Solutions Manual For Linear Algebra And Its Applications (4th Edition)
Solutions%20Manual%20for%20Linear%20Algebra%20and%20Its%20Applications%20(4th%20Edition)
Solutions%20Manual%20for%20Linear%20Algebra%20and%20Its%20Applications%20(4th%20Edition)
User Manual:
Open the PDF directly: View PDF 
.
Page Count: 485
| Download | |
| Open PDF In Browser | View PDF |
INSTRUCTOR
SOLUTIONS
MANUAL
INSTRUCTOR’S
SOLUTIONS MANUAL
THOMAS POLASKI
Winthrop University
JUDITH MCDONALD
Washington State University
L INEAR ALGEBRA
AND I TS
A PPLICATIONS
F OURTH E DITION
David C. Lay
University of Maryland
The author and publisher of this book have used their best efforts in preparing this book. These efforts include the
development, research, and testing of the theories and programs to determine their effectiveness. The author and
publisher make no warranty of any kind, expressed or implied, with regard to these programs or the documentation
contained in this book. The author and publisher shall not be liable in any event for incidental or consequential
damages in connection with, or arising out of, the furnishing, performance, or use of these programs.
Reproduced by Pearson Addison-Wesley from electronic files supplied by the author.
Copyright © 2012, 2006, 1997 Pearson Education, Inc.
Publishing as Pearson Addison-Wesley, 75 Arlington Street, Boston, MA 02116.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any
form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written
permission of the publisher. Printed in the United States of America.
ISBN-13: 978-0-321-38888-9
ISBN-10: 0-321-38888-7
1 2 3 4 5 6 BB 15 14 13 12 11
_____________________________________________________
Contents
CHAPTER 1
Linear Equations in Linear Algebra 1
CHAPTER 2
Matrix Algebra
87
CHAPTER 3
Determinants
167
CHAPTER 4
Vector Spaces
197
CHAPTER 5
Eigenvalues and Eigenvectors
273
CHAPTER 6
Orthogonality and Least Squares
357
CHAPTER 7
Symmetric Matrices and Quadratic Forms
CHAPTER 8
The Geometry of Vector Spaces
405
453
iii
1.1
SOLUTIONS
Notes: The key exercises are 7 (or 11 or 12), 19–22, and 25. For brevity, the symbols R1, R2,…, stand
for row 1 (or equation 1), row 2 (or equation 2), and so on. Additional notes are at the end of the section.
1.
x1 + 5 x2 = 7
−2 x1 − 7 x2 = −5
ª 1
« −2
¬
5
−7
7º
−5»¼
Replace R2 by R2 + (2)R1 and obtain:
x1 + 5 x2 = 7
3x2 = 9
x1 + 5 x2 = 7
Scale R2 by 1/3:
x2 = 3
x1
Replace R1 by R1 + (–5)R2:
= −8
x2 = 3
ª1
«0
¬
5
ª1
«0
¬
5
ª1
«0
¬
0
ª1
«5
¬
2
7º
9 »¼
3
7º
3 »¼
1
1
−8 º
3»¼
The solution is (x1, x2) = (–8, 3), or simply (–8, 3).
2.
3x1 + 6 x2 = −3
5 x1 + 7 x2 = 10
ª3
«5
¬
6
7
−3 º
10 »¼
Scale R1 by 1/3 and obtain:
Replace R2 by R2 + (–5)R1:
Scale R2 by –1/3:
Replace R1 by R1 + (–2)R2:
x1 + 2 x2 = −1
5 x1 + 7 x2 = 10
x1 + 2 x2 = −1
−3x2 = 15
x1 + 2 x2 = −1
x2 = −5
x1
= 9
x2 = −5
ª1
«0
¬
−1º
10 »¼
7
2
−3
ª1
«0
¬
2
ª1
«0
¬
0
1
1
−1º
15»¼
−1º
−5»¼
9º
−5 »¼
The solution is (x1, x2) = (9, –5), or simply (9, –5).
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1
2
CHAPTER 1
• Linear Equations in Linear Algebra
3. The point of intersection satisfies the system of two linear equations:
x1 + 2 x2 = 4
ª1
«1
¬
x1 − x2 = 1
2
−1
4º
1»¼
x1 + 2 x2 = 4
Replace R2 by R2 + (–1)R1 and obtain:
−3 x2 = −3
x1 + 2 x2 = 4
Scale R2 by –1/3:
x2 = 1
= 2
x1
Replace R1 by R1 + (–2)R2:
x2 = 1
ª1
«0
¬
ª1
«0
¬
ª1
«0
¬
2
−3
4º
−3»¼
4º
1 »¼
2
1
2º
1»¼
0
1
The point of intersection is (x1, x2) = (2, 1).
4. The point of intersection satisfies the system of two linear equations:
x1 + 2 x2 = −13
3 x1 − 2 x2 =
1
ª1
«3
¬
2
−2
−13º
1»¼
Replace R2 by R2 + (–3)R1 and obtain:
Scale R2 by –1/8:
Replace R1 by R1 + (–2)R2:
x1 + 2 x2 = − 13
− 8 x2 =
40
x1 + 2 x2 = − 13
x1
x2 =
−5
=
−3
x2 =
−5
ª1
«0
¬
2
−8
ª1
«0
¬
2
ª1
«0
¬
0
1
1
−13º
40 »¼
−13º
−5»¼
−3 º
−5 »¼
The point of intersection is (x1, x2) = (–3, –5).
5. The system is already in “triangular” form. The fourth equation is x4 = –5, and the other equations do
not contain the variable x4. The next two steps should be to use the variable x3 in the third equation to
eliminate that variable from the first two equations. In matrix notation, that means to replace R2 by
its sum with –4 times R3, and then replace R1 by its sum with 3 times R3.
6. One more step will put the system in triangular form. Replace R4 by its sum with –4 times R3, which
4
0 −1º
ª 1 −6
«0
2 −7
0
4 »»
«
. After that, the next step is to scale the fourth row by –1/7.
produces
«0
0
1
2 −3»
«
»
0
0 −7 14 ¼
¬0
7. Ordinarily, the next step would be to interchange R3 and R4, to put a 1 in the third row and third
column. But in this case, the third row of the augmented matrix corresponds to the equation 0 x1 + 0
x2 + 0 x3 = 1, or simply, 0 = 1. A system containing this condition has no solution. Further row
operations are unnecessary once an equation such as 0 = 1 is evident. The solution set is empty.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.1
• Solutions
8. The standard row operations are:
ª1
«0
«
«0
«
¬0
−5
4
0
1
0
0
0
3
0
1
0
2
ª1
«0
~«
«0
«
¬0
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
−5
1
0
0
0
1
0
0
0
0
0
1
−5
4
0
1
0
0
0
3
0
1
0
1
0
1
0
0
0
1
0
0
0
0
0
1
0º ª 1
0»» ««0
~
0» «0
» «
0¼ ¬0
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
−5
4
0
1
0
0
0
3
0
0
0
1
0º ª 1
0 »» ««0
~
0 » «0
» «
0¼ ¬0
−5
4
0
1
0
0
0
1
0
0
0
1
0º
0 »»
0»
»
0¼
0º
0»»
0»
»
0¼
The solution set contains one solution: (0, 0, 0, 0).
9. The system has already been reduced to triangular form. Begin by replacing R3 by R3 + (3)R4:
ª1
«0
«
«0
«
¬0
−1
1
0
−2
0
0
0
0
1
0
−3
1
−5º ª 1
−7 »» «« 0
~
2» «0
» «
4¼ ¬0
−1
1
0
−2
0
0
0
0
1
0
0
1
−5º
−7 »»
14 »
»
4¼
Next, replace R2 by R2 + (2)R3. Finally, replace R1 by R1 + R2:
ª1
«0
~«
«0
«
¬0
−1
1
0
0
0
0
0
0
1
0
0
1
−5º ª 1
21»» «« 0
~
14 » « 0
» «
4 ¼ ¬0
0
1
0
0
0
0
1
0
16 º
21»»
0 14 »
»
1 4¼
0
0
The solution set contains one solution: (16, 21, 14, 4).
10. The system has already been reduced to triangular form. Use the 1 in the fourth row to change the 3
and –2 above it to zeros. That is, replace R2 by R2 + (-3)R4 and replace R1 by R1 + (2)R4. For the
final step, replace R1 by R1 + (-3)R2.
ª1
«0
«
«0
«
¬0
3
1
0
0
−2
3
0
0
1
0
0
1
−7 º ª 1
6 »» «« 0
~
2» «0
» «
−2 ¼ ¬ 0
3
1
0
0
0
0
0
0
1
0
0
1
−11º ª 1
12 »» ««0
~
2» «0
» «
−2 ¼ ¬0
0
1
0
0
0
0
0
0
1
0
0
1
−47 º
12»»
2»
»
−2 ¼
The solution set contains one solution: (–47, 12, 2, –2).
11. First, swap R1 and R2. Then replace R3 by R3 + (–2)R1. Finally, replace R3 by R3 + (1)R2.
ª0
«1
«
¬« 2
1
4
7
5
3
1
−4º ª 1
−2 »» ~ «« 0
−2 »¼ «¬ 2
4
1
7
3
5
1
−2º ª 1
−4»» ~ «« 0
−2 »¼ «¬0
4
1
−1
3
5
−5
−2 º ª 1
−4 »» ~ «« 0
2 »¼ «¬ 0
4
1
0
3
5
0
−2º
−4»»
−2»¼
The system is inconsistent, because the last row would require that 0 = –2 if there were a solution.
The solution set is empty.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3
4
CHAPTER 1
• Linear Equations in Linear Algebra
12. Replace R2 by R2 + (–2)R1 and replace R3 by R3 + (2)R1. Finally, replace R3 by R3 + (3)R2.
−5
−7
1
ª 1
« 2
«
«¬ −2
4
3
7
−3º ª 1
−2 »» ~ ««0
−1»¼ «¬0
−5
3
−9
4
−5
15
−3º ª 1
4»» ~ «« 0
−7 »¼ «¬0
−5
3
0
4
−5
0
−3º
4»»
5»¼
The system is inconsistent, because the last row would require that 0 = 5 if there were a solution.
The solution set is empty.
ª1
13. «« 2
«¬ 0
ª1
~ ««0
«¬0
ª2
14. «« 0
¬« 3
ª1
~ ««0
«¬0
−3
9
5
0
2
1
0
1
0
8º ª 1
7 »» ~ ««0
−2 »¼ «¬0
−3
5
1
−6
2
−2
0
1
6
0
1
0
8º ª 1
−2 »» ~ ««0
−1»¼ «¬0
−8º ª 1
3»» ~ ««0
−4 ¼» ¬« 3
−3
2
1
−4º ª 1
3»» ~ ««0
2 »¼ «¬0
−3
15
5
0
2
1
0
1
0
0
0
1
−3
2
−2
0
1
6
0
1
0
−3
0
1
8º ª 1
−9 »» ~ ««0
−2 »¼ «¬ 0
−3
5
15
0
1
2
8º ª 1
−2 »» ~ ««0
−9»¼ «¬ 0
0
1
0
−3
5
5
8º
−2 »»
−5»¼
0
1
0
−3
2
−5
−4º
3»»
−10¼»
5º
3»» . The solution is (5, 3, –1).
−1»¼
−4º ª 1
3»» ~ ««0
−4¼» ¬«0
−4º ª 1
−1»» ~ ««0
2 »¼ «¬0
−3
2
7
0
1
6
0
1
0
0
0
1
−4º ª 1
3»» ~ «« 0
8¼» ¬« 0
2º
−1»» . The solution is (2, –1, 2).
2 »¼
15. First, replace R3 by R3 + (1)R1, then replace R4 by R4 + (1)R2, and finally replace R4 by R4 + (–
1)R3.
0 0 5º ª 1 −6
0 0 5º ª 1 −6
0 0 5º ª 1 −6
0 0
5º
ª 1 −6
« 0
»
«
»
«
»
«
1 −4 1 0 » « 0
1 −4 1 0 » «0
1 −4 1 0 » « 0
1 −4 1
0 »»
«
~
~
~
« −1
6
1 5 3» « 0
0
1 5 8» «0
0
1 5 8» « 0
0
1 5
8»
«
» «
» «
» «
»
5 4 0 ¼ ¬ 0 −1
5 4 0 ¼ ¬0
0
1 5 0¼ ¬0
0
0 0 −8¼
¬ 0 −1
The system is inconsistent, because the last row would require that 0 = –8 if there were a solution.
16. First replace R4 by R4 + (3/2)R1 and replace R4 by R4 + (–2/3)R2. (One could also scale R1 and R2
before adding to R4, but the arithmetic is rather easy keeping R1 and R2 unchanged.) Finally, replace
R4 by R4 + (–1)R3.
ª 2 0 0 −4 −10 º ª 2 0 0 −4 −10 º ª 2 0 0 −4 −10º ª 2 0 0 −4 −10 º
« 0 3 3
0
0 »» «« 0 3 3
0
0»» «« 0 3 3
0
0»» «« 0 3 3
0
0 »»
«
~
~
~
« 0 0 1
−1» « 0 0 1
−1» « 0 0 1
−1» « 0 0 1
−1»
4
4
4
4
«
» «
» «
» «
»
−9 ¼
1
5¼ ¬ 0 2 3 −5 −10¼ ¬ 0 0 1 −5 −10¼ ¬ 0 0 0 −9
¬ −3 2 3
The system is now in triangular form and has a solution. In fact, using the argument from Example 2,
one can see that the solution is unique.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.1
• Solutions
5
17. Row reduce the augmented matrix corresponding to the given system of three equations:
3 −1º ª 2
3 −1º ª 2
3 −1º
ª2
«6
»
«
»
«
5
0 » ~ « 0 −4
3» ~ « 0 −4
3»»
«
«¬ 2 −5
7 »¼ «¬ 0 −8
8»¼ «¬ 0
0
2»¼
The third equation, 0 = 2, shows that the system is inconsistent, so the three lines have no point in
common.
18. Row reduce the augmented matrix corresponding to the given system of three equations:
ª2
«0
«
¬« 2
4
1
3
4
−2
0
4º ª2
−2»» ~ «« 0
0¼» ¬« 0
4
1
−1
4
−2
−4
4º ª 2
−2»» ~ «« 0
−4¼» ¬« 0
4
1
0
4
−2
−6
4º
−2»»
−6¼»
The system is consistent, and using the argument from Example 2, there is only one solution. So the
three planes have only one point in common.
4º
h
ª 1 h 4 º ª1
19. «
~«
»
» Write c for 6 – 3h. If c = 0, that is, if h = 2, then the system has no
¬3 6 8¼ ¬ 0 6 − 3h −4 ¼
solution, because 0 cannot equal –4. Otherwise, when h ≠ 2, the system has a solution.
h −5 º ª 1
h
−5 º
ª1
Write c for −8 − 2h. If c = 0, that is, if h = –4, then the system
20. «
~«
»
6 ¼ ¬ 0 −8 − 2h 16 »¼
¬ 2 −8
has no solution, because 0 cannot equal 16. Otherwise, when h ≠ –4, the system has a solution.
4
−2 º
ª 1 4 −2 º ª 1
21. «
~«
»
» Write c for h − 12 . Then the second equation cx2 = 0 has a solution
¬3 h −6 ¼ ¬ 0 h − 12 0 ¼
for every value of c. So the system is consistent for all h.
ª −4
hº «
~
−3»¼ « 0
¬
and only if h = 6.
ª −4
22. «
¬ 2
12
−6
12
0
º
h
h »» The system is consistent if and only if −3 + = 0, that is, if
2
−3 +
2¼
h
23. a. True. See the remarks following the box titled Elementary Row Operations.
b. False. A 5 × 6 matrix has five rows.
c. False. The description applies to a single solution. The solution set consists of all possible
solutions. Only in special cases does the solution set consist of exactly one solution. Mark a
statement True only if the statement is always true.
d. True. See the box before Example 2.
24. a. False. The definition of row equivalent requires that there exist a sequence of row operations that
transforms one matrix into the other.
b. True. See the box preceding the subsection titled Existence and Uniqueness Questions.
c. False. The definition of equivalent systems is in the second paragraph after equation (2).
d. True. By definition, a consistent system has at least one solution.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6
CHAPTER 1
ª 1
25. «« 0
¬« −2
• Linear Equations in Linear Algebra
−4
3
5
−4
3
−3
g º ª1
h »» ~ «« 0
k ¼» ¬«0
7
−5
−9
7
−5
5
−4
3
0
g º ª1
h »» ~ «« 0
k + 2 g ¼» ¬« 0
7
−5
0
g
º
»
h
»
k + 2 g + h ¼»
Let b denote the number k + 2g + h. Then the third equation represented by the augmented matrix
above is 0 = b. This equation is possible if and only if b is zero. So the original system has a solution
if and only if k + 2g + h = 0.
26. Row reduce the augmented matrix for the given system:
ª2
«c
¬
4
d
f º ª1
~
g »¼ «¬ c
2
d
f / 2 º ª1
~
g »¼ «¬0
2
º
g − c ( f / 2) »¼
f /2
d − 2c
This shows that d – 2c must be nonzero, since f and g are arbitary. Otherwise, for some choices of f
and g the second row would correspond to an equation of the form 0 = b, where b is nonzero. Thus
d 2c.
27. Row reduce the augmented matrix for the given system. Scale the first row by 1/a, which is possible
since a is nonzero. Then replace R2 by R2 + (–c)R1.
ªa
«
¬c
b
d
f º ª1
~
g ¼» ¬« c
f / a º ª1
~
g ¼» ¬« 0
b/a
d
º
g − c( f / a ) ¼»
b/a
f /a
d − c(b / a )
The quantity d – c(b/a) must be nonzero, in order for the system to be consistent when the quantity
g – c( f /a) is nonzero (which can certainly happen). The condition that d – c(b/a) ≠ 0 can also be
written as ad – bc ≠ 0, or ad ≠ bc.
28. A basic principle of this section is that row operations do not affect the solution set of a linear
system. Begin with a simple augmented matrix for which the solution is obviously (3, –2, –1), and
then perform any elementary row operations to produce other augmented matrices. Here are three
examples. The fact that they are all row equivalent proves that they all have the solution set (3, –2, –
1).
ª1
«
«0
«¬0
0
1
0
0
0
1
3º ª 1
−2»» ~ «« 2
−1»¼ «¬ 0
0
1
0
0
0
1
3º ª 1
4»» ~ «« 2
−1»¼ «¬ 2
0
1
0
0
0
1
3º
4»»
5»¼
29. Swap R1 and R3; swap R1 and R3.
30. Multiply R3 by –1/5; multiply R3 by –5.
31. Replace R3 by R3 + (–4)R1; replace R3 by R3 + (4)R1.
32. Replace R3 by R3 + (–4)R2; replace R3 by R3 + (4)R2.
33. The first equation was given. The others are:
T2 = (T1 + 20 + 40 + T3 )/4,
or
4T2 − T1 − T3 = 60
T3 = (T4 + T2 + 40 + 30)/4,
or
4T3 − T4 − T2 = 70
T4 = (10 + T1 + T3 + 30)/4,
or
4T4 − T1 − T3 = 40
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.1
• Solutions
7
Rearranging,
4T1
−
T2
−T1
+
4T2
−T2
−
−T1
T4
= 30
− T3
+ 4T3
−
T4
= 60
= 70
−
+ 4T4
= 40
T3
34. Begin by interchanging R1 and R4, then create zeros in the first column:
ª 4
« −1
«
« 0
«
¬ −1
−1
0
−1
4
−1
−1
4
0
−1
0
−1
4
30 º ª −1
60 » « −1
»~«
70 » « 0
» «
40 ¼ ¬ 4
0
−1
4
4
−1
−1
4
0
−1
−1
0
−1
40 º ª −1
60 » « 0
»~«
70 » « 0
» «
30 ¼ ¬ 0
0
−1
4
−1
0
4
−1
−4
40 º
20 »
−4
»
−1 70 »
»
15 190 ¼
4
Scale R1 by –1 and R2 by 1/4, create zeros in the second column, and replace R4 by R4 + R3:
ª1
«0
«
~
«0
«
¬0
0
1
−4
1
−1
0
4
−1
−1
−1
−4
15
−40 º ª 1
5» « 0
»~«
70 » « 0
» «
190 ¼ ¬ 0
0
1
−4
1
0
0
4
−1
−2
0
−4
14
−40 º ª 1
5» « 0
»~«
75» « 0
» «
195¼ ¬ 0
0
1
−4
1
0
0
4
−1
−2
0
0
12
−40 º
5»
»
75»
»
270 ¼
Scale R4 by 1/12, use R4 to create zeros in column 4, and then scale R3 by 1/4:
ª1
«0
~«
«0
«
¬0
0
1
−4
1
0
0
4
−1
−2
0
0
1
−40 º ª 1
5» «0
»~«
75» «0
» «
22.5¼ ¬0
0
1
0
1
0
0
4
0
0
0
0
1
50 º ª 1
27.5» « 0
»~«
120 » «0
» «
22.5¼ ¬0
0
1
0
1
0
0
1
0
0
0
0
1
50 º
27.5»
»
30 »
»
22.5¼
The last step is to replace R1 by R1 + (–1)R3:
ª1
«0
~«
«0
«
¬0
0
0
0
1
0
0
1
0
0
0
0
1
20.0 º
27.5»
» . The solution is (20, 27.5, 30, 22.5).
30.0 »
»
22.5¼
Notes: The Study Guide includes a “Mathematical Note” about statements, “If … , then … .”
This early in the course, students typically use single row operations to reduce a matrix. As a result,
even the small grid for Exercise 34 leads to about 80 multiplications or additions (not counting operations
with zero). This exercise should give students an appreciation for matrix programs such as MATLAB.
Exercise 14 in Section 1.10 returns to this problem and states the solution in case students have not
already solved the system of equations. Exercise 31 in Section 2.5 uses this same type of problem in
connection with an LU factorization.
For instructors who wish to use technology in the course, the Study Guide provides boxed MATLAB
notes at the ends of many sections. Parallel notes for Maple, Mathematica, and the TI-83+/84+/89
calculators appear in separate appendices at the end of the Study Guide. The MATLAB box for Section
1.1 describes how to access the data that is available for all numerical exercises in the text. This feature
has the ability to save students time if they regularly have their matrix program at hand when studying
linear algebra. The MATLAB box also explains the basic commands replace, swap, and scale.
These commands are included in the text data sets, available from the text web site,
www.pearsonhighered.com/lay.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
8
CHAPTER 1
1.2
• Linear Equations in Linear Algebra
SOLUTIONS
Notes: The key exercises are 1–20 and 23–28. (Students should work at least four or five from Exercises
7–14, in preparation for Section 1.5.)
1. Reduced echelon form: a and b. Echelon form: d. Not echelon: c.
2. Reduced echelon form: a. Echelon form: b and d. Not echelon: c.
ª1
3. «« 2
«¬ 3
2
4
6
4
6
9
ª1
~ ««0
«¬0
ª1
4. «« 2
«¬ 4
2
4
5
8º ª 1
8»» ~ ««0
12 »¼ «¬0
ª1
~ ««0
«¬0
8º ª 1
4 »» ~ ««0
0 »¼ «¬0
2
0
0
5 º ª1
4 »» ~ ««0
2 »¼ «¬0
2
0
−3
4
−3
−12
2
1
0
4
4
1
5º ª 1
6»» ~ «« 0
2 »¼ «¬0
º
0 »¼
ª
5. «
¬0
* º ª
,
»¼ «¬ 0
*º ª0
,
0 »¼ «¬ 0
ª1
7. «
¬3
3
9
7º ª 1
~
6 »¼ «¬ 0
4
7
4
−2
−3
4
1
0
2
0
0
4
5
4
2
0
0
3
0
2
1
0
8º ª 1
−8»» ~ ««0
−12 »¼ «¬0
0
1
0
2
0
0
−8º
4»» . Pivot cols 1 and 3.
0»¼
5 º ª1
−6 »» ~ ««0
−18»¼ «¬0
4
0
1
2
−3
0
5º ª 1
−2 »» ~ ««0
2 »¼ «¬0
7º ª 1
~
−15»¼ «¬0
Corresponding system of equations:
x1
3
0
* º ª
»» , «« 0
0 ¼» ¬« 0
+ 3 x2
x3
2
4
6
*º ª0
0 »» , ««0
0¼» ¬«0
7º ª 1
~
3»¼ «¬ 0
8º
8»»
12 »¼
4
6
9
1º
Pivot cols
−2 »» .
1, 2, and 3
2 »¼
0
0
1
4
1
ª1
«2
«
«¬ 3
5 º ª1
− 18»» ~ «« 0
−6 »¼ «¬ 0
4
− 12
−3
0
1
0
ª
6. «« 0
¬« 0
4
−5
8º
4»»
−12»¼
4
1
−3
2
1
0
4
4
−3
ª1
«2
«
«¬ 4
2
4
5
5º
6»»
−6»¼
4
5
4
5º
4 »»
2 »¼
º
0 »»
0 »¼
3
0
0
1
−5 º
3»¼
= −5
=
3
The basic variables (corresponding to the pivot positions) are x1 and x3. The remaining variable x2 is
free. Solve for the basic variables in terms of the free variable. The general solution is
x1 = −5 − 3 x2
°
® x2 is free
°x = 3
¯ 3
Note: Exercise 7 is paired with Exercise 10.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.2
8.
ª 1
« −3
¬
−3
−5 º ª 1 −3 0 −5 º ª 1
~
~
−2 0 −6 »¼ «¬ 0
7 0
1 0
3»¼ «¬ 0
x
= 4
Corresponding system of equations: 1
x2
= 3
0
−5 º ª 1
~
9 »¼ «¬ 0
−3
0
0
0
1
0
• Solutions
9
4º
3»¼
The basic variables (corresponding to the pivot positions) are x1 and x2. The remaining variable x3 is
free. Solve for the basic variables in terms of the free variable. In this particular problem, the basic
variables do not depend on the value of the free variable.
x1 = 4
°
General solution: ® x2 = 3
° x is free
¯ 3
Note: A common error in Exercise 8 is to assume that x3 is zero. To avoid this, identify the basic
variables first. Any remaining variables are free. (This type of computation will arise in Chapter 5.)
ª0
9. «
¬1
1
−2
−3
4
3º ª 1
~
−6 »¼ «¬0
Corresponding system:
−3
4
1
−2
x1
x2
−6 º ª 1
~
3»¼ «¬ 0
− 2 x3
= 3
− 2 x3
= 3
0
−2
1
−2
3º
3»¼
x1 = 3 + 2 x3
°
Basic variables: x1, x2; free variable: x3. General solution: ® x2 = 3 + 2 x3
° x is free
¯ 3
ª 1
10. «
¬ −2
−2
−1
4
−5
4º ª 1
~
6 »¼ «¬0
Corresponding system:
x1
−2
−1
0
−7
4º ª 1
~
14 »¼ «¬ 0
− 2 x2
= 2
x3
−2
0
0
1
2º
−2 »¼
= −2
x1 = 2 + 2 x2
°
Basic variables: x1, x3; free variable: x2. General solution: ® x2 is free
° x = −2
¯ 3
ª3
11. ««9
«¬6
−2
−6
−4
4
12
8
−2
0
0
0º ª 3
0 »» ~ ««0
0 »¼ «¬0
x1
Corresponding system:
−
4
0
0
0º ª 1
0»» ~ «« 0
0»¼ «¬ 0
2
x2
3
+
4
x3
3
0
0
−2 3
0
0
43
0
0
0º
0»»
0»¼
= 0
= 0
= 0
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
10
CHAPTER 1
• Linear Equations in Linear Algebra
2
4
° x1 = 3 x2 − 3 x3
°
Basic variable: x1; free variables x2, x3. General solution: ® x2 is free
° x is free
° 3
¯
12. Since the bottom row of the matrix is equivalent to the equation 0 = 1, the system has no solutions.
ª1
«0
13. «
«0
«
¬0
−3
0
−1
0
1
0
0
0
0
1
−4
9
0
0
0
0
−2 º ª 1
1» «0
»~«
4» «0
» «
0¼ ¬0
−3
0
0
9
1
0
0
0
0
1
−4
9
0
0
0
0
x1
x2
Corresponding system:
x4
2º ª 1
1» « 0
»~«
4» «0
» «
0¼ ¬0
− 3 x5
= 5
− 4 x5
=
0
0
0
−3
1
0
0
0
0
1
−4
9
0
0
0
0
5º
1»
»
4»
»
0¼
1
+ 9 x5 = 4
0 = 0
x1 = 5 + 3x5
° x = 1 + 4x
5
°° 2
Basic variables: x1, x2, x4; free variables: x3, x5. General solution: ® x3 is free
°x = 4 − 9x
5
° 4
°̄ x5 is free
Note: The Study Guide discusses the common mistake x3 = 0.
ª1
«0
14. «
«0
«
¬0
0
1
−5
4
0
−1
−8
0
0
0
0
0
0
0
1
0
3º ª 1
6 »» ««0
~
0 » «0
» «
0 ¼ ¬0
−5
4
0
−1
0
0
0
0
0
0
0
0
1
0
− 5 x3
x1
Corresponding system:
0
1
x2
+ 4 x3
3º
6 »»
0»
»
0¼
= 3
−
x4
= 6
x5 = 0
0 = 0
x1 = 3 + 5 x3
°x = 6 − 4x + x
3
4
°° 2
Basic variables: x1, x2, x5; free variables: x3, x4. General solution: ® x3 is free
° x is free
° 4
°̄ x5 = 0
15. a. The system is consistent. There are many solutions because x3 is a free variable.
b. The system is consistent. There are many solutions because x1 is a free variable.
16. a. The system is inconsistent. (The rightmost column of the augmented matrix is a pivot column).
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.2
• Solutions
11
b. The system is consistent. There are many solutions because x2 is a free variable.
4 º
ª 1 −1 4 º ª 1 −1
17. «
The system has a solution for all values of h since the augmented
~«
»
3 h ¼ ¬0 1 h + 8»¼
¬ −2
column cannot be a pivot column.
1º ª 1
−3
1º
ª 1 −3
18. «
If 3h + 6 is zero, that is, if h = –2, then the system has a
~«
»
6 −2 ¼ ¬ 0 3h + 6 − h − 2 »¼
¬h
solution, because 0 equals 0. When h ≠ −2, the system has a solution since the augmented column
cannot be a pivot column. Thus the system has a solution for all values of h.
2 º
h
ª 1 h 2 º ª1
19. «
~«
»
»
¬ 4 8 k ¼ ¬ 0 8 − 4 h k − 8¼
a. When h = 2 and k ≠ 8, the augmented column is a pivot column, and the system is inconsistent.
b. When h ≠ 2, the system is consistent and has a unique solution. There are no free variables.
c. When h = 2 and k = 8, the system is consistent and has many solutions.
−3
1 º
ª 1 −3 1º ª1
20. «
~«
»
h k ¼ ¬ 0 h + 6 k − 2 »¼
¬2
a. When h = –6 and k ≠ 2, the system is inconsistent, because the augmented column is a pivot
column.
b. When h ≠ −6, the system is consistent and has a unique solution. There are no free variables.
c. When h = –6 and k = 2, the system is consistent and has many solutions.
21. a.
b.
c.
d.
e.
False. See Theorem 1.
False. See the second paragraph of the section.
True. Basic variables are defined after equation (4).
True. This statement is at the beginning of Parametric Descriptions of Solution Sets.
False. The row shown corresponds to the equation 5x4 = 0, which does not by itself lead to a
contradiction. So the system might be consistent or it might be inconsistent.
22. a. True. See Theorem 1.
b. False. See Theorem 2.
c. False. See the beginning of the subsection Pivot Positions. The pivot positions in a matrix are
determined completely by the positions of the leading entries in the nonzero rows of any echelon
form obtained from the matrix.
d. True. See the paragraph just before Example 4.
e. False. The existence of at least one solution is not related to the presence or absence of free
variables. If the system is inconsistent, the solution set is empty. See the solution of Practice
Problem 2.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
12
CHAPTER 1
• Linear Equations in Linear Algebra
23. Since there are four pivots (one in each row), the augmented matrix must reduce to the form
ª1
«0
«
«0
«
¬0
0
1
0
0
0
0
0
0
1
0
0
1
x1
aº
»
b»
and so
c»
»
d¼
x2
x3
x4
=
=
=
a
b
c
=
d
No matter what the values of a, b, c, and d, the solution exists and is unique.
24. The system is consistent because there is not a pivot in column 5, which means that there is not a row
of the form [0 0 0 0 1]. Since the matrix is the augmented matrix for a system, Theorem 2 shows
that the system has a solution.
25. If the coefficient matrix has a pivot position in every row, then there is a pivot position in the bottom
row, and there is no room for a pivot in the augmented column. So, the system is consistent, by
Theorem 2.
26. Since the coefficient matrix has three pivot columns, there is a pivot in each row of the coefficient
matrix. Thus the augmented matrix will not have a row of the form [0 0 0 0 0 1], and the
system is consistent.
27. “If a linear system is consistent, then the solution is unique if and only if every column in the
coefficient matrix is a pivot column; otherwise there are infinitely many solutions. ”
This statement is true because the free variables correspond to nonpivot columns of the coefficient
matrix. The columns are all pivot columns if and only if there are no free variables. And there are no
free variables if and only if the solution is unique, by Theorem 2.
28. Every column in the augmented matrix except the rightmost column is a pivot column, and the
rightmost column is not a pivot column.
29. An underdetermined system always has more variables than equations. There cannot be more basic
variables than there are equations, so there must be at least one free variable. Such a variable may be
assigned infinitely many different values. If the system is consistent, each different value of a free
variable will produce a different solution, and the system will not have a unique solution. If the
system is inconsistent, it will not have any solution.
30. Example:
x1
2 x1
+
x2
+ 2 x2
+
x3
+ 2 x3
= 4
= 5
31. Yes, a system of linear equations with more equations than unknowns can be consistent.
x1 +
x2 = 2
Example (in which x1 = x2 = 1): x1 −
x2 = 0
3 x1 + 2 x2 = 5
32. According to the numerical note in Section 1.2, when n = 20 the reduction to echelon form takes
about 2(20)3/3 5,333 flops, while further reduction to reduced echelon form needs at most (20)2 =
400 flops. Of the total flops, the “backward phase” is about 400/5733 = .07 or about 7%. When n =
200, the estimates are 2(200)3/3 5,333,333 flops for the reduction to echelon form and (200)2 =
40,000 flops for the backward phase. The fraction associated with the backward phase is about
(4×104) /(5.3×106) = .007, or about .7%.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.2
• Solutions
13
33. For a quadratic polynomial p(t) = a0 + a1t + a2t2 to exactly fit the data (1, 6), (2, 15), and (3, 28), the
coefficients a0, a1, a2 must satisfy the systems of equations given in the text. Row reduce the
augmented matrix:
ª1
«1
«
«¬1
1
2
3
ª1
~ ««0
¬«0
6º ª 1
15»» ~ ««0
28»¼ «¬0
1
4
9
1
1
0
0
0
1
1
1
2
4º ª 1
3»» ~ ««0
2 ¼» ¬«0
6º ª 1
9»» ~ ««0
22»¼ «¬0
1
3
8
0
1
0
1
1
0
1
3
2
6º ª 1
9»» ~ «« 0
4»¼ «¬ 0
1
1
0
1
3
1
6º
9»»
2»¼
1º
3»»
2 »¼
0
0
1
The polynomial is p(t) = 1 + 3t + 2t2.
34. [M] The system of equations to be solved is:
a0
+
a1 ⋅ 0
+
a2 ⋅ 02
+
a3 ⋅ 03
+
a4 ⋅ 0 4
+
a5 ⋅ 05
=
a0
+
a1 ⋅ 2
+
a2 ⋅ 22
+
a3 ⋅ 23
+
a4 ⋅ 2 4
+
a5 ⋅ 25
= 2.90
2
+
3
a3 ⋅ 4
+
a4 ⋅ 4
4
+
a5 ⋅ 4
5
= 14.8
+
a3 ⋅ 63
+
a4 ⋅ 6 4
+
a5 ⋅ 65
= 39.6
+
3
+
4
+
5
= 74.3
a0
+
a1 ⋅ 4
+
a2 ⋅ 4
a0
+
a1 ⋅ 6
+
a2 ⋅ 62
+
2
a0
a0
+
a1 ⋅ 8
a2 ⋅ 8
+ a1 ⋅ 10 + a2 ⋅ 10
2
a3 ⋅ 8
+ a3 ⋅ 10
3
a4 ⋅ 8
+ a4 ⋅ 10
4
a5 ⋅ 8
5
+ a5 ⋅ 10
=
0
119
The unknowns are a0, a1, …, a5. Use technology to compute the reduced echelon of the augmented
matrix:
ª1 0
«1 2
«
«1 4
«
«1 6
«1 8
«
¬«1 10
ª1
«0
«
«0
~«
«0
«0
«
«¬ 0
0
4
0
8
0
16
0
32
16
64
256
1024
36
64
216
512
1296
4096
7776
32768
102
103
104
105
0
0
0
0
0
2
4
8
16
32
0
0
8
0
48
48
224
576
960
4800
0
0
192
2688
26880
0
0
480
7680
90240
0º ª 1
2.9 » « 0
» «
14.8» « 0
»~«
39.6 » « 0
74.3» « 0
» «
119 ¼» «¬ 0
0º ª 1
2.9 » « 0
» «
9» «0
»~«
3.9 » « 0
8.7 » « 0
» «
14.5»¼ «¬ 0
0
2
0
4
0
8
0
16
0
32
0
8
48
224
960
0
0
0
24
48
80
192
480
960
1248
4032
9920
7680
32640
99840
0
0
0
0
0
2
4
8
16
32
0
0
8
0
48
48
224
576
960
4800
0
0
0
384
7680
0
0
0
1920
42240
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0º
2.9 »
»
9»
»
30.9 »
62.7 »
»
104.5»¼
0º
2.9 »
»
9»
»
3.9 »
−6.9 »
»
−24.5¼»
14
CHAPTER 1
ª1
«0
«
«0
~«
«0
«0
«
¬« 0
• Linear Equations in Linear Algebra
0
0
0
0
0
2
4
8
16
32
0
0
8
0
48
48
224
576
960
4800
0
0
0
384
7680
0
0
0
0
3840
ª1
«0
«
«0
~«
«0
«0
«
¬« 0
0º ª 1
2.9 » « 0
» «
9 » «0
»~«
3.9 » « 0
−6.9 » « 0
» «
10 ¼» ¬« 0
0
2
0
4
0
8
0
16
0
0
0
8
48
224
0
0
0
0
0
48
0
576
384
0
0
0
0
0
0
1
0
0
0
0
0
2
4
8
16
32
0
0
8
0
48
48
224
576
960
4800
0
0
0
384
7680
0
0
0
0
1
0º
ª1
»
«0
2.8167 »
«
«0
6.5000 »
» ~"~ «
−8.6000 »
«0
«0
−26.900 »
»
«
.002604 ¼»
¬«0
0º
2.9 »
»
9»
»
3.9 »
−6.9 »
»
.0026 ¼»
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
1
0
0
0
0
0
0
1
0º
1.7125»»
−1.1948»
»
.6615»
−.0701»
»
.0026 ¼»
Thus p(t) = 1.7125t – 1.1948t2 + .6615t3 – .0701t4 + .0026t5, and p(7.5) = 64.6 hundred lb.
Notes: In Exercise 34, if the coefficients are retained to higher accuracy than shown here, then p(7.5) =
64.8. If a polynomial of lower degree is used, the resulting system of equations is overdetermined. The
augmented matrix for such a system is the same as the one used to find p, except that at least column 6 is
missing. When the augmented matrix is row reduced, the sixth row of the augmented matrix will be
entirely zero except for a nonzero entry in the augmented column, indicating that no solution exists.
Exercise 34 requires 25 row operations. It should give students an appreciation for higher-level
commands such as gauss and bgauss, discussed in Section 1.4 of the Study Guide. The command
ref (reduced echelon form) is available, but I recommend postponing that command until Chapter 2.
The Study Guide includes a “Mathematical Note” about the phrase, “If and only if,” used in Theorem
2.
1.3
SOLUTIONS
Notes: The key exercises are 11–16, 19–22, 25, and 26. A discussion of Exercise 25 will help students
understand the notation [a1 a2 a3], {a1, a2, a3}, and Span{a1, a2, a3}.
ª −1º ª −3º ª−1 + ( −3) º ª −4 º
1. u + v = « » + « » = «
»=« ».
¬ 2 ¼ ¬ −1¼ ¬ 2 + ( −1) ¼ ¬ 1¼
Using the definitions carefully,
ª −1º
ª −3º ª −1º ª ( −2)(−3) º ª −1 + 6 º ª 5º
u − 2 v = « » + ( −2) « » = « » + «
»=«
» = « » , or, more quickly,
¬ 2¼
¬ −1¼ ¬ 2 ¼ ¬ ( −2)( −1) ¼ ¬ 2 + 2 ¼ ¬ 4 ¼
ª −1º
ª −3º ª −1 + 6 º ª 5 º
u − 2v = « » − 2 « » = «
» = « » . The intermediate step is often not written.
¬ 2¼
¬ −1¼ ¬ 2 + 2 ¼ ¬ 4 ¼
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.3
• Solutions
15
ª 3º ª 2 º ª 3 + 2 º ª 5 º
2. u + v = « » + « » = «
» =« ».
¬ 2 ¼ ¬ −1¼ ¬ 2 − 1 ¼ ¬ 1¼
Using the definitions carefully,
ª 3º
ª 2 º ª 3º ª ( −2)(2) º ª 3 − 4 º ª −1º
u − 2 v = « » + ( −2) « » = « » + «
»=«
» = « » , or, more quickly,
¬2¼
¬ −1¼ ¬ 2 ¼ ¬ ( −2)( −1) ¼ ¬ 2 + 2 ¼ ¬ 4 ¼
ª 3º
ª 2 º ª 3 − 4 º ª −1º
u − 2v = « » − 2 « » = «
» = « » . The intermediate step is often not written.
¬2¼
¬ −1¼ ¬ 2 + 2 ¼ ¬ 4 ¼
3.
4.
ª 3º
ª 5º ª 2 º
ª 3 x1 º ª 5 x2 º ª 2º
«
»
«
»
«
»
5. x1 « −2 » + x2 « 0» = « −3» , «« −2 x1 »» + «« 0 »» = «« −3»» ,
«¬ 8»¼
«¬ −9»¼ «¬ 8»¼
«¬ 8 x1 »¼ «¬ −9 x2 »¼ «¬ 8»¼
3 x1 + 5 x2 = 2
−2 x1
= −3
8 x1 − 9 x2 = 8
ª3 x1 + 5 x2 º ª 2º
« −2 x » = « −3»
1 »
«
« »
«¬8 x1 − 9 x2 »¼ «¬ 8»¼
Usually the intermediate steps are not displayed.
ª 3º
ª7 º
ª −2 º ª 0 º
6. x1 « » + x2 « » + x3 « » = « » ,
¬ −2 ¼
¬ 3¼
¬ 1¼ ¬ 0 ¼
3 x2 + 7 x2 − 2 x3 =
−2 x1 + 3 x2 +
x3 =
ª 3x1 º ª7 x2 º ª −2 x3 º ª0º
«
»+« »+«
»=« »,
¬ −2 x1 ¼ ¬ 3x2 ¼ ¬ x3 ¼ ¬0¼
0
0
ª3x1 + 7 x2 − 2 x3 º ª0º
«
»=« »
¬ −2 x1 + 3x2 + x3 ¼ ¬0¼
Usually the intermediate steps are not displayed.
7. See the figure below. Since the grid can be extended in every direction, the figure suggests that every
vector in R2 can be written as a linear combination of u and v.
To write a vector a as a linear combination of u and v, imagine walking from the origin to a along
the grid "streets" and keep track of how many "blocks" you travel in the u-direction and how many in
the v-direction.
a. To reach a from the origin, you might travel 1 unit in the u-direction and –2 units in the vdirection (that is, 2 units in the negative v-direction). Hence a = u – 2v.
b. To reach b from the origin, travel 2 units in the u-direction and –2 units in the v-direction. So
b = 2u – 2v. Or, use the fact that b is 1 unit in the u-direction from a, so that
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
16
CHAPTER 1
• Linear Equations in Linear Algebra
b = a + u = (u – 2v) + u = 2u – 2v
c. The vector c is –1.5 units from b in the v-direction, so
c = b – 1.5v = (2u – 2v) – 1.5v = 2u – 3.5v
d. The “map” suggests that you can reach d if you travel 3 units in the u-direction and –4 units in
the v-direction. If you prefer to stay on the paths displayed on the map, you might travel from the
origin to –3v, then move 3 units in the u-direction, and finally move –1 unit in the v-direction. So
d = –3v + 3u – v = 3u – 4v
Another solution is
d = b – 2v + u = (2u – 2v) – 2v + u = 3u – 4v
d
b
c
u
0
–v
–2v
2v
v
a
w
y
–u
x
z
Figure for Exercises 7 and 8
8. See the figure above. Since the grid can be extended in every direction, the figure suggests that every
vector in R2 can be written as a linear combination of u and v.
w. To reach w from the origin, travel –1 units in the u-direction (that is, 1 unit in the negative
u-direction) and travel 2 units in the v-direction. Thus, w = (–1)u + 2v, or w = 2v – u.
x. To reach x from the origin, travel 2 units in the v-direction and –2 units in the u-direction. Thus,
x = –2u + 2v. Or, use the fact that x is –1 units in the u-direction from w, so that
x = w – u = (–u + 2v) – u = –2u + 2v
y. The vector y is 1.5 units from x in the v-direction, so
y = x + 1.5v = (–2u + 2v) + 1.5v = –2u + 3.5v
z. The map suggests that you can reach z if you travel 4 units in the v-direction and –3 units in the
u-direction. So z = 4v – 3u = –3u + 4v. If you prefer to stay on the paths displayed on the “map,”
you might travel from the origin to –2u, then 4 units in the v-direction, and finally move –1 unit
in the u-direction. So
z = –2u + 4v – u = –3u + 4v
x2
+ 6 x2
+ 5 x3
− x3
= 0
= 0,
+
− 8 x3
= 0
ª x2 + 5 x3 º ª0 º
« 4 x + 6 x − x » = «0»
2
3 »
« 1
« »
«¬ − x1 + 3 x2 − 8 x3 »¼ «¬0 »¼
ª 0º ª x2 º ª 5 x3 º ª 0º
« 4 x » + «6 x » + « − x » = « 0» ,
3»
« 1» « 2 » «
« »
«¬ − x1 »¼ «¬ 3 x2 »¼ «¬ −8 x3 »¼ «¬ 0»¼
ª 0º
ª 1º
ª 5º ª 0 º
«
»
«
»
x1 « 4 » + x2 «6» + x3 «« −1»» = ««0 »»
«¬ −1»¼
«¬ 3»¼
«¬ −8»¼ «¬0 »¼
9. 4 x1
− x1
3 x2
Usually, the intermediate calculations are not displayed.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.3
• Solutions
17
Note: The Study Guide says, “Check with your instructor whether you need to “show work” on a
problem such as Exercise 9.”
3 x1
10. −2 x1
− 2 x2
− 7 x2
+ 4 x3
+ 5 x3
=
=
5 x1
+ 4 x2
−
= 2
3 x3
ª 3x1 − 2 x2 + 4 x3 º ª 3 º
« −2 x − 7 x + 5 x » = «1 »
2
3»
« 1
« »
«¬ 5 x1 + 4 x2 − 3x3 »¼ «¬ 2 »¼
3
1,
ª 3 x1 º ª −2 x2 º ª 4 x3 º ª 3º
» « »
«
» «
» «
« −2 x1 » + « −7 x2 » + « 5 x3 » = « 1» ,
«¬ 5 x1 »¼ «¬ 4 x2 »¼ «¬ −3 x3 »¼ «¬ 2 »¼
ª 3º
ª −2º
ª 4 º ª 3º
«
»
«
»
x1 « −2 » + x2 « −7 » + x3 «« 5»» = «« 1»»
«¬ 5»¼
«¬ 4»¼
«¬ −3»¼ «¬ 2 »¼
Usually, the intermediate calculations are not displayed.
11. The question
Is b a linear combination of a1, a2, and a3?
is equivalent to the question
Does the vector equation x1a1 + x2a2 + x3a3 = b have a solution?
The equation
ª 1º
ª0º
ª 5º ª 2 º
«
»
«
»
x1 −2 + x2 1 + x3 « −6 » = « −1»
« »
« »
« » « »
«¬ 0 »¼
«¬ 2 »¼
«¬ 8»¼ «¬ 6 »¼
↑
↑
↑
↑
a1
a2
a3
b
has the same solution set as the linear system whose augmented matrix is
5
2º
ª 1 0
«
M = « −2 1 −6 −1»»
«¬ 0 2
8
6 »¼
Row reduce M until the pivot positions are visible:
ª 1 0 5 2 º ª 1 0 5 2º
M ~ ««0 1 4 3»» ~ ««0 1 4 3»»
«¬0 2 8 6 »¼ «¬0 0 0 0»¼
The linear system corresponding to M has a solution, so the vector equation (*) has a solution, and
therefore b is a linear combination of a1, a2, and a3.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
(*)
18
CHAPTER 1
• Linear Equations in Linear Algebra
12. The equation
ª 1º
ª −2 º
ª −6 º ª 11º
«
»
«
»
x1 « 0 » + x2 « 3» + x3 «« 7 »» = «« −5»»
«¬ 1»¼
«¬ −2 »¼
«¬ 5»¼ «¬ 9»¼
↑
↑
↑
↑
a1
a2
a3
b
(*)
has the same solution set as the linear system whose augmented matrix is
ª1
M = ««0
«¬ 1
−2
3
−2
−6
7
5
11º
−5 »»
9 »¼
Row reduce M until the pivot positions are visible:
ª1
M ~ ««0
«¬0
−2
3
0
−6
7
11
11º
−5»»
−2 »¼
The linear system corresponding to M has a solution, so the vector equation (*) has a solution, and
therefore b is a linear combination of a1, a2, and a3.
13. Denote the columns of A by a1, a2, a3. To determine if b is a linear combination of these columns,
use the boxed fact in the subsection Linear Combinations. Row reduce the augmented matrix
[a1 a2 a3 b] until you reach echelon form:
2
3º ª 1 −4 2
3º
ª 1 −4
«
»
«
3
5 −7 » ~ «0
3 5 −7 »»
[a1 a2 a3 b] = « 0
«¬ −2
8 −4 −3»¼ «¬0
0 0
3»¼
The system for this augmented matrix is inconsistent, so b is not a linear combination of the columns
of A.
14. Row reduce the augmented matrix [a1 a2 a3 b] until you reach echelon form:
[a1 a2 a3
ª 1
b] = «« −2
«¬ 0
0
1
2
5
−6
8
2º ª 1
−1»» ~ «« 0
6 »¼ «¬0
0
1
2
5
4
8
2º ª 1
3»» ~ «« 0
6 »¼ «¬ 0
0
1
0
5
4
0
2º
3»» .
0»¼
The linear system corresponding to this matrix has a solution, so b is a linear combination of the
columns of A.
3º ª 1 −5
3 º ª 1 −5
3 º ª 1 −5
3 º
ª 1 −5
«
»
«
»
«
»
«
7 −14 » ~ «0
1
−2 » ~ «0
1 −2 »» . The vector b
15. [a1 a2 b] = « 3 −8 −5» ~ « 0
2
h ¼» ¬«0 −3 h + 3¼» ¬«0 −3 h + 3¼» ¬«0
0 h − 3¼»
¬« −1
is in Span{a1, a2} when h – 3 is zero, that is, when h = 3.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.3
h º ª 1 −2
h º ª1
ª 1 −2
«
»
«
1 −3» ~ «0
1
−3 »» ~ «« 0
16. [v1 v2 y] = « 0
«¬ −2
7 −5»¼ «¬0
3 −5 + 2h »¼ «¬0
Span{v1, v2} when 4 + 2h is zero, that is, when h = –2.
−2
1
0
• Solutions
19
h º
−3 »» . The vector y is in
4 + 2h »¼
17. Noninteger weights are acceptable, of course, but some simple choices are 0 v1 + 0 v2 = 0, and
ª 3º
1 v1 + 0 v2 = «« 1»» , 0 v1 + 1 v2 =
«¬ 2 »¼
ª −4 º
« 0» , 1 v + 1 v =
1
2
« »
«¬ 1»¼
ª −1º
« 1» , 1 v – 1 v =
1
2
« »
«¬ 3»¼
ª7 º
« 1»
« »
«¬ 1»¼
18. Some likely choices are 0 v1 + 0 v2 = 0, and
ª 1º
1 v1 + 0 v2 = «« 1»» , 0 v1 + 1 v2 =
«¬ −2 »¼
ª −2 º
« 3» , 1 v + 1 v =
1
2
« »
«¬ 0 »¼
ª −1º
« 4» , 1 v – 1 v =
1
2
« »
«¬ −2 »¼
ª 3º
« −2 »
« »
«¬ −2 »¼
19. By inspection, v2 = (3/2)v1. Any linear combination of v1 and v2 is actually just a multiple of v1. For
instance,
av1 + bv2 = av1 + b(3/2)v1 = (a + 3b/2)v1
So Span{v1, v2} is the set of points on the line through v1 and 0.
Note: Exercises 19 and 20 prepare the way for ideas in Sections 1.4 and 1.7.
20. Span{v1, v2} is a plane in R3 through the origin, because neither vector in this problem is a multiple
of the other.
h º
ªhº
ª 2 2 hº ª2 2
21. Let y = « » . Then [u v y] = «
~«
»
» . This augmented matrix
¬k ¼
¬ −1 1 k ¼ ¬ 0 2 k + h / 2 ¼
corresponds to a consistent system for all h and k. So y is in Span{u, v} for all h and k.
22. Construct any 3×4 matrix in echelon form that corresponds to an inconsistent system. Perform
sufficient row operations on the matrix to eliminate all zero entries in the first three columns.
23. a. False. The alternative notation for a (column) vector is (–4, 3), using parentheses and commas.
ª −5º
b. False. Plot the points to verify this. Or, see the statement preceding Example 3. If « » were on
¬ 2¼
ª −2 º
ª −2 º
ª −5º
the line through « » and the origin, then « » would have to be a multiple of « » , which is
¬ 5¼
¬ 5¼
¬ 2¼
not the case.
c. True. See the line displayed just before Example 4.
d. True. See the box that discusses the matrix in (5).
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
20
CHAPTER 1
• Linear Equations in Linear Algebra
e. False. The statement is often true, but Span{u, v} is not a plane when v is a multiple of u, or
when u is the zero vector.
24. a.
b.
c.
d.
e.
False. Span{u, v} can be a plane.
True. See the beginning of the subsection Vectors in Rn.
True. See the comment following the definition of Span{v1, …, vp}.
False. (u – v) + v = u – v + v = u.
False. Setting all the weights equal to zero results in a legitimate linear combination of a set of
vectors.
25. a. There are only three vectors in the set {a1, a2, a3}, and b is not one of them.
b. There are infinitely many vectors in W = Span{a1, a2, a3}. To determine if b is in W, use the
method of Exercise 13.
[a1 a2 a3
ª 1
b] = «« 0
«¬ −2
0
3
6
−4
−2
3
4º ª 1
1»» ~ ««0
−4 »¼ «¬0
0
3
6
−4
−2
−5
4º ª 1
1»» ~ ««0
4»¼ «¬0
0
3
0
−4
−2
−1
4º
1»»
2»¼
The system for this augmented matrix is consistent, so b is in W.
c. a1 = 1a1 + 0a2 + 0a3. See the discussion in the text following the definition of Span{v1, …, vp}.
0 6 10 º ª 1
0 3 5º ª 1
0
3 5º ª 1
ª 2
«
»
«
»
«
8 5
3 ~ −1
8 5 3 ~ 0
8
8 8» ~ « 0
26. a. [a1 a2 a3 b] = −1
«
» «
» «
» «
«¬ 1 −2 1 7 »¼ «¬ 1 −2 1 7 »¼ «¬0 −2 −2 2 »¼ «¬ 0
No, b is not a linear combination of the columns of A, that is, b is not in W.
b. The second column of A is in W because a2 = 0 a1 + 1 a2 + 0 a3.
5º
0
3
8
8
8»
0
0
4 »¼
»
27. a. 5v1 is the output of 5 days’ operation of mine #1.
ª 240 º
b. The total output is x1v1 + x2v2, so x1 and x2 should satisfy x1v1 + x2 v 2 = «
».
¬ 2824 ¼
40
240 º ª 1 0 1.73º
ª 30
c. [M] Reduce the augmented matrix «
»~«
».
¬600 380 2824 ¼ ¬ 0 1 4.70 ¼
Operate mine #1 for 1.73 days and mine #2 for 4.70 days. (This is an approximate solution.)
28. a. The amount of heat produced when the steam plant burns x1 tons of anthracite and x2 tons of
bituminous coal is 27.6x1 + 30.2x2 million Btu.
b. The total output produced by x1 tons of anthracite and x2 tons of bituminous coal is given by the
ª 27.6 º
ª 30.2 º
«
»
vector x1 «3100 » + x2 «« 6400»» .
«¬ 250 »¼
«¬ 360 »¼
ª 27.6 º
ª 30.2 º ª 162º
«
»
c. [M] The appropriate values for x1 and x2 satisfy x1 «3100 » + x2 ««6400»» = «« 23,610»» .
«¬ 250 »¼
«¬ 360 »¼ «¬ 1,623»¼
To solve, row reduce the augmented matrix:
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.3
ª 27.6
«3100
«
¬« 250
30.2
6400
360
162 º ª1.000
23610 »» ~ ««
0
1623¼» ¬«
0
• Solutions
3.900º
1.800»»
0»¼
0
1.000
0
The steam plant burned 3.9 tons of anthracite coal and 1.8 tons of bituminous coal.
29. The total mass is 4 + 2 + 3 + 5 = 14. So v = (4v1 +2v2 + 3v3 + 5v4)/14. That is,
§ ª 2º
ª −4 º
ª 4º
ª 1º ·
ª 8 − 8 + 12 + 5º ª 17 14º ª 1.214º
1¨ « »
¸ 1 «
«
»
«
»
«
»
v = ¨ 4 « −2 » + 2 « 2 » + 3 « 0 » + 5 « −6 » ¸ = « −8 + 4 + 0 − 30»» = «« −17 7 »» ≈ «« −2.429»»
14 ¨
¸ 14 « 16 + 6 − 6 + 0» « 8 7 » « 1.143»
¬« 3¼»
¬« −2 ¼»
¬« 0¼» ¹
¬
¼ ¬
¼ ¬
¼
© ¬« 4 ¼»
30.
Let m be the total mass of the system. By definition,
m
1
m
(m1v1 + " + mk v k ) = 1 v1 + " + k v k
m
m
m
The second expression displays v as a linear combination of v1, …, vk, which shows that v is in
Span{v1, …, vk}.
v=
31. a. The center of mass is
ª8º
ª 2 º · ª10 / 3º
1 § ª0 º
¨1 ⋅ « » + 1 ⋅ « » + 1 ⋅ « » ¸ = «
».
3 © ¬1 ¼
¬1¼
¬4¼ ¹ ¬ 2 ¼
b. The total mass of the new system is 9 grams. The three masses added, w1, w2, and w3, satisfy the
equation
ª0º
ª8º
ª 2º · ª 2º
1§
¨ ( w1 + 1) ⋅ « » + ( w2 + 1) ⋅ « » + ( w3 + 1) ⋅ « » ¸ = « »
9©
¬1 ¼
¬1¼
¬ 4¼ ¹ ¬ 2¼
which can be rearranged to
ª0 º
ª8 º
ª 2º
ª18º
¬ ¼
¬ ¼
¬ ¼
¬ ¼
( w1 + 1) ⋅ «1 » + ( w2 + 1) ⋅ «1» + ( w3 + 1) ⋅ « 4» = «18»
and
ª0º
ª8 º
ª2º ª 8 º
w1 ⋅ « » + w2 ⋅ « » + w3 ⋅ « » = « »
¬1 ¼
¬1 ¼
¬ 4 ¼ ¬12 ¼
The condition w1 + w2 + w3 = 6 and the vector equation above combine to produce a system of
three equations whose augmented matrix is shown below, along with a sequence of row
operations:
ª1 1
«0 8
«
¬« 1 1
1 6º ª 1
2
8»» ~ ««0
4 12 ¼» ¬«0
1
8
0
1
2
3
6º ª 1
8»» ~ ««0
6 ¼» ¬«0
1
8
0
1
2
1
6º
8»»
2 »¼
ª1
~ ««0
«¬0
1
8
0
0
0
1
4º ª 1
4»» ~ «« 0
2»¼ «¬0
0
8
0
0
0
1
3.5º ª 1
4»» ~ «« 0
2»¼ «¬ 0
0
1
0
0
0
1
3.5º
.5»»
2 »¼
Answer: Add 3.5 g at (0, 1), add .5 g at (8, 1), and add 2 g at (2, 4).
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
21
22
CHAPTER 1
• Linear Equations in Linear Algebra
Extra problem: Ignore the mass of the plate, and distribute 6 gm at the three vertices to make the center
of mass at (2, 2). Answer: Place 3 g at (0, 1), 1 g at (8, 1), and 2 g at (2, 4).
32. See the parallelograms drawn on the figure from the text that accompanies this exercise. Here c1, c2,
c3, and c4 are suitable scalars. The darker parallelogram shows that b is a linear combination of v1
and v2, that is
c1v1 + c2v2 + 0 v3 = b
The larger parallelogram shows that b is a linear combination of v1 and v3, that is,
c4v1 + 0 v2 + c3v3 = b
So the equation x1v1 + x2v2 + x3v3 = b has at least two solutions, not just one solution. (In fact, the
equation has infinitely many solutions.)
v3
c3 v 3
c2v2
b
v2
0
v1
c1v1
c 4v 1
33. a. For j = 1,…, n, the jth entry of (u + v) + w is (uj + vj) + wj. By associativity of addition in R, this
entry equals uj + (vj + wj), which is the jth entry of u + (v + w). By definition of equality of
vectors, (u + v) + w = u + (v + w).
b. For any scalar c, the jth entry of c(u + v) is c(uj + vj), and the jth entry of cu + cv is cuj + cvj (by
definition of scalar multiplication and vector addition). These entries are equal, by a distributive
law in R. So c(u + v) = cu + cv.
34. a. For j = 1,…, n, uj + (–1)uj = (–1)uj + uj = 0, by properties of R. By vector equality,
u + (–1)u = (–1)u + u = 0.
b. For scalars c and d, the jth entries of c(du) and (cd )u are c(duj) and (cd )uj, respectively. These
entries in R are equal, so the vectors c(du) and (cd)u are equal.
Note: When an exercise in this section involves a vector equation, the corresponding technology data (in
the data files on the web) is usually presented as a set of (column) vectors. To use MATLAB or other
technology, a student must first construct an augmented matrix from these vectors. The MATLAB note in
the Study Guide describes how to do this. The appendices in the Study Guide give corresponding
information about Maple, Mathematica, and the TI calculators.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.4
1.4
• Solutions
23
SOLUTIONS
Notes: Key exercises are 1–20, 27, 28, 31 and 32. Exercises 29, 30, 33, and 34 are harder. Exercise 34
anticipates the Invertible Matrix Theorem but is not used in the proof of that theorem.
1. The matrix-vector Ax is not defined because the number of columns (2) in the 3×2 matrix A does not
match the number of entries (3) in the vector x.
2. The matrix-vector Ax is not defined because the number of columns (1) in the 3×1 matrix A does not
match the number of entries (2) in the vector x.
2º
ª 1º
ª 2 º ª −2 º ª 6º ª 4º
ª −2 º
»
«
»
1» « » = (−2) « −3» + 3 «« 1»» = «« 6 »» + «« 3»» = «« 9»» , and
3
«¬ 1»¼
«¬ 6 »¼ «¬ −2 »¼ «¬18»¼ «¬16»¼
6»¼ ¬ ¼
ª 1
3. Ax = «« −3
«¬ 1
2º
ª 1 ⋅ (−2) + 2 ⋅ 3 º ª 4 º
ª −2 º «
»
1» « » = «(−3) ⋅ (−2) + 1 ⋅ 3»» = «« 9 »»
3
6¼» ¬ ¼ ¬« 1 ⋅ (−2) + 6 ⋅ 3 »¼ «¬16 »¼
ª 1
«
Ax = « −3
¬« 1
ª1
4. Ax = «
¬3
ª1
Ax = «
¬3
3
2
3
2
ª 1º
−4 º « »
ª 1º
ª 3º
ª −4º ª1 + 6 − 4º ª3º
2
=
1
⋅
+
2
⋅
+
1
⋅
=
=
, and
»
«
»
«
»
«
1¼ « »
3¼
2¼
1»¼ «¬ 3 + 4 + 1»¼ «¬8»¼
¬
¬
¬
«¬ 1»¼
ª 1º
−4 º « » ª1 ⋅ 1 + 3 ⋅ 2 + (−4) ⋅ 1º ª3º
2 =
=
1¼» « » ¬« 3 ⋅1 + 2 ⋅ 2 + 1 ⋅ 1 ¼» ¬«8¼»
«¬ 1»¼
5. On the left side of the matrix equation, use the entries in the vector x as the weights in a linear
combination of the columns of the matrix A:
ª 1º
ª 2º
ª −3 º
ª 1º ª −4 º
2 ⋅ « » − 1⋅ « » + 1⋅ « » − 1⋅ « » = « »
¬ −2 ¼
¬ −3 ¼
¬ 1¼
¬ −1¼ ¬ 1¼
6. On the left side of the matrix equation, use the entries in the vector x as the weights in a linear
combination of the columns of the matrix A:
ª 2º
ª −3º ª −21º
« 3»
« 2 » « 1»
»
−3 ⋅ « » + 5 ⋅ « » = «
« 8»
« −5» « −49 »
« »
« » «
»
¬ −2 ¼
¬ 1¼ ¬ 11¼
7. The left side of the equation is a linear combination of three vectors. Write the matrix A whose
columns are those three vectors, and create a variable vector x with three entries:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
24
CHAPTER 1
• Linear Equations in Linear Algebra
ª −5 º
« 3»
« »
« −5 »
« »
¬ 1¼
ª ª 4º
«« »
−1
A = «« »
«« 7»
«« »
«¬ ¬ −4 ¼
−5
ª 4
« −1
«
« 7
«
¬ −4
3
−5
1
ª 7ºº ª 4
« −8» » « −1
« »» = «
« 0» » « 7
« »» «
¬ 2 ¼ »¼ ¬ −4
−5
3
−5
1
7º
ª x1 º
−8»
» and x = « x2 » . Thus the equation Ax = b is
« »
0»
«¬ x3 »¼
»
2¼
7º
ª 6º
ª x1 º « »
»
−8 « »
−8
» x2 = « »
0» « » « 0»
» « x3 » « »
2 ¼ ¬ ¼ ¬ −7 ¼
For your information: The unique solution of this equation is (5, 7, 3). Finding the solution by hand
would be time-consuming.
Note: The skill of writing a vector equation as a matrix equation will be important for both theory and
application throughout the text. See also Exercises 27 and 28.
8. The left side of the equation is a linear combination of four vectors. Write the matrix A whose
columns are those four vectors, and create a variable vector with four entries:
ª −1º
« 5»
¬ ¼
ª ª 2º
A = «« »
¬ ¬ −4¼
ª 2
is «
¬ −4
−1
−4
5
3
ª −4º
« 3»
¬ ¼
ª 0º º ª 2
« 2» » = « −4
¬ ¼¼ ¬
−1
5
−4
3
ª z1 º
«z »
0º
2
, and z = « » . Then the equation Az = b
»
«
z 3»
2¼
« »
¬« z4 ¼»
ª z1 º
0 º «« z2 »» ª 5º
=
.
2 »¼ « z3 » «¬12 »¼
« »
¬ z4 ¼
For your information: One solution is (8, 7, 1, 3). The general solution is z1 = 37/6 + (17/6)z3 –
(1/3)z4, z2 = 22/3 +(5/3)z3 – (2/3)z4, with z3 and z4 free.
9. The system has the same solution set as the vector equation
ª 5º
ª 1º
ª −3 º ª 8 º
x1 « » + x2 « » + x3 « » = « »
¬0¼
¬2¼
¬ 4 ¼ ¬0¼
and this equation has the same solution set as the matrix equation
ª5
«0
¬
1
2
ªx º
−3º « 1 » ª 8º
x2 =
4»¼ « » «¬0»¼
«¬ x3 »¼
10. The system has the same solution set as the vector equation
ª4º
ª −1º ª 8º
«
»
x1 « 5» + x2 «« 3»» = «« 2 »»
«¬ 3»¼
«¬ −1»¼ «¬ 1»¼
and this equation has the same solution set as the matrix equation
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.4
• Solutions
25
−1º
ª 8º
ª x1 º « »
»
3» « » = « 2»
x
−1»¼ ¬ 2 ¼ «¬ 1»¼
ª4
«5
«
«¬ 3
11. To solve Ax = b, row reduce the augmented matrix [a1 a2 a3 b] for the corresponding linear
system:
ª 1
« 1
«
«¬ −3
−4
2
6
3
5
−7
−2 º ª 1
4 »» ~ ««0
12 »¼ «¬0
−4
6
−6
3
2
2
−2º ª 1
6»» ~ ««0
6 »¼ «¬0
3
1
0
−4
3
1
−2º ª 1
3»» ~ ««0
0»¼ «¬0
x1 = −11
°
3 . As a vector, the solution is x =
The solution is ® x2 =
°x =
0
¯ 3
3
1
0
0
0
1
−2º ª 1
3»» ~ «« 0
0»¼ «¬0
0
1
0
0
0
1
ª x1 º ª −11º
« x » = « 3» .
« 2» «
»
«¬ x3 »¼ «¬ 0»¼
12. To solve Ax = b, row reduce the augmented matrix [a1 a2 a3 b] for the corresponding linear
system:
ª 1
« −3
«
¬« 5
2
−4
2
−1
2
3
ª1
~ ««0
«¬0
2
1
0
−1
−1
1
x1
°
The solution is ® x2
°x
¯ 3
1º ª 1
2»» ~ ««0
−3¼» ¬«0
1º ª 1
1»» ~ ««0
3»¼ «¬0
2
2
−8
2
1
0
0
0
1
−1
−1
8
1º ª 1
5»» ~ ««0
−8¼» «¬0
4 º ª1
4»» ~ ««0
3»¼ «¬0
0
1
0
2
−8
2
0
0
1
−1
8
−1
1º ª 1
−8»» ~ ««0
5»¼ «¬0
2
1
2
−1
−1
−1
1º
1»»
5»¼
−4º
4 »»
3 »¼
= −4
= 4 . As a vector, the solution is x =
= 3
ª x1 º ª −4 º
«x » = « 4 » .
« 2» « »
«¬ x3 »¼ «¬ 3 »¼
13. The vector u is in the plane spanned by the columns of A if and only if u is a linear combination of
the columns of A. This happens if and only if the equation Ax = u has a solution. (See the box
preceding Example 3 in Section 1.4.) To study this equation, reduce the augmented matrix [A u]
1 4º ª 1
1
4º ª 1 1 4º
ª 3 −5 0 º ª 1
« −2
»
«
»
«
6 4 » ~ « −2
6 4» ~ «0
8
12»» ~ «« 0 8 12»»
«
«¬ 1
1 4 »¼ «¬ 3 −5 0»¼ «¬ 0 −8 −12»¼ «¬ 0 0
0»¼
The equation Ax = u has a solution, so u is in the plane spanned by the columns of A.
For your information: The unique solution of Ax = u is (5/2, 3/2).
14. Reduce the augmented matrix [A u] to echelon form:
ª2
«0
«
«¬ 1
5
1
2
−1
−1
0
4º ª 1
−1»» ~ «« 0
4 »¼ «¬ 2
2
1
5
0
−1
−1
4º ª 1
−1»» ~ «« 0
4»¼ «¬ 0
2
1
1
0
−1
−1
4º ª 1
−1»» ~ «« 0
−4»¼ «¬ 0
2
1
0
0
−1
0
4º
−1»»
−3»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
−11º
3»»
0»¼
26
CHAPTER 1
• Linear Equations in Linear Algebra
The equation Ax = u has no solution, so u is not in the subset spanned by the columns of A.
b1 º
ª 3 −1 b1 º
ª 3 −1
15. The augmented matrix for Ax = b is «
, which is row equivalent to «
.
»
3 b2 ¼
0 b2 + 3b1 »¼
¬ −9
¬0
This shows that the equation Ax = b is not consistent when 3b1 + b2 is nonzero. The set of b for
which the equation is consistent is a line through the origin–the set of all points (b1, b2) satisfying b2
= –3b1.
ª 1
16. Row reduce the augmented matrix [A b]: A = «« −2
«¬ 4
ª 1
« −2
«
¬« 4
−2
2
−1
−1
0
3
ª1
~ ««0
¬«0
−2
−2
0
−1
b1
º ª1
» «
−2
b2 + 2b1
» = «0
0 b3 − 4b1 + (7 2)(b2 + 2b1 ) ¼» ¬«0
b1 º ª 1
b2 »» ~ ««0
b3 ¼» ¬«0
−2
−2
7
−1
−2
7
−2
2
−1
−1º
ª b1 º
»
0 » , b = ««b2 »» .
«¬ b3 »¼
3»¼
b1 º
b2 + 2b1 »»
b3 − 4b1 ¼»
−2
−2
0
−1
b1
º
»
−2
b2 + 2b1
»
0 3b1 + (7 2) b2 + b3 ¼»
The equation Ax = b is consistent if and only if 3b1 + (7/2)b2 + b3 = 0, or 6b1 + 7b2 + 2b3 = 0. The set
of such b is a plane through the origin in R3.
17. Row reduction shows that only three rows of A contain a pivot position:
ª 1
« −1
A= «
« 0
«
¬ 2
3
0
−1
−4
−1
2
0
3
3º ª 1
1»» «« 0
~
−8 » « 0
» «
−1¼ ¬ 0
3
0
2
−4
−1
2
−6
3
3º ª 1
4»» «« 0
~
−8 » « 0
» «
−7 ¼ ¬ 0
3
0
2
0
−1
0
0
0
3º ª 1
4 »» ««0
~
0» «0
» «
5¼ ¬ 0
3
0
2
0
−1
0
0
0
3º
4 »»
5»
»
0¼
Because not every row of A contains a pivot position, Theorem 4 in Section 1.4 shows that the
equation Ax = b does not have a solution for each b in R4.
18. Row reduction shows that only three rows of B contain a pivot position:
ª1
«0
B=«
«0
«
¬2
4
1
1
3
2
9
6
5
2º ª 1
−4 »» ««0
~
7 » «0
» «
−7 ¼ ¬0
4
1
1
3
2
1
6
3
2º ª 1
−4 »» ««0
~
7 » «0
» «
−11¼ ¬0
4
1
1
3
0
0
0
0
2º ª 1
−4 »» «« 0
~
15» « 0
» «
−7 ¼ ¬ 0
4
1
1
3
0
0
0
0
2º
−4 »»
15»
»
0¼
Because not every row of B contains a pivot position, Theorem 4 in Section 1.4 shows that not all
vectors in R4 can be written as a linear combination of the columns of B. The columns of B certainly
do not span R3, because each column of B is in R4, not R3. (This question was asked to alert students
to a fairly common misconception among students who are just learning about spanning.)
19. The work in Exercise 17 shows that statement (d) in Theorem 4 is false. So all four statements in
Theorem 4 are false. Thus, not all vectors in R4 can be written as a linear combination of the columns
of A. Also, the columns of A do not span R4.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.4
• Solutions
27
20. The work in Exercise 18 shows that statement (d) in Theorem 4 is false. So all four statements in
Theorem 4 are false. Thus, the equation Bx = y does not have a solution for each y in R4, and the
columns of B do not span R4.
21. Row reduce the matrix [v1 v2 v3] to determine whether it has a pivot in each row.
ª 1
« 0
«
« −1
«
¬ 0
0
−1
0
1
1º ª 1
0» «0
»~«
0» «0
» «
−1¼ ¬ 0
0
−1
0
1
1º ª 1
0» «0
»~«
1» « 0
» «
−1¼ ¬ 0
0
−1
0
0
1º ª 1
0» «0
»~«
1» « 0
» «
−1¼ ¬ 0
0
1
0
0
1º
0»
».
1»
»
0¼
The matrix [v1 v2 v3] does not have a pivot in each row, so the columns of the matrix do not span
R4, by Theorem 4. That is, {v1, v2, v3} does not span R4.
Note: Some students may realize that row operations are not needed, and thereby discover the principle
covered in Exercises 31 and 32.
22. Row reduce the matrix [v1 v2 v3] to determine whether it has a pivot in each row.
ª 0
« 0
«
«¬ −3
0
−3
9
4 º ª −3
−2 »» ~ «« 0
−6 »¼ «¬ 0
9
−3
0
−6º
−2 »»
4»¼
The matrix [v1 v2 v3] has a pivot in each row, so the columns of the matrix span R3, by Theorem 4.
That is, {v1, v2, v3} spans R3.
23. a.
b.
c.
d.
e.
f.
False. See the paragraph following equation (3). The text calls Ax = b a matrix equation.
True. See the box before Example 3.
False. See the warning following Theorem 4.
True. See Example 4.
True. See parts (c) and (a) in Theorem 4.
True. In Theorem 4, statement (a) is false if and only if statement (d) is also false.
24. a. True. This statement is in Theorem 3. However, the statement is true without any "proof"
because, by definition, Ax is simply a notation for x1a1 + ⋅ ⋅ ⋅ + xnan, where a1, …, an are the
columns of A.
b. True. See the box before Example 3.
c. True. See Example 2.
d. False. In Theorem 4, statement (d) is true if and only if statement (a) is true.
e. True. See Theorem 3.
f. False. In Theorem 4, statement (c) is false if and only if statement (a) is also false.
25. By definition, the matrix-vector product on the left is a linear combination of the columns of the
matrix, in this case using weights –3, –1, and 2. So c1 = –3, c2 = –1, and c3 = 2.
26. The equation in x1 and x2 involves the vectors u, v, and w, and it may be viewed as
ª x1 º
v ] « » = w. By definition of a matrix-vector product, x1u + x2v = w. The stated fact that
¬ x2 ¼
2u – 3v – w = 0 can be rewritten as 2u – 3v = w. So, a solution is x1 = 2, x2 = –3.
[u
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
28
CHAPTER 1
• Linear Equations in Linear Algebra
27. The matrix equation can be written as c1v1 + c2v2 + c3v3 + c4v4 + c5v5 = v6, where
c1 = –3, c2 = 1, c3 = 2, c4 = –1, c5 = 2, and
ª −3 º
ª 5º
ª −4 º
ª 9º
ª 7º
ª 11º
v1 = « » , v 2 = « » , v 3 = « » , v 4 = « » , v 5 = « » , v 6 = «
»
¬ 5¼
¬ 8¼
¬ 1¼
¬ −2 ¼
¬ −4 ¼
¬ −11¼
28. Place the vectors q1, q2, and q3 into the columns of a matrix, say, Q and place the weights x1, x2, and
x3 into a vector, say, x. Then the vector equation becomes
Qx = v, where Q = [q1 q2
ª x1 º
q3] and x = «« x2 »»
«¬ x3 »¼
Note: If your answer is the equation Ax = b, you need to specify what A and b are.
29. Start with any 3×3 matrix B in echelon form that has three pivot positions. Perform a row operation
(a row interchange or a row replacement) that creates a matrix A that is not in echelon form. Then A
has the desired property. The justification is given by row reducing A to B, in order to display the
pivot positions. Since A has a pivot position in every row, the columns of A span R3, by Theorem 4.
30. Start with any nonzero 3×3 matrix B in echelon form that has fewer than three pivot positions.
Perform a row operation that creates a matrix A that is not in echelon form. Then A has the desired
property. Since A does not have a pivot position in every row, the columns of A do not span R3, by
Theorem 4.
31. A 3×2 matrix has three rows and two columns. With only two columns, A can have at most two pivot
columns, and so A has at most two pivot positions, which is not enough to fill all three rows. By
Theorem 4, the equation Ax = b cannot be consistent for all b in R3. Generally, if A is an m×n matrix
with m > n, then A can have at most n pivot positions, which is not enough to fill all m rows. Thus,
the equation Ax = b cannot be consistent for all b in R3.
32. A set of three vectors in R4 cannot span R4. Reason: the matrix A whose columns are these three
vectors has four rows. To have a pivot in each row, A would have to have at least four columns (one
for each pivot), which is not the case. Since A does not have a pivot in every row, its columns do not
span R4, by Theorem 4. In general, a set of n vectors in Rm cannot span Rm when n is less than m.
33. If the equation Ax = b has a unique solution, then the associated system of equations does not have
any free variables. If every variable is a basic variable, then each column of A is a pivot column. So
ª1 0 0 º
«0 1 0»
».
the reduced echelon form of A must be «
«0 0 1 »
«
»
«¬ 0 0 0 »¼
Note: Exercises 33 and 36 are difficult in the context of this section because the focus in Section 1.4 is on
existence of solutions, not uniqueness. However, these exercises serve to review ideas from Section 1.2,
and they anticipate ideas that will come later.
34. Given Au1 = v1 and Au2 = v2, you are asked to show that the equation Ax = w has a solution, where
w = v1 + v2. Observe that w = Au1 + Au2 and use Theorem 5(a) with u1 and u2 in place of u and v,
respectively. That is, w = Au1 + Au2 = A(u1 + u2). So the vector x = u1 + u2 is a solution of w = Ax.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.4
• Solutions
29
35. Suppose that y and z satisfy Ay = z. Then 5z = 5Ay. By Theorem 5(b), 5Ay = A(5y). So 5z = A(5y),
which shows that 5y is a solution of Ax = 5z. Thus, the equation Ax = 5z is consistent.
36. If the equation Ax = b has a unique solution, then the associated system of equations does not have
any free variables. If every variable is a basic variable, then each column of A is a pivot column. So
ª1 0 0 0 º
«0 1 0 0»
» . Now it is clear that A has a pivot position in
the reduced echelon form of A must be «
«0 0 1 0»
«
»
¬0 0 0 1 ¼
each row. By Theorem 4, the columns of A span R4.
2
−5
8º ª 7
2
−5
8º
ª 7 2 −5 8º ª 7
« −5 −3 4 −9 » « 0 −11/ 7 3/ 7 −23/ 7 » « 0 −11/ 7
3/ 7
−23/ 7 »»
»~«
»~«
37. [M] «
« 6 10 −2 7 » « 0 58 / 7 16 / 7
1/ 7 » « 0
0 50 /11 −189 /11»
«
» «
» «
»
−3
11
23»¼ «¬ 0
0
0
0 »¼
¬« −7 9 2 15¼» «¬ 0
ª7
«0
or, approximately «
«0
«
¬« 0
−5
.429
8º
−3.29 »»
, to three significant figures. The original matrix does
0 4.55 −17.2 »
»
0
0
0 ¼»
not have a pivot in every row, so its columns do not span R4, by Theorem 4.
8º ª 4
−5
−1
8º ª 4
−5
−1
8º
»
«
»
«
2 » « 0 −13 / 4 −13 / 4
−4 » « 0 −13 / 4 −13 / 4
−4»»
~
~
−6 − 1 4 » « 0
−6 » « 0
1/ 4
1/ 4
0
0 −82 / 13»
» «
» «
»
1 10 7 ¼ ¬ 0
49 / 4
49 / 4 −11¼ ¬ 0
0
0
0¼
With pivots only in the first three rows, the original matrix has columns that do not span R4, by
Theorem 4.
ª4
«3
38. [M] «
«5
«
¬9
ª 10
« −8
39. [M] «
« −7
«
¬ 3
−5
−7
2
−1.57
−1
−4
−7
4
1
−6
4
−10
11
−1
−5
10
−1
12
6 º ª10
−3»» «« 0
~
−8» « 0
» «
12 ¼ ¬ 0
−7
−8 / 5
1
−26 / 5
4
−34 / 5
61 / 10
11 / 10
−43 / 10
97 / 10
9/5
54 / 5
6º
9 / 5»»
−19 / 5»
»
51 / 5¼
−7
1
4
6 º ª10
−7
1
4
6º
ª10
« 0 −8 / 5
»
«
9 / 5» « 0 −8 / 5
9 / 5»»
−26 / 5
−34 / 5
−26 / 5
−34 / 5
~«
~
« 0
0 −193 / 8 −193 / 8 49 / 16 » « 0
0 −193 / 8 −193 / 8
49 / 16 »
«
» «
»
0
49 / 8
49 / 8 183 / 16 ¼ ¬ 0
0
0
0 4715 / 386 ¼
¬ 0
The original matrix has a pivot in every row, so its columns span R4, by Theorem 4.
ª5
« −7
40. [M] «
« 11
«
¬ −3
11
−3
−6
−4
−7
6
5
4
6
−7
−9
2
12 º ª 5
−9 »» «« 0
~
−3» « 0
» «
7 ¼ ¬0
11
62 / 5
−6
−62 / 5
−7
−19 / 5
−96 / 5
53 / 5
96 / 5
−53 / 5
32 / 5
−11 / 5
12 º
39 / 5 »»
−147 / 5»
»
71 / 5 ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
30
CHAPTER 1
ª5
«0
~«
«0
«
¬0
• Linear Equations in Linear Algebra
11
62 / 5
−6
−62 / 5
−7
−19 / 5
0
0
0
0
16 / 31
65 / 62
12 º ª 5
39 / 5 »» «« 0
~
−537 / 31» « 0
» «
467 / 62 ¼ ¬ 0
11
62 / 5
−6
−62 / 5
−7
−19 / 5
0
0
0
0
16 / 31
0
12 º
39 / 5 »»
−537 / 31»
»
1367 / 62 ¼
The original matrix has a pivot in every row, so its columns span R4, by Theorem 4.
41. [M] Examine the calculations in Exercise 39. Notice that the fourth column of the original matrix,
say A, is not a pivot column. Let Ao be the matrix formed by deleting column 4 of A, let B be the
echelon form obtained from A, and let Bo be the matrix obtained by deleting column 4 of B. The
sequence of row operations that reduces A to B also reduces Ao to Bo. Since Bo is in echelon form, it
shows that Ao has a pivot position in each row. Therefore, the columns of Ao span R4.
It is possible to delete column 3 of A instead of column 4. In this case, the fourth column of A
becomes a pivot column of Ao, as you can see by looking at what happens when column 3 of B is
deleted. For later work, it is desirable to delete a nonpivot column.
Note: Exercises 41 and 42 help to prepare for later work on the column space of a matrix. (See Section
2.9 or 4.6.) The Study Guide points out that these exercises depend on the following idea, not explicitly
mentioned in the text: when a row operation is performed on a matrix A, the calculations for each new
entry depend only on the other entries in the same column. If a column of A is removed, forming a new
matrix, the absence of this column has no affect on any row-operation calculations for entries in the other
columns of A. (The absence of a column might affect the particular choice of row operations performed
for some purpose, but that is not being considered here.)
42. [M] Examine the calculations in Exercise 40. The third column of the original matrix, say A, is not a
pivot column. Let Ao be the matrix formed by deleting column 3 of A, let B be the echelon form
obtained from A, and let Bo be the matrix obtained by deleting column 3 of B. The sequence of row
operations that reduces A to B also reduces Ao to Bo. Since Bo is in echelon form, it shows that Ao has
a pivot position in each row. Therefore, the columns of Ao span R4.
It is possible to delete column 2 of A instead of column 3. (See the remark for Exercise 41.)
However, only one column can be deleted. If two or more columns were deleted from A, the
resulting matrix would have fewer than four columns, so it would have fewer than four pivot
positions. In such a case, not every row could contain a pivot position, and the columns of the matrix
would not span R4, by Theorem 4.
Notes: At the end of Section 1.4, the Study Guide gives students a method for learning and mastering
linear algebra concepts. Specific directions are given for constructing a review sheet that connects the
basic definition of “span” with related ideas: equivalent descriptions, theorems, geometric interpretations,
special cases, algorithms, and typical computations. I require my students to prepare such a sheet that
reflects their choices of material connected with “span”, and I make comments on their sheets to help
them refine their review. Later, the students use these sheets when studying for exams.
