._- ---_.._._ ...__ .-.,-_... ..• -.--.- .. ...
...
-
\0(')
CHAPMAN, o G
BERNHARDT,L.J.
HAUSSLtNG H J
1
1
1
1.15
1.15
1.15
CLOCK
THU.MAR.26,1964
ENGER.TOM
ENGER,TOM
SKODON. JOHN G.
SKODON. JOHN G.
ENGER,TOM
ENGER,TOM
ENGER,TOM
ENGER,TOM
ENGER,TOM
CHAMPY,J.A.
HAUSSLtNG H J
GOLKA, J. J.
HAUSSLING H J
CHAPMAN, D G.
GOLKA, J. J.
HAUSSLING H J
CHAPMAN, D G
BATOR RICHARD
BATOR RICHARD
GOLKA, J. J.
HAUSSLING H J
BATOR RICHARD
GOLKA, J. J.
GOLKA. J. J.
HAUSSLtNG H J
1
1
I
I
5
5
1
1
1
I
r
1.15
1
GTA
1
1
1.15
1.09
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1.15
1
t
18
:3
I
18
18
1
1
I
CLOCK
1
',./
9.00 PM
1.15
1.15
5011
5011
1.15
1.15
1.15
1
1
1
1
1
3
3
1
1
/f'-""
/f-~
\',.-y
THU.MAR.26,1964 10.00 PM
ENGER,TOM
GARDNER,D •.
SKODON, . JOHN G.
SKODON, JOHN G.
SKODON. JOHN G.
SKODOH, JOHN G.
HARMAN , JACK
HARMAN t JACK
HARMAN
JACK
HARMAN t JACK
HARMAN • JACK
HARMAN t JACK
LEARY 0 F
LEARY 0 F
LEARY 0 F
,
1
I
I
I
t
1
1
1
1
1
1
POWER-OFF
5
5
5
5
S
5
S
S
S
5
S
1
1
1
1.15
1673F
5011
5011
5011
5011
9407
9407
9407
9407
9407
9407
1.15
1.15
1.15
FRl.MAR.27,1964 12.45 AM
NUMBER OF ENTRIES 344
(-"',
392
'I. '<5'-/7
11
to check source programs in FORTRAN II-D. Since SSGO does
not require Format statements l ED.IT was moditied to by-pass'
this type ot statement.
Each statement in which an error is detected is
punched out together with a diagnostic report. It is a
usetul debugging techniq~e as ~ell as an aid in program
development. It is most usetul to novice programmers~'
It no errors are deteoted the message "NO ERRORS DETECTED"
is t-yped out. Shown in Fig. 6 is' a list ot the error
messages.
SSGO is a modified version of a LOAD and GO FORTRAN
whioh was based upon IBMt s first 16~O CARD FORTRAN WITHOUT
FORMAT. This is a one-pass operati~n with 3 options. Any,
all or none of the'se options may be exercised by inserting,
immediately preceding t~e Data oardl the foll'owing:
1.
: 2.
3.
*
Trace M to N
M is the number of the source statement where
tracing is to begin and N is the number ot the
source statement where tracing 1s to oease.
* Map.
Causes a storage map of the symbol table together
with the unused memory to be punched out.
.:10 Cards
The original version 'of this FORTRAN was modifiea
so that the compiler was located in the ,2ndModu/l e while the
compiled object program was stored in the first 20K of core,
thus conserving both time and cards. The Cards card was,
therefore, used to obtain a punched object deck. This of
course finds little use today sinoe it. is more expedient to
load on disk any program that might be needed later.
ERROR MESSAGES
l~
EDIT
UN~FFINED STATFMENT NUMAERS
UNREFEREN(FD .S TATEMFNT NUMBERS
UNACCFPTARLf FORM TO LEFT.OF = SIGN
MULTIPLE = SIGNS.
SUCCFSIVE OPERATION SYMBOLS OR OPERATORS
MISSING OPERATION SYMBOL OR OPERAND
RIGHT PARENTHESIS ENCOUNTERED BEFORE LEFT PARENTHESIS'
MISSING RIGHT PARENTHESIS
MIXED MODF FXPRFSSJON
NO VARIABLE TO LEFT OF = STGN
EXPONENTIATION OF A FIXED POINT VARIABLE OR CONSTANT
VARIABLE USED HAS NOT BEEN DEFINED
SURSCRIPTF.D VARIABLE HAS NOT APPEARED IN A DIMENSION STATEMENT
fRROR IN SUBSCRIPT
NOT ALL OF THE INDICES ARE LEGAL
SECOND INDfX IS LFSS THAN FIRST
THIRD INDEX' IS ILLEGAL
STATEMENT NUMBER NOT ACCEPTABLE OR IS MISSING
DO STATEMENT J S 'INCOMPLETE
DO LOOPS INCORRECTLY NESTED
DC LOOP TERMINATES ON TRANSFER STATEMENT
DECIMAL POINT'OMITTED FROM FLOATING POINT CONSTANT
DECIMAL EXPONENT FOLLOWING AN E IS INCORRECTLY EXPRESSED'
THE EXPONENT FOLLOWING AN E HAS BEEN OMITTED
FLOATING POINT NUMBER FOLLOWED BY CHARACTER NOT AN E
STATEMENT NUMRER GREATER THAN 9999
STATEMENT NUMBFR PREVIOUSLY DEFINED
U~NUMBERED CONTINUE
STATF~ENT (ANNOT BE ·REACHED BY PROGRAM
ERROR IN STATFMFNT NUMAFR(S)
MISSING COMMA AFTER RIGHT PARENTHESIS
INDEX IS ILLEGAL OR MISSING
NON-NUMERIC CHARACTER FOLLOWS RIGHT PARENTHESIS
LEFT PARENTHESIS DOES NOT FOLLOW WORD IF
NO EXPRESSION WITHIN PARENTHESIS
MISSPELLED OR UNACCEPTABLE NONARITHMETIC STATEMENT
STATEMENT CONTAINS UNACCEPTABLE CHARACTER
STATEMENT cONTAINS DFCIMAL POINT NOT IN A CONSTANT
nIMENSION STATEMENT IS IN INCORRECT FORM
UNNUMAERflJ FORMAl' STATEMFNT
INCOMPLETE' FORMAT STATEMENT
VARIABLE IN DIMENSION STATEMENT PREVIOUSLY DEFINED
FORMAT STATEMENT NUMBER MISSING OR INCORRECTLY STATED
FIRST OR LAST CHARACTER IN LIST IS NOT ACCEPTABLE'
THREE DIMENSIONS HAVE BEEN SPECIFIED. OR IN INCORRECT FORM
COMMA MISSING. OR LIST IS MISSING OR INVALID
SYMBOL TABLE FULL
FIXED POINT CONSTANT GREATER THAN 9999
FLOATING POINT CONSTANT OUT. OF RANGE
VARIABLE NAME LONGER THAN 5 ,CHARACTERS
39
f(
Fig. 6
n ,(",.,..
~ y
Martin J. Goldberg
Computer Applications Engineer
Advanced Engineering Section
Norden Division ot United Aircraft Corporation
Norwalk, Connecticut
.Presented at the Eastern Reg1ona1 1620
Users Group meeting in Washington D.C.,
'on May 7, 1964.
422
22 June 1964
x36017
C'
NETWORK ANALYSIS ON THE 1620
(Presented at the Eastern Regiona.l 1620 Users Group meeting
1n WaShington D.C., on May 7, 1964.)
by:
1l1art1n J. Goldberg
Computer Applications Engineer
Norden Division
.United Aircraft Corporation
Helen Street
Norwalk, Connecticut
. ABSTRACT
A computer program for analyzing electronic circuits on a 40K
1620 with 1311 disc has been developed under a cooperative agreement
between the Norden Division of United Aircraft Corporation and IBM.
(
".
•... '.'.
The Program is divided into four ·sections: language interpreter,
DC analysis, AC analysis, and trans1ent analysis. The analysis sections
are limited to linear circuits, although the transient-analysis section
can approximate some non-linearities. The language interpreter section
1nterprets input data on the circuit to be analyzed. The data is written in a user-oriented language so tha,t the Program can be used by individuals not fam1l1.ar with programming, and also so that man-machine
communication is more efficient. The primary input is an equivalent
circuit schematic. Input data cards, prepared from the schematic,
describe the circuit topology and parameter values.
From the above data, each analysis section sets up the proper
matrix equations describing the network and performs the desired
analysis.
The DC analysis section obtains the DC steady-state solution
for all node voltages and currents. In addition, there are several
optional routines such as worst-case analysis of node voltages, computation of standard deviations of node voltages, determination of
sensitiv1ties of node voltages to network parameters and the like.
The AC analysis section obtains the frequency response of the
network i'. e ., magnitude and phase of node voltages and of currents
versus frequency.
The transient analysis section obtains the time response of
node voltages and currents. It is capable of accepting time varying input sources and also has the capability of switching parameter
values from one state to another. This allows for simulation of
change of state of transistors, diodes, and other Bwitching elements.
423
4,2,'7·1
"
-2-
The Program h~a had a significant impact on circuit design
procedures at Norden. Engineers are relieved of routine calculations
and are left with more t:Lme to apply to creative tasks. A high
degree of confidence that a circuit will meet specifications 1s
obtained before manufacturing. This is especially important in
the case of 1ntegr~ted circuits. More reliable circuits are obtained
since statistical analy~eB which used to be too time-consum~ng to
perform at all now are done on the computer. In addition, overdesign is reduced since the designer can readily see the effect of
reduced toleranoes on perfo]wance.
rfI-~
',-_,.,v
Norden has already ex.tended the capability of the Program
beyond that of the version to ,be released by IBM" and is \'1Torklng
on further extensions such as adding a non-linear capab1lity.
/1'"-"',
I
\~).?
424
c
-3-
1. INTHODUCTION:
This report will discuss the main features of the Network
Analysis Program developed JOintly by the Norden Division of
the United Aircraft Corporation, and I.B,M. Details of operating
prooedures and the mathematical theory will not be dealt wlth at
length. The objective here is to describe the usefulness and
significance of the program.
A specific sample circuit will be discussed from preparation
of input data to obtaining the output from the computer.
The Network Analysis Program is used as a tool by the designer
of electronio circuits. It can also be used to review the design
of existing circuits for reliablity or optimization.