The MATLAB box for Section 1.4 introduces two useful commands gauss and bgauss that
allow a student to speed up row reduction while still visualizing all the steps involved. The command
B = gauss(A,1) causes MATLAB to find the left-most nonzero entry in row 1 of matrix A, and use
that entry as a pivot to create zeros in the entries below, using row replacement operations. The result is a
matrix that a student might write next to A as the first stage of row reduction, since there is no need to
write a new matrix after each separate row replacement. I use the gauss command frequently in lectures
to obtain an echelon form that provides data for solving various problems. For instance, if a matrix has 5
rows, and if row swaps are not needed, the following commands produce an echelon form of A:
B = gauss(A,1),
B = gauss(B,2),
B = gauss(B,3),
B = gauss(B,4)
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.5
• Solutions
31
If an interchange is required, I can insert a command such as B = swap(B,2,5) . The command
bgauss uses the left-most nonzero entry in a row to produce zeros above that entry. This command,
together with scale, can change an echelon form into reduced echelon form.
The use of gauss and bgauss creates an environment in which students use their computer
program the same way they work a problem by hand on an exam. Unless you are able to conduct your
exams in a computer laboratory, it may be unwise to give students too early the power to obtain reduced
echelon forms with one command—they may have difficulty performing row reduction by hand during an
exam. Instructors whose students use a graphic calculator in class each day do not face this problem. In
such a case, you may wish to introduce rref earlier in the course than Chapter 4 (or Section 2.8), which
is where I finally allow students to use that command.
1.5
SOLUTIONS
Notes: The geometry helps students understand Span{u, v}, in preparation for later discussions of subspaces. The parametric vector form of a solution set will be used throughout the text. Figure 6 will appear
again in Sections 2.9 and 4.8.
For solving homogeneous systems, the text recommends working with the augmented matrix, although no calculations take place in the augmented column. See the Study Guide comments on Exercise 7
that illustrate two common student errors.
All students need the practice of Exercises 1–14. (Assign all odd, all even, or a mixture. If you do not
assign Exercise 7, be sure to assign both 8 and 10.) Otherwise, a few students may be unable later to find
a basis for a null space or an eigenspace. Exercises 28–36 are important. Exercises 35 and 36 help
students later understand how solutions of Ax = 0 encode linear dependence relations among the columns
of A. Exercises 37–40 are more challenging. Exercise 37 will help students avoid the standard mistake of
forgetting that Theorem 6 applies only to a consistent equation Ax = b.
1. Reduce the augmented matrix to echelon form and circle the pivot positions. If a column of the
coefficient matrix is not a pivot column, the corresponding variable is free and the system of
equations has a nontrivial solution. Otherwise, the system has only the trivial solution.
ª 2
« −2
«
«¬ 4
−5
−7
2
8
1
7
0º ª2
0 »» ~ «« 0
0 »¼ «¬ 0
−5
−12
12
8
9
−9
0º ª 2
0»» ~ «« 0
0»¼ «¬ 0
−5
−12
0
8
9
0
0º
0»»
0»¼
The variable x3 is free, so the system has a nontrivial solution.
ª 1
2. «« −2
«¬ 2
−2
−3
−4
3
−4
9
0º ª 1
0 »» ~ ««0
0 »¼ «¬0
−2
−7
0
3
2
3
0º
0 »»
0 »¼
There is no free variable; the system has only the trivial solution.
4
−8 0 º
ª −3 4 −8 0 º ª −3
. The variable x3 is free; the system has nontrivial
3. «
~«
»
4 0 ¼ ¬ 0 7 / 3 28 / 3 0 »¼
¬ −2 5
solutions. An alert student will realize that row operations are unnecessary. With only two equations,
there can be at most two basic variables. One variable must be free. Refer to Exercise 29 in Section
1.2.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
32
CHAPTER 1
• Linear Equations in Linear Algebra
2 0º
−3
ª 5 −3 2 0 º ª 5
4. «
~«
»
» . The variable x3 is free; the system has nontrivial
¬ −3 −4 2 0 ¼ ¬ 0 −29 / 5 16 / 5 0 ¼
solutions. As in Exercise 3, row operations are unnecessary.
ª 2
5. «« −4
«¬ 0
x1
2
−4
−3
4
−8
−3
0º ª2
0 »» ~ «« 0
0 »¼ «¬ 0
2
0
−3
0º ª 1
0 »» ~ ««0
0 »¼ «¬0
4
0
−3
1
1
0
2
1
0
0º ª 1
0»» ~ ««0
0 »¼ «¬0
0
1
0
1
1
0
0º
0»»
0 »¼
+ x3 = 0
x2 + x 3 = 0 . The variable x3 is free, x1 = –x3, and x2 = –x3.
0 = 0
ª −1º
ª x1 º ª − x3 º
«
»
«
»
In parametric vector form, the general solution is x = « x2 » = « − x3 » = x3 «« −1»» .
«¬ 1»¼
«¬ x3 »¼ «¬ x3 »¼
2 −3 0º ª 1 2 −3 0º ª 1 0 −1 0º
ª 1 2 −3 0 º ª 1
«
»
«
3 0»» ~ «« 0 1 −1 0»» ~ «« 0 1 −1 0»»
6. « 2 1 −3 0 » ~ « 0 −3
0 0¼» ¬« 0
3 −3 0¼» ¬« 0 0
0 0¼» ¬« 0 0
0 0¼»
¬« −1 1
x1
− x3 = 0
x2 − x 3 = 0 . The variable x3 is free, x1 = x3, and x2 = x3.
0 = 0
ª x1 º ª x3 º
ª1º
«
»
«
»
In parametric vector form, the general solution is x = « x2 » = « x3 » = x3 ««1»» .
«¬1»¼
«¬ x3 »¼ «¬ x3 »¼
ª1
7. «
¬0
3
−3
7
1
−4
5
0º ª 1
~
0 »¼ «¬ 0
0
9
−8
1
−4
5
+ 9 x3 − 8 x4 = 0
0 º x1
.
»
0¼
x2 − 4 x3 + 5 x4 = 0
The basic variables are x1 and x2, with x3 and x4 free. Next, x1 = –9x3 + 8x4, and x2 = 4x3 – 5x4. The
general solution is
ª x1 º ª −9 x3 + 8 x4 º ª −9 x3 º ª 8 x4 º
ª −9 º
ª 8º
« x » « 4 x − 5 x » « 4 x » « −5 x »
« 4»
« −5»
4 »
4»
x = « 2» = « 3
=« 3 »+«
= x3 « » + x4 « »
« x3 » «
» « x3 » « 0 »
« 1»
« 0»
x3
« » «
» «
»
» «
« »
« »
x4
¬« x4 ¼» ¬«
¼» ¬« 0 ¼» ¬« x4 ¼»
¬« 0 ¼»
¬« 1¼»
ª1
8. «
¬0
−3
−8
5
1
2
−4
0º ª 1
~
0 »¼ ¬« 0
0
−2
−7
1
2
−4
− 2 x3 − 7 x4 = 0
0 º x1
.
0 »¼
x2 + 2 x3 − 4 x4 = 0
The basic variables are x1 and x2, with x3 and x4 free. Next, x1 = 2x3 + 7x4 and x2 = –2x3 + 4x4. The
general solution in parametric vector form is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.5
ª x1 º ª 2 x3 + 7 x4 º ª 2 x3 º ª 7 x4 º
ª 2º
ª
« x » « −2 x + 4 x » « −2 x » « 4 x »
« −2»
«
3
4»
3»
x = « 2» = «
=«
+ « 4 » = x3 « » + x4 «
« x3 » «
» « x3 » « 0 »
« 1»
«
x3
« » «
» «
»
» «
« »
«
x4
¬ 0¼
¬
¬ x4 ¼ ¬
¼ ¬ 0 ¼ ¬ x4 ¼
ª 3
9. «
¬ −2
−6
6
4
−2
x1 − 2 x2
0º ª 1
~
0 ¼» ¬« −2
−2
2
4
−2
0º ª 1
~
0 ¼» ¬« 0
−2
2
0
2
• Solutions
33
7º
4»»
0»
»
1¼
0º ª 1
~
0 ¼» ¬« 0
−2
0
0
1
0º
0 ¼»
= 0
.
x3 = 0
The solution is x1 = 2x2 , x3 = 0, with x2 free. In parametric vector form,
ª 2 x2 º ª 2 x2 º
ª 2º
«
»
«
»
x = « x2 » = « x2 » = x2 «« 1»» .
«¬ x3 »¼ «¬ 0 »¼
«¬ 0»¼
ª −1
10. «
¬ 2
−4
0
−4
−8
0
8
x1
x2
0º ª 1
~
0 ¼» ¬« 2
4
0
4
−8
0
8
0º ª 1
~
0 ¼» ¬« 0
4
0
4
1
0
0
0º ª 1
~
0 ¼» ¬« 0
0
0
4
1
0
0
0º
0 ¼»
+ 4 x4 = 0
.
= 0
The basic variables are x1 and x2, so x3 and x4 are free. (Note that x3 is not zero.) Also, x1 = –4x4. The
general solution is
ª x1 º ª −4 x4 º ª 0 º ª −4 x4 º
ª0º
ª −4 º
«x » « 0 » « 0 » « 0 »
«0»
« »
»=« »+«
»=x « »+x « 0 »
x = « 2» = «
« x3 » « x3 » « x3 » « 0 » 3 «1 » 4 « 0 »
« » «
» « » «
»
« »
« »
¬0¼
¬1¼
¬ x4 ¼ ¬ x4 ¼ ¬ 0 ¼ ¬ x4 ¼
11.
ª1
«0
«
«0
«
«¬0
−4
0
−2
1
0
0
3
0
−5
−1
0
0
0
0
0
0
1
0
−4
0
x1 − 4 x2
0º ª 1
0 »» ««0
~
0 » «0
» «
0 »¼ «¬0
−4
0
−2
1
0
0
0
0
7
−1
0
0
0
0
0
0
1
0
−4
0
0º ª 1
0 »» «« 0
~
0» «0
» «
0 »¼ «¬ 0
−4
0
0
1
0
0
0
0
5
−1
0
0
0
0
0
0
1
0
−4
0
+ 5 x6 = 0
x3
− x6 = 0
.
x5 − 4 x6 = 0
0 = 0
The basic variables are x1, x3, and x5. The remaining variables are free. In particular, x4 is free (and
not zero as some may assume). The solution is x1 = 4x2 – 5x6, x3 = x6, x5 = 4x6, with x2, x4, and x6 free.
In parametric vector form,
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0º
0 »»
0»
»
0»¼
34
CHAPTER 1
• Linear Equations in Linear Algebra
ª x1 º ª 4 x2 − 5 x6 º ª 4 x2 º ª 0 º ª −5 x6 º
ª 4º
ª0º
ª −5º
«x » « x
» « x » «0» « 0 »
« 1»
«0»
« 0»
2
« 2» «
» « 2 » « » «
»
« »
« »
« »
« x3 » «
» « 0 » « 0 » « x6 »
« 0»
«0»
« 1»
x6
x=« »=«
»=«
» = x2 « » + x4 « » + x6 « »
»+« »+«
« x4 » « x4
» « 0 » « x4 » « 0 »
« 0»
« 1»
« 0»
« x5 » « 4 x6 » « 0 » « 0 » « 4 x6 »
« 0»
«0»
« 4»
« » «
» «
»
» « » «
« »
« »
« »
x6
«¬0 »¼
«¬ 1»¼
¬« 0¼»
¬« x6 ¼» ¬«
¼» ¬« 0 ¼» ¬« 0 ¼» ¬« x6 ¼»
Note: The Study Guide discusses two mistakes that students often make on this type of problem.
12.
ª1
«
«0
«0
«
¬0
−2
0
3
0
−6
1
5
4
0
−6
0
0
0
0
0
0
0
0
1
0
x1 − 2 x2 + 3x3
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
+ 29 x5
x4 + 4 x5
−2
0
3
0
−6
1
5
4
0
0
0
0
0
0
0
0
0
0
1
0
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
−2
0
3
0
0
1
29
4
0
0
0
0
0
0
0
0
0
0
1
0
= 0
= 0
.
x6 = 0
0 = 0
The basic variables are x1, x4, and x6; the free variables are x2, x3, and x5. The general solution is
x1 = 2x2 – 3x3 – 29x5, x4 = – 4x5, and x6 = 0. In parametric vector form, the solution is
ª x1 º ª 2 x2 − 3 x3 − 29 x5 º ª 2 x2 º ª −3 x3 º ª −29 x5 º
ª −3º
ª −29 º
ª2º
«x » «
«
»
« 0»
»
«
»
«
»
«
»
«
»
x2
« 2» «
« 0»
«
»
» « x2 » « 0 » « 0 »
« 1»
« x3 » «
« 1»
« 0»
» « 0 » « x3 » « 0 »
«0»
x3
x=« »=«
»
»=«
»+«
»+«
» = x2 « » + x3 « » + x5 «
−4 x5
« x4 » «
« 0»
« −4 »
» « 0 » « 0 » « −4 x5 »
«0»
« x5 » «
« 0»
« 1»
» « 0 » « 0 » « x5 »
«0»
x5
« » «
« »
«
»
» «
» «
» «
»
« »
0
«¬ 0 ¼»
«¬ x6 »¼ ¬«
¬« 0 »¼
¼» ¬« 0 ¼» ¬« 0 ¼» ¬« 0 ¼»
¬« 0 ¼»
13. To write the general solution in parametric vector form, pull out the constant terms that do not
involve the free variable:
ª x1 º ª 5 + 4 x3 º ª 5º ª 4 x3 º ª 5º
ª 4º
«
»
«
»
«
»
«
»
«
»
x = « x2 » = «−2 − 7 x3 » = « −2 » + « −7 x3 » = « −2 » + x3 «« −7 »» = p + x3q.
«¬ x3 »¼ «¬ x3 »¼ «¬ 0 »¼ «¬ x3 »¼ «¬ 0 »¼
«¬ 1»¼
↑
↑
p
q
ª 5º
Geometrically, the solution set is the line through «« −2 »» in the direction of
«¬ 0 »¼
ª 4º
« −7 » .
« »
«¬ 1»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0º
0»»
0»
»
0¼
1.5
• Solutions
35
14. To write the general solution in parametric vector form, pull out the constant terms that do not
involve the free variable:
ª x1 º ª 5 x4 º ª0 º ª 5 x4 º ª 0 º
ª 5º
« x » «3 − 2 x » « 3 » « −2 x » « 3 »
« −2 »
4»
4»
x = « 2» = «
=« »+«
= « » + x4 « » = p + x4q
« x3 » « 2 + 5 x4 » « 2 » « 5 x4 » « 2 »
« 5»
« » «
» « » «
» « »
« »
¬ 1¼
¬ x4 ¼ ¬ x4 ¼ ¬0 ¼ ¬ x4 ¼ ¬ 0 ¼
↑
↑
p
q
The solution set is the line through p in the direction of q.
15. Solve x1 + 5x2 – 3x3 = –2 for the basic variable: x1 = –2 – 5x2 + 3x3, with x2 and x3 free. In vector
form, the solution is
ª x1 º ª −2 − 5 x2 + 3 x3 º ª −2 º ª −5 x2 º ª3 x3 º ª −2º
ª −5º
ª 3º
«
»
«
»
«
»
«
»
«
»
«
»
«
»
« »
x = « x2 » = «
x2
» = « 0 » + « x2 » + « 0 » = « 0» + x2 « 1» + x3 « 0»
«¬ x3 »¼ «¬
»¼ «¬ 0 »¼ «¬ 0 »¼ «¬ x3 »¼ «¬ 0»¼
«¬ 0 »¼
«¬ 1»¼
x3
The solution of x1 + 5x2 – 3x3= 0 is x1 = – 5x2 + 3x3, with x2 and x3 free. In vector form,
ª x1 º ª −5 x2 + 3 x3 º ª −5 x2 º ª3 x3 º
ª −5º
ª 3º
«
»
«
»
«
»
«
»
«
»
« »
x = « x2 » = «
x2
» = « x2 » + « 0 » = x2 « 1» + x3 « 0» = x2u + x3v
«¬ x3 »¼ «¬
»¼ «¬ 0 »¼ «¬ x3 »¼
«¬ 0»¼
«¬ 1»¼
x3
The solution set of the homogeneous equation is the plane through the origin in R3 spanned by
u and v. The solution set of the nonhomogeneous equation is parallel to this plane and passes through
ª −2 º
the point p = «« 0 »» .
¬« 0 ¼»
16. Solve x1 – 2x2 + 3x3 = 4 for the basic variable: x1 = 4 + 2x2 – 3x3, with x2 and x3 free. In vector form,
the solution is
ª x1 º ª 4 + 2 x2 − 3 x3 º ª 4 º ª 2 x2 º ª −3x3 º ª 4º
ª 2º
ª −3º
«
»
«
»
«
»
«
»
«
»
«
»
«
»
« »
x = « x2 » = «
x2
» = « 0 » + « x2 » + « 0 » = « 0» + x2 « 1» + x3 « 0»
«¬ x3 »¼ «¬
»¼ «¬ 0 »¼ «¬ 0 »¼ «¬ x3 »¼ «¬ 0»¼
«¬ 0»¼
«¬ 1»¼
x3
The solution of x1 – 2x2 + 3x3 = 0 is x1 = 2x2 – 3x3, with x2 and x3 free. In vector form,
ª x1 º ª 2 x2 − 3x3 º ª 2 x2 º ª −3 x3 º
ª 2º
ª −3º
«
»
«
»
«
»
«
»
«
»
« »
x = « x2 » = « x2
» = « x2 » + « 0 » = x2 « 1» + x3 « 0» = x2u + x3v
«¬ x3 »¼ «¬ x3
»¼ «¬ 0 »¼ «¬ x3 »¼
«¬ 0»¼
«¬ 1»¼
The solution set of the homogeneous equation is the plane through the origin in R3 spanned by u and
v. The solution set of the nonhomogeneous equation is parallel to this plane and passes through the
ª 4º
point p = «« 0 »» .
«¬ 0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
36
CHAPTER 1
• Linear Equations in Linear Algebra
17. Row reduce the augmented matrix for the system:
2
4
8º ª 2
2
4
8º ª 2
ª 2
« −4 −4 −8 −16 » ~ « 0
0
0
0»» ~ «« 0
«
» «
12 ¼» ¬« 0 −3 −3 12¼» ¬« 0
¬« 0 −3 −3
x1
4
8º ª 1 1 2
4º ª 1 0 1
8º
»
«
»
«
−3 12» ~ « 0 1 1 −4» ~ « 0 1 1 −4»»
0
0¼» ¬« 0 0 0
0¼» ¬« 0 0 0
0¼»
8 . Thus x1 = 8 – x3, x2 = –4 – x3, and x3 is free. In parametric vector form,
+ x3 =
x2 + x3 =
−4
0 =
0
2
−3
0
ª −1º
ª x1 º ª 8 − x3 º ª 8º ª − x3 º ª 8º
«
»
«
»
«
»
«
»
«
»
x = « x2 » = « −4 − x3 » = « −4 » + « − x3 » = « −4» + x3 «« −1»»
«¬ x3 »¼ «¬ x3 »¼ «¬ 0 »¼ «¬ x3 »¼ «¬ 0»¼
«¬ 1»¼
ª 8º
The solution set is the line through «« −4 »» , parallel to the line that is the solution set of the
«¬ 0 »¼
homogeneous system in Exercise 5.
18. Row reduce the augmented matrix for the system:
ª 1
« 2
«
«¬ −1
x1
2
1
1
−3
−3
0
5º ª 1
13»» ~ «« 0
−8»¼ «¬ 0
2
−3
3
−3
3
−3
5º ª 1
3»» ~ «« 0
−3»¼ «¬ 0
2
1
0
−3
−1
0
5º ª 1
−1»» ~ «« 0
0»¼ «¬ 0
0
1
0
−1
−1
0
7º
−1»»
0»¼
− x3 =
7 . Thus x1 = 7 + x3, x2 = –1 + x3, and x3 is free. In parametric vector form,
x2 − x3 = − 1
0 =
0
ª1º
ª x1 º ª 7 + x3 º ª 7 º ª x3 º ª 7 º
«
»
«
»
«
»
«
»
«
»
x = « x2 » = « −1 + x3 » = « −1» + « x3 » = « −1» + x3 ««1»»
«¬ x3 »¼ «¬ x3 »¼ «¬ 0 »¼ «¬ x3 »¼ «¬ 0»¼
«¬1»¼
ª 7º
The solution set is the line through «« −1»» , parallel to the line that is the solution set of the
«¬ 0»¼
homogeneous system in Exercise 6.
19. The line through a parallel to b can be written as x = a + t b, where t represents a parameter:
x1 = −2 − 5t
ª x º ª −2 º ª −5 º
x = « 1 » = « » + t « » , or ®
¬ x2 ¼ ¬ 0 ¼ ¬ 3¼
¯ x2 = 3t
20. The line through a parallel to b can be written as x = a + tb, where t represents a parameter:
x1 = 3 − 7t
ª x º ª 3º ª −7 º
x = « 1 » = « » + t « » , or ®
¬ x2 ¼ ¬ −2 ¼ ¬ 6 ¼
¯ x2 = −2 + 6t
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.5
• Solutions
37
ª 3º
ª4º
21. The line through p and q is parallel to q – p. So, given p = « » and q = « » , form
¬ −3¼
¬ 1¼
ª 4 − 3 º ª 1º
ª 3º ª 1º
= « » , and write the line as x = p + t(q – p) = « » + t « » .
q −p = «
»
¬1 − ( −3) ¼ ¬ 4 ¼
¬ −3¼ ¬ 4¼
ª −3º
ª 0º
22. The line through p and q is parallel to q – p. So, given p = « » and q = « » , form
¬ 2¼
¬ −3 ¼
ª 0 − ( −3) º ª 3º
ª −3 º ª 3 º
= « » , and write the line as x = p + t(q – p) = « » + t « »
q−p = «
»
¬ −3 − 2 ¼ ¬ −5 ¼
¬ 2 ¼ ¬ −5 ¼
Note: Exercises 21 and 22 prepare for Exercise 26 in Section 1.8.
23. a. True. See the first paragraph of the subsection titled Homogeneous Linear Systems.
b. False. The equation Ax = 0 gives an implicit description of its solution set. See the subsection
entitled Parametric Vector Form.
c. False. The equation Ax = 0 always has the trivial solution. The box before Example 1 uses the
word nontrivial instead of trivial.
d. False. The line goes through p parallel to v. See the paragraph that precedes Fig. 5.
e. False. The solution set could be empty! The statement (from Theorem 6) is true only when there
exists a vector p such that Ap = b.
24. a. False. The trivial solution is always a solution to a homogeneous system of linear equations.
b. False. A nontrivial solution of Ax = 0 is any nonzero x that satisfies the equation. See the
sentence before Example 2.
c. True. See the paragraph following Example 3.
d. True. If the zero vector is a solution, then b = Ax = A0 = 0.
e. True. See Theorem 6.
25. Suppose p satisfies Ax = b. Then Ap = b. Theorem 6 says that the solution set of Ax = b equals the
set S ={w : w = p + vh for some vh such that Avh = 0}. There are two things to prove: (a) every vector
in S satisfies Ax = b, (b) every vector that satisfies Ax = b is in S.
a. Let w have the form w = p + vh, where Avh = 0. Then
Aw = A(p + vh) = Ap + Avh. By Theorem 5(a) in section 1.4
=b+0=b
So every vector of the form p + vh satisfies Ax = b.
b. Now let w be any solution of Ax = b, and set vh = w − p. Then
Avh = A(w – p) = Aw – Ap = b – b = 0
So vh satisfies Ax = 0. Thus every solution of Ax = b has the form w = p + vh.
26. When A is the 3×3 zero matrix, every x in R3 satisfies Ax = 0. So the solution set is all vectors in R3.
27. (Geometric argument using Theorem 6.) Since Ax = b is consistent, its solution set is obtained by
translating the solution set of Ax = 0, by Theorem 6. So the solution set of Ax = b is a single vector if
and only if the solution set of Ax = 0 is a single vector, and that happens if and only if Ax = 0 has
only the trivial solution.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
38
CHAPTER 1
• Linear Equations in Linear Algebra
(Proof using free variables.) If Ax = b has a solution, then the solution is unique if and only if there
are no free variables in the corresponding system of equations, that is, if and only if every column of
A is a pivot column. This happens if and only if the equation Ax = 0 has only the trivial solution.
28. a. When A is a 3×3 matrix with three pivot positions, the equation Ax = 0 has no free variables and
hence has no nontrivial solution.
b. With three pivot positions, A has a pivot position in each of its three rows. By Theorem 4 in
Section 1.4, the equation Ax = b has a solution for every possible b. The term "possible" in the
exercise means that the only vectors considered in this case are those in R3, because A has three
rows.
29. a. When A is a 4×4 matrix with three pivot positions, the equation Ax = 0 has three basic variables
andone free variable. So Ax = 0 has a nontrivial solution.
b. With only three pivot positions, A cannot have a pivot in every row, so by Theorem 4 in Section
1.4, the equation Ax = b cannot have a solution for every possible b (in R4).
30. a. When A is a 2×5 matrix with two pivot positions, the equation Ax = 0 has two basic variables and
three free variables. So Ax = 0 has a nontrivial solution.
b. With two pivot positions and only two rows, A has a pivot position in every row. By Theorem 4
in Section 1.4, the equation Ax = b has a solution for every possible b (in R2).
31. a. When A is a 3×2 matrix with two pivot positions, each column is a pivot column. So the equation
Ax = 0 has no free variables and hence no nontrivial solution.
b. With two pivot positions and three rows, A cannot have a pivot in every row. So the equation Ax
= b cannot have a solution for every possible b (in R3), by Theorem 4 in Section 1.4.
32. No. If the solution set of Ax = b contained the origin, then 0 would satisfy A0= b, which is not true
since b is not the zero vector.
33. Look for A = [a1 a2 a3] such that 1 a1 + 1 a2 + 1 a3 = 0. That is, construct A so that each row sum
(the sum of the entries in a row) is zero.
34. Look for A = [a1 a2 a3] such that 2 a1 – 1 a2 + 1 a3 = 0. That is, construct A so that subtracting the
third column from the second column is twice the first column.
ª −1º
ª −3º
«
»
35. Look at x1 « 7 » + x2 «« 21»» and notice that the second column is 3 times the first. So suitable values
«¬ −2 »¼
«¬ −6»¼
ª 3º
for x1 and x2 would be 3 and –1 respectively. (Another pair would be 6 and –2, etc.) Thus x = « »
¬ −1¼
satisfies Ax = 0.
36. Inspect how the columns a1 and a2 of A are related. The second column is –2/3 times the first. Put
ª2º
another way, 2a1 + 3a2 = 0. Thus « » satisfies Ax = 0.
¬ 3¼
Note: Exercises 35 and 36 set the stage for the concept of linear dependence.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.6
• Solutions
39
37. Since the solution set of Ax = 0 contains the point (4,1), the vector x = (4,1) satisfies Ax = 0. Write
this equation as a vector equation, using a1 and a2 for the columns of A:
4 a1 + 1 a2 = 0
Then a2 = –4a1. So choose any nonzero vector for the first column of A and multiply that column by
ª1 −4 º
– 4 to get the second column of A. For example, set A = «
».
¬1 −4 ¼
Finally, the only way the solution set of Ax = b could not be parallel to the line through (1,4) and the
origin is for the solution set of Ax = b to be empty. This does not contradict Theorem 6, because that
theorem applies only to the case when the equation Ax = b has a nonempty solution set. For b, take
any vector that is not a multiple of the columns of A.
Note: In the Study Guide, a “Checkpoint” for Section 1.5 will help students with Exercise 37.
38. If w satisfies Ax = 0, then Aw = 0. For any scalar c, Theorem 5(b) in Section 1.4 shows that
A(cw)=cAw = c0 = 0.
39. Suppose Av = 0 and Aw = 0. Then, since A(v + w) = Av + Aw by Theorem 5(a) in Section 1.4,
A(v + w) = Av + Aw = 0 + 0 = 0.
Now, let c and d be scalars. Using both parts of Theorem 5,
A(cv + dw) = A(cv) + A(dw) = cAv + dAw = c0 + d0 = 0.
40. No. If Ax = b has no solution, then A cannot have a pivot in each row. Since A is 3×3, it has at most
two pivot positions. So the equation Ax = y for any y has at most two basic variables and at least one
free variable. Thus, the solution set for Ax = y is either empty or has infinitely many elements.
Note: The MATLAB box in the Study Guide introduces the zeros command, in order to augment a
matrix with a column of zeros.
1.6
SOLUTIONS
1. Fill in the exchange table one column at a time. The entries in a column describe where a sector's
output goes. The decimal fractions in each column sum to 1.
Distribution of
Output From:
Goods Services
Purchased by:
output ↓
↓
input
.2
.7
→
Goods
.8
.3
→
Services
Denote the total annual output (in dollars) of the sectors by pG and pS. From the first row, the total
input to the Goods sector is .2 pG + .7 pS. The Goods sector must pay for that. So the equilibrium
prices must satisfy
income
expenses
pG
= .2pG + .7 pS
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
40
CHAPTER 1
• Linear Equations in Linear Algebra
From the second row, the input (that is, the expense) of the Services sector is .8 pG + .3 pS.
The equilibrium equation for the Services sector is
income
expenses
pS
= .8pG + .3 pS
Move all variables to the left side and combine like terms:
.8 pG − .7 pS = 0
−.8 pG + .7 pS = 0
Row reduce the augmented matrix:
ª .8
« −.8
¬
−.7
.7
0 º ª.8
~
0 »¼ «¬ 0
−.7
0
0º ª 1
~
0 »¼ «¬ 0
−.875
0
0º
0 »¼
The general solution is pG = .875 pS, with pS free. One equilibrium solution is pS = 1000 and pG =
875. If one uses fractions instead of decimals in the calculations, the general solution would be
written pG = (7/8) pS, and a natural choice of prices might be pS = 80 and pG = 70. Only the ratio of
the prices is important: pG = .875 pS. The economic equilibrium is unaffected by a proportional
change in prices.
2. Take some other value for pS, say 200 million dollars. The other equilibrium prices are then
pC = 188 million, pE = 170 million. Any constant nonnegative multiple of these prices is a set of
equilibrium prices, because the solution set of the system of equations consists of all multiples of one
vector. Changing the unit of measurement to another currency such as Japanese yen has the same
effect as multiplying all equilibrium prices by a constant. The ratios of the prices remain the same,
no matter what currency is used.
3. a. Fill in the exchange table one column at a time. The entries in a column describe where a sector’s
output goes. The decimal fractions in each column sum to 1.
Distribution of Output From:
output
Fuels and Power
Manufacturing
Services
↓
.10
.80
↓
.10
.10
↓
.20
.40
.10
.80
.40
Purchased by :
input
→
→
→
Fuels and Power
Manufacturing
Services
b. Denote the total annual output (in dollars) of the sectors by pF, pM, and pS. From the first row of
the table, the total input to the Fuels & Power sector is .1pF + .1pM + .2pS. So the equilibrium
prices must satisfy
income
expenses
pF
= .1pF + .1 pM + .2 pS
From the second and third rows of the table, the income/expense requirements for the
Manufacturing sector and the Services sector are, respectively,
pM = .8 pF + .1 pM + .4 pS
pS = .1 p F + .8 pM + .4 pS
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.6
• Solutions
41
Move all variables to the left side and combine like terms:
.9 pF – .1 pM – .2 pS = 0
–.8 pF + .9 pM – .4 pS = 0
ª .9
« −.8
«
«¬ −.1
–.1 pF – .8 pM + .6 pS = 0
−.1
.9
−.8
−.2
−.4
.6
0º
0 »»
0 »¼
c. [M] You can obtain the reduced echelon form with a matrix program.
ª .9 −.1 −.2 0 º ª 1 0 −.301 0 º The number of decimal
«
.9 −.4 0 »» ~ ««0 1 −.712 0 »» places displayed is
« −.8
«¬ −.1 −.8
.6 0 »¼ «¬0 0
0 0 »¼ somewhat arbitrary.
The general solution is pF = .301 pS, pM = .712 pS, with pS free. If pS is assigned the value of 100,
then pF = 30.1 and pM = 71.2. Note that only the ratios of the prices are determined. This makes
sense, for if they were converted from, say, dollars to yen or Euros, the inputs and outputs of each
sector would still balance. The economic equilibrium is not affected by a proportional change in
prices.
4. a. Fill in the exchange table one column at a time. The entries in each column must sum to 1.
Distribution of Output From:
output
Mining
Lumber
Energy
Transportation
↓
.30
↓
.15
↓
.20
↓
.20
.10
.15
.15
.10
.60
0
.50
.20
.45
.20
.50
.20
Purchased by :
input
→
→
→
→
Mining
Lumber
Energy
Transportation
b. [M] Denote the total annual output of the sectors by pM, pL, pE, and pT, respectively. From the first
row of the table, the total input to Agriculture is .30pM + .15pL + .20pE + .20 pT. So the
equilibrium prices must satisfy
income
expenses
pM = .30 pM + .15pL + .20 pE + .20 pT
From the second, third, and fourth rows of the table, the equilibrium equations are
pL = .10 pM + .15 pL + .15 pE + .10 pT
pE = .60 pM + .50 pL + .45 p E + .50 pT
pT =
.20 pL + .20 pE + .20 pT
Move all variables to the left side and combine like terms:
.70 pM − .15 pL − .20 pE − .20 pT = 0
−.10 pM + .85 pL − .15 pE − .10 pT = 0
−.60 pM − .50 pL + .55 pE − .50 pT = 0
−.20 pL − .20 pE + .80 pT = 0
Reduce the augmented matrix to reduced echelon form:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
42
CHAPTER 1
ª .70
«
« −.10
« −.60
«
¬ 0
• Linear Equations in Linear Algebra
−.15
.85
−.20
−.15
−.20
−.10
−.50
−.20
.55
−.20
−.50
.80
0º ª 1
0 »» ««0
~
0 » «0
» «
0 ¼ ¬0
0
1
0
0
−1.37
−.84
0
0
1
0
−3.16
0
0º
0 »»
0»
»
0¼
Solve for the basic variables in terms of the free variable pT, and obtain pM = 1.37pT, pL = .84pT,
and pE = 3.16pT. The data probably justifies at most two significant figures, so take pT = 100 and
round off the other prices to pM = 137, pL = 84, and pE = 316.
5. a. Fill in the exchange table one column at a time. The entries in each column must sum to 1.
Distribution of Output From:
Purchased by :
Agriculture
Manufacturing
Services
Transportation
↓
.20
↓
.35
↓
.10
↓
.20
input
→
.20
.10
.20
.30
→
Manufacturing
.30
.30
.35
.20
.50
.20
.20
.30
→
→
Services
Transportation
output
Agriculture
b. [M] Denote the total annual output of the sectors by pA, pM, pS, and pT, respectively. The
equilibrium equations are
pA
=
.20 p A + .35 pM + .10 pS + .20 pT
pM
pS
=
=
.20 p A + .10 pM + .20 pS + .30 pT
.30 p A + .35 pM + .50 pS + .20 pT
pT
=
.30 p A + .20 pM + .20 pS + .30 pT
Move all variables to the left side and combine like terms:
.80 p A − .35 pM − .10 pS − .20 pT = 0
−.20 p A + .90 pM − .20 pS − .30 pT = 0
−.30 p A − .35 pM + .50 pS − .20 pT = 0
−.30 p A − .20 pM − .20 pS + .70 pT = 0
Reduce the augmented matrix to reduced echelon form:
ª .80
« −.20
«
« −.30
«
¬ −.30
−.35
.90
−.10
−.20
−.20
−.30
−.35
−.20
.50
−.20
−.20
.70
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
0
1
0
0
−.799
−.836
0
0
1
0
−1.465
0
0º
0 »»
0»
»
0¼
Solve for the basic variables in terms of the free variable pT, and obtain pA = .799pT, pM = .836pT,
and pS = 1.465pT. Take pT = $10.00 and round off the other prices to pA = $7.99, pM = $8.36, and
pS = $14.65 per unit.
c. Construct the new exchange table one column at a time. The entries in each column must sum to 1.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.6
• Solutions
43
Distribution of Output From:
Purchased by :
Agriculture
Manufacturing
Services
Transportation
↓
.20
↓
.35
↓
.10
↓
.20
input
→
.10
.10
.20
.30
→
Manufacturing
.40
.30
.35
.20
.50
.20
.20
.30
→
→
Services
Transportation
output
Agriculture
d. [M] The new equilibrium equations are
pA
=
.20 p A + .35 pM + .10 pS + .20 pT
pM
pS
=
=
.10 p A + .10 pM + .20 pS + .30 pT
.40 p A + .35 pM + .50 pS + .20 pT
pT
=
.30 p A + .20 pM + .20 pS + .30 pT
Move all variables to the left side and combine like terms:
.80 p A − .35 pM − .10 pS − .20 pT = 0
−.10 p A + .90 pM − .20 pS − .30 pT = 0
−.40 p A − .35 pM + .50 pS − .20 pT = 0
−.30 p A − .20 pM − .20 pS + .70 pT = 0
Reduce the augmented matrix to reduced echelon form:
ª .80
« −.10
«
« −.40
«
¬ −.30
−.35
.90
−.10
−.20
−.20
−.30
−.35
−.20
.50
−.20
−.20
.70
0º ª 1
0 »» ««0
~
0 » «0
» «
0 ¼ ¬0
0
1
0
0
−.781
−.767
0
0
1
0
−1.562
0
0º
0 »»
0»
»
0¼
Solve for the basic variables in terms of the free variable pT, and obtain pA = .781pT, pM = .767pT,
and
pS = 1.562pT. Take pT = $10.00 and round off the other prices to pA = $7.81, pM = $7.67, and
pS = $15.62 per unit. The campaign has caused unit prices for the Agriculture and
Manufacturing sectirs to go down slightly, while increasing the unit price for the Services sector
to increase by $.10 per unit. The campaign has benefited the Services sector the most.
6. The following vectors list the numbers of atoms of aluminum (Al), oxygen (O), and carbon (C):
ª2º
ª0º
ª1 º
«
»
«
»
Al2 O3 : « 3 » , C: « 0 » , Al: «« 0 »» , CO 2 :
«¬ 0 »¼
«¬1 »¼
«¬ 0 »¼
ª 0 º aluminum
« 2 » oxygen
« »
«¬ 1 »¼ carbon
The coefficients in the equation x1⋅Al2O3 + x2⋅C → x3⋅Al + x4⋅CO2 satisfy
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
44
CHAPTER 1
• Linear Equations in Linear Algebra
ª2º
ª0º
ª1 º
ª0º
«
»
«
»
«
»
x1 « 3 » + x2 «0» = x3 « 0» + x4 «« 2 »»
«¬ 0 »¼
«¬1 »¼
«¬ 0»¼
«¬ 1 »¼
Move the right terms to the left side (changing the sign of each entry in the third and fourth vectors)
and row reduce the augmented matrix of the homogeneous system:
0 0 º ª 1 0 −1 / 2
0 0 º ª 1 0 −1 / 2
ª 2 0 −1
«3 0
»
«
0 −2 0 » ~ « 3 0
0 −2 0»» ~ ««0 0
3/ 2
«
«¬ 0 1
0 −1 0 »¼ «¬0 1
0 −1 0 »¼ «¬0 1
0
0 0º ª 1 0 −1 / 2
0 0º ª 1 0
ª 1 0 −1 / 2
«
»
«
~ «0 1
0 −1 0» ~ «0 1
0
−1 0»» ~ «« 0 1
«¬0 0
3 / 2 −2 0 »¼ «¬0 0
1 −4 / 3 0»¼ «¬ 0 0
0º
0»»
0 »¼
0
−2
−1
0
0
1
−2 / 3
−1
−4 / 3
0º
0»»
0»¼
The general solution is x1 = (2/3)x4, x2 = x4, x3 = (4/3)x4, with x4 free. Take x4 = 3. Then x1 = 2,
x2 = 3, and x3 = 4. The balanced equation is
2Al2O3 + 3C → 4Al + 3CO2
7. The following vectors list the numbers of atoms of sodium (Na), hydrogen (H), carbon (C), and
oxygen (O):
ª1º
«1»
NaHCO3 : « » , H 3C6 H 5 O7 :
«1»
« »
¬3¼
ª0 º
«8 »
« » , Na 3C6 H 5O7 :
«6 »
« »
¬7 ¼
ª3º
ª0 º
ª0 º
«5 »
« 2»
«0 »
« » , H 2 O : « » , CO 2 : « »
«6 »
«0 »
«1 »
« »
« »
« »
¬7 ¼
¬1 ¼
¬2 ¼
sodium
hydrogen
carbon
oxygen
The order of the various atoms is not important. The list here was selected by writing the elements in
the order in which they first appear in the chemical equation, reading left to right:
x1 · NaHCO3 + x2 · H3C6H5O7 → x3 · Na3C6H5O7 + x4 · H2O + x5 · CO2.
The coefficients x1, …, x5 satisfy the vector equation
ª1º
ª0º
ª 3º
ª0º
ª0 º
«1»
«8 »
«5»
« 2»
«0 »
x1 « » + x2 « » = x3 « » + x4 « » + x5 « »
«1»
«6»
«6»
«0»
«1 »
« »
« »
« »
« »
« »
«¬3»¼
«¬ 7 »¼
«¬ 7 »¼
«¬ 1 »¼
«¬ 2 »¼
Move all the terms to the left side (changing the sign of each entry in the third, fourth, and fifth
vectors) and reduce the augmented matrix:
ª1
«1
«
«1
«
¬3
0
8
−3
−5
0
−2
0
0
6
7
−6
−7
0
−1
−1
−2
0 º ª1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
0
1
0
0
0
0
−1
−1 / 3
0
0
1
0
0
1
−1 / 3
−1
0º
0 »»
0»
»
0¼
The general solution is x1 = x5, x2 = (1/3)x5, x3 = (1/3)x5, x4 = x5, and x5 is free. Take x5 = 3. Then x1 =
x4 = 3, and x2 = x3 = 1. The balanced equation is
3NaHCO3 + H3C6H5O7 → Na3C6H5O7 + 3H2O + 3CO2
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.6
• Solutions
45
8. The following vectors list the numbers of atoms of hydrogen (H), oxygen (O), calcium (Ca), and
carbon (C):
ª3º
ª0 º
ª2º
ª0º
«1 »
«3»
«1 »
«0»
H 3O: « » , CaCO 3 : « » , H 2 O: « » , Ca: « » , CO 2 :
«0 »
«1 »
«0»
«1 »
« »
« »
« »
« »
¬0 ¼
¬1 ¼
¬0¼
¬0¼
The coefficients in the chemical equation
ª 0 º hydrogen
« 2 » oxygen
« »
« 0 » calcium
« »
¬ 1 ¼ carbon
x1⋅H3O + x2⋅CaCO3 → x3⋅H2O + x4⋅Ca + x5⋅CO2
satisfy the vector equation
ª3º
«1 »
x1 « » + x2
«0 »
« »
¬0 ¼
ª0º
ª 2º
ª0º
ª0º
« 3»
«1 »
«0»
« »
« » = x « » + x « » + x « 2»
«1 » 3 « 0 » 4 «1 » 5 « 0 »
« »
« »
« »
« »
¬1 ¼
¬0¼
¬0¼
¬1 ¼
Move the terms to the left side (changing the sign of each entry in the last three vectors) and reduce
the augmented matrix:
ª3
«1
«
«0
«
¬0
0
3
−2
−1
0
0
0
−2
1
1
0
0
−1
0
0
−1
0 º ª1
0 »» ««0
~
0 » «0
» «
0 ¼ ¬0
0
1
0
0
0
0
−2
−1
0
0
1
0
0
1
−3
−1
0º
0 »»
0»
»
0¼
The general solution is x1 = 2x5, x2 = x5, x3 = 3x5, x4 = x5, and x5 is free. Take x5 = 1. Then x1 = 2, and
x2 = x4 = 1, and x3 = 3. The balanced equation is
2H3O + CaCO3 → 3H2O + Ca + CO2
9.The following vectors list the numbers of atoms of boron (B), sulfur (S), hydrogen (H), and oxygen
(O):
ª2 º
«3 »
B2S3 : « » , H 2 O:
«0 »
« »
¬0 ¼
ª0 º
«0 »
« » , H 3 BO3 :
«2»
« »
¬1 ¼
ª1 º
«0 »
« » , H 2S:
«3»
« »
¬3¼
ª0 º
«1 »
« »
«2»
« »
¬0 ¼
boron
sulfur
hydrogen
oxygen
The coefficients in the equation x1⋅B2S3 + x2⋅H2O → x3⋅H3BO3 + x4⋅H2S satisfy
ª2º
ª0º
ª1 º
ª0º
«3»
«0»
«0»
«1 »
x1 « » + x2 « » = x3 « » + x4 « »
«0»
« 2»
« 3»
«2»
« »
« »
« »
« »
¬0¼
¬1 ¼
¬ 3¼
¬0¼
Move the terms to the left side (changing the sign of each entry in the third and fourth vectors) and
row reduce the augmented matrix of the homogeneous system:
ª2
«3
«
«0
«
¬0
0
0
−1
0
0
−1
2
1
−3
−3
−2
0
0º ª 1
0 »» ««0
~
0 » «0
» «
0 ¼ ¬0
0
1
0
0
−1 / 3
−2
0
0
1
0
−2 / 3
0
0º
0 »»
0»
»
0¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
46
CHAPTER 1
• Linear Equations in Linear Algebra
The general solution is x1 = (1/3) x4, x2 = 2x4, x3 = (2/3) x4, with x4 free. Take x4 = 3. Then x1 = 1,
x2 = 6, and x3 = 2. The balanced equation is
B2S3 + 6H2O → 2H3BO3 + 3H2S
10. [M] Set up vectors that list the atoms per molecule. Using the order lead (Pb), nitrogen (N),
chromium (Cr), manganese (Mn), and oxygen (O), the vector equation to be solved is
ª1 º
ª0 º
ª3º
ª0 º
ª0º
ª0 º
«6 »
«0 »
«0 »
«0 »
«0 »
«1 »
« »
« »
« »
« »
« »
« »
x1 «0 » + x2 «1 » = x3 «0 » + x4 « 2 » + x5 «0 » + x6 «0»
« »
« »
« »
« »
« »
« »
«0 »
«2»
«0 »
«0 »
«1 »
«0 »
«¬0 »¼
«¬8 »¼
«¬ 4 »¼
«¬ 3»¼
«¬ 2 »¼
«¬1 »¼
lead
nitrogen
chromium
manganese
oxygen
The general solution is x1 = (1/6)x6, x2 = (22/45)x6, x3 = (1/18)x6, x4 = (11/45)x6, x5 = (44/45)x6, and x6
is free. Take x6 = 90. Then x1 = 15, x2 = 44, x3 = 5, x4 = 22, and x5 = 88. The balanced equation is
15PbN6 + 44CrMn2O8 → 5Pb3O4 + 22Cr2O3 + 88MnO2 + 90NO
11. [M] Set up vectors that list the atoms per molecule. Using the order manganese (Mn), sulfur (S),
arsenic (As), chromium (Cr), oxygen (O), and hydrogen (H), the vector equation to be solved is
ª1 º
ª0º
ª0 º
ª1 º
ª0º
ª0º
ª0 º
«1 »
«0»
«1 »
«0 »
«0»
«3»
«0 »
« »
« »
« »
« »
« »
« »
« »
«0 »
«2»
«0 »
«0 »
«1 »
«0»
«0 »
x1 « » + x2 « » + x3 « » = x4 « » + x5 « » + x6 « » + x7 « »
«0 »
«10 »
«0 »
«0 »
«0»
«1»
«0 »
«0 »
«35»
«4»
«4»
«0»
«12»
«1 »
« »
« »
« »
« »
« »
« »
« »
«¬0 »¼
«¬ 0 »¼
«¬ 2 »¼
«¬1 »¼
«¬3 »¼
«¬ 0 »¼
«¬ 2¼»
manganese
sulfur
arsenic
chromium
oxygen
hydrogen
In rational format, the general solution is x1 = (16/327)x7, x2 = (13/327)x7, x3 = (374/327)x7,
x4 = (16/327)x7, x5 = (26/327)x7, x6 = (130/327)x7, and x7 is free. Take x7 = 327 to make the other
variables whole numbers. The balanced equation is
16MnS + 13As2Cr10O35 + 374H2SO4 → 16HMnO4 + 26AsH3 + 130CrS3O12 + 327H2O
Note that some students may use decimal calculation and simply "round off" the fractions that relate
x1, ..., x6 to x7. The equations they construct may balance most of the elements but miss an atom or
two. Here is a solution submitted by two of my students:
5MnS + 4As2Cr10O35 + 115H2SO4 → 5HMnO4 + 8AsH3 + 40CrS3O12 + 100H2O
Everything balances except the hydrogen. The right side is short 1 hydrogen atom. Perhaps the
students thought that it escaped!
12. Write the equations for each intersection:
Intersection
Flow in
A
x1 + x4
=
x2
B
x2
=
x3 + 100
=
x4
C
x3 + 80
Rearrange the equations:
Flow out
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.6
x1
−
x2
x2
−
+
x4
=
=
0
100
−
x4
=
−80
x3
x3
• Solutions
Reduce the augmented matrix:
ª1
«
«0
«¬0
−1
1
0
0
−1
1
0º
ª1
»
100 » ~ ⋅⋅⋅ ~ ««0
«¬0
−80 »¼
1
0
−1
0
1
0
0
0
1
0
−1
−1
20º
20»»
−80»¼
The general solution (written in the style of Section 1.2) is
x1 = 20
° x = 20 + x
° 2
4
®
=
80
+
x
−
x4
° 3
°̄ x4 is free
Since x3 cannot be negative, the minimum value of x4 is 80.
13. Write the equations for each intersection:
Intersection Flow in
Flow out
A
x2 + 30 =
x1 + 80
B
x3 + x5
=
x2 + x4
x6 + 100 =
x5 + 40
C
D
x4 + 40 =
x6 + 90
E
x1 + 60 =
x3 + 20
Rearrange the equations:
x1 − x2
x2 − x3
+
x4
x4
−
x1
−
x5
x5
−
−
=
=
=
=
=
x6
x6
x3
−50
0
60
50
−40
Reduce the augmented matrix:
ª1
«0
«
«0
«
«0
«¬ 1
−1
1
0
0
0
0
0
0
0
−1
0
0
−1
1
0
1
0
−1
1
0
0
0
−1
−1
0
−50 º
ª1
»
«0
0»
«
60 » ~ ⋅⋅⋅ ~ «0
»
«
50 »
«0
«¬0
−40 »¼
0
1
0
0
0
−1
−1
0
0
0
0
0
0
0
1
0
0
0
0
1
0
0
−1
−1
0
−40 º
10 »»
50 »
»
60 »
0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
47
48
CHAPTER 1
• Linear Equations in Linear Algebra
x1 = x3 − 40
° x = x + 10
3
° 2
°° x3 is free
a. The general solution is ®
° x4 = x6 + 50
° x5 = x6 + 60
°
°̄ x6 is free
b. To find minimum flows, note that since x1 cannot be negative, x3 > 40. This implies that
x2 > 50. Also, since x6 cannot be negative, x4 > 50 and x5 > 60. The minimum flows are
x2 = 50, x3 = 40, x4 = 50, x5 = 60 (when x1 = 0 and x6 = 0).
14. Write the equations for each intersection:
Intersection
A
B
C
D
Flow in
80
x1 + x2 + 100
Flow out
=
=
=
=
x3
x4 + x5
x1 + x5
x4
x2 + 90
x3 + 90
Rearrange the equations:
+
x1
x1
+
x2
x2
−
−
x4
−
x4
x5
x3
x3
−
x5
=
80
=
=
−100
−90
=
−90
Reduce the augmented matrix:
ª1
«1
«
«0
«
¬0
0
1
0
0
0
−1
1
0
1
0
−1
1
0
−1
0
−1
80 º
ª1
»
«0
−100 »
~ ⋅⋅⋅ ~ «
«0
−90 »
»
«
−90 ¼
¬0
0
1
0
0
0
−1
1
−1
0
0
1
0
−1
0
−1
0
80 º
−180 »»
−90 »
»
0¼
x1 = 80 − x5
° x = x + x − 180
4
5
°° 2
a. The general solution is ® x3 = x4 + x5 − 90
° x is free
° 4
°̄ x5 is free
x1 = 80
° x = x − 180
°
4
b. If x5 = 0, then the general solution is ® 2
=
x
x
4 − 90
° 3
°̄ x4 is free
c. Since x2 cannot be negative, the minimum value of x4 when x5 = 0 is 180.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.6
15.
• Solutions
Write the equations for each intersection.
Intersection
A
B
C
D
E
F
Flow in
x6 + 60
=
=
=
=
=
=
x1
x2 + 100
x3
x4 + 80
x5
Rearrange the equations:
x1
x1 − x2
x2 − x3
x3
Flow out
x1
x2 + 70
x3
x4 + 90
x5
x6 + 80
−
−
x6
−
x6
=
=
=
=
=
=
x4
x4
−
x5
x5
60
70
− 100
90
− 80
80
Reduce the augmented matrix:
ª1
«1
«
«0
«
«0
«0
«
«¬0
0
−1
1
0
0
0
0
0
−1
1
0
0
0
0
0
−1
1
0
0
0
0
0
−1
1
−1
0
0
0
0
−1
60 º
ª1
»
«0
70 »
«
«0
−100 »
» ~ ⋅⋅⋅ ~ «
90 »
«0
»
«0
−80
»
«
80 »¼
«¬0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
−1
−1
−1
−1
−1
0
60 º
−10 »»
90 »
»
0»
80 »
»
0 »¼
x1 = 60 + x6
° x = −10 + x
6
° 2
°° x3 = 90 + x6
.
The general solution is ®
=
x
x
4
6
°
° x5 = 80 + x6
°
°̄ x6 is free
Since x2 cannot be negative, the minimum value of x6 is 10.
Note: The MATLAB box in the Study Guide discusses rational calculations, needed for
balancing the chemical equations in Exercises 10 and 11. As usual, the appendices cover this
material for Maple, Mathematica, and the TI calculators.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
49
50
CHAPTER 1
1.7
• Linear Equations in Linear Algebra
SOLUTIONS
Note: Key exercises are 9–20 and 23–30. Exercise 30 states a result that could be a theorem in the text.
There is a danger, however, that students will memorize the result without understanding the proof, and
then later mix up the words row and column. Exercises 37 and 38 anticipate the discussion in Section 1.9
of one-to-one transformations. Exercise 44 is fairly difficult for my students.
1. Use an augmented matrix to study the solution set of x1u + x2v + x3w = 0 (*), where u, v, and w are
7
9 0º ª 5 7 9 0º
ª5
«
2
4 0»» ~ «« 0 2 4 0»» , there are no free variables. So
the three given vectors. Since «0
«¬0 −6 −8 0»¼ «¬0 0 4 0»¼
the homogeneous equation (*) has only the trivial solution. The vectors are linearly independent.
2. Use an augmented matrix to study the solution set of x1u + x2v + x3w = 0 (*), where u, v, and w are
0 −1 0 º ª 2
0
3 0º
ª0
«
»
«
0
3 0 » ~ « 0 −8 −7 / 2 0»» , there are no free
the three given vectors. Since « 2
«¬ 3 −8
−1 0»¼
1 0 »¼ «¬ 0
0
variables. So the homogeneous equation (*) has only the trivial solution. The vectors are linearly
independent.
3. Use the method of Example 3 (or the box following the example). By comparing entries of the
vectors, one sees that the second vector is –2 times the first vector. Thus, the two vectors are linearly
dependent.
ª −1º ª −3º
4. From the first entries in the vectors, it seems that the second vector of the pair « » , « » may be 3
¬ 3¼ ¬ − 9 ¼
times the first vector. But there is a sign problem with the second entries. So neither of the vectors is
a multiple of the other. The vectors are linearly independent.
5. Use the method of Example 2. Row reduce the augmented matrix for Ax = 0:
9 0 º ª 1 −4 −2 0 º ª 1 −4 −2 0 º ª 1 −4 −2 0 º ª 1 −4 −2
ª 0 −3
« 2
1 −7 0 »» «« 2
1 −7 0 »» ««0
9 −3 0»» «« 0
9 −3 0»» «« 0
9 −3
«
~
~
~
~
« −1
4 −5 0 » « −1
4 −5 0 » «0
0 −7 0 » « 0
0 −7 0» « 0
0 −7
«
» «
» «
» «
» «
9 0 ¼ ¬0 −3
9 0¼ ¬0
0
8 0¼ ¬0
0
0
¬ 1 −4 −2 0 ¼ ¬ 0 −3
There are no free variables. The equation Ax = 0 has only the trivial solution and so the columns of A
are linearly independent.
0º
0»»
0»
»
0¼
6. Use the method of Example 2. Row reduce the augmented matrix for Ax = 0:
ª −4
«
« 0
« 1
«
¬ 2
−3
−1
0
1
1
−5
−10
5
0º ª 1
0 »» «« 0
~
0 » « −4
» «
0¼ ¬ 2
−1
−5
5
−3
1
0
−10
1
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
−1
−5
5
1
−1
−20
0
1
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
−1
−5
5
0
0
−15
−5
1
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
−1
−5
5
0
0
−15
0
1
0º
0 »»
0»
»
0¼
1.7
• Solutions
51
There are no free variables. The equation Ax = 0 has only the trivial solution and so the columns of A
are linearly independent.
7. Study the equation Ax = 0. Some people may start with the method of Example 2:
ª 1
« −2
«
«¬ −4
4
−7
−5
−3
5
7
0
1
5
0º ª 1 4
0 »» ~ ««0
1
»
«
0 ¼ ¬0 11
−3
−1
−5
0
1
5
0º ª 1
0»» ~ ««0
0 »¼ «¬0
−3
−1
6
4
1
0
0
1
−6
0º
0»»
0»¼
But this is a waste of time. There are only 3 rows, so there are at most three pivot positions. Hence, at
least one of the four variables must be free. So the equation Ax = 0 has a nontrivial solution and the
columns of A are linearly dependent.
8. Same situation as with Exercise 7. The (unnecessary) row operations are
ª 1
« −2
«
«¬ 0
−2
4
1
3
−6
−1
2
2
3
0º ª 1
0»» ~ ««0
0 »¼ «¬0
−2
0
1
3
0
−1
2
6
3
0º ª 1
0»» ~ «« 0
0»¼ «¬ 0
−2
1
0
3
−1
0
2
3
6
0º
0»»
0»¼
Again, because there are at most three pivot positions yet there are four variables, the equation
Ax = 0 has a nontrivial solution and the columns of A are linearly dependent.
9. a. The vector v3 is in Span{v1, v2} if and only if the equation x1v1 + x2v2 = v3 has a solution. To find
out, row reduce [v1 v2 v3], considered as an augmented matrix:
ª 1
« −3
«
«¬ 2
−3
9
−6
5º ª 1
−7 »» ~ ««0
h »¼ «¬0
−3
0
0
5 º
8 »»
h − 10 »¼
At this point, the equation 0 = 8 shows that the original vector equation has no solution. So v3 is
in Span{v1, v2} for no value of h.
b. For {v1, v2, v3} to be linearly independent, the equation x1v1 + x2v2 + x3v3 = 0 must have only the
trivial solution. Row reduce the augmented matrix [v1 v2 v3 0]
ª 1
« −3
«
«¬ 2
−3
9
−6
5
−7
h
0º ª 1
0»» ~ ««0
0»¼ «¬0
−3
0
0
5
8
h − 10
0º ª 1
0 »» ~ «« 0
0 »¼ «¬ 0
−3
0
0
5
8
0
0º
0»»
0»¼
For every value of h, x2 is a free variable, and so the homogeneous equation has a nontrivial
solution. Thus {v1, v2, v3} is a linearly dependent set for all h.
10. a. The vector v3 is in Span{v1, v2} if and only if the equation x1v1 + x2v2 = v3 has a solution. To find
out, row reduce [v1 v2 v3], considered as an augmented matrix:
ª 1
« −3
«
¬« −5
−3
9
15
2º ª 1
−5»» ~ ««0
h »¼ «¬0
−3
0
0
2 º
1 »»
h + 10»¼
At this point, the equation 0 = 1 shows that the original vector equation has no solution. So v3 is
in Span{v1, v2} for no value of h.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
52
CHAPTER 1
• Linear Equations in Linear Algebra
b. For {v1, v2, v3} to be linearly independent, the equation x1v1 + x2v2 + x3v3 = 0 must have only the
trivial solution. Row reduce the augmented matrix [v1 v2 v3 0]
ª 1
« −3
«
«¬ −5
−3
9
15
2
−5
h
0º ª 1
0»» ~ «« 0
0»¼ «¬ 0
−3
0
0
2
1
h + 10
0º ª 1
0»» ~ «« 0
0»¼ «¬ 0
−3
0
0
0º
0»»
0»¼
2
1
0
For every value of h, x2 is a free variable, and so the homogeneous equation has a nontrivial
solution. Thus {v1, v2, v3} is a linearly dependent set for all h.
11. To study the linear dependence of three vectors, say v1, v2, v3, row reduce the augmented matrix
[v1 v2 v3 0]:
ª 2
« −2
«
¬« 4
4
−6
7
−2
2
h
0º ª2
0 »» ~ «« 0
0 ¼» ¬« 0
4
−2
−1
−2
0
h+4
0º ª 2
0»» ~ «« 0
0¼» ¬« 0
4
−2
0
−2
0
h+4
0º
0»»
0¼»
The equation x1v1 + x2v2 + x3v3 = 0 has a nontrivial solution if and only if h + 4 = 0 (which
corresponds to x3 being a free variable). Thus, the vectors are linearly dependent if and only if h = –4.
12. To study the linear dependence of three vectors, say v1, v2, v3, row reduce the augmented matrix
[v1 v2 v3 0]:
ª 3
«
« −6
¬« 1
−6
4
−3
0º ª3
0 »» ~ ««0
0 ¼» «¬0
9
h
3
−6
−1
0
9
0
h + 18
0º
0»»
0»¼
The equation x1v1 + x2v2 + x3v3 = 0 has a nontrivial solution if and only if h + 18 = 0 (which
corresponds to x3 being a free variable). Thus, the vectors are linearly dependent if and only if h = –
18.
13. To study the linear dependence of three vectors, say v1, v2, v3, row reduce the augmented matrix
[v1 v2 v3 0]:
ª 1
« 5
«
«¬ −3
−2
−9
6
3
h
−9
0º ª 1
0»» ~ ««0
0 »¼ «¬0
−2
1
0
3
h − 15
0
0º
0»»
0»¼
The equation x1v1 + x2v2 + x3v3 = 0 has a free variable and hence a nontrivial solution no matter what
the value of h. So the vectors are linearly dependent for all values of h.
14. To study the linear dependence of three vectors, say v1, v2, v3, row reduce the augmented matrix
[v1 v2 v3 0]:
ª 1
« −2
«
«¬ −4
−3
7
6
2
1
h
0º ª 1
0»» ~ ««0
0»¼ «¬0
−3
1
−6
2
5
h +8
0º ª 1
0»» ~ «« 0
0 »¼ «¬0
−3
1
0
2
5
h + 38
0º
0»»
0 »¼
The equation x1v1 + x2v2 + x3v3 = 0 has a nontrivial solution if and only if h + 38 = 0 (which
corresponds to x3 being a free variable). Thus, the vectors are linearly dependent if and only
if h = –38.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.7
• Solutions
53
15. The set is linearly dependent, by Theorem 8, because there are four vectors in the set but only two
entries in each vector.
16. The set is linearly dependent because the second vector is –3/2 times the first vector.
17. The set is linearly dependent, by Theorem 9, because the list of vectors contains a zero vector.
18. The set is linearly dependent, by Theorem 8, because there are four vectors in the set but only two
entries in each vector.
19. The set is linearly independent because neither vector is a multiple of the other vector. [Two of the
entries in the first vector are – 4 times the corresponding entry in the second vector. But this multiple
does not work for the third entries.]
20. The set is linearly dependent, by Theorem 9, because the list of vectors contains a zero vector.
21. a.
b.
c.
d.
False. A homogeneous system always has the trivial solution. See the box before Example 2.
False. See the warning after Theorem 7.
True. See Fig. 3, after Theorem 8.
True. See the remark following Example 4.
22. a. True. See Theorem 7.
b. True. See Example 4.
ª 1º
ª 2º
«
»
c. False. For instance, the set consisting of « −2 » and «« –4»» is linearly dependent. See the warning
«¬ 3»¼
«¬ 6»¼
after Theorem 8.
d. False. See Example 3(a).
ª * º ª 0 º ª 0
23. «
»,«
»,«
¬ 0 0¼ ¬0 0 ¼ ¬0
0º
0 »¼
ª * * º
24. «« 0 * »»
«¬ 0 0 »¼
ª * º
ª0 º
« 0 »
«
»
» and «0 0 »
25. «
«0 0»
«0 0 »
«
»
«
»
¬« 0 0 ¼»
¬«0 0 »¼
ª * * º
«0 *»
» . The columns must be linearly independent, by Theorem 7, because the first column is
26. «
« 0 0 »
«
»
¬« 0 0 0 ¼»
not zero, the second column is not a multiple of the first, and the third column is not a linear
combination of the preceding two columns (because a3 is not in Span{a1, a2}).
27. All four columns of the 6×4 matrix A must be pivot columns. Otherwise, the equation Ax = 0 would
have a free variable, in which case the columns of A would be linearly dependent.
28. If the columns of a 4×6 matrix A span R4, then A has a pivot in each row, by Theorem 4. Since each
pivot position is in a different column, A has four pivot columns.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
54
CHAPTER 1
• Linear Equations in Linear Algebra
29. A: any 3×2 matrix with one column a multiple of the other.
B: any 3×2 matrix with two nonzero columns such that neither column is a multiple of the other. In
this case the columns are linearly independent and so the equation Bx = 0 has only the trivial
solution.
30. a. n
b. The columns of A are linearly independent if and only if the equation Ax = 0 has only the trivial
solution. This happens if and only if Ax = 0 has no free variables, which in turn happens if and
only if every variable is a basic variable, that is, if and only if every column of A is a pivot
column.
31. Think of A = [a1 a2 a3]. The text points out that a3 = a1 + a2. Rewrite this as a1 + a2 – a3 = 0. As a
matrix equation, Ax = 0 for x = (1, 1, –1).
32. Think of A = [a1 a2 a3]. The text points out that a1 – 3a2 = a3. Rewrite this as a1 – 3a2 – a3 = 0. As
a matrix equation, Ax = 0 for x = (1, –3, –1).
33. True, by Theorem 7. (The Study Guide adds another justification.)
34. False. The vector v1 could be the zero vector.
35. True, by Theorem 9.
36. False. Counterexample: Take v1 and v2 to be multiples of one vector. Take v3 to be not a multiple of
that vector. For example,
ª1º
ª 2º
ª1 º
«
»
«
»
v1 = «1» , v 2 = « 2» , v 3 = «« 0»»
«¬1»¼
«¬ 2»¼
«¬ 0»¼
37. True. A linear dependence relation among v1, v2, v3 may be extended to a linear dependence relation
among v1, v2, v3, v4 by placing a zero weight on v4.
38. True. If the equation x1v1 + x2v2 + x3v3 = 0 had a nontrivial solution (with at least one of x1, x2, x3
nonzero), then so would the equation x1v1 + x2v2 + x3v3 + 0¸v4 = 0. But that cannot happen because
{v1, v2, v3, v4} is linearly independent. So {v1, v2, v3} must be linearly independent. This problem can
also be solved using Exercise 37, if you know that the statement there is true.
39. If for all b the equation Ax = b has at most one solution, then take b = 0, and conclude that the
equation Ax = 0 has at most one solution. Then the trivial solution is the only solution, and so the
columns of A are linearly independent.
40. An m×n matrix with n pivot columns has a pivot in each column. So the equation Ax = b has no free
variables. If there is a solution, it must be unique.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.7
ª 3
« −5
«
A
=
41. [M]
« 4
«
¬ 8
ª3
«0
~«
«0
«
¬0
−4
−3
10
7
−7
−11
3
−7
5
23
2
4
−4 º ª 3
15»» «« 0
~
1» « 0
» «
15¼ ¬ 0
−4
−29 / 3
10
29 / 3
7
2/3
0
0
0
0
−196 / 29
−418 / 29
−4
−29 / 3
−4 º
29 / 3
2 / 3 25 / 3»»
25 / 3 −25 / 3 −22 / 3 19 / 3»
»
11 / 3 −11 / 3 −44 / 3 77 / 3¼
−4 º ª 3
−4
10
7
»
«
25 / 3» « 0 −29 / 3 29 / 3
2/3
~
392 / 29 » « 0
0
0 −196 / 29
» «
836 / 29 ¼ ¬ 0
0
0
0
ª 3
« −5
Use the pivot columns of A to form B = «
« 4
«
¬ 8
ª 12
« −7
«
42. [M] « 9
«
« −4
«¬ 8
10
−6
9
−3
7
−6
4
−9
−1
−5
8
−5
9
0
6
4
−7
9
−8
1
• Solutions
−4
−3
3
−7
−14 º
ª12
»
«0
9»
«
−18» ~ ⋅⋅⋅ ~ « 0
»
«
1»
«0
«¬ 0
−11»¼
ª 12
« −7
«
Use the pivot columns of A to form B = « 9
«
« −4
«¬ 8
10
−6
9
−3
7
10
55
7
−4 º
25 / 3»»
392 / 29 »
»
0¼
7º
−11»»
. Other choices are possible.
2»
»
4¼
10
−1 / 6
0
0
0
−6
4
−9
−1
−5
−6
1/ 2
−2
0
0
8
−1 / 3
2
0
0
4
−14 / 3
−16
−36
0
−14 º
5 / 6 »»
−2 »
»
0 »
0 »¼
4º
−7 »»
9 » . Other choices are possible.
»
−8»
1»¼
43. [M] Make v any one of the columns of A that is not in B and row reduce the augmented matrix
[B v]. The calculations will show that the equation Bx = v is consistent, which means that v is a
linear combination of the columns of B. Thus, each column of A that is not a column of B is in the set
spanned by the columns of B.
44. [M] Calculations made as for Exercise 43 will show that each column of A that is not a column of B
is in the set spanned by the columns of B. Reason: The original matrix A has only four pivot
columns. If one or more columns of A are removed, the resulting matrix will have at most four pivot
columns. (Use exactly the same row operations on the new matrix that were used to reduce A to
echelon form.) If v is a column of A that is not in B, then row reduction of the augmented matrix
[B v] will display at most four pivot columns. Since B itself was constructed to have four pivot
columns, adjoining v cannot produce a fifth pivot column. Thus the first four columns of [B v] are
the pivot columns. This implies that the equation Bx = v has a solution.
Note: At the end of Section 1.7, the Study Guide has another note to students about “Mastering Linear
Algebra Concepts.” The note describes how to organize a review sheet that will help students form a
mental image of linear independence. The note also lists typical misuses of terminology, in which an
adjective is applied to an inappropriate noun. (This is a major problem for my students.) I require my
students to prepare a review sheet as described in the Study Guide, and I try to make helpful comments on
their sheets. I am convinced, through personal observation and student surveys, that the students who
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
56
CHAPTER 1
• Linear Equations in Linear Algebra
prepare many of these review sheets consistently perform better than other students. Hopefully, these
students will remember important concepts for some time beyond the final exam.
1.8
SOLUTIONS
Notes: The key exercises are 17–20, 25 and 31. Exercise 20 is worth assigning even if you normally
assign only odd exercises. Exercise 25 (and 26) can be used to make a few comments about computer
graphics, even if you do not plan to cover Section 2.6. For Exercise 31, the Study Guide encourages
students not to look at the proof before trying hard to construct it. Then the Guide explains how to create
the proof.
Exercises 19 and 20 provide a natural segue into Section 1.9. I arrange to discuss the homework on
these exercises when I am ready to begin Section 1.9. The definition of the standard matrix in Section 1.9
follows naturally from the homework, and so I’ve covered the first page of Section 1.9 before students
realize we are working on new material.
The text does not provide much practice determining whether a transformation is linear, because the
time needed to develop this skill would have to be taken away from some other topic. If you want your
students to be able to do this, you may need to supplement Exercises 23, 24, 32 and 33.
If you skip the concepts of one-to-one and “onto” in Section 1.9, you can use the result of Exercise 31
to show that the coordinate mapping from a vector space onto Rn (in Section 4.4) preserves linear
independence and dependence of sets of vectors. (See Example 6 in Section 4.4.)
ª2
1. T(u) = Au = «
¬0
0 º ª 1º ª 2 º
, T(v) =
=
2 »¼ «¬ −3»¼ «¬ −6 »¼
ª2
«0
¬
0 º ª a º ª 2a º
=
2 »¼ «¬ b »¼ «¬ 2b »¼
0
0º ª 3º ª 1º
ª1 3
«
0»» «« 6»» = «« 2 »» , T(v) =
2. T(u) = Au = « 0 1 3
0 1 3¼» ¬« −9¼» ¬« −3¼»
¬« 0
ª 1
3. [ A b ] = «« −3
«¬ 2
ª1
«
~ «0
«¬0
0
1
0
−3
−3
1
ª1
«
4. [ A b ] = « 0
«¬ 2
ª1
~ ««0
«¬0
−3
6
−1
0
1
−2
−2º ª 1
3»» ~ ««0
−1»¼ ¬«0
−2 º ª 1
−3»» ~ ««0
3»¼ «¬0
−2
1
−5
−2
1
0
3
−3
6
3
−3
1
0
1
0
0
0
1
−6 º ª 1
−4 »» ~ ««0
−5»¼ «¬0
−6 º ª 1
−4 »» ~ ««0
−1»¼ «¬0
0
1
−2
−2º ª 1
−3»» ~ ««0
3¼» «¬0
7º
ª
»
6 » x = ««
«¬
1»¼
−2
1
−1
0
1
0
−3
−3
5
0
0
1
3
−3
0
0
0º ª a º ª a
ª1 3
« 0 13
0 »» «« b »» = «« b
«
«¬ 0
0 1 3»¼ «¬ c »¼ «¬ c
−3
−3
−1
−2º
−3»»
−3»¼
7º
6 »» , unique solution
1»¼
−6 º ª 1
−4»» ~ ««0
7 »¼ «¬0
−17 º
−7 »»
−1»¼
0
1
0
3º
3»»
3»¼
−2
1
0
3
−3
−3
−6º
−4»»
3»¼
ª −17 º
x = «« −7 »» , unique solution
«¬ −1»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.8
ª 1
5. [ A b ] = «
¬ −3
−5
−7
7
5
−2 º ª 1
~
−2 »¼ «¬ 0
−5
−7
1
2
−2 º ª 1
~
1»¼ «¬0
0
3
1
2
• Solutions
57
3º
1»¼
ª 3º
Note that a solution is not « » . To avoid this common error, write the equations:
¬ 1¼
x1
x2
+ 3 x3
+ 2 x3
x1 = 3 − 3 x3
= 3
°
and solve for the basic variables: ® x2 = 1 − 2 x3
= 1
° x is free
¯ 3
ª x1 º ª3 − 3x3 º ª3º
ª −3º
«
»
«
»
«
»
The general solution is x = « x2 » = «1 − 2 x3 » = «1 » + x3 ««−2»» . For a particular solution, one might
«¬ x3 »¼ «¬ x3 »¼ «¬0 »¼
«¬ 1»¼
ª3º
choose x3 = 0 and x = ««1 »» .
«¬ 0 »¼
ª1
«3
6. [ A b ] = «
«0
«
¬1
equations:
x1
x2
−3
−8
2
8
1
0
2
8
1º ª 1
6 »» «« 0
~
3» « 0
» «
10 ¼ ¬ 0
−3
1
2
2
1
3
2
6
1º ª 1
3»» «« 0
~
3» « 0
» «
9¼ ¬0
−3
1
2
2
0
0
0
0
1º ª 1
3»» ««0
~
0 » «0
» «
0 ¼ ¬0
0
1
8
2
0
0
0
0
10 º
3»»
Write the
0»
»
0¼
x1 = 10 − 8 x3
°
and solve for the basic variables: ® x2 = 3 − 2 x3
3
° x is free
¯ 3
+ 8 x3
= 10
+ 2 x3
=
ª −8º
ª x1 º ª10 − 8 x3 º ª10 º
«
»
«
»
«
»
The general solution is x = « x2 » = « 3 − 2 x3 » = « 3» + x3 «« −2»» . For a particular solution, one might
«¬ x3 »¼ «¬ x3 »¼ «¬ 0 »¼
«¬ 1»¼
ª10 º
choose x3 = 0 and x = «« 3 »» .
«¬ 0 »¼
7. The value of a is 5. The domain of T is R5, because a 6×5 matrix has 5 columns and for Ax to be
defined, x must be in R5. The value of b is 6. The codomain of T is R6, because Ax is a linear
combination of the columns of A, and each column of A is in R6.
8. The matrix A must have 7 rows and 5 columns. For the domain of T to be R5, A must have five
columns so that Ax is defined for x in R5. For the codomain of T to be R7, the columns of A must
have seven entries (in which case A must have seven rows), because Ax is a linear combination of the
columns of A.
ª1
«
9. Solve Ax = 0: « 0
«¬ 2
−3
1
−4
5
−3
4
−5
5
−4
0º ª 1
0 »» ~ ««0
0 »¼ «¬0
−3
1
2
5
−3
−6
−5
5
6
0º ª 1
0»» ~ «« 0
0»¼ «¬ 0
−3
1
0
5
−3
0
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
−5
5
−4
0º
0»»
0»¼
58
CHAPTER 1
ª1
~ ««0
«¬0
0
1
0
• Linear Equations in Linear Algebra
−4
−3
0
0º
0»»
0»¼
0
0
1
x1
x2
− 4 x3
− 3 x3
x4
x1 = 4 x3
= 0 °
° x2 = 3x3
= 0, ®
x3 is free
= 0 °
°̄ x4 = 0
ª x1 º ª 4 x3 º
ª4º
« x » «3x »
« 3»
x = « 2 » = « 3 » = x3 « »
« x3 » « x3 »
« 1»
« » « »
« »
¬0¼
¬ x4 ¼ ¬ 0 ¼
ª3
«1
«
10. Solve Ax = 0.
«0
«
¬1
ª1
«0
~«
«0
«
¬0
0
1
2
2
−4
3
2
4
4
8
6
12
x1
x2
2
0
10
2
−6
−4
1
2
3
4
10
8
0º ª 1
0 »» «« 3
~
0 » «0
» «
0¼ ¬ 1
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
+ 2 x3
− 4 x4
= 0
+ 2 x3
+
= 0
3 x4
0
1
2
2
−4
3
0
0
0
0
0
0
0
2
2
10
−4
−6
1
2
3
4
10
8
0º ª 1
0 »» «« 0
~
0» «0
» «
0 ¼ ¬0
0
2
2
4
−4
6
1
2
3
4
8
12
0º
0 »»
0»
»
0¼
0º
0 »»
0»
»
0¼
x1 = −2 x3 + 4 x4
° x = −2 x − 3 x
° 2
3
4
®
° x3 is free
°̄ x4 is free
ª −2 x3 º ª 4 x4 º
ª −2 º
ª 4º
« −2 x » « −3 x »
« −2 »
« −3»
4»
3»
«
«
«
»
x=
+
=x
+x « »
« x3 » « 0 » 3 « 1» 4 « 0 »
»
«
» «
« »
« »
¬ 0 ¼ ¬ x4 ¼
¬ 0¼
¬ 1¼
11. Is the system represented by [A b] consistent? Yes, as the following calculation shows.
ª1
«0
«
¬« 2
−3
1
−4
5
−3
4
−5
5
−4
−1º ª 1
1»» ~ ««0
0 ¼» ¬« 0
−3
1
2
5
−3
−6
−5
5
6
−1º ª 1
1»» ~ «« 0
2¼» ¬« 0
−3
1
0
5
−3
0
−5
5
−4
−1º
1»»
0»¼
The system is consistent, so b is in the range of the transformation x 6 Ax .
12. Is the system represented by [A b] consistent?
ª3
«1
«
«0
«
¬1
ª1
«0
~«
«0
«
¬0
2
0
10
2
−6
−4
1
4
2
10
3
8
0
1
2
2
−4
3
2
4
4
8
6
12
−1º ª 1
3»» «« 3
~
−1» « 0
» «
4¼ ¬ 1
3º ª 1
−1»» «« 0
~
−10 » « 0
» «
1¼ ¬ 0
2
10
−4
−6
1 2
4 10
3
8
0
2
3º ª 1
−1»» ««0
~
−1» «0
» «
4 ¼ ¬0
0
1
2
2
−4
3
0
0
0
0
0
0
0
2
2
4
−4
6
1
4
2
8
3
12
3º
−10 »»
−1»
»
1¼
3º
−1»»
−8»
»
5¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.8
• Solutions
59
The system is inconsistent, so b is not in the range of the transformation x 6 Ax .
13.
14.
A reflection through the origin.
A scaling by the factor 2.
The transformation in Exercise 13 may also be described as a rotation of π radians about the origin or
a rotation of –π radians about the origin.
15.
16.
A reflection through the line x2 = x1.
axis.
A scaling by a factor of 2 and a projection onto the x2
ª 4 º ª 8º
17. T(2u) = 2T(u) = 2 « » = « » , T(3v) = 3T(v) =
¬ 1¼ ¬ 2 ¼
ª −1º ª −3º
3 « » = « » , and
¬ 3¼ ¬ 9 ¼
ª 8 º ª −3 º ª 5 º
T(2u + 3v) = 2T(u) + 3T(v) = « » + « » = « » .
¬ 2 ¼ ¬ 9 ¼ ¬11¼
18. Draw a line through w parallel to v, and draw a line through w parallel to u. See the left part of the
figure below. From this, estimate that w = u + 2v. Since T is linear, T(w) = T(u) + 2T(v). Locate T(u)
and 2T(v) as in the right part of the figure and form the associated parallelogram to locate T(w).
x2
w
2v v
x2
u
2T(v)
T(w)
T(v)
x1
x1
T(u)
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
60
CHAPTER 1
• Linear Equations in Linear Algebra
19. All we know are the images of e1 and e2 and the fact that T is linear. The key idea is to write
ª 5º
ª 1º ª 0 º
x = « » = 5 « » − 3 « » = 5 e1 − 3 e 2 . Then, from the linearity of T, write
¬ −3 ¼
¬ 0 ¼ ¬ 1¼
ª 2 º ª −1º ª13º
T(x) = T(5e1 – 3e2) = 5T(e1) – 3T(e2) = 5y1 – 3y2 = 5 « » − 3 « » = « » .
¬ 5¼ ¬ 6¼ ¬ 7 ¼
ªx º
ªx º
ª 1º
ª0º
To find the image of « 1 » , observe that x = « 1 » = x1 « » + x2 « » = x1e1 + x2e 2 . Then
¬0¼
¬ 1¼
¬ x2 ¼
¬ x2 ¼
ª2º
ª −1º ª 2 x − x º
T(x) = T(x1e1 + x2e2) = x1T(e1) + x2T(e2) = x1 « » + x2 « » = « 1 2 »
¬ 5¼
¬ 6 ¼ ¬5 x1 + 6 x2 ¼
20. Use the basic definition of Ax to construct A. Write
ª x º ª −3
T ( x) = x1 v1 + x2 v 2 = [ v1 v 2 ] « 1 » = «
¬ x2 ¼ ¬ 5
7º
ª −3
x, A = «
»
−2 ¼
¬ 5
7º
−2 »¼
21. a. True. Functions from Rn to Rm are defined before Fig. 2. A linear transformation is a function
with certain properties.
b. False. The domain is R5. See the paragraph before Example 1.
c. False. The range is the set of all linear combinations of the columns of A. See the paragraph
before Example 1.
d. False. See the paragraph after the definition of a linear transformation.
e. True. See the paragraph following the box that contains equation (4).
22. a. True. See the subsection on Matrix Transformations.
b. True. See the subsection on Linear Transformations.
c. False. The question is an existence question. See the remark about Example 1(d), following the
solution of Example 1.
d. True. See the discussion following the definition of a linear transformation.
e. True. T(0) = 0. See the box after the definition of a linear transformation.
23. a. When b = 0, f (x) = mx. In this case, for all x,y in R and all scalars c and d,
f (cx + dy) = m(cx + dy) = mcx + mdy = c(mx) + d(my) = c f (x) + d f (y)
This shows that f is linear.
b. When f (x) = mx + b, with b nonzero, f(0) = m(0) = b = b ≠ 0. This shows that f is not linear,
because every linear transformation maps the zero vector in its domain into the zero vector in the
codomain. (In this case, both zero vectors are just the number 0.) Another argument, for instance,
would be to calculate f (2x) = m(2x) + b and 2f (x) = 2mx + 2b. If b is nonzero, then f (2x) is not
equal to 2f (x) and so f is not a linear transformation.
c. In calculus, f is called a “linear function” because the graph of f is a line.
24. Let T(x) = Ax + b for x in Rn. If b is not zero, T(0) = A0 + b = b ≠ 0. Actually, T fails both
properties
of a linear transformation. For instance, T(2x) = A(2x) + b = 2Ax + b, which is not the same as 2T(x)
= 2(Ax + b) = 2Ax + 2b. Also,
T(x + y) = A(x + y) + b = Ax + Ay + b
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.8
• Solutions
61
which is not the same as
T(x) + T(y) = Ax + b + Ay + b
25. Any point x on the line through p in the direction of v satisfies the parametric equation
x = p + tv for some value of t. By linearity, the image T(x) satisfies the parametric equation
T(x) = T(p + tv) = T(p) + tT(v)
(*)
If T(v) = 0, then T(x) = T(p) for all values of t, and the image of the original line is just a single
point. Otherwise, (*) is the parametric equation of a line through T(p) in the direction of T(v).
26. a. From the figure following Exercise 22 in Section 1.5, the line through p and q is in the direction
of q – p, and so the equation of the line is x = p + t(q – p) = p + tq – tp = (1 – t)p + tq.
b. Consider x = (1 – t)p + tq for t such that 0 < t < 1. Then, by linearity of T,
T(x) = T((1 – t)p + tq) = (1 – t)T(p) + tT(q) 0 < t < 1
(*)
If T(p) and T(q) are distinct, then (*) is the equation for the line segment between T(p) and T(q),
as shown in part (a) Otherwise, the set of images is just the single point T(p), because
(1 – t)T(p) + tT(q) =(1 – t)T(p) + tT(p) = T(p)
27. Any point x on the plane P satisfies the parametric equation x = su + tv for some values of s and t.
By linearity, the image T(x) satisfies the parametric equation
T(x) = sT(u) + tT(v)
(s, t in R)
The set of images is just Span{T(u), T(v)}. If T(u) and T(v) are linearly independent, Span{T(u),
T(v)} is a plane through T(u), T(v), and 0. If T(u) and T(v) are linearly dependent and not both zero,
then Span{T(u), T(v)} is a line through 0. If T(u) = T(v) = 0, then Span{T(u), T(v)} is {0}.
28. Consider a point x in the parallelogram determined by u and v, say x = au + bv for 0 < a < 1,
0 < b < 1. By linearity of T, the image of x is
T(x) = T(au + bv) = aT(u) + bT(v), for 0 < a < 1, 0 < b < 1
This image point lies in the parallelogram determined by T(u) and T(v). Special “degenerate” cases
arise when T(u) and T(v) are linearly dependent. If one of the images is not zero, then the
“parallelogram” is actually the line segment from 0 to T(u) + T(v). If both T(u) and T(v) are zero,
then the parallelogram is just {0}. Another possibility is that even u and v are linearly dependent, in
which case the original parallelogram is degenerate (either a line segment or the zero vector). In this
case, the set of images must be degenerate, too.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
62
CHAPTER 1
• Linear Equations in Linear Algebra
29.
30. Given any x in Rn, there are constants c1, …, cp such that x = c1v1 +
Rn. Then, from property (5) of a linear transformation,
T(x) = c1T(v1) + + cpT(vp) = c10 + + cp0 = 0
cpvp, because v1, …, vp span
31. (The Study Guide has a more detailed discussion of the proof.) Suppose that {v1, v2, v3} is linearly
dependent. Then there exist scalars c1, c2, c3, not all zero, such that
c1v1 + c2v2 + c3v3 = 0
Then T(c1v1 + c2v2 + c3v3) = T(0) = 0. Since T is linear,
c1T(v1) + c2T(v2) + c3T(v3) = 0
Since not all the weights are zero, {T(v1), T(v2), T(v3)} is a linearly dependent set.
32. Take any vector (x1, x2) with x2 ≠ 0, and use a negative scalar. For instance, T(0, 1) = (–2, –4), but
T(–1 (0, 1)) = T(0, –1) = (–2, 4) ≠ (–1) T(0, 1).
33. One possibility is to show that T does not map the zero vector into the zero vector, something that
every linear transformation does do. T(0, 0) = (0, –3, 0).
34. Take u and v in R3 and let c and d be scalars. Then
cu + dv = (cu1 + dv1, cu2 + dv2, cu3 + dv3). The transformation T is linear because
T(cu + dv) = (cu1 + dv1, cu2 + dv2, – (cu3 + dv3)) = (cu1 + dv1, cu2 + dv2, cu3 – dv3)
= (cu1, cu2, –cu3) + (dv1, dv2, –dv3) = c(u1, u2, –u3) + d(v1, v2, –v3)
= cT(u) + dT(v)
35. Take u and v in R3 and let c and d be scalars. Then
cu + dv = (cu1 + dv1, cu2 + dv2, cu3 + dv3). The transformation T is linear because
T(cu + dv) = (cu1 + dv1, 0, cu3 + dv3) = (cu1, 0, cu3) + (dv1, 0, dv3)
= c(u1, 0, u3) + d(v1, 0, v3)
= cT(u) + dT(v)
36. Suppose that {u, v} is a linearly independent set in Rn and yet T(u) and T(v) are linearly dependent.
Then there exist weights c1, c2, not both zero, such that c1T(u) + c2T(v) = 0 . Because T is linear,
T(c1u + c2v) = 0. That is, the vector x = c1u + c2v satisfies T(x) = 0. Furthermore, x cannot be the
zero vector, since that would mean that a nontrivial linear combination of u and v is zero, which is
impossible because u and v are linearly independent. Thus, the equation T(x) = 0 has a nontrivial
solution.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.8
ª 2
« −7
37. [M] «
« −3
«
¬ −9
3
7
5
0
−5
0
4
3
1
−6
3
−4
ª3
«5
38. [M] «
«6
«
¬9
4
−8
−7
7
0
4
−8
−7
6
−2
4
0
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
0º ª 1
0 »» «« 0
~
0» «0
» «
0¼ ¬0
0
1
1
1
0
0
0
0
0
0
1
0
0
1
0
0
1
1
0
0
1
0
1
0
0º
0 »»
,
0»
»
0¼
0º
0 »»
,
0»
»
0¼
x1 = – x3
°x = – x
° 2
3
®
° x3 is free
°¯ x4 = 0
x1 = − x4
°
° x2 = − x4
®
° x3 = − x4
°̄ x4 is free
• Solutions
63
ª −1º
« −1»
x = x3 « »
« 1»
« »
¬ 0¼
ª −1º
« −1»
x = x4 « »
« −1»
« »
¬ 1¼
5 −5
8º ª 1 0 1 0 1º
ª 2 3
« −7 7
0
0
7 »» ««0 1 1 0 2 »»
~
, yes, b is in the range of the transformation,
39. [M] «
« −3 4
1
3
5» «0 0 0 1 0 »
«
» «
»
¬ −9 3 −6 −4 −3¼ ¬0 0 0 0 0 ¼
because the augmented matrix shows a consistent system. In fact,
x1 = 1 – x3
ª1 º
°x = 2 – x
«2»
° 2
3
; when x3 = 0 a solution is x = « » .
the general solution is ®
«0 »
is
free
x
° 3
« »
°¯ x4 = 0
¬0 ¼
4 −7 0
4 º ª 1 0 0 1 1º
ª3
« 5 −8
7 4 −4 »» «« 0 1 0 1 2 »»
~
40. [M] «
, yes, b is in the range of the transformation,
«6 −8
6 4 −4 » « 0 0 1 1 1»
«
» «
»
¬9 −7 −2 0 −7 ¼ ¬ 0 0 0 0 0 ¼
because the augmented matrix shows a consistent system. In fact,
x1 = 1 − x4
ª 1º
°x = 2 − x
«2»
° 2
4
; when x4 = 0 a solution is x = « » .
the general solution is ®
« 1»
1
x
=
−
x
4
° 3
« »
°̄ x4 is free
¬0¼
Notes: At the end of Section 1.8, the Study Guide provides a list of equations, figures, examples, and
connections with concepts that will strengthen a student’s understanding of linear transformations. I
encourage my students to continue the construction of review sheets similar to those for “span” and
“linear independence,” but I refrain from collecting these sheets. At some point the students have to
assume the responsibility for mastering this material.
If your students are using MATLAB or another matrix program, you might insert the definition of
matrix multiplication after this section, and then assign a project that uses random matrices to explore
properties of matrix multiplication. See Exercises 34–36 in Section 2.1. Meanwhile, in class you can
continue with your plans for finishing Chapter 1. When you get to Section 2.1, you won’t have much to
do. The Study Guide’s MATLAB note for Section 2.1 contains the matrix notation students will need for
a project on matrix multiplication. The appendices in the Study Guide have the corresponding material for
Mathematica, Maple, and the TI-83+/84+/89 calculators.
.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
64
CHAPTER 1
1.9
• Linear Equations in Lineaar Algebra
SOLUTIONS
Notes: This section is optional if youu plan to treat linear transformations only lightlyy, but many
instructors will want to cover at least T
Theorem 10 and a few geometric examples. Exercisees 15 and 16
illustrate a fast way to solve Exercises 17–22 without explicitly computing the images of the standard
basis.
m (in Section
The purpose of introducing one-to-one and onto is to prepare for the term isomorphism
4.4) and to acquaint math majors with thhese terms. Mastery of these concepts would require a substantial
digression, and some instructors prefer to omit these topics (and Exercises 25–40). In this caase, you can
or space onto
use the result of Exercise 31 in Section 1.8 to show that the coordinate mapping from a vecto
E
6 in
Rn (in Section 4.4) preserves linear indeependence and dependence of sets of vectors. (See Example
Section 4.4.) The notions of one-to-one and onto appear in the Invertible Matrix Theorem (S
Section 2.3),
but can be omitted there if desired
Exercises 25–28 and 31–36 offer fairly easy writing practice. Exercises 31, 32, and 35 provide
important links to earlier material.
ª3
«1
1. A = [T(e1) T(e2)] = «
«3
«
¬« 1
−5 º
2 »»
0»
»
0 ¼»
ª1
2. A = [T(e1) T(e2) T(e3)] = «
¬4
−2
9
3º
−8 »¼
ª 1º
3. T(e1) = e1 – 3e2 = « » , T(e2) = e2, A =
¬ −3¼
ª2º
4. T(e1) = e1, T(e2) = e2 + 2e1 = « » , A =
¬ 1¼
ª 1
« −3
¬
ª1
«0
¬
0º
1»¼
2º
1»¼
5. T(e1) = e2, T(e2) = –e1. A = [e 2
ª0
−e1 ] = «
¬1
− 1º
0 »¼
6. T(e1) = e2, T(e2) = –e1. A = [e 2
ª0
−e1 ] = «
¬1
− 1º
0 »¼
7. Follow what happens to e1 and e2. S
Since e1 is on the unit
circle in the plane, it rotates throughh –3 /4 radians into a
point on the unit circle that lies in thhe third quadrant and
on the line x2 = x1 (that is, y = x inn more familiar notation).
The point (–1,–1) is on the ine x2 = x1 , but its distance
from the origin is 2. So the rotatioonal image of e1 is
(–1/ 2, –1/ 2) . Then this image reeflects in the horizontal
axis to (–1/ 2,1/ 2) . Similarly, e2 rotates into a point on
the unit circle that lies in the secondd quadrant and on the
Copyright © 2012 Pearrson Education, Inc. Publishing as Addison-Wesley.
1.9
• Solutions
65
line x2 = − x1 , namely, (1 / 2, –1 / 2) . Then this image reflects in the horizontal axis to
(1 / 2,1 / 2) . When the two calculations described above are written in vertical vector notation, the
transformation’s standard matrix [T(e1) T(e2)] is easily seen:
ª −1/ 2 º ª −1/ 2 º
ª 1/ 2 º ª1/ 2 º
ª −1/ 2
e1 → «
»→«
» , e2 → «
»→«
», A= «
«¬ 1/ 2
¬« −1/ 2 »¼ ¬« 1/ 2 ¼»
¬« −1/ 2 ¼» ¬«1/ 2 ¼»
1/ 2 º
»
1/ 2 ¼»
8. The horizontal shear maps e1 into e1, and then the reflection in the line x2 = –x1 maps e1 into –e2. (See
Table 1.) The horizontal shear maps e2 into e2 into e2 + 2e1. To find the image of e2 + 2e1 when it is
reflected in the line x2 = –x1, use the fact that such a reflection is a linear transformation. So, the
image of e2 + 2e1 is the same linear combination of the images of e2 and e1, namely,
–e1 + 2(–e2) = – e1 – 2e2. To summarize,
ª 0
e1 → e1 → −e 2 and e 2 → e 2 + 2e1 → −e1 − 2e 2 , so A = «
¬ −1
ª 0
9. e1 → e1 → −e2 and e 2 → −e 2 → −e1 , so A = «
¬ −1
10. e1 → e1 → e 2 and e 2 → −e 2 → −e1 , so A = [e 2
−1º
−2 »¼
−1º
0 »¼
ª0
−e1 ] = «
¬1
−1º
0 »¼
11. The transformation T described maps e1 → e1 → −e1 and maps e2 → −e 2 → −e 2 . A rotation through
π radians also maps e1 into –e1 and maps e2 into –e2. Since a linear transformation is completely
determined by what it does to the columns of the identity matrix, the rotation transformation has the
same effect as T on every vector in 2 .
12. The transformation T in Exercise 10 maps e1 → e1 → e 2 and maps e2 → −e2 → −e1 . A rotation about
the origin through π / 2 radians also maps e1 into e2 and maps e2 into –e1. Since a linear
transformation is completely determined by what it does to the columns of the identity matrix, the
rotation transformation has the same effect as T on every vector in 2 .
13. Since (2, 1)=2 e1+ e2, the image of (2, 1) under T is 2T(e1) + T(e2), by linearity of T. On the figure in
the exercise, locate 2T(e1) and use it with T(e2) to form the parallelogram shown below.
14. Since T ( x) = Ax = [a1 a 2 ] x = x1a1 + x2a 2 = a1 − 2a 2 , when x = (1, –2), the image of x is located by
forming the parallelogram shown below.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
66
CHAPTER 1
• Linear Equations in Linear Algebra
ª2
15. By inspection, «« 1
«¬ 0
−4
0
−1
0 º ª x1 º ª 2 x1 − 4 x2 º
−1»» «« x2 »» = «« x1 − x3 »»
3»¼ ¬« x3 ¼» ¬« − x2 + 3x3 ¼»
ª3
16. By inspection, «« 1
«¬0
−2 º
ª3 x1 − 2 x2 º
ª x1 º «
»
4 » « » = « x1 + 4 x2 »»
x
1»¼ ¬ 2 ¼ «¬ x2 »¼
17. To express T(x) as Ax , write T(x) and x as column vectors, and then fill in the entries in A by
inspection, as done in Exercises 15 and 16. Note that since T(x) and x have four entries, A must be a
4×4 matrix.
ª x1 + 2 x2 º ª
« 0 » «
»=«
T(x) = «
« 2 x2 + x4 » «
«
» «
¬ x2 − x4 ¼ ¬
A
º ª x1 º ª 1
» « x » «0
»« 2» = «
» « x3 » « 0
»« » «
¼ ¬ x4 ¼ ¬ 0
2
0
0
0
2
1
0
0
0 º ª x1 º
0 »» «« x2 »»
1» « x3 »
»« »
−1¼ ¬ x4 ¼
18. As in Exercise 17, write T(x) and x as column vectors. Since x has 2 entries, A has 2 columns. Since
T(x) has 4 entries, A has 4 rows.
ª x1 + 4 x2 º ª
º
ª1
« 0 » «
» x
«
«
» = « A » ª 1 º = «0
« x1 − 3 x2 » «
» «¬ x2 »¼ « 1
«
» «
»
«
¼
¬1
¬ x1 ¼ ¬
4º
0»» ª x1 º
−3» «¬ x2 »¼
»
0¼
19. Since T(x) has 2 entries, A has 2 rows. Since x has 3 entries, A has 3 columns.
ª x1 − 5 x2 + 4 x3 º ª
«
»=«
¬ x2 − 6 x3 ¼ ¬
ª x1 º
º « » ª1
» « x2 » = «
¼ « x » ¬0
¬ 3¼
A
ª x1 º
4º « »
x2
−6 ¼» « »
«¬ x3 »¼
−5
1
20. Since T(x) has 1 entry, A has 1 row. Since x has 4 entries, A has 4 columns.
[3 x1 + 4 x3 − 2 x4 ] = [
A
ª x1 º
«x »
] « 2 » = [3
« x3 »
« »
¬ x4 ¼
0
4
ª x1 º
«x »
−2] « 2 »
« x3 »
« »
¬ x4 ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.9
• Solutions
67
ª 3º
ª x +x º ª
º ª x º ª 1 1º ª x1 º
21. T(x) = « 1 2 » = « A » « 1 » = «
. To solve T(x) = « » , row reduce the augmented
«
»
»
¬8 ¼
¼ ¬ x2 ¼ ¬ 4 5¼ ¬ x2 ¼
¬ 4 x1 + 5 x2 ¼ ¬
3º ª 1 0
7º
ª 1 1 3º ª 1 1
ª 7º
matrix: «
~«
~«
, x=« ».
»
»
»
¬ 4 5 8¼ ¬ 0 1 −4 ¼ ¬ 0 1 −4 ¼
¬ −4 ¼
ª 2 x1 − x2 º ª
º
ª 2
ª x1 º «
«
»
«
»
22. T(x) = « −3 x1 + x2 » = « A » « » = « −3
x
«¬ 2 x1 − 3 x2 »¼ «¬
»¼ ¬ 2 ¼ «¬ 2
augmented matrix:
ª 2
« −3
«
«¬ 2
−1
1
−3
0º ª2
−1»» ~ «« 0
−4 »¼ «¬ 0
−1
−1 2
−2
0º ª 2
−1»» ~ «« 0
−4»¼ «¬ 0
−1º
ªx º
1»» « 1 » . To solve T(x) =
x
−3»¼ ¬ 2 ¼
−1
1
−2
0º ª 2
2»» ~ «« 0
−4»¼ «¬ 0
−1
1
0
ª 0º
« −1» , row reduce the
« »
«¬ −4 »¼
0º ª 2
2»» ~ «« 0
0»¼ «¬ 0
0
1
0
2º ª 1
2»» ~ ««0
0»¼ «¬0
0
1
0
1º
2»» ,
0»¼
ª 1º
x = « ».
¬ 2¼
23. a.
b.
c.
d.
True. See Theorem 10.
True. See Example 3.
False. See the paragraph before Table 1.
False. See the definition of onto. Any function from Rn to Rm maps each vector onto another
vector.
e. False. See Example 5.
24. a.
b.
c.
d.
False. See Theorem 12.
True. See Theorem 10.
True. See Theorem 10.
False. See the definition of one-to-one. Any function from Rn to Rm maps a vector onto a single
(unique) vector.
e. False. See Table 3.
25. A row interchange and a row replacement on the standard matrix A of the transformation T in
ª1 2 0 0 º
«0 1 0 −1»
» . This matrix shows that A has only three pivot positions, so
Exercise 17 produce «
«0 0 0 3 »
«
»
¬0 0 0 0 ¼
the equation Ax = 0 has a nontrivial solution. By Theorem 11, the transformation T is not one-to-one.
Also, since A does not have a pivot in each row, the columns of A do not span R4. By Theorem 12, T
does not map R4 onto R4.
26. The standard matrix A of the transformation T in Exercise 2 is 2×3. Its columns are linearly
dependent because A has more columns than rows. So T is not one-to-one, by Theorem 12. Also, A is
3º
ª 1 −2
2
row equivalent to «
» , which shows that the rows of A span R . By Theorem 12, T maps
−
0
17
20
¬
¼
R3onto R2.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
68
CHAPTER 1
• Linear Equations in Linear Algebra
4º
ª 1 −5
27. The standard matrix A of the transformation T in Exercise 19 is «
. The columns of A
1 −6 »¼
¬0
are linearly dependent because A has more columns than rows. So T is not one-to-one, by Theorem
12. Also, A has a pivot in each row, so the rows of A span R2. By Theorem 12, T maps R3 onto R2.
28. The standard matrix A of the transformation T in Exercise 14 has linearly independent columns,
because the figure in that exercise shows that a1 and a2 are not multiples. So T is one-to-one, by
Theorem 12. Also, A must have a pivot in each column because the equation Ax = 0 has no free
ª * º
variables. Thus, the echelon form of A is «
» . Since A has a pivot in each row, the columns of A
¬ 0 ¼
span R2. So T maps R2 onto R2. An alternate argument for the second part is to observe directly from
the figure in Exercise 14 that a1 and a2 span R2. This is more or less evident, based on experience
with grids such as those in Figure 8 and Exercise 7 of Section 1.3.
29. By Theorem 12, the columns of the standard matrix A must be linearly independent and hence the
equation Ax = 0 has no free variables. So each column of A must be a pivot column:
ª * * º
«0 *»
» . Note that T cannot be onto because of the shape of A.
A~ «
« 0 0 »
«
»
¬« 0 0 0 ¼»
30. By Theorem 12, the columns of the standard matrix A must span 3. By Theorem 4, the matrix must
have a pivot in each row. There are four possibilities for the echelon form:
ª * * *º ª * * * º ª * * * º ª 0 * * º
« 0 * *» , « 0 * * » , « 0 0 * » , « 0 0 * »
«
» «
» «
» «
»
«¬ 0 0 *»¼ «¬ 0 0 0 »¼ «¬ 0 0 0 »¼ «¬ 0 0 0 »¼
Note that T cannot be one-to-one because of the shape of A.
31. “T is one-to-one if and only if A has n pivot columns.” By Theorem 12(b), T is one-to-one if and only
if the columns of A are linearly independent. And from the statement in Exercise 30 in Section 1.7,
the columns of A are linearly independent if and only if A has n pivot columns.
32. The transformation T maps Rn onto Rm if and only if the columns of A span Rm, by Theorem 12. This
happens if and only if A has a pivot position in each row, by Theorem 4 in Section 1.4. Since A has m
rows, this happens if and only if A has m pivot columns. Thus, “T maps Rn onto Rm if and only A has
m pivot columns.”
33. Define T : n → m by T(x) = Bx for some m×n matrix B, and let A be the standard matrix for T.
By definition, A = [T(e1) ⋅ ⋅ ⋅ T(en)], where ej is the jth column of In. However, by matrix-vector
multiplication, T(ej) = Bej = bj, the jth column of B. So A = [b1 ⋅ ⋅ ⋅ bn] = B.
34. Take u and v in Rp and let c and d be scalars. Then
T(S(cu + dv)) = T(c⋅S(u) + d⋅S(v))
because S is linear
= c⋅T(S(u)) + d⋅T(S(v))
because T is linear
This calculation shows that the mapping x → T(S(x)) is linear. See equation (4) in Section 1.8.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.10
• Solutions
69
35. If T : n → m maps n onto m , then its standard matrix A has a pivot in each row, by Theorem
12 and by Theorem 4 in Section 1.4. So A must have at least as many columns as rows. That is, m <
n. When T is one-to-one, A must have a pivot in each column, by Theorem 12, so m > n.
36. The transformation T maps Rn onto Rm if and only if for each y in Rm there exists an x in Rn such that
y = T(x).
6
1º
−5
−6 º
ª −5
ª1 0 0
« 8
»
«
3
8»
0 1 0 −1»»
−3
«
«
~ ⋅⋅⋅ ~
. There is no pivot in the fourth column of
37. [M]
« 2
« 0 0 1 −1»
9
5 −12 »
«
»
«
»
2
7 −12 ¼
0¼
¬ −3
¬0 0 0
the standard matrix A, so the equation Ax = 0 has a nontrivial solution. By Theorem 11, the
transformation T is not one-to-one. (For a shorter argument, use the result of Exercise 31.)
5
9 −9 º
ª 7
ª 1 0 0 0º
« 5
»
«0 1 0 0 »
6
4 −4 »
» . Yes. There is a pivot in every column of the
~ ⋅⋅⋅ ~ «
38. [M] «
« 4
«0 0 1 0 »
8
0
7»
«
»
«
»
6
5¼
¬ −6 −6
¬0 0 0 1¼
standard matrix A, so the equation Ax = 0 has only the trivial solution. By Theorem 11, the transformation T is one-to-one. (For a shorter argument, use the result of Exercise 31.)
3
7
5º
0
0
5
0º
ª 4 −7
ª 1
« 6 −8
»
«
5 12 −8»
1
0
1
0 »»
«
« 0
39. [M] « −7 10 −8 −9 14 » ~ ⋅⋅⋅ ~ « 0
0
1 −2
0 » . There is not a pivot in every row,
«
»
«
»
4
2 −6 »
0
0
0
1»
« 3 −5
« 0
«¬ −5
«¬ 0
6 −6 −7
3»¼
0
0
0
0 »¼
5
so the columns of the standard matrix do not span R . By Theorem 12, the transformation T does not
map R5 onto R5.
5
6 −1º
ª 9 43
ª 1 0 0 0 0º
« 14 15 −7 −5
»
«0 1 0 0 0»
4»
«
«
»
40. [M] « −8 −6 12 −5 −9 » ~ ⋅⋅⋅ ~ « 0 0 1 0 0 » . There is a pivot in every row, so the
«
»
«
»
9
8»
« −5 −6 −4
«0 0 0 1 0»
«¬ 13 14 15
«¬ 0 0 0 0 1»¼
3 11»¼
5
columns of the standard matrix span R . By Theorem 12, the transformation T maps R5 onto R5.
1.10
SOLUTIONS
1. a. If x1 is the number of servings of Cheerios and x2 is the number of servings of 100% Natural
Cereal, then x1 and x2 should satisfy
ª nutrients º
ª nutrients º ª quantities º
x1 « per serving » + x2 «per serving of » = «of nutrients »
«¬of Cheerios »¼
«¬100% Natural »¼ «¬ required »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
70
CHAPTER 1
• Linear Equations in Linear Algebra
That is,
ª110 º
ª130 º ª 295º
« 4»
« 3» « 9 »
»
x1 « » + x2 « » = «
« 20 »
« 18» « 48»
« »
« » «
»
¬« 2 ¼»
¬« 5¼» ¬« 8¼»
ª110
« 4
b. The equivalent matrix equation is «
« 20
«
«¬ 2
augmented matrix for this equation.
ª110
« 4
«
« 20
«
«¬ 2
ª1
«0
~«
«0
«
¬« 0
130
3
18
5
295º ª 2
9 »» «« 4
~
48» « 20
» «
8»¼ «¬110
2.5
−7
−16
−145
4º ª 1
−7 »» «« 0
~
−16 » « 0
» «
−145¼» ¬« 0
5
3
18
130
2.5
1
0
0
130 º
ª 295º
»
3» ª x1 º «« 9 »»
=
. To solve this, row reduce the
18» «¬ x2 »¼ « 48»
»
«
»
5»¼
«¬ 8»¼
8º ª 1 2.5
9 »» «« 4
3
~
48» « 10
9
» «
295»¼ «¬110 130
4º ª 1
1»» «« 0
~
0» «0
» «
0 ¼» ¬« 0
0
1
0
0
4º
9 »»
24 »
»
295»¼
1.5º
1 »»
0»
»
0 ¼»
The desired nutrients are provided by 1.5 servings of Cheerios together with 1 serving of 100%
Natural Cereal.
2. Set up nutrient vectors for one serving of Shredded Wheat (SW) and Kellogg's Crispix (Crp):
Nutrients:
SW
Crp
calories
ª160 º
« 5»
« »
« 6»
« »
¬ 1¼
ª110 º
« 2» .
« »
« .1»
« »
¬ .4 ¼
protein
fiber
fat
ª160 110 º
« 5
2 »»
ª 3º
, u=« ».
a. Let B = [SW Crp ] = «
« 6
.1»
¬2¼
«
»
.4 ¼
¬ 1
Then Bu lists the amounts of calories, protein, carbohydrate, and fat in a mixture of three servings
of Shredded Wheat and two servings of Crispix.
b. [M] Let u1 and u2 be the number of servings of Shredded Wheat and Crispix, respectively. Can
ª 120 º
« 3.2 »
ªu º
» ? To find out, row reduce the augmented
these numbers satisfy the equation B « 1 » = «
«
»
u
2.46
¬ 2¼
«
»
¬ .64 ¼
matrix
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.10
ª160 110
« 5
2
«
« 6
.1
«
.4
¬ 1
130 º ª 1
3.2 »» ««0
~
2.46 » «0
» «
.64 ¼ ¬0
.4
0
−2.3
46
.64 º ª 1
0 »» ««0
~
−1.38» «0
» «
27.6 ¼ ¬0
.4
46
−2.3
0
.64 º ª 1
27.6 »» ««0
~
−1.38» «0
» «
0 ¼ ¬0
• Solutions
.4 .64 º ª 1
1
.6 »» ««0
~
0
0 » «0
» «
0
0 ¼ ¬0
71
0 .4 º
1 .6 »»
0 0»
»
0 0¼
Since the system is consistent, it is possible for a mixture of the two creals to provide the desired
nutrients. The mixture is .4 servings of Shredded Wheat and .6 servings of Crispix.
3.
a. [M] Let x1, x2, and x3 be the number of servings ofAnnies’s Mac and Cheese, broccoli, and
chicken, respectively, needed for the lunch. The values of x1, x2, and x3 should satisfy
nutrients
ª
º
ª nutrients º
ª nutrients º ª quantities º
x1 «
per serving
» + x2 « per serving » + x3 « per serving » = « of nutrients »
¬«of Mac and Cheese ¼»
¬« of broccoli ¼»
¬« of chicken ¼» ¬« required ¼»
From the given data,
ª 270 º
ª 51º
ª 70 º ª 400º
«
»
«
»
x1 10 + x2 5.4 + x3 « 15» = « 30»
« »
« »
« » « »
«¬ 2 »¼
«¬5.2 »¼
«¬ 0»¼ «¬ 10»¼
To solve, row reduce the corresponding augmented matrix:
ª 270
« 10
«
«¬ 2
0 .99º
0 1.54»»
1 .79»¼
ª .99º ªservings of Mac and Cheese º
x ≈ ««1.54»» = ««
servings of broccoli »»
«¬ .74»¼ «¬
servings of chicken »¼
51
5.4
5.2
70
15
0
400º
ª1
»
30 » ~ ... ~ ««0
«¬0
10 »¼
0
1
0
b. [M] Changing from Annie’s Mac and Cheese to Annie’s Whole Wheat Shells and White Cheddar
changes the vector equation to
ª 260 º
ª 51º
ª 70 º ª 400º
«
»
«
»
x1
9 + x 5.4 + x « 15» = « 30 »
« » 2« » 3« » « »
«¬ 5»¼
«¬5.2 »¼
«¬ 0»¼ «¬ 10»¼
To solve, row reduce the corresponding augmented matrix:
ª 260 51 70 400º
ª1 0
« 9 5.4 15
»
30 » ~ ... ~ ««0 1
«
«¬ 5 5.2
«¬0 0
0
10 »¼
ª1.09 º ª servings of Shells º
x ≈ «« .88»» = ««servings of broccoli »»
¬«1.03¼» ¬« servings of chicken ¼»
0 1.09º
0
.88»»
1 1.03»¼
Notice that the number of servings of broccoli has decreased as was desired.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
72
CHAPTER 1
• Linear Equations in Linear Algebra
4. Here are the data, assembled from Table 1 and Exercise 4:
Mg of Nutrients/Unit
soy
soy
milk flour whey prot.
Nutrient
Nutrients
Required
(milligrams)
protein
carboh.
36
52
51
34
13
74
80
0
33
45
fat
calcium
0
1.26
7
.19
1.1
.8
3.4
.18
3
.8
a. Let x1, x2, x3, x4 represent the number of units of nonfat milk, soy flour, whey, and isolated soy
protein, respectively. These amounts must satisfy the following matrix equation
51
13
34
74
7
1.1
.19
.8
80 º ª x1 º ª 33º
»« » « »
0 » « x2 » « 45»
=
3.4 » « x3 » « 3 »
»« » « »
.18 »¼ «¬ x4 »¼ «¬ .8 »¼
ª 36
«
52
b. [M] «
« 0
«
¬«1.26
51
13
80
34
74
0
7
1.1
3.4
.19
.8
.18
ª 36
«
« 52
« 0
«
«¬1.26
ª1
33 º
»
«
45»
0
~ ⋅⋅⋅ ~ «
«0
3»
»
«
.8 ¼»
¬«0
0
0
0
1
0
0
0
1
0
0
0
1
.64 º
»
.54 »
−.09 »
»
−.21¼»
The “solution” is x1 = .64, x2 = .54, x3 = –.09, x4 = –.21. This solution is not feasible, because the
mixture cannot include negative amounts of whey and isolated soy protein. Although the
coefficients of these two ingredients are fairly small, they cannot be ignored. The mixture of .64
units of nonfat milk and .54 units of soy flour provide 50.6 g of protein, 51.6 g of carbohydrate,
3.8 g of fat, and .9 g of calcium. Some of these nutrients are nowhere close to the desired
amounts.
5. Loop 1: The resistance vector is
ª 11º
« »
−5
r1 = « »
« 0»
« »
¬ 0¼
Total of RI voltage drops for current I 1
Voltage drop for I 2 is negative; I 2 flows in opposite direction
Current I 3 does not flow in loop 1
Current I4 does not flow in loop 1
Loop 2: The resistance vector is
ª −5º
« »
10
r2 = « »
« −1»
« »
¬ 0¼
Voltage drop for I1 is negative; I1 flows in opposite direction
Total of RI voltage drops for current I 2
Voltage drop for I 3 is negative; I 3 flows in opposite direction
Current I 4 does not flow in loop 2
ª 0º
« −1»
Also, r3 = « » , r4 =
« 9»
« »
¬ −2 ¼
ª 0º
« 0»
« » , and R = [r1 r2 r3 r4] =
« −2 »
« »
¬ 10 ¼
ª 11
«
« −5
« 0
«
¬ 0
−5
10
−1
0
0
−1
9
−2
0º
»
0»
.
−2 »
»
10 ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.10
• Solutions
73
Notice that each off-diagonal entry of R is negative (or zero). This happens because the loop current
directions are all chosen in the same direction on the figure. (For each loop j, this choice forces the
currents in other loops adjacent to loop j to flow in the direction opposite to current Ij.)
ª 50º
«
»
−40»
. The voltages in loops 2 and 4 are negative because the battery orientation in
Next, set v = «
« 30»
«
»
¬ −30¼
each loop is opposite to the direction chosen for positive current flow. Thus, the equation Ri = v
becomes
ª 11
«
« −5
« 0
«
¬ 0
−5
10
−1
0
0
−1
9
−2
0 º ª I1 º ª 50 º
»« » «
»
0 » « I 2 » « −40 »
=
.
−2 » « I 3 » « 30 »
»« » «
»
10 ¼ ¬ I 4 ¼ ¬ −30 ¼
ª I1 º ª 3.68º
«I » «
»
−1.90 »
[M]: The solution is i = « 2 » = «
.
« I 3 » « 2.57 »
« » «
»
¬ I 4 ¼ ¬ −2.49 ¼
6. Loop 1: The resistance vector is
ª 6º
« »
−
r1 = « 1»
« 0»
« »
¬ 0¼
Total of RI voltage drops for current I1
Voltage drop for I 2 is negative; I 2 flows in opposite direction
Current I 3 does not flow in loop 1
Current I 4 does not flow in loop 1
Loop 2: The resistance vector is
ª −1º
« »
9
r2 = « »
« −4 »
« »
¬ 0¼
Voltage drop for I 1 is negative; I 1 flows in opposite direction
Total of RI voltage drops for current I 2
Voltage drop for I 3 is negative; I 3 flows in opposite direction
Current I 4 does not flow in loop 2
ª 0º
ª 0º
« −4 »
« 0»
Also, r3 = « » , r4 = « » , and R = [r1 r2 r3 r4]=
« 7»
« −2 »
« »
« »
¬ −2 ¼
¬ 7¼
v becomes
ª 6
« −1
«
« 0
«
¬ 0
−1
9
0
−4
−4
0
7
−2
ª 6
« −1
«
« 0
«
¬ 0
−1
9
0
−4
−4
0
7
−2
0º
0 »»
. Set v =
−2 »
»
7¼
0 º ª I1 º ª 30 º
ª I1 º ª 6.36 º
«
»
« I » « 8.14 »
»
«
»
0 » « I 2 » « 20 »
».
=
. [M]: The solution is i = « 2 » = «
« I 3 » «11.73»
−2 » « I 3 » « 40 »
« » «
»
»« » « »
7 ¼ ¬ I 4 ¼ ¬ 10 ¼
¬ I 4 ¼ ¬ 4.78¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
ª 30 º
« 20 »
« » . Then Ri =
« 40 »
« »
¬ 10 ¼
74
CHAPTER 1
• Linear Equations in Linear Algebra
7. Loop 1: The resistance vector is
ª 12 º
« »
−7
r1 = « »
« 0»
« »
¬ −4 ¼
Total of RI voltage drops for current I 1
Voltage drop for I 2 is negative; I 2 flows in opposite direction
Current I 3 does not flow in loop 1
Voltage drop for I 4 is negative; I 4 flows in opposite direction
Loop 2: The resistance vector is
ª −7 º
« »
15
r2 = « »
« −6 »
« »
¬ 0¼
Voltage drop for I1 is negative; I1 flows in opposite direction
Total of RI voltage drops for current I 2
Voltage drop for I 3 is negative; I 3 flows in opposite direction
Current I 4 does not flow in loop 2
ª 0º
« −6 »
Also, r3 = « » , r4 =
« 14 »
« »
¬« −5¼»
ª −4 º
« 0»
« » , and R = [r1 r2 r3 r4] =
« −5»
« »
¬« 13¼»
ª 12
« −7
«
« 0
«
¬« −4
−7
15
0
−6
−6
0
14
−5
−4 º
0 »»
.
−5»
»
13¼»
Notice that each off-diagonal entry of R is negative (or zero). This happens because the loop current
directions are all chosen in the same direction on the figure. (For each loop j, this choice forces the
currents in other loops adjacent to loop j to flow in the direction opposite to current Ij.)
ª 40 º
« 30 »
» . Note the negative voltage in loop 4. The current direction chosen in loop 4 is
Next, set v = «
« 20 »
«
»
¬« −10 ¼»
opposed by the orientation of the voltage source in that loop. Thus Ri = v becomes
ª 12
« −7
«
« 0
«
«¬ −4
−7
15
0
−6
−6
0
14
−5
−4 º ª I1 º ª 40 º
ª I1 º ª11.43º
«
»
« I » «10.55»
»
«
»
0 » « I 2 » « 30 »
».
=
. [M]: The solution is i = « 2 » = «
« I 3 » « 8.04 »
−5» « I 3 » « 20 »
« » «
»
»« » «
»
13»¼ «¬ I 4 »¼ «¬ −10 »¼
«¬ I 4 »¼ «¬ 5.84 »¼
8. Loop 1: The resistance vector is
ª 9º
« »
« −1»
r1 = « 0»
« »
« −1»
«¬ −4»¼
Total of RI voltage drops for current I1
Voltage drop for I 2 is negative; I 2 flows in opposite direction
Current I 3 does not flow in loop 1
Voltage drop for I 4 is negative; I 4 flows in opposite direction
Voltage drop for I 5 is negative; I 5 flows in opposite direction
Loop 2: The resistance vector is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.10
ª −1º
« »
« 7»
r2 = « −2»
« »
« 0»
«¬ −3»¼
• Solutions
75
Voltage drop for I1 is negative; I1 flows in opposite direction
Total of RI voltage drops for current I 2
Voltage drop for I 3 is negative; I 3 flows in opposite direction
Current I4 does not flow in loop 2
Voltage drop for I5 is negative; I5 flows in opposite direction
0 −1 −4 º
ª 0º
ª −4 º
ª 50º
ª −1º
ª 9 −1
« −2 »
« −3»
« −30»
« 0»
« −1
»
7 −2
0 −3»
« »
« »
«
»
« »
«
Also, r3 = « 10 » , r4 = « −3» , r5 = « −3» , and R = « 0 −2 10 −3 −3» . Set v = « 20» . Note
« »
«
»
« »
«
»
« »
0 −3
7 −2 »
« −3»
« −40»
« 7»
« −1
« −2 »
«¬ −3»¼
«¬ 0»¼
«¬ −2 »¼
«¬ −4 −3 −3 −2 12 »¼
«¬ 12 »¼
the negative voltages for loops where the chosen current direction is opposed by the orientation of
the voltage source in that loop. Thus Ri = v becomes:
0 −1 −4 º ª I1 º ª 50 º
ª I1 º ª 4.00 º
ª 9 −1
«
» «
«
»
»
« −1
»
«
»
7 −2
0 −3» « I 2 » « −30 »
« I 2 » « −4.38»
«
« 0 −2 10 −3 −3» « I 3 » = « 20 » . [M] The solution is « I 3 » = « −.90 » .
« » «
»
«
»« » «
»
0 −3
7 −2 » « I 4 » « −40 »
« I 4 » « −5.80 »
« −1
«¬ I 5 »¼ «¬ −.96 »¼
¬« −4 −3 −3 −2 12¼» ¬« I 5 ¼» ¬« 0 ¼»
9. The population movement problems in this section assume that the total population is constant, with
no migration or immigration. The statement that “about 7% of the city’s population moves to the
suburbs” means also that the rest of the city’s population (93%) remain in the city. This determines
the entries in the first column of the migration matrix (which concerns movement from the city).
From:
City Suburbs
To:
ª .93
«.07
¬
º
»
¼
City
Suburbs
Likewise, if 5% of the suburban population moves to the city, then the other 95% remain in the
ª .93 .05º
suburbs. This determines the second column of the migration matrix:, M = «
» . The
¬.07 .95¼
ª800, 000 º
difference equation is xk+1 = Mxk for k = 0, 1, 2, …. Also, x0 = «
»
¬500, 000 ¼
ª .93
The population in 2011 (when k = 1) is x1 = Mx0 = «
¬.07
.05º ª800,000 º ª 769, 000 º
=
.95»¼ «¬ 500,000 »¼ «¬ 531, 000 »¼
ª .93
The population in 2012 (when k = 2) is x2 = Mx1 = «
¬.07
.05º ª 769,000 º ª 741,720 º
=
.95»¼ «¬ 531,000 »¼ «¬558, 280 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
76
CHAPTER 1
• Linear Equations in Linear Algebra
10. The data in the first sentence implies that the migration matrix has the form:
From:
City Suburbs To:
ª
«.06
¬
.04 º
»
¼
City
Suburbs
The remaining entries are determined by the fact that the numbers in each column must sum to 1.
(For instance, if 6% of the city people move to the suburbs, then the rest, or 94%, remain in the city.)
ª10, 000,000 º
ª.94 .04 º
. The initial population is x0 = «
So the migration matrix is M = «
».
»
¬.06 .96 ¼
¬ 800,000 ¼
ª.94
The population in 2011 (when k = 1) is x1 = Mx0 = «
¬.06
.04 º ª10,000,000 º ª9, 432,000 º
=
.96 »¼ «¬ 800,000 »¼ «¬ 1,368,000 »¼
ª.94
The population in 2012 (when k = 2) is x2 = Mx1 = «
¬.06
.04 º ª 9, 432,000 º ª8,920,800 º
=
.96 »¼ «¬ 1,368,000 »¼ «¬1,879, 200 »¼
11. The problem concerns two groups of people–those living in California and those living outside
California (and in the United States). It is reasonable, but not essential, to consider the people living
inside California first. That is, the first entry in a column or row of a vector will concern the people
living in California. With this choice, the migration matrix has the form:
From:
Calif. Outside To:
ª
«
¬
º
»
¼
Calif.
Outside
a. For the first column of the migration matrix M, compute
persons
{Calif.
who moved }
{Total Calif. pop.}
=
516,100
= .016372
31,524,000
The other entry in the first column is 1 – .016372 = .983628. The exercise requests that 5 decimal
places be used. So this number should be rounded to .98363. Whatever number of decimal places
is used, it is important that the two entries sum to 1. So, for the first fraction, use .01637.
For the second column of M, compute
persons
{outside
who moved }
{Total outside pop.}
=
381,262
= .00167 . The other
228,680,000
entry is 1 – .00167 = .99833. Thus, the migration matrix is
From:
Calif.
Outside To:
ª .98363
«.01637
¬
.00167 º
.99833»¼
Calif.
Outside
b. [M] The initial vector is x0 = (31.524, 228.680), with data in millions of persons. Since x0
describes the population in 1994, and x1 describes the population in 1995, the vector x6 describes
the projected population for the year 2000, assuming that the migration rates remain constant and
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1.10
• Solutions
77
there are no deaths, births, or migration. Here are the vectors x0 through x6 with the first 5 figures
displayed. Numbers are in millions of persons:
ª 31.524 º ª 31.390 º ª 31.258º ª 31.129 º ª 31.002 º ª30.877 º ª 30.755º
x0 = «
»,«
»,«
»,«
»,«
»,«
»,«
» = x6.
¬ 228.68 ¼ ¬ 228.82 ¼ ¬ 228.95¼ ¬ 229.08¼ ¬ 229.20 ¼ ¬ 229.33¼ ¬ 229.45¼
ª.97
12. Set M = ««.00
«¬.03
ª.97
x2 = ««.00
«¬.03
.10º
ª 295º
ª.97 .05 .10 º ª 295º ª304º
»
«
»
.05» and x 0 = « 55» . Then x1 = ««.00 .90 .05»» «« 55»» ≈ «« 57 »» , and
«¬150»¼
«¬.03 .05 .85»¼ «¬150 »¼ «¬139»¼
.85»¼
.10 º ª304º ª312º
.05»» «« 57 »» ≈ «« 58»» . The entries in x2 give the approximate distribution of cars on
.85»¼ «¬139»¼ «¬130»¼
.05
.90
.05
.05
.90
.05
Wednesday, two days after Monday.
13. [M] The order of entries in a column of a migration matrix must match the order of the columns. For
instance, if the first column concerns the population in the city, then the first entry in each column
must be the fraction of the population that moves to (or remains in) the city. In this case, the data in
ª 600, 000 º
ª.95 .03º
and x0 = «
the exercise leads to M = «
»
»
¬.05 .97 ¼
¬ 400, 000 ¼
a. Some of the population vectors are
ª 523, 293º
ª 472,737 º
ª 439, 417 º
ª 417, 456 º
x5 = «
, x10 = «
, x15 = «
, x 20 = «
»
»
»
»
¬ 476,707 ¼
¬ 527, 263¼
¬ 560,583¼
¬ 582,544 ¼
The data here shows that the city population is declining and the suburban population is
increasing, but the changes in population each year seem to grow smaller.
ª 350,000 º
b. When x0 = «
» , the situation is different. Now
¬ 650,000 ¼
ª 358,523º
ª 364,140 º
ª 367,843º
ª 370, 283º
x5 = «
, x10 = «
, x15 = «
, x 20 = «
»
»
»
»
¬ 641, 477 ¼
¬ 635,860 ¼
¬ 632,157 ¼
¬ 629,717 ¼
The city population is increasing slowly and the suburban population is decreasing. No other
conclusions are expected. (This example will be analyzed in greater detail later in the text.)
14. Here are Figs. (a) and (b) for Exercise 13, followed by the figure for Exercise 34 in Section 1.1:
0˚
0˚
20˚
20˚
1
2
4
3
20˚
20˚
(a)
0˚
0˚
0˚
10˚
10˚
0˚
1
2
4
3
10˚
10˚
(b)
40˚
40˚
10˚
10˚
20˚
20˚
1
2
4
3
30˚
30˚
40˚
40˚
Section 1.1
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4T1 = 0 + 20 + T2 + T4
4T2 = T1 + 20 + 0 + T3
For Fig. (a), the equations are
4T3 = T4 + T2 + 0 + 20
4T4 = 0 + T1 + T3 + 20
To solve the system, rearrange the equations and row reduce the augmented matrix. Interchanging
rows 1 and 4 speeds up the calculations. The first five steps are shown in detail.
ª4
« −1
«
«0
«¬ −1
ª1
«0
~«
«0
«0
¬
−1
0
−1
4
−1
0
−1
4
−1
0
−1
4
20 º
ª1
»
« −1
20
»~«
20 » « 0
» «
20 ¼ ¬ 4
0
1
−4
1
0
−1
0
4
−2
0
−4
14
−20 º
ª1
»
«
0
0
»~«
20 » « 0
» «
100 ¼ ¬ 0
0
1
−4
4
−1
0
−1
4
−1
−1
0
−1
0
1
−4
1
0
−1
0
4
−2
0
0
12
−20 º
ª1
»
«0
20
»~«
20 » « 0
» «
20 ¼ ¬ 0
−20 º
ª1
»
«0
0
»~«
20 » « 0
» «
100 ¼ ¬ 0
0
1
−4
4
0
−4
−1
4
−1
−1
−4
15
0
0
0
1
0
0
10 »
0
1
0
10 »
0
0
1
10 ¼
−20 º
ª1
»
«0
0
» ~ ⋅⋅⋅ ~ «
20 »
«0
»
«
120 ¼
¬0
−20 º
0
1
−4
1
0
−1
0»
−1
4
−1
20 »
−1
−4
15
100 ¼
»
»
10 º
»
»
4T1 = 10 + 0 + T2 + T4
For Fig (b), the equations are
4T2 = T1 + 0 + 40 + T3
4T3 = T4 + T2 + 40 + 10
4T4 = 10 + T1 + T3 + 10
Rearrange the equations and row reduce the augmented matrix:
ª 4
« −1
«
« 0
«
¬ −1
−1
0
−1
4
−1
0
−1
4
−1
0
−1
4
10 º
ª1
»
«0
40
» ~ ⋅⋅⋅ ~ «
50 »
«0
»
«
20 ¼
¬0
0
0
0
1
0
0
0
1
0
0
0
1
º
17.5»
»
20 »
»
12.5¼
10
a. Here are the solution temperatures for the three problems studied:
Fig. (a) in Exercise 14 of Section 1.10:
(10, 10, 10, 10)
Fig. (b) in Exercise 14 of Section 1.10:
(10, 17.5, 20, 12.5)
Figure for Exercises 34 in Section 1.1
(20, 27.5, 30, 22.5)
When the solutions are arranged this way, it is evident that the third solution is the sum of the first
two solutions. What might not be so evident is that list of boundary temperatures of the third
problem is the sum of the lists of boundary temperatures of the first two problems. (The
temperatures are listed clockwise, starting at the left of T1.)
Fig. (a):
( 0, 20, 20, 0, 0, 20, 20, 0)
Fig. (b):
(10, 0, 0, 40, 40, 10, 10, 10)
Fig. from Section 1.1:
(10, 20, 20, 40, 40, 30, 30, 10)
b. When the boundary temperatures in Fig. (a) are multiplied by 3, the new interior temperatures are
also multiplied by 3.
c. The correspondence from the list of eight boundary temperatures to the list of four interior
temperatures is a linear transformation. A verification of this statement is not expected. However,
it can be shown that the solutions of the steady-state temperature problem here satisfy a
superposition principle. The system of equations that approximate the interior temperatures can
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 1 • Supplementary Exercises
79
be written in the form Ax = b, where A is determined by the arrangement of the four interior
points on the plate and b is a vector in R4 determined by the boundary temperatures.
Note: The MATLAB box in the Study Guide for Section 1.10 discusses scientific notation and shows
how to generate a matrix whose columns list the vectors x0, x1, x2, …, determined by an equation
xk+1 = Mxk for k = 0 , 1, ….
Chapter 1 SUPPLEMENTARY EXERCISES
1. a. False. (The word “reduced” is missing.) Counterexample:
ª1
A=«
¬3
2º
ª1
, B=«
»
4¼
¬0
2º
ª1
, C=«
»
−2 ¼
¬0
2º
1»¼
The matrix A is row equivalent to matrices B and C, both in echelon form.
b. False. Counterexample: Let A be any n×n matrix with fewer than n pivot columns. Then the
equation Ax = 0 has infinitely many solutions. (Theorem 2 in Section 1.2 says that a system has
either zero, one, or infinitely many solutions, but it does not say that a system with infinitely
many solutions exists. Some counterexample is needed.)
c. True. If a linear system has more than one solution, it is a consistent system and has a free
variable. By the Existence and Uniqueness Theorem in Section 1.2, the system has infinitely
many solutions.
d. False. Counterexample: The following system has no free variables and no solution:
x1 + x2 = 1
x2 = 5
x1 + x2 = 2
e. True. See the box after the definition of elementary row operations, in Section 1.1. If [A b] is
transformed into [C d] by elementary row operations, then the two augmented matrices are row
equivalent.
f. True. Theorem 6 in Section 1.5 essentially says that when Ax = b is consistent, the solution sets
of the nonhomogeneous equation and the homogeneous equation are translates of each other. In
this case, the two equations have the same number of solutions.
g. False. For the columns of A to span Rm, the equation Ax = b must be consistent for all b in Rm,
not for just one vector b in Rm.
h. False. Any matrix can be transformed by elementary row operations into reduced echelon form,
but not every matrix equation Ax = b is consistent.
i. True. If A is row equivalent to B, then A can be transformed by elementary row operations first
into B and then further transformed into the reduced echelon form U of B. Since the reduced
echelon form of A is unique, it must be U.
j. False. Every equation Ax = 0 has the trivial solution whether or not some variables are free.
k. True, by Theorem 4 in Section 1.4. If the equation Ax = b is consistent for every b in Rm, then A
must have a pivot position in every one of its m rows. If A has m pivot positions, then A has m
pivot columns, each containing one pivot position.
l. False. The word “unique” should be deleted. Let A be any matrix with m pivot columns but more
than m columns altogether. Then the equation Ax = b is consistent and has m basic variables and
at least one free variable. Thus the equation does not does not have a unique solution.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
m. True. If A has n pivot positions, it has a pivot in each of its n columns and in each of its n rows.
The reduced echelon form has a 1 in each pivot position, so the reduced echelon form is the n×n
identity matrix.
n. True. Both matrices A and B can be row reduced to the 3×3 identity matrix, as discussed in the
previous question. Since the row operations that transform B into I3 are reversible, A can be
transformed first into I3 and then into B.
o. True. The reason is essentially the same as that given for question f.
p. True. If the columns of A span Rm, then the reduced echelon form of A is a matrix U with a pivot
in each row, by Theorem 4 in Section 1.4. Since B is row equivalent to A, B can be transformed
by row operations first into A and then further transformed into U. Since U has a pivot in each
row, so does B. By Theorem 4, the columns of B span Rm.
q. False. See Example 5 in Section 1.7.
r. True. Any set of three vectors in R2 would have to be linearly dependent, by Theorem 8 in
Section 1.7.
s. False. If a set {v1, v2, v3, v4} were to span R5, then the matrix A = [v1 v2 v3 v4] would have
a pivot position in each of its five rows, which is impossible since A has only four columns.
t. True. The vector –u is a linear combination of u and v, namely, –u = (–1)u + 0v.
u. False. If u and v are multiples, then Span{u, v} is a line, and w need not be on that line.
v. False. Let u and v be any linearly independent pair of vectors and let w = 2v. Then w = 0u + 2v,
so w is a linear combination of u and v. However, u cannot be a linear combination of v and w
because if it were, u would be a multiple of v. That is not possible since {u, v} is linearly
independent.
w. False. The statement would be true if the condition v1 is not zero were present. See Theorem 7 in
Section 1.7. However, if v1 = 0, then {v1, v2, v3} is linearly dependent, no matter what else might
be true about v2 and v3.
x. True. “Function” is another word used for “transformation” (as mentioned in the definition of
“transformation” in Section 1.8), and a linear transformation is a special type of transformation.
y. True. For the transformation x 6 Ax to map R5 onto R6, the matrix A would have to have a pivot
in every row and hence have six pivot columns. This is impossible because A has only five
columns.
z. False. For the transformation x 6 Ax to be one-to-one, A must have a pivot in each column.
Since A has n columns and m pivots, m might be less than n.
2. If a ≠ 0, then x = b/a; the solution is unique. If a = 0, and b ≠ 0, the solution set is empty, because
0x = 0 ≠ b. If a = 0 and b = 0, the equation 0x = 0 has infinitely many solutions.
3. a. Any consistent linear system whose echelon form is
ª * *
«0 *
«
«¬ 0 0 0
*º
ª * * *º
ª 0 * *º
»
«
»
*» or « 0 0 *» or «« 0 0 *»»
«¬ 0 0 0 0»¼
«¬ 0 0 0 0»¼
0»¼
b. Any consistent linear system whose coefficient matrix has reduced echelon form I3.
c. Any inconsistent linear system of three equations in three variables.
4. Since there are three pivots (one in each row), the augmented matrix must reduce to the form
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 1 • Supplementary Exercises
81
ª * * *º
« 0 * *» . A solution of Ax = b exists for all b because there is a pivot in each row of A. Each
«
»
«¬ 0 0 *»¼
solution is unique because there are no free variables.
3
k º
ª 1 3 k º ª1
5. a. «
~«
»
» . If h = 12 and k v 2, the second row of the augmented matrix
¬ 4 h 8 ¼ ¬ 0 h − 12 8 − 4 k ¼
indicates an inconsistent system of the form 0x2 = b, with b nonzero. If h = 12, and k = 2, there is
only one nonzero equation, and the system has infinitely many solutions. Finally, if h v 12, the
coefficient matrix has two pivots and the system has a unique solution.
1º
h
ª −2 h 1 º ª −2
b. «
~«
»
» . If k + 3h = 0, the system is inconsistent. Otherwise, the
¬ 6 k −2 ¼ ¬ 0 k + 3h 1¼
coefficient matrix has two pivots and the system has a unique solution.
ª4º
ª −2 º
ª 7º
ª −5 º
6. a. Set v1 = « » , v 2 = « » , v 3 = « » , and b = « » . “Determine if b is a linear combination of v1,
¬ 8¼
¬ −3 ¼
¬10 ¼
¬ −3 ¼
v2, v3.” Or, “Determine if b is in Span{v1, v2, v3}.” To do this, compute
7 −5 º ª 4 −2
7 −5 º
ª 4 −2
. The system is consistent, so b is in Span{v1, v2, v3}.
« 8 −3 10 −3» ~ « 0
1 −4
7 »¼
¬
¼ ¬
7º
ª 4 −2
ª −5 º
b. Set A = «
, b = « » . “Determine if b is a linear combination of the columns of A.”
»
¬ 8 −3 10 ¼
¬ −3 ¼
c. Define T(x) = Ax. “Determine if b is in the range of T.”
ª 2º
ª −4º
ª −2º
ª b1 º
«
»
«
»
«
»
7. a. Set v1 = « −5» , v 2 = « 1» , v3 = « 1» and b = ««b2 »» . “Determine if v1, v2, v3 span R3.” To do this,
«¬ 7 »¼
«¬ −5»¼
«¬ −3»¼
«¬ b3 »¼
row reduce [v1 v2 v3]:
ª 2 −4 −2º ª 2 −4 −2º ª 2 −4 −2 º
« −5
1
1»» ~ «« 0 −9 −4»» ~ «« 0 −9 −4 »» . The matrix does not have a pivot in each row,
«
«¬ 7 −5 −3»¼ «¬ 0
9
4»¼ «¬ 0
0
0 »¼
3
so its columns do not span R , by Theorem 4 in Section 1.4.
ª 2
b. Set A = «« −5
«¬ 7
−4
1
−5
−2 º
1»» . “Determine if the columns of A span R3.”
−3»¼
c. Define T(x) = Ax. “Determine if T maps R3 onto R3.”
ª * *º ª * * º ª 0 * º
8. a. «
», «
», «
»
¬ 0 *¼ ¬ 0 0 ¼ ¬ 0 0 ¼
ª * * º
b. «« 0 * »»
«¬ 0 0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
ª1 º
ª2º
9. The first line is the line spanned by « » . The second line is spanned by « » . So the problem is to
¬2¼
¬1 ¼
ª5º
ª1 º
ª2º
write « » as the sum of a multiple of « » and a multiple of « » . That is, find x1 and x2 such that
¬6¼
¬2¼
¬1 ¼
ª 2º
ª 1º ª 5º
x1 « » + x2 « » = « » . Reduce the augmented matrix for this equation:
¬ 1¼
¬ 2¼ ¬6¼
ª2
«1
¬
1
2
5º ª 1
~
6 »¼ «¬ 2
2
1
6º ª 1
~
5»¼ «¬ 0
2
−3
6º ª 1
~
−7 »¼ «¬ 0
6º ª 1
~
7 / 3»¼ «¬ 0
2
1
0
1
4 / 3º
7 / 3»¼
ª 5º
ª 2º
ª 1º
ª 5 º ª 8 / 3º ª 7 / 3º
Thus, « » = 4 « » + 7 « » or « » = «
»+«
».
¬6 ¼ 3 ¬ 1¼ 3 ¬ 2 ¼
¬6 ¼ ¬ 4 / 3¼ ¬14 / 3¼
10. The line through a1 and the origin and the line through a2 and the origin determine a “grid” on the
x1x2-plane as shown below. Every point in R2 can be described uniquely in terms of this grid. Thus, b
can be reached from the origin by traveling a certain number of units in the a1-direction and a certain
number of units in the a2-direction.
x2
b
a1
x1
a2
11. A solution set is a line when the system has one free variable. If the coefficient matrix is 2×3, then
ª 1 2 *º
two of the columns should be pivot columns. For instance, take «
» . Put anything in column
¬ 0 3 *¼
3. The resulting matrix will be in echelon form. Make one row replacement operation on the second
ª 1 2 1º ª1 2 1º
row to create a matrix not in echelon form, such as «
»~«
»
¬ 0 3 1¼ ¬1 5 2 ¼
12. A solution set is a plane where there are two free variables. If the coefficient matrix is 2×3, then only
one column can be a pivot column. The echelon form will have all zeros in the second row. Use a
ª1 2 3º
row replacement to create a matrix not in echelon form. For instance, let A = «
».
¬1 2 3¼
ª1
13. The reduced echelon form of A looks like E = ««0
¬«0
0
1
0
*º
*»» . Since E is row equivalent to A, the
0 »¼
ª1
equation Ex = 0 has the same solutions as Ax = 0. Thus ««0
«¬0
0
1
0
*º ª 3º ª0º
*»» «« −2»» = ««0»» .
0 »¼ «¬ 1»¼ «¬0»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 1 • Supplementary Exercises
ª1
By inspection, E = ««0
«¬0
0
1
0
83
−3º
2»» .
0»¼
ª 1º
ª a º ª0 º
14. Row reduce the augmented matrix for x1 « » + x2 «
» = « » (*).
¬a ¼
¬ a + 2 ¼ ¬0 ¼
ª1
«a
¬
a
a+2
0 º ª1
~«
0 »¼ ¬0
a
a+2−a
2
0 º ª1
»=«
0¼ ¬0
a
(2 − a)(1 + a)
0º
0»¼
The equation (*) has a nontrivial solution only when (2 – a)(1 + a) = 0. So the vectors are linearly
independent for all a except a = 2 and a = –1.
15. a. If the three vectors are linearly independent, then a, c, and f must all be nonzero. (The converse is
true, too.) Let A be the matrix whose columns are the three linearly independent vectors. Then A
must have three pivot columns. (See Exercise 30 in Section 1.7, or realize that the equation
Ax = 0 has only the trivial solution and so there can be no free variables in the system of
equations.) Since A is 3×3, the pivot positions are exactly where a, c, and f are located.
b. The numbers a, …, f can have any values. Here's why. Denote the columns by v1, v2, and v3.
Observe that v1 is not the zero vector. Next, v2 is not a multiple of v1 because the third entry of v2
is nonzero. Finally, v3 is not a linear combination of v1 and v2 because the fourth entry of v3 is
nonzero. By Theorem 7 in Section 1.7, {v1, v2, v3} is linearly independent.
16. Denote the columns from right to left by v1, …, v4. The “first” vector v1 is nonzero, v2 is not a
multiple of v1 (because the third entry of v2 is nonzero), and v3 is not a linear combination of v1 and
v2 (because the second entry of v3 is nonzero). Finally, by looking at first entries in the vectors, v4
cannot be a linear combination of v1, v2, and v3. By Theorem 7 in Section 1.7, the columns are
linearly independent.
17. Here are two arguments. The first is a “direct” proof. The second is called a “proof by contradiction.”
i. Since {v1, v2, v3} is a linearly independent set, v1 ≠ 0. Also, Theorem 7 shows that v2 cannot be a
multiple of v1, and v3 cannot be a linear combination of v1 and v2. By hypothesis, v4 is not a linear
combination of v1, v2, and v3. Thus, by Theorem 7, {v1, v2, v3, v4} cannot be a linearly dependent
set and so must be linearly independent.
ii. Suppose that {v1, v2, v3, v4} is linearly dependent. Then by Theorem 7, one of the vectors in the
set is a linear combination of the preceding vectors. This vector cannot be v4 because v4 is not in
Span{v1, v2, v3}. Also, none of the vectors in {v1, v2, v3} is a linear combinations of the preceding
vectors, by Theorem 7. So the linear dependence of {v1, v2, v3, v4} is impossible. Thus {v1, v2, v3,
v4} is linearly independent.
18. Suppose that c1 and c2 are constants such that
c1v1 + c2(v1 + v2) = 0 (*)
Then (c1 + c2)v1 + c2v2 = 0. Since v1 and v2 are linearly independent, both c1 + c2 = 0 and c2 = 0. It
follows that both c1 and c2 in (*) must be zero, which shows that {v1, v1 + v2} is linearly independent.
19. Let M be the line through the origin that is parallel to the line through v1, v2, and v3. Then v2 – v1 and
v3 – v1 are both on M. So one of these two vectors is a multiple of the other, say v2 – v1 = k(v3 – v1).
This equation produces a linear dependence relation (k – 1)v1 + v2 – kv3 = 0.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
A second solution: A parametric equation of the line is x = v1 + t(v2 – v1). Since v3 is on the line,
there is some t0 such that v3 = v1 + t0(v2 – v1) = (1 – t0)v1 + t0v2. So v3 is a linear combination of v1
and v2, and {v1, v2, v3} is linearly dependent.
20. If T(u) = v, then since T is linear,
T(–u) = T((–1)u) = (–1)T(u) = –v.
21. Either compute T(e1), T(e2), and T(e3) to make the columns of A, or write the vectors vertically in the
definition of T and fill in the entries of A by inspection:
ª?
Ax = ««?
«¬?
?
A
?
?º ª x1 º ª x1 º
?»» «« x2 »» = «« − x2 »» ,
?»¼ «¬ x3 »¼ «¬ x3 »¼
ª1
A = «« 0
«¬ 0
0º
0 »»
1»¼
0
−1
0
22. By Theorem 12 in Section 1.9, the columns of A span R3. By Theorem 4 in Section 1.4, A has a pivot
in each of its three rows. Since A has three columns, each column must be a pivot column. So the
equation Ax = 0 has no free variables, and the columns of A are linearly independent. By Theorem 12
in Section 1.9, the transformation x 6 Ax is one-to-one.
− b º ª 4 º ª 5º
4a
= « » implies that
»
«
»
a ¼ ¬ 3¼ ¬ 0 ¼
3a
ªa
23. «
¬b
−3
ª4
«
¬3
5 º ª4
~
0 ¼» «¬ 0
4
−3
−
+
º ª4
~
−15 / 4 »¼ «¬ 0
5
25 / 4
=
=
3b
4b
−3
5
. Solve:
0
5 º ª4
~
−3/ 5»¼ «¬ 0
1
0
1
16 / 5º ª 1
~
−3/ 5»¼ «¬ 0
0
1
4 / 5º
−3/ 5»¼
Thus a = 4/5 and b = –3/5.
24. The matrix equation displayed gives the information 2 a − 4b = 2 5 and 4a + 2b = 0. Solve for a and
ª2
b: «
¬4
−4
2
2 5 º ª2
»~«
0 ¼ «¬ 0
−4
10
2 5º ª1
»~«
−4 5 »¼ «¬0
−2
1
5º ª1
»~«
−2 / 5 »¼ «¬0
0
1
1/ 5 º
»
−2 / 5 »¼
So a = 1/ 5, b = −2 / 5.
25. a. The vector lists the number of three-, two-, and one-bedroom apartments provided when x1 floors
of plan A are constructed.
ª 3º
ª 4º
ª5º
«
»
«
»
b. x1 «7 » + x2 « 4» + x3 «« 3»»
«¬ 8»¼
«¬ 8»¼
«¬9»¼
ª 3º
ª 4º
ª5º ª 66º
«
»
«
»
c. [M] Solve x1 «7 » + x2 « 4» + x3 «« 3»» = «« 74»»
«¬ 8»¼
«¬ 8»¼
«¬9»¼ «¬136»¼
ª3
«7
«
«¬ 8
4
4
8
5
3
9
66º
ª1
»
74» ~ ⋅⋅⋅ ««0
«¬0
136»¼
0
1
0
−1/ 2
13/ 8
0
2º
15»»
0»¼
x1
x2
−
+
(1/ 2) x3
(13/ 8) x3
0
=
=
=
The general solution is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2
15
0
Chapter 1 • Supplementary Exercises
85
ª x1 º ª 2 + (1/ 2) x3 º ª 2º
ª 1/ 2 º
«
»
«
»
«
»
x = « x2 » = «15 − (13/ 8) x3 » = «15» + x3 «« −13/8»»
«¬ x3 »¼ «¬
»¼ «¬ 0»¼
«¬ 1 »¼
x3
However, the only feasible solutions must have whole numbers of floors for each plan. Thus, x3
must be a multiple of 8, to avoid fractions. One solution, for x3 = 0, is to use 2 floors of plan A
and 15 floors of plan B. Another solution, for x3 = 8, is to use 6 floors of plan A , 2 floors of plan
B, and 8 floors of plan C. These are the only feasible solutions. A larger positive multiple of 8 for
x3 makes x2 negative. A negative value for x3, of course, is not feasible either.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.1
SOLUTIONS
Notes: The definition here of a matrixx product AB gives the proper view of AB for nearlly all matrix
calculations. (The dual fact about the roows of A and the rows of AB is seldom needed, maiinly because
vectors here are usually written as coolumns.) I assign Exercise 13 and most of Exercisees 17–22 to
reinforce the definition of AB.
Exercises 23 and 24 are used in the proof of the Invertible Matrix Theorem, in Section 2..3. Exercises
23–25 are mentioned in a footnote in Seection 2.2. A class discussion of the solutions of Exerrcises 23–25
can provide a transition to Section 2.2. O
Or, these exercises could be assigned after starting Secction 2.2.
Exercises 27 and 28 are optional, buut they are mentioned in Example 4 of Section 2.4. Ou
uter products
also appear in Exercises 31–34 of Sectioon 4.6 and in the spectral decomposition of a symmetrric matrix, in
Section 7.1. Exercises 29–33 provide good training for mathematics majors.
ª2
1. −2 A = ( −2) «
¬4
0
−5
ª7
B − 2A = «
¬1
−1º ª −4
=
2 »¼ «¬ −8
−5
−4
0
110
1º ª −4
+
−3»¼ «¬ −8
2º
. Next, use B – 2A = B + (–2A):
−4 »¼
0
10
2º ª 3
=
−4 »¼ «¬ −7
−5
6
3º
−7 »¼
The product AC is not defined becauuse the number of columns of A does not match the nu
umber of
1
2
3
5
1
⋅
3
+
2(
−
1)
1
⋅
5
+
2
⋅
4
1
13
ª
ºª
º ª
º ª
º
rows of C. CD = «
=«
=«
»
«
»
»
» . For mental
¬ −2 1¼ ¬ −1 4 ¼ ¬ −2 ⋅ 3 + 1( −1) −2 ⋅ 5 + 1 ⋅ 4 ¼ ¬ −7 −6 ¼
computation, the row-column rule is probably easier to use than the definition.
0 −1º
1º ª 2 + 21
0 − 15 −1 + 3º ª 23 −15
2º
ª2
ª 7 −5
2. A + 3B = «
+ 3«
=«
=«
»
»
»
2 − 9 ¼ ¬ 7 −17 −7 »¼
2¼
¬ 4 −5
¬ 1 −4 −3¼ ¬ 4 + 3 −5 − 12
The expression 2C – 3E is not definned because 2C has 2 columns and –3E has only 1 colu
umn.
1º ª 3 ⋅ 7 + 5 ⋅ 1
3( −5) + 5( −4)
3 ⋅ 1 + 5( −3) º ª 26 −35 −12 º
ª 3 5 º ª 7 −5
=«
DB = «
»
«
»
»=«
»
¬ −1 4 ¼ ¬ 1 −4 −3¼ ¬ −1 ⋅ 7 + 4 ⋅ 1 −1( −5) + 4( −4) −1 ⋅ 1 + 4( −3) ¼ ¬ −3 −11 −13¼
The product EC is not defined becauuse the number of columns of E does not match the nu
umber of
rows of C.
Copyright © 2012 Pearrson Education, Inc. Publishing as Addison-Wesley.
87
88
CHAPTER 2
• Matrix Algebra
ª3
3. 3I 2 − A = «
¬0
0º ª2
−
3»¼ «¬ 3
−5 º ª 3 − 2
=
−2 »¼ «¬ 0 − 3
−5 º ª 6
=
−2 »¼ «¬ 9
ª2
(3 I 2 ) A = 3( I 2 A) = 3 «
¬3
ª3
(3I 2 ) A = «
¬0
ª 5
4. A − 5 I 3 = « −4
«
«¬ −3
0º ª 2
3»¼ «¬ 3
−1
1
5
0
−1
ª 5
(5 I 3 ) A = 5( I 3 A) = 5 A = 5 «« −4
«¬ −3
ª5
(5 I 3 ) A = «« 0
«¬ 0
0
5
0
ª 5⋅5 + 0 + 0
= «« 0 + 5( −4) + 0
«¬ 0 + 0 + 5( −3)
ª −1
5. a. Ab 1 = « 2
«
«¬ 5
AB = [ Ab1
ª −1
«
b. 2
«
¬« 5
AB = [ Ab1
1
1
0 + 5⋅3 + 0
0 + 0 + 5 ⋅1
1
3º
−6 »»
−3»¼
−5
15
5
15 º
−30 »» , or
10 »¼
5 ⋅ 3 + 0 + 0 º ª 25
0 + 5( −6) + 0 »» = «« −20
0 + 0 + 5 ⋅ 2 »¼ «¬ −15
ª −1
Ab 2 = «« 2
«¬ 5
−5
15
5
15 º
−30 »»
10 »¼
3º
ª 11º
ª −2 º «
»
4 » « » = « 8»»
¬ 3¼ « 19 »
−3»¼
¬− ¼
11º
8 »»
−19 »¼
ª −1 ⋅ 4 + 3( −2)
−2 º «
= 2 ⋅ 4 + 4( −2)
3»¼ «
¬« 5 ⋅ 4 − 3( −2)
−3 º
ª −5 º
ª 1º « »
»
5 » « » = « 12 » ,
3
1»¼ ¬ ¼ «¬ 3»¼
ª −5
Ab 2 ] = «« 12
«¬ 3
−2
3º ª 25
−6 »» = «« −20
2 »¼ «¬ −15
5( −1) + 0 + 0
ª −10
Ab 2 ] = «« 0
«¬ 26
−1
−15 º
−6 »¼
3º
−6 »»
2 »¼
3
3º
ª −10 º
ª 4º «
»
4 » « » = « 0 »» ,
¬ −2 ¼ « 26 »
−3»¼
¬
¼
3º
ª 4
4 »» «
¬ −2
−3¼»
ª 4
6. a. Ab1 = « −3
«
«¬ 0
3
−1
0º ª 5
0 »» «« −4
5 »¼ «¬ −3
3( −5) + 0 º ª 6
=
0 + 3( −2) »¼ «¬ 9
0º ª 0
0 »» = «« −4
5 »¼ «¬ −3
0
5º
5»¼
−15 º
, or
−6 »¼
−5 º ª 3 ⋅ 2 + 0
=
−2 »¼ «¬ 0 + 3 ⋅ 3
3º ª 5
−6 »» − «« 0
2 »¼ «¬ 0
3
0 − ( −5) º ª 1
=
3 − ( −2) »¼ «¬ −3
−1( −2) + 3 ⋅ 3º ª −10
2( −2) + 4 ⋅ 3»» = «« 0
5( −2) − 3 ⋅ 3¼» ¬« 26
ª 4
Ab 2 = «« −3
«¬ 0
11º
8»»
−19 ¼»
−3º
ª 22 º
ª 4º «
»
5» « » = « −22 »»
−2
1»¼ ¬ ¼ «¬ −2 »¼
22 º
−22 »»
−2 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.1
ª 4
«
b. −3
«
¬« 0
−3 º
ª1
5 »» «
3
1¼» ¬
ª 4 ⋅1 − 3 ⋅ 3
4º «
= −3 ⋅ 1 + 5 ⋅ 3
−2 ¼» «
¬« 0 ⋅ 1 + 1 ⋅ 3
4 ⋅ 4 − 3( −2) º ª −5
−3 ⋅ 4 + 5( −2) »» = «« 12
0 ⋅ 4 + 1( −2) ¼» ¬« 3
• Solutions
89
22 º
−22 »»
−2 ¼»
7. Since A has 3 columns, B must match with 3 rows. Otherwise, AB is undefined. Since AB has 7
columns, so does B. Thus, B is 3×7.
8. The number of rows of B matches the number of rows of BC, so B has 5 rows.
12 º
ª 1 9 º ª 2 3 º ª −7
ª 2 3º ª 1 9 º ª −7 18 + 3k º
9. AB = «
=«
=«
, while BA = «
»
«
»
».
»
«
»
»
¬ −3 k ¼ ¬ −1 1¼ ¬ −6 − k −9 + k ¼
¬ −1 1¼ ¬ −3 k ¼ ¬ −4 −9 + k ¼
Then AB = BA if and only if 18 + 3k = 12 and –4 = –6 – k, which happens if and only if k = –2.
ª 3
10. AB = «
¬ −1
−6 º ª − 1
2 »¼ «¬ 3
ª1
11. AD = « 2
«
¬« 3
2
ª5
DA = «« 0
¬« 0
4
5
0
3
0
3º ª 5
5»» «« 0
6 ¼» ¬« 0
0º ª 1
0 »» «« 2
2 ¼» ¬« 3
1º ª −21
=
4 »¼ «¬ 7
0
3
0
2
4
5
−21º
ª 3
, AC = «
»
7¼
¬ −1
0º ª 5
0 »» = ««10
2 ¼» ¬«15
3º ª 5
5 »» = «« 6
6 ¼» ¬« 6
6
12
15
10
12
10
−6 º ª −3
2 »¼ «¬ 2
−5º ª −21
=
1»¼ «¬ 7
−21º
7 »¼
6º
10 »»
12 ¼»
15º
15 »»
12 ¼»
Right-multiplication (that is, multiplication on the right) by the diagonal matrix D multiplies each
column of A by the corresponding diagonal entry of D. Left-multiplication by D multiplies each row
of A by the corresponding diagonal entry of D. To make AB = BA, one can take B to be a multiple of
I3. For instance, if B = 4I3, then AB and BA are both the same as 4A.
12. Consider B = [b1 b2]. To make AB = 0, one needs Ab1 = 0 and Ab2 = 0. By inspection of A, a
suitable
ª2 6º
ª2º
ª 2º
b1 is « » , or any multiple of « » . Example: B = «
».
¬ 1 3¼
¬ 1¼
¬ 1¼
13. Use the definition of AB written in reverse order: [Ab1 ⋅ ⋅ ⋅ Abp] = A[b1 ⋅ ⋅ ⋅ bp]. Thus
[Qr1 ⋅ ⋅ ⋅ Qrp] = QR, when R = [r1 ⋅ ⋅ ⋅ rp].
14. By definition, UQ = U[q1 ⋅ ⋅ ⋅ q4] = [Uq1 ⋅ ⋅ ⋅ Uq4]. From Example 6 of Section 1.8, the vectorUq1
lists the total costs (material, labor, and overhead) corresponding to the amounts of products B andC
specified in the vector q1. That is, the first column of UQ lists the total costs for materials, labor, and
overhead used to manufacture products B and C during the first quarter of the year. Columns 2, 3,and
4 of UQ list the total amounts spent to manufacture B and C during the 2nd, 3rd, and 4th quarters,
respectively.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
90
CHAPTER 2
• Matrix Algebra
15. a. False. See the definition of AB.
b. False. The roles of A and B should be reversed in the second half of the statement. See the box
after Example 3.
c. True. See Theorem 2(b), read right to left.
d. True. See Theorem 3(b), read right to left.
e. False. The phrase “in the same order” should be “in the reverse order.” See the box after Theorem
3.
16. a. True. See the box after Example 4.
b. False. AB must be a 3×3 matrix, but the formula given here implies that it is a 3×1 matrix. The
plus signs should just be spaces (between columns). This is a common mistake.
c. True. Apply Theorem 3(d) to A2=AA
d. False. The left-to-right order of (ABC)T, is CTBTAT. The order cannot be changed in general.
e. True. This general statement follows from Theorem 3(b).
ª −3
17. Since «
¬ 1
ª −1º
Ab 2 ] , the first column of B satisfies the equation Ax = « » . Row
¬ 6¼
−3 −3 º ª 1 0 3 º
ª 3º
. So b1 = « » . Similarly,
~«
»
»
5
1¼ ¬ 0
1 2¼
¬2¼
ª 1
Ab1 ] ~ «
¬ −3
ª 1 −3 −11º ª 1
Ab 2 ] ~ «
~
5
17 »¼ «¬ 0
¬ −3
reduction: [ A
[A
−11º
= AB = [ Ab1
17 »¼
0
1
1º
and b2 =
4 »¼
ª 1º
«4» .
¬ ¼
Note: An alternative solution of Exercise 17 is to row reduce [A Ab1 Ab2] with one sequence of row
operations. This observation can prepare the way for the inversion algorithm in Section 2.2.
18. The third column of AB is also all zeros because Ab3 = A0 = 0
19. (A solution is in the text). Write B = [b1 b2 b3]. By definition, the third column of AB is Ab3. By
hypothesis, b3 = b1 + b2. So Ab3 = A(b1 + b2) = Ab1 + Ab2, by a property of matrix-vector
multiplication. Thus, the third column of AB is the sum of the first two columns of AB.
20. The first two columns of AB are Ab1 and Ab2. They are equal since b1 and b2 are equal.
21. Let bp be the last column of B. By hypothesis, the last column of AB is zero. Thus, Abp = 0.
However, bp is not the zero vector, because B has no column of zeros. Thus, the equation Abp = 0 is a
linear dependence relation among the columns of A, and so the columns of A are linearly dependent.
Note: The text answer for Exercise 21 is, “The columns of A are linearly dependent. Why?” The Study
Guide supplies the argument above, in case a student needs help.
22. If the columns of B are linearly dependent, then there exists a nonzero vector x such that Bx = 0.
From this, A(Bx) = A0 and (AB)x = 0 (by associativity). Since x is nonzero, the columns of AB must
be linearly dependent.
23. If x satisfies Ax = 0, then CAx = C0 = 0 and so Inx = 0 and x = 0. This shows that the equation Ax = 0
has no free variables. So every variable is a basic variable and every column of A is a pivot column.
(A variation of this argument could be made using linear independence and Exercise 30 in Section
1.7.) Since each pivot is in a different row, A must have at least as many rows as columns.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.1
• Solutions
91
24. Write I3 =[e1 e2 e3] and D = [d1 d2 d3]. By definition of AD, the equation AD = I3 is equivalent
|to the three equations Ad1 = e1, Ad2 = e2, and Ad3 = e3. Each of these equations has at least one
solution because the columns of A span R3. (See Theorem 4 in Section 1.4.) Select one solution of
each equation and use them for the columns of D. Then AD = I3.
25. By Exercise 23, the equation CA = In implies that (number of rows in A) > (number of columns), that
is, m > n. By Exercise 24, the equation AD = Im implies that (number of rows in A) < (number of
columns), that is, m < n. Thus m = n. To prove the second statement, observe that CAD = (CA)D =
InD = D, and also CAD = C(AD) = CIm = C. Thus C = D. A shorter calculation is
C =C In = C(AD) = (CA)D = In D= D
26. Take any b in Rm. By hypothesis, ADb = Imb = b. Rewrite this equation as A(Db) = b. Thus, the
vector x = Db satisfies Ax = b. This proves that the equation Ax = b has a solution for each b in Rm.
By Theorem 4 in Section 1.4, A has a pivot position in each row. Since each pivot is in a different
column, A must have at least as many columns as rows.
27. The product uTv is a 1×1 matrix, which usually is identified with a real number and is written
without the matrix brackets.
u v = [ −3
T
2
ª −3 º
uv = «« 2 »» [ a
«¬ −5 »¼
T
ªaº
«
vu = « b »» [ −3
«¬ c »¼
T
ªa º
−5] «« b »» = −3a + 2b − 5c , vT u = [ a
«¬ c »¼
ª − 3a
c ] = «« 2 a
«¬ −5a
b
ª −3a
−5] = «« −3b
«¬ −3c
2
−3b
2b
−5b
2a
2b
2c
b
ª −3 º
c ] «« 2 »» = −3a + 2b − 5c
«¬ −5 »¼
−3c º
2c »»
−5c »¼
−5 a º
−5b »»
−5c »¼
28. Since the inner product uTv is a real number, it equals its transpose. That is,
uTv = (uTv)T = vT (uT)T = vTu, by Theorem 3(d) regarding the transpose of a product of matrices and
by Theorem 3(a). The outer product uvT is an n×n matrix. By Theorem 3, (uvT)T = (vT)TuT = vuT.
29. The (i, j)-entry of A(B + C) equals the (i, j)-entry of AB + AC, because
n
n
n
k =1
k =1
k =1
¦ aik (bkj + ckj ) = ¦ aik bkj + ¦ aik ckj
The (i, j)-entry of (B + C)A equals the (i, j)-entry of BA + CA, because
n
n
n
k =1
k =1
k =1
¦ (bik + cik )akj = ¦ bik akj + ¦ cik akj
30. The (i, j))-entries of r(AB), (rA)B, and A(rB) are all equal, because
n
n
n
k =1
k =1
k =1
r ¦ aik bkj = ¦ (raik )bkj = ¦ aik (rbkj )
31. Use the definition of the product ImA and the fact that Imx = x for x in Rm.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
92
CHAPTER 2
• Matrix Algebra
ImA = Im[a1 ⋅ ⋅ ⋅ an] = [Ima1 ⋅ ⋅ ⋅ Iman] = [a1 ⋅ ⋅ ⋅ an] = A
32. Let ej and aj denote the jth columns of In and A, respectively. By definition, the jth column of AIn is
Aej, which is simply aj because ej has 1 in the jth position and zeros elsewhere. Thus corresponding
columns of AIn and A are equal. Hence AIn = A.
33. The (i, j)-entry of (AB)T is the ( j, i)-entry of AB, which is
a j1b1i + ⋅⋅⋅ + a jnbni
The entries in row i of BT are b1i, … , bni, because they come from column i of B. Likewise, the
entries in column j of AT are aj1, …, ajn, because they come from row j of A. Thus the (i, j)-entry in
BTAT is a j1b1i + " + a jnbni , as above.
34. Use Theorem 3(d), treating x as an n×1 matrix: (ABx)T = xT(AB)T = xTBTAT.
35. [M] The answer here depends on the choice of matrix program. For MATLAB, use the help
command to read about zeros, ones, eye, and diag. For other programs see the
appendices in the Study Guide. (The TI calculators have fewer single commands that produce
special matrices.)
36. [M] The answer depends on the choice of matrix program. In MATLAB, the command
rand(5,6) creates a 5×6 matrix with random entries uniformly distributed between 0 and 1. The
command
round(19*(rand(4,4)–.5))
creates a random 4×4 matrix with integer entries between –9 and 9. The same result is produced by
the command randomint in the Laydata4 Toolbox on text website. For other matrix programs
see the appendices in the Study Guide.
37. [M] The equality AB = BA is very likely to be false for 4×4 matrices selected at random.
38. [M] (A + I)(A – I) – (A2 – I) = 0 for all 5×5 matrices. However, (A + B)(A – B) – A2 – B2 is the zero
matrix only in the special cases when AB = BA. In general,
(A + B)(A – B) = A(A – B) + B(A – B) = AA – AB + BA – BB.
39. [M] The equality (AT+BT)=(A+B)T and (AB)T=BTATshould always be true, whereas (AB)T = ATBT is
very likely to be false for 4×4 matrices selected at random.
40. [M] The matrix S “shifts” the entries in a vector (a, b, c, d, e) to yield (b, c, d, e, 0). The entries in S2
result from applying S to the columns of S, and similarly for S 3 , and so on. This explains the patterns
of entries in the powers of S:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.2
ª0
«0
«
S 2 = «0
«
«0
«¬ 0
0
0
0
0
0
1
0
0
0
0
0
1
0
0
0
0º
ª0
»
«0
0»
«
1 » , S 3 = «0
»
«
0»
«0
»
«¬ 0
0¼
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0º
ª0
»
«0
1»
«
0» , S 4 = «0
»
«
0»
«0
»
«¬ 0
0¼
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
• Solutions
93
1º
0 »»
0»
»
0»
0 »¼
S 5 is the 5×5 zero matrix. S 6 is also the 5×5 zero matrix.
ª.3339
41. [M] A5 = «.3349
«
«¬.3312
.3349
.3351
.3300
.3312 º
ª .333341
»
10
.3300 » , A = ««.333344
«¬.333315
.3388 »¼
.333344
.333350
.333306
.333315 º
.333306 »»
.333379 »¼
20
The entries in A all agree with .3333333333 to 8 or 9 decimal places. The entries in A30 all agree
with .33333333333333 to at least 14 decimal places. The matrices appear to approach the matrix
ª1/ 3 1/ 3 1/ 3º
«1/ 3 1/ 3 1/ 3» . Further exploration of this behavior appears in Sections 4.9 and 5.2.
«
»
«¬1/ 3 1/ 3 1/ 3»¼
Note: The MATLAB box in the Study Guide introduces basic matrix notation and operations,
including the commands that create special matrices needed in Exercises 35, 36 and elsewhere. The
Study Guide appendices treat the corresponding information for the other matrix programs.
2.2
SOLUTIONS
Notes: The text includes the matrix inversion algorithm at the end of the section because this topic is
popular. Students like it because it is a simple mechanical procedure. However, I no longer cover it in my
classes because technology is readily available to invert a matrix whenever needed, and class time is
better spent on more useful topics such as partitioned matrices. The final subsection is independent of the
inversion algorithm and is needed for Exercises 35 and 36.
Key Exercises: 8, 11–24, 35. (Actually, Exercise 8 is only helpful for some exercises in this section.
Section 2.3 has a stronger result.) Exercises 23 and 24 are used in the proof of the Invertible Matrix
Theorem (IMT) in Section 2.3, along with Exercises 23 and 24 in Section 2.1. I recommend letting
students work on two or more of these four exercises before proceeding to Section 2.3. In this way
students participate in the proof of the IMT rather than simply watch an instructor carry out the proof.