The Program i6 written in FORTRAN II for a 40K, 1620 with
1311 Disk Storage Drive, 1622 Card Read Punch, Automatic Divide
and Indirect Addressing. The Program also requires the Monitor I
System. :
2.
Advantages of Network Ana1Yeis Program.
Figure 1 lists some of the advantae;es of the Program over
conventional design techniques of hand calculations, breadboarding
and testing. Points land 2 (figure 1) are obvious advantages.
Point 3 emphasizes that conventional techniques cannot give accurate
statistical information on the effects of component variations on
circuit performance. Because the designer 'does not have a high
degree of confidence that the circuit will perform satisfactorily
under all extremes of component variation and environment, he is
forced t6 over-design.
The Network AnalYSis Program can perform complicated statistical
analyses on the circuit, analyses much too time consuming to perform
by hand calculation. With this tool the designer can optimize·the
oircuits1noe the program readily shows the effects of tolerances
on performance.
Alternative c1rcuit configurations (point 4) oan easily be
studied on the computer and the beet configuration c.hosen. In fact,
designs incorporat1ng new or expensive oomponents can be studied
even .1f the actual components are not.available for breadboarding.
0",
It is often possible to .obta.in .insi~h. t i. n.. t.. o the. OP.·· eratlon of
a circuit by use of the Program (point 5). Alao, the effects of
parameters, such as transistor current gain, can be determined.
On a breadboard one cannot vary the gain of a transistor direotly.
"
425
1. ROUTINE CALCULATIONS ELIMINATED
2. MORE TIME FOR CREATIVE DESIGN
3. EFFECTS OF COMPONENT VARIATIONS
t
,
.t::-
4. ALTERNATIVE DESIGN STUDIED
5. INSIGHT INTO COMPLEX INTERACTIONS
6. BREADBOARDING & TESTING REDUCED
f'
(':)
-:=r
+
Figure 1
~
N
<:iJ
~
~< ~\
~
j
fF~
\t -/
-5The factors mentioned contribute to reduc1ng the need for breadboarding (point 6), and even when breadboarding is still required the
amount ot testing is reduced.
3.
Br1ef Description of Analysis Proced\1re
Before discussing ~he details of the Program, 1t 1s appropriate
to obta1n an overall picture of the procedure and flow ot information
involved in analyzing a circuit by use of\the Program. Figure 2 1s a
flow-ohart describing the procedure.
From the specifications and requirements on the circuit, the designer develops a oircu1t schemat1c describing thecontigurat10n he
plans to use. An equivalent circuit schematic is then prepared trom
the original cirouit schematic. The equivalent oircuit schematio
oonta1ns equivalent circuit representations for transistors and diodes.
The equivalent circuit of a transistor or diode can be thought of as
the model used by the Program to simulate them. The Program will accept
any equivalent cirouit chosen by the designer.
The input data cards are prepared by referring to the equivalent
schematic, using a user-oriented language to describe the
circuit.
circ~it
The Network Analysis Program is divided into tour main sections:
language interpreter". DC analysis, AC analysis, and transient analysis.
The language interpreter translates the input data cards, setting up
lists of 1nform,atlon for the analy.s1,s sections. It then calls on the
proper analysis section as specified by the user, a.nd ,the results of
the analysis are computed.
.
The user-oriented language used to prepare the input Data cards
is probably one of the most 1mportant features of the program. It
allows the Program to be used. by engineers not familial" with the machine
or with programming. Also, it allows easy and tastoommunioatlon between the c1rcuit designer and the oomputer. Very l1ttle of the full
potential of the Program would be realized wlthoutthe oapab1l1t1es
prov1ded by the u·ser-or1ented language.
4.
Basic Network Branch Defination
The bas1c entity 1n the matrix approach used to set up and solve
the matrix equations of the network 1s the network branoh. Figure 3
shows the basic branch with a de1'ln1t1'on of the important variables.
As shown" the branch 1s composed of three network elements: a 'passive
element (resistor, capacito)',) or inductor); voltage source; and current
source. ,The terminat10n 'po~ints of the branch are defined as ,nodes. We
may then imagine a network as composed of many branches.
427
~'.'''''''''''''''''''"'''''''''''''~.<''.''''''''''''''''~'''~'"''''"'.'''''''''>''.'
.• ·.... ·,..
~,'ft
...., _ _ _ ,............................. _~ ................ _ _ _ __
o
J
n
~
'0 !
Y
=
CONDUCTAN€£ Of PASSIVE ELEMENT
E = VOLTAGE SOURCE
I
---::l
I = CURRENT SOURCE
I
i = BRANCH CURRENT
e = BRANCH VOLTAGE
el
~
NODE TO DATUM VOLTAGE
BRANCH DATA CARD:
Norden
U
DIVISION OF UNITED AIRCRAFT CORPORATION
R
Figure 3
:F
~
...:t
...:J.
~
N
C7.~
-I
-8-
Then the variables defined in figure 3 may be thought of aa vectors
or column matrices with elemcntp correaponding to each branch of the
network. The primary output of each of the analysis sections 1s the
node voltage vector, e', and the branch current vector, i.
The arrows on the diagram (figure 3) define the positive convention for current flow and the positive convention for the voltage
source with polarity as showh. This convention should be kept in
mind until later when preparation of input data cards for a specific
example 1s described •.
Also, shown in rlgure 3 1s a typical branch data card prepared
in the ~ser-oriented language. It contains all necessary information
on a branch. The Bl indicates the data is for branch number 1. The
N = (0,1) indicates that the branch is connected to nodes 0 and 1
and that the assumed direction of current is from node 0 to node 1.
Then the values for the three branch elements are given.
5.
Nodal Eg,uations of NetvI0rl{*
Figure !~ shows the nodal equations which are Bet up automatically
by the program. 1'hese same equations would be Bet up by using the
oonventional manual technique of Bumming the currents at each node
using Kirchhoff's current law.
/" -,
I
'"" ..
The only item which has not been defined 1s matrix A. It is
Bufficient for our purposes to say that it defines the topology of
the network, that is, which nodes each branch is connected to and
the assumed current direction for the branch. It is developed from
the nodal data supplied for each branch.
The e t vector of the node voltages is solved for. This is done
by inverting the triple product (At YA) and multiplying it timeB the
right hand side of the equations, where the superscript t indicates
the transpose of A. Thin gives a nominal solution for the node voltages. New solutions corresponding to a change in value of one or
more parameters are obtained by a very fasE update technique. This
1s a method of modifying the inverse of (A YA) rather than starting
the solution over from the beginning. An example of the use of this
r,eohnique will be shown later.
6.
Preparation of Input
l~f~~nple
Problem.
A specific example will help to clarify the preceding discussion
and should nil()\'l the relative ease with which a problem can be Bet up
for analyn1s by use of the user-oriented language.
* Technical Report TR 00.855, March 30, 1962. Machine Analysis of
Networks and Its Applications. Franklin H. Branin, Jr. IBM Data
Systems Division Poughkeepsie, New York.
42U
c,~J
-9-
430
-10-
Figure 5 is the schematic of a one transistor amplifier. From the
circuit schematic the equivalent circuit schematic (figure 6) is prepared.
The equivalent circuit of the transist()l't is enclosed within the dotted
lines. Numbers enclosed 1n squareesre branch numbers and the circled
labels are node deBl~n. atlons,. The arrows indicate assumed current dl:£!ections. The term G(4,3) indioates tha. t .there is a transconductance
from branch 4 to branch 3 •. ·The transconductance is used to model the
gain of the transistor and Is related to· the ourrent gain.
The language interprete~ nas:a dual function of interpreting specific data on the circuit and also interpreting certain control funtions
such as whichanalysls ieto be performed, which optional outputs are to
be obtained and which 1nput-output·devices are to be used. Thus there
are two types otlnput:. control statements and data statements.
The input·oard,,·requ1red to perform a DC analysis of theeample
circuit are shol'ln1nfigure 6 .•. :The first oard, DCNODE, is an example
of a contrql.stateinent and indicates that a DC analysis·ie to.be performed. The next card 1s an example of a data statement. The Bl indicates
that the data is foJ;' branch number one of the circuit. The N = (0,1)
is nodal data indicating that the branoh is connec.ted between nodes 0
and 1 and since node 0 is mentioned first, the assumed direction of
ourrent flow is from node 0 to node 1. The R = 1. (.01) indicates
tha·t the nominal value of resistance is 1 ohm and that the tolerance
is 1 percent. Thetoleral1ce is specified as a decimal percent tolerance enclosed 1n parenthesls •. ~he E = -30. indicates that there is
a 30 volt voltage'source in the branch and it 1s negative according
to the basic branch definition of section 4. Data on the other branches
is prepared in a s1milarma.nner. Note that the resistance of. branch
3 is expressed as R~5.E + 06 wh16h is equivalent to5. x 106 ohms.
~,
~IY
Skipping to the card~abeled Tlwe have another type of d~ta.
statement for preserttatlonofdata ontranslstor number one, The
B = (4,,) meane that the tran~donductance'lB from branch 4 to branch
3. Next, the valu~ ot that~anBconductance is given. It has a nominal value of .0364. Tqe tolerances are specif1edin an optional. form
in which the minimum 18.0361 and the maximum 1s .0367.
,.
.
..
Figure 7 shows ,addition~lcontrolBtatements and is a continuation
of figure 6. The SENSITIVITY statement indicates that an optional output
consisting of the sensitivities of each node in the circuit with respect
to each parameter is to be computed.' These outputs a.re discussed further
in section 7. The POWER control statement asks for the power dissipation
in each branch tobecompu·t·ed •.. The STATISTICAL ANALYSIS statement calls
on a worst-case analysis 'and.computat1on of the standard deviation of
eaoh node voltage. The MODIFY .control stat.ement and the statement immediately following itlnd,lcate·tt)at the resistance 1n branch 7 is to be
modified from 35K to.40K.1n'5equal step~.
43]
C
!
i#brltf.ffri"$ """ dt" t"· " II
" "Ft" 8H +BrW±
j" • . rirlitldJ"
-
III.
-11-
c:
~'
C
/
432
DCNODE
N = (0,1)
!,
N = (1,2)
I
N = (2,3)
N
Tl
= (3,4)
f
~
~
r=»
E = 30.
I
R:::: 2000.{.10)
R == 5.E
R:: 27.