Also, this activity will help students understand why the theorem is true.
ª8
1. «
¬5
6º
4»¼
ª3
2. «
¬8
2º
5»¼
ª 7
3. «
¬ −6
−1
=
1 ª 4
32 − 30 «¬ −5
−6º ª 2
=
8»¼ «¬ −5/ 2
=
1 ª 5
15 − 16 «¬ −8
−2º ª −5
=
3»¼ «¬ 8
−1
3º
−3»¼
−1
=
ª −3
1
«
−21 − ( −18) ¬ 6
−3º
4 »¼
2º
−3»¼
−3º
1 ª −3
=− «
»
7¼
3¬ 6
−3º
ª 1
or «
»
7¼
¬ −2
1º
−7 / 3»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
94
CHAPTER 2
ª2
4. «
¬4
−4º
−6»¼
• Matrix Algebra
−1
=
ª −6
1
−12 − ( −16 ) «¬ −4
4º 1 ª −6
=
2»¼ 4 «¬ −4
4º
1 º
ª −3 / 2
or «
»
»
2¼
¬ −1 1/ 2¼
ª8
5. The system is equivalent to Ax = b, where A = «
¬5
6º
ª 2º
and b = « » , and the solution is
»
4¼
¬ −1¼
−3 º ª 2 º ª 7 º
=
. Thus x1 = 7 and x2 = –9.
4 »¼ «¬ −1»¼ «¬ −9 »¼
ª 2
x = A–1b = «
¬ −5 / 2
3º
ª 7
ª −9 º
–1
6. The system is equivalent to Ax = b, where A = «
and b = « » , and the solution is x = A b.
»
−
−
6
3
4
¬
¼
¬ ¼
To compute this by hand, the arithmetic is simplified by keeping the fraction 1/det(A) in front of the
matrix for A–1. (The Study Guide comments on this in its discussion of Exercise 7.) From Exercise 3,
1 ª −3
x = A–1b = − «
3¬ 6
ª 1 2º
7. a. «
»
¬5 12¼
−1
=
x = A–1b1 =
−3 º ª −9 º
1 ª 15 º ª −5º
. Thus x1 = −5 and x2 = 26/3.
=− «
=
»
«
»
7 ¼ ¬ 4¼
3 ¬ −26 »¼ «¬ 26 / 3»¼
−2º 1 ª 12
=
1»¼ 2 «¬ −5
ª 12
1
1⋅12 − 2 ⋅ 5 «¬ −5
1 ª 12
2 «¬ −5
−2º
ª 6
or «
»
1¼
¬ −2.5
−1º
.5»¼
−2 º ª −1º 1 ª −18º ª −9 º
. Similar calculations give
=
=
1»¼ «¬ 3»¼ 2 «¬ 8 »¼ «¬ 4 »¼
ª 11º
ª 6º
ª 13º
A−1b 2 = « » , A−1b 3 = « » , A−1b 4 = « » .
¬ −5 ¼
¬ −2 ¼
¬ −5 ¼
−1
3
ª1 2
b. [A b1 b2 b3 b4] = «
¬5 12
ª1
~«
¬0
2
2
−1
8
1
−10
ª1
~«
¬0
0
1
−9
4
11
−5
2
−4
6
−2
ª −9 º
The solutions are « » ,
¬ 4¼
1
−5
2
6
3º ª 1
~
−10 »¼ «¬ 0
2
1
3º
5 »¼
−1
4
1
−5
2
−2
3º
−5 »¼
13º
−5 »¼
ª 11º
« −5 » ,
¬ ¼
ª 6º
ª 13º
« −2 » , and « −5» , the same as in part (a).
¬ ¼
¬ ¼
Note: The Study Guide also discusses the number of arithmetic calculations for this Exercise 7, stating
that when A is large, the method used in (b) is much faster than using A–1.
8. Left-multiply each side of A = PBP–1 by P–1:
P–1A = P–1PBP–1, P–1A = IBP–1, P–1A = BP–1
Then right-multiply each side of the result by P:
P–1AP = BP–1P, P–1AP = BI, P–1AP = B
Parentheses are routinely suppressed because of the associative property of matrix multiplication.
9. a. True, by definition of invertible.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.2
• Solutions
95
b. False. See Theorem 6(b).
ª 1 1º
c. False. If A = «
» , then ab – cd = 1 – 0 ≠ 0, but Theorem 4 shows that this matrix is not
¬0 0 ¼
invertible, because ad – bc = 0.
d. True. This follows from Theorem 5, which also says that the solution of Ax = b is unique, for
each b.
e. True, by the box just before Example 6.
10. a. False. The last part of Theorem 7 is misstated here.
b. True, by Theorem 6(a).
c. False. The product matrix is invertible, but the product of inverses should be in the reverse order.
See Theorem 6(b).
d. True. See the subsection “Another View of Matrix Inversion”.
e. True, by Theorem 7.
11. (The proof can be modeled after the proof of Theorem 5.) The n×p matrix B is given (but is
arbitrary). Since A is invertible, the matrix A–1B satisfies AX = B, because A(A–1B) = A A–1B = IB =
B. To show this solution is unique, let X be any solution of AX = B. Then, left-multiplication of each
side by A–1 shows that X must be A–1B:
A–1 (AX) = A–1B, IX = A–1B, and X = A–1B.
12. Left-multiply each side of the equation AD = I by A–1 to obtain
A–1AD = A–1I, ID = A–1, and D = A–1.
Parentheses are routinely suppressed because of the associative property of matrix multiplication.
13. Left-multiply each side of the equation AB = AC by A–1 to obtain
A–1AB = A–1AC, IB = IC, and B = C.
This conclusion does not always follow when A is singular. Exercise 10 of Section 2.1 provides a
counterexample.
14. Right-multiply each side of the equation (B – C)D = 0 by D–1 to obtain
(B – C)DD–1 = 0D–1, (B – C)I = 0, B – C = 0, and B = C.
15. If you assign this exercise, consider giving the following Hint: Use elementary matrices and imitate
the proof of Theorem 7. The solution in the Instructor’s Edition follows this hint. Here is another
solution, based on the idea at the end of Section 2.2.
Write B = [b1 ⋅ ⋅ ⋅ bp] and X = [u1 ⋅ ⋅ ⋅ up]. By definition of matrix multiplication,
AX = [Au1 ⋅ ⋅ ⋅ Aup]. Thus, the equation AX = B is equivalent to the p systems:
Au1 = b1, … Aup = bp
Since A is the coefficient matrix in each system, these systems may be solved simultaneously,
placing the augmented columns of these systems next to A to form [A b1 ⋅ ⋅ ⋅ bp] = [A B]. Since A
is invertible, the solutions u1, …, up are uniquely determined, and [A b1 ⋅ ⋅ ⋅ bp] must row reduce to
[I u1 ⋅ ⋅ ⋅ up] = [I X]. By Exercise 11, X is the unique solution A–1B of AX = B.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
96
CHAPTER 2
• Matrix Algebra
16. Let C = AB. Then CB–1 = ABB–1, so CB–1 = AI = A. This shows that A is the product of invertible
matrices and hence is invertible, by Theorem 6.
Note: The Study Guide warns against using the formula (AB) –1 = B–1A–1 here, because this formula can
be used only when both A and B are already known to be invertible.
17. The box following Theorem 6 suggests what the inverse of ABC should be, namely, C–1B–1A–1. To
verify that this is correct, compute:
(ABC) C–1B–1A–1 = ABCC–1B–1A–1 = ABIB–1A–1 = ABB–1A–1 = AIA–1 = AA–1 = I
and
C–1B–1A–1 (ABC) = C–1B–1A–1ABC = C–1B–1IBC = C–1B–1BC = C–1IC = C–1C = I
18. Right-multiply each side of AB = BC by B–1:
ABB–1 = BCB–1, AI = BCB–1, A = BCB–1.
19. Unlike Exercise 18, this exercise asks two things, “Does a solution exist and what is it?” First, find
what the solution must be, if it exists. That is, suppose X satisfies the equation C–1(A + X)B–1 = I.
Left-multiply each side by C, and then right-multiply each side by B:
CC–1(A + X)B–1 = CI, I(A + X)B–1 = C, (A + X)B–1B = CB, (A + X)I = CB
Expand the left side and then subtract A from both sides:
AI + XI = CB, A + X = CB, X = CB – A
If a solution exists, it must be CB – A. To show that CB – A really is a solution, substitute it for X:
C–1[A + (CB – A)]B–1 = C–1[CB]B–1 = C–1CBB–1 = II = I.
Note: The Study Guide suggests that students ask their instructor about how many details to include in
their proofs. After some practice with algebra, an expression such as CC–1(A + X)B–1 could be simplified
directly to (A + X)B–1 without first replacing CC–1 by I. However, you may wish this detail to be included
in the homework for this section.
20. a. Left-multiply both sides of (A – AX)–1 = X–1B by X to see that B is invertible because it is the
product of invertible matrices.
b. Invert both sides of the original equation and use Theorem 6 about the inverse of a product
(which applies because X–1 and B are invertible):
A – AX = (X–1B)–1 = B–1(X–1)–1 = B–1X
Then A = AX + B–1X = (A + B–1)X. The product (A + B–1)X is invertible because A is invertible.
Since X is known to be invertible, so is the other factor, A + B–1, by Exercise 16 or by an
argument similar to part (a). Finally,
(A + B–1)–1A = (A + B–1)–1(A + B–1)X = X
Note: This exercise is difficult. The algebra is not trivial, and at this point in the course, most students
will not recognize the need to verify that a matrix is invertible.
21. Suppose A is invertible. By Theorem 5, the equation Ax = 0 has only one solution, namely, the zero
solution. This means that the columns of A are linearly independent, by a remark in Section 1.7.
22. Suppose A is invertible. By Theorem 5, the equation Ax = b has a solution (in fact, a unique solution)
for each b. By Theorem 4 in Section 1.4, the columns of A span Rn.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.2
• Solutions
97
23. Suppose A is n×n and the equation Ax = 0 has only the trivial solution. Then there are no free
variables in this equation, and so A has n pivot columns. Since A is square and the n pivot positions
must be in different rows, the pivots in an echelon form of A must be on the main diagonal. Hence A
is row equivalent to the n×n identity matrix.
24. If the equation Ax = b has a solution for each b in Rn, then A has a pivot position in each row, by
Theorem 4 in Section 1.4. Since A is square, the pivots must be on the diagonal of A. It follows that A
is row equivalent to In. By Theorem 7, A is invertible.
ªa b º
ª 0 0 º ª x1 º ª 0 º
25. Suppose A = «
and ad – bc = 0. If a = b = 0, then examine «
» « » = « » This has the
»
¬c d ¼
¬ c d ¼ ¬ x2 ¼ ¬ 0 ¼
ª dº
solution x1 = « » . This solution is nonzero, except when a = b = c = d. In that case, however, A is
¬−c ¼
ª −b º
the zero matrix, and Ax = 0 for every vector x. Finally, if a and b are not both zero, set x2 = « » .
¬ a¼
ª a b º ª −b º ª − ab + ba º ª 0º
Then Ax 2 = «
»« » = «
» = « » , because –cb + da = 0. Thus, x2 is a nontrivial solution
¬ c d ¼ ¬ a ¼ ¬ − cb + da ¼ ¬ 0¼
of Ax = 0. So, in all cases, the equation Ax = 0 has more than one solution. This is impossible when A
is invertible (by Theorem 5), so A is not invertible.
ª d
26. «
¬−c
−b º ª a
a »¼ «¬ c
ªa
«c
¬
b º ª da − bc
=
d »¼ «¬ 0
bº ª d
d »¼ «¬ − c
0
º
. Divide both sides by ad – bc to get CA = I.
− cb + ad »¼
− b º ª ad − bc
=
a »¼ «¬ 0
0
º
.
− cb + da »¼
Divide both sides by ad – bc. The right side is I. The left side is AC, because
1 ªa
ad − bc «¬ c
bº ª d
d »¼ «¬ −c
−b º ª a
=
a »¼ «¬ c
bº 1 ª d
d »¼ ad − bc «¬ −c
−b º
= AC
a »¼
27. a. Interchange A and B in equation (1) after Example 6 in Section 2.1: rowi (BA) = rowi (B)⋅A. Then
replace B by the identity matrix: rowi (A) = rowi (IA) = rowi (I)⋅A.
b. Using part (a), when rows 1 and 2 of A are interchanged, write the result as
ª row 2 ( A) º ª row 2 ( I ) ⋅ A º ª row 2 ( I ) º
«
» «
» «
»
« row 1 ( A) » = « row1 ( I ) ⋅ A » = « row1 ( I ) » A = EA
¬« row 3 ( A) ¼» ¬« row 3 ( I ) ⋅ A ¼» ¬« row 3 ( I ) ¼»
Here, E is obtained by interchanging rows 1 and 2 of I. The second equality in (*) is a
consequence of the fact that rowi (EA) = rowi (E)¸A.
c. Using part (a), when row 3 of A is multiplied by 5, write the result as
ª row1 ( A) º ª row1 ( I ) ⋅ A º ª row1 ( I ) º
« row ( A) » = « row ( I ) ⋅ A » = « row ( I ) » A = EA
2
2
2
«
» «
» «
»
«¬5 ⋅ row 3 ( A) »¼ «¬5 ⋅ row 3 ( I ) ⋅ A »¼ «¬5 ⋅ row 3 ( I ) »¼
Here, E is obtained by multiplying row 3 of I by 5.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
(*)
98
CHAPTER 2
• Matrix Algebra
28. When row 2 of A is replaced by row2(A) – 3⋅row1(A), write the result as
row1 ( A)
row1 ( I ) ⋅ A
ª
º ª
º
« row ( A) − 3 ⋅ row ( A) » = « row ( I ) ⋅ A − 3 ⋅ row ( I ) ⋅ A »
2
1
2
1
«
» «
»
«¬
»¼ «¬
»¼
row 3 ( A)
row 3 ( I ) ⋅ A
row1 ( I ) ⋅ A
row1 ( I )
ª
º ª
º
«
»
«
= «[row 2 ( I ) − 3 ⋅ row 1 ( I )] ⋅ A » = « row 2 ( I ) − 3 ⋅ row1 ( I ) »» A = EA
row 3 ( I ) ⋅ A
row 3 ( I )
¬«
¼» ¬«
¼»
Here, E is obtained by replacing row2(I) by row2(I) – 3⋅row1(I).
−3
−9
ª1
29. [ A I ] = «
¬4
ª −3
A–1 = «
¬ −4 / 3
30. [ A
ª3
I] = «
¬4
ª1
~«
¬0
31. [ A
32. [ A
ª 1
I ] = « −3
«
¬« 2
0º ª 1
~
1¼» ¬« 0
1
−4
−3
−4
0
3
1
0
0º ª1
~
1»¼ «¬ 4
1/ 3
4/3
0º ª 1
~
−1»¼ «¬ 0
0
1
−3
−2
4
4
1
0
0
2
7
0º ª 1
~
1 »¼ «¬ 0
1/ 3
0
0
1
−7 / 3
4/3
0º ª 1
0» ~ «0
» «
1¼» ¬« 0
0
1
0
2
−1
ª −7 / 3
A −1 = «
¬ 4/3
2º
.
−1»¼
0
1
−3
−2
−2
8
1
3
−2
0
1
0
8
10
7
3
4
3
1º
1»
»
1»¼
0
1
0
−2
−2
2
1
3
7
ª1
~ «0
«
«¬ 0
0
1
0
0
0
1
8
10
7/2
3
4
3/ 2
2
−1
1
0
−7
3
0
1
−6
4
0
0
2
−1
1
0
1
−1
4
1
0
0
10
2
0º
0 »» . The matrix A is not invertible.
1»¼
ª1
~ «« 0
«¬ 0
0
1
0
1
3
0º ª 1
0 » ~ «0
» «
1¼» ¬« 0
1 º
1 ».
»
1/ 2 »¼
0
1
0
0
0
2
ª 8
A = « 10
«
«¬ 7 / 2
3
4
3/ 2
−1
0º ª 1
0 »» ~ «« 0
1¼» ¬« 0
0º
1»¼
1/ 3
−4 / 3
ª1
~ «0
«
¬« 0
ª 1
I ] = «« −4
¬« −2
1º ª 1
~
1¼» ¬« 0
1º
1 / 3»¼
6
7
2
1
−3
3
0º ª 1
~
1¼» ¬« 0
1
0
2º
−1»¼
0º
0»
»
1»¼
1 º
1 »
»
1/ 2 »¼
2
−1
1
0
1
−1
4
1
−2
2
2
0
0º
0 »»
1¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
−3
−4 / 3
1º
1 / 3¼»
2.2
• Solutions
99
0 0 " 0º
ª 1
« −1
1 0
0 »»
«
» , and for j = 1, …, n, let aj, bj, and ej denote the jth columns of A, B,
33. Let B = « 0 −1 1
«
»
% % #»
« #
«¬ 0
0 " −1 1»¼
and I, respectively. Note that for j = 1, …, n – 1, aj – aj+1 = ej (because aj and aj+1 have the same
entries except for the jth row), bj = ej – ej+1 and an = bn = en.
To show that AB = I, it suffices to show that Abj = ej for each j. For j = 1, …, n – 1,
Abj = A(ej – ej+1) = Aej – Aej+1 = aj – aj+1 = ej
and Abn = Aen = an = en. Next, observe that aj = ej + ⋅ ⋅ ⋅ + en for each j. Thus,
Baj = B(ej + ⋅ ⋅ ⋅ + en) = bj + ⋅ ⋅ ⋅ + bn
= (ej – ej+1) + (ej+1 – ej+2) + ⋅ ⋅ ⋅ + (en–1 – en) + en = ej
This proves that BA = I. Combined with the first part, this proves that B = A–1.
Note: Students who do this problem and then do the corresponding exercise in Section 2.4 will appreciate
the Invertible Matrix Theorem, partitioned matrix notation, and the power of a proof by induction.
34. Let
ª1
«2
«
A = «3
«
«#
«¬ n
0
2
3
n
0
0
3
n
"
%
"
0º
ª1
»
« −1
0»
«
0 » , and B = « 0
»
«
#»
«#
«¬ 0
n »¼
0
1/ 2
−1 / 2
0
0
0
1/ 3
%
"
%
−1 / (n − 1)
0 º
»
»
»
»
# »
1 / n »¼
and for j = 1, …, n, let aj, bj, and ej denote the jth columns of A, B, and I, respectively. Note that for
1
1
j = 1, …, n–1, aj = jej +(j+1)ej+1 ⋅ ⋅ ⋅ +n en, an = n en, bj = (e j − e j +1 ) , and b n = en .
n
j
To show that AB = I, it suffices to show that Abj = ej for each j. For j = 1, …, n–1,
1
§1
·
Abj = A ¨ (e j − e j +1 ) ¸ = (a j − a j +1 )
j
© j
¹
=
1
1
ª¬ ( je j + ( j + 1) e j+1 + ! + ne n ) – (( j + 1)e j+1 + ! + ne n ) º¼ = je j = e j .
j
j
§1 · 1
Also, Abn = A ¨ e n ¸ = a n = e n .
©n ¹ n
Moreover,
Ba j = jBe j + ( j + 1) Be j+1 + ! + nBen = jb j + ( j + 1)b j +1 + ! + nb n
= (e j − e j+1 ) + (e j+1 − e j+2 ) + ! + (e n −1 − e n ) + e n = e j .
which proves that BA = I. Combined with the first part, this proves that B = A–1.
Note: If you assign Exercise 34, you may wish to supply a hint using the notation from Exercise 33:
Express each column of A in terms of the columns e1, …, en of the identity matrix. Do the same for B.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
100
CHAPTER 2
• Matrix Algebra
35. Row reduce [A e3]:
ª −1
« 2
«
«¬ 1
−7
−3
15
6
3
2
0º ª 1
0 »» ~ «« 0
1»¼ «¬ 0
7
3
1
0
−4
−1
0º ª 1
0 »» ~ «« 0
1»¼ «¬ 0
7
3
1
0
0
−1
0º ª 1
0 »» ~ «« 0
1»¼ «¬ 0
0
3
1
0
0
1
0º ª 1
0 »» ~ «« 0
−1»¼ «¬ 0
0
0
1
0
0
1
3º
0 »»
−1»¼
ª 3º
Answer: The third column of A–1 is « 0 » .
« »
«¬ −1»¼
36. [M] Write B = [A F], where F consists of the last two columns of I3, and row reduce:
ª −25
B = « 536
«
«¬ 154
−9
−27
0
185
537
1
52
143
0
0º
ª1
»
0 » ~ «« 0
«¬ 0
1»¼
0
0
.1126
1
0
−.5611
0
1
.0828
ª .1126
The last two columns of A are ~ « −.5611
«
«¬ .0828
–1
−.1559 º
1.0077 »»
−.1915 »¼
−.1559 º
1.0077 »»
−.1915 »¼
ª 1 1 −1º
37. There are many possibilities for C, but C = «
is the only one whose entries are 1, –1,
0 »¼
¬ −1 1
and 0. With only three possibilities for each entry, the construction of C can be done by trial and
error. This is probably faster than setting up a system of 4 equations in 6 unknowns. The fact that A
cannot be invertible follows from Exercise 25 in Section 2.1, because A is not square.
ª 1 1º
ª 1 0º
«0 1»
« 1 1»
«
»
» are
and D = «
38. Write AD = A[d1 d2] = [Ad1 Ad2]. The structure of A shows that D =
«0 0 »
« 1 1»
«
»
«
»
¬0 0 ¼
¬0 1¼
two possibilities. There are 9 possible answers. However, there is no 4×2 matrix C such that CA =
I4. If this were true, then CAx would equal x for all x in R4. This cannot happen because the columns
of A are linearly dependent and so Ax = 0 for some nonzero vector x. For such an x,
CAx = C(0) = 0. An alternate justification would be to cite Exercise 23 or 25 in Section 2.1.
ª.011
39. y = Df = «.003
«
«¬.001
.003
.009
.003
.001º ª 40 º ª.62 º
.003»» «« 50 »» = ««.66 »» . The deflections are .62 in., .66 in., and .52 in. at points
.011»¼ «¬ 30 »¼ «¬.52 »¼
1, 2, and 3, respectively.
40. [M] The stiffness matrix is D–1. Use an “inverse” command to produce
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.3
ª 3
100 «
D =
−1
3 «
«¬ 0
–1
−1
4
−1
• Solutions
101
0º
−1»»
3»¼
To find the forces (in pounds) required to produce a deflection of .04 cm at point 3, most students
will use technology to solve Df = (0, 0, .04) and obtain (0, –4/3, 4).
Here is another method, based on the idea suggested in Exercise 42. The first column of D–1 lists the
forces required to produce a deflection of 1 in. at point 1 (with zero deflection at the other points).
Since the transformation y 6 D–1y is linear, the forces required to produce a deflection of .04 cm at
point 3 is given by .04 times the third column of D–1, namely (.04)(100/3) times (0, –1, 3), or (0, –
4/3, 4) pounds.
41. To determine the forces that produce deflections of .07, .12, .16, and .12 cm at the four points on the
beam, use technology to solve Df = y, where y = (.07, .12, .16, .12). The forces at the four points are
.95, 6.19, 11.43, and 3.81 newtons, respectively.
42. [M] To determine the forces that produce a deflection of .22 cm at the second point on the beam, use
technology to solve Df = y, where y = (0, .22, 0, 0). The forces at the four points are –10.476,
31.429,
–10.476, and 0 newtons, respectively (to three significant digits). These forces are .22 times the
entries in the second column of D–1. Reason: The transformation y 6 D −1y is linear, so the forces
required to produce a deflection of .22 cm at the second point are .22 times the forces required to
produce a deflection of 1 cm at the second point. These forces are listed in the second column of D–1.
Another possible discussion: The solution of Dx = (0, 1, 0, 0) is the second column of D–1.
Multiply both sides of this equation by .22 to obtain D(.22x) = (0, .22, 0, 0). So .22x is the solution
of Df = (0, .22, 0, 0). (The argument uses linearity, but students may not mention this.)
Note: The Study Guide suggests using gauss, swap, bgauss, and scale to reduce [A I]
because I prefer to postpone the use of ref (or rref) until later. If you wish to introduce ref now,
see the Study Guide’s technology notes for Sections 2.8 or 4.3. (Recall that Sections 2.8 and 2.9 are only
covered when an instructor plans to skip Chapter 4 and get quickly to eigenvalues.)
2.3
SOLUTIONS
Notes: This section ties together most of the concepts studied thus far. With strong encouragement from
an instructor, most students can use this opportunity to review and reflect upon what they have learned,
and form a solid foundation for future work. Students who fail to do this now usually struggle throughout
the rest of the course. Section 2.3 can be used in at least three different ways.
(1) Stop after Example 1 and assign exercises only from among the Practice Problems and Exercises
1 to 28. I do this when teaching “Course 3” described in the text's “Notes to the Instructor. ” If you did not
cover Theorem 12 in Section 1.9, omit statements (f) and (i) from the Invertible Matrix Theorem.
(2) Include the subsection “Invertible Linear Transformations” in Section 2.3, if you covered Section
1.9. I do this when teaching “Course 1” because our mathematics and computer science majors take this
class. Exercises 29–40 support this material.
(3) Skip the linear transformation material here, but discusses the condition number and the
Numerical Notes. Assign exercises from among 1–28 and 41–45, and perhaps add a computer project on
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
102
CHAPTER 2
• Matrix Algebra
the condition number. (See the projects on our web site.) I do this when teaching “Course 2” for our
engineers.
The abbreviation IMT (here and in the Study Guide) denotes the Invertible Matrix Theorem (Theorem
8).
7º
ª 5
1. The columns of the matrix «
» are not multiples, so they are linearly independent. By (e) in
¬ −3 −6 ¼
the IMT, the matrix is invertible. Also, the matrix is invertible by Theorem 4 in Section 2.2 because
the determinant is nonzero.
2º
ª −4
2. The fact that the columns of «
» are multiples of each other is one way to show that this matrix
¬ 6 −3 ¼
is not invertible. Another is to check the determinant. In this case it is easily seen to be zero. By
Theorem 4 in Section 2.2, the matrix is not invertible.
3. Row reduction to echelon form is trivial because there is really no need for arithmetic calculations:
0
0º ª 3
0
0º ª5
0
0º
ª 3
« −3 −4
»
«
»
«
0 » ~ «0 −4
0 » ~ « 0 −4
0 »» The 3×3 matrix has 3 pivot positions and hence is
«
«¬ 8
5 −3»¼ «¬0
5 −3»¼ «¬ 0
0 −3»¼
invertible, by (c) of the IMT. [Another explanation could be given using the transposed matrix. But
see the note below that follows the solution of Exercise 14.]
ª −5 1 4 º
4. The matrix « 0 0 0 » cannot row reduce to the identity matrix since it already contains a row of
«
»
«¬ 1 4 9 »¼
zeros. Hence the matrix is not invertible (or singular) by (b) in the IMT.
ª 3 0 − 3º
5. The matrix « 2 0
4 »» obviously has linearly dependent columns (because one column is zero),
«
«¬ −4 0
7 »¼
and so the matrix is not invertible (or singular) by (e) in the IMT.
ª 1
6. « 0
«
¬« −3
−3
4
6
−6 º ª 1
3»» ~ «« 0
0 ¼» ¬«0
−3
4
−3
−6 º ª 1
3»» ~ «« 0
−18 ¼» ¬« 0
−3
4
1
−6 º ª 1
3»» ~ «« 0
6 ¼» ¬« 0
−3
−6 º
3»»
−21¼»
4
0
The matrix is invertible because it is row equivalent to the identity matrix.
ª −1
« 3
7. «
« −2
«
¬« 0
−3
5
0
8
−6
−1
3
2
1º ª −1
−3»» «« 0
~
2» « 0
» «
1¼» ¬« 0
−3
−4
0
8
0
−1
3
2
1º ª −1
0»» «« 0
~
0» « 0
» «
1¼» ¬« 0
−3
−4
0
8
0
0
3
0
1º
0 »»
0»
»
1¼»
The 4×4 matrix has four pivot positions and so is invertible by (c) of the IMT.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.3
ª3
«0
8. The 4×4 matrix «
«0
«
¬0
4
1
7
4
0
0
2
0
ª 4
« −6
9. [M] Using technology, «
« 7
«
¬ −1
• Solutions
103
4º
6 »»
is invertible because it has four pivot positions, by (c) of the IMT.
8»
»
1¼
0
9
−3
9
−5
2
10
4
−7 º ª 1
9 »» «« 0
~
19» « 0
» «
−1¼ ¬ 0
0
1
0
0
0
0
1
0
0º
0 »»
.
0»
»
1¼
The 4×4 matrix is invertible because it has four pivot positions, by (c) of the IMT.
7
9º ª 5
3
1
7
9º
ª5 3 1
«6 4 2
»
«
8 −8» «0
.4
.8
−.4 −18.8»»
«
10. [M] «7 5 3 10
9 » ~ «0
.8 1.6
.2
−3.6 »
«
» «
»
.6 2.2 −21.6 −21.2 »
« 9 6 4 −9 −5» «0
«¬ 8 5 2 11
4 »¼ «¬0
.2
.4
−.2 −10.4 »¼
ª5
«0
«
~ «0
«
«0
«¬0
3
.4
0
0
0
1
.8
0
1
0
7
−.4
1
−21
0
9º ª5
−18.8»» ««0
34 » ~ « 0
» «
7 » «0
−1»¼ «¬ 0
3
.4
0
0
0
1
.8
1
0
0
7
−.4
−21
1
0
9º
−18.8»»
7»
»
34 »
−1»¼
The 5×5 matrix is invertible because it has five pivot positions, by (c) of the IMT.
11. a.
b.
c.
d.
True, by the IMT. If statement (d) of the IMT is true, then so is statement (b).
True. If statement (h) of the IMT is true, then so is statement (e).
False. Statement (g) of the IMT is true only for invertible matrices.
True, by the IMT. If the equation Ax = 0 has a nontrivial solution, then statement (d) of the IMT
is false. In this case, all the lettered statements in the IMT are false, including statement (c),
which means that A must have fewer than n pivot positions.
e. True, by the IMT. If AT is not invertible, then statement (1) of the IMT is false, and hence
statement (a) must also be false.
12. a. True. If statement (k) of the IMT is true, then so is statement ( j). Use the first box after the IMT.
b. False. Notice that (i) if the IMT uses the work onto rather than the word into.
c. True. If statement (e) of the IMT is true, then so is statement (h).
d. False. Since (g) if the IMT is true, so is (f).
e. False, by the IMT. The fact that there is a b in n such that the equation Ax = b is consistent, does
not imply that statement (g) of the IMT is true, and hence there could be more than one solution.
Note: The solutions below for Exercises 13–30 refer mostly to the IMT. In many cases, however, part or
all of an acceptable solution could also be based on various results that were used to establish the IMT.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
104
CHAPTER 2
• Matrix Algebra
13. If a square upper triangular n×n matrix has nonzero diagonal entries, then because it is already in
echelon form, the matrix is row equivalent to In and hence is invertible, by the IMT. Conversely, if
the matrix is invertible, it has n pivots on the diagonal and hence the diagonal entries are nonzero.
14. If A is lower triangular with nonzero entries on the diagonal, then these n diagonal entries can be
used as pivots to produce zeros below the diagonal. Thus A has n pivots and so is invertible, by the
IMT. If one of the diagonal entries in A is zero, A will have fewer than n pivots and hence be
singular.
Notes: For Exercise 14, another correct analysis of the case when A has nonzero diagonal entries is to
apply the IMT (or Exercise 13) to AT. Then use Theorem 6 in Section 2.2 to conclude that since AT is
invertible so is its transpose, A. You might mention this idea in class, but I recommend that you not spend
much time discussing AT and problems related to it, in order to keep from making this section too lengthy.
(The transpose is treated infrequently in the text until Chapter 6.)
If you do plan to ask a test question that involves AT and the IMT, then you should give the students
some extra homework that develops skill using AT. For instance, in Exercise 14 replace “columns” by
“rows.” Also, you could ask students to explain why an n×n matrix with linearly independent columns
must also have linearly independent rows.
15. Part (h) of the IMT shows that a 4×4 matrix cannot be invertible when its columns do not span R4.
16. If A is invertible, so is AT, by (l) of the IMT. By (e) of the IMT applied to AT, the columns of AT are
linearly independent.
17. If A has two identical columns then its columns are linearly dependent. Part (e) of the IMT shows that
A cannot be invertible.
18. If A contains two identical rows, then it cannot be row reduced to the identity because subtracting
one row from the other creates a row of zeros. By (b) of the IMT, such a matrix cannot be invertible.
19. By (e) of the IMT, D is invertible. Thus the equation Dx = b has a solution for each b in R7, by (g) of
the IMT. Even better, the equation Dx = b has a unique solution for each b in R7, by Theorem 5 in
Section 2.2. (See the paragraph following the proof of the IMT.)
20. By (g) of the IMT, A is invertible. Hence, each equation Ax = b has a unique solution, by Theorem 5
in Section 2.2. This fact was pointed out in the paragraph following the proof of the IMT.
21. The matrix C cannot be invertible, by Theorem 5 in Section 2.2 or by the box following the IMT. So
(h) of the IMT is false and the columns of C do not span Rn.
22. By the box following the IMT, E and F are invertible and are inverses. So FE = I = EF, and so E and
F commute.
23. Statement (g) of the IMT is false for F, so statement (d) is false, too. That is, the equation Fx = 0 has
a nontrivial solution.
24. Statement (b) of the IMT is false for G, so statements (e) and (h) are also false. That is, the columns
of G are linearly dependent and the columns do not span Rn.
25. Suppose that A is square and AB = I. Then A is invertible, by the (k) of the IMT. Left-multiplying
each side of the equation AB = I by A–1, one has
A–1AB = A–1I, IB = A–1, and B = A–1.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.3
• Solutions
105
By Theorem 6 in Section 2.2, the matrix B (which is A–1) is invertible, and its inverse is (A–1)–1,
which is A.
26. If the columns of A are linearly independent, then since A is square, A is invertible, by the IMT. So
A2, which is the product of invertible matrices, is invertible. By the IMT, the columns of A2 span Rn.
27. Let W be the inverse of AB. Then ABW = I and A(BW) = I. Since A is square, A is invertible, by (k) of
the IMT.
Note: The Study Guide for Exercise 27 emphasizes here that the equation A(BW) = I, by itself, does not
show that A is invertible. Students are referred to Exercise 38 in Section 2.2 for a counterexample.
Although there is an overall assumption that matrices in this section are square, I insist that my students
mention this fact when using the IMT. Even so, at the end of the course, I still sometimes find a student
who thinks that an equation AB = I implies that A is invertible.
28. Let W be the inverse of AB. Then WAB = I and (WA)B = I. By (j) of the IMT applied to B in place of
A, the matrix B is invertible.
29. Since the transformation x 6 A x is one-to-one, statement (f) of the IMT is true. Then (i) is also true
and the transformation x 6 Ax does map Rn onto Rn. Also, A is invertible, which implies that the
transformation x 6 Ax is invertible, by Theorem 9.
30. Since the transformation x 6 Ax is not one-to-one, statement (f) of the IMT is false. Then (i) is also
false and the transformation x 6 Ax does not map Rn onto Rn. Also, A is not invertible, which
implies that the transformation x 6 Ax is not invertible, by Theorem 9.
31. Since the equation Ax = b has a solution for each b, the matrix A has a pivot in each row (Theorem 4
in Section 1.4). Since A is square, A has a pivot in each column, and so there are no free variables in
the equation Ax = b, which shows that the solution is unique.
Note: The preceding argument shows that the (square) shape of A plays a crucial role. A less revealing
proof is to use the “pivot in each row” and the IMT to conclude that A is invertible. Then Theorem 5 in
Section 2.2 shows that the solution of Ax = b is unique.
32. If Ax = 0 has only the trivial solution, then A must have a pivot in each of its n columns. Since A is
square (and this is the key point), there must be a pivot in each row of A. By Theorem 4 in Section
1.4, the equation Ax = b has a solution for each b in Rn.
Another argument: Statement (d) of the IMT is true, so A is invertible. By Theorem 5 in Section
2.2, the equation Ax = b has a (unique) solution for each b in Rn.
9º
ª −5
33. (Solution in Study Guide) The standard matrix of T is A = «
» , which is invertible because
¬ 4 −7 ¼
det A ≠ 0. By Theorem 9, the transformation T is invertible and the standard matrix of T–1 is A–1.
ª7 9 º
From the formula for a 2×2 inverse, A−1 = «
» . So
¬ 4 5¼
ª7
T −1 ( x1 , x2 ) = «
¬4
9 º ª x1 º
= ( 7 x1 + 9 x2 , 4 x1 + 5 x2 )
5 »¼ ¬« x2 ¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
106
CHAPTER 2
• Matrix Algebra
ª 2
34. The standard matrix of T is A = «
¬ −2
−8 º
, which is invertible because det A = -2 ≠ 0. By Theorem
7 »¼
1 ª7
9, T is invertible, and T −1 ( x ) = Bx, where B = A−1 = − «
2 ¬2
1 ª7
T −1 ( x1 , x2 ) = − «
2 ¬2
8º
. Thus
2 »¼
8 º ª x1 º § 7
·
= ¨ − x1 − 4 x2 , − x1 − x2 ¸
«
»
»
2 ¼ ¬ x2 ¼ © 2
¹
35. (Solution in Study Guide) To show that T is one-to-one, suppose that T(u) = T(v) for some vectors u
and v in Rn. Then S(T(u)) = S(T(v)), where S is the inverse of T. By Equation (1), u = S(T(u)) and
S(T(v)) = v, so u = v. Thus T is one-to-one. To show that T is onto, suppose y represents an arbitrary
vector in Rn and define x = S(y). Then, using Equation (2), T(x) = T(S(y)) = y, which shows that T
maps Rn onto Rn.
Second proof: By Theorem 9, the standard matrix A of T is invertible. By the IMT, the columns of A
are linearly independent and span Rn. By Theorem 12 in Section 1.9, T is one-to-one and maps Rn
onto Rn.
36. Let A be the standard matrix of T. By hypothesis, T is not a one-to-one mapping. So, by Theorem 12
in Section 1.9, the standard matrix A of T has linearly dependent columns. Since A is square, the
columns of A do not span Rn. By Theorem 12, again, T cannot map Rn onto Rn.
37. Let A and B be the standard matrices of T and U, respectively. Then AB is the standard matrix of the
mapping x 6 T (U (x)) , because of the way matrix multiplication is defined (in Section 2.1). By
hypothesis, this mapping is the identity mapping, so AB = I. Since A and B are square, they are
invertible, by the IMT, and B = A–1. Thus, BA = I. This means that the mapping x 6 U (T (x)) is the
identity mapping, i.e., U(T(x)) = x for all x in Rn.
38. Given any v in Rn, we may write v = T(x) for some x, because T is an onto mapping. Then, the
assumed properties of S and U show that S(v) = S(T(x)) = x and U(v) = U(T(x)) = x. So S(v) and U(v)
are equal for each v. That is, S and U are the same function from Rn into Rn.
39. If T maps Rn onto Rn, then the columns of its standard matrix A span Rn, by Theorem 12 in Section
1.9. By the IMT, A is invertible. Hence, by Theorem 9 in Section 2.3, T is invertible, and A–1 is the
standard matrix of T–1. Since A–1 is also invertible, by the IMT, its columns are linearly independent
and span Rn. Applying Theorem 12 in Section 1.9 to the transformation T–1, we conclude that T–1 is a
one-to-one mapping of Rn onto Rn.
40. Given u, v in n, let x = S(u) and y = S(v). Then T(x)=T(S(u)) = u and T(y) = T(S(v)) = v, by
equation (2). Hence
S (u + v) = S (T (x) + T (y ))
= S (T (x + y ))
=x+y
= S (u ) + S ( v )
Because T is linear
By equation (1)
So, S preserves sums. For any scalar r,
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.3
S ( r u ) = S (rT (x)) = S (T (r x))
= rx
• Solutions
107
Because T is linear
By equation (1)
= rS (u)
So S preserves scalar multiples. Thus S is a linear transformation.
41. [M] a. The exact solution of (3) is x1 = 3.94 and x2 = .49. The exact solution of (4) is x1 = 2.90 and
x2 = 2.00.
b. When the solution of (4) is used as an approximation for the solution in (3) , the error in using the
value of 2.90 for x1 is about 26%, and the error in using 2.0 for x2 is about 308%.
c. The condition number of the coefficient matrix is 3363. The percentage change in the solution
from (3) to (4) is about 7700 times the percentage change in the right side of the equation. This is
the same order of magnitude as the condition number. The condition number gives a rough
measure of how sensitive the solution of Ax = b can be to changes in b. Further information about
the condition number is given at the end of Chapter 6 and in Chapter 7.
Note: See the Study Guide’s MATLAB box, or a technology appendix, for information on condition
number. Only the TI-83+ and TI-89 lack a command for this.
42. [M] MATLAB gives cond(A) 10, which is approximately 101. If you make several trials with
MATLAB, which records 16 digits accurately, you should find that x and x1 agree to at least 14 or 15
significant digits. So about 1 significant digit is lost. Here is the result of one experiment. The
vectors were all computed to the maximum 16 decimal places but are here displayed with only four
decimal places:
ª.9501 º
«.2311 »
» , b = Ax =
x = rand(4,1) = «
«.6068 »
«
»
¬.4860 ¼
ª −1.4219º
« 6.2149»
«
» . The MATLAB solution is x1 = A\b =
« 20.7973»
«
»
¬ 1.4535¼
ª.9501 º
«.2311 »
«
».
«.6068 »
«
»
¬.4860 ¼
ª −.2220 º
« −.2220 »
» ×10–15. The computed solution x1 is accurate to about
However, x – x1 = «
«0
»
«
»
¬ −.1665 ¼
14 decimal places.
43. [M] MATLAB gives cond(A) = 69,000. Since this has magnitude between 104 and 105, the
estimated accuracy of a solution of Ax = b should be to about four or five decimal places less than
the 16 decimal places that MATLAB usually computes accurately. That is, one should expect the
solution to be accurate to only about 11 or 12 decimal places. Here is the result of one experiment.
The vectors were all computed to the maximum 16 decimal places but are here displayed with only
four decimal places:
ª.8214 º
«.4447 »
«
»
x = rand(5,1) = «.6154 » , b = Ax =
«
»
«.7919 »
«¬.9218 »¼
ª 19.8965º
ª.8214 º
« 6.8991»
«.4447 »
«
»
«
»
« 26.0354 » . The MATLAB solution is x1 = A\b = «.6154 » .
«
»
«
»
«.7919 »
« 0.7861»
«¬.9218 »¼
«¬ 22.4242 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
108
CHAPTER 2
• Matrix Algebra
ª −1679 º
« .3578»
«
»
However, x – x1 = « −.1775» × 10−11 . The computed solution x1 is accurate to about 11 decimal
«
»
« −.0084 »
«¬ .0002 »¼
places.
44. [M] Solve Ax = (0, 0, 0, 0, 1). MATLAB shows that cond( A) ≈ 4.8 × 105. Since MATLAB
computes numbers accurately to 16 decimal places, the entries in the computed value of x should be
accurate to at least 11 digits. The exact solution is (630, –12600, 56700, –88200, 44100).
45. [M] Some versions of MATLAB issue a warning when asked to invert a Hilbert matrix of order 12
or larger using floating-point arithmetic. The product AA–1 should have several off-diagonal entries
that are far from being zero. If not, try a larger matrix.
Note: All matrix programs supported by the Study Guide have data for Exercise 45, but only MATLAB
and Maple have a single command to create a Hilbert matrix.
Notes: The Study Guide for Section 2.3 organizes the statements of the Invertible Matrix Theorem in a
table that imbeds these ideas in a broader discussion of rectangular matrices. The statements are arranged
in three columns: statements that are logically equivalent for any m×n matrix and are related to existence
concepts, those that are equivalent only for any n×n matrix, and those that are equivalent for any n×p
matrix and are related to uniqueness concepts. Four statements are included that are not in the text’s
official list of statements, to give more symmetry to the three columns. You may or may not wish to
comment on them.
I believe that students cannot fully understand the concepts in the IMT if they do not know the correct
wording of each statement. (Of course, this knowledge is not sufficient for understanding.) The Study
Guide’s Section 2.3 has an example of the type of question I often put on an exam at this point in the
course. The section concludes with a discussion of reviewing and reflecting, as important steps to a
mastery of linear algebra.
2.4
SOLUTIONS
Notes: Partitioned matrices arise in theoretical discussions in essentially every field that makes use of
matrices. The Study Guide mentions some examples (with references).
Every student should be exposed to some of the ideas in this section. If time is short, you might omit
Example 4 and Theorem 10, and replace Example 5 by a problem similar to one in Exercises 1–10. (A
sample replacement is given at the end of these solutions.) Then select homework from Exercises 1–13,
15, and 21–24.
The exercises just mentioned provide a good environment for practicing matrix manipulation. Also,
students will be reminded that an equation of the form AB = I does not by itself make A or B invertible.
(The matrices must be square and the IMT is required.)
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.4
• Solutions
109
1. Apply the row-column rule as if the matrix entries were numbers, but for each product always write
the entry of the left block-matrix on the left.
ªI
«E
¬
B º ª IA + 0C
=
D »¼ «¬ EA + IC
0º ª A
I »¼ «¬ C
IB + 0 D º ª A
=
EB + ID »¼ «¬ EA + C
B º
EB + D »¼
2. Apply the row-column rule as if the matrix entries were numbers, but for each product always write
the entry of the left block-matrix on the left.
ªE
«0
¬
0 º ªP
F ¼» «¬ R
Q º ª EP + 0 R
=
S »¼ «¬ 0 P + FR
EQ + 0 S º ª EP
=
0Q + FS »¼ «¬ FR
EQ º
FS »¼
3. Apply the row-column rule as if the matrix entries were numbers, but for each product always write
the entry of the left block-matrix on the left.
ª0
«
¬I
B º ª 0 A + IC
=
D ¼» ¬« IA + 0C
Iº ªA
0 ¼» ¬«C
0 B + ID º ªC
=
IB + 0 D »¼ «¬ A
Dº
B »¼
4. Apply the row-column rule as if the matrix entries were numbers, but for each product always write
the entry of the left block-matrix on the left.
ª I
«− E
¬
0 º ªW
I »¼ «¬ Y
X º ª IW + 0Y
=
Z »¼ «¬ − EW + IY
IX + 0 Z º ª W
=
− EX + IZ »¼ «¬ − EW + Y
X
º
− EX + Z »¼
5. Compute the left side of the equation:
ªA
«
¬C
Bº ª I
0 »¼ «¬ X
0 º ª AI + BX
=
Y »¼ «¬ CI + 0 X
A0 + BY º
C 0 + 0Y »¼
Set this equal to the right side of the equation:
ª A + BX
« C
¬
BY º ª 0
=
0 »¼ «¬ Z
Iº
so that
0 »¼
A + BX = 0
C=Z
BY = I
0=0
Since the (2, 1) blocks are equal, Z = C. Since the (1, 2) blocks are equal, BY = I. To proceed further,
assume that B and Y are square. Then the equation BY =I implies that B is invertible, by the IMT, and
Y = B–1. (See the boxed remark that follows the IMT.) Finally, from the equality of the (1, 1) blocks,
BX = –A, B–1BX = B–1(–A), and X = –B–1A.
The order of the factors for X is crucial.
Note: For simplicity, statements (j) and (k) in the Invertible Matrix Theorem involve square matrices
C and D. Actually, if A is n×n and if C is any matrix such that AC is the n×n identity matrix, then C must
be n×n, too. (For AC to be defined, C must have n rows, and the equation AC = I implies that C has n
columns.) Similarly, DA = I implies that D is n×n. Rather than discuss this in class, I expect that in
Exercises 5–8, when students see an equation such as BY = I, they will decide that both B and Y should be
square in order to use the IMT.
6. Compute the left side of the equation:
ªX
«Y
¬
0º ªA
Z »¼ «¬ B
0 º ª XA + 0 B
=
C »¼ «¬ YA + ZB
X 0 + 0C º ª XA
=
Y 0 + ZC »¼ «¬YA + ZB
0 º
ZC »¼
Set this equal to the right side of the equation:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
110
CHAPTER 2
• Matrix Algebra
ª XA
«YA + ZB
¬
0 º ªI
=
ZC »¼ «¬ 0
0º
XA = I
so that
»
I¼
YA + ZB = 0
0=0
ZC = I
To use the equality of the (1, 1) blocks, assume that A and X are square. By the IMT, the equation
XA =I implies that A is invertible and X = A–1. (See the boxed remark that follows the IMT.)
Similarly, if C and Z are assumed to be square, then the equation ZC = I implies that C is invertible,
by the IMT, and Z = C–1. Finally, use the (2, 1) blocks and right-multiplication by A–1:
YA = –ZB = –C–1B, YAA–1 = (–C–1B)A–1, and Y = –C–1BA–1
The order of the factors for Y is crucial.
7. Compute the left side of the equation:
ªX
«Y
¬
0
0
ªA
0º «
0
I ¼» «
¬« B
Zº
ª XA + 0 + 0 B
0 »» = «
YA + 0 + IB
I ¼» ¬
XZ + 0 + 0 I º
YZ + 0 + II ¼»
Set this equal to the right side of the equation:
ª XA
«YA + B
¬
XZ º ª I
=
YZ + I »¼ «¬ 0
0º
XA = I
so that
»
I¼
YA + B = 0
XZ = 0
YZ + I = I
To use the equality of the (1, 1) blocks, assume that A and X are square. By the IMT, the equation XA
=I implies that A is invertible and X = A–1. (See the boxed remark that follows the IMT) Also, X is
invertible. Since XZ = 0, X – 1 XZ = X – 1 0 = 0, so Z must be 0. Finally, from the equality of the (2, 1)
blocks, YA = –B. Right-multiplication by A–1 shows that YAA–1 = –BA–1 and Y = –BA–1. The order of
the factors for Y is crucial.
8. Compute the left side of the equation:
ªA
«0
¬
Bº ª X
I »¼ «¬ 0
Y
0
Z º ª AX + B 0
=
I »¼ «¬ 0 X + I 0
AY + B 0
0Y + I 0
AZ + BI º
0 Z + II »¼
Set this equal to the right side of the equation:
ª AX
« 0
¬
AY
0
AZ + B º ª I
=
I »¼ «¬0
0
0
0º
I »¼
To use the equality of the (1, 1) blocks, assume that A and X are square. By the IMT, the equation XA
=I implies that A is invertible and X = A–1. (See the boxed remark that follows the IMT. Since AY =
0, from the equality of the (1, 2) blocks, left-multiplication by A–1 gives A–1AY = A–10 = 0, so Y = 0.
Finally, from the (1, 3) blocks, AZ = –B. Left-multiplication by A–1 gives A–1AZ = A–1(–B), and Z = –
A–1B. The order of the factors for Z is crucial.
Note: The Study Guide tells students, “Problems such as 5–10 make good exam questions. Remember to
mention the IMT when appropriate, and remember that matrix multiplication is generally not
commutative.” When a problem statement includes a condition that a matrix is square, I expect my
students to mention this fact when they apply the IMT.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.4
• Solutions
9. Compute the left side of the equation:
ª I
«A
« 21
«¬ A31
0
I
0
B12 º ª IB11 + 0 B21 + 0 B31
B22 »» = «« A21 B11 + IB21 + 0 B31
B32 »¼ «¬ A31 B11 + 0 B21 + IB31
0 º ª B11
0 »» «« B21
I »¼ «¬ B31
IB12 + 0 B22 + 0 B32 º
A21 B12 + IB22 + 0 B32 »»
A31 B12 + 0 B22 + IB32 »¼
Set this equal to the right side of the equation:
B11
ª
«A B + B
21
« 21 11
«¬ A31 B11 + B31
º ªC11
A21 B12 + B22 »» = «« 0
A31 B12 + B32 »¼ «¬ 0
B12
B11 = C11
so that A21 B11 + B21 = 0
A31B11 + B31 = 0
C12 º
C22 »»
C32 »¼
B12 = C12
A21B12 + B22 = C22
A31B12 + B32 = C32
Since the (2,1) blocks are equal, A21 B11 + B21 = 0 and A21 B11 = − B21. Since B11 is invertible, right
−1
gives A21 = − B21B11−1. Likewise since the (3,1) blocks are equal,
multiplication by B11
A31 B11 + B31 = 0 and A31 B11 = − B31 . Since B11 is invertible, right multiplication by
B11−1 gives A31 = − B31B11−1. Finally, from the (2,2) entries,
A21B12 + B22 = C22 . Since A21 = − B21B11−1 , C22 = − B21B11−1B12 + B22 .
10. Since the two matrices are inverses,
ªI
«A
«
«¬ B
0
I
D
0º ª I
0 »» «« P
I »¼ «¬Q
0
I
R
0º ª I
0 »» = «« 0
I »¼ «¬ 0
0
I
0
0º
0 »»
I »¼
Compute the left side of the equation:
ªI
«A
«
«¬ B
0
I
D
0º ª I
0 »» «« P
I »¼ «¬Q
0
I
R
0 º ª II + 0 P + 0Q
0 »» = «« AI + IP + 0Q
I »¼ «¬ BI + DP + IQ
I 0 + 0I + 0R
A0 + II + 0 R
B 0 + DI + I R
I 0 + 00 + 0 I º
A0 + I 0 + 0 I »»
B 0 + D 0 + I I »¼
Set this equal to the right side of the equation:
I
ª
« A+ P
«
¬« B + DP + Q
0
I
D+R
I=I
so that
A+ P = 0
B + DP + Q = 0
0º ª I
0 »» = «« 0
I ¼» ¬« 0
0
I
0
0=0
I=I
D+R=0
0º
0 »»
I ¼»
0=0
0=0
I =I
Since the (2,1) blocks are equal, A + P = 0 and P = − A . Likewise since the (3, 2) blocks are equal,
D + R = 0 and R = − D. Finally, from the (3,1) entries, B + DP + Q = 0 and Q = − B − DP.
Since P = − A, Q = − B − D(− A) = − B + DA .
11. a. True. See the subsection Addition and Scalar Multiplication.
b. False. See the paragraph before Example 3.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
111
112
CHAPTER 2
• Matrix Algebra
12. a. False. The both AB and BA are defined.
b. False. The RT and QT also need to be switched.
13. You are asked to establish an if and only if statement. First, supose that A is invertible,
ªD
and let A−1 = «
¬F
ªB
«0
¬
Eº
. Then
G »¼
0 º ªD
C »¼ «¬ F
E º ª BD
=
G »¼ «¬ CF
BE º ª I
=
CG »¼ «¬ 0
0º
I »¼
Since B is square, the equation BD = I implies that B is invertible, by the IMT. Similarly, CG = I
implies that C is invertible. Also, the equation BE = 0 imples that E = B −1 0 = 0. Similarly F = 0.
Thus
−1
0 º
(*)
»
C −1 »¼
This proves that A is invertible only if B and C are invertible. For the “if ” part of the statement,
suppose that B and C are invertible. Then (*) provides a likely candidate for A−1 which can be used
to show that A is invertible. Compute:
ªB
A−1 = «
¬0
0º
ªD
=«
»
C¼
¬E
E º ª B −1
=«
G »¼ «¬ 0
0 º ª B −1
«
C »¼ «¬ 0
0 º ª BB −1
0 º ª I 0º
=
»
«
»=«
»
C −1 »¼ «¬ 0
CC −1 »¼ ¬0 I ¼
Since A is square, this calculation and the IMT imply that A is invertible. (Don’t forget this final
sentence. Without it, the argument is incomplete.) Instead of that sentence, you could add the
equation:
ªB
«0
¬
ª B −1
«
¬« 0
0 º ªB
»«
C −1 ¼» ¬ 0
0 º ª B −1B
=«
C ¼» ¬« 0
0 º ªI
»=«
C −1C ¼» ¬0
0º
I »¼
14. You are asked to establish an if and only if statement. First suppose that A is invertible. Example 5
shows that A11 and A22 are invertible. This proves that A is invertible only if A11 A22 are invertible. For
the if part of this statement, suppose that A11 and A22 are invertible. Then the formula in Example 5
provides a likely candidate for A−1 which can be used to show that A is invertible . Compute:
ª A11
« 0
¬
−1
A12 º ª A 11
«
A22 »¼ « 0
¬
−1
−1
A12 A −221 º ª A11 A 11 + A12 0
− A 11
«
=
»
−1
A −221
¼» «¬ 0 A 11 + A22 0
−1
−1
−1
º
A11 ( − A 11
) A 12 A 22
+ A12 A 22
»
−1
−1
−1
0(− A 11 ) A12 A 22 + A 22 A 22 »¼
ªI
=«
¬0
−1
−( A11 A 11
) A12 A −221 + A12 A −221 º
»
I
¼
ªI
=«
¬0
− A12 A −221 + A 12 A −221 º ª I
»=«
I
¼ ¬0
0º
I »¼
Since A is square, this calculation and the IMT imply that A is invertible.
15. The column-row expansions of Gk and Gk+1 are:
Gk = X k X kT
= col1 ( X k ) row1 ( X kT ) + ... + colk ( X k ) row k ( X kT )
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.4
• Solutions
113
and
Gk +1 = X k +1 X kT+1
= col1 ( X k +1 ) row1 ( X kT+1 ) + ... + colk ( X k +1 ) row k ( X kT+1 ) + colk +1 ( X k +1 ) row k +1 ( X kT+1 )
= col1 ( X k ) row1 ( X kT ) + ... + colk ( X k ) row k ( X kT ) + colk +1 ( X k +1 ) row k +1 ( X kT )
= Gk + colk +1 ( X k +1 ) row k +1 ( X kT )
since the first k columns of Xk+1 are identical to the first k columns of Xk. Thus to update Gk to
produce Gk+1, the matrix colk+1 (Xk+1) rowk+1 ( X kT ) should be added to Gk.
16. Compute the right side of the equation:
ªI
«X
¬
0º ª A11
I »¼ «¬ 0
0º ªI
S »¼ «¬0
Y º ª A11
=
I »¼ «¬ X A 11
0º ªI
S »¼ «¬0
Y º ª A11
=«
I »¼ ¬ X A11
A11Y º
X A11Y + S »¼
Set this equal to the left side of the equation:
ª A 11
«X A
¬
11
A 11Y º ª A 11
=
X A 11Y + S »¼ «¬ A 21
A 11 = A 11
A 12 º
so that
»
X A11 = A 21
A 22 ¼
A 11Y = A 12
X A 11 Y + S = A 22
−1
gives
Since the (1, 2) blocks are equal, A 11Y = A 12. Since A11 is invertible, left multiplication by A 11
−1
A12. Likewise since the (2,1) blocks are equal, X A11 = A21. Since A11 is invertible, right
Y = A 11
multiplication by A11−1 gives that X = A21 A11−1. One can check that the matrix S as given in the exercise
satisfies the equation X A11Y + S = A22 with the calculated values of X and Y given above.
17. Suppose that A and A11 are invertible. First note that
ªI
«X
¬
0º ª I
I »¼ «¬ − X
0º ª I
=
I »¼ «¬ 0
0º
I »¼
and
ªI
«0
¬
Y º ªI
I »¼ «¬ 0
−Y º ª I
=
I »¼ «¬ 0
0º
I »¼
ª I 0º
ªI Y º
Since the matrices «
and «
»
»
¬X I ¼
¬0 I ¼
are square, they are both invertible by the IMT. Equation (7) may be left multipled by
ªI
«X
¬
0º
I »¼
−1
ª A11
«0
¬
ªI
and right multipled by «
¬0
0º ª I
=
S »¼ «¬ X
−1
−1
Yº
to find
I »¼
−1
0º
ªI Y º
A«
»
»
I¼
¬0 I ¼
0º
ªA
is invertible as the product of invertible matrices. Finally,
Thus by Theorem 6, the matrix « 11
S »¼
¬ 0
Exercise 13 above may be used to show that S is invertible.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
114
CHAPTER 2
• Matrix Algebra
18. Since W = [ X x0 ] ,
ªX T º
ª X T X X T x0 º
W T W = « T » [ X x0 ] = « T
»
xT0 x0 »¼
«¬ x0 »¼
«¬ x0 X
By applying the formula for S from Exercise 15, S may be computed:
S = xT0 x 0 − xT0 X ( X T X ) −1 X T x 0
= xT0 ( I m − X ( X T X ) −1 X T )x 0
= xT0 Mx0
19. The matrix equation (8) in the text is equivalent to
( A − sI n ) x + Bu = 0 and Cx + u = y
Rewrite the first equation as ( A − sI n ) x = − Bu. When A − sI n is invertible,
x = ( A − sI n )−1 (− Bu) = −( A − sI n )−1 Bu
Substitute this formula for x into the second equation above:
C (−( A − sI n )−1 Bu) + u = y ,so that I m u − C ( A − sI n )−1 Bu = y
Thus y = ( I m − C ( A − sI n )−1 B)u. If W (s) = I m − C ( A − sI n )−1 B, then y = W (s)u. The matrix W(s) is
the Schur complement of the matrix A − sI n in the system matrix in equation (8)
20. The matrix in question is
Bº
ª A − BC − sI n
«
−C
I m »¼
¬
By applying the formula for S from Exercise 16, S may be computed:
S = I m − (−C )( A − BC − sI n ) −1 B
= I m + C ( A − BC − sI n ) −1 B
ª1
21. a. A2 = «
¬2
0º ª 1
−1»¼ «¬ 2
ªA
b. M 2 = «
¬I
0º ª A
− A»¼ «¬ I
0º ª 1 + 0
=«
−1»¼ ¬ 2 − 2
0 + 0º ª 1
»=«
0 + (−1) 2 ¼ ¬0
0 º ª A2 + 0
=«
− A»¼ ¬« A − A
0 + 0 º ªI
»=«
0 + (− A) 2 ¼» ¬0
ª I2
22. Let C be any nonzero 2×2 matrix. Define M = « 0
«
«¬ C
ªI2
M = «« 0
«¬ C
2
0
I2
0
0 º ªI2
0 »» «« 0
− I 2 »¼ «¬ C
0
I2
0
0º
1»¼
0 º ª I2
0 »» = «« 0
− I 2 »¼ «¬ C − C
0º
I »¼
0 º
0 »» . Then
− I 2 »¼
0
I2
0
0
I2
0
0 º ª I2
0 »» = «« 0
I 2 »¼ «¬ 0
0
I2
0
0º
0 »»
I 2 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.4
• Solutions
115
23. The product of two 1×1 “lower triangular” matrices is “lower triangular.” Suppose that for n = k, the
product of two k×k lower triangular matrices is lower triangular, and consider any (k+1)× (k+1)
matrices A1 and B1. Partition these matrices as
ª a 0T º
ª b 0T º
=
,
A1 = «
B
»
«
»
1
¬v A ¼
¬w B ¼
where A and B are k×k matrices, v and w are in Rk, and a and b are scalars. Since A1 and B1 are lower
triangular, so are A and B. Then
T
T
T
0T º ª ab + 0 w a0 + 0 B º ª ab
0T º
=
»
»=«
«
»
B ¼ «¬ vb + Aw
v0T + AB »¼ ¬bv + Aw AB ¼
Since A and B are k×k, AB is lower triangular. The form of A1B1 shows that it, too, is lower
triangular. Thus the statement about lower triangular matrices is true for n = k +1 if it is true for n =
k. By the principle of induction, the statement is true for all n > 1.
ªa
A1B1 = «
¬v
0T º ª b
»«
A ¼ ¬w
Note: Exercise 23 is good for mathematics and computer science students. The solution of Exercise 23 in
the Study Guide shows students how to use the principle of induction. The Study Guide also has an
appendix on “The Principle of Induction,” at the end of Section 2.4. The text presents more applications
of induction in Section 3.2 and in the Supplementary Exercises for Chapter 3.
ª1
«1
«
24. Let An = «1
«
«#
«¬1
0
1
1
1
0 "
0
1
%
1 "
0º
ª 1
»
« −1
0»
«
0 » , Bn = « 0
»
«
»
« #
«¬ 0
1»¼
0
1
−1
0
0
1
%
"
"
%
−1
0º
0 »»
0» .
»
»
1»¼
By direct computation A2B2 = I2. Assume that for n = k, the matrix AkBk is Ik, and write
ª1
Ak +1 = «
¬v
ª1
0T º
» and Bk +1 = «
Ak ¼
¬w
0T º
»
Bk ¼
where v and w are in Rk , vT = [1 1 ⋅ ⋅ ⋅ 1], and wT = [–1 0 ⋅ ⋅ ⋅ 0]. Then
T
0T + 0T Bk º ª1 0T º
0T º ª1 + 0 w
=
«
»=«
»
» = I k +1
Bk ¼ «¬ v + Ak w v0T + Ak Bk »¼ ¬0 I k ¼
The (2,1)-entry is 0 because v equals the first column of Ak., and Akw is –1 times the first column of
Ak. By the principle of induction, AnBn = In for all n > 2. Since An and Bn are square, the IMT shows
that these matrices are invertible, and Bn = An−1.
ª1
Ak +1 Bk +1 = «
¬v
0T º ª 1
»«
Ak ¼ ¬ w
Note: An induction proof can also be given using partitions with the form shown below. The details are
slightly more complicated.
ª Ak
Ak +1 = « T
¬v
0º
ª Bk
» and Bk +1 = « T
1¼
¬w
0º
»
1¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
missing
2.4
For [ j=10, j<=19, j++,
A [[ i,j ]] = B [[ i-4, j-9 ]] ] ];
• Solutions
117
Colon suppresses
output.
0 º
ªA
c. To create B = «
with MATLAB, build B out of four blocks:
T »
¬0 A ¼
B = [A zeros(20,20); zeros(30,30) A’];
Another method: first enter B = A ; and then enlarge B with the command
B(21:50, 31:50) = A’;
This places AT in the (2, 2) block of the larger B and fills in the (1, 2) and (2, 1) blocks with zeros.
For Maple:
B := matrix(50,50,0):
copyinto(A, B, 1, 1):
copyinto( transpose(A), B, 21, 31):
For Mathematica:
B = BlockMatrix[ {{A, ZeroMatrix[20,20]}, ZeroMatrix[30,30],
Transpose[A]}} ]
27. a. [M] Construct A from four blocks, say C11, C12, C21, and C22, for example with C11 a 30×30
matrix and C22 a 20×20 matrix.
MATLAB:
C11 = A(1:30, 1:30) + B(1:30, 1:30)
C12 = A(1:30, 31:50) + B(1:30, 31:50)
C21 = A(31:50, 1:30)+ B(31:50, 1:30)
C22 = A(31:50, 31:50) + B(31:50, 31:50)
C = [C11 C12; C21 C22]
The commands in Maple and Mathematica are analogous, but with different syntax. The first
commands are:
C11 := submatrix(A, 1..30, 1..30) + submatrix(B, 1..30,
Maple:
1..30)
Mathematica:
C11 := Take[ A, {1,30), {1,30} ] + Take[B, {1,30), {1,30}
]
b. The algebra needed comes from block matrix multiplication:
ªA
AB = « 11
¬ A21
A12 º ª B11
A22 »¼ «¬ B21
B12 º ª A11 B11 + A12 B21
=
B22 »¼ «¬ A21 B11 + A22 B21
A11 B12 + A12 B22 º
A21 B12 + A22 B22 »¼
Partition both A and B, for example with 30×30 (1, 1) blocks and 20×20 (2, 2) blocks. The four
necessary submatrix computations use syntax analogous to that shown for (a).
0 º ª x1 º ª b1 º
ªA
c. The algebra needed comes from the block matrix equation « 11
» « » = « » , where x1
¬ A21 A22 ¼ ¬ x 2 ¼ ¬b 2 ¼
and b1 are in R20 and x2 and b2 are in R30. Then A1 1x1 = b1, which can be solved to produce x1.
Once x1 is found, rewrite the equation A21x1 + A22x2 = b2 as A22x2 = c, where c = b2 – A21x1, and
solve A22x2 = c for x2.
Notes: The following may be used in place of Example 5:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.4
• Solutions
118
Example 5: Use equation (*) to find formulas for X, Y, and Z in terms of A, B, and C. Mention any
assumptions you make in order to produce the formulas.
ªX
«Y
¬
0º ª I
Z »¼ «¬ A
0º ª I
=
B »¼ «¬C
0º
I »¼
(*)
Solution:
This matrix equation provides four equations that can be used to find X, Y, and Z:
X + 0 = I,
0=0
YI + ZA = C,
Y0 + ZB = I
(Note the order of the factors.)
The first equation says that X = I. To solve the fourth equation, ZB = I, assume that B and Z are
square. In this case, the equation ZB = I implies that B and Z are invertible, by the IMT. (Actually, it
suffices to assume either that B is square or that Z is square.) Then, right-multiply each side of ZB = I
to get ZBB–1 = IB–1 and Z = B–1. Finally, the third equation is Y + ZA = C. So, Y + B–1A = C, and Y =
C – B–1A.
The following counterexample shows that Z need not be square for the equation (*) above to be true.
ª1
«0
«
«1
«
«¬ 3
0
0
0
1
0
0
2
1
3
4
1
0
ª1
0º «
0
0 »» «
«1
1» «
» 1
−1»¼ «
«¬ 1
0
0
1
0
1
2
1
−1
−1
2
0º
ª1
0 »» «
0
5» = «
» «6
−3» «
«¬ 3
4 »¼
0
0
1
0
5
1
6
0
0º
0 »»
0»
»
1»¼
Note that Z is not determined by A, B, and C, when B is not square. For instance, another Z that
5 0º
ª 3
works in this counterexample is Z = «
».
¬ −1 −2 0 ¼
2.5
SOLUTIONS
Notes: Modern algorithms in numerical linear algebra are often described using matrix factorizations.
For practical work, this section is more important than Sections 4.7 and 5.4, even though matrix
factorizations are explained nicely in terms of change of bases. Computational exercises in this section
emphasize the use of the LU factorization to solve linear systems. The LU factorization is performed
using the algorithm explained in the paragraphs before Example 2, and performed in Example 2. The text
discusses how to build L when no interchanges are needed to reduce the given matrix to U. An appendix
in the Study Guide discusses how to build L in permuted unit lower triangular form when row
interchanges are needed. Other factorizations are introduced in Exercises 22–26.
ª 1
1. L = « −1
«
¬« 2
0
1
−5
ª 1
[ L b ] = «« −1
«¬ 2
0º
ª3
»
0 » , U = «« 0
«¬ 0
1¼»
0
0
1
0
−5
1
−7
−2
0
−7 º ª 1
5 »» ~ «« 0
2 »¼ «¬ 0
−2 º
ª −7 º
»
−1» , b = «« 5 »» . First, solve Ly = b.
−1¼»
¬« 2 ¼»
0
0
1
0
−5
1
−7 º
−2 »» . The only arithmetic is in column 4
16 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.5
ª1
∼ «« 0
«¬ 0
0
0
1
0
0
1
• Solutions
−7 º
ª −7 º
»
−2 » , so y = «« −2 »» .
«¬ 6 »¼
6 »¼
Next, solve Ux = y, using back-substitution (with matrix notation).
ª3
[U y ] = «« 0
«¬ 0
ª3
~ «« 0
¬« 0
−7
−2
−2
−1
0
−1
−7
0
1
0
0
1
−7 º ª 3
−2 »» ∼ ««0
6 »¼ «¬ 0
−19 º ª 3
4 »» ∼ «« 0
−6 »¼ «¬ 0
−7
−2
−2
−1
0
1
0
0
1
0
0
1
−7 º ª 3
−2 »» ∼ «« 0
−6 »¼ «¬ 0
9º ª 1
4 »» ∼ «« 0
−6 »¼ «¬ 0
−7
0
−2
0
0
1
0
0
1
0
0
1
−19 º
−8»»
−6 »¼
3º
4 »» , So x =
−6 »¼
ª3º
« 4 ».
« »
«¬ −6 »¼
To confirm this result, row reduce the matrix [A b]:
ª 3
«
[ A b ] = « −3
¬« 6
−7
−2
5
1
−4
0
−7 º ª 3
5 »» ∼ ««0
2 ¼» ¬«0
−7
−2
−2
−1
10
4
−7 º ª 3
−2 »» ∼ «« 0
16 ¼» ¬« 0
−7
−2
−2
−1
0
−1
−7 º
−2 »»
6 ¼»
From this point the row reduction follows that of [U y] above, yielding the same result.
ª 1
2. L = « −2
«
«¬ 0
0
1
1
ª 1
[ L b ] = «« −2
¬« 0
0º
ª2
»
0 » , U = «« 0
«¬ 0
1»¼
0
0
1
0
1
1
−6
4º
ª 2º
»
8 » , b = «« −4 »» . First, solve Ly = b:
«¬ 6 »¼
−2 »¼
−4
0
2º ª 1
−4 »» ∼ «« 0
6 ¼» ¬« 0
0
0
1
0
0
1
2º
ª2º
»
0 » , so y = «« 0 »» .
6 ¼»
¬« 6 ¼»
Next solve Ux = y, using back-substitution (with matrix notation):
ª2
[U y ] = «« 0
«¬ 0
ª2
∼ «« 0
«¬ 0
0
0
1
0
0
1
−6
4
−4
8
0
−2
2º ª2
0 »» ∼ «« 0
6 »¼ «¬ 0
−22 º ª 1
−6 »» ∼ «« 0
−3»¼ «¬ 0
0
0
1
0
0
1
−6
0
−4
0
0
1
14 º
24 »»
−3»¼
−11º
ª −11º
»
−6 » , so x = «« −6 »» .
«¬ −3 »¼
−3»¼
To confirm this result, row reduce the matrix [A b]:
ª 2
[ A b ] = «« −4
«¬ 0
−6
4
8
0
−4
6
2º ª2
−4 »» ∼ «« 0
6 »¼ «¬ 0
−6
4
−4
8
0
−2
2º
0 »»
6 »¼
From this point the row reduction follows that of [U y] above, yielding the same result.
ª 1
3. L = « −2
«
«¬ 3
0
1
−1
0º
ª2
»
0 » , U = «« 0
«¬ 0
1»¼
−4
−3
0
2º
ª6º
»
6 » , b = «« 0 »» . First, solve Ly = b:
«¬ 6 »¼
1»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
119
120
CHAPTER 2
• Matrix Algebra
6º ª 1 0 0 6º
ª6º
ª 1 0 0 6º ª 1 0 0
«
»
«
»
«
»
[ L b ] = « −2
1 0 0» ∼ «0
1 0
12 » ∼ « 0 1 0 12 » , so y = ««12 »» .
«¬ 3 −1 1 6 »¼ «¬ 0 −1 1 −12 »¼ «¬ 0 0 1 0 »¼
«¬ 0 »¼
Next solve Ux = y, using back-substitution (with matrix notation):
−4
ª2
[U y ] = «« 0
«¬ 0
−3
0
−4
6º ª 2
6 12 »» ∼ «« 0
1 0 »¼ «¬ 0
2
−3
0
6º ª 2
0 12 »» ∼ «« 0
1 0 »¼ «¬ 0
0
0
0
1
0
0
1
−10 º ª 1
−4 »» ∼ «« 0
0 »¼ «¬ 0
0
0
1
0
0
1
−5 º
−4 »» ,
0 »¼
ª −5º
« »
so x = −4 .
« »
«¬ 0 »¼
ª1
4. L = «1
«
¬«3
1
−5
ª1
[ L b ] = ««1
«¬3
−1
0º
ª1
»
0 » ,U = ««0
«¬0
1¼»
0
0
0
1
0
−5
1
2º
ª 0º
»
−1» , b = «« −5»» . First, solve Ly = b:
−6 ¼»
¬« 7 ¼»
−2
0
0º ª 1
−5 »» ∼ «« 0
7 »¼ «¬ 0
0
0
1
0
−5
1
0º ª 1
−5 »» ∼ ««0
7 »¼ «¬0
0
0
1
0
0
1
0º
ª 0º
»
−5 » , so y = «« −5»» .
«¬ −18»¼
−18 »¼
Next solve Ux = y, using back-substitution (with matrix notation):
2
0 º ª 1 −1
2
0 º ª 1 −1 0
ª 1 −1
«
»
«
[U y ] = «0 −2 −1
−5 » ∼ « 0 −2 −1 −5»» ∼ «« 0 −2 0
«¬0
0 −6 −18 »¼ «¬ 0
0
1
3»¼ «¬ 0
0 1
ª1
«
∼ «0
«¬ 0
−1
0
1
0
0
1
ª 1
« 3
5. L = «
« −1
«
¬ −3
0
1
0
4
−6 º ª 1
1»» ∼ «« 0
3»¼ «¬ 0
0
0
1
0
0
1
0º
ª1
»
«0
0»
,U = «
«0
1 0»
»
«
−2 1 ¼
¬0
0
0
ª 1
« 3
[ L b] = «
« −1
«
¬ −3
ª1
«0
∼«
«0
«
¬0
0
1
0
4
0
1
0
0
0
0
0
0
1 0
−2 1
0
0
0
0
1 0
−2 1
−6 º
−2 »»
3»¼
ª −5 º
−5 º
« »
1»» , so x = « 1 » .
3»¼
«¬ 3 »¼
−2
−3
−2
6
0
0
2
0
−3º
ª 1º
»
«6»
0»
, b = « » . First solve Ly = b:
«0 »
4»
»
« »
1¼
¬ 3¼
1º ª 1
6 »» ««0
∼
0 » «0
» «
3¼ ¬0
0
1
0
4
1 0
−2 1
1º ª 1
3»» ««0
∼
1» «0
» «
−6 ¼ ¬0
0
1
0
0
0
0
0
0
1
0
0
1
0
0
0
0
1º
3»»
1»
»
6¼
1º
ª 1º
« 3»
»
3»
, so y = « » .
« 1»
1»
« »
»
−4 ¼
¬ −4 ¼
Next solve Ux = y, using back-substitution (with matrix notation):
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.5
ª1
«0
«
[U y ] =
«0
«
¬0
ª1
«0
«
∼
«0
«
¬0
−2
−3
−2
6
−3
0
0
0
2
0
4
1
1º ª 1
3»» ««0
∼
1» «0
» «
−4 ¼ ¬0
−2
−3
0
0
0
0
1
0
6º
ª1
»
«0
−48»
«
∼∼
«0
0 17 / 2 »
»
«
−4 ¼
1
¬0
0
1
0
0
−3
4
1
−1
ª 1
« −2
6. L = «
« 3
«
¬ −5
0
0
0º
ª1
»
«0
0»
,U = «
«0
0»
»
«
1¼
¬0
ª 1
« −2
[ L b] = «
« 3
«
¬ −5
0
1
0
0
0
0
−3
4
1
−1
0
1
ª1
«0
∼«
«0
«
¬0
matrix notation):
0
1
0
0
0
0
0
0
1
0
0
1
ª1
«0
[U y ] = «
«0
«
¬0
3
3
2
0
0
12
0
0
−2
0
0
1
ª1
«0
∼«
«0
«
¬0
0
1
0
0
0
0
0
0
1
0
0
1
−2
−3
−2
6
0
0
0
0
2
0
0
1
−2
1
0
0
0
0
1
0
3
3
2
0
0
0
−2
0
−11º ª 1
3»» ««0
∼
17 » «0
» «
−4 ¼ ¬0
6º ª 1
16 »» «« 0
∼
0 17 / 2 » « 0
» «
−4 ¼ ¬ 0
1
0
0
0
1
0
0
0
0
1
0
−2
−3
−2
6
0
0
1
0
• Solutions
121
−11º
3»»
0 17 / 2 »
»
−4¼
1
0
0
ª 38 º
38º
« 16 »
»
16»
».
, so x = «
0 17 / 2»
«17 / 2 »
»
«
»
−4¼
1
¬ −4 ¼
0
0
0º
ª 1º
»
« −2»
12 »
, b = « » . First, solve Ly = b:
« −1»
0»
»
« »
1¼
¬ 2¼
1º ª 1
−2 »» ««0
∼
−1» «0
» «
2¼ ¬0
0
1
0
0
0
0
−3
4
1
−1
0
1
1º ª 1
0 »» «« 0
∼
−4 » «0
» «
7 ¼ ¬0
0
1
0
0
0
0
0
0
1
−1
0
1
1º
0»»
−4»
»
7¼
1º
ª1º
«0»
»
0»
, so y = « » . Next solve Ux = y, using back-substitution (with
« −4 »
−4 »
»
« »
3¼
¬3¼
1º ª 1
0 »» ««0
∼
−4 » « 0
» «
3¼ ¬ 0
3
3
2
0
0
0
0
0
−2
0
0
1
1º ª 1
−36 »» «« 0
∼
−4 » «0
» «
3¼ ¬0
3
3
0
0
0
0
0
0
1
0
0
1
ª 33 º
33º
« −12 »
−12 »»
».
, so x = «
2»
« 2 »
»
«
»
3¼
¬ 3 ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
−3º
−36 »»
2»
»
3¼
122
CHAPTER 2
• Matrix Algebra
5º
ª 2
7. Place the first pivot column of «
» into L, after dividing the column by 2 (the pivot), then add
¬ −3 −4 ¼
3/2 times row 1 to row 2, yielding U.
5º ª 2
5º
ª 2
A=«
~«
»
» =U
¬ −3 − 4 ¼ ¬ 0 7/2 ¼
ª 2º
« −3»
¬ ¼ [7 / 2 ]
÷2
÷7 / 2
ª 1
«−3 /2
¬
º
ª 1
, L=«
»
1¼
¬ −3 /2
0º
1 »¼
8. Row reduce A to echelon form using only row replacement operations. Then follow the algorithm in
Example 2 to find L.
ª6
A=«
¬12
4 º ª6
∼
5 »¼ «¬0
ª6º
«12 »
¬ ¼
4º
=U
−3»¼
[ −3]
÷6 ÷ − 3
ª 3
9. A = « −9
«
¬« 9
ª 3º
« −9 »
« »
«¬ 9 »¼
÷3
ª1
«2
¬
º
ª1
, L=«
»
1¼
¬2
1
2º ª 3
−4 »» ∼ «« 0
14 ¼» ¬« 0
0
9
ª3º
«6 »
¬ ¼
÷3
0º
1 »¼
1
3
6
2º ª 3
2 »» ~ «« 0
8 ¼» ¬« 0
1
3
0
2º
2 »» = U
4 ¼»
[ 4]
÷4
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.5
ª 1
« −3
«
«¬ 3
ª −5
10. A = « 10
«
«¬ 10
º
ª 1
» , L = « −3
»
«
«¬ 3
1»¼
1
2
4 º ª −5
−5 »» ~ «« 0
16 »¼ «¬ 0
0
2
10
0
2
10
123
0º
0 »»
1»¼
0
1
2
4 º ª −5
3»» ~ «« 0
24 »¼ «¬ 0
• Solutions
0
2
0
4º
3»» = U
9 »¼
ª −5º
«10 »
« » ª2º
«¬10 »¼ «¬10 »¼ [9]
÷ −5 ÷ 2 ÷9
ª 1
«
« −2
«¬ −2
ª 3
11. A = « 6
«
«¬ −3
1
5
2º ª 3
4 »» ∼ «« 0
3»¼ «¬ 0
7
19
−2
ª 3º
« 6»
« »
«¬ −3 »¼
÷3
ª 1
« 2
«
«¬−1
º
ª 1
» , L = « −2
»
«
«¬ −2
1»¼
7
5
5
0
1
5
2º ª 3
0 »» ∼ «« 0
5 »¼ «¬ 0
0º
0 »»
1»¼
7
5
0
2º
0 »» = U
5 »¼
ª5º
«5»
¬ ¼ [5]
÷5
÷5
1
1
º
ª 1
», L = « 2
»
«
«¬−1
1»¼
0
1
1
0º
0 »»
1»¼
12. Row reduce A to echelon form using only row replacement operations. Then follow the algorithm in
Example 2 to find L. Use the last column of I3 to make L unit lower triangular.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
124
CHAPTER 2
• Matrix Algebra
ª 2
A = «« 4
¬« −6
3
13
5
ª2º
«4»
« »
«¬ −6 »¼
÷2
ª 1
« 2
«
¬« −3
2º ª2
9 »» ∼ «« 0
4 ¼» ¬« 0
2º ª 2
5»» ∼ «« 0
10 ¼» ¬« 0
3
7
14
3
7
0
2º
5»» = U
0 ¼»
ª7º
«14 »
¬ ¼
÷7
1
2
º
ª 1
», L = « 2
»
«
1¼»
¬« −3
0
1
2
0º
0 »»
1¼»
3 −5 −3º ª 1 3 −5 −3º
ª 1 3 −5 −3º ª 1
« −1 −5 8 4 » «0 −2 3 1» « 0 −2 3 1»
»∼«
»∼«
» = U No more pivots!
13. «
« 4 2 −5 −7 » «0 −10 15 5» «0 0 0 0 »
«
» «
» «
»
2 −3 −1¼ ¬ 0 0 0 0¼
¬ −2 −4 7 5 ¼ ¬0
ª 1º
« −1»
« »
« 4»
« »
¬ −2 ¼
÷1
ª −2 º
« −10 »
«
»
¬« 2 ¼» Use the last two columns of I 4 to make L unit lower triangular.
÷ –2
ª 1
º
ª 1 0
« −1 1
»
«
«
» , L = « −1 1
« 4 5 1 »
« 4 5
«
»
«
«¬ −2 −1 0 1»¼
«¬ −2 −1
ª 1 3 1 5º ª 1 3
« 5 20 6 31» «0 5
»∼«
14. A = «
« −2 −1 −1 −4 » «0 5
«
» «
¬ −1 7 1 7 ¼ ¬0 10
0 0º
0 0 »»
1 0»
»
0 1»¼
5º ª 1
6 »» «« 0
∼
1 6» «0
» «
2 12 ¼ ¬ 0
1
1
3 1 5º
5 1 6»»
=U
0 0 0»
»
0 0 0¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.5
ª1º
«5»
« »
« −2 »
« »
¬ −1¼
÷1
ª 1
« 5
«
« −2
«
¬ −1
ª5º
«5»
« »
«¬10 »¼ Use the last two columns of I 4 to make L unit lower triangular.