+ 06
I
, R
=
1000.(.10)
N :: {3,0}
I
R
=
10000.(.10)
N = (1,3)
I
R = 39.E3U 0)
I
G = .0364(.0361
B :::: (4,3)
,
I:: -.01 E - 06
I
I
.......
I\)
I
E = -.6
N == (1,0)
Norden
~
R= 1.(.01)
I
.0367)
U
DIVISiON ()pr UNITED .... "CRAFT CORPORATION
A
Figure 6
~
~
~
c
f-~~
'\
.)-
r~~
\', )
tiE
tn t
ttll
1}-
()
Norden
-14-
It is implied that the node voltages at each step are tobe computed.
These new solutions are obtained at great speed by making use of
the updatetechn1que dj.Elcussed in Sect10n 5. The EXECUTE statement
causes all previous control statement~ to be executed in the order
1n wh1ch they appeared. The END statement returns oontrol to the
monitor system in the 1620.
From the above discussion it should be evident that preparation
of in~ut is relatively simple using the user-oriented language. In
fact the rules tor prepal't1ng the input cards from the equivalent
oircuit schematic can eas.11y be taught to an aid or keypunch operator.
7.
Discussion Of Output
or
Sample Problem.
Figure 8 shows the output'of the node voltages and the sensitivities for the sample problem. There are four nodes -in the problem,
. not countlngthe ground or dat;um node.' The values of each node .
voltage are outputed from left to right.
The sensltiv1tles sre ,the percent change in each node voltage
with respect to a positive one percent change in each parameter of
the network. 'rhe first set of sensi t1 vi ties are with respect to
the resistance in branch 1. They read from left tc;> right for nodes 1
through 4 respectively. The sensi ti vi ties with respect- to the other
parameters such aa transc~ndue'tanceB, voltage and current sources are
outputed in a similar fOl'1Jl1.· .'rhe sensi ti vi ties are used to detect
critical parameters in the network. For example, scanning the first
column of sensitivities one can see that node 1 is most sensitive
- td the voltage source, El. The sensitivities are also used in the'
worst-case and standard deviation computation.
The method used in obtaining the pal~tial derivatives, from which
the sensitivities are computed, is a relatively efficient method.
They are obtained from explic:lt matrix fo:rmulas obtained by taking
the partial derivat1~es-6f_ the nodal equatioris described in Section
5. This is a much faster technique than some others which require
obtaining new solutions for the node voltages with the parameters
increased by small amounts.
.
Figure 9 shows the res\llts of a stress analysis which was called
for by the POWER control statement. Here the current, voltage and
dissipated power for each branch are outputed. Also, shown are the
results of the worst-case and standard deviatlon(l-sigma} computations
for each node voltage. Thus, the nominal, W01'st-caae maximum and
minimum and standard deviati.:>n for each node voltage is listed. These
are based on the tolerances of theparametera and the computed sensitivities. FI'om this output is 1s evident that the designer, can readily
see the effects ot component tolerances on the circuit_and can optimize
the circuit ty selecting the proper tolerances.
435
t# tf1b' "triM""'·;!
I
t
WF't""'UIf'"
-15':"
NO. B~ANCHES =
NO. NODES = 4
1
NOMINAL ~ODE TO DATUM VOLTAGES
2.999E+Ol
2.062E+Ol
5.493E-OO
4.164E-OO
SENSITIVITIES
PERCENT CHANGE IN NODE VOLTAGES FOR A ONE PERCENT CHANGE IN PARAMETERS
R 1
1.170E-04
-l.683E-04
-1.141E-04
-l.961E-04
4.690E-Ol
-4.530E-Ol
-2.385E-03
,-2.611E-03
1.112E-O~
3.493E'-01)
-S.962E~OO
-6.693E-OO
3.623E-06
1.O!>GE-C; ;~
2.809E-03
-2.313E-02
1.342E-04
3.891E-Ol
1.040e'-Ol
1.431E-Ol
1.192E-04
-3.564£=01
6.993E-Ol
'l.SS0E-Ol
1.572E-04
4, 3
1.l11E-03
4.076E-()1
-1.991E-Ol
'-8.978E-Ol
-3.491E-OO
5.958E-OO
6.689E-OO
9.999E-Ol
9~509E-Ol
9.868E-Ol
1.107E .... OO
R 2
R 3
R 4
C"
R S
R 6
R 1
G
't
E 1
E 4
If·
1.690E-OS
4. 90GE'~OZ
1.310E-Q2
-1.079E-Ol
2.S02E-09
-7. 4~~6t>.. O6
1.212E-OS
1.428E-05
3
Figure 8
436
-16rr-'"
\~'"/-
STRESS
BRANCH
1
2
3
4
5
6
7
A~'ALVSIS
CURRtNT
.5312E-02
.4685E-02
.4685E-02
.4764E-02
.4764E-02
.5493E-03
.6282E-03
VOLTAGE
-.2999E+02
.9371E+Ol
.1513E+02
• 7286E .... OO
.4764E+01
.5493E+01
.2450E+02
POWER
.1593E-OO
.4391E-Ol
.• 7089E-01
.3471E-02
.2270E-01
.3017E-02
.1539E-Ol
/f"'"".
WORST CASE
NODE
1·
2
3
4
At~ALYSIS
NOMINAL
.2999t+02
.2062E+02
.5493E+Ol
.4764E+Ol
i
AND STANDARD DEVIATION
MAX
.3299E+02
.2649E+02
.7187E+Ol
.6426E+Ol
Figure 9.
MIN
.2699E+02
.1475E+02
.3799E+Ol
.3102E+Ol
"'--"
SIGMA
.9998E-OO
.8795E-OO
.2809E-OO
.2739E-OO
437
C· :
!
w"
r""
"t"
n"
"td,.tfMM
II
f
*
f
tnt
'"ft H" Hri#tttMftd W"hHj*""ftH&frti+±ith*"i6fifiAiti£*:ft9 ""i'frl"t;-#itiPdt£&bdtW;"tbWS¥fdW""
"fflNNtiift"w±Pifw"" p-wJ"!! ["""TfI"'pn-W''''li'!\llwmf,'rTT"'P'!"R"" "'1'"I"U
-17-
NODE VOLT AG:t:S f'OR MO.D I F tED 1 NPUT DA T A .
R 7 = 3. 500E+'04
NODE VOL TAGE.S
2.999E+Ol
1.968£+01
5.985£-00
5.234E-OO
1.993E+ol
5.853£-00
5.105E-OO
2.017E+Ol
5.728E-OO
4.982E-OO
2.040£+01
5.608E-OO
4.864E-OO
2.062E+Ol
5.493E-OO
4.752£-00
2.083£+01
5.382E-OO
4.644E-OO
R 7 = 3.600E+04
NODE VOLTAGES.
2.999E+Ol
R 7
~
3.700E+'04
NODE VOLT AGE.S
2.999E+Ol
=
R 7
3.800E+04
NODE VOLTAGES
2.999E+Ol
R 7
= 3.900E+G4
NODE VOLTAGES
2.999E+Ol
R 7 = 4.
E+U4
NODE VOLTAGES
2.999E+Ol
-18Figure 10 shows t'he output obtained as a result of the MODIFY
control statement. Sets of node voltages for each value of R7 are
obtained. The five solutions are obtained in about the time it takes
to get one nominal solution, since the high speed update technique
is used to modify the original nominal solution.
~
,~
8.
Discussion of AC and Transient Analyses
The AC and transient analyses will not be discussed in as
much detail ·as the sample DC problem. The outputs obtained from these
analyses are shown graphically in figure 11. In the AC analysis the
primary output is the frequency response of a network. This consists
of the magnitude and phase of each of the node voltages as a function
of frequency. The input is prepared', :in a similar'··manner to the DC
analysis, except that additional input such as values of capacitors,
inductors and mutual inductances must be given."",'Also, the· range of
frequencies over which the output is to be obtained is specified in
a Dlanner similar to the method used ~n the sample problem for modifying
a resistance.
The primary output of the transient analysis is the time response
ot the node voltages of the network. The transient'analysis is
capable of handling time varying sou:rceawhich may be inputs to a
circuit. Thus, input waveforems such as square waves, sawtooths, sine
waves, etc., can be handled. A certain amount olnon-linear capability exists in that switching of parameters from tone value to another
1s possible. Therefore, change of state of transistors and diodes
can be simulated. The switching is accomplished automatically and
the times at which switching occurs can be made dependedt on computed
~~
quantities such as a certain node voltage. For example, switching,,,"~
ot ~ertain parameter values from one value to another can be made
to occur when a specified node reach~s, say, 10 volts.
9.
Pro~ram
Extentions
Norden has extended the Network Analyeis Program beyond the vv
version to 'be released by IBM. Some' of these extensions are shown
in figure 12.
The Failure Effect Analysis searches for failures of components
whioh may be over-stressed when certain other components fail. For
example, each resistor in a network 1s alternately shorted and opened
and the corresponding stress levels in the other components are checked
to see that they do not exceed recommended values beyond which a
failure may occur.
The Derating Analysis compares the stress levels computed in
the stress Analysis (described in the sample problem) with recommended
rated values which are stored for each standard part on the 1311 disk
file. Percent operating to recommended and percent operating to
rated stress are then computed. Aleo, the failure :nate of each
component and tor the entire circuit is computed.
439
C
-19":
c:\
o
44u
I
-
,
-20-
EQUIVALENT CIRCUIT STORAGE
Norden
U
R
DIVISION OF UNITED AIRCRAFT C;ORPOIIIIATION
441
Figure 12
c
-21The Degradation Analysis accounts for the effects of aging or
some other criterion on a circuit. For example, the nominal tolerances
on the parameters are adjusted to account for degradation with aging
and the worst-case and standard deviations are computed with the
adjusted tolerances.
One of the disadvantages of the program as it presently exists
is the necessity of preparing an equlvalent circuit schematic including the equivalent circuit models of the transistors and diodes.
Norden 1s extending the Program to include storage of the equivalent
circuit so that the user need only specify the type transistor or
diode he is using. The necessary data would then be pulled automatically from storage. This will greatly facilitate the setting up
of a problem for the computer.
One of the most important limitations of the Program 1s the
lack of non~linear capability. This problem makes itself evident
in cases where it is necessary to account for the non-linear characteristics of transistors and diodes. At present, the user-must
select an operating point for a transistor from which the parameter
values for the equivalent circuit model are selected. If after
analysis, the operating point does not burn out to be the same as
that which was expected, the model will not be accurate. This
problem also occurs in the transient case when transistors change
state and it may be necessary to account for the non-linear transition, Bay, from the off state through the active to the saturate,i
state. Norden is working on extending the program to'account for
the above non-linearities.
C:
.
"
I
I.
~
l .)
i-
AN EXPERIMENTAL PERSONALIZED
ARRAY TRANSLATOR SYSTEM
H. Hellerman
Watson Research Center
1111 Corporation
4~3
c
iHii. .W'W·blJifi,· 9'6.·· ··..-¥···trlifflt&t+··LW· .. ·-1;w,;;. . 8··r '·j·WMtlilWill··
-
NffI"!fI"!i"·8DlilI!jIIfill i iP'iI"Ie'Hl"fffiJ'Qf"W:i!hi.·O'"m'
CI
AN EXPERIMENTAL PERSONALIZED ARRAY TRANSLATOR SYSTEM
by
H. Hellerman
IBM Watson Research Center
Yorktown Heights, New York
ABSTRACT: A system de.igned for intimate man-machine interaction
in a general purpose problem- solving environment is experimentally
operational. The system utilizes an array-oriented symbolic source
language containing powerful statement types. These include Numeric,
Boolean, Relational, and Selection operators on operands which can be
entire arrays. The system also permits simple specification of test
and argument arrays in single statements.
The completely symbolic operating system includes display and entry
of program and data. Sequence control is aided by an interrupt switch
which allows the user to interact with the program during execution.
In addition to normal stored program sequencing, the system provides
trace options and the ability to enter any statement for immediate
execution.
Present implementation of the system i8 with an interpretive translator
on an IBM 1620 computer.
Research Paper
RC-1091
December 19. 1963
(
.--
DEBUGGING IN THE FOR II SYSTEM
bY'
R. D. Burgees
Mechanical Technology, he.
445
Debugging in the FOR II System
This paper briefly describes the 1620 FOR II programming system.
Machine
language features pertinent to FOR II, difference between FORTRAN with FORMAT
and FOR II, indexing in FOR II, subroutine linkage and construction of the
object deck are covered.
Stress is placed on debugging techniques such as use
of the Symbol Table, Instruction Register No.2, and general debugging aids.
Knowledge of FORTRAN with FORMAT is assumed.
c
446
.~--
----_._.,.. '", ..• _..."".,.
SECTION,
APPENDICES
Table of Contents
Registration List. • • • • • • • • • • • • • • • • • • 5.1.1
List of Programs Teams Chairmen • • • • • • • • • • • • 5.2.1
Report of Meeting of Programs Teams Chairmen • • • • • 5.3.1
Survey of Type 1620 Programming Systems • • • • • • • •
S.h.l
447
5.0.0
G
EASTERN REGION 1620 USERS GROUP
WASHINGTON, D. C., MAY 6 - 8, 1964
REGISTRATION
ALMAN, John E.
Boston University
700 Commonwealth Ave.
Boston, Mass.
BEST, A. H.
The Glidden Co.
P. O. Box 389
Jacksonville, Fla.
ANDERSON, Jay Martin
Bryn Mawr & Haverford College
Bryn Mawr, Pa.
BICKFORD, Paul A.
O.U, Medical Research
800 N.E. 13th Street
Oklahoma City, Okla.
ARENA, Angelo
IBM
White Plains, N. Y.
BLAI CHER , He rh
Jersey Central - N.J. P & C
Madison Ave.
Morris Plains, N. J.
ARNOLD" Stewart
J.M. Patterson State
P.O. Box 6199
Montgomery, Ala
BLADES, Co Lo
Baltimore Gas and Electric
Baltimore 3, Md.
AUSTIN, Milton T.
Mauch1y Associates, Inc.
P.O. Box 475
Ft. Washington, Pa
BOCKHOFF, J. W.
Gen. Am. Trans. Co.
2501 N. Natchez Avenue
Chicago, Ill.
BABER, Gaye M.
Na ti onal Ed. Assoc eo.
1201 16th St., N. W.
Washington, D. C.
BOLLACASA, Dario
General Electric
P. O. Box 1072
Schenectady, N. Y.
BAILEY J C.
IBM Dept. 808
P. B. #6
Endicott~ N. Y.
BONNETTE, Mrs. Robert W.
Norwich University
Northfield, vt.
BAKER, A. DO'yle
Kentucky Utilities
120 South Limestone
Lexington, Ky.
BRADY, Kenneth W.
Long Island Lgt. Co.
250 Old Country Road
Mineola, Long Island, N. Yo
BANTA, E. S.
Wesley College
Dover, Dela.
BRAUN, Carol A.
Harrison Radiator Div. GMC
1400 N. Goodman St.
Rochester, N. Y.
BERGER, Kenneth R.
National Ed. Assoc.
1201 16th st., N. W.
Washington, D. C.
c
BRIDGEMAN, Keene
General Dynamic/Electronics
Rochester, N. Y.
BECK, W. H.
Baltimore Gas & Electric
Baltimore 3, Maryland
BRIESEMEISTER, R. E.
Consolidated Edison
4 Irving Place
New York, New York
5.1.1.
BRIGHT, Harold F.
George Washington Univ.
Washington, D. C.
CASLIN, James C.
US Air Force
318 Whitmore Ave.
Dayton, ,Ohio"
BRITTAIN, MAl USN, Paul
U. S. Naval Academy
Annapolis, Maryland
CHANCEY, Harold R.
Hercules Powder Company
Kenvil, New Jersey
BROWN, Charles C.
Paper Mfg. Co.
Roanoke Rapids, N. C"
Alb~marle
CHAPUT, Robert L,.
Comm. of Mass
100 Nashua St.
Boston, Mass.
BRZOZOWSKI, w.
Comp-Tron Co.
1035 N. Calvert St.
Baltimore, Md.
CHURCH" John O.
General Foods Corp.
555 S. Broadway'
Tarrytown, N. Y.
BURGESS, Robert D.
Mechanical Tech. Inc.
968 Albany Shaker Road
Latham, New York
CIPOLLA, Jo Ann
E. I. DuPont de Nemours & Co.
Wilmington, Dela.
BURLEI.GH, W. F.
Tidewater Oil Co.
Delaware City, Dela.
CLARK, Stanley A.
Public Service Co. of N. H.
1087 Elm Street
Manchester, New Hampshire
BURROWS, W. Ao
Dravo Corp.
Pittsburgh, Pat
COOK, Leroy L.
U.,S. Public Health Service
1901 Chapman Ave.
Rockville, Maryland
BUSCH-PETERSEN, Bent
Houdry Process & Chern. Co.
1339 Chestnut St.
Philadelphia, Pat
COOPER"Richard A.
Riley Stoker Corp.
9 Neponset st.
Wor\!ester, Mass.
CALLAHAN , E. R.
E. I. Dupont
Textile Fibers Dept.
Chattanooga, Tenn.
COX, Ed,
Memphis Light Gas & Water
P. O. Box 388
Memphis, Tenn.
CANNER, Arnold K,
Transitron Electronic Corp.
168 Albion Street
Wakefield, Mass.
CROMWELL, John E.
Sun Pipe Line Co.
1608 Walnut Street
Philadelphia, Pat
CARLSON, C. Eo
New York State Dept. of
·Public Works
Albany 5, New York
DABE, Rodney
Consoer Townsend
360 E. Grand Ave.
Chicago, Ill.
CARR, Wendell
US Pub. Health Service
1901 Chapman Ave.
Rockville, Maryland
DAVIDSON, C. H.
Univ. of Wisconsin
Madison, Wisconsin
5.1.2
,----.-----""",'-"
-------".--.-~~---------
c
"/"W""Nf:iHtfrlrliittli'F'riif H*
"" 'b"'&ii"" Wf""#bH*WWftlffLWiiWWIi4i!litPWg'W"PPHlI"P"HliI'OC.ti"Uf!W"i!!W
DAVIDSON, E. A.
IBM
425 Park Ave.
New York, New York
EVANS, John H.
Southwestern Public Service
Box 1261
Amarillo, Texas
DAVIDSON, James N.
Long Island Lighting Co.
Hicksville, New York
FALES, Henry N.
General Electric
100 Plastics Ave.
Pittsfield, Mass.
DAVIES, R. W.
Baltimore Gas & Electric
Gas and Electric Building
Baltimore 3, Maryland
FARINA, Mario
General Electric
600 Main St.
Johnson City, N. Y.
DAVIS, Russell H.
State Roads Comma of Md.
300 W. Preston Street
Baltimore, Maryland
DELHOMMER, Harold
Petroleum Industrial Computer Center
132 Demanade Blvd.
Lafayette, Louisiana
DICKINSON, Frank N.
Biometrical Services - USDA
Beltsville, Maryland
DOUGHERTY, Edward
Broome Technical Community College
Upper Front Street
Binghamton, N. Y.
DRISCOLL, Thomas L.
US Army Trans. Research Comma
Fort Eustis
Newport News, Va.
FELL, Henry W.
Princeton-Penna Accelerator
Forrestal Rech. Center
Route #1
Princeton, New Jersey
FINNEGAN, Katherine
Remington Arms Co.
939 Barum Avenue
Bridgeport, Conn.
FISCHER, Michael
Worthington Corp.
Harrison, N. J.
FITCH, Harold L.
Long Island Lighting Co.
250 Old Country Road
Mineola, Long Island, New York
FLAGG, Donald
Budd Company
Hunting Park Ave.
Philadelphia, Pa.
EISEMANN, Dr. Kurt
Catholic University
Michigan Ave. & 4th St., N. E.
Washington, D. C.
FLEISCH, Sylvia
Boston University
700 Commonwealth
Boston, Mass.
EISENBERG, Jerry
IBM
1120 Conn. Ave., N. W.
Washington, D. C.
FOLSE, Paul
Tampa Electric Co.
P. O. Box III
Tampa, Fla.
ELLIS, Robert w.
Houdry Process & Chem. Co.
1528 Walnut Street
Philadelphia, Pa.
FOX, J. Baine
Washington & Lee Univ.
Lexington, Va.
ELROD, J. C.
Georgia Experiment Station
Experiment, Ga.
GANGI, Vincent J.
The Armstrong Rubber Co.
475 Elm St.
West Haven, Conn.
5.1.3
450
- - - - - - - - - - - - - - - - - - - . -...- .•.......