÷5
1
1
2
1
0
º
ª 1
»
«
», L = « 5
»
« −2
»
«
1¼
¬ −1
0
0
1
0
1
2
1
0
0º
0 »»
0»
»
1¼
5 2º ª 2 0 5 2º ª 2 0 5 2º
ª 2 0
«
15. A = −6 3 −13 −3» ∼ « 0 3 2 3» ∼ « 0 3 2 3» = U
«
» «
» «
»
«¬ 4 9 16 17 »¼ «¬ 0 9 6 13»¼ «¬ 0 0 0 4 »¼
ª 2º
« −6 »
« »
«¬ 4 »¼
÷2
ª 3º
«9 » 4
¬ ¼ [ ]
÷3 ÷4
ª 1
« −3
«
«¬ 2
1
3
ª 2
« −4
«
16. A = « 6
«
« −6
«¬ 8
−3
8
−5
9
−6
ª 2º
« »
« −4 »
« 6»
« »
« −6 »
«¬ 8»¼
º
ª 1
» , L = « −3
»
«
«¬ 2
1»¼
4º ª 2
−7 »» «« 0
14 » ~ « 0
» «
−12 » « 0
19 »¼ «¬ 0
−3
2
4
0
6
0
1
3
0º
0»»
1»¼
4º ª2
1»» «« 0
2» ~ «0
» «
0» «0
3»¼ «¬ 0
−3
2
0
0
0
4º
1»»
0» = U
»
0»
0 »¼
ª 2º
« 4»
« »
« 0»
« »
¬ 6 ¼ Use the last three columns of I 5 to make L unit lower triangular.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
• Solutions
125
126
CHAPTER 2
• Matrix Algebra
÷2
÷2
ª 1
« −2
«
« 3
«
« −3
«¬ 4
ª 1
17. L = « −2
«
«¬ 0
reduce [L
1
2
0
3
1
0
0
1
1
so U
1
−1
0
0
1
0
1
0
0
1
1
1
0
0
0º ª 1
0 »» ∼ «« 0
1»¼ «¬ 0
−6
4
1
0
−4
8
0
1
0
−2
0
0
0
1
−3 / 2
1
0
0
−1 / 4
0
1
0
0
ª1 / 2
= «« 0
«¬ 0
0
0
1
0
1
0
2
1
0
1
−2
−1
−3 / 4
−1 / 4
0
0º
0 »» = [ I L−1 ],
1»¼
−3 / 4
−1 / 4
0
−4
0º
ª2
»
0 » , U = «« 0
«¬ 0
1»¼
1
−1
ª 1
«
I ] = « −2
¬« 3
−6
0
1
0
−4
0
0
1
0
−2
0
0
2º
4 »»
1»¼
0
0
1/ 2
−3 / 4
1
0
0
−1 / 4
0
1
0
0
−2 º
−1»» = [ I U −1 ],
−1 / 2 »¼
−2 º
−1»» . Thus
−1 / 2 »¼
− 1 −1
0
0º ª2
0 »» ∼ «« 0
1»¼ «¬ 0
−4 º ª 1
−1»» ∼ «« 0
−1 / 2 »¼ «¬ 0
ª1 / 2
A = U L = «« 0
«¬ 0
−1
[L
0º
0 »»
0»
»
0»
1»¼
0
0
0
1
0
4º
–1
8»» To find L , use the method of Section 2.2; that is, row
−2 »¼
−4
0
0
ª 1
18. L = « −2
«
«¬ 3
0
0
1
0
0
0º
–1
0 »» . Likewise to find U , row reduce [U I ] :
1¼»
0
ª2
[U I ] = «« 0
«¬ 0
−1
−6
0
1
2
0
3
I ]:
−1
ª2
«
∼ «0
«¬ 0
1
0
0º
ª2
»
0 » , U = «« 0
«¬ 0
1»¼
0
ª 1
[ L I ] = «« −2
«¬ 0
ª 1
so L = « 2
«
¬« −2
º
ª 1
»
« −2
»
«
», L = « 3
»
«
»
« −3
«¬ 4
1»¼
−3
0
0
0
1
0
1
0
0
1
−1
1
0
0
−2 º ª 1
−1»» «« 2
−1 / 2 »¼ «¬ −2
0
1
−1
0º ª 3
0 »» = ««3 / 2
1»¼ «¬ 1
5/4
3/ 4
1/ 2
−2 º
−1»»
−1 / 2 »¼
2º
6 »» To find L−1 , row reduce[ L I ] :
1»¼
0º ª 1
0 »» ~ «« 0
1¼» ¬« 0
0
0
1
0
1
0
2
1
−1
1
−3
0
0º
0 »»
1¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
ª1
~ «« 0
«¬ 0
0
0
1
0
1
0
2
1
0
1
−1
1
0º
0 »» = ª¬ I
1»¼
ª 1
L º¼ , so L = «« 2
«¬ −1
−1
Likewise tofind U −1 , row reduce [U
[U
ª2
I ] = «« 0
«¬ 0
ª2
«
~ «0
«¬ 0
−4
2
1
0
−3
6
0
1
0
1
0
0
0
0
1
−4 / 3
1
0
0
−1 / 3
0
1
0
0
ª1 / 2
= «« 0
«¬ 0
1
1
−4
0
1
0
−3
0
0
1
0
1
0
0
0
0
1/ 2
−2 / 3
1
0
0
−1 / 3
0
1
0
0
−2 º ª 2
−6 »» ~ «« 0
1»¼ «¬ 0
0
1
0
1
0
0
−1 / 3
0
1
0
0
−2 º
2 »»
1»¼
3º
−1
2 »» = [ I U ],
1»¼
−2 / 3
3º
so U
−1 / 3 2 »» . Thus
0 1»¼
ª1 / 2 −2 / 3
−1
− 1 −1
A = U L = «« 0 −1 / 3
«¬ 0
0
−1
−4
127
I ]:
0º ª2
0 »» ~ «« 0
1»¼ «¬ 0
6º ª 1
2 »» ~ «« 0
1»¼ «¬ 0
• Solutions
0º
0 »» .
1»¼
0
−1
2.5
3º ª 1
2 »» «« 2
1»¼ «¬ −1
0
1
1
0 º ª −23 / 6
0 »» = «« −8 / 3
1»¼ «¬
−1
7/3
5/3
1
3º
2 »»
1»¼
19. Let A be a lower-triangular n × n matrix with nonzero entries on the diagonal, and consider the
augmented matrix [A I].
a. The (1, 1)-entry can be scaled to 1 and the entries below it can be changed to 0 by adding
multiples of row 1 to the rows below. This affects only the first column of A and the first column
of I. So the (2, 2)-entry in the new matrix is still nonzero and now is the only nonzero entry of
row 2 in the first n columns (because A was lower triangular). The (2, 2)-entry can be scaled to
1, the entries below it can be changed to 0 by adding multiples of row 2 to the rows below. This
affects only columns 2 and n + 2 of the augmented matrix. Now the (3, 3) entry in A is the only
nonzero entry of the third row in the first n columns, so it can be scaled to 1 and then used as a
pivot to zero out entries below it. Continuing in this way, A is eventually reduced to I, by scaling
each row with a pivot and then using only row operations that add multiples of the pivot row to
rows below.
b. The row operations just described only add rows to rows below, so the I on the right in [A I]
changes into a lower triangular matrix. By Theorem 7 in Section 2.2, that matrix is A–1.
20. Let A = LU be an LU factorization for A. Since L is unit lower triangular, it is invertible by Exercise
19. Thus by the Invertible Matrix Theroem, L may be row reduced to I. But L is unit lower triangular,
so it can be row reduced to I by adding suitable multiples of a row to the rows below it, beginning
with the top row. Note that all of the described row operations done to L are row-replacement
operations. If elementary matrices E1, E2, … Ep implement these row-replacement operations, then
E p ... E2 E1 A = ( E p ... E2 E1 ) LU = IU = U
This shows that A may be row reduced to U using only row-replacement operations.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
128
CHAPTER 2
• Matrix Algebra
21. (Solution in Study Guide.) Suppose A = BC, with B invertible. Then there exist elementary matrices
E1, …, Ep corresponding to row operations that reduce B to I, in the sense that Ep … E1B = I.
Applying the same sequence of row operations to A amounts to left-multiplying A by the product Ep
… E1. By associativity of matrix multiplication.
E p ... E1 A = E p ... E1BC = IC = C
so the same sequence of row operations reduces A to C.
22. First find an LU factorization for A. Row reduce A to echelon form using only row replacement
operations:
−4
−9
−7
−2
3
ª 2
«
« 6
A=« 2
«
« 4
«¬ −6
ª2
«0
«
~ «0
«
«0
«¬ 0
−4
3
0
0
0
−2
−5
−3
−2
3
−2
1
0
0
0
3º ª 2
8»» «« 0
9» ~ « 0
» «
−1» « 0
4 »¼ «¬ 0
−4
3
−3
6
−9
−2
1
−1
2
−3
3º ª 2
−1»» «« 0
6» ~ « 0
» «
−7 » « 0
13»¼ «¬ 0
−4
3
0
0
0
−2
1
0
0
0
3º
−1»»
5»
»
−5»
10»¼
3º
−1»»
5» = U
»
0»
0 »¼
then follow the algorithm in Example 2 to find L. Use the last two columns of I5 to make L unit lower
triangular.
ª 2º
« 6»
« »
« 2»
« »
« 4»
«¬ −6 »¼
÷2
ª 1
« 3
«
« 1
«
« 2
«¬ −3
ª 3º
« −3»
« »
« 6»
« »
¬ −9 ¼
÷3
1
−1
2
−3
ª 5º
« −5»
« »
«¬ 10 »¼
÷5
1
−1
2
1
0
º
ª 1
»
« 3
»
«
», L = « 1
»
«
»
« 2
«¬ −3
1»¼
0
1
−1
2
−3
0
0
1
−1
2
0
0
0
1
0
0º
0 »»
0»
»
0»
1»¼
Now notice that the bottom two rows of U contain only zeros. If one uses the row-column method to
find LU, the entries in the final two columns of L will not be used, since these entries will be
multiplied by zeros from the bottom two rows of U. So let B be the first three columns of L and let C
be the top three rows of U. That is,
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.5
ª 1
«
« 3
B=« 1
«
« 2
«¬ −3
0º
0 »»
ª2
1» , C = «« 0
»
«¬ 0
−1»
»
2¼
0
1
−1
2
−3
−4
3
−2
1
0
0
• Solutions
129
3º
−1»»
5»¼
Then B and C have the desired sizes and BC = LU = A. We can generalize this process to the case
where A is m × n, A = LU, and U has only three non-zero rows: let B be the first three columns of L
and let C be the top three rows of U.
23. a. Express each row of D as the transpose of a column vector. Then use the multiplication rule for
partitioned matrices to write
A = CD = [ c1
c2
c3
ª d1T
« T
«d 2
c4 ] « T
« d3
« T
¬«d 4
º
»
»
T
T
T
T
» = c1d1 + c 2d 2 + c3d3 + c 4d4
»
»
¼»
which is the sum of four outer products.
b. Since A has 400 × 100 = 40000 entries, C has 400 × 4 = 1600 entries and D has 4 × 100 = 400
entries, to store C and D together requires only 2000 entries, which is 5% of the amount of entries
needed to store A directly.
24. Since Q is square and QTQ = I, Q is invertible by the Invertible Matrix Theorem and Q–1 = QT. Thus
A is the product of invertible matrices and hence is invertible. Thus by Theorem 5, the equation
Ax = b has a unique solution for all b. From Ax = b, we have QRx = b, QTQRx = QTb, Rx = QTb, and
finally x = R–1QTb. A good algorithm for finding x is to compute QTb and then row reduce the matrix
[ R QTb ]. See Exercise 11 in Section 2.2 for details on why this process works. The reduction is fast
in this case because R is a triangular matrix.
25. A = UDV T . Since U and V T are square, the equations U T U = I and V T V = I imply that U and V T
are invertible, by the IMT, and hence U – 1 = U T and (VT)–1 = V. Since the diagonal entries σ 1 ,! , σ n
in D are nonzero, D is invertible, with the inverse of D being the diagonal matrix with σ1−1 ,!, σ −n 1 on
the diagonal. Thus A is a product of invertible matrices. By Theorem 6, A is invertible and A–1 =
(UDV T ) –1 = (VT)–1D–1U–1 = VD–1UT.
26. If A = PDP–1, where P is an invertible 3 × 3 matrix and D is the diagonal matrix
ª2
D = «« 0
«¬ 0
0
3
0
0º
0 »»
1»¼
then
A2 = ( PDP −1 )( PDP −1 ) = PD( P −1P) DP −1 = PDIDP −1 = PD 2 P −1
and since
ª2
D = «« 0
«¬ 0
2
0
3
0
0º ª2
0 »» «« 0
1»¼ «¬ 0
0
3
0
0º ª4
0 »» = «« 0
1»¼ «¬ 0
0
9
0
0º
ª4
»
2
0 » , A = P «« 0
«¬ 0
1»¼
0
9
0
0º
0 »» P −1
1»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
130
CHAPTER 2
• Matrix Algebra
Likewise, A3 = PD3P –1, so
ª 23
«
A3 = P « 0
«
¬« 0
0º
ª8
» −1
0 » P = P ««0
»
«¬0
13 ¼»
0
3
3
0
0
27
0
0º
0 »» P −1
1»¼
In general, Ak = PDkP–1, so
ª2k
0 0º
«
»
Ak = P « 0 3k 0 » P −1
« 0 0 1»
«¬
»¼
27. First consider using a series circuit with resistance R1 followed by a shunt circuit with resistance R2
for the network. The transfer matrix for this network is
ª 1
« −1/R
2
¬
− R1 º ª 1
=
1 »¼ «¬ −1/R2
0º ª1
1 »¼ «¬ 0
− R1
º
( R1 + R2 ) / R2 »¼
For an input of 12 volts and 6 amps to produce an output of 9 volts and 4 amps, the transfer matrix
must satisfy
ª 1
« −1/R
2
¬
− R1
12 − 6 R1
º ª12 º ª
º ª9 º
=«
=
»
«
»
( R1 + R2 )/ R2 ¼ ¬ 6 ¼ ¬ ( −12 + 6 R1 + 6 R2 )/ R2 »¼ «¬ 4 »¼
Equate the top entries and obtain R1 = 12 ohm. Substitute this value in the bottom entry and solve to
obtain R2 = 92 ohms. The ladder network is
a.
i1
v1
i2
1/ 2 ohm
v2
i2
i3
9/ 2
ohms
v3
Next consider using a shunt circuit with resistance R1 followed by a series circuit with resistance R2
for the network. The transfer matrix for this network is
ª1
«0
¬
− R2 º ª 1
1 »¼ ¬« −1/R1
0 º ª ( R1 + R2 )/ R1
=
1 ¼» ¬« −1/R1
− R2 º
1 ¼»
For an input of 12 volts and 6 amps to produce an output of 9 volts and 4 amps, the transfer matrix
must satisfy
ª( R1 + R2 ) /R1
« −1/R
1
¬
− R2 º ª12 º ª (12 R1 + 12 R2 ) /R1 − 6 R2 º ª 9 º
=
» = «4»
1 »¼ «¬ 6 »¼ «¬
−12 /R1 + 6
¼ ¬ ¼
Equate the bottom entries and obtain R1 = 6 ohms. Substitute this value in the top entry and solve to
obtain R2 = 34 ohms. The ladder network is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.5
b.
i1
i2
6
ohms
v1
i2
v2
• Solutions
131
i3
3/4 ohm
v3
28. The three shunt circuits have transfer matrices
ª 1
« −1/R
1
¬
0º ª 1
,
1 »¼ «¬ −1/ R2
0º
, and
1 »¼
ª 1
« −1/ R
3
¬
0º
1 »¼
respectively. To find the transfer matrix for the series of circuits, multiply these matrices
ª 1
«
¬ −1 /R3
0º
1 ¼»
ª 1
«
¬ −1 / R2
0º
1 ¼»
ª 1
«
¬ −1 /R1
1
0º ª
=«
»
1 ¼ ¬ −(1 /R1 + 1 /R2 + 1 /R3 )
Thus the resulting network is itself a shunt circuit with resistance
0º
1 ¼»
R1 R2 R3
R1R2 + R1R3 + R2 R3
ª1 − R1 º
29. a. The first circuit is a series circuit with resistance R1 ohms, so its transfer matrix is «
.
1 »¼
¬0
The second circuit is a shunt circuit with resistance R2 ohms, so its transfer matrix is
0º
ª 1
. The third circuit is a series circuit with resistance R3 ohms so its transfer matrix is
« −1 / R
1 »¼
¬
2
ª1 − R3 º
.The transfer matrix of the network is the product of these matrices, in right-to-left
«0
1 »¼
¬
order:
−R1 º
ª1 + R3 / R2 − R1 −R3 − R1 R3 / R2 º
== «
»
»
1 + R1 / R2
1 ¼
¬ −1 / R2
¼
b. To find a ladder network with a structure like that in part (a) and with the given transfer matrix A,
we must find resistances R1, R2, and R3 such that
ª1
«0
¬
− R3 º ª 1
1 »¼ «¬ −1 / R2
ª 3
A=«
¬ −1 / 3
0 º ª1
1 »¼ «¬ 0
−12 º ª1 + R3 / R2
=
5 / 3»¼ «¬ −1 / R2
− R1 −R3 − R1 R3 / R2 º
»
1 + R1 / R2
¼
From the (2, 1) entries, R2 = 3 ohms. The (1, 1) entries now give 1 + R3 /R2 = 3, which may be
solved to obtain R3 = 6 ohms. Likewise the (2, 2) entries gives 1 + R1 /R2 = 5 / 3, which also may
be solved to obtain R1 = 2 ohms. Thus the matrix A may be factored as
ª1
A=«
¬0
ª1
=«
¬0
− R3 º ª 1
1 »¼ «¬ −1 / R2
−6 º ª 1
1 »¼ «¬ −1 / 3
−R1 º
1 »¼
0 º ª1
1 »¼ «¬ 0
0 º ª1
1 »¼ «¬ 0
−2 º
1 »¼
ª 3
¬−1/ 3
30. Answers may vary. For example, «
−12 º ª1
=
5 / 3»¼ «¬0
−6º ª 1
1 »¼ «¬ −1/ 6
0º ª 1
1»¼ «¬ −1/ 6
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0º ª1
1»¼ «¬0
−2º
1 »¼
132
CHAPTER 2
• Matrix Algebra
The network corresponding to this factorization consists of a series circuit, followed by two shunts,
followed by another series circuit. The resistances would be R1=2, R2=6, R3=6, R4=6.
Note: The Study Guide’s MATLAB box for Section 2.5 suggests that for most LU factorizations in this
section, students can use the gauss command repeatedly to produce U, and use paper and mental
arithmetic to write down the columns of L as the row reduction to U proceeds. This is because for
Exercises 7–16 the pivots are integers and other entries are simple fractions. However, for Exercises 31
and 32 this is not reasonable, and students are expected to solve an elementary programming problem.
(The Study Guide provides no hints.)
31. [M] Store the matrix A in a temporary matrix B and create L initially as the 8×8 identity matrix. The
following sequence of MATLAB commands fills in the entries of L below the diagonal, one column
at a time, until the first seven columns are filled. (The eighth column is the final column of the
identity matrix.)
L(2:8, 1) = B(2:8, 1)/B(1, 1)
B = gauss(B, 1)
L(3:8, 2) = B(3:8, 2)/B(2, 2)
B = gauss(B, 2)
#
L(8:8, 7) = B(8:8, 7)/B(7, 7)
U = gauss(B,7)
Of course, some students may realize that a loop will speed up the process. The for..end syntax
is illustrated in the MATLAB box for Section 5.6. Here is a MATLAB program that includes the
initial setup of B and L:
B = A
L = eye(8)
for j=1:7
L(j+1:8, j) = B(j+1:8, j)/B(j, j)
B = gauss(B, j)
end
U = B
a. To four decimal places, the results of the LU decomposition are
ª 1
« −.25
«
« −.25
«
0
L=«
« 0
«
« 0
« 0
«
¬« 0
0
1
−.0667
−.2667
0
0
0
0
0
0
1
−.2857
−.2679
0
0
0
0
0
0
1
−.0833
−.2917
0
0
0
0
0
0
1
−.2921
−.2697
0
0
0
0
0
0
1
−.0861
−.2948
0
0
0
0
0
0
1
−.2931
0º
0 »»
0»
»
0»
0»
»
0»
0»
»
1 ¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.5
ª4
«0
«
«0
«
0
U =«
«0
«
«0
«0
«
«¬ 0
−1
3.75
0
0
0
0
0
0
−1
−.25
3.7333
0
0
0
0
0
0
−1
−1.0667
3.4286
0
0
0
0
0
0
−1
−.2857
3.7083
0
0
0
0
0
0
−1
−1.0833
3.3919
0
0
0
0
0
0
−1
−.2921
3.7052
0
• Solutions
133
0 º
0 »»
0 »
»
0 »
0 »
»
−1 »
−1.0861»
»
3.3868 »¼
b. The result of solving Ly = b and then Ux = y is
x = (27.1292, 19.2344, 29.2823, 19.8086, 30.1914, 20.7177, 30.7656, 22.8708)
ª.2953
«.0866
«
«.0945
«
.0509
−1
c. A = «
«.0318
«
«.0227
«.0010
«
«¬.0082
ª 3
« −1
32. [M] A = «
« 0
«
¬ 0
produce
ª1
«− 1
L=« 3
«0
«
¬« 0
ª3
«0
U =«
«0
«
«¬0
.0866
.2953
.0509
.0945
.0227
.0318
.0082
.0100
−1
3
−1
0
0
1
− 83
0
1
8
− 21
−1
0
8
3
−1
0
21
8
0
0
.0509
.0945
.1093
.3271
.0591
.1045
.0227
.0318
.0318
.0227
.1045
.0591
.3271
.1093
.0945
.0509
.0227
.0318
.0591
.1045
.1093
.3271
.0509
.0945
.0010
.0082
.0318
.0227
.0945
.0509
.2953
.0866
.0082 º
.0100 »»
.0227 »
»
.0318 »
.0509 »
»
.0945 »
.0866 »
»
.2953»¼
0º
0»»
. The commands shown for Exercise 31, but modified for 4×4 matrices,
−1»
»
3¼
0
−1
3
−1
0
0
.0945
.0509
.3271
.1093
.1045
.0591
.0318
.0227
0º
0 »»
0»
»
1 ¼»
0º
0 »»
−1»
55 »
¼
21 »
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.4
• Solutions
134
b. Let sk+1 be the solution of Lsk+1 = tk for k = 0, 1, 2, …. Then tk+1 is the solution of Utk+1 = sk+1
for k = 0, 1, 2, …. The results are
ª10.0000 º
ª 7.0000º
ª 7.0000º
ª5.0000º
«18.3333 »
«11.0000»
«13.3333»
«8.0000»
«
»
«
»
«
»
»,
s1 =
,t =
,s =
,t = «
« 21.8750» 1 «11.0000» 2 «16.0000» 2 «8.0000»
«
»
«
»
«
»
«
»
¬18.3333 ¼
¬ 7.0000¼
¬13.0952¼
¬5.0000¼
ª 5.0000 º
ª3.6000º
ª 3.6000º
ª 2.6000 º
« 9.6667 »
«5.8000»
« 7.0000»
«
»
» ,t = «
» ,s = «
» , t = « 4.2000 » .
s3 = «
«11.6250 » 3 «5.8000» 4 « 8.4250» 4 « 4.2000 »
«
»
«
»
«
»
«
»
¬ 9.4286 ¼
¬3.6000¼
¬ 6.8095¼
¬ 2.6000 ¼
2.6
SOLUTIONS
Notes: This section is independent of Section 1.10. The material here makes a good backdrop for the
series expansion of (I–C)–1 because this formula is actually used in some practical economic work.
Exercise 8 gives an interpretation to entries of an inverse matrix that could be stated without the economic
context.
1. The answer to this exercise will depend upon the order in which the student chooses to list the
sectors. The important fact to remember is that each column is the unit consumption vector for the
appropriate sector. If we order the sectors manufacturing, agriculture, and services, then the
consumption matrix is
ª.10
C = ««.30
«¬.30
.60 º
.20
0 »»
.10 .10 »¼
The intermediate demands created by the production vector x are given by Cx. Thus in this case the
intermediate demand is
ª.10 .60 .60 º ª 0 º ª 60 º
Cx = ««.30 .20 .00 »» ««100 »» = «« 20 »»
«¬.30 .10 .10 »¼ «¬ 0 »¼ «¬10 »¼
.60
2. Solve the equation x = Cx + d for d:
ª x1 º ª.10 .60 .60 º ª x1 º ª .9 x1
«
d = x − C x = « x2 »» − ««.30 .20 .00 »» «« x2 »» = «« −.3 x1
¬« x3 ¼» ¬«.30 .10 .10 »¼ ¬« x3 ¼» ¬« −.3 x1
This system of equations has the augmented matrix
ª .90
« −.30
«
«¬ −.30
−.60
−.60
.80
.00
−.10
.90
0 º ª1
20 »» ~ «« 0
0 »¼ «¬0
0
0
1
0
0
1
−.6 x2
+.8 x2
−.1x2
37.03º
38.89 »» , so x =
16.67 »¼
−.6 x3 º ª 0 º
» = « 20 »
» « »
+.9 x3 ¼» «¬ 0 »¼
ª 37.03 º
«38.89 » .
«
»
«¬16.67 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.6
• Solutions
3. Solving as in Exercise 2:
ª x1 º ª.10 .60 .60 º ª x1 º ª .9 x1
d = x − C x = «« x2 »» − ««.30 .20 .00 »» «« x2 »» = «« −.3 x1
¬« x3 ¼» ¬«.30 .10 .10 ¼» «¬ x3 »¼ «¬ −.3 x1
This system of equations has the augmented matrix
ª .90
« −.30
«
«¬ −.30
−.60
−.60
.80
.00
−.10
.90
20 º ª1
0 »» ~ «« 0
0 »¼ «¬ 0
0
0
1
0
0
1
−.6 x2
+.8 x2
−.1x2
−.6 x3 º ª 20 º
»=«0»
» « »
+.9 x3 »¼ ¬« 0 ¼»
44.44 º
ª 44.44 º
«16.67 » .
»
,
so
x
=
16.67 »
«
»
«¬16.67 »¼
16.67 »¼
4. Solving as in Exercise 2:
ª x1 º ª.10
«
d = x − Cx = « x2 »» − ««.30
¬« x3 ¼» ¬«.30
.60
.20
.10
.60 º ª x1 º ª .9 x1
.00 »» «« x2 »» = «« −.3 x1
.10 »¼ «¬ x3 »¼ «¬ −.3 x1
−.6 x2
+.8 x2
−.1x2
−.6 x3 º ª 20 º
» = « 20 »
» « »
+.9 x3 »¼ «¬ 0 »¼
This system of equations has the augmented matrix
ª .90
« −.30
«
«¬ −.30
−.60
−.60
.80
.00
−.10
.90
20 º ª1
20 »» ~ «« 0
0 »¼ «¬ 0
0
0
1
0
0
1
81.48 º
ª81.48 º
»
55.56 » , so x = ««55.56 »» .
«¬ 33.33 »¼
33.33»¼
Note: Exercises 2–4 may be used by students to discover the linearity of the Leontief model.
−1
ª 1
5. x = ( I − C ) d = «
¬ −.6
−.5º ª50º ª1.6
=
.8»¼ «¬30»¼ «¬1.2
ª .8
6. x = ( I − C )−1 d = «
¬ −.6
−.5º ª16º
ª.9
1
=
»
«
»
.9¼ ¬12¼ (.72 − .30) «¬.6
−1
1 º ª50º ª110º
=
2»¼ «¬30»¼ «¬120»¼
−1
.5º ª16º ª 48.57º
=
.8»¼ «¬12»¼ «¬ 45.71»¼
7. a. From Exercise 5,
ª1.6
( I − C ) −1 = «
¬1.2
1º
2 »¼
so
ª1.6
x1 = ( I − C ) −1 d1 = «
¬1.2
1 º ª1 º ª1.6 º
=
2 »¼ «¬0 »¼ «¬1.2 »¼
which is the first column of ( I − C )−1.
b.
ª1.6
x 2 = ( I − C ) −1 d 2 = «
¬1.2
1 º ª 51º ª111.6 º
=
2 »¼ «¬30 »¼ «¬121.2 »¼
ª50 º
ª110 º
c. From Exercise 5, the production x corressponding to d = « » is x = « » .
¬30 ¼
¬120 ¼
Note that d 2 = d + d1 . Thus
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
135
136
CHAPTER 2
• Matrix Algebra
x 2 = ( I − C ) −1 d 2
= ( I − C )−1 (d + d1 )
= ( I − C )−1 d + ( I − C ) −1 d1
= x + x1
8. a. Given ( I − C )x = d and ( I − C )∆ x = ∆d,
( I − C )(x + ∆x) = ( I − C )x + ( I − C )∆ x = d + ∆ d
Thus x + ∆ x is the production level corresponding to a demand of d + ∆ d .
b. Since ∆ x = ( I − C )−1 ∆ d and ∆ d is the first column of I, ∆x will be the first column of
( I − C )−1.
9. In this case
ª .8
I − C = «« −.3
¬« −.1
Row reduce
ª .8
« −.3
«
«¬ −.1
−.2
.9
.0
.0 º
−.3»»
.8 ¼»
[ I − C d ] to find
−.2
.0
.9
−.3
.0
.8
40.0 º ª 1
60.0 »» ~ «« 0
80.0 »¼ «¬ 0
0
1
0
82.8º
0 131.0 »»
1 110.3»¼
0
So x = (82.8, 131.0, 110.3).
10. From Exercise 8, the (i, j) entry in (I – C)–1 corresponds to the effect on production of sector i when
the final demand for the output of sector j increases by one unit. Since these entries are all positive,
an increase in the final demand for any sector will cause the production of all sectors to increase.
Thus an increase in the demand for any sector will lead to an increase in the demand for all sectors.
11. (Solution in study Guide) Following the hint in the text, compute pTx in two ways. First, take the
transpose of both sides of the price equation, p = C Tp + v, to obtain
pT = (CT p + v)T = (CT p)T + vT = pT C + vT
and right-multiply by x to get
pTx = (pT C + vT )x = pT Cx + vT x
Another way to compute pTx starts with the production equation x = Cx + d. Left multiply by pT to
get
pT x = pT (Cx + d) = pT C x + pT d
The two expression for pTx show that
pT C x + vT x = pT C x + pT d
so vTx = pTd. The Study Guide also provides a slightly different solution.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.6
• Solutions
12. Since
Dm+1 = I + C + C 2 + ... + C m+1 = I + C ( I + C + ... + C m ) = I + CDm
Dm +1 may be found iteratively by Dm +1 = I + CDm .
13. [M] The matrix I – C is
ª 0.8412
« −0.0057
«
« −0.0264
«
« −0.3299
« −0.0089
«
« −0.1190
« −0.0063
¬
−0.0064
0.7355
−0.0025
−0.0436
−0.0304
−0.0099
−0.0014
−0.0083
−0.0083
−0.0201
−0.1506
−0.0565
−0.0081
−0.0901
0.6443
−0.0495
−0.0333
−0.0996
−0.0139
0.6364
−0.0295
−0.1260
−0.0142
−0.0204
0.6588
−0.1722
−0.0070
−0.0483
−0.0237
0.7632
−0.0126
−0.0196
−0.0098
−0.0064
−0.0132
−0.1594 º
−0.3413»»
−0.0236 »
»
−0.0649 »
−0.0020 »
»
−0.3369 »
0.9988»¼
so the augmented matrix [ I − C d ] may be row reduced to find
ª 0.8412
« −0.0057
«
« −0.0264
«
« −0.3299
« −0.0089
«
« −0.1190
« −0.0063
¬
−0.0064
0.7355
−0.0025
−0.0436
−0.0304
−0.0099
−0.0014
−0.0083
−0.0083
−0.0201
−0.1594
−0.3413
−0.1506
−0.0565
−0.0081
−0.0901
0.6443
−0.0495
−0.0333
−0.0996
−0.0139
0.6364
−0.0295
−0.1260
−0.0142
−0.0204
0.6588
−0.1722
−0.0070
−0.0483
−0.0237
0.7632
−0.0236
−0.0649
−0.0020
−0.3369
−0.0126
−0.0196
−0.0098
−0.0064
−0.0132
0.9988
ª1
«0
«
«0
«
~ «0
«0
«
«0
«0
¬
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
1
0
0
0
0
1
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
74000 º
56000 »»
10500 »
»
25000 »
17500 »
»
196000 »
5000 »¼
99576 º
97703»»
51231»
»
131570 »
49488»
»
329554 »
13835»¼
so x = (99576, 97703, 51231, 131570, 49488, 329554, 13835). Since the entries in d seem to be
accurate to the nearest thousand, a more realistic answer would be x = (100000, 98000, 51000,
132000, 49000, 330000, 14000).
14. [M] The augmented matrix [ I − C d ] in this case may be row reduced to find
ª 0.8412
« −0.0057
«
« −0.0264
«
« −0.3299
« −0.0089
«
« −0.1190
« −0.0063
¬
−0.0064
0.7355
−0.0025
−0.0436
−0.0304
−0.0099
−0.0014
−0.0083
−0.0083
−0.0201
−0.1594
−0.3413
−0.1506
−0.0565
−0.0081
−0.0901
0.6443
−0.0495
−0.0333
−0.0996
−0.0139
0.6364
−0.0295
−0.1260
−0.0142
−0.0204
0.6588
−0.1722
−0.0070
−0.0483
−0.0237
0.7632
−0.0236
−0.0649
−0.0020
−0.3369
−0.0126
−0.0196
−0.0098
−0.0064
−0.0132
0.9988
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
99640 º
75548»»
14444 »
»
33501»
23527 »
»
263985»
6526 »¼
137
138
CHAPTER 2
• Matrix Algebra
ª1
«0
«
«0
«
~ «0
«0
«
«0
«0
¬
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
1
0
0
0
0
1
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
134034 º
131687 »»
69472 »
»
176912 »
66596 »
»
443773»
18431»¼
so x = (134034, 131687, 69472, 176912, 66596, 443773, 18431). To the nearest thousand, x =
(134000, 132000, 69000, 177000, 67000, 444000, 18000).
15. [M] Here are the iterations rounded to the nearest tenth:
x(0) = (74000.0, 56000.0, 10500.0, 25000.0, 17500.0, 196000.0, 5000.0)
x(1) = (89344.2, 77730.5, 26708.1, 72334.7, 30325.6, 265158.2, 9327.8)
x(2) = (94681.2, 87714.5, 37577.3, 100520.5, 38598.0, 296563.8, 11480.0)
x (3) = (97091.9, 92573.1, 43867.8, 115457.0, 43491.0, 312319.0, 12598.8)
x(4) = (98291.6, 95033.2, 47314.5, 123202.5, 46247.0, 320502.4, 13185.5)
x (5) = (98907.2, 96305.3, 49160.6, 127213.7, 47756.4, 324796.1, 13493.8)
x(6) = (99226.6, 96969.6, 50139.6, 129296.7, 48569.3, 327053.8, 13655.9)
x(7) = (99393.1, 97317.8, 50656.4, 130381.6, 49002.8, 328240.9, 13741.1)
x (8) = (99480.0, 97500.7, 50928.7, 130948.0, 49232.5, 328864.7, 13785.9)
x (9) = (99525.5, 97596.8, 51071.9, 131244.1, 49353.8, 329192.3, 13809.4)
x (10) = (99549.4, 97647.2, 51147.2, 131399.2, 49417.7, 329364.4, 13821.7)
x (11) = (99561.9, 97673.7, 51186.8, 131480.4, 49451.3, 329454.7, 13828.2)
x (12) = (99568.4, 97687.6, 51207.5, 131523.0, 49469.0, 329502.1, 13831.6)
so x(12) is the first vector whose entries are accurate to the nearest thousand. The calculation of x(12)
takes about 1260 flops, while the row reduction above takes about 550 flops. If C is larger than
20 × 20, then fewer flops are required to compute x(12) by iteration than by row reduction. The
advantage of the iterative method increases with the size of C. The matrix C also becomes more
sparse for larger models, so fewer iterations are needed for good accuracy.
2.7
SOLUTIONS
Notes: The content of this section seems to have universal appeal with students. It also provides practice
with composition of linear transformations. The case study for Chapter 2 concerns computer graphics –
see this case study (available as a project on the website) for more examples of computer graphics in
action. The Study Guide encourages the student to examine the book by Foley referenced in the text. This
section could form the beginning of an independent study on computer graphics with an interested
student.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.7
• Solutions
139
1. Refer to Example 5. The representation in homogenous coordinates can be written as a partitioned
ª A 0º
matrix of the form « T
» , where A is the matrix of the linear transformation. Since in this case
1¼
¬0
ª 1 .25 º
A=«
, the representation of the transformation with respect to homogenous coordinates is
1»¼
¬0
ª1
«0
«
«¬ 0
0º
0 »»
1»¼
.25
1
0
0º
» on the vector
1¼
ªA
Note: The Study Guide shows the student why the action of «
T
¬0
ªx º
« 1 » corresponds to
¬ ¼
the action of A on x.
ª −1
2. The matrix of the transformation is A = «
¬ 0
ª −1
AD = «
¬ 0
0º ª4
1»¼ «¬ 0
2
2
5 º ª −4
=
3»¼ «¬ 0
−2
2
0º
, so the transformed data matrix is
1»¼
− 5º
3»¼
3. Following Examples 4–6,
ª0
«1
«
«¬ 0
ª1 / 2
4. « 0
«
«¬ 0
−1
0
0
0
3/ 2
0
0 º ª1
0 »» «« 0
1»¼ «¬ 0
0º ª 1
0 »» «« 0
1»¼ «¬ 0
ª 2/2
«
5. « 2 / 2
«
0
«¬
− 2/2
ª1
6. ««0
«¬0
0º ª 2 / 2
«
0 »» « 2 / 2
1»¼ ««
0
¬
0
−1
0
2/2
0
0
1
0
0
1
0
0º ª 1
»
0 » ««0
»
1» «¬0
¼
2º ª0
1 »» = «« 1
1 »¼ «¬ 0
−1º ª1 / 2
4 »» = «« 0
1»¼ «¬ 0
0
−1
0
− 2/2
2/2
0
−1
0
0
−1º
2 »»
1»¼
0
3/ 2
0
0º ª 2 / 2
«
0 »» = « 2 / 2
«
1»¼ «
0
¬
0º ª 2 / 2
» «
0» = «− 2 / 2
1»» ««
0
¼ ¬
−1 / 2 º
6 »»
1»¼
2/2
− 2/2
0
− 2/2
− 2/2
0
0º
»
0»
»
1»
¼
0º
»
0»
1»»
¼
7. A 60° rotation about the origin is given in homogeneous coordinates by the matrix
ª 1/ 2 − 3 / 2 0 º
«
»
1/ 2 0 » . To rotate about the point (6, 8), first translate by (–6, –8), then rotate about
« 3/2
«
»
0
0 1»
¬«
¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
140
CHAPTER 2
• Matrix Algebra
the origin, then translate back by (6, 8) (see the Practice Problem in this section). A 60° rotation
about (6, 8) is thus is given in homogeneous coordinates by the matrix
ª1
«0
«
«¬0
0
1
0
6 º ª 1/ 2
«
8 »» « 3 / 2
1 »¼ ««
0
¬
− 3/2
1/ 2
0
0º ª 1
»
0 » «« 0
1»» «¬0
¼
0
1
0
−6 º ª 1/ 2
«
−8»» = « 3 / 2
1»¼ ««
0
¬
− 3/2
1/ 2
0
3 + 4 3º
»
4 − 3 3»
1»»
¼
8. A 45° rotation about the origin is given in homogeneous coordinates by the matrix
ª 2 / 2 − 2 / 2 0º
«
»
2 / 2 0 » . To rotate about the point (3, 7), first translate by (–3, –7), then rotate about
« 2 /2
«
»
0
0 1»
¬«
¼
the origin, then translate back by (3, 7) (see the Practice Problem in this section). A 45° rotation
about (3, 7) is thus is given in homogeneous coordinates by the matrix
ª1
«0
«
«¬0
0
1
0
3º ª 2 / 2
«
7 »» « 2 / 2
1 »¼ ««
0
¬
− 2 /2
2/2
0
0º ª 1
»
0 » ««0
1»» «¬0
¼
0
1
0
−3º ª 2 / 2
«
−7 »» = « 2 / 2
1»¼ ««
0
¬
− 2/2
2/2
0
3 + 2 2º
»
7 − 5 2»
1»»
¼
9. To produce each entry in BD two multiplications are necessary. Since BD is a 2 × 100 matrix, it will
take 2 × 2 × 100 = 400 multiplications to compute BD. By the same reasoning it will take 2 × 2 × 100 =
400 multiplications to compute A(BD). Thus to compute A(BD) from the beginning will take
400 + 400 = 800 multiplications.
To compute the 2 × 2 matrix AB it will take 2 × 2 × 2 = 8 multiplications, and to compute
(AB)D it will take 2 × 2 × 100 = 400 multiplications. Thus to compute (AB)D from the beginning will
take 8 + 400 = 408 multiplications.
For computer graphics calculations that require applying multiple transformations to data
matrices, it is thus more efficient to compute the product of the transformation matrices before
applying the result to the data matrix.
10. Let the transformation matrices in homogeneous coordinates for the dilation, rotation, and translation
be called respectively D, and R, and T. Then for some value of s, ϕ, h, and k,
ªs
D = «« 0
¬« 0
0º
ªcos ϕ
»
0 » , R = «« sin ϕ
1 ¼»
¬« 0
0
s
0
− sin ϕ
cos ϕ
0
0º
ª1
»
0 » , T = «« 0
«¬ 0
1 ¼»
0
1
0
hº
k »»
1 ¼»
Compute the products of these matrices:
ª s cos ϕ
DR = «« s sin ϕ
«¬ 0
ªs
DT = «« 0
«¬ 0
0
s
0
− s sin ϕ
s cos ϕ
0
0º
ª s cos ϕ
»
0 » , RD = «« s sin ϕ
«¬ 0
1 »¼
sh º
ªs
»
sk » , TD = «« 0
«¬ 0
1 »¼
0
s
0
− s sin ϕ
s cos ϕ
0
0º
0 »»
1 »¼
hº
k »»
1 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.7
ª cos ϕ
«
RT = « sin ϕ
¬« 0
− sin ϕ
h cos ϕ − k sin ϕ º
ª cos ϕ
»
«
h sin ϕ + k cos ϕ » , TR = « sin ϕ
1
¼»
¬« 0
cos ϕ
0
− sin ϕ
cos ϕ
0
• Solutions
141
hº
k »»
1 ¼»
Since DR = RD, DT ≠ TD and RT ≠ TR, D and R commute, D and T do not commute and R and T do
not commute.
11. To simplify A 2 A 1 completely, the following trigonometric identities will be needed:
sin ϕ
1. − tan ϕ cos ϕ = − cos
cos ϕ = − sin ϕ
ϕ
sin ϕ
sin ϕ
ϕ
sin ϕ = 1−cos
= cos
= cos ϕ
2. sec ϕ − tan ϕ sin ϕ = cos1 ϕ − cos
ϕ
ϕ
cos ϕ
2
2
Using these identities,
ªsec ϕ
A2 A1 = «« 0
«¬ 0
− tan ϕ
1
0
0º ª 1
0 »» ««sin ϕ
1 »¼ «¬ 0
ªsec ϕ − tan ϕ sin ϕ
= ««
sin ϕ
«¬
0
ªcos ϕ
= «« sin ϕ
«¬ 0
− sin ϕ
cos ϕ
0
0º
0 »»
1 »¼
0
cos ϕ
0
− tan ϕ cos ϕ
cos ϕ
0
0º
0 »»
1 »¼
0º
0 »»
1 »¼
which is the transformation matrix in homogeneous coordinates for a rotation in
2
.
12. To simplify this product completely, the following trigonometric identity will be needed:
tan ϕ / 2 =
1 − cos ϕ
sin ϕ
=
sin ϕ
1 + cos ϕ
This identity has two important consequences:
1 − (tan ϕ / 2)(sin ϕ ) = 1 −
1 − cos ϕ
sin ϕ = cos ϕ
sin ϕ
(cos ϕ )(− tan ϕ / 2) − tan ϕ / 2 = −(cos ϕ + 1) tan ϕ / 2 = −(cos ϕ + 1)
sin ϕ
= − sin ϕ
1 + cos ϕ
The product may be computed and simplified using these results:
ª1
«0
«
«¬ 0
− tan ϕ / 2
1
0
0º ª 1
0 »» ««sin ϕ
1 »¼ «¬ 0
ª1 − (tan ϕ / 2)(sin ϕ )
= ««
sin ϕ
«¬
0
0
1
0
0 º ª1
0 »» «« 0
1 »¼ «¬ 0
− tan ϕ / 2
1
0
− tan ϕ / 2
0 º ª1
0 »» ««0
1 »¼ «¬0
1
0
0º
0 »»
1 »¼
− tan ϕ / 2
1
0
0º
0 »»
1 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
142
CHAPTER 2
• Matrix Algebra
ª cos ϕ
= «« sin ϕ
«¬ 0
ª cos ϕ
= «« sin ϕ
«¬ 0
− tan ϕ / 2
0 º ª1 − tan ϕ / 2 0 º
1
0 »» «« 0
1
0 »»
0
1 »¼ «¬ 0
0
1 »¼
(cos ϕ )( − tan ϕ / 2) − tan ϕ / 2 0 º
−(sin ϕ )(tan ϕ / 2) + 1
0 »»
0
1 »¼
ªcos ϕ
= «« sin ϕ
«¬ 0
− sin ϕ
cos ϕ
0
0º
0 »»
1 »¼
which is the transformation matrix in homogeneous coordinates for a rotation in
2
.
13. Consider first applying the linear transformation on 2 whose matrix is A, then applying a translation
by the vector p to the result. The matrix representation in homogeneous coordinates of the linear
ª A 0º
transformation is « T
» , while the matrix representation in homogeneous coordinates of the
1¼
¬0
pº
ªI
translation is « T
» . Applying these transformations in order leads to a transformation whose
1¼
¬0
matrix representation in homogeneous coordinates is
ª I p º ª A 0º ª A
« T
»«
»=«
1 ¼ ¬0T 1 ¼ ¬0T
¬0
which is the desired matrix.
pº
»
1¼
14. The matrix for the transformation in Exercise 7 was found to be
ª 1/ 2
«
« 3/2
«
0
«¬
− 3/2
1/ 2
0
3 + 4 3º
»
4 − 3 3»
»
1»
¼
ªA
This matrix is of the form « T
¬0
ª 1/ 2
A=«
«¬ 3 / 2
pº
» , where
1¼
ª3 + 4 3 º
− 3 / 2º
» ,p = «
»
1/ 2 »¼
¬« 4 − 3 3 ¼»
By Exercise 13, this matrix may be written as
ª I p º ª A 0º
« T
»«
»
1 ¼ ¬0 T 1 ¼
¬0
that is, the composition of a linear transformation on 2 and a translation. The matrix A is the matrix
of a rotation of 60 about the origin in 2. Thus the transformation in Exercise 7 is the composition of
ª3 + 4 3 º
a rotation about the origin and a translation by p = «
».
¬« 4 − 3 3 ¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.7
1
), the corresponding point in
15. Since ( X , Y , Z , H ) = ( 12 , − 41 , − 81 , 24
3
• Solutions
143
has coordinates
1
§ X Y Z · § 1 −1 − ·
( x, y, z ) = ¨ , , ¸ = ¨¨ 12 , 14 , 18 ¸¸ = (12, −6, −3)
© H H H ¹ © 24 24 24 ¹
16. The homogeneous coordinates (1, –2, −3, 4) represent the point
(1/ 4, − 2 / 4, − 3 / 4) = (1/ 4, − 1/ 2, − 3 / 4)
while the homogeneous coordinates (10, –20,− 30, 40) represent the point
(10 / 40, − 20 / 40, − 30 / 40) = (1/ 4, − 1/ 2, − 3 / 4)
so the two sets of homogeneous coordinates represent the same point in
3
.
17. Follow Example 7a by first constructing that 3 × 3 matrix for this rotation. The vector e1 is not
changed by this rotation. The vector e2 is rotated 60° toward the positive z-axis, ending up at the
point (0, cos 60°, sin 60°) = (0, 1/ 2, 3 / 2). The vector e3 is rotated 60° toward the negative y-axis,
stopping at the point
(0, cos 150°, sin 150°) = (0, − 3 / 2, 1/ 2). The matrix A for this rotation is thus
ª1
«
A = «0
«0
¬
0
1/ 2
3/2
0º
»
− 3 / 2»
1/ 2 »¼
so in homogeneous coordinates the transformation is represented by the matrix
ªA
« T
¬0
ª1
«
0 º «0
»=
1 ¼ «0
«
«¬ 0
0
0
1/ 2
− 3/2
3/2
0
1/ 2
0
0º
»
0»
0 »»
1»¼
18. First construct the 3 × 3 matrix for the rotation. The vector e1 is rotated 30° toward the negative yaxis, ending up at the point (cos(–30)°, sin (–30)°, 0) = ( 3 / 2, −1/ 2, 0). The vector e2 is rotated 30°
toward the positive x-axis, ending up at the point (cos 60°, sin 60°, 0) = (1/ 2, 3 / 2, 0). The vector e3
is not changed by the rotation. The matrix A for the rotation is thus
ª 3/2
1/ 2 0 º
«
»
A = « −1/ 2
3 / 2 0»
«
»
0
0 1»
«¬
¼
so in homogeneous coordinates the rotation is represented by the matrix
ªA
« T
¬0
ª 3/2
«
0 º « −1/ 2
=
»
1¼ «
0
«
«¬
0
1/ 2
0
3/2
0
0
0
1
0
0º
»
0»
0 »»
1»¼
Following Example 7b, in homogeneous coordinates the translation by the vector (5, –2, 1) is
represented by the matrix
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
144
CHAPTER 2
ª1
«0
«
«0
«
¬«0
• Matrix Algebra
0
1
0
0
0
0
1
0
5º
−2 »»
1»
»
1¼»
Thus the complete transformation is represented in homogeneous coordinates by the matrix
ª1
«0
«
«0
«
¬« 0
0
1
0
0
5º ª 3 / 2
«
−2 »» « −1/ 2
1» «
0
»«
1¼» «¬
0
0
0
1
0
1/ 2
0
3/2
0
0
0
1
0
0º ª 3 / 2
» «
0 » « −1/ 2
=
0 »» ««
0
1»¼ «¬
0
1/ 2
0
3/2
0
0
0
1
0
5º
»
−2 »
1»»
1»¼
19. Referring to the material preceding Example 8 in the text, we find that the matrix P that performs a
perspective projection with center of projection (0, 0, 10) is
ª1
«0
«
«0
«
¬«0
0
1
0
0
0
0
0
−.1
0º
0 »»
0»
»
1¼»
The homogeneous coordinates of the vertices of the triangle may be written as (4.2, 1.2, 4, 1), (6, 4,
2, 1), and (2, 2, 6, 1), so the data matrix for S is
ª 4.2 6 2 º
« 1.2 4 2 »
«
»
« 4 2 6»
«
»
«¬ 1 1 1»¼
and the data matrix for the transformed triangle is
ª1
«0
«
«0
«
¬«0
0 º ª 4.2 6 2 º ª 4.2 6 2º
0»» «« 1.2 4 2»» «« 1.2 4 2»»
=
0
0 0» « 4 2 6» « 0 0 0»
»«
» «
»
0 −.1 1¼» «¬ 1 1 1»¼ «¬ .6 .8 .4»¼
Finally, the columns of this matrix may be converted from homogeneous coordinates by dividing by
the final coordinate:
(4.2, 1.2, 0, .6) → (4.2 / .6, 1.2 / .6, 0 / .6) = (7, 2, 0)
(6, 4, 0, .8)
→ (6/.8, 4/.8, 0/.8) = (7.5, 5, 0)
(2, 2, 0, .4)
→ (2 / .4, 2 / .4, 0 / .4) = (5, 5, 0)
0
1
0
0
So the coordinates of the vertices of the transformed triangle are (7, 2, 0), (7.5, 5, 0), and (5, 5, 0).
20. As in the previous exercise, the matrix P that performs the perspective projection is
ª1
«0
«
«0
«
«¬0
0
1
0
0
0
0
0
−.1
0º
0 »»
0»
»
1»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
s 3OLUTIONS
The homogeneous coordinates of the vertices of the triangle may be written as (7, 3, –5, 1), (12, 8, 2,
1), and (1, 2, 1, 1), so the data matrix for S is
ª 7 12 1º
« 3
8 2»»
«
« −5
2 1»
«
»
1 1¼
¬ 1
and the data matrix for the transformed triangle is
ª1
«0
«
«0
«
¬0
0º ª 7 12 1º ª 7 12 1º
0»» «« 3 8 2»» «« 3 8 2»»
=
0
0 0» « −5 2 1» « 0
0 0»
»«
» «
»
0 −.1 1¼ ¬ 1
1 1¼ ¬1.5 .8 .9¼
Finally, the columns of this matrix may be converted from homogeneous coordinates by dividing by
the final coordinate:
(7, 3, 0, 1.5) → (7 / 1.5,3 / 1.5,0 / 1.5) = (4.67, 2,0)
(12, 8, 0, .8) → (12 / .8,8 / .8,0 / .8) = (15,10,0)
(1, 2, 0, .9) → (1 / .9, 2 / .9, 0 / .9) = (1.11, 2.22, 0)
0
1
0
0
So the coordinates of the vertices of the transformed triangle are (4.67, 2, 0), (15, 10, 0),
and (1.11, 2.22, 0).
21. [M] Solve the given equation for the vector (R, G, B), giving
ª R º ª .61
« » «
«G » = «.35
¬« B ¼» ¬«.04
.29
.59
.12
.15º
.063»»
.787 »¼
−1
ª X º ª 2.2586
« Y » = « −1.3495
« » «
«¬ Z »¼ «¬ .0910
−1.0395
2.3441
−.3046
−.3473º ª X º
.0696 »» «« Y »»
1.2777 ¼» ¬« Z ¼»
22. [M] Solve the given equation for the vector (R, G, B), giving
ª R º ª.299
« » «
«G » = «.596
¬« B ¼» ¬«.212
2.8
.587
−.275
−.528
.114 º
−.321»»
.311¼»
−1
ª Y º ª1.0031
« » «
« I » = « .9968
¬«Q ¼» ¬«1.0085
.9548
−.2707
−1.1105
.6179º ªY º
−.6448»» «« I »»
1.6996 »¼ «¬Q »¼
SOLUTIONS
Notes: Cover this section only if you plan to skip most or all of Chapter 4. This section and the next
cover everything you need from Sections 4.1–4.6 to discuss the topics in Section 4.9 and Chapters 5–7
(except for the general inner product spaces in Sections 6.7 and 6.8). Students may use Section 4.2 for
review, particularly the Table near the end of the section. (The final subsection on linear transformations
should be omitted.) Example 6 and the associated exercises are critical for work with eigenspaces in
Chapters 5 and 7. Exercises 31–36 review the Invertible Matrix Theorem. New statements will be added
to this theorem in Section 2.9.
Key Exercises: 5–20 and 23–26.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
146
CHAPTER 2
• Matrix Algebra
1. The set is closed under sums but not under multiplication
by a negative scalar. A counterexample to the subspace
condition is shown at the right.
u
(–1)u
Note: Most students prefer to give a geometric counterexample, but some may choose an algebraic calculation. The four exercises here should help students develop an understanding of subspaces, but they may
be insufficient if you want students to be able to analyze an unfamiliar set on an exam. Developing that
skill seems more appropriate for classes covering Sections 4.1–4.6.
2.The set is closed under scalar multiples but not sums.
For example, the sum of the vectors in H shown
here is not in H.
3. No. The set is not closed under sums or scalar multiples. See the diagram.
4. No. The set is closed under sums, but not under multiplication by a
negative scalar.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.8
• Solutions
147
5. The vector w is in the subspace generated by v1 and v2 if and only if the vector equation x1v1 + x2v2
= w is consistent. The row operations below show that w is in the subspace generated by v1 and v2.
[ v1 v 2
−2
ª 1
w ] ~ «« 3
«¬ −4
−3 º ª 1
−3»» ~ «« 0
10 »¼ «¬ 0
−3
7
−2
−3 º ª 1
6 »» ~ «« 0
−2 »¼ «¬ 0
3
−1
−2
3
0
−3º
6 »»
0 »¼
6. The vector u is in the subspace generated by {v1, v2, v3} if and only if the vector equation x1v1 + x2v2
+ x3v3 = u is consistent. The row operations below show that u is not in the subspace generated by
{v1, v2, v3}.
[ v1 v 2
ª1
«0
«
«0
«
¬0
ª 1
« −3
v 3 u] ~ «
« 2
«
¬ 3
4
1
5
2
0
0
−4
2
4
−4
5
−3
5
7
6
5
−1º ª 1
−1»» «« 0
−2 » « 0
» «
−2 ¼ ¬ 0
−1º ª 1
−7 »» ««0
~
−1» «0
» «
2 ¼ ¬0
4
1
5
2
0
0
1
−4
4
8
5
12
−3
−5
−4
−10
−1º ª 1
1»» «« 0
−1» « 0
» «
−2¼ ¬ 0
4
1
5
2
0
0
1
0
−1º ª 1
−10 »» ««0
~
1» «0
» «
5¼ ¬0
4
1
5
2
8
−3
12
−4
−1º
−1»»
−10»
»
1¼
−1º
1»»
−1»
»
−6 ¼
Note: For a quiz, you could use w = (9, –7, 11, 12), which is in Span{v1, v2, v3}.
7. a. There are three vectors: v1, v2, and v3 in the set {v1, v2, v3}.
b. There are infinitely many vectors in Span{v1, v2, v3} = Col A.
c. Deciding whether p is in Col A requires calculation:
ª 2
«
[ A p ] ~ « −8
¬« 6
−3
−4
8
6
−7
−7
6º ª2
−10 »» ~ «« 0
11¼» ¬« 0
−3
−4
−4
−10
2
5
−3
−4
−4
−10
0
0
−2
0
3
−5
0
0
−6 º
1»»
0 »¼
6º ª 2
14 »» ~ «« 0
−7 ¼» ¬« 0
6º
14 »»
0 ¼»
The equation Ax = p has a solution, so p is in Col A.
ª −2
«
8. [ A p ] = 0
«
«¬ 6
−2
0
3
−5
3
5
−6 º ª −2
1»» ~ «« 0
17 »¼ ¬« 0
−2
0
3
−5
−3
5
−6 º ª −2
1»» ~ «« 0
−1¼» «¬ 0
Yes, the augmented matrix [A p] corresponds to a consistent system, so p is in Col A.
9. To determine whether p is in Nul A, simply compute Ap. Using A and p as in Exercise 7,
ª 2
Ap = « −8
«
«¬ 6
−3
8
−7
−4 º ª 6 º ª −2 º
6 »» «« −10 »» = «« −62 »» . Since Ap ≠ 0, p is not in Nul A.
−7 »¼ «¬ 11»¼ «¬ 29 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
148
CHAPTER 2
• Matrix Algebra
10. To determine whether u is in Nul A, simply compute Au. Using A as in Exercise 7 and u = (–5, 5, 3),
ª −2
Au = « 0
«
«¬ 6
−2
3
3
0 º ª −5 º ª 0 º
−5 »» «« 5»» = «« 0 »» . Yes, u is in Nul A.
5 »¼ «¬ 3»¼ «¬ 0 »¼
11. p = 4 and q = 3. Nul A is a subspace of R4 because solutions of Ax = 0 must have 4 entries, to match
the columns of A. Col A is a subspace of R3 because each column vector has 3 entries.
12. p = 3 and q = 5. Nul A is a subspace of R3 because solutions of Ax = 0 must have 3 entries, to match
the columns of A. Col A is a subspace of R5 because each column vector has 5 entries.
13. To produce a vector in Col A, select any column of A. For Nul A, solve the equation Ax = 0. (Include
an augmented column of zeros, to avoid errors.)
ª 3
« −9
«
«¬ 9
ª3
~ ««0
«¬0
2
1
−5
−4
1
7
2
−5
1
2
1
−5
1
2
−4
0
0
0
0º ª 3
0 »» ~ ««0
0 »¼ «¬0
0º ª 1
0 »» ~ ««0
0 »¼ «¬0
2
1
−5
2
4
−8
−4
−8
16
0
−1
1
1
2
−4
0
0
0
0º ª 3
0 »» ~ «« 0
0 »¼ «¬ 0
x1
0º
0 »» ,
0 »¼
2
1
−5
2
4
−8
0
0
0
0º
0 »»
0 »¼
− x3 + x4 = 0
x2 + 2 x3 − 4 x4 = 0
0 = 0
The general solution is x1 = x3 – x4, and x2 = –2x3 + 4x4, with x3 and x4 free. The general solution in
parametric vector form is not needed. All that is required here is one nonzero vector. So choose any
values for x3 and x4 (not both zero). For instance, set x3 = 1 and x4 = 0 to obtain the vector (1, –2, 1,
0) in Nul A.
Note: Section 2.8 of Study Guide introduces the ref command (or rref, depending on the technology), which produces the reduced echelon form of a matrix. This will greatly speed up homework for
students who have a matrix program available.
14. To produce a vector in Col A, select any column of A. For Nul A, solve the equation Ax = 0:
ª 1
« 4
«
« −5
«
« 2
«¬ 3
2
5
3
7
−1 0
7 11
3
4
0º ª 1
0 »» ««0
0 » ~ «0
» «
0 » «0
0 »¼ «¬0
2
−3
3
−5
9
3
−3
15
5
−5
0º ª 1
«
0 »» « 0
0» ~ «0
» «
0» «0
0 »¼ «¬ 0
2
3
1
0
5/3
0
0
0
0
0
0º ª 1
0 »» « 0
«
0» ~ «0
» «
0» «0
0 »¼ «¬ 0
0
1
−1 / 3
5/3
0
0
0
0
0
0
0º
0 »»
0»
»
0»
0 »¼
The general solution is x1 = (1/3)x3 and x2 = (–5/3) x3, with x3 free. The general solution in parametric
vector form is not needed. All that is required here is one nonzero vector. So choose any nonzero
value of x3. For instance, set x3 = 3 to obtain the vector (1, –5, 3) in Nul A.
15. Yes. Let A be the matrix whose columns are the vectors given. Then A is invertible because its
determinant is nonzero, and so its columns form a basis for R2, by the Invertible Matrix Theorem (or
by Example 5). (Other reasons for the invertibility of A could be given.)
16. No. One vector is a multiple of the other, so they are linearly dependent and hence cannot be a basis
for any subspace.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.8
• Solutions
149
17. Yes. Place the three vectors into a 3×3 matrix A and determine whether A is invertible:
ª 0
A = «« 0
«¬ −2
6 º ª −2
3»» ~ «« 0
2 »¼ «¬ 0
5
0
4
2º
6 »»
3»¼
4
5
0
The matrix A has three pivots, so A is invertible by the IMT and its columns form a basis for R3 (as
pointed out in Example 5).
18. No. Place the three vectors into a 3×3 matrix A and determine whether A is invertible:
ª 1
A = «« 1
«¬ −3
3
−1
2
5º ª 1
1»» ~ «« 0
−4 »¼ «¬ 0
3
−4
11
5º ª 1
−4 »» ~ «« 0
11»¼ «¬ 0
3
−4
0
5º
−4 »»
0 »¼
The matrix A has two pivots, so A is not invertible by the IMT and its columns do not form a basis
for R3 (as pointed out in Example 5).
19. No. The vectors cannot be a basis for R3 because they only span a plan in R3. Or, point out that the
6º
ª 3
3
«
columns of the matrix −8
2 »» cannot possibly span R because the matrix cannot have a pivot in
«
«¬ 1 −5 »¼
every row. So the columns are not a basis for R3.
Note: The Study Guide warns students not to say that the two vectors here are a basis for R2.
20. No. The vectors are linearly dependent because there are more vectors in the set than entries in each
vector. (Theorem 8 in Section 1.7.) So the vectors cannot be a basis for any subspace.
21. a. False. See the definition at the beginning of the section. The critical phrases “for each” are
missing.
b. True. See the paragraph before Example 4.
c. False. See Theorem 12. The null space is a subspace of Rn, not Rm.
d. True. See Example 5.
e. True. See the first part of the solution of Example 8.
22. a. False. See the definition at the beginning of the section. The condition about the zero vector is
only one of the conditions for a subspace.
b. False. See the warning that follows Theorem 13.
c. True. See Example 3.
d. False. Since y need not be in H, it is not gauranteed by the definition of a subspace that x+y will
be in H.
e. False. See the paragraph after Example 4.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
150
CHAPTER 2
• Matrix Algebra
−5 º
5 1
1 5 −6 »» . The echelon form identifies
4 8
0 0
0 »¼
ª 4º ª 5º
columns 1 and 2 as the pivot columns. A basis for Col A uses columns 1 and 2 of A: « 6 » , « 5 » . This
« » « »
«¬ 3»¼ «¬ 4 »¼
ª4
23. (Solution in Study Guide) A = « 6
«
«¬ 3
5
9
−2 º ª 1
12 »» ~ ««0
−3»¼ «¬ 0
2
6
is not the only choice, but it is the “standard” choice. A wrong choice is to select columns 1 and 2 of
the echelon form. These columns have zero in the third entry and could not possibly generate the
columns displayed in A.
For Nul A, obtain the reduced (and augmented) echelon form for Ax = 0:
x1
− 4 x3 + 7 x4 = 0
7 0º
ª 1 0 −4
«0
»
x2 + 5 x3 − 6 x4 = 0 .
1
5 −6 0 » . This corresponds to:
«
«¬ 0 0
0 = 0
0
0 0 »¼
Solve for the basic variables and write the solution of Ax = 0 in parametric vector form:
ª x1 º ª 4 x3 − 7 x4 º
ª 4º
ª −7 º
« x » « −5 x + 6 x »
« −5»
« 6»
3
4»
« 2» = «
= x3 « » + x4 « » . Basis for Nul A:
« x3 » «
»
« 1»
« 0»
x3
« » «
»
« »
« »
x4
¬« 0 ¼»
¬« 1¼»
¬« x4 ¼» ¬«
¼»
ª 4 º ª −7 º
« −5 » « 6 »
« », « »
« 1» « 0 »
« » « »
¬« 0 ¼» ¬« 1¼»
Notes: (1) A basis is a set of vectors. For simplicity, the answers here and in the text list the vectors
without enclosing the list inside set brackets. This style is also easier for students. I am careful,
however, to distinguish between a matrix and the set or list whose elements are the columns of the
matrix.
(2) Recall from Chapter 1 that students are encouraged to use the augmented matrix when solving Ax
= 0, to avoid the common error of misinterpreting the reduced echelon form of A as itself the augmented
matrix for a nonhomogeneous system.
(3) Because the concept of a basis is just being introduced, I insist that my students write the
parametric vector form of the solution of Ax = 0. They see how the basis vectors span the solution space
and are obviously linearly independent. A shortcut, which some instructors might introduce later in the
course, is only to solve for the basic variables and to produce each basis vector one at a time. Namely, set
all free variables equal to zero except for one free variable, and set that variable equal to a suitable
nonzero number.
ª3
«
24. A = 2
«
¬« 3
−6
9
−4
7
−6
6
0º ª 1
2 »» ~ «« 0
−6 ¼» ¬« 0
−2
5
0
3
0
0
4º
6 »» . Basis for Col A:
0 ¼»
ª 3º ª 9 º
«2» , «7 » .
« » « »
¬« 3¼» ¬« 6 ¼»
For Nul A, obtain the reduced (and augmented) echelon form for Ax = 0:
ª1
«0
«
«¬0
−2
0
−6
0
1
2
0
0
0
x1 − 2 x2
0º
»
0 » . This corresponds to:
0 »¼
− 6 x4 = 0
x3 + 2 x4 = 0 .
0 = 0
Solve for the basic variables and write the solution of Ax = 0 in parametric vector form:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.8
ª x1 º ª 2 x2 − 6 x4 º
ª 2º
ª −6 º
«x » «
»
«
»
« »
x2
« 2» = «
» = x « 1» + x « 0 » . Basis for Nul A:
« x3 » « −2 x4 » 2 « 0 » 4 « −2 »
« » «
»
« »
« »
x4
¬ 0¼
¬ 1¼
¬ x4 ¼ ¬
¼
ª 1
« −1
25. A = «
« −2
«
«¬ 3
4
2
8
7
−3
3
2
6
9
9
5
−5
−7 º ª 1
4 »» ««0
~
5» «0
» «
−2 »¼ «¬0
4
2
8
5
0
0
0
0
0
0
1
0
• Solutions
151
ª 2 º ª −6 º
« 1» « 0»
« », « »
« 0 » « −2 »
« » « »
¬ 0 ¼ ¬ 1¼
5º
−1»»
. Basis for Col A:
4»
»
0 »¼
ª 1º ª
« −1» «
« », «
« −2 » «
« » «
«¬ 3»¼ «¬
4º ª −3º
2»» «« 3»»
,
.
2 » « 5»
» « »
6»¼ «¬ −5»¼
For Nul A, obtain the reduced (and augmented) echelon form for Ax = 0:
ª1
«0
[ A 0] ~ «
«0
«
¬«0
0
1
−2
2.5
0
0
7
−.5
0
0
0
0
1
0
4
0
0º
0»»
.
0»
»
0¼»
x1
− 2 x3
x2 + 2.5 x3
+ 7 x5 =
− .5 x5 =
x4 + 4 x5 =
0 =
0
0
.
0
0
ª x1 º ª 2 x3 − 7 x5 º
ª 2 º
ª −7 º
« x » « −2.5 x + .5 x »
« −2.5»
« .5»
3
5»
« 2» «
«
»
« »
» = x3 « 1 » + x5 « 0 » .
Thesolution of Ax = 0in parametric vector form : « x3 » = «
x3
« » «
»
«
»
« »
−4 x5
« x4 » «
»
« 0 »
« −4 »
«¬ x5 »¼ «¬
»¼
«¬ 0 »¼
«¬ 1»¼
x5
u
v
Basis for Nul A: {u, v}.
Note: The solution above illustrates how students could write a solution on an exam, when time is
precious, namely, describe the basis by giving names to appropriate vectors found in the calculations.
ª3
«3
26. A = «
«0
«
¬6
−1
1
3
3
−3
3
9
9
−1
0
−1
−2
8º ª 3
2»» ««0
~
−4» «0
» «
6¼ ¬0
−1
2
0
0
−3
6
0
0
0
0
−1
0
6º
−4 »»
. Basis for Col A:
2»
»
0¼
ª 3º ª −1º ª −1º
« 3» « 1» « 0»
« », « », « » .
«0» « 3» « −1»
« » « » « »
¬6¼ ¬ 3¼ ¬ −2¼
For Nul A,
ª1
«0
[ A 0] ~ ««
0
«
¬0
0
1
0
0
0
3
0
0
0
0
1
0
4/3
−2
−2
0
0º
0»»
.
0»
»
0¼
x1
+ 4 / 3x5
2 x5
x2 + 3x3
−
2 x5
x4 −
0
=
=
=
=
0
0
0
0
The solution of Ax = 0 in parametric vector form:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
152
CHAPTER 2
• Matrix Algebra
ª x1 º ª −4 / 3x5 º
ª0º
ª −4 / 3º
« x » « −3 x + 2 x »
« −3»
«
2 »»
3
5»
« 2» «
« »
«
« x3 » = «
» = x3 « 1 » + x5 «
x3
0»
« » «
»
« »
«
»
2 » . Basis for Nul A: {u, v}.
« x4 » « 2 x5
»
«0»
«
«¬ x5 »¼ «¬
»¼
«¬ 0 »¼
«¬
x5
1»¼
u
v
27. Construct a nonzero 3×3 matrix A and construct b to be almost any convenient linear combination of
the columns of A.
28. The easiest construction is to write a 3×3 matrix in echelon form that has only 2 pivots, and let b be
any vector in R3 whose third entry is nonzero.
29. (Solution in Study Guide) A simple construction is to write any nonzero 3×3 matrix whose columns
are obviously linearly dependent, and then make b a vector of weights from a linear dependence
relation among the columns. For instance, if the first two columns of A are equal, then b could be (1,
–1, 0).
30. Since Col A is the set of all linear combinations of a1, … , ap, the set {a1, … , ap} spans Col A.
Because {a1, … , ap} is also linearly independent, it is a basis for Col A. (There is no need to discuss
pivot columns and Theorem 13, though a proof could be given using this information.)
31. If Col F ≠ R5, then the columns of F do not span R5. Since F is square, the IMT shows that F is not
invertible and the equation Fx = 0 has a nontrivial solution. That is, Nul F contains a nonzero vector.
Another way to describe this is to write Nul F ≠ {0}.
32. If Col B = R7, then the columns of B span R7. Since B is square, the IMT shows that B is invertible
and the equation Bx = b has a solution for each b in R7. Also, each solution is unique, by Theorem 5
in Section 2.2.
33. If Nul C = {0}, then the equation Cx = 0 has only the trivial solution. Since C is square, the IMT
shows that C is invertible and the equation Cx = b has a solution for each b in R6. Also, each solution
is unique, by Theorem 5 in Section 2.2.
34. If the columns of A form a basis, they are linearly independent. This means that A cannot have more
columns than rows. Since the columns also span Rm, A must have a pivot in each row, which means
that A cannot have more rows than columns. As a result, A must be a square matrix.
35. If Nul B contains nonzero vectors, then the equation Bx = 0 has nontrivial solutions. Since B is
square, the IMT shows that B is not invertible and the columns of B do not span R5. So Col B is a
subspace of R5, but Col B ≠ R5.
36. If the columns of C are linearly independent, then the equation Cx = 0 has only the trivial (zero)
solution. That is, Nul C = {0}.
37. [M] Use the command that produces the reduced echelon form in one step (ref or rref depending
on the program). See the Section 2.8 in the Study Guide for details. By Theorem 13, the pivot
columns of A form a basis for Col A.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.8
ª 3
« −7
A=«
« −5
«
¬« 3
−5
9
7
−7
0
−4
−2
−3
−1
9
5
4
3º ª 1
−11»» «« 0
~
−7 » «0
» «
0¼» ¬«0
0
1
0
0
−4.5
−2.5
0
0
2.5
1.5
0
0
• Solutions
3.5º
1.5»»
Basis for Col A:
0»
»
0 ¼»
ª 3º ª −5º
« −7 » « 9 »
« », « »
« −5 » « 7 »
« » « »
¬« 3¼» ¬« −7 ¼»
For Nul A, obtain the solution of Ax = 0 in parametric vector form:
x1
+ 2.5 x3 − 4.5 x4 + 3.5 x5 = 0
x2 + 1.5 x3 − 2.5 x4 + 1.5 x5 = 0
x1 = −2.5 x3 + 4.5 x4 − 3.5 x5
°
Solution: ® x2 = −1.5 x3 + 2.5 x4 − 1.5 x5
° x , x , and x are free
5
¯ 3 4
ª x1 º ª −2.5 x3 + 4.5 x4 − 3.5 x5 º
ª −2.5º
ª 4.5º
ª −3.5º
« x » « −1.5 x + 2.5 x − 1.5 x »
« −1.5»
« 2.5»
« −1.5»
3
4
5»
« 2» «
«
»
« »
«
»
» = x3 « 1 » + x4 « 0 » + x5 « 0 » = x3u + x4v + x5w
x = « x3 » = «
x3
« » «
»
«
»
« »
«
»
x4
« x4 » «
»
« 0 »
«1 »
« 0 »
«¬ x5 »¼ «¬
»¼
«¬ 0 »¼
«¬ 0 »¼
«¬ 1 »¼
x5
By the argument in Example 6, a basis for Nul A is {u, v, w}.
ª 5
« 4
«
A
=
38. [M]
« 5
«
¬ −7
3
1
2
3
−6
−8
1
−5
4
−2
5
8
−8º ª 1
−7 »» ««0
~
19 » « 0
» «
5¼ ¬ 0
0
1
1
−1
0
0
0
0
0
0
1
0
0º
0 »»
.
0»
»
1¼
ª 5º ª 3º ª − 6º ª −8 º
« 4 » « 1» « −8» « −7 »
»,« » .
The pivot columns of A form a basis for Col A: « » , « » , «
« 5» « 1» « 5» «19 »
« » « » «
» « »
¬ −7 ¼ ¬ −5¼ ¬ 8¼ ¬ 5 ¼
x1
For Nul A, solve Ax = 0:
x1
°x
° 2
° x3
°
Solution: ® x4
°x
° 5
°
°
¯
+ x3
= 0
x2 − x3
= 0
= 0
x5 = 0
x4
= − x3
= x3
is free
= 0
= 0
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
153
154
CHAPTER 2
• Matrix Algebra
ª x1 º
ª −1º
«x »
«1»
« 2»
« »
x = « x3 » = x3 « 1 » = x3 u
« »
« »
« x4 »
«0»
«¬ x5 »¼
«¬ 0 »¼
By the method of Example 6, a basis for Nul A is {u}.
Note: The Study Guide for Section 2.8 gives directions for students to construct a review sheet for the
concept of a subspace and the two main types of subspaces, Col A and Nul A, and a review sheet for the
concept of a basis. I encourage you to consider making this an assignment for your class.
2.9
SOLUTIONS
Notes: This section contains the ideas from Sections 4.4–4.6 that are needed for later work in Chapters
5–7. If you have time, you can enrich the geometric content of “coordinate systems” by discussing crystal
lattices (Example 3 and Exercises 35 and 36 in Section 4.4.) Some students might profit from reading
Examples 1–3 from Section 4.4 and Examples 2, 4, and 5 from Section 4.6. Section 4.5 is probably not a
good reference for students who have not considered general vector spaces.
Coordinate vectors are important mainly to give an intuitive and geometric feeling for the
isomorphism between a k-dimensional subspace and Rk. If you plan to omit Sections 5.4, 5.6, 5.7 and 7.2,
you can safely omit Exercises 1–8 here.
Exercises 1–16 may be assigned after students have read as far as Example 2. Exercises 19 and 20 use
the Rank Theorem, but they can also be assigned before the Rank Theorem is discussed.
The Rank Theorem in this section omits the nontrivial fact about Row A which is included in the
Rank Theorem of Section 4.6, but that is used only in Section 7.4. The row space itself can be introduced
in Section 6.2, for use in Chapter 6 and Section 7.4.
Exercises 9–16 include important review of techniques taught in Section 2.8 (and in Sections 1.2 and
2.5). They make good test questions because they require little arithmetic. My students need the practice
here. Nearly every time I teach the course and start Chapter 5, I find that at least one or two students
cannot find a basis for a two-dimensional eigenspace!
ª 3º
1. If [x]B = « » , then x is formed from b1 and b2 using
¬ 2¼
weights 3 and 2:
ª 1º
ª 2º ª
x = 3b1 + 2b2 = 3 « » + 2 « » = «
¬ 1¼
¬ −1¼ ¬
7º
1»¼
x2
3b 1
2b 1
b1
x
x1
b2
2b 2
ª −1º
2. If [x]B = « » , then x is formed from b1 and b2 using weights –1 and 2:
¬ 2¼
ª −3 º
ª 3º ª9 º
x = (–1)b1 + 2b2 = ( −1) « » + 2 « » = « »
¬ 1¼
¬ 2 ¼ ¬ 3¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.9
• Solu
utions
3. To find c1 and c2 that satisfy x = c1b1 + c2b2, row reduce the augmented matrix:
ª 2 −1 0 º ª 1 −1 / 2 0 º ª 1 0 1º
[b1 b 2 x ] = «
»~«
»~«
» . Or, one can write a matrix equation as
¬ −3 5 7 ¼ ¬ 0 7 / 2 7 ¼ ¬ 0 1 2 ¼
suggested by Exercise 1 and solve uusing the matrix inverse. In either case,
ª c º ª 1º
[x]B = « 1 » = « » .
¬ c2 ¼ ¬ 2 ¼
ª 1
4. As in Exercise 3, [b1 b 2 x] = «
¬ −5
−2
3
1º ª 1
~
9 »¼ «¬ 0
−2
−7
−2
2º ª 1
1»» ~ «« 0
−1»¼ «¬ 0
0
−3 º
,
−2 »¼
1º ª 1
~
14 »¼ «¬ 0
0
1
4º
1»» . [x]B =
0 »¼
ª c1 º ª 4 º
« c » = « 1» .
¬ 2¼ ¬ ¼
ª c º ª −3 º
and [x]B = « 1 » = « » .
¬ c2 ¼ ¬ − 2 ¼
5. [b1 b 2
ª 1
x ] = «« 4
«¬ −3
ª −3
6. [b1 b 2 x] = « 2
«
«¬ −4
−2
−7
5
7
−3
5
2º ª 1
9 »» ~ «« 0
−7 »¼ «¬ 0
1
−1
5º ª 1
1
»
«
0» ~ «0
1
»
«
−2 ¼ ¬ 0 13 / 3
1
0
5º ª 1
2 »» ~ «« 0
26 / 3»¼ «¬ 0
0
1
0
3º
2 »» ,
0 »¼
ª c º ª 3º
[x]B = « 1 » = « » .
¬ c2 ¼ ¬ 2 ¼
7. Fig. 1 suggests that w = 2b1 – b2 annd x = 1.5b1 + .5b2, in which case,
ª 2º
ª1.5 º
[w]B = « » and [x]B = « » . To cconfirm [x]B, compute
¬ −1¼
¬ .5 ¼
ª 3º
ª − 1º ª 4 º
1.5 b1 + .5 b 2 = 1.5 « » + .5 « » = « » = x
¬ 0¼
¬ 2 ¼ ¬1 ¼
b2
x
0
b1
w
Figure 1
Figure 2
Note: Figures 1 and 2 display what Secttion 4.4 calls B-graph paper.
8. Fig. 2 suggests that x = b1 + b2, y = 1/3b1 - b2, and z = –4/3 b1 +2b2. If so, then
ª1º
[x]B = « » , [y]B =
¬1¼
ª1 / 3 º
« − 1» , and [z]B =
¬
¼
ª −4 / 3 º
. To confirm [y]B and [z]B, compute
«
2 »¼
¬
Copyright © 2012 Pearrson Education, Inc. Publishing as Addison-Wesley.
155
156
CHAPTER 2
• Matrix Algebra
ª 0 º ª 2 º ª −2 º
ª 0º
ª 2º ª4 º
(1 / 3)b1 − b 2 = (1 / 3) « » − « » = « » = y and ( − 4 / 3)b1 + 2 b 2 = −4 / 3 « » + 2 « » = « » = z .
¬ 3 ¼ ¬ 2 ¼ ¬ −1 ¼
¬ 3¼
¬ 2¼ ¬ 0 ¼
ª 1 3 2 −6 º ª 1
« 3 9 1 5» « 0
»~«
9. The information A = «
« 2 6 −1 9 » « 0
«
» «
¬ 5 15 0 14¼ ¬ 0
ª 1º ª 2 º ª −6º
« 3» « 1» « 5»
of A form a basis for Col A: « » , « » , « » .
« 2 » « −1» « 9»
« » « » « »
¬ 5¼ ¬ 0 ¼ ¬ 14¼
3
0
0
0
3 2º
5 −7 »»
is enough to see that columns 1, 3, and 4
0
5»
»
0 0¼
Columns 1, 2 and 4, of the echelon form certainly cannot span Col A since those vectors all have zero
in their fourth entries. For Nul A, use the reduced echelon form, augmented with a zero column to
insure that the equation Ax = 0 is kept in mind:
ª1
«0
«
«0
«
¬0
3
0
0
1
0
0
0
0
0
0
1
0
x1 + 3x2
0º
0 »»
.
0»
»
0¼
= 0
x3
= 0
,
x4 = 0
x2 is the free variable
ª x1 º ª −3 x2 º
ª −3º
ª −3º
«x » « x »
«1»
«1»
2»
2 »
«
«
«
»
x
.
=
=
x=
So « » is
« x3 » « 0 » 2 « 0 »
«0»
« » «
« »
»
« »
¬0¼
¬0¼
¬ x4 ¼ ¬ 0 ¼
a basis for Nul A. From this information, dim Col A = 3 (because A has three pivot columns) and dim
Nul A = 1 (because the equation Ax = 0 has only one free variable).
4 º ª 1 −2 −1 2
6 »» ««0
1
1 0
~
0 −2
1 −6 » «0
0
0 1
» «
1
4
1
5¼ ¬ 0
0
0 0
ª 1º ª −2º ª5º ª 4 º
« 2 » « −1» «5» « 6 »
and 5 of A form a basis for Col A: « » , « » , « » , « » . For Nul A,
« −2 » « 0» « 1» « −6»
« » « » « » « »
¬ 3¼ ¬ 1¼ ¬ 1¼ ¬ 5 ¼
ª 1
« 2
10. The information A = «
« −2
«
¬ 3
ª1
«0
~
A
0
[
] ««
0
«
¬0
−2
−1
0
1
1
1
0
0
0
0
0
0
0
0
1
0
0
1
−1
1
0º
0 »»
.
0»
»
0¼
5
5
x1
+ x3
x2 + x3
x4
=
=
=
x5 =
0º
3»»
shows that columns 1, 2,4
0»
»
1¼
0
0
0
0
x3 is a free variable
ª x1 º ª − x3 º
ª −1º
« x » «− x »
« −1»
« 2» « 3»
« »
x = « x3 » = « x3 » = x3 « 1» . Basis for Nul A:
« » «
»
« »
« x4 » « 0 »
« 0»
«¬ x5 »¼ «¬ 0 »¼
«¬ 0 »¼
ª −1º
« −1»
« »
« 1» .
« »
« 0»
«¬ 0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.9
• Solutions
157
From this, dim Col A = 4 and dim Nul A = 1.
ª 2
« 3
11. The information A = «
« 0
«
¬ −3
4
6
−5
−8
2
3
0
−6
9
−7
0
−3
−3º ª 1
−5»» ««0
~
9 » «0
» «
−10 ¼ ¬0
2
0
−5
5
1
0
0
0
0
0
0
0
−4 º
5»»
shows that columns 1and
0»
»
0¼
ª 2 º ª −5º
« 3» « −8»
3 of A form a basis for Col A: « » , « » .
« 0» « 9»
« » « »
¬ −3¼ ¬ −7 ¼
ª1
«0
For Nul A, [ A 0] ~ «
«0
«
¬0
2
0
0
1
1
0
1
1
0
0
0
0
0
0
0
0
0º
x1 + 2 x2
+ x4 + x5 = 0
0 »»
.
x3
+ x5 = 0
0»
» x2 , x4 and x5 are free variables
0¼
ª x1 º ª −2 x2 − x4 − x5 º
ª −2 º
ª −1º
ª −1º
«x » «
»
«
»
«
»
« 0»
x2
« 2» «
»
« 1»
« 0»
« »
» = x2 « 0 » + x4 « 0 » + x5 « −1» . Basis for Nul A:
x = « x3 » = «
− x5
« » «
»
« »
« »
« »
x4
« x4 » «
»
« 0»
« 1»
« 0»
«¬ x5 »¼ «¬
»¼
«¬ 0 »¼
«¬ 0 »¼
«¬ 1»¼
x5
ª −2 º ª
« 1» «
« » «
« 0» , «
« » «
« 0» «
«¬ 0 »¼ «¬
−1º ª
0 »» ««
0» , «
» «
1» «
0 »¼ «¬
−1º
0 »»
−1» .
»
0»
1»¼
From this, dim Col A = 2 and dim Nul A = 3.
ª1
«5
12. The information A = «
«4
«
¬3
−4
−9
6º ª 1 2 8 4
10 »» ««0 2 3 4
~
6 −9 12 15» «0 0 5 0
» «
4 −5 8 9 ¼ ¬0 0 0 0
ª 1º ª 2 º ª −4 º
« 5» « 1» « −9 »
and 3 of A form a basis for Col A: « » , « » , « » . For Nul A
« 4 » « 6 » « −9 »
« » « » « »
¬ 3 ¼ ¬ 4 ¼ ¬ −5 ¼
ª1
«0
[ A 0] ~ ««0
«
¬0
2
1
0 0º
1 0 »»
.
0 1 0 −1 0 »
»
0 0 0 0 0¼
0 0 0
1 0 2
4
2
x1
−6º
−1»»
shows that columns 1, 2,
−5»
»
0¼
= 0
x2
+ 2 x4 + x5 = 0
x3
− x5 = 0
x4 and x5 are free variables
0
ª x1 º ª
º
ª 0º
ª
« x » « −2 x − x »
« −2 »
«
4
5»
« 2» «
« »
«
» = x4 « 0 » + x5 «
x = « x3 » = «
x5
« » «
»
« »
«
x4
« x4 » «
»
« 1»
«
«¬ x5 »¼ «¬
»¼
«¬ 0 »¼
«¬
x5
0º
−1»»
1» . Basis for Nul A:
»
0»
1»¼
ª 0º ª
« −2» «
« » «
« 0» , «
« » «
« 1» «
«¬ 0»¼ «¬
0º
−1»»
1» .
»
0»
1»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
158
CHAPTER 2
• Matrix Algebra
From this, dim Col A = 3 and dim Nul A = 2.
13. The four vectors span the column space H of a matrix that can be reduced to echelon form:
ª 1
« −3
«
« 2
«
«¬ −4
−3
9
2
−1
−6
12
4
2
−4 º ª 1
5»» ««0
~
−3» «0
» «
7 »¼ «¬0
−3
0
2
5
0
0
0
10
−4º ª 1
−7 »» ««0
~
5» « 0
» «
−9 »¼ «¬0
−3
0
2
5
0
0
0
0
−4º ª 1
−7 »» ««0
~
5» « 0
» «
5»¼ «¬0
−3
0
2
5
0
0
0
0
−4º
−7 »»
5»
»
0 »¼
Columns 1, 3, and 4 of the original matrix form a basis for H, so dim H = 3.
Note: Either Exercise 13 or 14 should be assigned because there are always one or two students who
confuse Col A with Nul A. Or, they wrongly connect “set of linear combinations” with “parametric vector
form” (of the general solution of Ax = 0).
14. The five vectors span the column space H of a matrix that can be reduced to echelon form:
ª 1
« −1
«
« −2
«
¬ 3
2
−3
0
−1
−1
4
−1
4
3
−2
−7
7
3º ª 1
−7 »» «« 0
~
6» «0
» «
−9¼ ¬ 0
2
−1
0
−1
−1
3
3
−2
3
−2
−9
10
3º ª 1
−4 »» «« 0
~
12 » « 0
» «
−18¼ ¬ 0
2
−1
0
−1
−1
3
0
0
0
0
4
0
3º
−4»»
−10»
»
0¼
Columns 1,2 and 4 of the original matrix form a basis for H, so dim H = 3.
15. Col A = R4, because A has a pivot in each row and so the columns of A span R4. Nul A cannot equal
R2, because Nul A is a subspace of R6. It is true, however, that Nul A is two-dimensional. Reason: the
equation Ax = 0 has two free variables, because A has six columns and only four of them are pivot
columns.
16. Col A cannot be R3 because the columns of A have four entries. (In fact, Col A is a 3-dimensional
subspace of R4, because the 3 pivot columns of A form a basis for Col A.) Since A has 7 columns and
3 pivot columns, the equation Ax = 0 has 4 free variables. So, dim Nul A = 4.
17. a. True. This is the definition of a B-coordinate vector.
b. False. Dimension is defined only for a subspace. A line must be through the origin in Rn to be a
subspace of Rn.
c. True. The sentence before Example 3 concludes that the number of pivot columns of A is the rank
of A, which is the dimension of Col A by definition.
d. True. This is equivalent to the Rank Theorem because rank A is the dimension of Col A.
e. True, by the Basis Theorem. In this case, the spanning set is automatically a linearly independent
set.
18. a. True. This fact is justified in the second paragraph of this section.
b. False. The dimension of Nul A is the number of free variables in the equation Ax = 0.
See Example 2.
c. True, by the definition of rank.
d. True. See the second paragraph after Fig. 1.
e. True, by the Basis Theorem. In this case, the linearly independent set is automatically a spanning
set.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.9
• Solutions
159
19. The fact that the solution space of Ax = 0 has a basis of three vectors means that dim Nul A = 3.
Since a 5×7 matrix A has 7 columns, the Rank Theorem shows that rank A = 7 – dim Nul A = 4.
Note: One can solve Exercises 19–22 without explicit reference to the Rank Theorem. For instance, in
Exercise 19, if the null space of a matrix A is three-dimensional, then the equation Ax = 0 has three free
variables, and three of the columns of A are nonpivot columns. Since a 5×7 matrix has seven columns, A
must have four pivot columns (which form a basis of Col A). So rank A = dim Col A = 4.
20. A 6×8 matrix A has 8 columns. By the Rank Theorem, rank A = 8 – dim Nul A. Since the null space
is three-dimensional, rank A = 5.
21. A 9×8 matrix has 8 columns. By the Rank Theorem, dim Nul A = 8 – rank A. Since the rank is seven,
dim Nul A = 1. That is, the dimension of the solution space of Ax = 0 is one.
22. Suppose that the subspace H = Span{v1, …, v5} is four-dimensional. If {v1, …, v5} were linearly
independent, it would be a basis for H. This is impossible, by the statement just before the definition
of dimension in Section 2.9, which essentially says that every basis of a p-dimensional subspace
consists of p vectors. Thus, {v1, …, v5} must be linearly dependent.
23. A 3×5 matrix A with a two-dimensional column space has two pivot columns. The remaining three
columns will correspond to free variables in the equation Ax = 0. So the desired construction is
possible. There are ten possible locations for the two pivot columns, one of which is
ª * * * * º
« 0 * * * » . A simple construction is to take two vectors in R3 that are obviously not
«
»
«¬ 0 0 0 0 0 »¼
linearly dependent, and put three copies of these two vectors in any order. The resulting matrix will
obviously have a two-dimensional column space. There is no need to worry about whether Nul A has
the correct dimension, since this is guaranteed by the Rank Theorem: dim Nul A = 5 – rank A.
24. A rank 1 matrix has a one-dimensional column space. Every column is a multiple of some fixed
vector. To construct a 3×4 matrix, choose any nonzero vector in R3, and use it for one column.
Choose any multiples of the vector for the other tthree columns.
25. The p columns of A span Col A by definition. If dim Col A = p, then the spanning set of p columns is
automatically a basis for Col A, by the Basis Theorem. In particular, the columns are linearly
independent.
26. If columns a1, a3, a4, a5, and a7 of A are linearly independent and if dim Col A = 5, then {a1, a3, a4,a5,
a7} is a linearly independent set in a 5-dimensional column space. By the Basis Theorem, this set of
five vectors is a basis for the column space.
27. a. Start with B = [b1 ⋅ ⋅ ⋅ bp] and A = [a1 ⋅ ⋅ ⋅ aq], where q > p. For j = 1, …, q, the vector aj is
in W. Since the columns of B span W, the vector aj is in the column space of B. That is, aj = Bcj
for some vector cj of weights. Note that cj is in Rp because B has p columns.
b. Let C = [c1 ⋅ ⋅ ⋅ cq]. Then C is a p×q matrix because each of the q columns is in Rp.
By hypothesis, q is larger than p, so C has more columns than rows. By a theorem, the columns of
C are linearly dependent and there exists a nonzero vector u in Rq such that Cu = 0.
c. From part (a) and the definition of matrix multiplication
A = [a1 ⋅ ⋅ ⋅ aq] = [Bc1 ⋅ ⋅ ⋅ Bcq] = BC
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
160
CHAPTER 2
• Matrix Algebra
From part (b), Au = (BC)u = B(Cu) = B0 = 0. Since u is nonzero, the columns of A are linearly
dependent.
28. If A contained more vectors than B, then A would be linearly dependent, by Exercise 27, because B
spans W. Repeat the argument with B and A interchanged to conclude that B cannot contain more
vectors than A.
29. [M] Apply the matrix command ref or rref to the matrix [v1 v2 x]:
ª 15
« −5
«
« 12
«
¬ 7
14
−10
13
17
16 º ª 1
0 »» ««0
~
11» «0
» «
−3¼ ¬0
0
1
0
0
2º
−1»»
0»
»
0¼
The equation c1v1 + c2v2 = x is consistent, so x is in the subspace H. Then c1 = 2 and c2 = -1. Thus,
the -coordinate of x is (2, -1).
30. [M] Apply the matrix command ref or rref to the matrix [v1 v2 v3 x]:
ª −6 8 −9 11 º ª 1 0 0 −2 º
«3
0
4
−2 »» ««0 1 0
1»»
«
~
« −9 7 −8 −17 » «0 0 1
1»
«
» «
»
0¼
−8 ¼ ¬0 0 0
¬ 4 −3 3
The first three columns of [v1 v2 v3 x] are pivot columns, so v1, v2 and v3 are linearly independent.
Thus v1, v2 and v3 form a basis B for the subspace H which they span. View [v1 v2 v3 x] as an
augmented matrix for c1v1 + c2v2 + c3v3 = x. The reduced echelon form shows that x is in H and
ª −2 º
[x]B = « 1» .
« »
«¬ 1»¼
Notes: The Study Guide for Section 2.9 contains a complete list of the statements in the Invertible Matrix
Theorem that have been given so far. The format is the same as that used in Section 2.3, with three
columns statements that are logically equivalent for any m×n matrix and are related to existence concepts,
those that are equivalent only for any n×n matrix, and those that are equivalent for any n×p matrix and
are related to uniqueness concepts. Four statements are included that are not in the text’s official list of
statements, to give more symmetry to the three columns.
The Study Guide section also contains directions for making a review sheet for “dimension” and
“rank
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 2
• Supplementary Exercises
161
Chapter 2 SUPPLEMENTARY EXERCISES
1. a. True. If A and B are m×n matrices, then BT has as many rows as A has columns, so ABT is
defined. Also, ATB is defined because AT has m columns and B has m rows.
b. False. B must have 2 columns. A has as many columns as B has rows.
c. True. The ith row of A has the form (0, …, di, …, 0). So the ith row of AB is (0, …, di, …, 0)B,
which is di times the ith row of B.
d. False. Take the zero matrix for B. Or, construct a matrix B such that the equation Bx = 0 has
nontrivial solutions, and construct C and D so that C ≠ D and the columns of C – D satisfy the
equation Bx = 0. Then B(C – D) = 0 and BC = BD.
ª 1 0º
ª0 0 º
e. False. Counterexample: A = «
and C = «
».
»
¬ 0 1¼
¬0 0 ¼
f. False. (A + B)(A – B) = A2 – AB + BA – B2. This equals A2 – B2 if and only if A commutes with B.
g. True. An n×n replacement matrix has n + 1 nonzero entries. The n×n scale and interchange
matrices have n nonzero entries.
h. True. The transpose of an elementary matrix is an elementary matrix of the same type.
i. True. An n×n elementary matrix is obtained by a row operation on In.
j. False. Elementary matrices are invertible, so a product of such matrices is invertible. But not
every square matrix is invertible.
k. True. If A is 3×3 with three pivot positions, then A is row equivalent to I3.
l. False. A must be square in order to conclude from the equation AB = I that A is invertible.
m. False. AB is invertible, but (AB)–1 = B–1A–1, and this product is not always equal to A–1B–1.
n. True. Given AB = BA, left-multiply by A–1 to get B = A–1BA, and then right-multiply by A–1 to
obtain BA–1 = A–1B.
o. False. The correct equation is (rA)–1 = r–1A–1, because
(rA)(r–1A–1) = (rr–1)(AA–1) = 1¸I = I.
ª 1º
p. True. If the equation Ax = «0» has a unique solution, then there are no free variables in this
« »
«¬0»¼
equation, which means that A must have three pivot positions (since A is 3×3). By the Invertible
Matrix Theorem, A is invertible.
2. C = (C – 1 ) –1 =
ª0
3. A = «1
«
¬« 0
0
0
1
1 ª 7
−2 «¬ −6
0º
0» ,
»
0 »¼
ª0
A = A ⋅ A = «1
«
«¬ 0
3
2
−5 º ª −7 / 2
=
4 »¼ «¬ 3
ª0
A = «1
«
¬« 0
2
0
0
1
0º ª0
0» «0
»«
0 »¼ «¬1
0
0
1
0
0
0
0º ª0
0 » «1
»«
0 ¼» ¬« 0
5 / 2º
−2 »¼
0
0
1
0º ª0
0» = «0
» «
0 »¼ «¬ 0
0º ª0
0» = «0
» «
0 ¼» ¬«1
0
0
0
0
0
0
0º
0»
»
0 »¼
0º
0»
»
0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
162
CHAPTER 2
• Matrix Algebra
Next, ( I − A)( I + A + A2 ) = I + A + A2 − A( I + A + A2 ) = I + A + A2 − A − A2 − A3 = I − A3 .
Since A3 = 0, ( I − A)( I + A + A2 ) = I .
4. From Exercise 3, the inverse of I – A is probably I + A + A2 + ¸ ¸ ¸ + An–1. To verify this, compute
( I − A)( I + A + " + An −1 ) = I + A + " + An −1 − A( I + A + " + An −1 ) = I − AAn −1 = I − An
If An = 0, then the matrix B = I + A + A2 + ¸ ¸ ¸ + An–1 satisfies (I – A)B = I. Since I – A and B are
square, they are invertible by the Invertible Matrix Theorem, and B is the inverse of I – A.
5. A2 = 2A – I. Multiply by A: A3 = 2A2 – A. Substitute A2 = 2A – I: A3 = 2(2A – I) – A = 3A – 2I.
Multiply by A again: A4 = A(3A – 2I) = 3A2 – 2A. Substitute the identity A2 = 2A – I again:
Finally, A4 = 3(2A – I) – 2A = 4A – 3I.
ª1
6. Let A = «
¬0
BA.
0º
ª0
and B = «
»
−1¼
¬1
1º
ª 0
2
2
. By direct computation, A = I, B = I, and AB = «
»
0¼
¬ −1
1º
=–
0 »¼
7. (Partial answer in Study Guide) Since A–1B is the solution of AX = B, row reduction of [A B] to
[I X] will produce X = A–1B. See Exercise 15 in Section 2.2.
[A
ª1
B ] = «« 2
¬« 1
ª1
~ «« 0
«¬ 0
3
8
−3
4
11
1
2
5
3
5º ª 1
1»» ~ «« 0
−3»¼ «¬ 0
3
8
−3
1
3
−6
0
1
−5
5º ª 1
5 »» ~ «« 0
4 ¼» ¬« 0
3
8
−3
−2
−5
7
−1
−3
6
29 º ª 1
10 »» ~ «« 0
−3»¼ «¬ 0
3
0
37
1
0
9
0
1
−5
5º ª 1
−5 »» ~ «« 0
−1¼» ¬« 0
3
8
−3
1
3
−6
−2
−5
7
0
0
10
1
0
9
0
1
−5
−1º
10 »»
−3»¼
−1º
10 »» .
−3¼»
ª 10
Thus, A B = « 9
«
¬« −5
–1
8. By definition of matrix multiplication, the matrix A satisfies
ª1
A«
¬3
2 º ª1
=
7 »¼ «¬1
3º
1»¼
ª1
Right-multiply both sides by the inverse of «
¬3
ª1
A=«
¬1
3º ª 7
1»¼ «¬ −3
ª 5
9. Given AB = «
¬ −2
ª 1
B −1 = «
¬ −2
−2 º ª −2
=
1»¼ «¬ 4
4º
ª7
and B = «
»
3¼
¬2
2º
. The left side becomes A. Thus,
7 »¼
1º
−1»¼
3º
, notice that ABB–1 = A. Since det B = 7 – 6 =1,
»
1¼
−3 º
ª 5
and A = ( AB ) B −1 = «
»
7¼
¬ −2
4º ª 1
3»¼ «¬ −2
− 3º ª −3
=
7 »¼ «¬ −8
13º
27 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5º
1»»
−5¼»
2.9
• Solutions
163
Note: Variants of this question make simple exam questions.
10. Since A is invertible, so is AT, by the Invertible Matrix Theorem. Then ATA is the product of
invertible matrices and so is invertible. Thus, the formula (ATA)–1AT makes sense. By Theorem 6 in
Section 2.2,
(ATA)–1¸AT = A–1(AT)–1AT = A–1I = A–1
An alternative calculation: (ATA)–1AT¸A = (ATA)–1(ATA) = I. Since A is invertible, this equation shows
that its inverse is (ATA)–1AT.
11. a. For i = 1,…, n, p(xi) = c0 + c1xi + ¸ ¸ ¸ +
cn −1 x in −1
ª c0 º
= row i (V ) ⋅ « # » = row i (V )c .
«
»
«¬ cn −1 »¼
By a property of matrix multiplication, shown after Example 6 in Section 2.1, and the fact that c
was chosen to satisfy Vc= y,
row i (V )c = row i (Vc) = row i ( y ) = yi
Thus, p(xi) = yi. To summarize, the entries in Vc are the values of the polynomial p(x) at x1, …, xn.
b. Suppose x1, …, xn are distinct, and suppose Vc = 0 for some vector c. Then the entries in c are the
coefficients of a polynomial whose value is zero at the distinct points x1, ..., xn. However, a
nonzero polynomial of degree n – 1 cannot have n zeros, so the polynomial must be identically
zero. That is, the entries in c must all be zero. This shows that the columns of V are linearly
independent.
c. (Solution in Study Guide) When x1, …, xn are distinct, the columns of V are linearly independent,
by (b). By the Invertible Matrix Theorem, V is invertible and its columns span Rn. So, for every
y = (y1, …, yn) in Rn, there is a vector c such that Vc = y. Let p be the polynomial whose
coefficients are listed in c. Then, by (a), p is an interpolating polynomial for (x1, y1), …, (xn, yn).
12. If A = LU, then col1(A) = L¸col1(U). Since col1(U) has a zero in every entry except possibly the first,
L¸col1(U) is a linear combination of the columns of L in which all weights except possibly the first
are zero. So col1(A) is a multiple of col1(L).
Similarly, col2(A) = L¸col2(U), which is a linear combination of the columns of L using the first
two entries in col2(U) as weights, because the other entries in col2(U) are zero. Thus col2(A) is a
linear combination of the first two columns of L.
13. a. P2 = (uuT)(uuT) = u(uTu)uT = u(1)uT = P, because u satisfies uTu = 1.
b. PT = (uuT)T = uTTuT = uuT = P
c. Q2 = (I – 2P)(I – 2P) = I – I(2P) – 2PI + 2P(2P)
= I – 4P + 4P2 = I, because of part (a).
ª0º
14. Given u = « 0 » , define P and Q as in Exercise 13 by
« »
«¬ 1 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
164
CHAPTER 2
• Matrix Algebra
ª0º
P = uu = « 0 » [ 0
« »
«¬1 »¼
T
0
ª0
1] = «0
«
«¬0
ª1 º
ª0
«
»
If x = 5 , then Px = « 0
« »
«
«¬3»¼
«¬ 0
0
0
0
0
0
0
0º
ª1
»
0 , Q = I − 2 P = «0
»
«
«¬ 0
1 »¼
0 º ª1 º ª 0 º
0 » «5 » = « 0 »
»« » « »
1 »¼ «¬3»¼ «¬ 3 »¼
ª1
and Qx = «0
«
«¬0
0º
ª0
»
0 − 2 «0
»
«
«¬ 0
1 »¼
0
1
0
0
1
0
0 º ª1
0 » = «0
» «
1 »¼ «¬0
0
0
0
0
1
0
0º
0»
»
−1»¼
0 º ª 1º ª 1º
0 » «5 » = « 5 » .
»« » « »
−1»¼ «¬ 3»¼ «¬ −3»¼
15. Left-multiplication by an elementary matrix produces an elementary row operation:
B ~ E1 B ~ E2 E1 B ~ E3 E2 E1 B = C
so B is row equivalent to C. Since row operations are reversible, C is row equivalent to B.
(Alternatively, show C being changed into B by row operations using the inverse of the Ei .)
16. Since A is not invertible, there is a nonzero vector v in Rn such that Av = 0. Place n copies of v into
an n×n matrix B. Then AB = A[v ⋅ ⋅ ⋅ v] = [Av ⋅ ⋅ ⋅ Av] = 0.
17. Let A be a 6×4 matrix and B a 4×6 matrix. Since B has more columns than rows, its six columns are
linearly dependent and there is a nonzero x such that Bx = 0. Thus ABx = A0 = 0. This shows that the
matrix AB is not invertible, by the IMT. (Basically the same argument was used to solve Exercise 22
in Section 2.1.)
Note: (In the Study Guide) It is possible that BA is invertible. For example, let C be an invertible 4×4
ªC º
matrix and construct A = « » and B = [C −1 0]. Then BA = I4, which is invertible.
¬0¼
18. By hypothesis, A is 5×3, C is 3×5, and CA = I3. Suppose x satisfies Ax = b. Then CAx = Cb. Since
CA = I, x must be Cb. This shows that Cb is the only solution of Ax = b.
ª.4 .2 .3º
ª .31
«
»
2
19. [M] Let A = .3 .6 .3 . Then A = «.39
«
»
«
«¬.3 .2 .4 »¼
«¬.30
the calculations by computing
ª.2875 .2834 .2874 º
A = A A = « .4251 .4332 .4251» ,
«
»
«¬.2874 .2834 .2875 »¼
To four decimal places, as k increases,
4
2
2
ª.2857
A → «.4286
«
«¬.2857
k
ª0
If B = « .1
«
¬«.9
.2
.6
.2
.3º
.3» ,
»
.4 ¼»
.2857
.4286
.2857
then
.26
.48
.26
.30 º
.39 » . Instead of computing A3 next, speed up
»
.31»¼
ª.2857
A = A A = «.4285
«
«¬.2857
8
4
4
.2857
.4286
.2857
.2857 º
ª2 / 7
»
k
.4286 , or, in rational format, A → « 3/ 7
»
«
«¬ 2 / 7
.2857 »¼
ª.29
B = «.33
«
¬«.38
2
.18
.44
.38
.2857 º
.4285 »
»
.2857 »¼
2/7
3/ 7
2/7
2 / 7º
3/ 7 » .
»
2 / 7 »¼
.18º
.33» ,
»
.49 ¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
2.9
ª.2119
B = ««.3663
«¬.4218
4
.1998 º
.3663»» ,
.4339 »¼
.1998
.3784
.4218
ª.2024
B = ««.3707
«¬.4269
8
• Solutions
.2022 º
.3707 »»
.4271»¼
.2022
.3709
.4269
To four decimal places, as k increases,
ª.2022
B → «.3708
«
¬«.4270
.2022 º
ª18 / 89
»
k
.3708 , or, in rational format, B → « 33 / 89
»
«
«¬38 / 89
.4270 ¼»
.2022
k
.3708
.4270
18 / 89
33/ 89
38 / 89
18 / 89 º
33/ 89 » .
»
38 / 89 ¼»
20. [M] The 4×4 matrix A4 is the 4×4 matrix of ones, minus the 4×4 identity matrix. The MATLAB
command is A4 = ones(4) – eye(4). For the inverse, use inv(A4).
ª −2 / 3
« 1/ 3
A4−1 = «
« 1/ 3
«
¬ 1/ 3
ª0
«1
A4 = «
«1
«
¬1
1
0
1
1
1
1
0
1
1º
1»»
,
1»
»
0¼
ª0
«1
«
A5 = « 1
«
«1
«¬ 1
1
0
1
1
1
1
1
0
1
1
1
1
1
0
1
1º
1»»
1» ,
»
1»
0 »¼
ª0
«1
«
«1
A6 = «
«1
«1
«
¬« 1
1
1
1
1
0
1
1
1
1
1
0
1
1
0
1
1
1
1
1
0
1
1
1
1
1/ 3
−2 / 3
1/ 3
1/ 3
ª −3/ 4
« 1/ 4
«
A5−1 = « 1/ 4
«
« 1/ 4
«¬ 1/ 4
1º
1»
»
1»
»,
1»
1»
»
0 ¼»
1/ 3
1/ 3
−2 / 3
1/ 3
1/ 4
−3/ 4
1/ 4
1/ 4
1/ 4
ª −4 / 5
« 1/ 5
«
« 1/ 5
−1
A6 = «
« 1/ 5
« 1/ 5
«
¬« 1/ 5
1/ 3º
1/ 3»»
1/ 3»
»
−2 / 3¼
1/ 4
1/ 4
−3/ 4
1/ 4
1/ 4
1/ 4
1/ 4
1/ 4
−3/ 4
1/ 4
1/ 4 º
1/ 4 »»
1/ 4 »
»
1/ 4 »
−3/ 4 »¼
1/ 5
1/ 5
1/ 5
1/ 5
−4 / 5
1/ 5
1/ 5
1/ 5
1/ 5
1/ 5
−4 / 5
1/ 5
1/ 5
−4 / 5
1/ 5
1/ 5
1/ 5
1/ 5
1/ 5
−4 / 5
1/ 5
1/ 5
1/ 5
1/ 5
1/ 5º
1/ 5»
»
1/ 5»
»
1/ 5»
1/ 5»
»
−4 / 5¼»
The construction of A6 and the appearance of its inverse suggest that the inverse is related to I6. In
fact, A6−1 + I 6 is 1/5 times the 6×6 matrix of ones. Let J denotes the n×n matrix of ones. The
conjecture is:
1
⋅ J − In
n −1
Proof: (Not required) Observe that J 2 = nJ and An J = (J – I ) J = J 2 – J = (n – 1) J. Now compute
An((n – 1)–1J – I) = (n – 1)–1A n J – An = J – (J – I) = I
Since An is square, An is invertible and its inverse is (n – 1)–1J – I.
An = J – In and
An−1 =
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
165
3.1
SOLUTIONS
Notes: Some exercises in this section provide practice in computing determinants, while others allow the
student to discover the properties of determinants which will be studied in the next section. Determinants
are developed through the cofactor expansion, which is given in Theorem 1. Exercises 33–36 in this
section provide the first step in the inductive proof of Theorem 3 in the next section.
1. Expanding along the first row:
3
2
0
3
0
5
4
3
2 =3
5
−1
2
2
−0
−1
0
2
2
+4
−1
0
3
= 3(−13) + 4(10) = 1
5
Expanding along the second column:
3
2
0
3
0
5
4
2
2 = (−1)1+ 2 ⋅ 0
0
−1
2
3
+ (−1) 2+ 2 ⋅ 3
−1
0
4
3
+ ( −1)3+ 2 ⋅ 5
−1
2
4
= 3(−3) − 5(−2) = 1
2
2. Expanding along the first row:
0
4
5
−3
2
4
1
−3
0 =0
4
1
0
4
−5
1
2
0
4
+1
1
2
−3
= −5(4) + 1(22) = 2
4
Expanding along the second column:
0
4
5
−3
2
4
1
4
0 = (−1)1+ 2 ⋅ 5
2
1
0
0
+ (−1) 2+ 2 ⋅ (−3)
1
2
1
0
+ (−1)3+ 2 ⋅ 4
1
4
1
= −5(4) − 3(−2) − 4( −4) = 2
0
3. Expanding along the first row:
2
3
−4
1
1
4
3
1
2 =2
4
−1
2
3
− (−4)
1
−1
2
3
+3
1
−1
1
= 2(−9) + 4(−5) + (3)(11) = −5
4
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
167
168
CHAPTER 3
• Determinants
Expanding along the second column:
2
3
−4
1
1
4
3
3
2 = (−1)1+ 2 ⋅ (−4)
1
−1
2
2
+ ( −1) 2+ 2 ⋅ 1
−1
1
3
2
+ (−1)3+ 2 ⋅ 4
−1
3
3
= 4(−5) + 1(−5) − 4(−5) = −5
2
4. Expanding along the first row:
1
2
3
1
3
4
5
1
1 =1
4
2
1
2
−3
2
3
1
2
+5
2
3
1
= 1(−2) − 3(1) + 5(5) = 20
4
Expanding along the second column:
1
2
3
1
3
4
5
2
1 = (−1)1+ 2 ⋅ 3
3
2
1
1
+ (−1)2 + 2 ⋅ 1
2
3
5
1
+ (−1)3+ 2 ⋅ 4
2
2
5
= −3(1) + 1(−13) − 4(−9) = 20
1
5. Expanding along the first row:
2
4
3
0
5
1
−4
0
5 =2
1
6
5
4
−3
6
5
5
4
+ (−4)
6
5
0
= 2(−5) − 3(−1) − 4(4) = −23
1
6. Expanding along the first row:
5
0
−2
3
2
−4
4
−5 = 5
7
3
−4
0
−5
− (−2)
7
2
0
−5
+4
7
2
3
= 5(1) + 2(10) + 4(−6) = 1
−4
7. Expanding along the first row:
4
6
3
5
9
7
0
5
2 =4
7
3
2
6
−3
3
9
2
6
+0
3
9
5
= 4(1) − 3(0) = 4
7
8. Expanding along the first row:
8
4
1
0
3
−2
6
0
3 =8
−2
5
3
4
−1
5
3
3
4
+6
5
3
0
= 8(6) − 1(11) + 6(−8) = −11
−2
9. First expand along the third row, then expand along the first row of the remaining matrix:
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3.1
6
0
0
5
1
7
2
−5
2
8
0
3
0
1
0
8
= (−1)
3+1
0
0
⋅2 7
3
2
1
5
7
−5 = 2 ⋅ (−1)1+3 ⋅ 5
3
8
2
1
• Solutions
169
= 10(1) = 10
10. First expand along the second row, then expand along either the third row or the second column of
the remaining matrix.
1
−2
5
2
0
0
3
0
2
5
−6
0
−7
4
5
4
= (−1)
§
−2
= (−3) ¨ (−1)3+1 ⋅ 5
−6
©
2+3
2
1
−2
2
⋅3 2
5
−6
0
5
4
+ (−1)3+3 ⋅ 4
5
−2 ·
¸ = (−3) (5(2) + 4(−2)) = −6
−6 ¹
1
2
or
1
−2
5
2
0
0
3
0
2
5
−6
0
−7
4
5
4
= (−1)
§
2
= (−3) ¨ (−1)1+ 2 ⋅ (−2)
5
©
2+3
5
4
1
−2
2
⋅3 2
5
−6
0
5
4
+ (−1)2 + 2 ⋅ (−6)
1
5
2 ·
¸ = (−3) ( 2(−17) − 6( −6) ) = −6
4 ¹
11. There are many ways to do this determinant efficiently. One strategy is to always expand along the
first column of each matrix:
3
5
−8
4
0
−2
3
−7
0
0
0
0
1
0
5
2
1+1
= (−1)
−2
3
⋅3 0
0
1
0
−7
1
5 = 3 ⋅ (−1)1+1 ⋅ (−2)
0
2
5
2
= 3(–2)(2) = –12
12. There are many ways to do this determinant efficiently. One strategy is to always expand along the
first row of each matrix:
4
0
0
0
7
−1
0
0
2
5
6
−8
3
4
0
−3
1+1
= (−1)
−1
0
⋅4 6
−8
3
4
0
3
0 = 4 ⋅ (−1)1+1 ⋅ (−1)
4
−3
0
−3
= 4(–1)( –9) = 36
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
170
CHAPTER 3
• Determinants
13. First expand along either the second row or the second column. Using the second row,
4
0
−7
3
−5
0
0
2
0
0
7
3
−6
4
−8 = (−1) 2+3 ⋅ 2
5
0
5
2
−3
0
0
9
−1
2
4
0
3
−5
7
3
4
−8
5
0
2
−3
0
0
−1
2
Now expand along the second column to find:
(−1)
2+3
⋅2
4
0
3
−5
7
3
4
−8
5
0
0
0
2
−1
§
4
¨
2+ 2
= −2 ¨ (−1) ⋅ 3 5
−3
¨
0
©
2
−5 ·
¸
−3 ¸
2 ¸¹
3
2
−1
Now expand along either the first column or third row. The first column is used below.
§
4
¨
2+ 2
−2 ¨ (−1) ⋅ 3 5
¨
0
©
−5
−3
2
3
2
−1
·
§
2
¸
1+1
¸ = −6 ¨ (−1) ⋅ 4 −1
©
¸
¹
−3
2
+ (−1)2 +1 ⋅ 5
3
−1
−5 ·
¸ = (−6)(4(1) − 5(1)) = 6
2 ¹
14. First expand along either the fourth row or the fifth column. Using the fifth column,
6
3
2
4
0
9
0
−4
1
0
8
−5
6
7
1 = (−1)3+5 ⋅ 1
3
0
0
0
0
4
2
3
2
0
6
3
2
4
9
0
−4
1
3
0
0
0
4
2
3
2
Now expand along the third row to find:
(−1)
3+ 5
⋅1
6
3
2
9
0
−4
3
4
0
2
0
3
4
§
3
¨
3+1
= 1¨ ( −1) ⋅ 3 0
0
¨
2
©
2
1
2
−4
3
4 ·
¸
1 ¸
2 ¸¹
Now expand along either the first column or second row. The first column is used below.
§
3
¨
3+1
1¨ (−1) ⋅ 3 0
¨
2
©
2
−4
3
4
1
2
·
§
−4
¸
1+1
¸ = 3 ¨ (−1) ⋅ 3 3
©
¸
¹
1
2
+ ( −1)3+1 ⋅ 2
2
−4
4 ·
¸ = (3)(3( −11) + 2(18)) = 9
1 ¹
3
0
4
15. 2
0
3
2 = (3)(3)(–1) + (0)(2)(0) + (4)(2)(5) – (0)(3)(4) – (5)(2)(3) – (–1)(2)(0)
5
−1
= –9 + 0 + 40 – 0 – 30 –0 = 1
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3.1
0
5
16. 4
−3
2
4
• Solutions
171
1
0 = (0)(–3)(1) + (5)(0)(2) + (1)(4)(4) – (2)(–3)(1) – (4)(0)(0) – (1)(4)(5)
1
= 0 + 0 + 16 – (–6) – 0 – 20 = 2
2
−4
17. 3
1
1
4
3
2 = (2)(1)(–1) + (–4)(2)(1) + (3)(3)(4) – (1)(1)(3) – (4)(2)(2) – (–1)(3)(–4)
−1
= –2 + (–8) + 36 – 3 – 16 – 12 = –5
1
3
5
18. 2
1
1 = (1)(1)(2) + (3)(1)(3) + (5)(2)(4) – (3)(1)(5) – (4)(1)(1) – (2)(2)(3)
3
4
2
= 2 + 9 + 40 – 15 – 4 – 12 = 20
19.
a
c
b
c
= ad − bc,
d
a
d
= cb − da = −( ad − bc)
b
The row operation swaps rows 1 and 2 of the matrix, and the sign of the determinant is reversed.
20.
a
c
b
a
= ad − bc,
d
kc
b
= a (kd ) − ( kc)b = kad − kbc = k (ad − bc)
kd
The row operation scales row 2 by k, and the determinant is multiplied by k.
21.
3
5
4
3
= 18 − 20 = −2,
6
5 + 3k
4
= 3(6 + 4k ) − (5 + 3k )4 = −2
6 + 4k
The row operation replaces row 2 with k times row 1 plus row 2, and the determinant is unchanged.
22.
b
a + kc
= ad − bc,
d
c
a
c
b + kd
= ( a + kc)d − c(b + kd ) = ad + kcd − bc − kcd = ad − bc
d
The row operation replaces row 1 with k times row 2 plus row 1, and the determinant is unchanged.
1
1
23. −3
8
2
−3
1
k
−4 = 1(4) − 1(2) + 1(−7) = −5, −3
k
8
2
−3
2
k
−4 = k (4) − k (2) + k (−7) = −5k
2
The row operation scales row 1 by k, and the determinant is multiplied by k.
a
24. 3
b
2
c
2 = a (2) − b(6) + c(3) = 2a − 6b + 3c,
6
5
6
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
172
CHAPTER 3
• Determinants
3
2
2
a
b
c = 3(6b − 5c) − 2(6a − 6c) + 2(5a − 6b) = −2a + 6b − 3c
6
5
6
The row operation swaps rows 1 and 2 of the matrix, and the sign of the determinant is reversed.
25. Since the matrix is triangular, by Theorem 2 the determinant is the product of the diagonal entries:
1
0
0
0
1
0 = (1)(1)(1) = 1
0
k
1
26. Since the matrix is triangular, by Theorem 2 the determinant is the product of the diagonal entries:
1
0
0
0
1
0 = (1)(1)(1) = 1
k
0
1
27. Since the matrix is triangular, by Theorem 2 the determinant is the product of the diagonal entries:
k
0
0
0
1
0 = (k )(1)(1) = k
0
0
1
28. Since the matrix is triangular, by Theorem 2 the determinant is the product of the diagonal entries:
1
0
0
0
k
0 = (1)(k )(1) = k
0
0
1
29. A cofactor expansion along row 1 gives
0
1
0
1
0
0 = −1
0
0
1
1
0
0
1
= −1
30. A cofactor expansion along row 1 gives
0
0
1
0
1
0 =1
1
0
0
0
1
1
0
= −1
31. A 3 × 3 elementary row replacement matrix looks like one of the six matrices
ª1
«k
«
«¬ 0
0
1
0
0º ª1
0 »» , «« 0
1 »¼ «¬ k
0
1
0
0 º ª1
0 »» , «« 0
1 »¼ «¬ 0
0
1
k
0 º ª1
0 »» , «« 0
1 »¼ «¬ 0
0
1
0
0 º ª1
k »» , «« 0
1 »¼ «¬ 0
0
1
0
k º ª1
0 »» , «« 0
1 »¼ «¬ 0
k
1
0
0º
0 »»
1 »¼
In each of these cases, the matrix is triangular and its determinant is the product of its diagonal
entries, which is 1. Thus the determinant of a 3 × 3 elementary row replacement matrix is 1.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3.1
• Solutions
173
32. A 3 × 3 elementary scaling matrix with k on the diagonal looks like one of the three matrices
ªk
«0
«
¬« 0
0
1
0
0 º ª1
0 »» , «« 0
1 ¼» ¬« 0
0
k
0
0 º ª1
0 »» , «« 0
1 ¼» ¬« 0
0
1
0
0º
0 »»
k ¼»
In each of these cases, the matrix is triangular and its determinant is the product of its diagonal
entries, which is k. Thus the determinant of a 3 × 3 elementary scaling matrix with k on the diagonal
is k.
ª0 1 º
ªa b º
ªc d º
33. E = «
, A=«
, EA = «
»
»
»
¬1 0 ¼
¬c d ¼
¬a b ¼
det E = –1, det A = ad – bc,
det EA = cb – da = –1(ad – bc) = (det E)(det A)
bº
ª1 0 º
ªa b º
ªa
34. E = «
, A=«
, EA = «
»
»
»
¬0 k ¼
¬c d ¼
¬ kc kd ¼
det E = k, det A = ad – bc,
det EA = a(kd) – (kc)b = k(ad – bc) = (det E)(det A)
ª1 k º
ªa b º
ª a + kc b + kd º
35. E = «
, A=«
, EA = «
»
»
d »¼
¬0 1 ¼
¬c d ¼
¬ c
det E = 1, det A = ad – bc,
det EA = (a + kc)d – c(b + kd) = ad + kcd – bc – kcd = 1(ad – bc) = (det E)(det A)
b º
ª1 0º
ªa b º
ª a
36. E = «
, A=«
, EA = «
»
»
»
¬k 1¼
¬c d ¼
¬ ka + c kb + d ¼
det E = 1, det A = ad – bc,
det EA = a(kb + d) – (ka + c)b = kab + ad – kab – bc = 1(ad – bc) = (det E)(det A)
ª3
37. A = «
¬4
1º
ª15
, 5A = «
»
2¼
¬ 20
5º
, det A = 2, det 5A = 50 ≠ 5det A
10 »¼
ªa b º
ª ka kb º
38. A = «
, kA = «
»
» , det A = ad – bc,
¬c d ¼
¬ kc kd ¼
det kA = (ka)(kd ) − (kb)(kc) = k 2 (ad − bc) = k 2 det A
39. a. True. See the paragraph preceding the definition of the determinant.
b. False. See the definition of cofactor, which precedes Theorem 1.
40. a. False. See Theorem 1.
b. False. See Theorem 2.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
174
CHAPTER 3
• Determinants
ª 3º
ª1 º
41. The area of the parallelogram determined by u = « » , v = « » , u + v, and 0 is 6, since the base of
¬0¼
¬2¼
the parallelogram has length 3 and the height of the parallelogram is 2. By the same reasoning, the
ª 3º
ª xº
area of the parallelogram determined by u = « » , x = « » , u + x, and 0 is also 6.
¬0¼
¬2¼
X2
V
2
X2
X
2
1
1
1
2
U
4
X1
1
2
U
X1
ª3 1 º
ª3 x º
v ] = det «
= 6, and det [u x ] = det «
»
» = 6. The determinant of the
¬0 2 ¼
¬0 2 ¼
matrix whose columns are those vectors which define the sides of the parallelogram adjacent to 0 is
equal to the area of the parallelogram
Also note that det [ u
ªa º
ªc º
42. The area of the parallelogram determined by u = « » , v = « » , u + v, and 0 is cb, since the base of
¬b ¼
¬0¼
the parallelogram has length c and the height of the parallelogram is b.
X2
b
U
V
a
c
X1
ªa c º
ªc a º
v ] = det «
= −cb , and det [ v u ] = det «
»
» = cb. The determinant of
¬b 0¼
¬0 b ¼
the matrix whose columns are those vectors which define the sides of the parallelogram adjacent to 0
either is equal to the area of the parallelogram or is equal to the negative of the area of the
parallelogram.
Also note that det [ u
43. [M] Answers will vary. The conclusion should be that det (A + B) ≠ det A + det B.
44. [M] Answers will vary. The conclusion should be that det (AB) = (det A)(det B).
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3.2
• Solutions
175
45. [M] Answers will vary. For 4 × 4 matrices, the conclusions should be that det AT = det A,
det(–A) = det A, det(2A) = 16det A, and det (10 A) = 104 det A . For 5 × 5 matrices, the conclusions
should be that det AT = det A, det(–A) = –det A, det(2A) = 32det A, and det (10 A) = 105 det A. For 6
× 6 matrices, the conclusions should be that det AT = det A , det(–A) = det A, det(2A) = 64det A, and
det (10 A) = 106 det A.
46. [M] Answers will vary. The conclusion should be that det A−1 = 1/ det A.
3.2
SOLUTIONS
Notes: This section presents the main properties of the determinant, including the effects of row
operations on the determinant of a matrix. These properties are first studied by examples in Exercises 1–
20. The properties are treated in a more theoretical manner in later exercises. An efficient method for
computing the determinant using row reduction and selective cofactor expansion is presented in this
section and used in Exercises 11–14. Theorems 4 and 6 are used extensively in Chapter 5. The linearity
property of the determinant studied in the text is optional, but is used in more advanced courses.
1. Rows 1 and 2 are interchanged, so the determinant changes sign (Theorem 3b.).
2. The constant 2 may be factored out of the Row 1 (Theorem 3c.).
3. The row replacement operation does not change the determinant (Theorem 3a.).
4. The row replacement operation does not change the determinant (Theorem 3a.).
1
5
−6
1
5
−6
−4
4 = 0
1
−2 = 0
1
−2 = 3
−7
9
3
−3
0
1
5
5. −1
−2
1
5
6. 3
−3
2
13
7.
8.
−6
−3
0
1
5
3 = 0
−18
−7
0
3
1
3
0
2
−2
−5
7
4
3
1
5
−1
2
2
1
−3
1
3
3
−4
0
1
2
−5
2
−3
5
−7
4
−5
−3
2
=
0
−3
1
5
12 = 6 0
−3
−1
0
−3
3
0
2
0
1
7
8
0
0
−4
−4
2
2
−5
−5
=
1
3
3
−4
0
1
2
−5
0
0
−1
2
−2
4
5
−10
1
5
2 =6 0
−3
−1
3
1
=
3
0
−3
2 = (6)(−3) = −18
0
1
1
3
0
2
0
1
7
8
0
0
0
0
30
30
27
27
1
3
3
−4
0
1
2
−5
0
0
0
0
0
0
0
0
=
=
1
3
0
2
0
1
7
8
0
0
0
0
30
0
27
0
=0
=0
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
176
CHAPTER 3
• Determinants
1
−1
−3
0
0
1
5
4
−1
3
2
−1
8
−2
5
3
1
3
−1
0
−2
1
3
−1
0
−2
1
0
2
−4
−1
−6
0
2
−4
−1
−6
10. −2
3
3
−6
2
3
9 = 0
0
0
3
7
−3
8
−7
0
−2
0
8
−1
5
5
2
7
0
−4
8
2
13
9.
=
1
−1
−3
0
0
1
5
4
0
0
1
2
5
7
5
3
1
3
−1
0
−2
0
2
−4
−1
−6
− 0
0
−4
7
0
0
0
3
5
0
0
0
0
1
=
1
−1
−3
0
1
−1
−3
0
0
1
5
4
0
1
5
4
0
0
0
0
0
−3
1
−5
0
0
0
0
−3
0
−5
1
3
−1
0
−2
0
2
−4
−1
−6
5 = 0
0
0
3
0
0
−4
7
−7
0
0
0
0
1
=−
= −(−3) = 3
5 =
−7 = −(−24) = 24
11. First use a row replacement to create zeros in the second column, and then expand down the second
column:
2
5
−3
−1
2
5
−3
−1
3
0
1
−3
3
0
1
−3
−6
4
0
10
−4
−4
9
−1
−6
0
0
0
−4
2
9
1
=
3
1
−3
= −5 −6
0
−4
2
9
1
Now use a row replacement to create zeros in the first column, and then expand down the first
column:
3
1
−5 −6
−4
0
2
−3
3
1
9 = −5 0
−2
1
0
2
−3
3 = (−5)(3)
−2
3
2
1
1
= (−5)(3)(−8) = 120
12. First use a row replacement to create zeros in the fourth column, and then expand down the fourth
column:
−1
2
3
0
−1
2
3
0
3
4
3
0
3
4
3
0
5
4
4
2
6
4
6
3
−3
4
0
2
−2
4
0
3
=
−1
2
3
=3 3
−3
4
0
3
−2
Now use a row replacement to create zeros in the first column, and then expand down the first
−1 2
3
−1
2
3
10
12
3 = 3 0 10
12 = 3(−1)
= 3(−1)(−38) = 114
column: 3 3 4
−6 −11
−3 0 −2
0 −6 −11
13. First use a row replacement to create zeros in the fourth column, and then expand down the fourth
column:
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3.2
2
5
4
1
2
5
4
1
4
7
6
2
0
−3
−2
0
6
−6
−2
7
−4
7
0
0
6
−6
−2
7
−4
7
0
0
=
0
−3
−2
= −1 6
−6
−2
7
−4
7
• Solutions
Now use a row replacement to create zeros in the first column, and then expand down the first
0 −3 −2
0 −3 −2
−3 −2
= (−1)(−6)(1) = 6
column: −1 6 −2 −4 = −1 6 −2 −4 = (−1)(−6)
5
3
−6
7
7
0
5
3
14. First use a row replacement to create zeros in the third column, and then expand down the third
column:
−3
−2
1
−4
−3
−2
1
−4
1
3
0
−3
1
3
0
−3
−3
3
4
−4
−2
0
8
4
−9
3
0
−4
0
0
0
4
=
1
3
−3
= 1 −9
3
0
−4
0
4
Now expand along the second row:
1
3
1 −9
0
3
−4
−3
0 = 1( −( −9))
4
3
−3
−4
4
= (1)(9)(0) = 0
a
d
b
e
c
a
f =5 d
b
e
c
f = 5(7) = 35
5g
5h
5i
g
h
i
a
16. 3d
b
3e
c
a
3f =3 d
b
e
c
f = 3(7) = 21
g
h
h
i
15.
i
g
a
17. g
b
h
c
a
i =− d
b
e
c
f = −7
d
e
f
g
h
i
g
18. a
d
h
b
e
i
a
c =− g
f
d
b
h
e
c
§ a
¨
i = −¨ − d
¨ g
f
©
a
19. 2d + a
g
b
2e + b
h
c
a
2 f + c = 2d
i
g
b
2e
h
b
e
h
c
f
i
·
¸
¸ = −(−7) = 7
¸
¹
c
a
2f =2 d
i
g
b
e
c
f = 2(7) = 14
h
i
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
177
178
CHAPTER 3
a+d
20.
d
• Determinants
b+e
e
g
c+ f
a
f = d
h
i
g
b
e
c
f =7
h
i
2
3
0
21. Since 1
3
4 = −1 ≠ 0 , the matrix is invertible.
1
2
1
−1
5
0
22. Since 1
−3
0
5
3
2
0
0
8
1
−7
−5
0
3
0
8
7
6
5
0
4
4
−7
−3
6
0
−7
2
6
7
−8
7
25. Since −4
5
−6
7
−5
3
2
−2
0
5
−6
−1
0
−6
4
0
7
3
0
0
−3
23. Since
24. Since
26. Since
−2 = 0 , the matrix is not invertible.
= 0 , the matrix is not invertible.
−5 = 11 ≠ 0 , the columns of the matrix form a linearly independent set.
0 = −1 ≠ 0 , the columns of the matrix form a linearly independent set.
= 0 , the columns of the matrix form a linearly dependent set.
27. a.
b.
c.
d.
True. See Theorem 3.
True. See the paragraph following Example 2.
True. See the paragraph following Theorem 4.
False. See the warning following Example 5.
28. a.
b.
c.
d.
True. See Theorem 3.
False. See the paragraphs following Example 2.
False. See Example 3.
False. See Theorem 5.
29. By Theorem 6, det B 5 = (det B )5 = (−2)5 = −32 .
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3.2
• Solutions
179
30. Suppose the two rows of a square matrix A are equal. By swapping these two rows, the matrix A is
not changed so its determinant should not change. But since swapping rows changes the sign of the
determinant, det A = – det A. This is only possible if det A = 0. The same may be proven true for
columns by applying the above result to AT and using Theorem 5.
31. By Theorem 6, (det A)(det A−1 ) = det I = 1 , so det A−1 = 1/ det A.
32. By factoring an r out of each of the n rows, det (rA) = r n det A.
33. By Theorem 6, det AB = (det A)(det B) = (det B)(det A) = det BA.
34. By Theorem 6 and Exercise 31,
det ( PAP −1 ) = (det P )(det A)(det P −1 ) = (det P )(det P −1 )(det A)
§ 1 ·
= (det P) ¨
¸ (det A) = 1det A
© det P ¹
= det A
35. By Theorem 6 and Theorem 5, det U T U = (det U T )(det U ) = (det U ) 2 . Since U T U = I ,
det U T U = det I = 1 , so (det U )2 = 1. Thus det U = ±1.
36. By Theorem 6 det A4 = (det A)4 . Since det A4 = 0 , then (det A)4 = 0 . Thus det A = 0, and A is not
invertible by Theorem 4.
ª 6
37. One may compute using Theorem 2 that det A = 3 and det B = 8, while AB = «
¬17
det AB = 24 = 3 × 8 = (det A)(det B).
ª 6
38. One may compute that det A = 0 and det B = –2, while AB = «
¬ −2
det AB = 0 = 0 × –2 = (det A)(det B).
0º
. Thus
0 »¼
39. a. By Theorem 6, det AB = (det A)(det B) = 4 × –3 = –12.
b. By Exercise 32, det 5 A = 53 det A = 125 × 4 = 500 .
c. By Theorem 5, det BT = det B = −3 .
d. By Exercise 31, det A−1 = 1/ det A = 1/ 4 .
e. By Theorem 6, det A3 = (det A)3 = 43 = 64 .
40. a. By Theorem 6, det AB = (det A)(det B) = –1 × 2 = –2.
b. By Theorem 6, det B 5 = (det B )5 = 25 = 32 .
c. By Exercise 32, det 2 A = 24 det A = 16 × −1 = −16 .
d. By Theorems 5 and 6, det AT A = (det AT )(det A) = (det A)(det A) = −1 × −1 = 1 .
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0º
. Thus
4 »¼
180
CHAPTER 3
• Determinants
e. By Theorem 6 and Exercise 31,
det B −1 AB = (det B −1 )(det A)(det B ) = (1/ det B )(det A)(det B) = det A = −1 .
41. det A = (a + e)d – c(b + f) = ad + ed – bc – cf = (ad – bc) + (ed – cf) = det B + det C.
1+ a
b
= (1 + a)(1 + d ) − cb = 1 + a + d + ad − cb = det A + a + d + det B , so
c
1+ d
det (A + B) = det A + det B if and only if a + d = 0.
42. det ( A + B) =
43. Compute det A by using a cofactor expansion down the third column:
det A = (u1 + v1 )det A13 − (u2 + v2 )det A23 + (u3 + v3 )det A33
= u1det A13 − u2 det A23 + u3det A33 + v1det A13 − v2 det A23 + v3 det A33
= det B + det C
44. By Theorem 5, det AE = det ( AE )T . Since ( AE )T = E T AT , det AE = det( E T AT ). Now E T is itself
an elementary matrix, so by the proof of Theorem 3, det ( E T AT ) = (det E T )(det AT ). Thus it is true
that det AE = (det E T )(det AT ), and by applying Theorem 5, det AE = (det E)(det A).
45. [M] Answers will vary, but will show that det AT A always equals 0 while det AAT should seldom
be zero. To see why AT A - should not be invertible (and thus det AT A = 0 ), let A be a matrix with
more columns than rows. Then the columns of A must be linearly dependent, so the equation Ax = 0
must have a non-trivial solution x. Thus ( AT A)x = AT ( Ax) = AT 0 = 0, and the equation ( AT A)x = 0
has a
non-trivial solution. Since AT A is a square matrix, the Invertible Matrix Theorem now says that
AT A is not invertible. Notice that the same argument will not work in general for AAT , since AT
has more rows than columns, so its columns are not automatically linearly dependent.
46. [M] One may compute for this matrix that det A = –4008 and cond A ≈ 16.3. Note that this is the A 2
condition number, which is used in Section 2.3. Since det A ≠ 0, it is invertible and
ª −837
«
1 « −750
A−1 = −
4008 « 171
«
¬ 21
−181
−207
−574
30
195
−187
−87
−81
297 º
654 »»
−1095»
»
639 ¼
The determinant is very sensitive to scaling, as det10 A = 104 det A = −40, 080,000 and
det 0.1A = (0.1) 4 det A = −0.4008. The condition number is not changed at all by scaling:
cond(10A) = cond(0.1A) = cond A ≈ 16.3. When A = I 4 , det A=1 and cond A = 1. As before the
determinant is sensitive to scaling: det10 A = 104 det A = 10, 000 and det 0.1A = (0.1) 4 det A = 0.0001.
Yet the condition number is not changed by scaling: cond(10A) = cond(0.1A) = cond A = 1.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3.3
3.3
• Solutions
181
SOLUTIONS
Notes: This section features several independent topics from which to choose. The geometric
interpretation of the determinant (Theorem 10) provides the key to changes of variables in multiple
integrals. Students of economics and engineering are likely to need Cramer’s Rule in later courses.
Exercises 1–10 concern Cramer’s Rule, exercises 11–18 deal with the adjugate, and exercises 19–32
cover the geometric interpretation of the determinant. In particular, Exercise 25 examines students’
understanding of linear independence and requires a careful explanation, which is discussed in the Study
Guide. The Study Guide also contains a heuristic proof of Theorem 9 for 2 × 2 matrices.
ª5 7 º
ª3º
and b = « » . We compute
1. The system is equivalent to Ax = b, where A = «
»
¬2 4¼
¬1 ¼
ª3 7 º
ª 5 3º
A1 (b) = «
, A2 (b) = «
»
» , det A = 6, det A1 (b ) = 5, det A2 (b) = −1,
¬1 4 ¼
¬ 2 1¼
x1 =
det A1 (b) 5
det A2 (b)
1
= , x2 =
=− .
det A
6
det A
6
ª4 1º
ª6º
and b = « » . We compute
2. The system is equivalent to Ax = b, where A = «
»
¬5 2¼
¬7 ¼
ª6 1 º
ª4 6º
A1 (b) = «
, A2 (b) = «
»
» , det A = 3, det A1 (b ) = 5, det A2 (b) = −2,
¬7 2¼
¬5 7 ¼
x1 =
det A1 (b) 5
det A2 (b)
2
= , x2 =
=− .
det A
3
det A
3
ª 3 −2 º
ª 7º
and b = « » . We compute
3. The system is equivalent to Ax = b, where A = «
»
6¼
¬ −5
¬ −5 ¼
7º
ª 7 −2 º
ª 3
A1 (b) = «
, A2 (b) = «
»
» , det A = 8, det A1 (b) = 32, det A2 (b) = 20,
6¼
¬ −5
¬ −5 −5 ¼
x1 =
det A1 (b) 32
det A2 (b) 20 5
=
= 4, x2 =
=
= .
det A
8
det A
8 2
3º
ª −5
ª 9º
and b = « » . We compute
4. The system is equivalent to Ax = b, where A = «
»
¬ 3 −1¼
¬ −5 ¼
3º
9º
ª 9
ª −5
A1 (b) = «
, A2 (b) = «
»
» , det A = −4, det A1 (b) = 6, det A2 (b) = −2,
¬ −5 −1¼
¬ 3 −5 ¼
x1 =
det A1 (b) 6
3
det A2 (b) −2 1
=
= − , x2 =
=
= .
−4
−4 2
det A
2
det A
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
182
CHAPTER 3
• Determinants
ª 2
5. The system is equivalent to Ax = b, where A = «« −3
«¬ 0
ª 7
A1 (b) = «« −8
«¬ −3
1
0
1
0º
ª 2
»
1» , A2 (b) = «« −3
«¬ 0
2 »¼
7
−8
−3
0º
ª 7º
»
0 1» and b = «« −8»» . We compute
«¬ −3»¼
1 2 »¼
0º
7º
ª 2 1
»
«
1» , A3 (b ) = « −3 0 −8»» ,
«¬ 0 1 −3»¼
2 »¼
1
det A = 4,det A1 (b ) = 6, det A2 (b ) = 16,det A3 (b ) = −14,
x1 =
det A3 (b) −14
det A1 (b) 6 3
det A2 (b) 16
7
= = , x2 =
= = 4, x3 =
=
=− .
det A
4 2
det A
4
det A
4
2
ª 2
6. The system is equivalent to Ax = b, where A = «« −1
«¬ 3
ª 4
A1 (b) = «« 2
«¬ −2
1
0
1
1º
ª 2
»
2 » , A2 (b) = «« −1
«¬ 3
3»¼
4
2
−2
1º
ª 4º
»
0 2 » and b = «« 2 »» . We compute
«¬ −2 »¼
1 3»¼
1º
4º
ª 2 1
»
«
2 » , A3 (b) = « −1 0
2 »» ,
«¬ 3 1 −2 »¼
3»¼
1
det A = 4, det A1 (b ) = −16, det A2 (b ) = 52, det A3 (b ) = −4,
x1 =
det A3 (b) −4
det A1 (b) −16
det A2 (b) 52
=
= −4, x2 =
=
= 13, x3 =
=
= −1.
det A
4
det A
4
det A
4
4º
ª6s
ª 5º
7. The system is equivalent to Ax = b, where A = «
and b = « » . We compute
»
¬ 9 2s ¼
¬ −2 ¼
4º
5º
ª 5
ª6s
A1 (b ) = «
, A2 (b ) = «
»
» , det A1 (b ) = 10 s + 8, det A2 (b ) = −12 s − 45.
¬ −2 2 s ¼
¬ 9 −2 ¼
Since det A = 12 s 2 − 36 = 12( s 2 − 3) ≠ 0 for s ≠ ± 3 , the system will have a unique solution when
s ≠ ± 3 . For such a system, the solution will be
det A1 (b)
10s + 8
5s + 4
det A2 (b) −12s − 45 −4s − 15
=
=
=
=
x1 =
, x2 =
.
2
2
det A
det A
12( s − 3) 6( s − 3)
12( s 2 − 3) 4( s 2 − 3)
ª 3 s −5 º
ª3º
and b = « » . We compute
8. The system is equivalent to Ax = b, where A = «
»
¬ 9 5s ¼
¬2¼
ª 3 −5 º
ª 3s 3º
A1 (b ) = «
, A2 (b ) = «
»
» , det A1 (b ) = 15s + 10, det A2 (b ) = 6 s − 27.
¬ 2 5s ¼
¬ 9 2¼
Since det A = 15s 2 + 45 = 15( s 2 + 3) ≠ 0 for all values of s, the system will have a unique solution for
all values of s. For such a system, the solution will be
det A1 (b) 15s + 10
3s + 2
det A2 (b)
6 s − 27
2s − 9
=
=
=
=
x1 =
, x2 =
.
2
2
2
det A
det A
15( s + 3) 3( s + 3)
15( s + 3) 5( s 2 + 3)
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3.3
• Solutions
183
ª s −2 s º
ª −1º
9. The system is equivalent to Ax = b, where A = «
and b = « » . We compute
»
6s ¼
¬3
¬ 4¼
ª −1 −2 s º
ª s −1º
A1 (b ) = «
, A2 (b ) = «
, det A1 (b ) = 2 s, det A2 (b ) = 4 s + 3.
»
6s ¼
4 »¼
¬ 4
¬3
Since det A = 6 s 2 + 6s = 6s ( s + 1) = 0 for s = 0, –1, the system will have a unique solution when s ≠ 0,
–1. For such a system, the solution will be
det A1 (b)
2s
1
det A2 (b)
4s + 3
=
=
=
x1 =
, x2 =
.
det A
6s ( s + 1) 3( s + 1)
det A
6s ( s + 1)
1º
ª2s
ª1 º
and b = « » . We compute
10. The system is equivalent to Ax = b, where A = «
»
¬ 3s 6 s ¼
¬2¼
1º
ª1
ª 2 s 1º
A1 (b ) = «
, A2 (b ) = «
»
» , det A1 (b ) = 6 s − 2, det A2 (b ) = s.
¬2 6s ¼
¬ 3s 2 ¼
Since det A = 12 s 2 − 3s = 3s(4 s − 1) = 0 for s = 0,1/4, the system will have a unique solution when
s ≠ 0,1/4. For such a system, the solution will be
s
det A1 (b)
6s − 2
det A2 (b)
1
=
=
=
x1 =
, x2 =
.
det A
3s (4s − 1)
det A
3s (4 s − 1) 3(4 s − 1)
11. Since det A = 3 and the cofactors of the given matrix are
C11 =
0
0
1
1
C21 = −
C31 =
= 0,
−2
−1
1
1
−2
−1
0
0
ª 0
adj A = «« −3
¬« 3
C12 = −
= 1,
= 0,
C22 =
3
0
−1
1
0
−1
−1
1
0
−1
3
0
C32 = −
−1
2
0
−1
1
C13 =
= −1,
C23 = −
= −3,
C33 =
0º
ª 0
1
»
−1
−3» and A =
adj A = «« −1
det A
6 ¼»
¬« 1
1
3
= −3,
= 3,
0
−2
−1
1
0
−2
3
0
= 2,
= 6,
0º
−1»» .
2 ¼»
1/ 3
−1/ 3
2/3
12. Since det A = 5 and the cofactors of the given matrix are
C11 =
−2
1
1
0
1
3
1
0
1
3
−2
1
C21 = −
C31 =
ª −1
adj A = «« 0
«¬ 2
3
0
−1
= −1,
C12 = −
= 3,
C22 =
= 7,
C32 = −
2
1
0
0
1
3
0
0
= 0,
= 0,
1
3
2
1
= 5,
C13 =
2
−2
0
1
C23 = −
C33 =
7º
ª −1/ 5
1
»
−1
5» and A =
adj A = ««
0
det A
«¬ 2 / 5
−4 »¼
1 1
0
1
1
1
2
−2
3/ 5
0
−1/ 5
= 2,
= −1,
= −4,
7 / 5º
1»» .
−4 / 5»¼
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
184
CHAPTER 3
• Determinants
13. Since det A = 6 and the cofactors of the given matrix are
C11 =
0
1
1 1
C21 = −
C31 =
= −1,
5
4
1
1
5
4
0
1
ª −1
adj A = «« 1
«¬ 1
= −1,
= 5,
−1
1 1
C12 = −
C22 =
2
1
3
4
2
1
C32 = −
= 1,
= −5,
3
4
1
1
= 1,
C13 =
7
0
2
1
C23 = −
C33 =
5º
ª −1/ 6
1
»
−1
1» and A =
adj A = «« 1/ 6
det A
«¬ 1/ 6
−5»¼
−5
1
= 1,
3
5
2
1
3
5
1
0
= 7,
= −5,
−1/ 6
5 / 6º
1/ 6 »» .
−5 / 6 »¼
−5 / 6
7/6
14. Since det A = –1 and the cofactors of the given matrix are
C11 =
2
1
3
4
C21 = −
C31 =
= 5,
6
7
3
4
6
7
2
1
= −8,
−3
ª 5
adj A = «« 2
«¬ −4
= −3,
C12 = −
C22 =
0
1
2
4
3
7
2
4
C32 = −
= 2,
= −2,
3
7
0
1
= −3,
C13 =
3
2
2
3
C23 = −
C33 =
−8º
ª −5
1
»
−1
−3» and A =
adj A = «« −2
det A
«¬ 4
6 »¼
−2
0
3
2
−3
= −4,
3
6
2
3
3
6
0
2
= 3,
= 6,
8º
3»» .
−6 »¼
15. Since det A = 6 and the cofactors of the given matrix are
C11 =
1
0
3
2
C21 = −
C31 =
= 2,
0
0
3
2
0
0
1
0
ª 2
adj A = «« 2
«¬ −1
= 0,
= 0,
C12 = −
C22 =
−1
0
−2
2
3
0
−2
2
C32 = −
= 2,
= 6,
3
0
−1
0
= 0,
C13 =
6
−9
1
−2
3
C23 = −
C33 =
0º
ª 1/ 3
1
»
−1
0 » and A =
adj A = «« 1/ 3
det A
«¬ −1/ 6
3»¼
0
−1
= −1,
3
0
−2
3
3
0
−1
1
= −9,
= 3,
0º
1
0 »» .
−3/ 2 1/ 2 »¼
0
16. Since det A = –9 and the cofactors of the given matrix are
C11 =
−3
1
0
3
2
4
0
3
C21 = −
= −9,
C12 = −
= −6,
C22 =
0
1
0
3
1
4
0
3
= 0,
= 3,
C13 =
0
−3
0
0
C23 = −
1
2
0
0
= 0,
= 0,
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3.3
C31 =
ª −9
adj A = «« 0
«¬ 0
2
4
−3
1
−6
3
0
= 14,
C32 = −
1
4
0
1
= −1,
14 º
ª1
1
»
−1
−1» and A =
adj A = «« 0
det A
«¬ 0
−3»¼
C33 =
2/3
−1/ 3
0
1
2
0
−3
• Solutions
185
= −3,
−14 / 9 º
1/ 9 »» .
1/ 3»¼
ªa
17. Let A = «
¬c
C22
bº
. Then the cofactors of A are C11 = d = d , C12 = − c = −c, C21 = − b = −b , and
d »¼
ª d −b º
= a = a . Thus adj A = «
. Since det A = ad – bc, Theorem 8 gives that
a »¼
¬ −c
A−1 =
1
1 ª d
adj A =
det A
ad − bc «¬ −c
−b º
. This result is identical to that of Theorem 4 in Section 2.2.
a »¼
18. Each cofactor of A is an integer since it is a sum of products of entries in A. Hence all entries in adj A
will be integers. Since det A = 1, the inverse formula in Theorem 8 shows that all the entries in A−1
will be integers.
ª5
19. The parallelogram is determined by the columns of A = «
¬2
|det A| = |8| = 8.
ª −1
20. The parallelogram is determined by the columns of A = «
¬ 3
is |det A| = |–7| = 7.
6º
, so the area of the parallelogram is
4 »¼
4º
, so the area of the parallelogram
−5»¼
21. First translate one vertex to the origin. For example, subtract (–1, 0) from each vertex to get a new
parallelogram with vertices (0, 0),(1, 5),(2, –4), and (3, 1). This parallelogram has the same area as
2º
ª1
the original, and is determined by the columns of A = «
» , so the area of the parallelogram is
¬ 5 −4 ¼
|det A| = |–14| = 14.
22. First translate one vertex to the origin. For example, subtract (0, –2) from each vertex to get a new
parallelogram with vertices (0, 0),(6, 1),(–3, 3), and (3, 4). This parallelogram has the same area as
ª 6 −3 º
, so the area of the parallelogram is
the original, and is determined by the columns of A = «
3»¼
¬1
|det A| = |21| = 21.
ª 1
23. The parallelepiped is determined by the columns of A = «« 0
«¬ −2
parallelepiped is |det A| = |22| = 22.
1
2
4
7º
1»» , so the volume of the
0 »¼
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
186
CHAPTER 3
• Determinants
ª1
24. The parallelepiped is determined by the columns of A = «« 4
«¬ 0
−2
−5
2
−1º
2 »» , so the volume of the
−1»¼
parallelepiped is |det A| = |–15| = 15.
25. The Invertible Matrix Theorem says that a 3 × 3 matrix A is not invertible if and only if its columns
are linearly dependent. This will happen if and only if one of the columns is a linear combination of
the others; that is, if one of the vectors is in the plane spanned by the other two vectors. This is
equivalent to the condition that the parallelepiped determined by the three vectors has zero volume,
which is in turn equivalent to the condition that det A = 0.
26. By definition, p + S is the set of all vectors of the form p + v, where v is in S. Applying T to a typical
vector in p + S, we have T(p + v) = T(p) + T(v). This vector is in the set denoted by T(p) + T(S). This
proves that T maps the set p + S into the set T(p) + T(S). Conversely, any vector in T(p) + T(S) has
the form T(p) + T(v) for some v in S. This vector may be written as T(p + v). This shows that every
vector in T(p) + T(S) is the image under T of some point p + v in p + S.
ª −2 −2 º
27. Since the parallelogram S is determined by the columns of «
, the area of S is
5»¼
¬ 3
6 −2
ª −2 −2 º
det «
= | −4 | = 4. The matrix A has det A =
= 6 . By Theorem 10, the area of T(S)
»
5¼
−3
2
¬ 3
is |det A|{area of S} = 6 ⋅ 4 = 24. Alternatively, one may compute the vectors that determine the
image, namely, the columns of
A [b1
ª 6
b2 ] = «
¬ −3
−2 º ª −2
2 »¼ «¬ 3
−2 º ª −18
=
5 »¼ «¬ 12
−22 º
16 »¼
The determinant of this matrix is –24, so the area of the image is 24.
ª 4 0º
28. Since the parallelogram S is determined by the columns of «
» , the area of S is
¬ −7 1¼
7 2
ª 4 0º
det «
= | 4 | = 4 . The matrix A has det A =
= 5 . By Theorem 10, the area of T(S) is
»
1 1
¬ −7 1¼
|det A|{area of S} =5 ⋅ 4 = 20. Alternatively, one may compute the vectors that determine the image,
namely, the columns of
A [b1
ª7
b2 ] = «
¬1
2º ª 4
1»¼ «¬ −7
0 º ª 14
=
1»¼ «¬ −3
2º
1»¼
The determinant of this matrix is 20, so the area of the image is 20.
29. The area of the triangle will be one half of the area of the parallelogram determined by v1 and v 2 .
By Theorem 9, the area of the triangle will be (1/2)|det A|, where A = [ v1
v 2 ].
30. Translate R to a new triangle of equal area by subtracting ( x3 , y3 ) from each vertex. The new triangle
has vertices (0, 0), ( x1 − x3 , y1 − y3 ) , and ( x2 − x3 , y2 − y3 ). By Exercise 29, the area of the triangle
will be
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3.3
ªx − x
1
det « 1 3
2
¬ y1 − y3
• Solutions
187
x2 − x3 º
.
y2 − y3 »¼
Now consider using row operations and a cofactor expansion to compute the determinant in the
formula:
ª x1
«
det « x2
«¬ x3
y1
y2
y3
1º
ª x1 − x3
»
«
1» = det « x2 − x3
«¬ x3
1»¼
y1 − y3
0º
ªx − x
0 »» = det « 1 3
¬ x2 − x3
1 »¼
y2 − y3
y3
y1 − y3 º
y2 − y3 »¼
By Theorem 5,
ªx − x
det « 1 3
¬ x2 − x3
y1 − y3 º
x2 − x3 º
ªx − x
= det « 1 3
»
»
y 2 − y3 ¼
¬ y1 − y3 y2 − y3 ¼
So the above observation allows us to state that the area of the triangle will be
ªx − x
1
det « 1 3
2
¬ y1 − y3
ª x1
x2 − x3 º 1
= det «« x2
»
y2 − y3 ¼ 2
«¬ x3
y1
y2
y3
1º
1»»
1»¼
ª u1 º
x12 x22 x32
31. a. To show that T(S) is bounded by the ellipsoid with equation 2 + 2 + 2 = 1 , let u = ««u2 »» and
a
b
c
«¬ u3 »¼
ª x1 º
let x = «« x2 »» = Au . Then u1 = x1 / a , u2 = x2 / b , and u3 = x3 / c , and u lies inside S (or
«¬ x3 »¼
u12 + u22 + u32 ≤ 1 ) if and only if x lies inside T(S) (or
x12 x22 x32
+ +
≤ 1 ).
a 2 b2 c 2
b. By the generalization of Theorem 10,
{volume of ellipsoid} = {volume of T ( S )}
= | det A | ⋅ {volume of S} = abc
4π 4πabc
=
3
3
32. a. A linear transformation T that maps S onto S ′ will map e1 to v1 , e 2 to v 2 , and e3 to v 3 ; that is,
T (e1 ) = v1 , T (e 2 ) = v 2 , and T (e3 ) = v 3 . The standard matrix for this transformation will be
A = [T (e1 )
T (e 2 )
T (e3 )] = [ v1
v2
v 3 ].
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
188
CHAPTER 3
• Determinants
b. The area of the base of S is (1/2)(1)(1) = 1/2, so the volume of S is (1/3)(1/2)(1) = 1/6. By part a.
T(S) = S ′ , so the generalization of Theorem 10 gives that the volume of S ′ is |det A|{volume of
S} = (1/6)|det A|.
33. [M] Answers will vary. In MATLAB, entries in B – inv(A) are approximately 10 −15 or smaller.
34. [M] Answers will vary, as will the commands which produce the second entry of x. For example, the
MATLAB command is x2 = det([A(:,1) b A(:,3:4)])/det(A) while the Mathematica
command is x2 = Det[{Transpose[A][[1]],b,Transpose[A][[3]],
Transpose[A][[4]]}]/Det[A].
35. [M] MATLAB Student Version 4.0 uses 57,771 flops for inv A and 14,269,045 flops for the inverse
formula. The inv(A) command requires only about 0.4% of the operations for the inverse formula.
Chapter 3 SUPPLEMENTARY EXERCISES
1. a. True. The columns of A are linearly dependent.
b. True. See Exercise 30 in Section 3.2.
c. False. See Theorem 3(c); in this case det 5 A = 53 det A .
ª2
d. False. Consider A = «
¬0
0º
ª1
, B=«
»
1¼
¬0
0º
ª3
, and A + B = «
»
3¼
¬0
0º
.
4 »¼
False. By Theorem 6, det A3 = 23 .
False. See Theorem 3(b).
True. See Theorem 3(c).
True. See Theorem 3(a).
False. See Theorem 5.
False. See Theorem 3(c); this statement is false for n × n invertible matrices with n an even
integer.
k. True. See Theorems 6 and 5; det AT A = (det A)2 .
l. False. The coefficient matrix must be invertible.
m. False. The area of the triangle is 5.
n. True. See Theorem 6; det A3 = (det A)3 .
o. False. See Exercise 31 in Section 3.2.
p. True. See Theorem 6.
e.
f.
g.
h.
i.
j.
12
13
14
12
13
2. 15
16
17 = 3
3
3 =0
18
19
20
6
6
a
b+c
6
1
a
b+c
3. 1
b
a+c = 0
1
c
a+b
1
0
14
b+c
1
a
b−a
a − b = (b − a )(c − a ) 0
1
−1 = 0
c−a
a−c
1
−1
0
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 3
a
b
c
a
b
c
a
b
c
4. a + x
b+x
c+x = x
x
x = xy 1
1
1 =0
a+ y
b+ y
c+ y
y
y
1
1
1
9
1
9
9
9
9
0
9
9
2
5. 4
9
6
0
0
5
0 = (−1)
0
3
9
0
0
0
7
0
= ( −1)( −2)(3)
4
5
6
7
y
9
9
9
2
4
0
5
0
9
3
9
0
6
0
7
0
• Supplementary Exercises
4
0
5
= (−1)(−2) 9
3
9
6
0
7
4
8
5
= (1)(2) 6
8
7 = (1)(2)(−3)
0
3
0
189
= ( −1)( −2)(3)( −2) = −12
4
8
8
8
5
0
1
0
0
0
6. 6
0
0
8
8
8
7 = (1)
8
8
3
0
8
2
0
0
4
8
8
5
6
8
8
7
0
8
3
0
0
2
0
0
7. Expand along the first row to obtain
1 x
y
x
y1
1 y1
1
1 x1 y1 = 1 1
−x
+y
x2 y2
1 y2
1
1 x2 y2
x1
x2
4
5
6
7
= (1)(2)(−3)(−2) = 12
= 0.
This is an equation of the form ax + by + c = 0, and since the points ( x1 , y1 ) and ( x2 , y2 ) are distinct,
at least one of a and b is not zero. Thus the equation is the equation of a line. The points ( x1 , y1 ) and
( x2 , y2 ) are on the line, because when the coordinates of one of the points are substituted for x and y,
two rows of the matrix are equal and so the determinant is zero.
8. Expand along the first row to obtain
1 x
y
x
y1
1 y1
1
1 x1 y1 = 1 1
−x
+y
1 m
0 m
0
0 1 m
x1
1
= 1(mx1 − y1 ) − x(m) + y (1) = 0. This equation may
be rewritten as mx1 − y1 − mx + y = 0, or y − y1 = m ( x − x1 ).
1
a
a2
a
a2
9. det T = 1
b
b2 = 0
b−a
b2 − a2
1
c
c2
c−a
c2 − a2
1
0
1
a
= (b − a )(c − a ) 0
0
1
1
a2
1
= 0
0
a2
(b − a )(b + a )
(c − a )(c + a )
a
b−a
c−a
1
a
b + a = (b − a )(c − a ) 0
c+a
0
1
0
a2
b + a = (b − a )(c − a )(c − b)
c−b
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
190
CHAPTER 3
• Determinants
10. Expanding along the first row will show that f (t ) = det V = c0 + c1t + c2t 2 + c3t 3 . By Exercise 9,
1
x1
x12
c3 = − 1
x2
x22 = ( x2 − x1 )( x3 − x1 )( x3 − x2 ) ≠ 0
1
x3
x32
since x1 , x2 , and x3 are distinct. Thus f (t) is a cubic polynomial. The points ( x1 ,0) , ( x2 ,0) , and
( x3 ,0) are on the graph of f, since when any of x1 , x2 or x3 are substituted for t, the matrix has two
equal rows and thus its determinant (which is f (t)) is zero. Thus f ( xi ) = 0 for i = 1, 2, 3.
11. To tell if a quadrilateral determined by four points is a parallelogram, first translate one of the
vertices to the origin. If we label the vertices of this new quadrilateral as 0, v1 , v 2 , and v 3 , then
they will be the vertices of a parallelogram if one of v1 , v 2 , or v 3 is the sum of the other two. In
this example, subtract (1, 4) from each vertex to get a new parallelogram with vertices 0 = (0, 0),
v1 = ( −2,1) , v 2 = (2,5) , and v 3 = (4, 4) . Since v 2 = v 3 + v1 , the quadrilateral is a parallelogram as
stated. The translated parallelogram has the same area as the original, and is determined by the
ª −2 4 º
columns of A = [ v1 v 3 ] = «
» , so the area of the parallelogram is |det A| = |–12| = 12.
¬ 1 4¼
12. A 2 × 2 matrix A is invertible if and only if the parallelogram determined by the columns of A has
nonzero area.
13. By Theorem 8, (adj A) ⋅
and (adj A)−1 =
1
A = A−1 A = I . By the Invertible Matrix Theorem, adj A is invertible
det A
1
A.
det A
ªA Oº
14. a. Consider the matrix Ak = «
» , where 1 ≤ k ≤ n and O is an appropriately sized zero matrix.
¬O I k ¼
We will show that det Ak = det A for all 1 ≤ k ≤ n by mathematical induction.
First let k = 1. Expand along the last row to obtain
ª A Oº
( n +1) + ( n +1)
det A1 = det «
⋅ 1 ⋅ det A = det A.
» = ( −1)
1
O
¬
¼
Now let 1 < k ≤ n and assume that det Ak −1 = det A. Expand along the last row of Ak to obtain
ªA Oº
det Ak = det «
= ( −1) ( n + k ) + ( n + k ) ⋅ 1 ⋅ det Ak −1 = det Ak −1 = det A. Thus we have proven the
»
¬O I k ¼
result, and the determinant of the matrix in question is det A.
Oº
ªI
b. Consider the matrix Ak = « k
» , where 1 ≤ k ≤ n, Ck is an n × k matrix and O is an
¬C k D ¼
appropriately sized zero matrix. We will show that det Ak = det D for all 1 ≤ k ≤ n by
mathematical induction.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 3
• Supplementary Exercises
191
First let k = 1. Expand along the first row to obtain
ª 1 Oº
= ( −1)1+1 ⋅ 1 ⋅ det D = det D.
det A1 = det «
»
¬C1 D ¼
Now let 1 < k ≤ n and assume that det Ak −1 = det D. Expand along the first row of Ak to obtain
Oº
ªI
1+1
det Ak = det « k
» = ( −1) ⋅ 1 ⋅ det Ak −1 = det Ak −1 = det D. Thus we have proven the result, and
C
D
¬ k
¼
the determinant of the matrix in question is det D.
c. By combining parts a. and b., we have shown that
ª A Oº §
ª A O º ·§
ª I Oº·
det «
= ¨ det «
¸¨ det «
»
»
» ¸ = (det A)(det D ).
¬C D ¼ ©
¬O I ¼ ¹©
¬C D ¼ ¹
From this result and Theorem 5, we have
T
ª AT
Oº
ªA Bº
ªA Bº
T
T
=
=
det «
det
det
« T
» = (det A )(det D ) = (det A)(det D ).
»
«
»
T
O
D
O
D
D ¼»
¬
¼
¬
¼
¬« B
15. a. Compute the right side of the equation:
ªI
«X
¬
Oº ª A
I »¼ «¬O
Bº ª A
=
Y »¼ «¬ XA
º
XB + Y »¼
B
Set this equal to the left side of the equation:
ªA
«C
¬
Bº ª A
=
D »¼ «¬ XA
º
so that XA = C
XB + Y »¼
B
XB + Y = D
Since XA = C and A is invertible, X = CA−1. Since XB + Y = D, Y = D − XB = D − CA−1B . Thus
by Exercise 14(c),
ªA
det «
¬C
ª I
Bº
= det « −1
»
D¼
¬CA
Oº
ªA
» det «
I¼
¬O
º
»
D − CA B ¼
B
−1
= (det A)(det ( D − CA−1 B ))
b. From part a.,
ªA
det «
¬C
Bº
= (det A)(det ( D − CA−1 B )) = det[ A( D − CA−1B )]
D »¼
= det[ AD − ACA−1 B ] = det[ AD − CAA−1 B]
= det[ AD − CB ]
16. a. Doing the given operations does not change the determinant of A since the given operations are
all row replacement operations. The resulting matrix is
ªa − b
« 0
«
« 0
«
« #
«¬ b
−a + b
0
…
a−b
−a + b
…
0
a−b
…
#
b
#
b
%
…
0º
0 »»
0»
»
#»
a »¼
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
192
CHAPTER 3
• Determinants
b. Since column replacement operations are equivalent to row operations on AT and det AT = det A ,
the given operations do not change the determinant of the matrix. The resulting matrix is
ªa − b
« 0
«
« 0
«
« #
«¬ b
…
º
»
a−b
0
0
…
»
»
a−b …
0
0
»
#
#
%
#
»
2b
3b
… a + (n − 1)b »¼
c. Since the preceding matrix is a triangular matrix with the same determinant as A,
0
0
0
det A = (a − b) n −1 (a + (n − 1)b).
17. First consider the case n = 2. In this case
a−b b
b
det B =
= a ( a − b), det C =
0
a
b
b
a
= ab − b 2 ,
so det A = det B + det C = a (a − b) + ab − b 2 = a 2 − b 2 = (a − b)( a + b) = (a − b) 2−1 (a + (2 − 1)b) , and
the formula holds for n = 2.
Now assume that the formula holds for all (k – 1) × (k – 1) matrices, and let A, B, and C be k × k
matrices. By a cofactor expansion along the first column,
det B = (a − b)
a
b
…
b
b
a
…
b
#
b
#
b
%
…
#
a
= (a − b)(a − b) k − 2 (a + (k − 2)b) = (a − b) k −1 (a + (k − 2)b)
since the matrix in the above formula is a (k – 1) × (k – 1) matrix. We can perform a series of row
operations on C to “zero out” below the first pivot, and produce the following matrix whose
determinant is det C:
ªb
«0
«
«#
«
¬«0
b
…
a −b
…
#
0
%
…
b º
0 »»
.
# »
»
a − b ¼»
Since this is a triangular matrix, we have found that det C = b(a − b) k −1 . Thus
det A = det B + det C = (a − b) k −1 (a + (k − 2)b) + b(a − b)k −1 = (a − b) k −1 (a + (k − 1)b),
which is what was to be shown. Thus the formula has been proven by mathematical induction.
18. [M] Since the first matrix has a = 3, b = 8, and n = 4, its determinant is
(3 − 8) 4−1 (3 + (4 − 1)8) = ( −5)3 (3 + 24) = (−125)(27) = −3375. Since the second matrix has a = 8, b =
3, and n = 5, its determinant is (8 − 3)5−1 (8 + (5 − 1)3) = (5)4 (8 + 12) = (625)(20) = 12,500.
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 3
• Supplementary Exercises
19. [M] We find that
1
1
1
1
2
2 = 1,
1
2
3
1
1
1
1
1
2
2
2
1
2
3
3
1
2
3
4
1
1
1
1
1
1
2
2
2
2
= 1, 1
2
3
3
3 = 1.
1
2
3
4
4
1
2
3
4
5
Our conjecture then is that
1
1
1
…
1
1
2
2
…
2
1
2
3
…
3 = 1.
#
#
#
#
1
2
3
%
…
n
To show this, consider using row replacement operations to “zero out” below the first pivot. The
resulting matrix is
ª1
«0
«
«0
«
«#
«¬0
1
1
…
1
1
…
1
2
…
#
#
%
1
2
…
1 º
1 »»
2 ».
»
# »
n − 1»¼
Now use row replacement operations to “zero out” below the second pivot, and so on. The final
matrix which results from this process is
ª1
«0
«
«0
«
«#
«¬0
1
1
1
1
0
1
#
#
0
0
… 1º
… 1»»
… 1» ,
»
% #»
… 1»¼
which is an upper triangular matrix with determinant 1.
20. [M] We find that
1
1
1
1
3
3 = 6,
1
3
6
1
1
1
1
1
3
3
3
1
3
6
6
1
3
6
9
1
1
1
1
1
1
3
3
3
3
= 18, 1
3
6
6
6 = 54.
1
3
6
9
9
1
3
6
9
12
Our conjecture then is that
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
193
194
CHAPTER 3
• Determinants
1
1
1
…
1
1
3
3
…
3
1
3
6
…
6
#
#
#
%
#
1
3
6
…
3(n − 1)
= 2 ⋅ 3n − 2.
To show this, consider using row replacement operations to “zero out” below the first pivot. The
resulting matrix is
ª1
«0
«
«0
«
«#
«¬ 0
1
1
…
2
2
…
2
5
…
#
#
%
2
5
…
º
»
2
»
».
5
»
#
»
3(n − 1) − 1»¼
1
Now use row replacement operations to “zero out” below the second pivot. The matrix which results
from this process is
ª1
«0
«
«0
«
«0
«0
«
«#
«0
¬
1
1
1
1
1
…
2
2
2
2
2
…
0
3
3
3
3
…
0
3
6
6
6
…
0
3
6
9
9
…
#
#
#
#
#
%
0
3
6
9
12
…
º
2 »»
3 »
»
6 ».
9 »
»
# »
3(n − 2) »¼
1
This matrix has the same determinant as the original matrix, and is recognizable as a block matrix of
the form
ªA
«O
¬
where
ª1
A=«
¬0
Bº
,
D »¼
ª3
«3
«
1º
«3
=
and
D
2 »¼
«
«#
«¬3
3
3
3
…
6
6
6
…
6
9
9
…
#
#
#
%
6
9
12
…
º
ª1
»
«1
6 »
«
9 » = 3 «1
»
«
# »
«#
«¬1
3(n − 2) »¼
3
ªA
As in Exercise 14(c), the determinant of the matrix «
¬O
Since D is an (n – 2) × (n – 2) matrix,
1
1
1
…
2
2
2
…
2
3
3
…
#
#
#
%
2
3
4
!
1 º
2 »»
3 ».
»
# »
n − 2 »¼
Bº
is (det A)(det D) = 2 det D.
D »¼
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 3
det D = 3
n−2
1
1
1
1
…
1
1
2
2
2
…
2
1
2
3
3
…
3
#
#
#
#
%
#
1
2
3
4
…
n−2
• Supplementary Exercises
= 3n − 2 (1) = 3n − 2
ªA
by Exercise 19. Thus the determinant of the matrix «
¬O
Bº
is 2 det D = 2 ⋅ 3n − 2.