GARDNER, Donald
General Foods Corp.
555 S. Broadway
Tarrytown, New York
GREEN, Bill
Arkansas Power & Light
6th & Pine
Pine Bluff, Arkansas
GARZANELL, James W.
Solvay Process-Allied Chem.
P: O. Box 271
Syracuse, New York
GROFT, Garth
York Div. Borg-Warner
Richland Ave.
York, Pa.
GELSI, Hector, P.
Chemical Construction
320 Park Ave.
New York, New York
GROVE, Dr. Richard E.
Randolph-Macon College
Ashland, Va.
----~-.-----.-
GUION, Richard L.
U.S. Dept. of Agriculture
14th & Ind. Ave., N. W.
Washington, D. C.
GERRISH, Everett
Orange and Rockland Utilities
125 High Street
Boston, Mass.
HAMILTON, E. Michael
George Washington University
300 North Washington Street
Alexandria, Va.
GLIMN, Arno
IBM
3424 Wilshire
Los Angeles, Calif.
HAMILTON, G. F.
Atlas Steels Co.
P. O. Box 20
Tracy, Quebec, Canada
GOLDBERG, Martin
United Aircraft Corp.
Helen Street
Norwalk, Conn.
HANAN, Sandra
GOLDEN, . J. T.
IBM
1730 Cambridge St.
Cambridge, Mass.
Rockland State Hospital
Orangeburg, New York
HARALAMPA, George S.
New England Electric System
441 Stuart St.
Boston, Mass.
GOLDMAN, Norman
Boston University
700 Commonwealth Ave.
Boston, Mass.
HAPLING, Reginald T.
A.F. Institute of Technology
Wright-Patterson AFB
Ohio
GOLDREICH, Estern
Young & Rubicam
285 Madison Ave.
New York, New York
HELLERMAN, Herbert
IBII
GOTTHEIM, Harold
N•y .• S. Dept. of Publ. Works
State Campus
Albany, New York
Yorktown Hts., New York
HELTZEL, William
Newport News Shipbuilding &
Dry Dock Co.
Newport News, ,va.
GRANT J Jay W.
IBM
San Jose, Calif.
HENDERSON, Frederick R.
Rochester Institute of Tech.
Rochester, New York
5.1.4
451
..-.. ............
~
.
(
'" "'\
I • • '
HENDRICKSON, MA3 USN, Lee
U. S. Naval Academy
Annapolis, Maryland
JACKSON, Arthur F.
A. and T. College
Greensboro, N. Carolina
HEWLETT, R. F.
Colorado School of Mines
Holfrn, Volotsfo
JARDINE, D'. A.
Dupont of Canada
Box 5000
Kingston, Ontario, Canada
HISE, William B.
George Washington Univ.
300 N. Washington St.
Alexandria, Va.
JASPEN, Dr. Nathan
National League of Nursing
10 Columbus Circle
New York, New York
HOFFMAN, Edward R.
St. Regis Paper Co.
Nyack, New York
JENSEN, Clayton, E.
Virginia Military Institute
Lexington Va.
HOFFMAN, L. L
Princeton Univ.
Guggenheim Labs.
Princeton, New Jersey
JOHNSON, James E.
Seton Hall Univ.
South Orange Ave.
South Orange, N.J.
HGFFMAN, Thomas R.
Union College
Schenectady, New York
JURNACK, Steve
Speer Carbon Co.
St. Marys, Pa.
HOKE, Tom
Oklahoma Gas & Electric
P. o. Box 1498
Oklahoma City, Okla.
KANE, J. L.
Guggenheim Labs.
Princeton University
Forrestal Research Center
Princeton, N. J.
HOULIHAN, Joseph
Riley Stoker Corp.
9 Neponset St.
Worcester, Mass.
KARR, Len
Bur.ns and Roe, Inc.
160 West Broadway
New York, New York
HUBBARD, J. C.
Baltimore Gas & Electric
Gas & Electric Building
Baltlmore'3, Md.
KELLMAN, Sidney
Naval Air Eng.• Center
Phila. Naval Base
Philadelphia 12, Pa.
HURT, Harry
U. S. Navy
Pensacola, Fla.
KELLER, John F.
Loyola University
New Orleans, La.
HUSSON, S. S.
IBM
Poughkeepsie, New York
KENNECOTT, Robert L.
Chas. H. Sells, Inc.
409 Manville Road
Pleasantville, N.' Y.
IRONS, R. C.
Naval Air Station
Pensacola, Fla.
Cl
lENT, John D
Worthington Corporation
P. O. Box 69
Buffalo, New York
IVESTER, Robert
Southern Engineering Co.
1000 Crescent Ave., N. E.
Atlanta, Ga.
5.1.5
452
...
......._.... ,,".... _
".""
.." .." .... _....
"~"
..•
~~--.-~--
..
~"---
'!
KINDRED, Alton R.
Ilanatee Junior College
Bradenton, Fla.
LIPSON, A. L.
Virginia Electr~c & Power
7th & Franklin Sts.
Richmond, Va.
KLEIN, Donald
V~llanova University
Villanova, Pa.
L'ITTLE,
Goucher
Dulaney
Towson,
DUEL, Martin
Burns an~ Roe, Inc.
160 west Broadway
New York, New York
Joyce II.
College
Valley Road
Maryland
LLOYD, Diana A.
FMC Corporation
P. O. Box 8
Princeton, New Jersey
KNIGHT, Mr. & Mrs. D. M.
st. Regis Paper Co.
Rt. 59;"'A
W. Nyack, 'New York
LOPEZ, Steve
IBM
112 E. Post Road
White Plains, N. Y.
LAPOSATA, Samuel M.
Ordance Research Lab.
P. O. Box 30
State College, Pa.
LOUIS, Jene
Long Island Lighting Co.
175 Old Country Road
Hicksville, New York
LARSON, Greta
Vapor Corporation
6420 W. Howard
Chicago'; Ill.
LOWRIMORE t G. R.
Hercules Powder Co.
Radford Army Ammunition Plant
Radford, Va.
LASKA,' Gene
Rockland State Hospital
Orangeburg, N. Y.
LUSSOW, Patricia
General Electric Compo
Ithaca, New York
LATTERNER, Thomas S.
Hercules Powder Co., Inc.
Rocky Hill, New Jersey
MAC LEOD" D. H.
stetson University
DeLand, Fla.
LEE, Ralph E.
Missouri 'School of Mines
Rolla, Mo.
MARTIN, C.
IBM
425 Park Avenue
New York, New York
LEillt, John F.
Ni agara . Mohawk
1100 Electric Bldg.
Buffalo, N. Y.
MARTIN, Samuel F.
US Aeronautical Charts
2nd Arsonal st.
St. Louis, Mo.
LIEBMAN, Judith
General Electric Compo
I~thaca, .New York
IIC GElID, Lt.T. L.
Naval
LIPPO,Veikko A.
West Pe,nn Power Co.
Cabin Hill Drive
Greensbu'rg, Pa.
Air Station
Fla.
Pen~acolaj
£.
lAUGHLIN, Jam.es J.
Aray )laterials Res. Agency
Watertown, Mass.
5.1 •.6
Baltimore Gas & Electric
Electric Building
Baltimore, Maryland
NUNBERG, Bennett
Leesona Moos Labs.
Lake Success Park
Great Neck, New York
MERIWEATHER, P. Wayne
NASA - MSFC
Huntsville, Ala.
OGILVIE, Norman
Airbourne Instruments Lab.
Melville, New York
MEADS, M. E.
,
MILLER, Adelaide
IBM
1120 Conn. Avenue, N. W.
Washington, D. C.
OLIVER, James R.
Univ. 'of Southwestern La.
Box 1333 USL Sta.
Lafayette, La.
MINKIN, Jean A.
Institute for Cancer Research
7701 Burholme Avenue
Philadelphia, Pa.
O'Neill, Brother B. E.,
College
Manhattan College Pkwy.
New York, New York
M~hattan
MOORHEAD, Patricia
John Hancock Mutual Life Ins.
200 Berkeley Street
Boston, Mass.
ORTH, Jr., E. J.
Southern Services, Inc.
P. 0, Box 2641
Birmingham, Ala.
MONAHAN, Jeanette
Rohn & Haas
5000 Richmond St.
Philadelphia, Pa.
ORTH, Richard
Worthington Corporation
Harrison, New Jersey
A. and T. College
Greensboro, N. C.
OWENS, John D.
New York Univeristy
University Heights
New York 53, New York
MORRISSON, Thomas E.
Speer Carbon Co.
47th and Packard Road
Niagara Falls, New York
PACINO, Salvatore
General Nuclear Eng. Corp.
P. O. Box 10
Dunedin, Fla.
MUNROE, M. E.
Universi'ty of New Hampshire
Durham, New Hampshire
PALMIERI, Anthony J.
Grumman Aircraft Eng. Corp.
South Oyster Bay Road
Bethpage, Long Island, N. Y.
MOORE, E. T.
NEWMAN, Edward
MIT
77 Massachusetts Ave.
Cambridge, Mass.
PARSONS, Jr. G. G.
Lublin, McGaughy & Assoc~
220 W. Freemason st.
Norfolk, Va.
NICHOLS, R. E.
Virginia Electric & Power Co.
7th and Franklin sts.
Richmond, Va.
:'· ~
c·
'"
PAQUIN, Nancy
U. S. Public Health Service
1901 Chapman Ave.
Rockville, Md.
NORTHROP, David T.
Fairchild Dumont Labs.
750 Bloomfield Ave.
Clifton, N. J.
PARKER, E. H.
Baltimore Gas & Electric
Gas and Electric Building
Baltimore 3, Maryland
5.1.7
PATERNITI, Richard
IBM
112 E. Post Road
White Plains, N. Y.
RAVER, Richard E.
W: R. Grace & Co·.
Clarksville, Maryland
REMILEN, Charles H.
Eastman Kodak
Lake Ave.
Rochester, New York
PAYNE, Gregson
F.T. Nicholls State College
Box 121
Tltibodaux, La.
PEARCE, D. J.
E. I. DuPont
Chattanooga, Tenn.
ROSENBARKER. I. E.
IBM
P. B. #6
Endicott, N. Y.
PETERS, Carl B.
Union County Tech. Inst.
423 Park Ave.
Scotch Plains, N. J.
RICHARDSON, F. W.
Jones & Lamson Mach. Co.