D »¼
Copyright ©!2012 Pearson Education, Inc. Publishing as Addison-Wesley.
195
4.1
SOLUTIONS
Notes: This section is designed to avoid the standard exercises in which a student is asked to check ten
howing that a
axioms on an array of sets. Theorem 1 pprovides the main homework tool in this section for sh
set is a subspace. Students should be tauught how to check the closure axioms. The exercises in
n this section
(and the next few sections) emphasize n, to give students time to absorb the abstract con
ncepts. Other
vectors do appear later in the chapter: thhe space of signals is used in Section 4.8, and the spaces n of
polynomials are used in many sections oof Chapters 4 and 6.
1. a. If u and v are in V, then their enttries are nonnegative. Since a sum of nonnegative num
mbers is
nonnegative, the vector u + v haas nonnegative entries. Thus u + v is in V.
ª2º
b. Example: If u = « » and c = –1, then u is in V but cu is not in V.
¬2¼
ªxº
ª x º ª cx º
2. a. If u = « » is in W, then the vecttor cu = c « » = « » is in W because (cx)(cy ) = c 2 ( xyy ) ≥ 0
¬ y¼
¬ y ¼ ¬cy ¼
since xy ≥ 0.
ª 2º
ª −1 º
b. Example: If u = « » and v = « » , then u and v are in W but u + v is not in W.
¬ 3¼
¬ −7 ¼
ª.5º
3. Example: If u = « » and c = 4, thenn u is in H but cu is not in H. Since H is not closed un
nder scalar
¬.5¼
multiplication, H is not a subspace oof 2.
4. Note that u and v are on the line L, but u + v is not.
u+v
u
v
L
Copyright © 2012 Pearrson Education, Inc. Publishing as Addison-Wesley.
197
198
CHAPTER 4
• Vector Spaces
5. Yes. Since the set is Span{t 2 } , the set is a subspace by Theorem 1.
6. No. The zero vector is not in the set.
7. No. The set is not closed under multiplication by scalars which are not integers.
8. Yes. The zero vector is in the set H. If p and q are in H, then (p + q)(0) = p(0) + q(0) = 0 + 0 = 0,
so p + q is in H. For any scalar c, (cp)(0) = c ⋅ p(0) = c ⋅ 0 = 0, so cp is in H. Thus H is a subspace by
Theorem 1.
ª −2 º
« »
9. The set H = Span {v}, where v = « 5» . Thus H is a subspace of
«¬ 3»¼
ª 3º
10. The set H = Span {v}, where v = «« 0»» . Thus H is a subspace of
«¬ −7 »¼
3
by Theorem 1.
3
by Theorem 1.
ª 2º
ª3º
« »
«
»
11. The set W = Span {u, v}, where u = « −1» and v = « 0 » . Thus W is a subspace of
«¬ 2 »¼
«¬ 0 »¼
ª 4º
ª2º
« 0»
«2»
12. The set W = Span {u, v}, where u = « » and v = « » . Thus W is a subspace of
«2»
« −3»
« »
« »
¬0¼
¬ 5¼
3
by Theorem 1.
4
by Theorem 1.
13. a. The vector w is not in the set {v1 , v 2 , v 3 } . There are 3 vectors in the set {v1 , v 2 , v 3 }.
b. The set Span{v1 , v 2 , v 3 } contains infinitely many vectors.
c. The vector w is in the subspace spanned by {v1 , v 2 , v 3 } if and only if the equation
x1 v1 + x2 v 2 + x3 v 3 = w has a solution. Row reducing the augmented matrix for this system of
linear equations gives
ª 1
« 0
«
«¬ −1
2
4
1
2
3
6
3º ª 1
1»» ∼ ««0
2»¼ «¬0
0
0
1
2
0
0
1º
1»» ,
0»¼
so the equation has a solution and w is in the subspace spanned by {v1 , v 2 , v 3 } .
14. The augmented matrix is found as in Exercise 13c. Since
ª1
«0
«
«¬ −1
2
4
1
2
3
6
1 º ª1
3 »» ∼ ««0
14 »¼ «¬0
0
0
1
2
0
0
−5º
3»» ,
0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.1
• Solutions
the equation x1 v1 + x2 v 2 + x3 v 3 = w has a solution, the vector w is in the subspace spanned by
{v1 , v 2 , v 3 }.
15. Since the zero vector is not in W, W is not a vector space.
16. Since the zero vector is not in W, W is not a vector space.
17. Since a vector w in W may be written as
ª 2º
ª −1º
ª 0º
« 0»
« 3»
« −1»
w = a« » + b« » + c« »
« −1»
« 0»
« 3»
« »
« »
« »
¬ 0¼
¬ 3¼
¬ 0¼
°
°
S =®
°
°
¯
ª 2 º ª −1º ª 0 º
« 0 » « 3» « −1»
« »,« »,« »
« −1» « 0 » « 3»
« » « » « »
¬ 0 ¼ ¬ 3¼ ¬ 0 ¼
½
°
°
¾
°
°
¿
is a set that spans W.
18. Since a vector w in W may be written as
ª 4º
ª 3º
ª 0º
« 0»
«0»
« 0»
«
»
«
»
w=a
+b
+c« »
« 1»
« 3»
« 1»
« »
« »
« »
¬ 0¼
¬ 3¼
¬ −2¼
°
°
S =®
°
°
¯
ª 4 º ª 3º ª 0 º
« 0 » «0 » « 0 »
« »,« »,« »
« 1» « 3» « 1»
« » « » « »
¬ 0 ¼ ¬ 3¼ ¬ −2¼
½
°
°
¾
°
°
¿
is a set that spans W.
19. Let H be the set of all functions described by y (t ) = c1cos ω t + c2 sin ω t . Then H is a subset of the
vector space V of all real-valued functions, and may be written as H = Span {cos ωt, sin ωt}. By
Theorem 1, H is a subspace of V and is hence a vector space.
20. a. The following facts about continuous functions must be shown.
1. The constant function f(t) = 0 is continuous.
2. The sum of two continuous functions is continuous.
3. A constant multiple of a continuous function is continuous.
b. Let H = {f in C[a, b]: f(a) = f(b)}.
1. Let g(t) = 0 for all t in [a, b]. Then g(a) = g(b) = 0, so g is in H.
2. Let g and h be in H. Then g(a) = g(b) and h(a) = h(b), and (g + h)(a) = g(a) + h(a) =
g(b) + h(b) = (g + h)(b), so g + h is in H.
3. Let g be in H. Then g(a) = g(b), and (cg)(a) = cg(a) = cg(b) = (cg)(b), so cg is in H.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
199
200
CHAPTER 4
• Vector Spaces
Thus H is a subspace of C[a, b].
21. The set H is a subspace of M 2×2 . The zero matrix is in H, the sum of two upper triangular matrices is
upper triangular, and a scalar multiple of an upper triangular matrix is upper triangular.
22. The set H is a subspace of M 2×4 . The 2 × 4 zero matrix 0 is in H because F0 = 0. If A and B are
matrices in H, then F(A + B) = FA + FB = 0 + 0 = 0, so A + B is in H. If A is in H and c is a scalar,
then F(cA) = c(FA) = c0 = 0, so cA is in H.
23. a. False. The zero vector in V is the function f whose values f(t) are zero for all t in .
b. False. An arrow in three-dimensional space is an example of a vector, but not every arrow is a
vector.
c. False. See Exercises 1, 2, and 3 for examples of subsets which contain the zero vector but are not
subspaces.
d. True. See the paragraph before Example 6.
e. False. Digital signals are used. See Example 3.
24. a.
b.
c.
d.
e.
True. See the definition of a vector space.
True. See statement (3) in the box before Example 1.
True. See the paragraph before Example 6.
False. See Example 8.
False. The second and third parts of the conditions are stated incorrectly. For example, part (ii)
does not state that u and v represent all possible elements of H.
25. 2, 4
26. a. 3
b. 5
c. 4
27. a.
b.
c.
d.
8
3
5
4
28. a.
b.
c.
d.
e.
4
7
3
5
4
29. Consider u + (–1)u. By Axiom 10, u + (–1)u = 1u + (–1)u. By Axiom 8, 1u + (–1)u = (1 + (–1))u =
0u. By Exercise 27, 0u = 0. Thus u + (–1)u = 0, and by Exercise 26 (–1)u = –u.
30. By Axiom 10 u = 1u. Since c is nonzero, c −1c = 1 , and u = (c −1c)u . By Axiom 9,
(c −1c )u = c −1 (cu) = c −1 0 since cu = 0. Thus u = c −1 0 = 0 by Property (2), proven in Exercise 28.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.1
• Solutions
201
31. Any subspace H that contains u and v must also contain all scalar multiples of u and v, and hence
must also contain all sums of scalar multiples of u and v. Thus H must contain all linear
combinations of u and v, or Span {u, v}.
Note: Exercises 32–34 provide good practice for mathematics majors because these arguments involve
simple symbol manipulation typical of mathematical proofs. Most students outside mathematics might
profit more from other types of exercises.
32. Both H and K contain the zero vector of V because they are subspaces of V. Thus the zero vector of V
is in H ∩ K. Let u and v be in H ∩ K. Then u and v are in H. Since H is a subspace u + v is in H.
Likewise u and v are in K. Since K is a subspace u + v is in K. Thus u + v is in H ∩ K. Let u be in H
∩ K. Then u is in H. Since H is a subspace cu is in H. Likewise u is in K. Since K is a subspace cu is
in K. Thus cu is in H ∩ K for any scalar c, and H ∩ K is a subspace of V.
The union of two subspaces is not in general a subspace. For an example in 2 let H be the x-axis and
let K be the y-axis. Then both H and K are subspaces of 2, but H ∪ K is not closed under vector
addition. The subset H ∪ K is thus not a subspace of 2.
33. a. Given subspaces H and K of a vector space V, the zero vector of V belongs to H + K, because 0 is
in both H and K (since they are subspaces) and 0 = 0 + 0. Next, take two vectors in H + K, say
w1 = u1 + v1 and w 2 = u 2 + v 2 where u1 and u 2 are in H, and v1 and v 2 are in K. Then
w1 + w 2 = u1 + v1 + u 2 + v 2 = (u1 + u 2 ) + ( v1 + v 2 )
because vector addition in V is commutative and associative. Now u1 + u 2 is in H and v1 + v 2 is
in K because H and K are subspaces. This shows that w1 + w 2 is in H + K. Thus H + K is closed
under addition of vectors. Finally, for any scalar c,
cw1 = c (u1 + v1 ) = cu1 + cv1
The vector cu1 belongs to H and cv1 belongs to K, because H and K are subspaces. Thus, cw1
belongs to H + K, so H + K is closed under multiplication by scalars. These arguments show that
H + K satisfies all three conditions necessary to be a subspace of V.
b. Certainly H is a subset of H + K because every vector u in H may be written as u + 0, where the
zero vector 0 is in K (and also in H, of course). Since H contains the zero vector of H + K, and H
is closed under vector addition and multiplication by scalars (because H is a subspace of V ), H is
a subspace of H + K. The same argument applies when H is replaced by K, so K is also a
subspace of H + K.
34. A proof that H + K = Span{u1 ,…, u p , v1 ,…, v q } has two parts. First, one must show that H + K is a
subset of Span{u1 ,…, u p , v1 ,…, v q }. Second, one must show that Span{u1 ,…, u p , v1 ,…, v q } is a
subset of H + K.
(1) A typical vector in H has the form c1u1 + … + c p u p and a typical vector in K has the form
d1v1 + …+ d q v q . The sum of these two vectors is a linear combination of u1 ,…, u p , v1 ,…, v q
and so belongs to Span{u1 ,…, u p , v1 ,…, v q }. Thus H + K is a subset of
Span{u1 ,…, u p , v1 ,…, v q }.
(2) Each of the vectors u1 ,…, u p , v1 ,…, v q belongs to H + K, by Exercise 33(b), and so any linear
combination of these vectors belongs to H + K, since H + K is a subspace, by Exercise 33(a).
Thus, Span{u1 ,…, u p , v1 ,…, v q } is a subset of H + K.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
202
CHAPTER 4
• Vector Spaces
35. [M] Since
ª 8
«
« −4
« −3
«
¬ 9
−4
3
−7
6
−2
−8
−5
−18
9º ª 1
−4 »» ««0
∼
−4 » «0
» «
7 ¼ ¬0
0
1
0
0
0
0
1
0
1º
−2»»
,
1»
»
0¼
w is in the subspace spanned by {v1 , v 2 , v 3 }.
36. [M] Since
[A
ª 3
« 8
y] = «
« −5
«
¬ 2
−5
7
−9
−6
−8
−2
3
−9
−4º ª 1
−8»» «« 0
∼
6» «0
» «
−5¼ ¬ 0
0
1
0
0
0
0
1
0
−1 / 5º
−2 / 5»»
,
3 / 5»
»
0¼
y is in the subspace spanned by the columns of A.
37. [M] The graph of f(t) is given below. A conjecture is that f(t) = cos 4t.
1
0.5
1
2
3
4
5
6
–0.5
–1
The graph of g(t) is given below. A conjecture is that g(t) = cos 6t.
1
0.5
1
2
3
4
5
6
–0.5
–1
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.2
• Solutions
203
38. [M] The graph of f(t) is given below. A conjecture is that f(t) = sin 3t.
1
0.5
1
2
3
4
5
6
–0.5
–1
The graph of g(t) is given below. A conjecture is that g(t) = cos 4t.
1
0.5
1
2
3
4
5
6
–0.5
–1
The graph of h(t) is given below. A conjecture is that h(t) = sin 5t.
1
0.5
1
2
3
4
5
6
–0.5
–1
4.2
SOLUTIONS
Notes: This section provides a review of Chapter 1 using the new terminology. Linear tranformations are
introduced quickly since students are already comfortable with the idea from n. The key exercises are
17–26, which are straightforward but help to solidify the notions of null spaces and column spaces.
Exercises 30–36 deal with the kernel and range of a linear transformation and are progressively more
advanced theoretically. The idea in Exercises 7–14 is for the student to use Theorems 1, 2, or 3 to
determine whether a given set is a subspace.
1. One calculates that
ª 3
Aw = «« 6
«¬ −8
−5
−2
4
−3º ª 1º ª 0º
0 »» «« 3»» = «« 0»» ,
1»¼ «¬ −4 »¼ «¬ 0»¼
so w is in Nul A.
2. One calculates that
ª 2
Aw = «« −3
«¬ −5
6
2
−4
4 º ª 1º ª 0º
5»» «« −1»» = ««0 »» ,
1»¼ «¬ 1»¼ «¬0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
204
CHAPTER 4
• Vector Spaces
so w is in Nul A.
3. First find the general solution of Ax = 0 in terms of the free variables. Since
[A
ª1
0] ∼ «
¬0
0
−2
4
1
3
−2
0º
,
0 »¼
the general solution is x1 = 2 x3 − 4 x4 , x2 = −3 x3 + 2 x4 , with x3 and x4 free. So
ª x1 º
ª 2º
ª −4º
«x »
« −3»
« 2»
x = « 2 » = x3 « » + x4 « » ,
« x3 »
« 1»
« 0»
« »
« »
« »
¬ 0¼
¬ 1¼
¬ x4 ¼
and a spanning set for Nul A is
°
°
®
°
°
¯
ª 2 º ª −4 º
« » « »
« −3» , « 2 »
« 1» « 0 »
« » « »
¬ 0 ¼ ¬ 1¼
½
°
°
¾.
°
°
¿
4. First find the general solution of Ax = 0 in terms of the free variables. Since
[A
ª1
0] ∼ «
¬0
−3
0
0
0
1
0
0º
,
0 »¼
the general solution is x1 = 3 x2 , x3 = 0 , with x2 and x4 free. So
ª x1 º
ª 3º
ª0º
«x »
«1 »
«0»
x = « 2 » = x2 « » + x4 « » ,
« x3 »
«0»
«0»
« »
« »
« »
¬0¼
¬1 ¼
¬ x4 ¼
and a spanning set for Nul A is
°
°
®
°
°
¯
ª 3º ª 0 º
«1 » « 0 »
« »,« »
«0» «0»
« » « »
¬ 0 ¼ ¬1 ¼
½
°
°
¾.
°
°
¿
5. First find the general solution of Ax = 0 in terms of the free variables. Since
[A
ª1
0] ∼ ««0
«¬0
−4
0
2
0
0
1
−5
0
0
0
0
1
0º
0 »» ,
0 »¼
the general solution is x1 = 4 x2 − 2 x4 , x3 = 5 x4 , x5 = 0 , with x2 and x4 free. So
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.2
• Solutions
205
ª x1 º
ª 4º
ª −2 º
«x »
«1 »
« 0»
« 2»
« »
« »
x = « x3 » = x2 « 0 » + x4 « 5» ,
« »
« »
« »
« x4 »
«0»
« 1»
«¬ x5 »¼
«¬ 0 »¼
«¬ 0 »¼
and a spanning set for Nul A is
°
°
°
®
°
°
°¯
ª 4 º ª −2 º
«1 » « 0 »
« » « »
« 0 » , « 5»
« » « »
« 0 » « 1»
«¬ 0 »¼ «¬ 0 »¼
½
°
°
°
¾.
°
°
°¿
6. First find the general solution of Ax = 0 in terms of the free variables. Since
[A
ª1
0] ∼ ««0
«¬0
0
5
−6
1
1
−3
1
0
0
0
0
0
0º
0 »» ,
0»¼
the general solution is x1 = −5 x3 + 6 x4 − x5 , x2 = 3 x3 − x4 , with x3 , x4 , and x5 free. So
ª x1 º
ª −5º
ª 6º
ª −1º
«x »
« 3»
« −1»
« 0»
« 2»
« »
« »
« »
x = « x3 » = x3 « 1» + x4 « 0 » + x5 « 0 » ,
« »
« »
« »
« »
« x4 »
« 0»
« 1»
« 0»
«¬ x5 »¼
«¬ 0 »¼
«¬ 0 »¼
«¬ 1»¼
and a spanning set for Nul A is
°
°
°
®
°
°
¯°
ª −5º ª 6 º ª −1º
« » « » « »
« 3» « −1» « 0 »
« 1» , « 0 » , « 0 »
« » « » « »
« 0 » « 1» « 0 »
«¬ 0 »¼ «¬ 0 »¼ «¬ 1»¼
7. The set W is a subset of
would be a subspace of
not a vector space.
8. The set W is a subset of
would be a subspace of
not a vector space.
½
°
°
°
¾.
°
°
¿°
3
. If W were a vector space (under the standard operations in 3), then it
. But W is not a subspace of 3 since the zero vector is not in W. Thus W is
3
3
. If W were a vector space (under the standard operations in 3), then it
3
. But W is not a subspace of 3 since the zero vector is not in W. Thus W is
9. The set W is the set of all solutions to the homogeneous system of equations p – 3q – 4s = 0,
0º
ª 1 − 3 −4
. Thus W is a subspace of 4 by
2p – s – 5r = 0. Thus W = Nul A, where A = «
»
0 −1 −5 ¼
¬2
Theorem 2, and is a vector space.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
206
CHAPTER 4
• Vector Spaces
10. The set W is the set of all solutions to the homogeneous system of equations 3a + b – c = 0,
0º
ª 3 1 −1
. Thus W is a subspace of
a + b + 2c – 2d = 0. Thus W = Nul A, where A = «
2 −2 »¼
¬1 1
Theorem 2, and is a vector space.
4
by
11. The set W is a subset of
would be a subspace of
not a vector space.
4
. If W were a vector space (under the standard operations in 4), then it
4
. But W is not a subspace of 4 since the zero vector is not in W. Thus W is
12. The set W is a subset of
would be a subspace of
not a vector space.
4
. If W were a vector space (under the standard operations in 4), then it
4
. But W is not a subspace of 4 since the zero vector is not in W. Thus W is
13. An element w on W may be written as
ª1 º
ª −6 º ª 1
«
»
w = c «0» + d «« 1»» = «« 0
¬«1 ¼»
¬« 0 ¼» ¬« 1
−6º
ªcº
1»» « »
d
0»¼ ¬ ¼
ª1
where c and d are any real numbers. So W = Col A where A = ««0
«¬ 1
by Theorem 3, and is a vector space.
−6 º
1»» . Thus W is a subspace of
0 »¼
14. An element w on W may be written as
ª −1º ª 3º ª −1
w = s «« 1»» + t «« −2»» = «« 1
«¬ 5»¼ «¬ −1»¼ «¬ 5
3º
ªsº
−2»» « »
t
−1»¼ ¬ ¼
ª −1
where a and b are any real numbers. So W = Col A where A = «« 1
«¬ 5
3
by Theorem 3, and is a vector space.
3º
−2 »» . Thus W is a subspace of
−1»¼
15. An element in this set may be written as
ª0º
ª 2º ª 1º ª 0
«1 »
« −1» « 2» « 1
r« » + s« » +t« » = «
« 3»
« 1» « 0» « 3
« »
« » « » «
¬ 2¼
¬ −1¼ ¬ −1¼ ¬ 2
2
−1
1
−1
1º
ªr º
2»» « »
s
0» « »
» «t »
−1¼ ¬ ¼
ª0
«1
where r, s and t are any real numbers. So the set is Col A where A = «
«3
«
¬2
2
−1
1
−1
1º
2 »»
.
0»
»
−1¼
16. An element in this set may be written as
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3
4.2
ª1 º
ª −1º
ª 0º ª 1
«2»
« 0»
« 3» « 2
b« » + c« » + d « » = «
«1 »
« 3»
« −3» « 1
« »
« »
« » «
¬0¼
¬ 1¼
¬ 1¼ ¬ 0
−1
0
3
1
• Solutions
207
0º
ªb º
3»» « »
c
−3» « »
» «d »
1¼ ¬ ¼
ª1
«2
where b, c and d are any real numbers. So the set is Col A where A = «
«1
«
¬0
−1
0
3
1
0º
3»»
.
−3»
»
1¼
17. The matrix A is a 4 × 2 matrix. Thus
(a) Nul A is a subspace of 2, and
(b) Col A is a subspace of 4.
18. The matrix A is a 4 × 3 matrix. Thus
(a) Nul A is a subspace of 3, and
(b) Col A is a subspace of 4.
19. The matrix A is a 2 × 5 matrix. Thus
(a) Nul A is a subspace of 5, and
(b) Col A is a subspace of 2.
20. The matrix A is a 1 × 5 matrix. Thus
(a) Nul A is a subspace of 5, and
(b) Col A is a subspace of 1 = .
21. Either column of A is a nonzero vector in Col A. To find a nonzero vector in Nul A, find the general
solution of Ax = 0 in terms of the free variables. Since
ª1
«0
0] ∼ «
«0
«
¬0
−2 / 3
0
0º
0 »»
A
, the general solution is x1 = (2 / 3) x2 , with x2 free. Letting x2 be a
[
0 0»
»
0 0¼
nonzero value (say x2 = 3 ) gives the nonzero vector
ª x º ª2º
x = « 1» = « »
¬ x2 ¼ ¬ 3 ¼
which is in Nul A.
22. Any column of A is a nonzero vector in Col A. To find a nonzero vector in Nul A, find the general
solution of Ax = 0 in terms of the free variables. Since
[A
ª1
«0
0] ∼ «
«0
«
¬0
0
1
1
1
0
0
0
0
0º
0 »»
,
0»
»
0¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
208
CHAPTER 4
• Vector Spaces
the general solution is x1 = − x3 , x2 = − x3 , with x3 free. Letting x3 be a nonzero value (say x3 = −1 )
gives the nonzero vector
ª x1 º ª 1º
x = «« x2 »» = «« 1»»
¬« x3 ¼» «¬ −1»¼
which is in Nul A.
23. Consider the system with augmented matrix [ A
[A
ª1
w] ∼ «
¬0
−2
w ] . Since
−1º
,
0 »¼
0
the system is consistent and w is in Col A. Also, since
ª −2
Aw = «
¬ −1
4 º ª2 º ª0º
=
2 »¼ «¬ 1 »¼ «¬ 0 »¼
w is in Nul A.
24. Consider the system with augmented matrix [ A
[A
ª1
«0
w] ∼ «
«0
«
¬0
0
1
0
0
−1
−1
0
0
1
0
0
0
w ] . Since
1º
1»»
,
0»
»
0¼
the system is consistent and w is in Col A. Also, since
ª10
« 0
Aw = «
« 1
«
¬ 1
−8
2
−2
2
−1
1
6
0
−2º ª 2º ª 0º
−2»» «« 2»» «« 0»»
=
0» « 0» «0»
»« » « »
−2¼ ¬ 2¼ ¬ 0¼
w is in Nul A.
25. a.
b.
c.
d.
e.
f.
True. See the definition before Example 1.
False. See Theorem 2.
True. See the remark just before Example 4.
False. The equation Ax = b must be consistent for every b. See #7 in the table on page 204.
True. See Figure 2.
True. See the remark after Theorem 3.
26. a.
b.
c.
d.
e.
f.
True. See Theorem 2.
True. See Theorem 3.
False. See the box after Theorem 3.
True. See the paragraph after the definition of a linear transformation.
True. See Figure 2.
True. See the paragraph before Example 8.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.2
• Solutions
209
27. Let A be the coefficient matrix of the given homogeneous system of equations. Since Ax = 0 for
ª 3º
x = «« 2 »» , x is in Nul A. Since Nul A is a subspace of 3, it is closed under scalar multiplication. Thus
«¬ −1»¼
ª 30 º
10x = «« 20 »» is also in Nul A, and x1 = 30 , x2 = 20 , x3 = −10 is also a solution to the system of
«¬ −10 »¼
equations.
28. Let A be the coefficient matrix of the given systems of equations. Since the first system has a
ª0º
« »
solution, the constant vector b = «1 » is in Col A. Since Col A is a subspace of 3, it is closed under
«¬9 »¼
ª 0º
« »
scalar multiplication. Thus 5b = « 5» is also in Col A, and the second system of equations must thus
«¬ 45»¼
have a solution.
29. a. Since A0 = 0, the zero vector is in Col A.
b. Since Ax + Aw = A(x + w ), A x + Aw is in Col A.
c. Since c( Ax) = A(cx), cAx is in Col A.
30. Since T (0V ) = 0W , the zero vector 0W of W is in the range of T. Let T(x) and T(w) be typical
elements in the range of T. Then since T (x) + T (w ) = T ( x + w ), T (x) + T (w ) is in the range of T and
the range of T is closed under vector addition. Let c be any scalar. Then since cT (x) = T (cx), cT (x)
is in the range of T and the range of T is closed under scalar multiplication. Hence the range of T is a
subspace of W.
31. a. Let p and q be arbitary polynomials in
2,
and let c be any scalar. Then
ª(p + q )(0) º ªp (0) + q (0) º ªp (0) º ªq (0) º
T (p + q ) = «
»=«
»=«
»+«
» = T (p ) + T ( q )
¬ (p + q )(1) ¼ ¬ p (1) + q (1) ¼ ¬ p (1) ¼ ¬ q (1) ¼
and
ª (cp )(0) º
ªp (0) º
= c«
T ( cp ) = «
»
» = cT (p )
¬ (cp )(1) ¼
¬ p(1) ¼
so T is a linear transformation.
b. Any quadratic polynomial q for which q(0) = 0 and q (1) = 0 will be in the kernel of T. The
ªx º
polynomial q must then be a multiple of p(t ) = t (t − 1). Given any vector « 1 » in 2, the
¬ x2 ¼
polynomial p = x1 + ( x2 − x1 )t has p (0) = x1 and p (1) = x2 . Thus the range of T is all of 2.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
210
CHAPTER 4
• Vector Spaces
32. Any quadratic polynomial q for which q(0) = 0 will be in the kernel of T. The polynomial q must
then be q = at + bt 2 . Thus the polynomials p1 (t ) = t and p 2 (t ) = t 2 span the kernel of T. If a vector
ªa º
is in the range of T, it must be of the form « » . If a vector is of this form, it is the image of the
¬a ¼
° ª a º
°½
polynomial p(t ) = a in 2. Thus the range of T is ® « » : a real ¾ .
°¯ ¬ a ¼
°¿
33. a. For any A and B in M 2×2 and for any scalar c,
T ( A + B) = ( A + B ) + ( A + B )T = A + B + AT + BT = ( A + AT ) + ( B + BT ) = T ( A) + T ( B )
and
T (cA) = (cA)T = c( AT ) = cT ( A)
so T is a linear transformation.
b. Let B be an element of M 2×2 with B T = B , and let A = 12 B. Then
1
1
1
1
1
1
T ( A) = A + AT = B + ( B)T = B + BT = B + B = B
2
2
2
2
2
2
c. Part b. showed that the range of T contains the set of all B in M 2×2 with B T = B. It must also be
shown that any B in the range of T has this property. Let B be in the range of T. Then B = T(A) for
some A in M 2×2 . Then B = A + AT , and
BT = ( A + AT )T = AT + ( AT )T = AT + A = A + AT = B
so B has the property that B T = B.
ªa
d. Let A = «
¬c
bº
be in the kernel of T. Then T ( A) = A + AT = 0 , so
d »¼
ªa
A + AT = «
¬c
b º ªa
+
d »¼ «¬ b
c º ª 2a
=
d »¼ «¬b + c
c + b º ª0
=
2d »¼ «¬ 0
0º
0 »¼
°
Solving it is found that a = d = 0 and c = −b . Thus the kernel of T is ®
°¯
ª0
« −b
¬
bº
°½
: b real ¾.
»
0¼
°¿
34. Let f and g be any elements in C[0, 1] and let c be any scalar. Then T(f) is the antiderivative F of f
with F(0) = 0 and T(g) is the antiderivative G of g with G(0) = 0. By the rules for antidifferentiation
F + G will be an antiderivative of f + g, and ( F + G )(0) = F (0) + G (0) = 0 + 0 = 0. Thus
T (f + g) = T ( f ) + T (g ). Likewise cF will be an antiderivative of cf, and (cF )(0) = cF (0) = c0 = 0.
Thus T (cf ) = cT (f ), and T is a linear transformation. To find the kernel of T, we must find all
functions f in C[0,1] with antiderivative equal to the zero function. The only function with this
property is the zero function 0, so the kernel of T is {0}.
35. Since U is a subspace of V, 0V is in U. Since T is linear, T (0V ) = 0W . So 0W is in T(U). Let T(x) and
T(y) be typical elements in T(U). Then x and y are in U, and since U is a subspace of V, x + y is also
in U. Since T is linear, T (x) + T (y ) = T (x + y ). So T ( x) + T (y ) is in T(U), and T(U) is closed under
vector addition. Let c be any scalar. Then since x is in U and U is a subspace of V, cx is in U. Since T
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.2
• Solutions
211
is linear, T (cx) = cT (x) and cT(x) is in T(U ). Thus T(U) is closed under scalar multiplication, and
T(U) is a subspace of W.
36. Since Z is a subspace of W, 0W is in Z. Since T is linear, T (0V ) = 0W . So 0V is in U. Let x and y be
typical elements in U. Then T(x) and T(y) are in Z, and since Z is a subspace of W, T ( x) + T (y ) is
also in Z. Since T is linear, T (x) + T (y ) = T (x + y ). So T (x + y ) is in Z, and x + y is in U. Thus U is
closed under vector addition. Let c be any scalar. Then since x is in U, T(x) is in Z. Since Z is a
subspace of W, cT(x) is also in Z. Since T is linear, cT ( x) = T (cx) and T(cx) is in T(U). Thus cx is in
U and U is closed under scalar multiplication. Hence U is a subspace of V.
37. [M] Consider the system with augmented matrix [ A
[A
ª1
«0
w] ∼ «
«0
«
«¬0
0
1
0
0
−1/ 95
39 /19
0
0
1
0
267 / 95
0
w ]. Since
1/ 95º
−20 /19»»
,
−172 / 95»
»
0»¼
the system is consistent and w is in Col A. Also, since
ª 7
« −5
Aw = «
« 9
«
«¬ 19
6
−1
−4
0
−11
−9
7
7
1º ª 1º ª14 º
−2»» «« 1»» «« 0 »»
=
−3» « −1» « 0 »
»« » « »
1»¼ «¬ −3»¼ «¬ 0 ¼»
w is not in Nul A.
38. [M] Consider the system with augmented matrix [ A
[A
ª1
«0
w] ∼ «
«0
«
«¬0
0
1
−1
−2
0
0
0
0
0
0
1
0
w ] . Since
−2 º
−3»»
,
1»
»
0»¼
the system is consistent and w is in Col A. Also, since
ª −8
« −5
Aw = «
« 10
«
¬« 3
5
2
−2
1
−8
−2
6
1
0º ª1 º ª0º
−2»» «« 2 »» «« 0»»
=
−3» «1 » «0»
»« » « »
0¼» ¬« 0 ¼» «¬0»¼
w is in Nul A.
39. [M]
a. To show that a3 and a5 are in the column space of B, we can row reduce the matrices [ B
and [ B
a3 ] :
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
a3 ]
212
CHAPTER 4
• Vector Spaces
ª1
«0
[ B a3 ] ∼ ««
0
«
¬«0
[B
ª1
«0
a5 ] ∼ «
«0
«
¬«0
0 0 1/ 3º
1 0 1/ 3»»
0 1 0 »
»
0 0 0 ¼»
0
1
0
0
0
0
1
0
10 / 3º
−26 / 3»»
−4»
»
0¼»
Since both these systems are consistent, a3 and a5 are in the column space of B. Notice that the
same conclusions can be drawn by observing the reduced row echelon form for A:
ª1
«0
A∼«
«0
«
«¬0
0 1/ 3
1 1/ 3
0
0
0
0
1
0
0
0
10 / 3º
−26 / 3»»
−4»
»
0»¼
b. We find the general solution of Ax = 0 in terms of the free variables by using the reduced row
echelon form of A given above: x1 = ( −1/ 3) x3 − (10 / 3) x5 , x2 = ( −1/ 3) x3 + (26 / 3) x5 , x4 = 4 x5
with x3 and x5 free. So
ª x1 º
ª −1/ 3º
ª −10 / 3º
«x »
« −1/ 3»
« 26 / 3»
« 2»
«
»
«
»
x = « x3 » = x3 «
1» + x5 «
0» ,
« »
«
»
«
»
0»
4»
« x4 »
«
«
«¬ x5 »¼
«¬
«¬
0 »¼
1»¼
and a spanning set for Nul A is
°
°
°
®
°
°
°¯
ª −1/ 3º ª −10 / 3º
« −1/ 3» « 26 / 3»
«
» «
»
«
1» , «
0»
«
» «
»
0» «
4»
«
«¬
0 »¼ «¬
1»¼
½
°
°
°
¾.
°
°
°¿
c. The reduced row echelon form of A shows that the columns of A are linearly dependent and do
not span 4. Thus by Theorem 12 in Section 1.9, T is neither one-to-one nor onto.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.3
• Solutions
213
40. [M] Since the line lies both in H = Span{v1 , v 2 } and in K = Span{v 3 , v 4 } , w can be written both as
c1v1 + c2 v 2 and c3 v 3 + c4 v 4 . To find w we must find the cj’s which solve
c1 v1 + c2 v 2 − c3 v 3 − c4 v 4 = 0 . Row reduction of [ v1
ª5
«3
«
«¬8
1
−2
0
3
1
12
4
−5
28
0º ª 1
0 »» ∼ ««0
0 »¼ «¬0
0
0
−10 / 3
1
0
26 / 3
0
1
−4
v2
− v3
− v4
0] yields
0º
0»» ,
0»¼
so the vector of cj’s must be a multiple of (10/3, –26/3, 4, 1). One simple choice is (10, –26, 12, 3),
which gives w = 10 v1 − 26 v 2 = 12 v 3 + 3v 4 = (24, −48, −24) . Another choice for w is (1, –2, –1).
4.3
SOLUTIONS
Notes: The definition for basis is given initially for subspaces because this emphasizes that the basis
elements must be in the subspace. Students often overlook this point when the definition is given for a
vector space (see Exercise 25). The subsection on bases for Nul A and Col A is essential for Sections 4.5
and 4.6. The subsection on “Two Views of a Basis” is also fundamental to understanding the interplay
between linearly independent sets, spanning sets, and bases. Key exercises in this section are Exercises
21–25, which help to deepen students’ understanding of these different subsets of a vector space.
1. Consider the matrix whose columns are the given set of vectors. This 3 × 3 matrix is in echelon form,
and has 3 pivot positions. Thus by the Invertible Matrix Theorem, its columns are linearly
independent and span 3. So the given set of vectors is a basis for 3.
2. Since the zero vector is a member of the given set of vectors, the set cannot be linearly independent
and thus cannot be a basis for 3. Now consider the matrix whose columns are the given set of
vectors. This 3 × 3 matrix has only 2 pivot positions. Thus by the Invertible Matrix Theorem, its
columns do not span 3.
3. Consider the matrix whose columns are the given set of vectors. The reduced echelon form of this
matrix is
ª 1 3 −2º ª 1 0 1º
« 0
1 −1»» ∼ ««0 1 −1»»
«
«¬ −3 −4
1»¼ «¬0 0 0»¼
so the matrix has only two pivot positions. Thus its columns do not form a basis for
vectors is linearly independent and does not span 3.
3
; the set of
4. Consider the matrix whose columns are the given set of vectors. The reduced echelon form of this
matrix is
ª 2
« −1
«
«¬ 1
2
−3
2
−8º ª 1
5»» ∼ ««0
4»¼ «¬0
0
1
0
0º
0 »»
1»¼
so the matrix has three pivot positions. Thus its columns form a basis for
3
.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
214
CHAPTER 4
• Vector Spaces
5. Since the zero vector is a member of the given set of vectors, the set cannot be linearly independent
and thus cannot be a basis for 3. Now consider the matrix whose columns are the given set of
vectors. The reduced echelon form of this matrix is
ª 3
« −3
«
«¬ 0
−3
0
7
0
0
0
0º ª 1
−3»» ∼ ««0
5»¼ «¬ 0
0
0
1
0
0
0
0º
0 »»
1»¼
so the matrix has a pivot in each row. Thus the given set of vectors spans
3
.
6. Consider the matrix whose columns are the given set of vectors. Since the matrix cannot have a pivot
in each row, its columns cannot span 3; thus the given set of vectors is not a basis for 3. The
reduced echelon form of the matrix is
ª 1
« 2
«
¬« −4
−4º ª 1
3»» ∼ ««0
6¼» ¬«0
0º
1»»
0¼»
so the matrix has a pivot in each column. Thus the given set of vectors is linearly independent.
7. Consider the matrix whose columns are the given set of vectors. Since the matrix cannot have a pivot
in each row, its columns cannot span 3; thus the given set of vectors is not a basis for 3. The
reduced echelon form of the matrix is
ª −2
« 3
«
«¬ 0
6º ª 1
−1»» ∼ ««0
5»¼ «¬0
0º
1»»
0»¼
so the matrix has a pivot in each column. Thus the given set of vectors is linearly independent.
8. Consider the matrix whose columns are the given set of vectors. Since the matrix cannot have a pivot
in each column, the set cannot be linearly independent and thus cannot be a basis for 3. The reduced
echelon form of this matrix is
ª 1
« −2
«
«¬ 3
0
2
3
−1
−1
5
0º ª 1
0»» ∼ ««0
−1»¼ «¬0
0
2
1
1
0
0
0º
0»»
1»¼
so the matrix has a pivot in each row. Thus the given set of vectors spans
3
.
9. We find the general solution of Ax = 0 in terms of the free variables by using the reduced echelon
form of A:
ª1
«0
«
«¬ 3
0
−2
1
1
−1
−7
−2º ª 1
4»» ∼ ««0
3»¼ «¬0
0
−2
1
1
0
0
0º
0 »» .
1»¼
So x1 = 2 x3 , x2 = − x3 , x4 = 0 , with x3 free. So
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.3
• Solutions
215
ª x1 º
ª 2º
«x »
« −1»
2»
«
x=
= x « »,
« x3 » 3 « 1»
« »
« »
¬ 0¼
¬ x4 ¼
and a basis for Nul A is
°
°
®
°
°
¯
ª 2º
« −1»
« »
« 1»
« »
¬ 0¼
½
°
°
¾.
°
°
¿
10. We find the general solution of Ax = 0 in terms of the free variables by using the reduced echelon
form of A:
ª1
«0
«
«¬0
1
−2
1
1
6
−1
0
−8
0
5º ª 1
−2»» ∼ ««0
16»¼ «¬0
0
0
2
1
0
−1
0
1
0
−9º
10»» .
−2»¼
So x1 = −2 x4 + 9 x5 , x2 = x4 − 10 x5 , x3 = 2 x5 , with x4 and x5 free. So
ª x1 º
ª −2 º
ª 9º
«x »
« 1»
«
»
« 2»
« »
« −10 »
x = « x3 » = x4 « 0 » + x5 « 2 » ,
« »
« »
«
»
« x4 »
« 1»
« 0»
«¬ x5 »¼
«¬ 0 »¼
«¬ 1»¼
and a basis for Nul A is
°
°
°
®
°
°
°¯
ª −2 º ª 9 º
« 1» « −10 »
« » «
»
« 0» , « 2»
« » «
»
« 1» « 0 »
«¬ 0 »¼ «¬ 1»¼
½
°
°
°
¾.
°
°
°¿
11. Let A = [1 −3 2] . Then we wish to find a basis for Nul A. We find the general solution of Ax = 0
in terms of the free variables: x = 3y – 2z with y and z free. So
ªxº
x = «« y »» =
«¬ z »¼
ª 3º
ª −2º
«
»
y « 1» + z «« 0»» ,
«¬0 »¼
«¬ 1»¼
and a basis for Nul A is
°
®
°
¯
ª 3 º ª −2 º
« 1» , « 0»
« » « »
«¬0 »¼ «¬ 1»¼
½
°
¾.
°
¿
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
216
CHAPTER 4
• Vector Spaces
12. We want to find a basis for the set of vectors in 2 in the line 3x + y = 0. Let A = [3 1] . Then we
wish to find a basis for Nul A. We find the general solution of Ax = 0 in terms of the free variables: y
= – 3x with x free. So
ªxº
ª 1º
x = « » = x « »,
¬ y¼
¬ −3¼
and a basis for Nul A is
°
®
°¯
ª 1º °½
« −3» ¾ .
¬ ¼ °¿
13. Since B is a row echelon form of A, we see that the first and second columns of A are its pivot
columns. Thus a basis for Col A is
°
®
°
¯
ª −2º ª 4 º
« 2» , « −6 »
« » « »
«¬ −3»¼ «¬ 8»¼
½
°
¾.
°
¿
To find a basis for Nul A, we find the general solution of Ax = 0 in terms of the free variables:
x1 = −6 x3 − 5 x4 , x2 = ( −5 / 2) x3 − (3 / 2) x4 , with x3 and x4 free. So
ª x1 º
ª −6º
ª −5º
«x »
« −5 / 2 »
«
»
» + x4 « −3/ 2 » ,
x = « 2 » = x3 «
« x3 »
«
«
1»
0»
« »
«
»
«
»
0 »¼
1»¼
«¬
«¬
«¬ x4 »¼
and a basis for Nul A is
°
°
®
°
°
¯
ª −6 º ª −5º
« −5 / 2 » « −3/ 2 »
«
»,«
»
«
1» «
0»
«
» «
»
0 »¼ «¬
1»¼
«¬
½
°
°
¾.
°
°
¿
14. Since B is a row echelon form of A, we see that the first, third, and fifth columns of A are its pivot
columns. Thus a basis for Col A is
°
°
®
°
°
¯
ª 1º ª 3º ª 8º
« 1» « 0 » « 8»
« »,« »,« »
« 2 » « −3» «9 »
« » « » « »
¬ 3¼ ¬ 0 ¼ ¬ 9 ¼
½
°
°
¾.
°
°
¿
To find a basis for Nul A, we find the general solution of Ax = 0 in terms of the free variables,
mentally completing the row reduction of B to get: x1 = −2 x2 − 2 x4 , x3 = 2 x4 , x5 = 0, with x2 and
x4 free. So
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.3
• Solutions
217
ª x1 º
ª −2 º
ª −2 º
«x »
« 1»
« 0»
« 2»
« »
« »
x = « x3 » = x2 « 0 » + x4 « 2 » ,
« »
« »
« »
« x4 »
« 0»
« 1»
«¬ x5 »¼
«¬ 0 »¼
«¬ 0 »¼
and a basis for Nul A is
ª −2 º ª −2 º
°« » « »
° « 1» « 0 »
°
® « 0» , « 2»
° « 0 » « 1»
°« » « »
°¯ «¬ 0 »¼ «¬ 0 »¼
½
°
°
°
¾.
°
°
°¿
15. This problem is equivalent to finding a basis for Col A, where A = [ v1
the reduced echelon form of A is
ª 1
« 0
«
« −2
«
¬ 3
0
1
2
−2
2
−1
2
3
−8
0
10
3
3º ª 1
−1»» ««0
∼
−6 » «0
» «
9¼ ¬0
0
1
2
−2
0
0
0
0
0
0
1
0
v2
v3
v4
v5 ] . Since
0º
0»»
,
0»
»
1¼
we see that the first, second, fourth and fifth columns of A are its pivot columns. Thus a basis for the
space spanned by the given vectors is
°
°
®
°
°
¯
ª 1º ª 0 º ª 2 º ª 3º
« 0 » « 1» « −1» « −1»
« »,« »,« »,« »
« −2 » « 2 » «10 » « −6 »
« » « » « » « »
¬ 3 ¼ ¬ 3¼ ¬ 3¼ ¬ 9 ¼
½
°
°
¾.
°
°
¿
16. This problem is equivalent to finding a basis for Col A, where A = [ v1
the reduced echelon form of A is
ª1
«0
«
«0
«
¬1
−2
0
3
−1
5
−3
0
2
1
−1
3
−4
2º ª 1
−1»» ««0
∼
1» «0
» «
0 ¼ ¬0
0
1
0
0
0
0
1
0
v2
v3
v4
v5 ] . Since
−5 / 2
3/ 4
0º
1 / 2»»
,
3
1»
»
0
0¼
we see that the first, second, and third columns of A are its pivot columns. Thus a basis for the space
spanned by the given vectors is
°
°
®
°
°
¯
ª1 º ª −2 º ª 3º
« 0 » « 0 » « −1»
« »,« »,« »
« 0 » « 0 » « 1»
« » « » « »
¬1 ¼ ¬ 2 ¼ ¬ −1¼
½
°
°
¾.
°
°
¿
17. [M] This problem is equivalent to finding a basis for Col A, where A = [ v1
Since the reduced echelon form of A is
v2
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
v3
v4
v5 ] .
218
CHAPTER 4
ª 2
« 0
«
« −4
«
« −6
«¬ 0
• Vector Spaces
4
0
−2
−4
8
4
2
−4
0
1
8
−3
4
−7
15
−8º ª 1
4 »» «« 0
0» ∼ «0
» «
0» «0
1»¼ «¬ 0
0
1
0
0
−1
2
0
0
1
0
−1
0
0
0
0
0º
0 »»
0» ,
»
1»
0 »¼
we see that the first, second, third, and fifth columns of A are its pivot columns. Thus a basis for the
space spanned by the given vectors is
°
°
°
®
°
°
°¯
ª 2 º ª 4 º ª −2 º ª −8º ½
« 0 » « 0 » « −4 » « 4 » °
« » « » « » « » °°
« −4 » , « 2 » , « 0 » , « 0 » ¾ .
« » « » « » « »°
« −6 » « −4 » « 1» « 0 » °
«¬ 0 »¼ «¬ 4 »¼ «¬ −7 »¼ «¬ 1»¼ °
¿
18. [M] This problem is equivalent to finding a basis for Col A, where A = [ v1
Since the reduced echelon form of A is
ª −3
« 2
«
« 6
«
« 0
«¬ −7
3
0
0
2
6
−2
−9
0
6
−4
0
−1
−14
0
13
−6 º ª 1
3»» «« 0
0» ∼ «0
» «
−1» « 0
0 »¼ «¬ 0
0
1
0
0
0
2
0
0
0
1
0
0
−1
0
0
v2
v3
v4
v 5 ].
0º
0 »»
0» ,
»
1»
0 »¼
we see that the first, second, third, and fifth columns of A are its pivot columns. Thus a basis for the
space spanned by the given vectors is
°
°
°
®
°
°
°¯
ª −3º ª 3º ª 0 º ª −6º
« 2 » « 0 » « 2 » « 3»
« » « » « » « »
« 6 » , « −9 » , « −4 » , « 0»
« » « » « » « »
« 0 » « 0 » « 0 » « −1»
«¬ −7 »¼ «¬ 6 »¼ «¬ −1»¼ «¬ 0»¼
½
°
°
°
¾.
°
°
°¿
19. Since 4 v1 + 5 v 2 − 3v 3 = 0, we see that each of the vectors is a linear combination of the others. Thus
the sets {v1 , v 2 }, {v1 , v 3 }, and {v 2 , v 3 } all span H. Since we may confirm that none of the three
vectors is a multiple of any of the others, the sets {v1 , v 2 }, {v1 , v 3 }, and {v 2 , v 3 } are linearly
independent and thus each forms a basis for H.
20. Since 2 v1 − v 2 − v 3 = 0, we see that each of the vectors is a linear combination of the others. Thus
the sets {v1 , v 2 }, {v1 , v 3 }, and {v 2 , v 3 } all span H. Since we may confirm that none of the three
vectors is a multiple of any of the others, the sets {v1 , v 2 }, {v1 , v 3 }, and {v 2 , v 3 } are linearly
independent and thus each forms a basis for H.
21. a. False. The zero vector by itself is linearly dependent. See the paragraph preceding Theorem 4.
b. False. The set {b1 ,…, b p } must also be linearly independent. See the definition of a basis.
c. True. See Example 3.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.3
• Solutions
219
d. False. See the subsection “Two Views of a Basis.”
e. False. See the box before Example 9.
22. a. False. The subspace spanned by the set must also coincide with H. See the definition of a basis.
b. True. Apply the Spanning Set Theorem to V instead of H. The space V is nonzero because the
spanning set uses nonzero vectors.
c. True. See the subsection “Two Views of a Basis.”
d. False. See the two paragraphs before Example 8.
e. False. See the warning after Theorem 6.
23. Let A = [ v1
v2
v3
v 4 ]. Then A is square and its columns span
4
since
4
= Span{v1 , v 2 , v 3 , v 4 }. So its columns are linearly independent by the Invertible Matrix Theorem,
and {v1 , v 2 , v 3 , v 4 } is a basis for 4.
24. Let A = [ v1
span
n
…
v n ]. Then A is square and its columns are linearly independent, so its columns
by the Invertible Matrix Theorem. Thus {v1 ,… , v n } is a basis for
n
.
25. In order for the set to be a basis for H, {v1 , v 2 , v 3 } must be a spanning set for H; that is,
H = Span{v1 , v 2 , v 3 }. The exercise shows that H is a subset of Span{v1 , v 2 , v 3 }. but there are vectors
in Span{v1 , v 2 , v 3 } which are not in H ( v1 and v3 , for example). So H ≠ Span{v1 , v 2 , v 3 }, and
{v1 , v 2 , v 3 } is not a basis for H.
26. Since sin t cos t = (1/2) sin 2t, the set {sin t, sin 2t} spans the subspace. By inspection we note that
this set is linearly independent, so {sin t, sin 2t} is a basis for the subspace.
27. The set {cos ωt, sin ωt} spans the subspace. By inspection we note that this set is linearly
independent, so {cos ωt, sin ωt} is a basis for the subspace.
28. The set {e −bt , te −bt } spans the subspace. By inspection we note that this set is linearly independent,
so {e −bt , te −bt } is a basis for the subspace.
29. Let A be the n × k matrix [ v1 … v k ] . Since A has fewer columns than rows, there cannot be a
pivot position in each row of A. By Theorem 4 in Section 1.4, the columns of A do not span n and
thus are not a basis for n.
30. Let A be the n × k matrix [ v1 … v k ] . Since A has fewer rows than columns, there cannot be a
pivot position in each column of A. By Theorem 8 in Section 1.7, the columns of A are not linearly
independent and thus are not a basis for n.
31. Suppose that {v1 ,…, v p } is linearly dependent. Then there exist scalars c1 ,…, c p not all zero with
c1 v1 + … + c p v p = 0.
Since T is linear,
T (c1v1 + …+ c p v p ) = c1T ( v1 ) + … + c pT ( v p )
and
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
220
CHAPTER 4
• Vector Spaces
T (c1 v1 + … + c p v p ) = T (0) = 0.
Thus
c1T ( v1 ) + … + c pT ( v p ) = 0
and since not all of the ci are zero, {T ( v1 ),…, T ( v p )} is linearly dependent.
32. Suppose that {T ( v1 ),…, T ( v p )} is linearly dependent. Then there exist scalars c1 ,…, c p not all zero
with
c1T ( v1 ) + … + c pT ( v p ) = 0.
Since T is linear,
T (c1 v1 + … + c p v p ) = c1T ( v1 ) + … + c pT ( v p ) = 0 = T (0)
Since T is one-to-one
T (c1 v1 + …+ c p v p ) = T (0)
implies that
c1 v1 + … + c p v p = 0.
Since not all of the ci are zero, {v1 ,…, v p } is linearly dependent.
33. Neither polynomial is a multiple of the other polynomial. So {p1 , p 2 } is a linearly independent set in
3. Note: {p1 , p 2 } is also a linearly independent set in 2 since p1 and p 2 both happen to be in 2.
34. By inspection, p 3 = p1 + p 2 , or p1 + p 2 − p 3 = 0 . By the Spanning Set Theorem,
Span{p1 , p 2 , p 3 } = Span{p1 , p 2 } . Since neither p1 nor p 2 is a multiple of the other, they are linearly
independent and hence {p1 , p 2 } is a basis for Span{p1 , p 2 , p 3 }.
35. Let {v1 , v 3 } be any linearly independent set in a vector space V, and let v 2 and v 4 each be linear
combinations of v1 and v 3 . For instance, let v 2 = 5 v1 and v 4 = v1 + v 3 . Then {v1 , v 3 } is a basis for
Span{v1 , v 2 , v 3 , v 4 }.
36. [M] Row reduce the following matrices to identify their pivot columns:
[u1
[ v1
u2
ª 1
« 2
u3 ] = «
« 0
«
¬ −1
v2
ª −2
« −2
v3 ] = «
« −1
«
¬ 3
0
2
−1
1
2
3
2
−6
3º ª 1
4 »» «« 0
∼
1» « 0
» «
−4 ¼ ¬ 0
−1º ª 1
4 »» ««0
∼
6» «0
» «
−2 ¼ ¬ 0
0
1
0
0
0
1
0
0
3º
−1»»
, so {u1 , u 2 } is a basis for H.
0»
»
0¼
0º
0 »»
, so {v1 , v 2 , v 3 } is a basis for K.
1»
»
0¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.3
[u1
u2
u3
v1
v2
ª 1
« 2
v3 ] = «
« 0
«
¬ −1
ª1
«0
∼«
«0
«
¬0
0
2
3
4
−2
−2
2
3
−1
1
1
−4
−1
3
2
−6
0
1
3
−1
−2
1
0
0
0
0
0
0
0
0
1
0
• Solutions
221
−1º
4 »»
6»
»
−2 ¼
0º
0 »»
, so {u1 , u 2 , v 2 , v 3 } is a basis for H + K.
0»
»
1¼
37. [M] For example, writing
c1 ⋅ t + c2 ⋅ sin t + c3cos 2t + c4sin t cos t = 0
with t = 0, .1, .2, .3 gives the following coefficent matrix A for the homogeneous system Ac = 0 (to
four decimal places):
ª0
« .1
A=«
«.2
«
¬«.3
sin 0
sin .1
cos 0
cos .2
sin .2
sin .3
cos .4
cos .6
sin 0 cos 0 º ª 0
sin .1 cos .1»» ««.1
=
sin .2 cos .2 » «.2
» «
sin .3 cos .3¼» ¬«.3
0
.0998
1
.9801
.1987
.2955
.9211
.8253
0 º
.0993»»
.
.1947 »
»
.2823¼»
This matrix is invertible, so the system Ac = 0 has only the trivial solution and {t, sin t, cos 2t,
sin t cos t} is a linearly independent set of functions.
38. [M] For example, writing
c1 ⋅ 1 + c2 ⋅ cos t + c3 ⋅ cos 2t + c4 ⋅ cos3t + c5 ⋅ cos 4t + c6 ⋅ cos5t + c7 ⋅ cos 6t = 0
with t = 0, .1, .2, .3, .4, .5, .6 gives the following coefficent matrix A for the homogeneous system Ac
= 0 (to four decimal places):
ª1
«
«1
«
«1
A = «1
«
«1
«
«1
«
¬«1
ª1
«1
«
«1
«
= «1
«1
«
«1
«
¬1
cos 0
cos 2 0
cos3 0
cos 4 0
cos5 0
cos.1
cos 2 .1
cos3 .1
cos 4 .1
cos5 .1
cos.2
cos 2 .2
cos3 .2
cos 4 .2
cos5 .2
cos.3
cos 2 .3
cos3 .3
cos 4 .3
cos5 .3
cos.4
cos 2 .4
cos3 .4
cos 4 .4
cos5 .4
cos.5
cos 2 .5
cos3 .5
cos 4 .5
cos5 .5
cos.6
cos 2 .6
cos3 .6
cos 4 .6
cos5 .6
cos6 0 º
»
cos6 .1»
»
cos6 .2 »
cos6 .3»
»
cos6 .4 »
»
cos 6 .5 »
»
cos6 .6 ¼»
1
1
1
1
1
1 º
.9950 .9900 .9851 .9802 .9753 .9704»»
.9801 .9605 .9414 .9226 .9042 .8862 »
»
.9553 .9127 .8719 .8330 .7958 .7602»
.9211 .8484 .7814 .7197 .6629 .6106»
»
.8776 .7702 .6759 .5931 .5205 .4568»
.8253 .6812 .5622 .4640 .3830 .3161»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
222
CHAPTER 4
• Vector Spaces
This matrix is invertible, so the system Ac = 0 has only the trivial solution and {1, cos t, cos2t, cos3t,
cos4t, cos5t, cos6t} is a linearly independent set of functions.
4.4
SOLUTIONS
Notes: Section 4.7 depends heavily on this section, as does Section 5.4. It is possible to cover the
n
parts
of the two later sections, however, if the first half of Section 4.4 (and perhaps Example 7) is covered. The
linearity of the coordinate mapping is used in Section 5.4 to find the matrix of a transformation relative to
two bases. The change-of-coordinates matrix appears in Section 5.4, Theorem 8 and Exercise 27. The
concept of an isomorphism is needed in the proof of Theorem 17 in Section 4.8. Exercise 25 is used in
Section 4.7 to show that the change-of-coordinates matrix is invertible.
1. We calculate that
ª 3º
ª −4 º ª 3º
x = 5« » + 3« » = « ».
¬ −5 ¼
¬ 6 ¼ ¬ −7 ¼
2. We calculate that
ª3º
ª −4 º ª −26 º
x = ( −2) « » + 5 « » = «
».
¬2¼
¬ 1¼ ¬ 1¼
3. We calculate that
ª 1º
ª 5º
ª 4º ª −7 º
«
»
«
»
x = 1 « −2 » + 0 « 0 » + (−2) «« −3»» = «« 4»» .
«¬ 3»¼
«¬ −2 »¼
«¬ 0»¼ «¬ 3»¼
4. We calculate that
ª −2 º
ª 3º
ª 4 º ª 8º
«
»
«
»
x = (−3) « 2 » + 2 « 0 » + (−1) «« −1»» = «« −5»» .
«¬ 0 »¼
«¬ 2 »¼
«¬ 3»¼ «¬ 1»¼
5. The matrix [b1
b2
ª1
x] row reduces to «
¬0
0
6. The matrix [b1
b2
ª1
x] row reduces to «
¬0
0
7. The matrix [b1
b2
b3
1
1
ª1
«
x ] row reduces to «0
«¬0
2º
ª 2º
, so [ x ]B = « » .
»
−1¼
¬ −1¼
3º
ª 3º
, so [ x ]B = « » .
»
−2 ¼
¬ −2 ¼
0
0
1
0
0
1
−1º
ª −1º
»
−1» , so [x]B = «« −1»» .
«¬ 3»¼
3»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.4
8. The matrix [b1
b2
b3
ª1
«
x ] row reduces to «0
«¬0
0
0
1
0
0
1
2
is
10. The change-of-coordinates matrix from B to the standard basis in
2
1º
ª3
«
2 −2 »» .
PB = [b1 b 2 b3 ] = «0
«¬6 −4
3»¼
3
is
−1
converts x into its B-coordinate vector, we find that
ª 1
[x]B = PB x = «
¬ −2
−1
12. Since PB
−1
223
1º
ª 1º
»
−1» , so [x]B = «« −1»» .
«¬ 1»¼
1»¼
9. The change-of-coordinates matrix from B to the standard basis in
2º
ª 1
PB = [b1 b 2 ] = «
».
¬ −3 −5 ¼
11. Since PB
• Solutions
−1
−3º ª 2º ª −5
=
5»¼ «¬ −5»¼ «¬ −2
−3º ª 2º ª5º
=
.
−1»¼ «¬ −5»¼ «¬ 1»¼
converts x into its B-coordinate vector, we find that
ª 1
[x]B = PB x = «
¬ −1
−1
−1
2º ª 2º ª −1
=
−1»¼ «¬ 3»¼ «¬ 1
−2º ª 2º ª −8º
=
.
1»¼ «¬ 3»¼ «¬ 5»¼
13. We must find c1 , c2 , and c3 such that
c1 (1 + t 2 ) + c2 (t + t 2 ) + c3 (1 + 2t + t 2 ) = p(t ) = 1 + 4t + 7t 2 .
Equating the coefficients of the two polynomials produces the system of equations
c1 +
c3 = 1
c2 + 2c3 = 4
c1 + c2 + c3 = 7
We row reduce the augmented matrix for the system of equations to find
ª1
«0
«
«¬1
0
1
1
2
1
1
1º ª 1
4 »» ∼ «« 0
7 »¼ «¬ 0
0
0
1
0
0
1
2º
ª 2º
»
6» , so [p]B = «« 6»» .
«¬ −1»¼
−1»¼
One may also solve this problem using the coordinate vectors of the given polynomials relative to the
standard basis {1, t , t 2 }; the same system of linear equations results.
14. We must find c1 , c2 , and c3 such that
c1 (1 − t 2 ) + c2 (t − t 2 ) + c3 (1 − t + t 2 ) = p(t ) = 2 + 3t − 6t 2 .
Equating the coefficients of the two polynomials produces the system of equations
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
224
CHAPTER 4
c1
−c1
• Vector Spaces
+
c3
c2
− c2
− c3
+ c3
=
2
=
3
= −6
We row reduce the augmented matrix for the system of equations to find
ª 1
« 0
«
«¬ −1
0
1
1
−1
−1
1
2º ª 1
3»» ∼ ««0
−6 »¼ «¬0
0
0
1
0
0
1
3º
ª 3º
»
2» , so [p]B = «« 2»» .
«¬ −1»¼
−1»¼
One may also solve this problem using the coordinate vectors of the given polynomials relative to the
standard basis {1, t , t 2 }; the same system of linear equations results.
15. a. True. See the definition of the B-coordinate vector.
b. False. See Equation (4).
c. False. 3 is isomorphic to 4. See Example 5.
16. a. True. See Example 2.
b. False. By definition, the coordinate mapping goes in the opposite direction.
c. True. If the plane passes through the origin, as in Example 7, the plane is isomorphic to
2
.
ª 1º
ª 2º
ª −3º ª1º
17. We must solve the vector equation x1 « » + x2 « » + x3 « » = « » . We row reduce the augmented
¬ −3 ¼
¬ −8 ¼
¬ 7 ¼ ¬1¼
matrix for the system of equations to find
ª 1
« −3
¬
2
−3
−8
7
1º ª 1
∼
1»¼ «¬ 0
0
−5
1
1
5º
.
−2 »¼
Thus we can let x1 = 5 + 5 x3 and x2 = −2 − x3 , where x3 can be any real number. Letting x3 = 0 and
ª1º
x3 = 1 produces two different ways to express « » as a linear combination of the other vectors:
¬1¼
5 v1 − 2 v 2 and 10 v 1 − 3v 2 + v 3 . There are infintely many correct answers to this problem.
18. For each k, b k = 0 ⋅ b1 + ⋅ ⋅⋅ + 1 ⋅ b k + ⋅ ⋅⋅ + 0 ⋅ b n , so [b k ]B = (0,…,1,…,0) = e k .
19. The set S spans V because every x in V has a representation as a (unique) linear combination of
elements in S. To show linear independence, suppose that S = {v1 ,…, v n } and that
c1 v1 + ⋅ ⋅ ⋅ + cn v n = 0 for some scalars c1 , …, cn . The case when c1 = ⋅ ⋅⋅ = cn = 0 is one possibility.
By hypothesis, this is the unique (and thus the only) possible representation of the zero vector as a
linear combination of the elements in S. So S is linearly independent and is thus a basis for V.
20. For w in V there exist scalars k1 , k2 , k3 , and k4 such that
w = k1 v1 + k2 v 2 + k3 v 3 + k 4 v 4
(1)
because {v1 , v 2 , v 3 , v 4 } spans V. Because the set is linearly dependent, there exist scalars c1 , c2 , c3 ,
and c4 not all zero, such that
0 = c1 v1 + c2 v 2 + c3 v 3 + c4 v 4
Adding (1) and (2) gives
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
(2)
4.4
• Solutions
225
w = w + 0 = ( k1 + c1 ) v1 + ( k2 + c2 ) v 2 + ( k3 + c3 ) v 3 + ( k 4 + c4 ) v 4
(3)
At least one of the weights in (3) differs from the corresponding weight in (1) because at least one of
the ci is nonzero. So w is expressed in more than one way as a linear combination of v1 , v 2 , v 3 ,
and v 4 .
ª 1
−1
21. The matrix of the transformation will be PB = «
¬ −4
22. The matrix of the transformation will be PB
−1
−1
−2 º
ª9
=«
»
9¼
¬4
2º
.
1»¼
= [b1 ⋅ ⋅⋅ b n ] .
−1
23. Suppose that
ª c1 º
[u]B = [w ]B = «« # »» .
«¬cn »¼
By definition of coordinate vectors,
u = w = c1b1 + ⋅ ⋅ ⋅ + cn b n .
Since u and w were arbitrary elements of V, the coordinate mapping is one-to-one.
24. Given y = ( y1 ,…, yn ) in n, let u = y1b1 + ⋅ ⋅⋅ + yn b n . Then, by definition, [u ]B = y . Since y was
arbitrary, the coordinate mapping is onto n.
25. Since the coordinate mapping is one-to-one, the following equations have the same solutions
c1 ,…, c p :
c1u1 + ⋅⋅ ⋅ + c p u p = 0
(the zero vector in V )
ª¬ c1u1 + ⋅ ⋅ ⋅ + c p u p º¼ = [ 0 ]B
B
(the zero vector in
n
)
(4)
(5)
Since the coordinate mapping is linear, (5) is equivalent to
ª0º
c1[u1 ]B + ⋅⋅⋅ + c p [u p ]B = «« # »»
«¬0 »¼
(6)
Thus (4) has only the trivial solution if and only if (6) has only the trivial solution. It follows that
{u1 ,…, u p } is linearly independent if and only if {[u1 ]B ,…,[u p ]B } is linearly independent. This
result also follows directly from Exercises 31 and 32 in Section 4.3.
26. By definition, w is a linear combination of u1 ,…, u p if and only if there exist scalars c1 ,…, c p such
that
w = c1u1 + ⋅ ⋅ ⋅ + c p u p
(7)
Since the coordinate mapping is linear,
[w ]B = c1[u1 ]B + ⋅⋅ ⋅ + c p [u p ]B
Conversely, (8) implies (7) because the coordinate mapping is one-to-one. Thus w is a linear
combination of u1 ,…, u p if and only if [ w ]B is a linear combination of [u1 ]B ,…,[u p ]B .
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
(8)
226
CHAPTER 4
• Vector Spaces
Note: Students need to be urged to write not just to compute in Exercises 27–34. The language in the
Study Guide solution of Exercise 31 provides a model for the students. In Exercise 32, students may have
difficulty distinguishing between the two isomorphic vector spaces, sometimes giving a vector in 3 as an
answer for part (b).
27. The coordinate mapping produces the coordinate vectors (1, 0, 0, 2), (2, 1, –3, 0), and (0, –1, 2, –1)
respectively. We test for linear independence of these vectors by writing them as columns of a matrix
and row reducing:
ª1
«0
«
«0
«
¬2
0º ª 1
−1»» ««0
∼
2» «0
» «
−1¼ ¬0
2
1
−3
0
0º
0 »»
.
1»
»
0¼
0
1
0
0
Since the matrix has a pivot in each column, its columns (and thus the given polynomials) are
linearly independent.
28. The coordinate mapping produces the coordinate vectors (1, 0, –2, –1), (0, 1, 0, 2), and (1, 1, –2, 0)
respectively. We test for linear independence of these vectors by writing them as columns of a matrix
and row reducing:
ª 1
« 0
«
« −2
«
¬ −1
1º ª 1
1»» «« 0
∼
−2 » « 0
» «
0¼ ¬0
0
1
0
2
0º
0»»
.
1»
»
0¼
0
1
0
0
Since the matrix has a pivot in each column, its columns (and thus the given polynomials) are
linearly independent.
29. The coordinate mapping produces the coordinate vectors (1, –2, 1, 0), (0, 1, –2, 1), and (1, –3, 3, –1)
respectively. We test for linear independence of these vectors by writing them as columns of a matrix
and row reducing:
ª 1
« −2
«
« 1
«
¬ 0
0
1
−2
1
1º ª 1
−3»» ««0
∼
3» «0
» «
−1¼ ¬0
0
1
0
0
1º
−1»»
.
0»
»
0¼
Since the matrix does not have a pivot in each column, its columns (and thus the given polynomials)
are linearly dependent.
30. The coordinate mapping produces the coordinate vectors (8, –12, 6, –1), (9, –6, 1, 0), and (1, 6, –5,1)
respectively. We test for linear independence of these vectors by writing them as columns of a matrix
and row reducing:
ª 8
« −12
«
« 6
«
¬ −1
9
−6
1
0
1º ª 1
6 »» ««0
∼
−5» « 0
» «
1¼ ¬ 0
0
1
0
0
−1º
1»»
.
0»
»
0¼
Since the matrix does not have a pivot in each column, its columns (and thus the given polynomials)
are linearly dependent.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.4
• Solutions
227
31. In each part, place the coordinate vectors of the polynomials into the columns of a matrix and reduce
the matrix to echelon form.
ª 1
« −3
a. «
«¬ 5
−3
−4
5
5
−7
−6
1º ª 1
0 »» ∼ ««0
−1»¼ «¬0
−3
−4
−4
−7
0
0
1º
3»»
0»¼
Since there is not a pivot in each row, the original four column vectors do not span
isomorphism between 3 and 2, the given set of polynomials does not span 2.
3
. By the
1 −3 2 º ª 1 −2 2
0º
ª0
« 5 −8 4 −3» ∼ «0 2 −6 −3»
b. «
» «
»
«¬ 1 −2 2 0 ¼» «¬0 0 0 7 / 2 »¼
Since there is a pivot in each row, the original four column vectors span
between 3 and 2, the given set of polynomials spans 2.
3
. By the isomorphism
32. a. Place the coordinate vectors of the polynomials into the columns of a matrix and reduce the
1º ª 1 0 1º
ª1 0
«0
1 1»» ∼ ««0 1 1»»
matrix to echelon form: «
«¬ 1 −3 −3»¼ «¬0 0 −1»¼
The resulting matrix is invertible since it row equivalent to I 3 . The original three column vectors
form a basis for 3 by the Invertible Matrix Theorem. By the isomorphism between 3 and 2, the
corresponding polynomials form a basis for 2.
b. Since [q ]B = ( −1, 1, 2), q = −p1 + p 2 + 2p 3 . One might do the algebra in 2 or choose to compute
1º ª −1º ª 1º
ª1 0
«0
1 1»» «« 1»» = «« 3»» . This combination of the columns of the matrix corresponds to the
«
¬« 1 −3 −3¼» ¬« 2 ¼» ¬« −10¼»
same combination of p1 , p 2 , and p 3 . So q(t ) = 1 + 3t − 10t 2 .
33. The coordinate mapping produces the coordinate vectors (3, 7, 0, 0), (5, 1, 0, –2), (0, 1, –2, 0) and
(1, 16, –6, 2) respectively. To determine whether the set of polynomials is a basis for 3, we
investigate whether the coordinate vectors form a basis for 4. Writing the vectors as the columns of
a matrix and row reducing
1º ª 1
ª3 5 0
«7
1 1 16 »» ««0
«
∼
« 0 0 −2 −6 » «0
«
» «
¬« 0 −2 0 2 ¼» ¬«0
0 0 2º
1 0 −1»»
,
0 1 3»
»
0 0 0 ¼»
we find that the matrix is not row equivalent to I 4 . Thus the coordinate vectors do not form a basis
for 4. By the isomorphism between 4 and 3, the given set of polynomials does not form a basis for
3.
34. The coordinate mapping produces the coordinate vectors (5, –3, 4, 2), (9, 1, 8, –6), (6, –2, 5, 0), and
(0, 0, 0, 1) respectively. To determine whether the set of polynomials is a basis for 3, we investigate
whether the coordinate vectors form a basis for 4. Writing the vectors as the columns of a matrix,
and row reducing
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
228
CHAPTER 4
ª 5
« −3
«
« 4
«
¬« 2
• Vector Spaces
9
1
6
−2
8
−6
5
0
0º ª 1
0 »» ««0
∼
0» «0
» «
1¼» ¬«0
0
1
3/ 4
1/ 4
0
0
0
0
0º
0 »»
1»
»
0¼»
we find that the matrix is not row equivalent to I4. Thus the coordinate vectors do not form a basis for
4
. By the isomorphism between 4 and 3, the given set of polynomials does not form a basis for 3.
35. To show that x is in H = Span{v1 , v 2 }, we must show that the vector equation x1 v1 + x2 v 2 = x has a
solution. The augmented matrix [ v1
ª 11
« −5
«
« 10
«
¬« 7
14
−8
13
10
19 º ª 1
−13»» ««0
∼
18» «0
» «
15¼» ¬«0
0
1
0
0
v2
x] may be row reduced to show
−5/ 3º
8/ 3»»
.
0»
»
0 ¼»
Since this system has a solution, x is in H. The solution allows us to find the B-coordinate vector for
ª −5 / 3 º
x: since x = x1 v1 + x2 v 2 = ( −5 / 3) v1 + (8 / 3) v 2 , [ x ]B = «
».
¬ 8 / 3¼
36. To show that x is in H = Span{v1 , v 2 , v 3 } , we must show that the vector equation
x1 v1 + x2 v 2 + x3 v 3 = x has a solution. The augmented matrix [ v1
reduced to show
ª −6
« 4
«
« −9
«
¬« 4
8
−3
−9
5
7
−3
−8
3
4º ª 1
7 »» ««0
∼
−8» «0
» «
3¼» ¬«0
0
1
0
0
0
0
1
0
v2
v3
x ] may be row
3º
5»»
.
2»
»
0¼»
The first three columns show that B is a basis for H. Moreover, since this system has a solution, x is
in H. The solution allows us to find the B-coordinate vector for x: since
ª3º
« »
x = x1 v1 + x2 v 2 + x3 v 3 = 3v1 + 5 v 2 + 2 v 3 , [ x]B = « 5 » .
¬« 2 ¼»
ª1/ 2 º
°
«
»
37. We are given that [x]B = «1/ 4 » , where B = ®
°
«¬1/ 6 »¼
¯
ª 2.6 º ª0 º ª 0 º ½
« −1.5» , « 3» , « 0 » ° . To find the coordinates of x
«
» « » « »¾
«¬ 0 »¼ «¬0»¼ «¬ 4.8»¼ ¿°
relative to the standard basis in 3, we must find x. We compute that
ª 2.6
x = PB [x]B = «« −1.5
¬« 0
0
3
0
0 º ª1/ 2º ª1.3 º
0 »» ««1/ 4»» = «« 0 »» .
4.8¼» ¬«1/ 6 ¼» ¬« 0.8¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.5
• Solutions
229
ª1/ 2 º
°
«
»
38. We are given that [x]B = «1/ 2 » , where B = ®
°
«¬1/ 3»¼
¯
ª 2.6 º ª0 º ª 0 º ½
« −1.5» , « 3» , « 0 » ° . To find the coordinates of x
«
» « » « »¾
«¬ 0 »¼ «¬0»¼ «¬ 4.8»¼ ¿°
relative to the standard basis in 3, we must find x. We compute that
ª 2.6
x = PB [x]B = «« −1.5
«¬ 0
4.5
0
3
0
0 º ª1/ 2 º ª 1.3º
0 »» ««1/ 2»» = «« 0.75»» .
4.8»¼ «¬1/ 3»¼ «¬ 1.6»¼
SOLUTIONS
Notes: Theorem 9 is true because a vector space isomorphic to
n
has the same algebraic properties as
; a proof of this result may not be needed to convince the class. The proof of Theorem 9 relies upon the
fact that the coordinate mapping is a linear transformation (which is Theorem 8 in Section 4.4). If you
have skipped this result, you can prove Theorem 9 as is done in Introduction to Linear Algebra by Serge
Lang (Springer-Verlag, New York, 1986). There are two separate groups of true-false questions in this
section; the second batch is more theoretical in nature. Example 4 is useful to get students to visualize
subspaces of different dimensions, and to see the relationships between subspaces of different
dimensions. Exercises 31 and 32 investigate the relationship between the dimensions of the domain and
the range of a linear transformation; Exercise 32 is mentioned in the proof of Theorem 17 in Section 4.8.
n
ª1 º
ª −2º
«
»
« »
1. This subspace is H = Span{v1 , v 2 }, where v1 = «1 » and v 2 = « 1» . Since v1 and v 2 are not
«¬ 0 »¼
«¬ 3»¼
multiples of each other, {v1 , v 2 } is linearly independent and is thus a basis for H. Hence the
dimension of H is 2.
ª 2º
ª 0º
«
»
2. This subspace is H = Span{v1 , v 2 }, where v1 = « 0 » and v 2 = «« −4»» . Since v1 and v 2 are not
«¬ −2 »¼
«¬ 0»¼
multiples of each other, {v1 , v 2 } is linearly independent and is thus a basis for H. Hence the
dimension of H is 2.
ª0º
ª 0º
ª 2º
«1 »
« −1»
« 0»
«
»
«
»
, v =
, and v 3 = « » . Theorem 4 in
3. This subspace is H = Span{v1 , v 2 , v 3 }, where v1 =
«0 » 2 « 1»
« −3»
« »
« »
« »
«¬1 »¼
«¬ 2»¼
«¬ 0 »¼
Section 4.3 can be used to show that this set is linearly independent: v1 ≠ 0, v 2 is not a multiple of
v1 , and (since its first entry is not zero) v 3 is not a linear combination of v1 and v 2 . Thus
{v1 , v 2 , v 3 } is linearly independent and is thus a basis for H. Alternatively, one can show that this set
is linearly independent by row reducing the matrix [ v1
subspace is 3.
v2
v3
0]. Hence the dimension of the
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
230
CHAPTER 4
• Vector Spaces
ª 1º
ª 2º
« −1»
« 0»
«
»
4. This subspace is H = Span{v1 , v 2 }, where v1 =
and v 2 = « » . Since v1 and v 2 are not
« 3»
« −1»
« »
« »
¬ 1¼
¬ 1¼
multiples of each other, {v1 , v 2 } is linearly independent and is thus a basis for H. Hence the
dimension of H is 2.
ª 1º
ª −2 º
ª0º
« 2»
« 0»
« 5»
5. This subspace is H = Span{v1 , v 2 , v 3 }, where v1 = « » , v 2 = « » , and v 3 = « » . The matrix A
« 0»
« −2 »
«2»
« »
« »
« »
¬ −3¼
¬ 0¼
¬6¼
ª 1 −2 0 º ª 1 0 0º
« 2
0 5»» ««0 1 0»»
∼
. There is a pivot in
with these vectors as its columns row reduces to «
« 0 −2 2 » «0 0 1»
«
» «
»
0 6 ¼ ¬0 0 0 ¼
¬ −3
each column, so {v1 , v 2 , v 3 } is linearly independent and is thus a basis for H. Hence the dimension
of H is 3.
ª 3º
ª 0º
ª −1º
« 0»
« −1»
« −3»
«
»
«
»
, v2 =
, and v 3 = « » . The matrix A
6. This subspace is H = Span{v1 , v 2 , v 3 }, where v1 =
« −7 »
« 6»
« 5»
« »
« »
« »
¬ −3¼
¬ 0¼
¬ 1¼
0 −1º ª 1 0 0 º
ª 3
« 0 −1 −3» «0 1 0 »
»∼«
» . There is a pivot in
with these vectors as its columns row reduces to «
« −7
6
5» «0 0 1»
«
» «
»
0
1¼ ¬0 0 0 ¼
¬ −3
each column, so {v1 , v 2 , v 3 } is linearly independent and is thus a basis for H. Hence the dimension
of H is 3.
1º
ª 1 −3
ª1 0 0 0 º
«
»
1 −2 » . Since [ A 0] ∼ ««0 1 0 0 »» , the
7. This subspace is H = Nul A, where A = «0
«¬0
«¬0 0 1 0 »¼
2 −1»¼
homogeneous system has only the trivial solution. Thus H = Nul A = {0}, and the dimension of H is
0.
8. From the equation a – 3b + c = 0, it is seen that (a, b, c, d) = b(3, 1, 0, 0) + c(–1, 0, 1, 0) + d(0, 0, 0,
1). Thus the subspace is H = Span{v1 , v 2 , v 3 }, where v1 = (3,1, 0, 0), v 2 = ( −1,0,1, 0), and
v 3 = (0, 0,0,1). It is easily checked that this set of vectors is linearly independent, either by appealing
to Theorem 4 in Section 4.3, or by row reducing [ v1
subspace is 3.
v2
v3
0]. Hence the dimension of the
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.5
• Solutions
231
ªa º
ª0 º
½
ª1 º
°
°« »
« »
«
»
9. This subspace is H = ® «b » : a, b in
¾ = Span{v1 , v 2 }, where v1 = « 0 » and v 2 = «1 » . Since v1
°
° «a »
«¬0 »¼
«¬1 »¼
¿
¯¬ ¼
and v 2 are not multiples of each other, {v1 , v 2 } is linearly independent and is thus a basis for H.
Hence the dimension of H is 2.
10. The matrix A with these vectors as its columns row reduces to
ª 1 −2 −3º ª 1 −2 −3º
.
« −5 10 15» ∼ « 0
0
0 »¼
¬
¼ ¬
There is one pivot column, so the dimension of Col A (which is the dimension of H) is 1.
11. The matrix A with these vectors as its columns row reduces to
ª1
«0
«
¬« 2
3
−2
1
−1
1
1
5º ª 1
2 »» ∼ «« 0
2 ¼» ¬« 0
0
1
1
−1
0
0
0º
0 »» .
1¼»
There are three pivot columns, so the dimension of Col A (which is the dimension of the subspace
spanned by the vectors) is 3.
12. The matrix A with these vectors as its columns row reduces to
ª 1
« −2
«
«¬ 0
−3
−2
−6
3
6
5
−3º ª 1
5»» ∼ «« 0
5»¼ «¬0
0
0
1
0
0
1
−1º
0 »» .
1»¼
There are three pivot columns, so the dimension of Col A (which is the dimension of the subspace
spanned by the vectors) is 3.
13. The matrix A is in echelon form. There are three pivot columns, so the dimension of Col A is 3.
There are two columns without pivots, so the equation Ax = 0 has two free variables. Thus the
dimension of Nul A is 2.
14. The matrix A is in echelon form. There are four pivot columns, so the dimension of Col A is 4. There
are three columns without pivots, so the equation Ax = 0 has three free variables. Thus the dimension
of Nul A is 3.
15. The matrix A is in echelon form. There are three pivot columns, so the dimension of Col A is 3.
There are two columns without pivots, so the equation Ax = 0 has two free variables. Thus the
dimension of Nul A is 2.
16. The matrix A row reduces to
ª 3 2 º ª1 0 º
« −6 5 » ∼ « 0 1 » .
¬
¼ ¬
¼
There are two pivot columns, so the dimension of Col A is 2. There are no columns without pivots,
so the equation Ax = 0 has only the trivial solution 0. Thus Nul A = {0}, and the dimension of Nul A
is 0.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
232
CHAPTER 4
• Vector Spaces
17. The matrix A is in echelon form. There are three pivot columns, so the dimension of Col A is 3.
There are no columns without pivots, so the equation Ax = 0 has only the trivial solution 0. Thus Nul
A = {0}, and the dimension of Nul A is 0.
18. The matrix A is in echelon form. There are two pivot columns, so the dimension of Col A is 2. There
is one column without a pivot, so the equation Ax = 0 has one free variable. Thus the dimension of
Nul A is 1.
19. a.
b.
c.
d.
e.
True. See the box before Example 5.
False. The plane must pass through the origin; see Example 4.
False. The dimension of n is n + 1; see Example 1.
False. The set S must also have n elements; see Theorem 12.
True. See Theorem 9.
20. a. False. The set 2 is not even a subset of 3.
b. False. The number of free variables is equal to the dimension of Nul A; see the box before
Example 5.
c. False. A basis could still have only finitely many elements, which would make the vector space
finite-dimensional.
d. False. The set S must also have n elements; see Theorem 12.
e. True. See Example 4.
21. The matrix whose columns are the coordinate vectors of the Hermite polynomials relative to the
standard basis {1, t , t 2 , t 3 } of 3 is
ª1
«0
A=«
«0
«
¬«0
0
2
−2
0
0
0
4
0
0º
−12»»
.
0»
»
8¼»
This matrix has 4 pivots, so its columns are linearly independent. Since their coordinate vectors form
a linearly independent set, the Hermite polynomials themselves are linearly independent in 3. Since
there are four Hermite polynomials and dim 3 = 4, the Basis Theorem states that the Hermite
polynomials form a basis for 3.
22. The matrix whose columns are the coordinate vectors of the Laguerre polynomials relative to the
standard basis {1, t , t 2 , t 3 } of 3 is
ª1
«0
A=«
«0
«
¬«0
1
−1
2
−4
0
0
1
0
6º
−18»»
.
9»
»
−1¼»
This matrix has 4 pivots, so its columns are linearly independent. Since their coordinate vectors form
a linearly independent set, the Laguerre polynomials themselves are linearly independent in 3. Since
there are four Laguerre polynomials and dim 3 = 4, the Basis Theorem states that the Laguerre
polynomials form a basis for 3.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.5
• Solutions
233
23. The coordinates of p(t ) = −1 + 8t 2 + 8t 3 with respect to B satisfy
c1 (1) + c2 (2t ) + c3 (−2 + 4t 2 ) + c4 ( −12t + 8t 3 ) = −1 + 8t 2 + 8t 3
Equating coefficients of like powers of t produces the system of equations
−
c1
12c4
=
=
=
−1
0
8
8c4
=
8
2c3
−
2c2
4c3
ª 3º
«6»
Solving this system gives c1 = 3, c2 = 6, c3 = 2, c4 = 1, and [p]B = « » .
«2»
« »
¬ 1¼
24. The coordinates of p(t ) = 5 + 5t − 2t 2 with respect to B satisfy
c1 (1) + c2 (1 − t ) + c3 (2 − 4t + t 2 ) = 5 + 5t − 2t 2
Equating coefficients of like powers of t produces the system of equations
c1 +
c2 + 2c3 =
5
−c2 − 4c3 =
5
c3 =
−2
ª 6º
« »
Solving this system gives c1 = 6, c2 = 3, c3 = −2, and [p]B = « 3» .
«¬ −2»¼
25. Note first that n ≥ 1 since S cannot have fewer than 1 vector. Since n ≥ 1, V ≠ 0. Suppose that S spans
V and that S contains fewer than n vectors. By the Spanning Set Theorem, some subset S ′ of S is a
basis for V. Since S contains fewer than n vectors, and S ′ is a subset of S, S ′ also contains fewer
than n vectors. Thus there is a basis S ′ for V with fewer than n vectors, but this is impossible by
Theorem 10 since dimV = n. Thus S cannot span V.