60 Clinton St.
Springfield, Vt.
PETERSON, Andrew I.
Univ. of Bridgeport
Park Ave.
Bridgeport, Conn.
RICKER, Guy W.
Jersey City State CDllege
Boulevard
Jersey City 5, New Jersey
PETERSON, H. C.
Air Reduction, Inc.
Murray Hill, N. J.
RIGGS, Heath K.
Univ. of Vermont
B~rlington, Vermont
PIXLEY, David C.
Bausch & Lomb, Inc.
st. Paul st.
Rochester, N. Y.
RILEY, R. B.
A.D. Margison & Assoc., Ltd.
1155 Leslie St.
Don Mill s I Ont. Canada
PLEINES, W. w.
Baltimore Gas & Electric
Gas and Electric Building
Baltimore 3, Maryland
RODGER.S, .r. A.
Baltimore Gas & Electric
Gas and Electric Building
Baltimore 3, Maryland
POINTER J John P.
Tenn. PolytechniC lnst.
Box 37
Cookerville, Tenn.
RUSSELL, J ame s H.
Gilbert Associates, Inc.
525 Lancaster Ave.
Re,ading, Pa.
POWELL, E. R.
Pan American World Airways
Patrick AFB, Fla.
RUVOLI S, Leo
York Junior College
Country Club Road
York, Pa.
PRATT, Richard L
Air Force Inst. of Tech.
Wright Patterson AFB, Ohio
SABINE, Theodore R.
Operations Research, Inc.
1400 Spring St.
Silver Spring, Md.
RANE, Dr., D. S.
Pennsylvania Military College
14th and Chestnut Sts.
Chester, Pa.
SALEK, Franklin
Hamden Testing Services,'Inc.
87 Valley Road
Montclair, New Jersey
5.1.8
c'
455
SAMSON, Stephen L.
US Atomic Energy Comm.
376 Hudson St.
New York, New York
SHIONO, Ryonosuke
Univ. of Pittsburgh
Pittsburg~ 13, Pa.
SHORE, Maynard N.
E. R. Squibb & Sons
Georges Road
New Brunswick, N. J.
SAMSTAG, Jay
U. S. Public Health Service
1901 Chapman Ave.
Rockville, Md.
SILVERMAN, Joan
IBM
555 Madison Ave.
New York, New York
SANDIN, E. M.
Baltimore Gas & Electric
Gas and Electric Building
Baltimore 3, Md.
SINAN,IAN, Ed
IBM
425 Park Ave.
New York, New York
SAWYER, Dr., John W.
Wake Forest College
Box 7401
Winston Salem, N. C.
SMITH, Walter J.
U. S. Naval Academy
AnnapoliS, Md.
SCALETTI, Henry M.
United Shoe Mach. Corp.
Balch St.
'
Beverly, Mass.
SMITH, Welborn H.
Airbourne Instruments Lab.
Melville, New York
SCHIFFMAN, Leonard
Berger Associates, Inc.
P. O. Box 1943
Harrisburg, Pa,
SOUCY, Robert
Transitron Electronic Corp.
168 Albion st.
Wakefield, Mass.
SCHLAGER, Dr. Gunther
The Jackson Laboratory
Bar Harbor, Maine
SPARVERO, 'Luke J.
Corporation
Dunham Road
Meadville, Pa.
FMC
SCHNECKER t Joe
Bethlehem Steel Co.
Bethlehem, Penna.
SPITALNY, Arnold
United Aircraft Corp.
Helen st.
Norwalk, Conn.
SCHRODEL, C. S
Sun Oil Co.
1608 Walnut St.
Philadelphia, Pa.
SPRINGER, Berl .M.
Southwestern Public Service
3rd & Polk Street
Amarillo, Texas
SCOTT, Thomas J.
Sun Oil Co.
1608 Walnut St.
Philadelphia 3, Pa.
STACD, Wanda
AT&T
32 Ave, of Amer.
New York, New York
SEEWALD, Mil ton
Texaco, Inc.
P. O. Box 509
Beacon, N. Y.
STAIANO, E. F.
Bucknell Univ.
LeWisburg, Pa.
SHEEHAN, Daniel M.
Virginia Polytechnic Inst.
Blackburg, Va.
5.1.9
456
STANSBURY, James C.
Halcon International, Inc.
2 Park Avenue
New York, New York
THOMPSON, E. E.
Arkansas Power & Light
6th and Pine
Pine Bluff, Arkansas
THOMPSON, William B.
Photo & Repro. Div.
General Aniline & Film Co.
Binghamton, N. Y.
STEINHART, R. F.
IBM
425 Park Ave.
, New York, New York
STONE, Eleanor
Brandeis University
Waltham, Mass.
TIERS, John
Bethlehem Steel
Bethlehem, Pat
STORRER, R. L.
Norwich Pharmacal Cq.
Box 191
Norwich 1, N. Y.
Gen. Am. Trans. Co.
7501 N. Natchez Ave.
Chicago, Ill.
SULLIVAN, Peter J.
United Illumination Co.
BO Temple street
New Haven., Conn.
TURNER, Don N.
Nat's Aero & Space Adm.
600 Independence Ave.
Washington, D, C.
SVIHLA, Edward B.
Automatic Electric Labs.
P. O. Box 17
Northlake, Ill.
VOYTOVICK, Sharon M.
Syracuse University Resl Corp.
P. O. Box 26
Syracuse, New York
TAYLOR, Carlis
National Naval Med. Center
Bethesda, Maryland
WAGNER, Thomas F.
IBM
425 Park Ave.
New York, New York
TRANTUM, J. W.
TAYLOR, CDR., Dean
U. S. Naval Academy
Annapolis, Maryland
WANG, Francesca
Stevens Inst. of Tech.
Hoboken, N. J.
TAYLOR, 'James S.
Data Corporation
7500 Old Xenia Pike
Dayton, Ohio
WEAVER, John
University of Dela,
Newar~, Dela.
TEEPLES, Thomas C.
George Washington Univ.
300N. Washington St.
Alexandria, Va.
WEINER, Myrna S.
Brandeis University
Waltham, Mass.
WEISS, Charles
USAF Aeronautical Chart
St.Louis, Mo.
THOMAS, Edward
Engineering Physics Co •.
5515 Randolph Road
Rockville, lid.
WELLS, Frank· J .
Long Island Lgt; Co.
175 Old Country Road
Hicksville, New York
TROMAS, Roy W.
Southwestern Public Service
Box 1261
Amarillo, .,Texas
c
WHELLER, R. C.
Airborne Instruments Lab.
Melville, New York
5.1.10
~57
c
WHITACRE, C. G.
Operations Research Group
Edgewood Arsenal, Md.
WYNN, W. Phil
George Washington Univ.
300 N. Washington St.
Alexandria, Va.
WHITE, Robert R.
L.A. Dept. Water & Power
207 S. Broadway
Lps Angeles, Calif.
YACKULICS, Charles W.
The Armstrong Rubber Co.
475 Elm Street
W. Haven, Conn.
WILLIAMS, Charles S.
Washington & Lee Univ.
Lexington, Va.
ZIMMERMAN, Leo F.
E.I. DuPont de Nemours & Co.
Savannah River Lab.
Aiken, s. 9.
WILLIAMS, D. D.
Baltimore Gas and Electric
Gas and Electric Building
Bal timore 3, lid.
ADDITIONS
,DAVIS, Richard B ..
Orange and Rockland utilities
125 High Street
Boston, Mass.
WILSON, George
Black & Veatch
1500 Meadowlake Pkwy.
Kansas City, Mo.
KRENZER, Theodore J.
Gleason Works
1000 University Ave.
Rochester, N. Y.
WINGERSKY, Boty G.
University of N. H.
Durham, New Hampshire
WINGERSKY, Marilyn S.
University of N. H.
Durham, New Hampshire
QUIGLEY, James J.
Orange & Rockland Utilities
10 North Broadway
Nyack, New York
WINGERT, Joseph T.
Trenton Junior College
101 w. State st.
Trenton, N. J.
. STUDENTS
CARLSON, Denni s
Wesley College
Dover, Dela.
WOLFF, Jack M.
Brooklyn College
Brooklyn 10, New York
BEALLY, J. Rodge r
Wesley College
Dover, Dela.
WOODRUFF, George H.
N. J. State Highway Dept.
1035 Parkway Ave.
Trenton, N. J.
CLEMENT, Ronald
Wesley College
Dover, Dela.
WRIGHT, Donald L.
George Washington Univ.
37th and 0 Streets, N. W.
Washington, D. C.
LIBERMAN, Nick
Wesley College
Dover, Dela.
WRIGHT, Lawrence
Sprague Electric
Marshall St.
North Adams, Mass.
MORROW, Robert
Wesley College
Dover, Dela.
WILSON, M. A.
Wesley Co~lege
Dover, Dela.
5.1.11
PROGRAMS TEAMS CHAIRMEN
~-""""
''\...,'/
1620 USERS GROUP - EASTERN REGION
Frank J. Wells, Chairman
Electric Utility Team
1620 Users Group - Eastern Region
Long Island Lighting Company
175 Old Countr.y Road
Hicksville, Long Island, New York
Mr.
Area ,16 - Phone:
WE 1-6300
Dr. James R. Oliver, Chairman
Operations Research Team
1620 Users Group - Eastern Region
USL Computer Center
Box 133 - USL Station
Lafayette, Louisiana 70,06
Dr. Frank N. Dickinson, Chairman
Statistics and Mathematics Team
1620 Users Group - Eastern Region
Agricultural Research Center
Biometrical Services - U.S.D.A.
Beltsville, Maryland
Mr. J. J. ," Owen, Chairman
1110 User~ Sub-Group
Union Bag-Camp Paper Corporation
P.O. Box 570
Savannah, Georgia
Area 912 - Phone:. AD 6-,111
Mr. Thomas Scott, Chairman
Structural and Oivil Engineering Team
1620 Users Group - Eastern Region
Sun Oil Company
1608 Walnut Street
Philadelphia 3, Pennsylvania
Phone: KI 6-1600
Mr. Arnold K. Canner, Chaiman
General Data Processing Team
1620 Users Group - Eastern Region
Transitron Electronic Corporation
168 Albion Street
Wakefield, Massachusetts
Area 611 - Phone: 24,-4500
Mr. Willi.~ ,H~lt~e;l., Altez-nate Qb~~
Structur~i~d
91v;i ,Et!g~eerjng ·Te~
1620 US$rs Qroup - Eas'tern ~egion
.
Newpor:t NewsSllipbuilding&Pry(j,Q.Cjk C()lIlP~
HewpQrt.¥~s, .Virginia
;S;~2.1
~'5"9
"..(,
,~,' ,
'"
o
Programs Teams Chairmen (Cont.)
Mr. N. Goldman, Chairman
Education Institutions Team
1620 Users Group - Eastern Region
Boston University
700 Commonwealth Avenue
Boston, Massachusetts
Phone: CO 7-1200
Mr. T. H. Korelitz, Chairman
Chemical Engineering Team
1620 Users Group - Eastern Region
The Badger Company
363 Third Street
Cambridge, Massachusetts
Mr. C. S. Schroedel, Acting Chair!T18n - 5/64
Process and Chemical Engineering Team
1620 Users Group - Eastern Region
Sun Oil Cbmpa.ny
1608 Walnut Street
Philadelphia, Pennsylvania
Phone: KI 6-1600
Mr. K. W. Brady, Chairman
Rate Engineering Team
1620 Users Group -'Eastern Region.
Long Island Lighting Company
250 Old Count~J Road
Mineola, New York
Area 516 - Phone: PI 7-1000 .
E~t. 422
Mr. Arnold Spitalny, Chairman
Electronics Team
1620 Users Group - Eastern Region
United Aircraft Corporation
Norden Division
Norwalk, Connecticut 06852
Area 203 -Phone: 838-4471
Mr. C. Bailey, Chairman
Plotting Team
1620 Users Group - Eastern Region
IBM Corporation
Endicott, New York
Mr. Martin Goldberg, Alternate Chairman and
Secretary
Electronics Team
1620 Users Group - Eastern Region
. United Aircraft Corporatio!l
Norden Division
Norwalk,Connecticut 068$2
Area 203 - Phone: 838-4471
Mr. E. E ...Newman, Alternate Chairman
Plotting Team
1620 Users Group - Eastern Region
Department of Civil Engineering - Room 1-179
Massachusetts Institute of Technology
Cambridge 39, Massachusetts
$.2.2
460
.~--.--
.....•.. ,.- ..-....", ......... ".
Hay 28, 1964
Report of Meeting of Programs Teams Chairmen
1620 Users Group - Eastern Region
Arligton, Virginia - May 7I 1964
A.t a meeting or eight of the eleven Programs Teams Chairmen
who ran sessions at this meeting of the 1620 Users Group - Eastern
Region, the following i tams were discussed as indicated:
1.
It was suggested that the attendants at these sessions be asked to
introduce themsel ves so that a spint of comradery could .be fostered
particularlY among those who wanted to have themselves considered as
"permanent" members of the Teams. The question of the continuity of
the Teams' efforts between meetings was discussed in this context.
2.
It was requested that rosters of the attendants at these sessions be
compiled with indications of those who wanted themselves to be oonsidered as "permanent" members of the Teams.
3. It was requested that a meeting secretary be appointed for each Programs Team to take notes on discussion topics and papers whose authors
did not want them published in toto. It was pOinted out that the
authors of papers presented at these meetings are not requested to
have them published in the Proceedings.
h.
A survey was asked for from each Programs Team Chairman of the inter-
ests which the "permanent" members ot his Team have in the ses SiODS
other Programs Teams. It was pointed out that this information
would be useful to the Program Chairman of the meeting in scheduling
the Programs Teams sessions. In this context, it v.s suggested that
the Programs Teams, in general, could be organized as far as their
interests are concerned along lines suggested by the program groupings
in the latest quick index.
or
,.
The structure of the meeting, namely programming and operating systems
during the first day and a half followed by' applicatiOns systems and
programs during the remaining day and a halt, vas reviewed with the
Programs Teams Chairmen with the suggestion that only those topiCS
in programming and operating systems or exolusive interest to a particular Programs Team be included in that Team'ssession agenda.
6. The matter ot continuity in Team. activities and leadership between
meetings was discussed. Allot the Programs Teams Chairmen at this
meeting agreed to continue their leadership at least until the next
meeting or the Users Group.
These items were subsequently reviewed with the three Programs
Teams Chairmen not present at this meeting.
461
D. D. Wllliams
Program Chatman
1964 Spring Meeting
1620 Users Group - Eastern Region
DDW:sk
5.3.1
SURVEY OF TYPE 1620 PROGRAMMING SYSTEMS
1620 Users Group - Eastern Region Meeting
Washington J D. C.
May 6, 7 and 8, 1964
Machine
Requirements
System
Detail
Page
lIM Corporation
Statement Languages
1. FORTRAN with Format (Card) - Fo-004, Ver IT 201
2. FORTRAN with Format (Tape) - Fe-004, Ver II 20K
3. FORTRAN II (Card) - FO-019
401
Indirect Addressing
Automatic Divide
4. GOTRAN (Card) - PR-011
201
S. GOTHAM (Tape ) - PR-OI0
20K
C:I
1
2
3
Sl!!!bol1c LanS!:aS!8
.,.>/
1. Symbolic Programming System (Card) SP-009
2. Symbolic Programming System (Tape) -
20K
4
201
S
3. 1620/1710 SPS (Card) - SP-020
20K
6
6
sp-ooB
4. 1620/1710 SPS (Tar) - sP-021
5. 1710 SPS II (Card - SP-007
6. 1710 SPS II (Tape) - SP-006
7. SPS III (Card) - SP-027
B. SPS III (Tape) - SP-028
Statement and
2. Monitor I (Tape) - PR-026
a. SPS II-D
b. FORTRAN II-D
0
201
201
20K
Indirect Addressing
201
Indirect Addressing
S~bol1c Lan~ales
1. Monitor I (Card) - PR-02S
a. SPS n-D
b. FORTRAN II-D
!
201
20K
1311 Disk Dri"
Indirect Addressing
A.utomatic Divide
201
1311 Disk Drive
Indirect Addressing
Automatic Divide
S.4.1
462
- Page 2 -
Statement Languages
1.
AFI'l' ~jrOVed FORTRAJ
(Card - 01.1.010
2.
Complex FORTRAN (Tape) -
06.0.008
3. FORTRAN 90 COllpUer and
Subroutines (Card) -
02.0.014
h. FORTRAN 1/2 (Tape) 01.6.037
S. FORTRAN I tor Magnetic
Tape System (Card) -
Machine
Author
Require_nta
Pratt
Air Force Institute
ot 'echno1og'
Maskiell
McGraw-Edison COIIlP&n7
Borst
Bausch &: Lomb Optical
Company
20K
Syatea
-Users
Anderson
Jllen-Bradle.y Co.p~
Wenrick
ACF IndUIJ~ie8, Inc.
02.0.018
6. FORTRAN I tor Magnetic
Tape S7Stem with Auto-
20K
20K
Automatic Divide
Indirect Addressing
Move Flag
20lC
201
Automatic Divide
Indirect Addressing
2 Tape Drives
Move Flag
60K
Wenrick
lC' Industries, Inc.
aatic noating Point
(Card) - 02.0.02S
Automatic Divide
Indirect Addres8ing
2 Tape Drives
Hove Flag
7. FORGO (Load &: Go FORTRAN)
(Card) - 02.0.008
8. FARGO (Card) - 02.0.027
9. FOR-TO-GO (2 Pass FOROO)
(Card) - 02.0.009
10. Load It Go FORTRAN
(Card) - 01.1.009
ll.
Load & Go FORTRAN
(Tape) - 01.1.008
12. NOVATlWf I (Tape) 02.0.012
13. trro FORTIWf (Card) 02.0.024
McClure
University ot Wisconsin
Heyworth
University of JJ..berta
McClure
UJ11T8rsit7 of Wisconsin
MailloU'X
ot Al.berta
Un1verait,.
hOI
Indirect Jddressing
Au.tomatic Divide
601
Indirect Addressing
AutcBa'tio Di'ride
1 MagnetiC Tape Drift
Move nag
hOI
Indirect Addressing
Automatic Di'ride
hOI
Indirect .lddressing
Mailloux
Un1:nrsity of Alberta
hOI
lorthrop
201
Fairchild Caaer.
&lnstr.
Lee
201
Univera!ty of Toronto
\(w~~
S.4.2
463
- Page 3 -
C\
sntea
14.
UTO FORTRAR Processor
and Subroutines (Tape) -
Author
Machine
Requirements
Field
Univerait,y of Toronto
201
Brittain
Sunstrand biation
Automatic Di'9'icIe
Indireot Jddre88iDg
J.ut~t1c noatiDg Point
Move Flag
02.0.011
IS.
.lutCllat1c noatiDI FOR'lRAI
PrOC888iD1 S78tea -
01.6.076
Xaak1ell
HcGraw-Edison ec.p...,
201
Automatic Dinde
201
16.
PDQ rORDAI (Card) -
17.
Mult1proc•• ainl rORftAI
(tape) - 1.1.00S
18.
601 Load It Go rCll'l'lWf
Mailloux
(Card) - 1.1.019
Univerait7 of Alberta
Indirect Addressing
Rumrill &£ rowler
Beyuk College of
201(
lutOllatic Dinde
2.0.031
19. ICI Load It Go rOR'l'RAI
(Tape) - 2.0.032
Carr
Indiana Tech. College
Engineering
601:
Indirect lddressing
Model 1
C)
.20.
Rei Load It Go FORnAl
(Ced) - 2.0.029
Ru1Ir11l It loyler
201
Newark Col1ep of
lngiDeer1ng
Automatic D1'Yid.e
Indirect Addressing
Modell
- Page
4Machine
Slstea
Author
Regu1r8l118nta
S_bollc Language.
1. Fast Assembler (Card) 02.0.028
HQWOrth
UniTersit7 of A'lberta
601
.lutca.tic Dirlde
Indirect Addressing
a'Magnetic Tape DriftS
MOft Flag
2. OSAP AlJa8JIb1l' S7at.
(Card) - 01.1.012
lumerical Caaputation
Laboratory
201
Indireot Addressing
Ohio State Uniftrsit7
3. SPS I for Magnetic Tape
S1St.. (Card) - 02.0.021
4. SPS II for Xagnetic Tape
Syst.a (Card) - 01.6.011
S.
LAMP (Card) - 1.1.001
6. Relocatab1e uS8mb17
System (Card) - 1.1.020
7.
AlI! SPS (Card) - 1.1.023
Wenrick
A'Cl Industries, Inc.
Wenrick
ICF Industrie., IDe.
Matth78
191
601
3 Magnetic
Tapes
Autoaatic noat1ng Point
201
~rporatiOD
Richardson
AUstralian !EO
Pratt
Air Force Institute
of Technolog
8. FAST .l'8seabler (Card) 2.0.028
60K,
Mailloux & Darla
UniT8rait7 of Alberta
60K
Indirect Addressing
401
Indirect Addressing
Automatic DiT1de
Move
nag
601
2 Magnetic Tapes
Indirect AddreSSing
Automatic Divide
Move nag
DetaU Page 1
PORTRAR with Foraat (Card)
C;
Machine
Author
'~~ProF-
Ind1reotAddreesiDg
1. Card rORTRAI Coapressor Ol.3.oo1~:! . ,
IBM Corporation
2. D.Jnamic !race - 01.4.011
Lesson
1111 Corporation
Indireot Addressing
Pope
Utah State tJDiftra1t7
Indirect Addressing
3. POROOM Subroutine - 01.6.0S1
4. For..at FORT.RAN Object Deck
Compressor - 01.2.009
S. FORTRlN PONat Checker 01.1.017
6. 0 _ Function SubroutiDe Positi" !rgaenw -
07.0.011.8
e,1
!ruh .
Regu1reaenta
Gi"t'Mr
Pont
Dept. ot Public Works
(Pta.no Rico)
Paohon
Automatic.Eleotric
Labe, Inc.
Vanscha1lc:
11M Corporation
8. Plot Subroutine - 01.6.0$6
i87Dolds
Sprague Ileatria CoapaDl'
10. Subroutine tor Plotting witb
a ClL-COMP Plotter -
nag
Brookqn College
7. Plot Subroutine - 06.0.U9
9. Randall . .bar Subroutine
07.0.021
Mow
201
Indireot AddreesiDg
Automatic Diyide
10 Automatic DiTide
P1Dk
Iltt Corporation
Passino
Todd Sh1pJVd Corp.
Automatic Divide
Special Instructiona
01.6.068
11. LiDk Subroutine - 13.0.002
12. not Sv.broutine - 13 .0.001
13. roRTRAR C0IIlpl'888or - Loader 1.2.007
lb.
FORDAH M-0Z7 Dap -
IS.
rCllTRAI Compressor and 7S
ColuD Daap - 1.3.008
1.3.004
Poore
201
Louisiana PI
Poore
Universi tJ' ot ICentuclq
201:
Indirect Addressing
Sabath
I ••tun Kodak
601:
tIaland
Dow Chemical Ccap8ll7
Indireot Jddrea8ing
Moore
Indirect Addres8iDg
1111 Corporation
C
S.4.~
466
Detail Page 1
(Cont1nued)
ProF16.
FORTlUI Coapres8or and
Programmer - 1.6.o4S
Multi-
17. FORCOM - 1.6.0Sl
18.
FORSTAD -
19.
Modified CDS FORCOMSub-
1.6.0S4
routine -1.6.096
Machine
Author
Reguirement.
Gardner
General rooda
201
Pope
Ut.ah State Uninrsit,.
Indirect.&ddressinc
ArJderson
General Electric
Webster
V1ctor1aElectrtc1t7
Indirect AddreSSing
CoD1s8ion
20. FOR!RAH Pre-Ccapiler ro-oo6
IB4: CO!'pOration
21. Plot Subroutine - MP-lbI
IEII Corporation
AutOilatic Divide
OAL-CctIP Plotter
467
Detail Page 2
FOR!RAII with Format (Tape)
(-~
",/
Machine
Prop-_
1. Bessel Punctioo Subroutine 07.0.031
2. CAL-OOMP S6S Plotter Subroutine Pen - 13.0.00$
3. n,namic Trace - 01.4.012
h. Random Haber Subroutine 07.0.022
S.
Random I_bar' Subroutine -
07.0.009
6. FORTRAN Lister - 1.1.007
luthor
SUDival1
General luclear
Requirements
201
Autanatic Dirlde
Eng. Corp.
Michel
Rational Lab.
Indirect Addressing
Argonne
Laeeon
IS( Corporation
. Indirect Addres8ing
Move Flag
F1n1c
IBM Corporation
Sanders
Abbott Laboratories
Pratt
Air Force Institute
of Technology
("'''I
,.,Il/
7. . FORTRAN Pre-Canpiler FO-OOS
Ill{ Corporation
C',·'\
"'/
S.4.7
469
-------------
------.",,,._.. _,,,,.
Detail page 3
FORTRAN II (Card)
Machine
Reguirements
Author
.-
Program
1. Dump Digit-X.'Un1oad MemolT
Program - Self Load,
II and Format
Subroutine - 01.6.092
Orth
Southern Services, Inc.
,I"
'I.",,'
60K
FORTRAN
2. FOReOM Subroutine - 01.6.074
3.
4.
,.
Webster
State Electrie1t7
Commission ot Victoria
(Australia)
n S1mbol Table
Punch - 01.6.067
McIlvain
Catalytic Construction Co.
Move Flag
FORTRAN II Diagnostician 01.6.019
Burgeson
Indirect Addressing
IlJ4: Corporation
FORTRAN
Format Statement Pres.lector 01.6.086
Morgan
WaslrlDgton State
Highway Depart.-nt
6. Relocatable Cos J Sin, and
Exp. Subroutine., -
White
Mqo
07.0.0)8
7. Relocatable Plot Subroutine 1).0.004
06.0.111
9.
Normal HUllber
Generator Subprogram -
RandOll
06.0.108
Clinic Coaputer
FacUity
Brq
Bray
Move Plag
Boeing Scientifio
Research Labs.
11. 1620 SOR'1'-LibrUT function 01.6.069
MndersoD
Sandia Corporation
011
Cards -
01.3.011
13. Auto...tic Floating Point
"
Boeing Soientific
Research Labs.
Fassino
Todd Shipyard, Inc.
S7llbo1 Table
(,c".,\
Cl1rd.c Coaputer
'acUit7
10. Subroutine tor Plotting with
CAL-COMP Plotter - 01.6.06S
12.
Automatic Ploating Point
Wbite
MaTo
8. RandClll Exponential IfuIlber
Generator Subprogram -
60K
QbristeDa8n
Move Flag
noating Point Hardware
'ltD' IDatructiOD
$11vania Ilectric
s,st.. -
West
Iat Corporation
Autoaat1c Floating Point
Subroutine. - IH-024
S.4.8
t!67
C
Detail Page 3
(COntinued)
c
ProE-
th.' blocatable DS and TBF
Subroutines - 1.6.063
is.
ALTER
n
Machine
luthor
Requirements
White
Move Flag
Mayo Clinic Computer
Fac1l1t 7
Levin
Hazeltine Reaearch Division
S.h.9
470
Szmbo11c PrOgra.!DS STate. (Card)
Prosr-
'Mach1Jle
Author
Requir. .nta
1. Additional Instruction.
Macro Subroutine - 01.1.002
Petersen
Washington State
H1ghvq Depart&ent
Indirect Addres'smg
Input-QQtput Subrout1D.. 01.6.028
D7kstra
General Foods Research
,20K
2.
Center
3. Multiple Precision Floating
Point .Arttbmetic Subroutine - 01.6.041
4. Relocator
PrOgrUl - 01.2.00S
S. SPS Modifier tor Magoetie
Tape Operation - 01.6.039
6. Multi-Purpose SPS Compressor '.
1.1.006
7. noat - 1.6.083
Hotfman
601
Boeing COIIlpatl7
Indirect Addressing
,Automatic Divide
"MoTe nag" Instruction
Glans
Purdue University
201
Indirect Addressing
aMove nag- Instruction
Hollina
-Bell TelephoDe Labs.
Bate
Wolt Research
..__ .-:;r
,and Developnent
:G1nsburg
D.S. P.B.S.
Indirect Addressing
JutClllatic Divide
Moft nag
S.h.10
4'7 1
Detail Page S
Symbolic Programming Slstem (Tape)
Machine
Program
1.
2.
SPS
Subroutines - 01.1.00)
Ass8mbq Program tar
Floating Po1ntlnput/Output
Subroutines - 01.6.02)
Author
Requirements
Pratt
Air Force Institute
ot TecbnoloU'
201
Pratt
201
Air Force Institute
of Teohnology
S.h.ll
472
1620/1710 SPS (Card)
Program
1.
Floating Point Conversion
Subroutines - 1.6.0$3
Author
Machine
Re9ui~menta
Pratt
.lit- rorce Inst1tut.
ot Technology
2.
Label Reterence Indexer -
1.1.014
3. Label Reterence Indexer
Without Indirect
Addressing- 1.1.013
4.
LlOS Floating to Fixed -
S.
L106 Floating to Filted -
1.6.048
1.6.049
6. Object Deck Analyser 1.6.060
7. Symbol Table Punch 1.3.010
8. Storage and Label Anal7aer 1.6.093
1.
Label Reterence Indexer -
1.1.022
Pratt
A,ir Force Institute
ot Teohnology
Indirect Addressing
Pratt
Air Force Institute
of Technology
Engelhardt
Ill( Corporation
Indirect AddressiDs
illpl.lla'rdt
1111 Corporation
Indirect AddreSSing
Beveh
BaacoclC
John
Bradshaw
Indireot .lddressing
IncB.rect lddre8,s1ng
Christian Brothers
Oollege
BOWMan .
IDdirect AddreSSing
IlJ! Corporation
Pratt
:l\r Porce Ill8titute
GtrecbDolGl7
,S.b.'12
473
Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.3 Linearized : No XMP Toolkit : Adobe XMP Core 4.2.1-c043 52.372728, 2009/01/18-15:56:37 Create Date : 2014:03:05 15:39:28-08:00 Modify Date : 2014:03:05 16:21:38-08:00 Metadata Date : 2014:03:05 16:21:38-08:00 Producer : Adobe Acrobat 9.55 Paper Capture Plug-in Format : application/pdf Document ID : uuid:7dac36a2-4e4e-aa4e-9470-10e942d7b560 Instance ID : uuid:e8ddce8d-122e-b24f-9be1-8334a4bee37e Page Layout : SinglePage Page Mode : UseNone Page Count : 481EXIF Metadata provided by EXIF.tools