26. If dimV = dim H = 0, then V = {0} and H = {0}, so H = V. Suppose that dim V = dim H > 0. Then H
contains a basis S consisting of n vectors. But applying the Basis Theorem to V, S is also a basis for
V. Thus H = V = SpanS.
27. Suppose that dim = k < ∞. Now n is a subspace of for all n, and dim k–1 = k, so dim k–1 = dim
. This would imply that k–1 = , which is clearly untrue: for example, p(t ) = t k is in but not in
cannot be finite.
k–1. Thus the dimension of
28. The space C( ) contains as a subspace. If C( ) were finite-dimensional, then would also be
finite-dimensional by Theorem 11. But is infinite-dimensional by Exercise 27, so C( ) must also
be infinite-dimensional.
29. a. True. Apply the Spanning Set Theorem to the set {v1 ,…, v p } and produce a basis for V. This
basis will not have more than p elements in it, so dimV ≤ p.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
234
CHAPTER 4
• Vector Spaces
b. True. By Theorem 11, {v1 ,…, v p } can be expanded to find a basis for V. This basis will have at
least p elements in it, so dimV ≥ p.
c. True. Take any basis (which will contain p vectors) for V and adjoin the zero vector to it.
30. a. False. For a counterexample, let v be a non-zero vector in
a linearly dependent set in 3, but dim 3 = 3 > 2 .
3
, and consider the set {v, 2v}. This is
b. True. If dimV ≤ p, there is a basis for V with p or fewer vectors. This basis would be a spanning
set for V with p or fewer vectors. If necessary, vectors in V could be added to this spanning set to
give a spanning set for V with exactly p vectors, which contradicts the assumption.
c. False. For a counterexample, let v be a non-zero vector in 3, and consider the set {v, 2v}. This is
a linearly dependent set in 3 with 3 – 1 = 2 vectors, and dim 3 = 3 .
31. Since H is a nonzero subspace of a finite-dimensional vector space V, H is finite-dimensional and has
a basis. Let {u1 ,…, u p } be a basis for H. We show that the set {T (u1 ),…, T (u p )} spans T(H). Let y
be in T(H). Then there is a vector x in H with T(x) = y. Since x is in H and {u1 ,…, u p } is a basis for
H, x may be written as x = c1u1 + … + c p u p for some scalars c1 ,…, c p . Since the transformation T is
linear,
y = T (x) = T (c1u1 + … + c p u p ) = c1T (u1 ) + … + c pT (u p )
Thus y is a linear combination of T (u1 ),…, T (u p ) , and {T (u1 ),…, T (u p )} spans T(H). By the
Spanning Set Theorem, this set contains a basis for T(H). This basis then has not more than p vectors,
and dimT(H) ≤ p = dim H.
32. Since H is a nonzero subspace of a finite-dimensional vector space V, H is finite-dimensional and has
a basis. Let {u1 ,…u p } be a basis for H. In Exercise 31 above it was shown that {T (u1 ),…, T (u p )}
spans T(H). In Exercise 32 in Section 4.3, it was shown that {T (u1 ),…, T (u p )} is linearly
independent. Thus {T (u1 ),…, T (u p )} is a basis for T(H), and dimT(H) = p = dim H.
33. [M]
a. To find a basis for
ª −9
« −7
«
« 8
«
« −5
«¬ 7
5
which contains the given vectors, we row reduce
9
6
1
0
0
0
4
7
0
1
0
0
1
−8
0
0
1
0
6
5
0
0
0
1
−7
−7
0
0
0
0
0º ª 1
0 »» ««0
0 » ∼ «0
» «
0 » «0
1»¼ «¬0
0
0
−1/ 3
0
0
1
1
0
0
0
0
1
0
1
−1/ 3
0
0
0
0
0
0
1
0
3
0
0
0
0
1
−9
3/ 7 º
5 / 7 »»
−3/ 7 » .
»
22 / 7 »
−53/ 7 »¼
The first, second, third, fifth, and sixth columns are pivot columns, so these columns of the
original matrix ( {v1 , v 2 , v 3 , e 2 , e3 } ) form a basis for 5:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.6
• Solutions
235
b. The original vectors are the first k columns of A. Since the set of original vectors is assumed
to be linearly independent, these columns of A will be pivot columns and the original set of
vectors will be included in the basis. Since the columns of A include all the columns of the
identity matrix, Col A = n.
34. [M]
a. The B-coordinate vectors of the vectors in C are the columns of the matrix
ª1
«0
«
«0
«
P = «0
«0
«
«0
«0
¬
0
−1
0
1
0
1
0
−3
0
5
0
2
0
−8
0
0
0
4
0
−20
0
0
0
8
0
0
0
0
0
16
0
0
0
0
0
−1º
0 »»
18»
»
0» .
−48»
»
0»
32 »¼
The matrix P is invertible because it is triangular with nonzero entries along its main
diagonal. Thus its columns are linearly independent. Since the coordinate mapping is an
isomorphism, this shows that the vectors in C are linearly independent.
b. We know that dim H = 7 because B is a basis for H. Now C is a linearly independent set, and
the vectors in C lie in H by the trigonometric identities. Thus by the Basis Theorem, C is a
basis for H.
4.6
SOLUTIONS
Notes: This section puts together most of the ideas from Chapter 4. The Rank Theorem is the main result
in this section. Many students have difficulty with the difference in finding bases for the row space and
the column space of a matrix. The first process uses the nonzero rows of an echelon form of the matrix.
The second process uses the pivots columns of the original matrix, which are usually found through row
reduction. Students may also have problems with the varied effects of row operations on the linear
dependence relations among the rows and columns of a matrix. Problems of the type found in Exercises
19–26 make excellent test questions. Figure 1 and Example 4 prepare the way for Theorem 3 in Section
6.1; Exercises 27–29 anticipate Example 6 in Section 7.4.
1. The matrix B is in echelon form. There are two pivot columns, so the dimension of Col A is 2. There
are two pivot rows, so the dimension of Row A is 2. There are two columns without pivots, so the
equation Ax = 0 has two free variables. Thus the dimension of Nul A is 2. A basis for Col A is the
pivot columns of A:
°
®
°
¯
ª 1º ª −4º
« −1» , « 2»
« » « »
¬« 5¼» ¬« −6¼»
½
°
¾.
°
¿
A basis for Row A is the pivot rows of B: {(1, 0, −1,5),(0, −2,5, −6)}. To find a basis for Nul A row
reduce to reduced echelon form:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
236
CHAPTER 4
ª1
A∼«
¬0
• Vector Spaces
0
−1
1
−5 / 2
5º
.
3»¼
The solution to Ax = 0 in terms of free variables is x1 = x3 − 5 x4 , x2 = (5 / 2) x3 − 3 x4 with x3 and x4
free. Thus a basis for Nul A is
°
°
®
°
°
¯
ª 1º ª −5º
«5 / 2 » « −3»
«
»,« »
« 1» « 0 »
«
» « »
«¬ 0 »¼ «¬ 1»¼
½
°
°
¾.
°
°
¿
2. The matrix B is in echelon form. There are three pivot columns, so the dimension of Col A is 3.
There are three pivot rows, so the dimension of Row A is 3. There are two columns without pivots,
so the equation Ax = 0 has two free variables. Thus the dimension of Nul A is 2. A basis for Col A is
the pivot columns
of A:
°
°
®
°
°
¯
ª 1º ª 4 º ª 2 º
« 2 » « 6 » « −3»
« »,« »,« »
« 3» « 3» « −3»
« » « » « »
¬ 3¼ ¬ 0 ¼ ¬ 0 ¼
½
°
°
¾.
°
°
¿
A basis for Row A is the pivot rows of B: {(1,3, 4, −1, 2),(0, 0,1, −1,1), (0,0,0, 0, −5)}. To find a basis
for Nul A row reduce to reduced echelon form:
ª1
«0
A∼«
«0
«
¬0
3
0
0
0
0
1
0
0
3
−1
0
0
0º
0»»
.
1»
»
0¼
The solution to Ax = 0 in terms of free variables is x1 = −3 x2 − 3 x4 , x3 = x4 , x5 = 0 , with x2 and x4
free. Thus a basis for Nul A is
°
°
°
®
°
°
°¯
ª −3º ª −3º
« 1» « 0 »
« » « »
« 0 » , « 1»
« » « »
« 0 » « 1»
«¬ 0 »¼ «¬ 0 »¼
½
°
°
°
¾.
°
°
°¿
3. The matrix B is in echelon form. There are three pivot columns, so the dimension of Col A is 3.
There are three pivot rows, so the dimension of Row A is 3. There are three columns without pivots,
so the equation Ax = 0 has three free variables. Thus the dimension of Nul A is 3. A basis for Col A
is the pivot columns of A:
°
°
®
°
°
¯
ª 2 º ª 6 º ª 3º
« −2 » « −3» « 0 »
« »,« »,« »
« 4 » « 9 » « 3»
« » « » « »
¬ −2 ¼ ¬ 3¼ ¬ 3¼
½
°
°
¾.
°
°
¿
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.6
• Solutions
237
A basis for Row A is the pivot rows of B: {(2,6, −6,6,3,6),(0,3,0,3,3,0),(0,0, 0,0,3, 0)} . To find a
basis for Nul A row reduce to reduced echelon form:
ª1
«0
A∼«
«0
«
¬0
0
1
0
0
−3
0
0
0
0
1
0
0
0
0
1
0
3º
0»»
.
0»
»
0¼
The solution to Ax = 0 in terms of free variables is x1 = 3 x3 − 3 x6 , x2 = − x4 , x5 = 0, with x3 , x4 ,
and x6 free. Thus a basis for Nul A is
°
°
°°
®
°
°
°
¯°
ª 3º ª 0 º ª −3º
« 0 » « −1» « 0 »
« » « » « »
« 1» « 0 » « 0 »
« » , « » ,« »
« 0 » « 1» « 0 »
«0» « 0 » « 0 »
« » « » « »
«¬ 0 »¼ «¬ 0 »¼ «¬ 1»¼
½
°
°
°°
¾.
°
°
°
¿°
4. The matrix B is in echelon form. There are five pivot columns, so the dimension of Col A is 5. There
are five pivot rows, so the dimension of Row A is 5. There is one column without a pivot, so the
equation Ax = 0 has one free variable. Thus the dimension of Nul A is 1. A basis for Col A is the
pivot columns of A:
°
°
°
®
°
°
¯°
ª1º ª 1º ª −2 º ª 1º ª −2º
«1» « 2 » « −3» « −2 » « −3»
«» « » « » « » « »
«1» , « −1» , « 0 » , « 1» , « 6»
«» « » « » « » « »
«1» « −2 » « 2 » « −3» « 0»
«¬1»¼ «¬ −2 »¼ «¬ 1»¼ «¬ 2 »¼ «¬ −1»¼
½
°
°
°
¾.
°
°
¿°
A basis for Row A is the pivot rows of B:
{(1, 1, − 2, 0, 1, − 2), (0, 1, − 1, 0, − 3, − 1), (0, 0, 1, 1, − 13, − 1), (0,0, 0, 0,1, −1),(0,0,0,0, 0,1)}.
To find a basis for Nul A row reduce to reduced echelon form:
ª1
«0
«
A ∼ «0
«
«0
«¬ 0
0º
1 0 1 0 0 »»
0 1 1 0 0» .
»
0 0 0 1 0»
0 0 0 0 1»¼
The solution to Ax = 0 in terms of free variables is x1 = − x4 , x2 = − x4 , x3 = − x4 , x5 = 0 , x6 = 0 ,
with x4 free. Thus a basis for Nul A is
0
0
1
0
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
238
CHAPTER 4
°
°
°°
®
°
°
°
¯°
ª −1º
« −1»
« »
« −1»
« »
« 1»
« 0»
« »
¬« 0 ¼»
• Vector Spaces
½
°
°
°°
¾.
°
°
°
¿°
5. By the Rank Theorem, dimNul A = 7 − rank A = 7 − 3 = 4. Since
dimRow A = rank A,dimRow A = 3. Since rank AT = dimCol AT = dimRow A, rankAT = 3.
6. By the Rank Theorem, dimNul A = 5 − rank A = 5 − 2 = 3. Since
dimRow A = rank A, dimRow A = 2. Since rank AT = dimCol AT = dimRow A, rank AT = 2.
7. Yes, Col A = 4. Since A has four pivot columns, dimCol A = 4. Thus Col A is a four-dimensional
subspace of 4, and Col A = 4.
No, Nul A ≠ 3. It is true that dimNul A = 3 , but Nul A is a subspace of 7.
8. Since A has four pivot columns, rank A = 4, and dimNul A = 8 − rank A = 8 − 4 = 4.
No. Col A ≠ 4. It is true that dimCol A = rank A = 4, but Col A is a subspace of 6.
9. Since dimNul A = 3, rank A = 6 − dimNul A = 6 − 3 = 3. So dimCol A = rank A = 3.
No. Col A ≠ 3. It is true that dimCol A = rank A = 3, but Col A is a subspace of 4.
10. Since dimNul A = 5, rank A = 7 − dimNul A = 7 − 5 = 2. So dimCol A = rank A = 2.
11. Since dimNul A = 3, rank A = 5 − dimNul A = 5 − 3 = 2. So dimRow A = dimCol A = rank A = 2.
12. Since dimNul A = 2, rank A = 4 − dimNul A = 4 − 2 = 2. So dimRow A = dimCol A = rank A = 2.
13. The rank of a matrix A equals the number of pivot positions which the matrix has. If A is either a
7 × 5 matrix or a 5 × 7 matrix, the largest number of pivot positions that A could have is 5. Thus the
largest possible value for rank A is 5.
14. The dimension of the row space of a matrix A is equal to rank A, which equals the number of pivot
positions which the matrix has. If A is either a 5 × 4 matrix or a 4 × 5 matrix, the largest number of
pivot positions that A could have is 4. Thus the largest possible value for dimRow A is 4.
15. Since the rank of A equals the number of pivot positions which the matrix has, and A could have at
most 3 pivot positions, rank A ≤ 3. Thus dimNul A = 7 − rank A ≥ 7 − 3 = 4.
16. Since the rank of A equals the number of pivot positions which the matrix has, and A could have at
most 5 pivot positions, rank A ≤ 5. Thus dimNul A = 5 − rank A ≥ 5 − 5 = 0.
17. a.
b.
c.
d.
True. The rows of A are identified with the columns of AT . See the paragraph before Example 1.
False. See the warning after Example 2.
True. See the Rank Theorem.
False. See the Rank Theorem.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.6
• Solutions
239
e. True. See the Numerical Note before the Practice Problem.
18. a.
b.
c.
d.
False. Review the warning after Theorem 6 in Section 4.3.
False. See the warning after Example 2.
True. See the remark in the proof of the Rank Theorem.
True. This fact was noted in the paragraph before Example 4. It also follows from the fact that the
rows of AT are the columns of ( AT )T = A.
e. True. See Theorem 13.
19. Yes. Consider the system as Ax = 0, where A is a 5 × 6 matrix. The problem states that
dimNulA = 1 . By the Rank Theorem, rank A = 6 − dimNul A = 5. Thus dim Col A = rank A = 5, and
since Col A is a subspace of 5, Col A = 5 So every vector b in 5 is also in Col A, and Ax = b, has
a solution for all b.
20. No. Consider the system as Ax = b, where A is a 6 × 8 matrix. The problem states that
dimNul A = 2. By the Rank Theorem, rank A = 8 − dimNul A = 6. Thus dimCol A = rank A = 6, and
since Col A is a subspace of 6, Col A = 6 So every vector b in 6 is also in Col A, and Ax = b has
a solution for all b. Thus it is impossible to change the entries in b to make Ax = b into an
inconsistent system.
21. No. Consider the system as Ax = b, where A is a 9 × 10 matrix. Since the system has a solution for
all b in 9, A must have a pivot in each row, and so rankA = 9. By the Rank Theorem,
dimNulA = 10 − 9 = 1. Thus it is impossible to find two linearly independent vectors in Nul A.
22. No. Consider the system as Ax = 0, where A is a 10 × 12 matrix. Since A has at most 10 pivot
positions, rankA ≤ 10. By the Rank Theorem, dimNulA = 12 − rankA ≥ 2. Thus it is impossible to
find a single vector in Nul A which spans Nul A.
23. Yes, six equations are sufficient. Consider the system as Ax = 0, where A is a 12 × 8 matrix. The
problem states that dimNul A = 2. By the Rank Theorem, rank A = 8 − dimNul A = 6. Thus
dimCol A = rank A = 6. So the system Ax = 0 is equivalent to the system Bx = 0, where B is an
echelon form of A with 6 nonzero rows. So the six equations in this system are sufficient to describe
the solution set of Ax = 0.
24. Yes, No. Consider the system as Ax = b, where A is a 7 × 6 matrix. Since A has at most 6 pivot
positions, rank A ≤ 6. By the Rank Theorem, dim Nul A = 6 − rank A ≥ 0. If dimNul A = 0, then the
system Ax = b will have no free variables. The solution to Ax = b, if it exists, would thus have to be
unique. Since rank A ≤ 6, Col A will be a proper subspace of 7. Thus there exists a b in 7 for
which the system Ax = b is inconsistent, and the system Ax = b cannot have a unique solution for
all b.
25. No. Consider the system as Ax = b, where A is a 10 × 12 matrix. The problem states that
dim Nul A = 3. By the Rank Theorem, dimCol A = rank A = 12 − dimNul A = 9. Thus Col A will be a
proper subspace of 10 Thus there exists a b in 10 for which the system Ax = b is inconsistent, and
the system Ax = b cannot have a solution for all b.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
240
CHAPTER 4
• Vector Spaces
26. Consider the system Ax = 0, where A is a m × n matrix with m > n. Since the rank of A is the
number of pivot positions that A has and A is assumed to have full rank, rank A = n. By the Rank
Theorem, dimNul A = n − rank A = 0. So Nul A = {0}, and the system Ax = 0 has only the trivial
solution. This happens if and only if the columns of A are linearly independent.
27. Since A is an m × n matrix, Row A is a subspace of n, Col A is a subspace of m, and Nul A is a
subspace of n. Likewise since AT is an n × m matrix, Row AT is a subspace of m, Col AT is a
subspace of
, and Nul AT is a subspace of
n
m
. Since Row A = Col AT and Col A = Row AT , there
are four dinstict subspaces in the list: Row A, Col A, Nul A, and Nul AT .
28. a. Since A is an m × n matrix and dimRow A = rank A,
dimRow A + dimNul A = rank A + dimNul A = n.
b. Since AT is an n × m matrix and dimCol A = dimRow A = dimCol AT = rank AT ,
dimCol A + dimNul AT = rank AT + dimNul AT = m.
29. Let A be an m × n matrix. The system Ax = b will have a solution for all b in m if and only if A has a
pivot position in each row, which happens if and only if dimCol A = m. By Exercise 28 b., dimCol A
= m if and only if dimNul AT = m − m = 0 , or Nul AT = {0}. Finally, Nul AT = {0} if and only if the
equation AT x = 0 has only the trivial solution.
30. The equation Ax = b is consistent if and only if rank [ A
equal if and only if b is not a pivot column of [ A
Section 1.2.
ª 2º
31. Compute that uv = «« −3»» [ a
«¬ 5»¼
T
b
ª 2a
c ] = «« −3a
«¬ 5a
2b
−3b
5b
b ] = rank A because the two ranks will be
b ]. The result then follows from Theorem 2 in
2c º
−3c »» . Each column of uvT is a multiple of u,
5c »¼
so dimCol uvT = 1 , unless a = b = c = 0, in which case uvT is the 3 × 3 zero matrix and
dimCol uvT = 0. In any case, rank uvT = dimCol uvT ≤ 1
32. Note that the second row of the matrix is twice the first row. Thus if v = (1, –3, 4), which is the first
row of the matrix,
ª1 º
uvT = « » [1
¬ 2¼
33. Let A = [u1
u2
−3
ª1
4] = «
¬2
−3
−6
4º
.
8»¼
u3 ] , and assume that rank A = 1. Suppose that u1 ≠ 0 . Then {u1} is basis for Col
A, since Col A is assumed to be one-dimensional. Thus there are scalars x and y with u 2 = xu1 and
ª1º
« »
u 3 = yu1 , and A = u1 vT , where v = « x » .
«¬ y »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.6
• Solutions
241
If u1 = 0 but u 2 ≠ 0 , then similarly {u 2 } is basis for Col A, since Col A is assumed to be one-
ª0º
dimensional. Thus there is a scalar x with u 3 = xu 2 , and A = u 2 v , where v = ««1 »» .
¬« x ¼»
T
ª0º
If u1 = u 2 = 0 but u 3 ≠ 0, then A = u3 v , where v = ««0»» .
«¬1 »¼
T
34. Let A be an m × n matrix with of rank r > 0, and let U be an echelon form of A. Since A can be
reduced to U by row operations, there exist invertible elementary matrices E1 , …, E p with
( E p ⋅⋅ ⋅ E1 ) A = U . Thus A = ( E p ⋅ ⋅⋅ E1 ) −1U , since the product of invertible matrices is invertible. Let
E = ( E p ⋅⋅ ⋅ E1 ) −1 ; then A = EU. Let the columns of E be denoted by c1 , … , c m . Since the rank of A is
r, U has r nonzero rows, which can be denoted d1T ,…, dTr . By the column-row expansion of A
(Theorem 10 in Section 2.4):
A = EU = [c1
ªd1T º
« »
« # »
« dT »
cm ] « r » = c1d1T + … + c r dTr ,
«0»
« »
« # »
«¬ 0 »¼
…
which is the sum of r rank 1 matrices.
35. [M]
a. Begin by reducing A to reduced echelon form:
ª1
«0
«
A ∼ «0
«
«0
«¬ 0
0
13/ 2
0
5
0
1
11/ 2
0
1/ 2
0
0
0
1
−11/ 2
0
0
0
0
0
1
0
0
0
0
0
−3º
2 »»
7» .
»
1»
0 »¼
A basis for Col A is the pivot columns of A, so matrix C contains these columns:
ª 7
« −4
«
C=« 5
«
« −3
«¬ 6
−9
5
6
−2
−7
5
5
−1
−8
4
−3º
−5»»
2» .
»
−4 »
9 »¼
A basis for Row A is the pivot rows of the reduced echelon form of A, so matrix R contains
these rows:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
242
CHAPTER 4
• Vector Spaces
ª1
«0
R=«
«0
«
¬«0
0 13/ 2
1 11/ 2
0
0
0
0
0
0
1
0
5
1/ 2
−11/ 2
0
0
0
0
1
−3º
2»»
.
7»
»
1¼»
To find a basis for Nul A row reduce to reduced echelon form, note that the solution to Ax = 0
in terms of free variables is x1 = −(13/ 2) x3 − 5 x5 + 3 x7 , x2 = −(11/ 2) x3 − (1/ 2) x5 − 2 x7 ,
x4 = (11/ 2) x5 − 7 x7 , x6 = − x7 , with x3 , x5 , and x7 free. Thus matrix N is
ª −13/ 2
« −11/ 2
«
«
1
«
N =«
0
«
0
«
0
«
«
0
¬
−5
−1/ 2
0
11/ 2
1
0
0
3º
−2 »»
0»
»
−7 » .
0»
»
−1»
1»¼
b. The reduced echelon form of AT is
ª1
«0
«
«0
«
T
A ∼ «0
«0
«
«0
«0
¬
0
0
0
1
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
−2 /11º
−41/11»»
0»
»
28 /11» ,
0»
»
0»
0 »¼
so the solution to AT x = 0 in terms of free variables is x1 = (2 /11) x5 , x2 = (41/11) x5 , x3 = 0,
x4 = −(28 /11) x5 , with x5 free. Thus matrix M is
ª 2 /11º
« 41/11»
«
»
M =«
0» .
«
»
« −28 /11»
«¬
1»¼
The matrix S = ª¬ RT
N º¼ is 7 × 7 because the columns of RT and N are in
+ dimNul A = 7. The matrix T = [C
7
and dimRow A
M ] is 5 × 5 because the columns of C and M are in
5
and dimCol A + dimNul AT = 5. Both S and T are invertible because their columns are linearly
independent. This fact will be proven in general in Theorem 3 of Section 6.1.
36. [M] Answers will vary, but in most cases C will be 6 × 4, and will be constructed from the first 4
columns of A. In most cases R will be 4 × 7, N will be 7 × 3, and M will be 6 × 2.
37. [M] The C and R from Exercise 35 work here, and A = CR.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.7
• Solutions
243
38. [M] If A is nonzero, then A = CR. Note that CR = [Cr1 Cr2 … Crn ] , where r1 , …, rn are the
columns of R. The columns of R are either pivot columns of R or are not pivot columns of R.
Consider first the pivot columns of R. The i th pivot column of R is e i , the i th column in the
identity matrix, so Cei is the i th pivot column of A. Since A and R have pivot columns in the same
locations, when C multiplies a pivot column of R, the result is the corresponding pivot column of A
in its proper location.
Now suppose r j is a nonpivot column of R. Then r j contains the weights needed to construct
the j th column of A from the pivot columns of A, as is discussed in Example 9 of Section 4.3 and in
the paragraph preceding that example. Thus r j contains the weights needed to construct the j th
column of A from the columns of C, and Cr j = a j .
4.7
SOLUTIONS
Notes: This section depends heavily on the coordinate systems introduced in Section 4.4. The row
reduction algorithm that produces P can also be deduced from Exercise 15 in Section 2.2, by row
c← B
reducing ª¬ PC PB º¼ . to ª I PC PB º . The change-of-coordinates matrix here is interpreted in Section 5.4
¬
¼
as the matrix of the identity transformation relative to two bases.
−1
ª 6º
ª 9º
ª 6
1. a. Since b1 = 6c1 − 2c 2 and b 2 = 9c1 − 4c 2 , [b1 ]C = « » , [b 2 ]C = « » , and P = «
←
C B
¬ −2 ¼
¬ −4 ¼
¬ −2
9º
.
−4 »¼
ª −3 º
b. Since x = −3b1 + 2b 2 , [ x ]B = « » and
¬ 2¼
ª 6
[ x ]C = P [ x]B = «
C ←B
¬ −2
9 º ª −3 º ª 0 º
=
−4 »¼ «¬ 2 »¼ «¬ −2 »¼
ª −2 º
ª 3º
ª −2
2. a. Since b1 = −2c1 + 4c 2 and b 2 = 3c1 − 6c 2 , [b1 ]C = « » , [b 2 ]C = « » , and P = «
C←B
¬ 4¼
¬ −6 ¼
¬ 4
ª2º
b. Since x = 2b1 + 3b 2 , [ x ]B = « » and
¬3¼
ª −2
[ x ]C = P [ x]B = «
C ←B
¬ 4
3º ª 2 º ª 5 º
=
−6 »¼ «¬ 3 »¼ «¬ −10 »¼
3. Equation (ii) is satisfied by P for all x in V.
4. Equation (i) is satisfied by P for all x in V.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
3º
.
−6 »¼
244
CHAPTER 4
• Vector Spaces
ª 4º
ª −1º
«
»
5. a. Since a1 = 4b1 − b 2 , a 2 = −b1 + b 2 + b 3 , and a3 = b 2 − 2b 3 , [a1 ]B = « −1» , [a 2 ]B = «« 1»» ,
«¬ 0»¼
«¬ 1»¼
0º
ª 0º
ª 4 −1
«
»
«
[a3 ]B = « 1» , and P = « −1
1
1»» .
B← A
1 −2¼»
¬« −2¼»
¬« 0
ª 3º
b. Since x = 3a1 + 4a 2 + a 3 , [x]A = «« 4»» and
«¬1 »¼
ª 4
[x]B = P = «« −1
B← A
«¬ 0
−1
0º ª 3º ª8 º
1»» «« 4»» = «« 2»»
−2»¼ «¬1 »¼ «¬ 2»¼
1
1
ª 2º
ª0º
«
»
« »
6. a. Since f1 = 2d1 − d 2 + d 3 , f 2 = 3d 2 + d3 , and f3 = −3d1 + 2d 3 , [f1 ]D = « −1» , [f2 ]D = « 3» ,
«¬ 1»¼
«¬1 »¼
ª −3º
ª 2 0 −3º
«
»
[f3 ]D = « 0» , and P = «« −1 3
0»» .
D←F
«¬ 2»¼
«¬ 1 1
2»¼
ª1º
b. Since x = f1 − 2f 2 + 2f3 , [ x]F = «« −2»» and
«¬ 2 »¼
ª 2
[x]D = P [x]F = «« −1
D← F
«¬ 1
0
3
1
−3º ª 1º ª −4º
0»» «« −2»» = «« −7 »»
2»¼ «¬ 2»¼ «¬ 3»¼
7. To find P , row reduce the matrix [c1
c2
0
−3
1
−5
1º
.
2 »¼
C←B
[c1
c2
b1
ª −3
Thus P = «
C←B
¬ −5
ª1
b2 ] ∼ «
¬0
1º
ª −2
, and P = P −1 = «
»
B ←C
C←B
2¼
¬ −5
8. To find P , row reduce the matrix [c1
C ←B
[c1
c2
b1
ª 9
Thus P = «
←
C B
¬ −10
ª1
b2 ] ∼ «
¬0
0
9
1
−10
c2
b2 ] :
b1
1º
.
3»¼
b1
b2 ] :
−8 º
.
9 »¼
−8 º
ª9
, and P = P −1 = «
»
←
←
B C
C B
9¼
¬10
8º
.
9 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.7
9. To find P , row reduce the matrix [c1
C ←B
[c1
c2
ª1
b2 ] ∼ «
¬0
b1
ª2
Thus P = «
←
C B
¬0
0
2
1
0
3º
ª1 / 2
, and P = P −1 = «
»
←
←
B C
C B
−1¼
¬ 0
C ←B
c2
ª1
b2 ] ∼ «
¬0
b1
ª 3
Thus P = «
C←B
¬ −2
b1
b2 ] :
3º
.
−1»¼
10. To find P , row reduce the matrix [c1
[c1
c2
• Solutions
0
3
1
−2
c2
3 / 2º
.
−1 »¼
b1
b2 ] :
1º
.
0 »¼
1º
ª0
, and P = P −1 = «
»
B ←C
C←B
0¼
¬1
−1 / 2 º
.
3 / 2 »¼
11. a. False. See Theorem 15.
b. True. See the first paragraph in the subsection “Change of Basis in
n
.”
12. a. True. The columns of P are coordinate vectors of the linearly independent set B. See the
C ←B
second paragraph after Theorem 15.
b. False. The row reduction is discussed after Example 2. The matrix P obtained there satisfies
[ x]C = P[ x]B
13. Let B = {b1 , b 2 , b3 } = {1 − 2t + t 2, 3 − 5t + 4t 2, 2t + 3t 2 } and let C = {c1 , c2 , c3 } = {1, t , t 2 }. The
C-coordinate vectors of b1 , b 2 , and b3 are
ª 1º
ª 3º
ª0º
«
»
«
»
[b1 ]C = « −2» , [b 2 ]C = « −5» , [b3 ]C = «« 2»»
«¬ 1»¼
«¬ 4»¼
«¬ 3»¼
So
ª 1
P = «« −2
←
C B
«¬ 1
3
−5
4
0º
2»»
3»¼
Let x = –1 + 2t. Then the coordinate vector [x]B satisfies
ª −1º
P [x]B = [x]C = «« 2»»
C←B
«¬ 0»¼
This system may be solved by row reducing its augmented matrix:
ª 1
« −2
«
«¬ 1
3
0
−5
4
2
3
−1º ª 1
2»» ∼ «« 0
0»¼ «¬0
0
0
1
0
0
0
5º
ª5º
»
−2» , so [x]B = «« −2»»
«¬ 1 »¼
1»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
245
246
CHAPTER 4
• Vector Spaces
14. Let B = {b1 , b 2 , b 3 } = {1 − 3t 2 , 2 + t − 5t 2 , 1 + 2t} and let C = {c1 , c2 , c3 } = {1, t , t 2 }. The C-coordinate
vectors of b1 , b 2 , and b3 are
ª 1º
ª 2º
ª1º
«
»
«
»
[b1 ]C = « 0» , [b 2 ]C = « 1» , [b3 ]C = «« 2»»
«¬ −3»¼
«¬ −5»¼
«¬ 0 »¼
So
ª 1
P = «« 0
C←B
«¬ −3
2
1
−5
1º
2»»
0»¼
Let x = t 2 . Then the coordinate vector [x]B satisfies
ª0 º
P [x]B = [x]C = ««0 »»
C←B
«¬1 »¼
This system may be solved by row reducing its augmented matrix:
ª 1
« 0
«
«¬ −3
2
1
1
−5
2
0
0º ª 1
0»» ∼ «« 0
1»¼ «¬ 0
0
0
1
0
0
0
3º
ª 3º
»
«
−2» , so [x]B = « −2»»
«¬ 1»¼
1»¼
and t 2 = 3(1 − 3t 2 ) − 2(2 + t − 5t 2 ) + (1 + 2t ).
15. (a)
(b)
(c)
(d)
B is a basis for V
the coordinate mapping is a linear transformation
of the product of a matrix and a vector
the coordinate vector of v relative to B
ª1 º
«0»
16. (a) [b1 ]C = Q[b1 ]B = Q « » = Qe1
«# »
« »
«¬0»¼
(b) [b k ]C
(c) [b k ]C = Q[b k ]B = Qe k
17. [M]
a. Since we found P in Exercise 34 of Section 4.5, we can calculate that
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.7
ª32
« 0
«
« 0
1 «
−1
P = « 0
32
« 0
«
« 0
« 0
¬
0
16
0
12
0
32
0
24
0
20
0
16
0
16
0
0
0
8
0
10
0
0
0
4
0
0
0
0
0
2
0
0
0
0
0
• Solutions
247
10 º
0 »»
15»
»
0» .
6»
»
0»
1»¼
b. Since P is the change-of-coordinates matrix from C to B, P −1 will be the change-ofcoordinates matrix from B to C. By Theorem 15, the columns of P −1 will be the C-coordinate
vectors of the basis vectors in B. Thus
1
cos 2t = (1 + cos 2t )
2
1
cos3t = (3cos t + cos 3t )
4
1
cos 4t = (3 + 4cos 2t + cos 4t )
8
1
cos5t = (10cos t + 5cos 3t + cos 5t )
16
1
cos6t = (10 + 15cos 2t + 6cos 4t + cos 6t )
32
18. [M] The C-coordinate vector of the integrand is (0, 0, 0, 5, –6, 5, –12). Using P −1 from the previous
exercise, the B- coordinate vector of the integrand will be
P −1 (0, 0, 0, 5, − 6, 5, − 12) = ( −6, 55 / 8, − 69 / 8, 45 /16, − 3, 5 /16, − 3/ 8)
Thus the integral may be rewritten as
55
69
45
5
3
cos t − cos 2t + cos 3t − 3cos 4t + cos 5t − cos 6t dt ,
8
8
16
16
8
which equals
³ −6 +
−6t +
55
69
15
3
1
1
sin t − sin 2t + sin 3t − sin 4t + sin 5t − sin 6t + C.
8
16
16
4
16
16
19. [M]
a. If C is the basis {v1 , v 2 , v 3 }, then the columns of P are [u1 ]C , [u 2 ]C , and [u 3 ]C . So
u j = [ v1
v2
[u1
v 3 ][u1 ]C , and [u1
u2
ª −2
u3 ] = «« 2
«¬ 3
u2
−8
5
2
u3 ] = [ v1
−7 º ª 1
2»» «« −3
6»¼ «¬ 4
2
−5
6
v2
v 3 ] P. In the current exercise,
−1º ª −6
0 »» = «« −5
1»¼ «¬ 21
−6
−9
32
−5º
0»» .
3»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
248
CHAPTER 4
• Vector Spaces
b. Analogously to part a., [ v1
[ v1
ª −2
«
w3 ] = « 2
«¬ 3
w2
ª −2
= «« 2
«¬ 3
20. a.
P = P
D← B
w2
v 3 ] P −1 . In the current exercise,
v2
[ w1
v 3 ] = [ w1
v2
−8
5
2
−8
5
2
−7 º ª 1
2 »» «« −3
6 »¼ «¬ 4
−7 º ª 5
2»» «« −3
6»¼ «¬ −2
−1º
0 »»
1»¼
2
−5
6
5º ª 28
−3»» = «« −9
−1»¼ «¬ −3
8
−5
−2
w 3 ] P, so [ w1
w2
w3 ] =
−1
38
−13
2
21º
−7 »» .
3»¼
P
D ←C C ← B
Let x be any vector in the two-dimensional vector space. Since P is the change-of-coordinates
C ←B
matrix from B to C and P is the change-of-coordinates matrix from C to D,
D←C
[ x ]C = P [x ]B and [ x ]D = P [x ]C = P
C←B
D ←C
P [ x ]B
D ←C C ← B
is the change-of-coordinates matrix from B to D,
But since P
D← B
[ x ]D = P [ x ]B
D← B
Thus
P [ x ]B = P
D← B
P [ x ]B
D ←C C ← B
2
, and P = P
for any vector [x]B in
D← B
P
D ←C C ← B
° ª 1º ª −2º
° ª7 º ª −3º °½
b. [M] For example, let B = ® « » , « » ¾ , C = ® « » , « »
°¯ ¬ 5 ¼ ¬ −1¼ °¿
¯° ¬ −5¼ ¬ 2¼
can calculate the change-of-coordinates matrices:
ª −1
« 8
¬
ª −1
« 8
¬
ª 1
« −5
¬
−2
7
2
5
1
1
−5
−5
−2 º ª 1
∼
2 »¼ «¬ 0
1
7
−5
5
−3 º ª 1
∼
−1»¼ «¬ 0
−3 º ª 1
∼
−1»¼ «¬ 0
°½
° ª −1º ª 1º
¾ , and D = ® « » , « »
°¯ ¬ 8¼ ¬ −5¼
¿°
0
−3
1
−5
0
0
1
1
−8 / 3 º
ª0
P =«
»
←
−14 / 3¼ D C ¬ 1
0
40 / 3
1
61/ 3
1º
ª −3
P =«
»
←
C
B
2¼
¬ −5
1º
2 »¼
− 8 / 3º
−14 / 3»¼
−16 / 3º
ª 40 / 3
P =«
»
−25 / 3¼ D ← B ¬ 61/ 3
−16 / 3º
−25 / 3»¼
One confirms easily that
ª 40 / 3
P =«
D ←B
¬ 61/ 3
−16 / 3º ª 0
=
−25 / 3»¼ «¬ 1
− 8 / 3 º ª −3
−14 / 3»¼ «¬ −5
1º
= P P
2 »¼ D ←C C ← B
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
°½
¾. Then we
°¿
4.8
4.8
• Solutions
249
SOLUTIONS
Notes: This is an important section for engineering students and worth extra class time. To spend only
one lecture on this section, you could cover through Example 5, but assign the somewhat lengthy
Example 3 for reading. Finding a spanning set for the solution space of a difference equation uses the
Basis Theorem (Section 4.5) and Theorem 17 in this section, and demonstrates the power of the theory of
Chapter 4 in helping to solve applied problems. This section anticipates Section 5.7 on differential
equations. The reduction of an n th order difference equation to a linear system of first order difference
equations was introduced in Section 1.10, and is revisited in Sections 4.9 and 5.6. Example 3 is the
background for Exercise 26 in Section 6.5.
1. Let yk = 2k . Then
yk + 2 + 2 yk +1 − 8 yk = 2k + 2 + 2(2k +1 ) − 8(2k )
= 2k (22 + 22 − 8)
= 2k (0) = 0 for all k
Since the difference equation holds for all k, 2 k is a solution.
Let yk = (−4) k . Then
yk + 2 + 2 yk +1 − 8 yk = (−4)k + 2 + 2(−4)k +1 − 8(−4)k
= (−4) k ((−4) 2 + 2(−4) − 8)
= (−4)k (0) = 0 for all k
Since the difference equation holds for all k, (−4)k is a solution.
2. Let yk = 5k . Then
yk + 2 − 25 yk = 5k + 2 − 25(5k )
= 5k (52 − 25)
= 5k (0) = 0 for all k
Since the difference equation holds for all k, 5k is a solution.
Let yk = ( −5) k . Then
yk + 2 − 25 yk = ( −5) k + 2 − 25( −5) k
= (−5)k ((−5)2 − 25)
= (−5) k (0) = 0 for all k
Since the difference equation holds for all k, (−5) k is a solution.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
250
CHAPTER 4
• Vector Spaces
3. The signals 2 k and (−4)k are linearly independent because neither is a multiple of the other; that is,
there is no scalar c such that 2k = c(−4) k for all k. By Theorem 17, the solution set H of the
difference equation yk + 2 + 2 yk +1 − 8 yk = 0 is two-dimensional. By the Basis Theorem, the two
linearly independent signals 2 k and (−4)k form a basis for H.
4. The signals 5k and (−5) k are linearly independent because neither is a multiple of the other; that is,
there is no scalar c such that 5k = c( −5)k for all k. By Theorem 17, the solution set H of the
difference equation yk + 2 − 25 yk = 0 is two-dimensional. By the Basis Theorem, the two linearly
independent signals 5k and (−5) k form a basis for H.
5. Let yk = ( −2)k . Then
yk + 2 + 4 yk +1 + 4 yk = (−2) k + 2 + 4( −2)k +1 + 4( −2)k
= (−2)k ((−2)2 + 4(−2) + 4)
= (−2)k (0) = 0 for all k
Since the difference equation holds for all k, (−2)k is in the solution set H.
Let yk = k (−2)k . Then
yk + 2 + 4 yk +1 + 4 yk = ( k + 2)(−2) k + 2 + 4(k + 1)(−2)k +1 + 4k (−2) k
= (−2) k ((k + 2)(−2)2 + 4(k + 1)(−2) + 4k )
= (−2)k (4k + 8 − 8k − 8 + 4k )
= (−2)k (0) = 0 for all k
Since the difference equation holds for all k, k (−2) k is in the solution set H.
The signals (−2)k and k (−2) k are linearly independent because neither is a multiple of the other;
that is, there is no scalar c such that (−2) k = ck (−2)k for all k and there is no scalar c such that
c(−2) k = k (−2) k for all k . By Theorem 17, dim H = 2, so the two linearly independent signals (−2)k
and k (−2) k form a basis for H by the Basis Theorem.
6. Let yk = 4 k cos k2π . Then
yk + 2 + 16 yk = 4 k + 2 cos
( k + 2)π
kπ ·
§
+ 16 ¨ 4 k cos
¸
2
2 ¹
©
( k + 2)π
kπ ·
§
= 4 k ¨ 42 cos
+ 16 cos
¸
2
2 ¹
©
§
kπ ·
§ kπ
·
= 16 ⋅ 4k ¨ cos ¨
+ π ¸ + cos
¸
2 ¹
© 2
¹
©
= 16 ⋅ 4k (0) = 0 for all k
since cos(t + π) = –cos t for all t. Since the difference equation holds for all k, 4 k cos k2π is in the
solution set H.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.8
• Solutions
251
Let yk = 4 k sin k2π . Then
yk + 2 + 16 yk = 4k + 2 sin
( k + 2)π
kπ ·
§
+ 16 ¨ 4 k sin
¸
2
2 ¹
©
( k + 2)π
kπ ·
§
= 4 k ¨ 4 2 sin
+ 16 sin
¸
2
2 ¹
©
§ § kπ
kπ ·
·
= 16 ⋅ 4 k ¨ sin ¨
+ π ¸ + sin
¸
2 ¹
¹
© © 2
= 16 ⋅ 4k (0) = 0 for all k
since sin(t + π) = –sin t for all t. Since the difference equation holds for all k, 4 k sin k2π is in the
solution set H.
The signals 4 k cos k2π and 4 k sin k2π are linearly independent because neither is a multiple of the
other. By Theorem 17, dim H = 2, so the two linearly independent signals 4 k cos k2π and 4 k sin k2π
form a basis for H by the Basis Theorem.
7. Compute and row reduce the Casorati matrix for the signals 1k , 2k , and (−2)k , setting k = 0 for
convenience:
ª10
« 1
«1
«2
¬«1
(−2)0 º ª1
»
(−2)1 » ∼ ««0
»
(−2)2 ¼» «¬0
20
1
2
22
0
1
0
0º
0 »»
1 »¼
This Casorati matrix is row equivalent to the identity matrix, thus is invertible by the IMT. Hence the
set of signals {1k , 2k ,(−2)k } is linearly independent in . The exercise states that these signals are in
the solution set H of a third-order difference equation. By Theorem 17, dim H = 3, so the three
linearly independent signals 1k , 2k , (−2)k form a basis for H by the Basis Theorem.
8. Compute and row reduce the Casorati matrix for the signals (−1)k , 2k , and 3k , setting k = 0 for
convenience:
ª( −1)0
«
1
« (−1)
«
2
¬«( −1)
20
1
2
22
30 º ª1
»
31 » ∼ ««0
»
32 ¼» ¬«0
0
1
0
0º
0 »»
1 »¼
This Casorati matrix is row equivalent to the identity matrix, thus is invertible by the IMT. Hence the
set of signals {(−1) k , 2k ,3k } is linearly independent in . The exercise states that these signals are in
the solution set H of a third-order difference equation. By Theorem 17, dim H = 3, so the three
linearly independent signals (−1)k , 2k , and 3k form a basis for H by the Basis Theorem.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
252
CHAPTER 4
• Vector Spaces
9. Compute and row reduce the Casorati matrix for the signals 2k , 5k cos k2 , and 5k sin k2 setting k = 0
for convenience:
ª 20
« 1
«2
« 2
¬« 2
50 sin 0 º ª1
»
51 sin π2 » ∼ ««0
»
52 sin π ¼» «¬0
50 cos 0
1
π
5 cos 2
52 cos π
0
1
0
0º
0 »»
1 »¼
This Casorati matrix is row equivalent to the identity matrix, thus is invertible by the IMT. Hence the
set of signals {2 k ,5k cos k2π ,5k sin k2π } is linearly independent in . The exercise states that these
signals are in the solution set H of a third-order difference equation. By Theorem 17, dim H = 3, so
the three linearly independent signals 2k , 5k cos k2 , and 5k sin k2 form a basis for H by the Basis
Theorem.
10. Compute and row reduce the Casorati matrix for the signals (−2)k , k ( −2) k , and 3k setting k = 0 for
convenience:
ª(−2)0
«
1
« (−2)
«
2
«¬(−2)
0(−2)0
1
1(−2)
2(−2) 2
30 º ª1
»
31 » ∼ ««0
»
32 »¼ «¬0
0
1
0
0º
0 »»
1 »¼
This Casorati matrix is row equivalent to the identity matrix, thus is invertible by the IMT. Hence the
set of signals {(−2) k , k (−2) k , 3k } is linearly independent in . The exercise states that these signals
are in the solution set H of a third-order difference equation. By Theorem 17, dim H = 3, so the three
linearly independent signals (−2)k , k ( −2) k , and 3k form a basis for H by the Basis Theorem.
11. The solution set H of this third-order difference equation has dim H = 3 by Theorem 17. The two
signals (−1) k and 2 k cannot possibly span a three-dimensional space, and so cannot be a basis for
H.
12. The solution set H of this fourth-order difference equation has dim H = 4 by Theorem 17. The two
signals 3k and (−2)k cannot possibly span a four-dimensional space, and so cannot be a basis for H.
13. The auxiliary equation for this difference equation is r 2 − r + 2 / 9 = 0. By the quadratic formula
(or factoring), r = 2/3 or r = 1/3, so two solutions of the difference equation are (2 / 3)k and (1/ 3) k .
The signals (2 / 3)k and (1/ 3) k are linearly independent because neither is a multiple of the other.
By Theorem 17, the solution space is two-dimensional, so the two linearly independent signals
(2 / 3)k and (1/ 3) k form a basis for the solution space by the Basis Theorem.
14. The auxiliary equation for this difference equation is r 2 − 5r + 6 = 0. By the quadratic formula (or
factoring), r = 2 or r = 3, so two solutions of the difference equation are 2 k and 3k . The signals 2 k
and 3k are linearly independent because neither is a multiple of the other. By Theorem 17, the
solution space is two-dimensional, so the two linearly independent signals 2 k and 3k form a basis for
the solution space by the Basis Theorem.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.8
• Solutions
253
15. The auxiliary equation for this difference equation is 6r 2 + r − 2 = 0. By the quadratic formula (or
factoring), r = 1/2 or r = –2/3, so two solutions of the difference equation are (1 / 2)k and (−2 / 3)k .
The signals (1 / 2)k and ( −2 / 3) k are linearly independent because neither is a multiple of the other.
By Theorem 17, the solution space is two-dimensional, so the two linearly independent signals
(1 / 2)k and ( −2 / 3) k form a basis for the solution space by the Basis Theorem.
16. The auxiliary equation for this difference equation is r 2 − 25 = 0. By the quadratic formula (or
factoring), r = 5 or r = –5, so two solutions of the difference equation are 5k and (−5) k . The signals
5k and (−5) k are linearly independent because neither is a multiple of the other. By Theorem 17, the
solution space is two-dimensional, so the two linearly independent signals 5k and (−5) k form a basis
for the solution space by the Basis Theorem.
17. Letting a = .9 and b = 4/9 gives the difference equation Yk + 2 − 1.3Yk +1 + .4Yk = 1. First we find a
particular solution Yk = T of this equation, where T is a constant. The solution of the equation T –
1.3T + .4T = 1 is T = 10, so 10 is a particular solution to Yk + 2 − 1.3Yk +1 + .4Yk = 1 . Next we solve the
homogeneous difference equation Yk + 2 − 1.3Yk +1 + .4Yk = 0. The auxiliary equation for this difference
equation is r 2 − 1.3r + .4 = 0. By the quadratic formula (or factoring), r = .8 or r = .5, so two
solutions of the homogeneous difference equation are .8k and .5k . The signals (.8) k and (.5) k are
linearly independent because neither is a multiple of the other. By Theorem 17, the solution space is
two-dimensional, so the two linearly independent signals (.8) k and (.5) k form a basis for the
solution space of the homogeneous difference equation by the Basis Theorem. Translating the
solution space of the homogeneous difference equation by the particular solution 10 of the
nonhomogeneous difference equation gives us the general solution of Yk + 2 − 1.3Yk +1 + .4Yk = 1 :
Yk = c1 (.8) k + c2 (.5)k + 10. As k increases the first two terms in the solution approach 0, so Yk
approaches 10.
18. Letting a = .9 and b = .5 gives the difference equation Yk + 2 − 1.35Yk +1 + .45Yk = 1. First we find a
particular solution Yk = T of this equation, where T is a constant. The solution of the equation
T – 1.35T + .45T = 1 is T = 10, so 10 is a particular solution to Yk + 2 − 1.35Yk +1 + .45Yk = 1 . Next we
solve the homogeneous difference equation Yk + 2 − 1.35Yk +1 + .45Yk = 0. The auxiliary equation for
this difference equation is r 2 − 1.35r + .45 = 0. By the quadratic formula (or factoring), r = .6 or
r = .75, so two solutions of the homogeneous difference equation are .6k and .75 k . The signals (.6) k
and (.75)k are linearly independent because neither is a multiple of the other. By Theorem 17, the
solution space is two-dimensional, so the two linearly independent signals (.6) k and (.75)k form a
basis for the solution space of the homogeneous difference equation by the Basis Theorem.
Translating the solution space of the homogeneous difference equation by the particular solution 10
of the nonhomogeneous difference equation gives us the general solution of
Yk + 2 − 1.35Yk +1 + .45Yk = 1 : Yk = c1 (.6)k + c2 (.75)k + 10.
19. The auxiliary equation for this difference equation is r 2 + 4 r + 1 = 0. By the quadratic formula,
r = −2 + 3 or r = −2 − 3, so two solutions of the difference equation are (−2 + 3)k and
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
254
CHAPTER 4
• Vector Spaces
(−2 − 3) k . The signals (−2 + 3)k and (−2 − 3) k are linearly independent because neither is a
multiple of the other. By Theorem 17, the solution space is two-dimensional, so the two linearly
independent signals (−2 + 3)k and (−2 − 3) k form a basis for the solution space by the Basis
Theorem. Thus a general solution to this difference equation is yk = c1 (−2 + 3)k + c2 (−2 − 3)k .
20. Let a = −2 + 3 and b = −2 − 3 . Using the solution from the previous exercise, we find that
y1 = c1a + c2 b = 5000 and y N = c1a N + c2 b N = 0. This is a system of linear equations with variables
c1 and c2 whose augmented matrix may be row reduced:
ªa
« N
¬a
b
bN
ª
1
5000 º «
«
»∼
0 ¼ «
«0
¬
0
1
5000b N º
»
bN a − aN b »
−5000a N »
»
bN a − aN b ¼
so
5000b N
−5000 a N
,
c
=
2
bN a − aN b
bN a − aN b
(Alternatively, Cramer’s Rule may be applied to get the same solution). Thus
c1 =
yk = c1a k + c2b k
=
5000( a k b N − a N b k )
bN a − aN b
21. The smoothed signal zk has the following values: z1 = (9 + 5 + 7) / 3 = 7, z2 = (5 + 7 + 3) / 3 = 5,
z3 = (7 + 3 + 2) / 3 = 4, z4 = (3 + 2 + 4) / 3 = 3, z5 = (2 + 4 + 6) / 3 = 4, z6 = (4 + 6 + 5) / 3 = 5,
z7 = (6 + 5 + 7) / 3 = 6, z8 = (5 + 7 + 6) / 3 = 6, z9 = (7 + 6 + 8) / 3 = 7, z10 = (6 + 8 + 10) / 3 = 8,
z11 = (8 + 10 + 9) / 3 = 9, z12 = (10 + 9 + 5) / 3 = 8, z13 = (9 + 5 + 7) / 3 = 7.
original data
10
smoothed data
8
6
4
2
2
4
6
8
10
12
14
22. a. The smoothed signal zk has the following values:
z0 = .35 y2 + .5 y1 + .35 y0 = .35(0) + .5(.7) + .35(3) = 1.4,
z1 = .35 y3 + .5 y2 + .35 y1 = .35( −.7) + .5(0) + .35(.7) = 0,
z2 = .35 y4 + .5 y3 + .35 y2 = .35( −.3) + .5( −.7) + .35(0) = −1.4,
z3 = .35 y5 + .5 y4 + .35 y3 = .35( −.7) + .5( −.3) + .35( −.7) = −2,
z4 = .35 y6 + .5 y5 + .35 y4 = .35(0) + .5( −.7) + .35( −.3) = −1.4,
z5 = .35 y7 + .5 y6 + .35 y5 = .35(.7) + .5(0) + .35( −.7) = 0,
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.8
• Solutions
255
z6 = .35 y8 + .5 y7 + .35 y6 = .35(3) + .5(.7) + .35(0) = 1.4,
z7 = .35 y9 + .5 y8 + .35 y7 = .35(.7) + .5(3) + .35(.7) = 2,
z8 = .35 y10 + .5 y9 + .35 y8 = .35(0) + .5(.7) + .35(3) = 1.4,…
b. This signal is two times the signal output by the filter when the input (in Example 3) was
y = cos(πt/4). This is expected because the filter is linear. The output from the input
2cos(πt/4) + cos(3πt/4) should be two times the output from cos(πt/4) plus the output from
cos(3πt/4) (which is zero).
23. a. yk +1 − 1.01 yk = −450, y0 = 10,000.
b. [M] MATLAB code to create the table:
pay = 450, y = 10000, m = 0, table = [0;y]
while y>450
y = 1.01*y-pay
m = m+1
table = [table [m;y]]
end
m,y
Mathematica code to create the table:
pay = 450; y = 10000; m = 0; balancetable = {{0, y}};
While[y > 450, {y = 1.01*y - pay; m = m + 1,
AppendTo[balancetable, {m, y}]}];
m
y
c. [M] At the start of month 26, the balance due is $114.88. At the end of this month the unpaid
balance will be (1.01)($114.88)=$116.03. The final payment will thus be $116.03. The total paid
by the borrower is (25)($450.00)+$116.03=$11,366.03.
24. a. yk +1 − 1.005 yk = 200, y0 = 1, 000.
b. [M] MATLAB code to create the table:
pay = 200, y = 1000, m = 0, table = [0;y]
for m = 1: 60
y = 1.005*y+pay
table = [table [m;y]]
end
interest = y-60*pay-1000
Mathematica code to create the table:
pay = 200; y = 1000; amounttable = {{0, y}};
Do[{y = 1.005*y + pay;
AppendTo[amounttable, {m, y}]},{m,1,60}];
interest = y-60*pay-1000
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
256
CHAPTER 4
• Vector Spaces
c. [M] The total is $6213.55 at k = 24, $12,090.06 at k = 48, and $15,302.86 at k = 60. When k = 60,
the interest earned is $2302.86.
25. To show that yk = k 2 is a solution of yk + 2 + 3 yk +1 − 4 yk = 10 k + 7, substitute yk = k 2 ,
yk +1 = (k + 1) 2 , and yk + 2 = (k + 2) 2 :
yk + 2 + 3k +1 − 4 yk = (k + 2)2 + 3(k + 1) 2 − 4k 2
= (k 2 + 4k + 4) + 3(k 2 + 2k + 1) − 4k 2
= k 2 + 4 k + 4 + 3k 2 + 6 k + 3 − 4k 2
= 10k + 7 for all k
The auxiliary equation for the homogeneous difference equation yk + 2 + 3 yk +1 − 4 yk = 0 is
r 2 + 3r − 4 = 0. By the quadratic formula (or factoring), r = –4 or r = 1, so two solutions of the
difference equation are (−4)k and 1k . The signals (−4)k and 1k are linearly independent because
neither is a multiple of the other. By Theorem 17, the solution space is two-dimensional, so the two
linearly independent signals (−4)k and 1k form a basis for the solution space of the homogeneous
difference equation by the Basis Theorem. The general solution to the homogeneous difference
equation is thus c1 (−4)k + c2 ⋅ 1k = c1 (−4) k + c2 . Adding the particular solution k 2 of the
nonhomogeneous difference equation, we find that the general solution of the difference equation
yk + 2 + 3 yk +1 − 4 yk = 10 k + 7 is yk = k 2 + c1 (−4) k + c2 .
26. To show that yk = 1 + k is a solution of yk + 2 − 6 yk +1 + 5 yk = −4, substitute yk = 1 + k ,
yk +1 = 1 + ( k + 1) = 2 + k , and yk + 2 = 1 + ( k + 2) = 3 + k :
yk + 2 − 6 yk +1 + 5 yk = (3 + k ) − 6(2 + k ) + 5(1 + k )
= 3 + k − 12 − 6k + 5 + 5k
= −4 for all k
The auxiliary equation for the homogeneous difference equation yk + 2 − 6 yk +1 + 5 yk = 0 is
r 2 − 6r + 5 = 0. By the quadratic formula (or factoring), r = 1 or r = 5, so two solutions of the
difference equation are 1k and 5k . The signals 1k and 5k are linearly independent because neither
is a multiple of the other. By Theorem 17, the solution space is two-dimensional, so the two linearly
independent signals 1k and 5k form a basis for the solution space of the homogeneous difference
equation by the Basis Theorem. The general solution to the homogeneous difference equation is thus
c1 ⋅1k + c2 ⋅ 5k . Adding the particular solution 1 + k of the nonhomogeneous difference equation, we
find that the general solution of the difference equation yk + 2 − 6 yk +1 + 5 yk = −4 is
yk = 1 + k + c1 ⋅ 1k + c2 ⋅ 5k .
27. To show that yk = k − 2 is a solution of yk + 2 − 4 yk = 8 − 3k , substitute yk = k − 2 and
yk + 2 = ( k + 2) − 2 = k :
yk + 2 − 4 yk = k − 4( k − 2) = k − 4k + 8 = 8 − 3k for all k
The auxiliary equation for the homogeneous difference equation yk + 2 − 4 yk = 0 is r 2 − 4 = 0. By the
quadratic formula (or factoring), r = 2 or r = − 2, so two solutions of the difference equation are 2 k
and (−2)k . The signals 2 k and (−2)k are linearly independent because neither is a multiple of the
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.8
• Solutions
257
other. By Theorem 17, the solution space is two-dimensional, so the two linearly independent signals
2 k and (−2)k form a basis for the solution space of the homogeneous difference equation by the
Basis Theorem. The general solution to the homogeneous difference equation is thus
c1 ⋅ 2 k + c2 ⋅ (−2) k . Adding the particular solution k − 2 of the nonhomogeneous difference equation,
we find that the general solution of the difference equation yk + 2 − 4 yk = 8 − 3k is
yk = k − 2 + c1 ⋅ 2k + c2 ⋅ (−2)k .
28. To show that yk = 1 + 2 k is a solution of yk + 2 + 25 yk = 30 + 52k , substitute yk = 1 + 2 k and
yk + 2 = 1 + 2( k + 2) = 5 + 2k :
yk + 2 + 25 yk = 5 + 2k + 25(1 + 2k ) = 5 + 2 k + 25 + 50k = 30 + 52k for all k
The auxiliary equation for the homogeneous difference equation yk + 2 + 25 yk = 0 is r 2 + 25 = 0. By
the quadratic formula (or factoring), r = 5i or r = –5i, so two solutions of the difference equation are
5k cos k2 and 5k sin k2 . The signals 5k cos k2 and 5k sin k2 are linearly independent because neither is
a multiple of the other. By Theorem 17, the solution space is two-dimensional, so the two linearly
independent signals 5k cos k2 and 5k sin k2 form a basis for the solution space of the homogeneous
difference equation by the Basis Theorem. The general solution to the homogeneous difference
equation is thus c1 ⋅ 5k cos k2 + c2 ⋅ 5k sin k2 . Adding the particular solution 1 + 2k of the
nonhomogeneous difference equation, we find that the general solution of the difference equation
yk + 2 + 25 yk = 30 + 52 k is yk = 1 + 2 k + c1 ⋅ 5k cos k2 + c2 ⋅ 5k sin k2 .
ª yk º
ª yk +1 º ª 0
«y »
« y » «0
k +2 »
k +1 »
=«
29. Let x k = «
. Then x k +1 = «
« yk + 2 »
« yk + 3 » « 0
«
»
«
» «
¬ yk + 3 ¼
¬ yk + 4 ¼ ¬ 2
ª yk +1 º ª 0
ª yk º
«
»
30. Let xk = « yk +1 » . Then x k +1 = «« yk + 2 »» = «« 0
¬« yk + 2 ¼»
¬« yk +3 ¼» ¬« −8
1
0
0
1
0
−6
0
8
1
0
0
0 º ª yk º
0 »» «« yk +1 »»
= Ax k .
1» « yk + 2 »
»
»«
−3¼ ¬ yk +3 ¼
0º ª yk º
1»» «« yk +1 »» = Axk .
5¼» ¬« yk + 2 ¼»
31. The difference equation is of order 2. Since the equation yk + 3 + 5 yk + 2 + 6 yk +1 = 0 holds for all k,
it holds if k is replaced by k − 1. Performing this replacement transforms the equation into
yk + 2 + 5 yk +1 + 6 yk = 0, which is also true for all k. The transformed equation has order 2.
32. The order of the difference equation depends on the values of a1 , a2 , and a3 . If a3 ≠ 0, then the
order is 3. If a3 = 0 and a2 ≠ 0, then the order is 2. If a3 = a2 = 0 and a1 ≠ 0, then the order is 1.
If a3 = a2 = a1 = 0, then the order is 0, and the equation has only the zero signal for a solution.
33. The Casorati matrix C(k) is
ª y
C (k ) = « k
¬ yk +1
zk º ª k 2
=«
zk +1 »¼ «¬(k + 1) 2
º
»
2(k + 1) | k + 1|»¼
2k | k |
In particular,
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
258
CHAPTER 4
• Vector Spaces
ª0 0 º
ª 1 −2 º
ª 4 −8 º
C (0) = «
, C (−1) = «
, and C (−2) = «
»
»
»
0¼
¬1 2 ¼
¬0
¬ 1 −2 ¼
none of which are invertible. In fact, C(k) is not invertible for all k, since
det C ( k ) = 2k 2 ( k + 1) | k + 1| −2( k + 1) 2 k | k | = 2k ( k + 1) ( k | k + 1| − ( k + 1) | k |)
If k = 0 or k = –1, det C(k) = 0. If k > 0, then k + 1 > 0 and k| k + 1 | – (k + 1)| k | = k(k + 1) – (k + 1)k
= 0, so det C(k) = 0. If k < –1, then k + 1 < 0 and k| k + 1 | – (k + 1)| k | = –k(k + 1) + (k + 1)k = 0, so
det C(k) = 0. Thus detC(k)=0 for all k, and C(k) is not invertible for all k. Since C(k) is not invertible
for all k, it provides no information about whether the signals { yk } and { zk } are linearly dependent
or linearly independent. In fact, neither signal is a multiple of the other, so the signals { yk } and { zk }
are linearly independent.
34. No, the signals could be linearly dependent, since the vector space V of functions considered on the
entire real line is not the vector space of signals. For example, consider the functions f (t) = sinπt,
g(t) = sin 2πt, and h(t) = sin 3πt. The functions f, g, and h are linearly independent in V since they
have different periods and thus no function could be a linear combination of the other two. However,
sampling the functions at any integer n gives f (n) = g(n) = h(n) = 0, so the signals are linearly
dependent in .
35. Let { yk } and { zk } be in , and let r be any scalar. The k th term of { yk } + { zk } is yk + zk , while the
k th term of r{ yk } is ryk . Thus
T ({ yk } + {zk }) = T { yk + zk }
= ( yk + 2 + zk + 2 ) + a ( yk +1 + zk +1 ) + b( yk + zk )
= ( yk + 2 + ayk +1 + byk ) + ( zk + 2 + azk +1 + bzk )
= T { yk } + T {zk },and
T (r{ yk }) = T {r yk }
= r yk + 2 + a ( r yk +1 ) + b( r yk )
= r ( yk + 2 + ayk +1 + byk )
= rT { yk }
so T has the two properties that define a linear transformation.
36. Let z be in V, and suppose that x p in V satisfies T (x p ) = z. Let u be in the kernel of T; then T(u) =
0. Since T is a linear transformation, T (u + x p ) = T (u) + T (x p ) = 0 + z = z, so the vector x = u + x p
satisfies the nonhomogeneous equation T(x) = z.
37. We compute that (TD)( y0 , y1 , y2 ,…) = T ( D ( y0 , y1 , y2 ,…)) = T (0, y0 , y1 , y2 ,…) = ( y0 , y1 , y2 ,…)
while ( DT )( y0 , y1 , y2 ,…) = D (T ( y0 , y1 , y2 ,…)) = D ( y1 , y2 , y3 ,…) = (0, y1 , y2 , y3 ,…)
Thus TD = I (the identity transformation on
0),
while DT ≠ I.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.9
4.9
• Solutions
259
SOLUTIONS
Notes: This section builds on the population movement example in Section 1.10. The migration matrix is
examined again in Section 5.2, where an eigenvector decomposition shows explicitly why the sequence of
state vectors x k tends to a steady state vector. The discussion in Section 5.2 does not depend on prior
knowledge of this section.
1. a. Let N stand for “News” and M stand for “Music.” Then the listeners’ behavior is given by the
table
From:
N
M
To:
.7
.6
N
.3
.4
M
ª.7
so the stochastic matrix is P = «
¬.3
.6 º
.
.4 »¼
ª1 º
b. Since 100% of the listeners are listening to news at 8: 15, the initial state vector is x0 = « » .
¬0 ¼
c. There are two breaks between 8: 15 and 9: 25, so we calculate x2 :
ª.7
x1 = P x0 = «
¬.3
.6 º ª1 º ª.7 º
=
.4 »¼ «¬ 0»¼ «¬.3 »¼
ª.7
x 2 = P x1 = «
¬.3
.6 º ª.7 º ª.67 º
=
.4 »¼ «¬.3 »¼ «¬.33 »¼
Thus 33% of the listeners are listening to music at 9:25.
2. a. Let the foods be labelled “1,” “2,” and “3.” Then the animals’ behavior is given by the table
From:
1
2
3
To:
.6
.2
.2
1
.2
.6
.2
2
.2
.2
.6
3
ª.6
so the stochastic matrix is P = ««.2
«¬.2
.2
.6
.2
.2 º
.2 »» .
.6 »¼
ª1 º
« »
b. There are two trials after the initial trial, so we calculate x2 . The initial state vector is « 0 » .
«¬ 0 »¼
ª.6
x1 = P x0 = ««.2
«¬.2
.2
.6
.2
.2 º ª1 º ª.6º
.2 »» ««0»» = ««.2»»
.6 »¼ «¬0»¼ «¬.2»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
260
CHAPTER 4
• Vector Spaces
ª.6
x 2 = P x1 = ««.2
«¬.2
.2
.6
.2
.2 º ª.6 º ª.44 º
.2 »» ««.2 »» = ««.28»»
.6 »¼ «¬.2 »¼ «¬.28»¼
Thus the probability that the animal will choose food #2 is .28.
3. a. Let H stand for “Healthy” and I stand for “Ill.” Then the students’ conditions are given by the
table
From:
H
I
To:
.95 .45
H
.05 .55
I
ª.95
so the stochastic matrix is P = «
¬.05
.45 º
.
.55 »¼
ª.8 º
b. Since 20% of the students are ill on Monday, the initial state vector is x0 = « » . For Tuesday’s
¬.2 ¼
percentages, we calculate x1 ; for Wednesday’s percentages, we calculate x2 :
ª.95
x1 = P x0 = «
¬.05
.45 º ª.8 º ª.85º
=
.55 »¼ «¬.2 »¼ «¬.15»¼
ª.95
x 2 = P x1 = «
¬.05
.45º ª.85º ª.875º
=
.55»¼ «¬.15»¼ «¬.125»¼
Thus 15% of the students are ill on Tuesday, and 12.5% are ill on Wednesday.
ª1 º
c. Since the student is well today, the initial state vector is x0 = « » . We calculate x2 :
¬0 ¼
ª.95
x1 = P x0 = «
¬.05
.45º ª1 º ª.95º
=
.55»¼ «¬ 0 »¼ «¬.05»¼
ª.95 .45º ª.95º ª.925º
x 2 = P x1 = «
»« » = «
»
¬.05 .55¼ ¬.05¼ ¬.075¼
Thus the probability that the student is well two days from now is .925.
4. a. Let G stand for good weather, I for indifferent weather, and B for bad weather. Then the change
in the weather is given by the table
From:
G
I
B
To:
.4
.5
.3
G
.3
.2
.4
I
.3
.3
.3
B
ª.4
so the stochastic matrix is P = ««.3
¬«.3
.5
.2
.3
.3º
.4 »» .
.3¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.9
• Solutions
261
ª.5º
« »
b. The initial state vector is «.5» . We calculate x1 :
«¬ 0 »¼
ª.4
x1 = P x0 = ««.3
«¬.3
.5
.2
.3
.3º ª.5º ª.45º
.4 »» ««.5»» = ««.25»»
.3»¼ «¬ 0 »¼ «¬.30»¼
Thus the chance of bad weather tomorrow is 30%.
ª0º
« »
c. The initial state vector is x0 = «.6 » . We calculate x2 :
«¬.4 »¼
ª.4
x1 = Px0 = ««.3
«¬.3
ª.4
x 2 = Px1 = ««.3
«¬.3
.5
.2
.3
.5
.2
.3
.3º ª 0 º ª.42 º
.4 »» ««.6»» = ««.28»»
.3»¼ «¬.4»¼ «¬.30 »¼
.3º ª.42º ª.398º
.4 »» ««.28»» = ««.302»»
.3»¼ «¬.30»¼ «¬.300»¼
Thus the chance of good weather on Wednesday is 39.8%, or approximately 40%.
ª −.9
5. We solve Px = x by rewriting the equation as (P – I )x = 0, where P − I = «
¬ .9
the augmented matrix for the homogeneous system (P – I )x = 0 gives
ª −.9
« .9
¬
.5
−.5
0º ª 1
∼
0 »¼ «¬ 0
−5 / 9
0
0º
0 »¼
ªx º
ª5 º
ª5 / 9 º
Thus x = « 1 » = x2 «
, and one solution is « » . Since the entries in
»
¬9 ¼
¬ 1¼
¬ x2 ¼
ª5 / 14 º
1/14 to obtain the steady-state vector q = «
».
¬9 / 14 ¼
ª5º
«9 » sum to 14, multiply by
¬ ¼
ª −.6
6. We solve Px = x by rewriting the equation as (P – I )x = 0, where P − I = «
¬ .6
the augmented matrix for the homogeneous system (P – I )x = 0 gives
ª −.6
« .6
¬
.8
−.8
0º ª 1
∼
0 »¼ «¬ 0
−4 / 3
0
.5º
. Row reducing
−.5»¼
.8º
. Row reducing
−.8»¼
0º
0 »¼
ªx º
ª4º
ª 4 / 3º
Thus x = « 1 » = x2 «
, and one solution is « » . Since the entries in
»
¬3¼
¬ 1¼
¬ x2 ¼
ª 4 / 7 º ª.571º
1/7 to obtain the steady-state vector q = «
»≈«
».
¬ 3 / 7 ¼ ¬.429 ¼
ª4º
« 3 » sum to 7, multiply by
¬ ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
262
CHAPTER 4
• Vector Spaces
.1
ª −.3
«
7. We solve Px = x by rewriting the equation as (P – I )x = 0, where P − I = « .2 −.2
«¬ .1
.1
reducing the augmented matrix for the homogeneous system (P – I )x = 0 gives
ª −.3
«
« .2
¬« .1
.1
.1
−.2
.2
.1
−.3
0º ª 1
0 »» ∼ ««0
0 ¼» ¬«0
0
−1
1
−2
0
0
0º
0»»
0¼»
ª x1 º
ª1 º
ª1 º
«
»
«
»
Thus x = « x2 » = x3 « 2» , and one solution is «« 2 »» . Since the entries in
«¬ x3 »¼
«¬1 »¼
«¬ 1 »¼
ª1/ 4º ª.25º
« » « »
to obtain the steady-state vector q = «1/ 2» = « .5» .
«¬1/ 4»¼ «¬.25»¼
ª1 º
« 2 » sum to 4, multiply by 1/4
« »
«¬1 »¼
.5
ª −.6
«
8. We solve Px = x by rewriting the equation as (P – I )x = 0, where P − I = « 0 −.5
«¬ .6
0
reducing the augmented matrix for the homogeneous system (P – I )x = 0 gives
ª −.6
« 0
«
¬« .6
.5
.8
−.5
.1
0
−.9
0º ª 1
0 »» ∼ ««0
0 ¼» ¬« 0
0
−3 / 2
1
−1 / 5
0
0
ª1
10. Since P k = «
¬«0
matrix.
.8º
.1»» . Row
−.9 »¼
0º
0»»
0¼»
ª x1 º
ª3 / 2 º
ª15 º
«
»
«
»
Thus x = « x2 » = x3 « 1 / 5» , and one solution is «« 2 »» . Since the entries in
«¬ x3 »¼
«¬ 1»¼
«¬10 »¼
ª15 / 27 º ª.556º
«
» «
»
by 1/27 to obtain the steady-state vector q = « 2 / 27 » ≈ «.074» .
«¬10 / 27 »¼ «¬.370»¼
ª.84
9. Since P 2 = «
¬.16
.1º
.2 »» . Row
−.3»¼
ª15 º
« 2 » sum to 27, multiply
« »
«¬10 »¼
.2 º
has all positive entries, P is a regular stochastic matrix.
.8»¼
1 − .7 k º
» will have a zero as its (2,1) entry for all k, P is not a regular stochastic
.7 k ¼»
ª.7
11. a. From Exercise 1, P = «
¬.3
augmented matrix gives
.6 º
ª −.3
, so P − I = «
»
.4 ¼
¬ .3
.6 º
. Solving (P – I )x = 0 by row reducing the
−.6 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.9
ª −.3
« .3
¬
.6
−.6
0º ª 1
∼
0 »¼ «¬ 0
−2
0
• Solutions
263
0º
0 »¼
ªx º
ª2º
ª2º
ª 2º
Thus x = « 1 » = x2 « » , and one solution is « » . Since the entries in « » sum to 3, multiply by 1/3
¬1 ¼
¬1 ¼
¬1 ¼
¬ x2 ¼
ª 2 / 3º ª.667 º
to obtain the steady-state vector q = «
»≈«
».
¬ 1/ 3 ¼ ¬.333¼
ª 2 / 3º
b. Since q = «
» , 2/3 of the listeners will be listening to the news at some time late in the day.
¬1 / 3 ¼
.2 º
ª −.4
»
.6 .2 » , so P − I = «« .2
«¬ .2
.2 .6 »¼
reducing the augmented matrix gives
ª.6
12. From Exercise 2, P = ««.2
«¬.2
ª −.4
« .2
«
«¬ .2
.2
.2
−.4
.2
.2
−.4
.2
0º ª 1
0 »» ∼ «« 0
0 »¼ «¬0
0
−1
1
−1
0
0
.2
−.4
.2
.2 º
.2 »» . Solving (P – I )x = 0 by row
−.4 »¼
0º
0 »»
0 »¼
ª x1 º
ª1º
ª1º
ª1º
«
»
«
»
«
»
Thus x = « x2 » = x3 «1» , and one solution is «1» . Since the entries in ««1»» sum to 3, multiply by 1/3 to
«¬ x3 »¼
«¬1»¼
«¬1»¼
«¬1»¼
ª1/ 3º ª.333º
« » «
»
obtain the steady-state vector q = «1/ 3» ≈ «.333» . Thus in the long run each food will be preferred
«¬1/ 3»¼ «¬.333»¼
equally.
ª −.05
ª.95 .45º
13. a. From Exercise 3, P = «
, so P − I = «
»
¬ .05
¬.05 .55¼
reducing the augmented matrix gives
ª −.05
« .05
¬
.45
−.45
0º ª 1
∼
0 ¼» «¬0
−9
0
.45º
. Solving (P – I )x = 0 by row
−.45»¼
0º
0 »¼
ªx º
ª9 º
ª9 º
Thus x = « 1 » = x2 « » , and one solution is « » . Since the entries in
¬1 ¼
¬1 ¼
¬ x2 ¼
ª9 /10 º ª.9 º
1/10 to obtain the steady-state vector q = «
» = « ».
¬1/10 ¼ ¬ .1¼
ª9 º
«1 » sum to 10, multiply by
¬ ¼
b. After many days, a specific student is ill with probability .1, and it does not matter whether that
student is ill today or not.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
264
CHAPTER 4
• Vector Spaces
.3º
ª −.6
»
.2 .4 » , so P − I = «« .3
«¬ .3
.3 .3»¼
reducing the augmented matrix gives
ª.4
14. From Exercise 4, P = ««.3
«¬.3
ª −.6
« .3
«
¬« .3
.5
.3
−.8
.4
.3
−.7
.5
0º ª 1
0 »» ∼ ««0
0 ¼» ¬«0
0
−4 / 3
1
−1
0
0
.5
−.8
.3
.3º
.4 »» . Solving (P – I )x = 0 by row
−.7 »¼
0º
0 »»
0 ¼»
ª x1 º
ª 4 / 3º
ª4º
ª4º
«
»
«
»
«
»
Thus x = « x2 » = x3 « 1» , and one solution is « 3 » . Since the entries in «« 3 »» sum to 10, multiply by
«¬ x3 »¼
«¬ 1»¼
«¬ 3 »¼
«¬ 3 »¼
ª 4 / 10 º ª.4º
«
» « »
1/10 to obtain the steady-state vector q = « 3 / 10 » = «.3» . Thus in the long run the chance that a day
«¬ 3 / 10 »¼ «¬.3»¼
has good weather is 40%.
ª −.0179
ª.9821 .0029 º
15. [M] Let P = «
, so P − I = «
»
¬ .0179
¬.0179 .9971¼
the augmented matrix gives
ª −.0179
« .0179
¬
.0029
−.0029
0º ª 1
∼
0 »¼ «¬ 0
−.162011
0
.0029 º
. Solving (P – I )x = 0 by row reducing
−.0029 »¼
0º
0 »¼
ªx º
ª.162011º
Thus x = « 1 » = x2 «
» , and one solution is
¬ 1 ¼
¬ x2 ¼
ª.162011º
« 1 » . Since the entries in
¬
¼
ª.162011º
« 1 » sum to
¬
¼
ª.139423º
1.162011, multiply by 1/1.162011 to obtain the steady-state vector q = «
» . Thus about
¬.860577 ¼
13.9% of the total U.S. population would eventually live in California.
ª −.10
ª.90 .01 .09º
«
«
»
16. [M] Let P = « .01 .90 .01» , so P − I = « .01
«¬.09 .09 .90»¼
«¬ .09
reducing the augmented matrix gives
ª −.10
« .01
«
¬« .09
.01
.09
−.10
.01
.09
−.1
0º ª 1
0 »» ∼ ««0
0 ¼» «¬0
.01
−.10
.09
0
−.919192
1
−.191919
0
0
.09º
.01»» . Solving (P – I )x = 0 by row
−.1»¼
0º
0 »»
0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
4.9
ª x1 º
ª.919192 º
«
»
«
»
Thus x = « x2 » = x3 «.191919 » , and one solution is
«¬ x3 »¼
«¬ 1 »¼
ª.919192º
«.191919» . Since the entries in
«
»
«¬ 1 »¼
• Solutions
265
ª.919192º
«.191919» sum
«
»
«¬ 1 »¼
ª.435407 º
«
»
to 2.111111, multiply by 1/2.111111 to obtain the steady-state vector q = «.090909 » . Thus on a
¬«.473684 ¼»
typical day, about (.090909)(2000) = 182 cars will be rented or available from the downtown
location.
17. a. The entries in each column of P sum to 1. Each column in the matrix P – I has the same entries as
in P except one of the entries is decreased by 1. Thus the entries in each column of P – I sum to 0,
and adding all of the other rows of P – I to its bottom row produces a row of zeros.
b. By part a., the bottom row of P – I is the negative of the sum of the other rows, so the rows of
P – I are linearly dependent.
c. By part b. and the Spanning Set Theorem, the bottom row of P – I can be removed and the
remaining (n – 1) rows will still span the row space of P – I. Thus the dimension of the row space
of P – I is less than n. Alternatively, let A be the matrix obtained from P – I by adding to the
bottom row all the other rows. These row operations did not change the row space, so the row
space of P – I is spanned by the nonzero rows of A. By part a., the bottom row of A is a zero row,
so the row space of P – I is spanned by the first (n – 1) rows of A.
d. By part c., the rank of P – I is less than n, so the Rank Theorem may be used to show that
dimNul(P – I ) = n – rank(P – I ) > 0. Alternatively the Invertible Martix Theorem may be used
since P – I is a square matrix.
ª1 0 º
18. If α = β = 0 then P = «
» . Notice that Px = x for any vector x in
¬0 1 ¼
two linearly independent steady-state vectors in this case.
ª −α
If α ≠ 0 or β ≠ 0, we solve (P – I )x = 0 where P − I = «
¬α
matrix gives
ª −α
«α
¬
β
−β
0º ªα
∼
0 »¼ «¬ 0
−β
0
ª1 º
ª0º
, and that « » and « » are
¬0¼
¬1 ¼
2
β º
. Row reducing the augmented
− β »¼
0º
0 »¼
ª x º ªβ º
So α x1 = β x2 , and one possible solution is to let x1 = β , x2 = α . Thus x = « 1 » = « » . Since the
¬ x2 ¼ ¬α ¼
ªβº
1 ªβ º
entries in « » sum to α + β, multiply by 1/(α + β ) to obtain the steady-state vector q =
.
α + β «¬α »¼
¬ α¼
19. a. The product Sx equals the sum of the entries in x. Thus x is a probability vector if and only if its
entries are nonnegative and Sx = 1.
b. Let P = [p1
p2
…
SP = [ Sp1
Sp 2
p n ] , where p1 , p 2 , …, p n are probability vectors. By part a.,
…
Sp n ] = [1 1 … 1] = S
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
266
CHAPTER 4
• Vector Spaces
c. By part b., S(Px) = (SP)x = Sx = 1. The entries in Px are nonnegative since P and x have only
nonnegative entries. By part a., the condition S(Px) = 1 shows that Px is a probability vector.
20. Let P = [p1
p2
…
p n ] , so P 2 = PP = [ Pp1
2
Pp 2
…
Pp n ]. By Exercise 19c., the columns
2
of P are probability vectors, so P is a stochastic matrix.
Alternatively, SP = S by Exercise 19b., since P is a stochastic matrix. Right multiplication by P gives
SP 2 = SP , so SP = S implies that SP 2 = S . Since the entries in P are nonnegative, so are the entries
in P 2 , and P 2 is stochastic matrix.
21. [M]
a. To four decimal places,
ª.2779
«.3368
P2 = «
«.1847
«
«¬.2005
.2780
.3355
.2941º
ª.2817
»
.3335» 3 ««.3356
,P =
«.1817
.1697 »
»
«
.2027 »¼
«¬.2010
.2803
.3357
.1861 .1833
.2004 .2007
ª.2816
«.3355
4
5
P =P =«
«.1819
«
«¬.2009
.2816
.3355
.2816
.3355
.1819
.2009
.1819
.2009
.2817
.3356
.2817
.3355
.1817
.2010
.1819
.2010
.2814º
.3352»»
,
.1825 »
»
.2009»¼
.2816 º
.3355»»
.1819 »
»
.2009 »¼
The columns of P k are converging to a common vector as k increases. The steady state vector q
ª.2816 º
«.3355»
» , which is the vector to which the columns of P k are converging.
for P is q = «
«.1819»
«
»
¬«.2009¼»
b. To four decimal places,
ª.8222
Q = ««.0324
«¬.1453
.4044
ª.7477
Q = ««.0783
«¬.1740
.6815
10
30
.3966
.1990
.1329
.1856
.5385º
ª.7674
»
20
.1666 » , Q = ««.0637
«¬.1688
.2949 »¼
.1964
.6690 º
.1326 »» ,
.1984 »¼
.7105º
ª .7401 .7140
»
40
.1074 » , Q = ««.0843 .1057
«¬.1756 .1802
.1821»¼
.7257 º
.0960»» ,
.1783»¼
ª.7372 .7269 .7315º
ª.7360
«
»
60
Q = «.0867 .0951 .0913» , Q = ««.0876
«¬ .1761 .1780 .1772 »¼
«¬.1763
50
Q
70
ª.7356
= ««.0880
«¬.1764
.7340
.0893
.1767
.6000
.2036
.7338º
.0909 .0894»» ,
.1771 .1767 »¼
.7320
.7347 º
ª.7354 .7348
»
80
.0887 » , Q = «« .0881 .0887
«¬.1764 .1766
.1766 »¼
.7351º
.0884 »» ,
.1765»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 4
116
Q
117
=Q
ª.7353
= ««.0882
«¬.1765
• Supplementary Exercises
267
.7353º
.0882 »»
.1765»¼
.7353
.0882
.1765
ª.7353 º
The steady state vector q for Q is q = ««.0882 »» Conjecture: the columns of P k , where P is a
«¬.1765 »¼
regular stochastic matrix, converge to the steady state vector for P as k increases.
c. Let P be an n × n regular stochastic matrix, q the steady state vector of P, and e j the j th column
of the n × n identity matrix. Consider the Markov chain {x k } where x k +1 = P x k and x 0 = e j . By
Theorem 18, x k = P k x0 converges to q as k → ∞. But P k x 0 = P k e j , which is the j th column of
P k . Thus the j th column of P k converges to q as k → ∞; that is, P k → [q
q
…
q] .
22. [M] Answers will vary.
MATLAB Student Version 4.0 code for Method (1):
A=randstoc(32); flops(0);
tic, x=nulbasis(A-eye(32));
q=x/sum(x); toc, flops
MATLAB Student Version 4.0 code for Method (2):
A=randstoc(32); flops(0);
tic, B=A^100; q=B(: ,1); toc, flops
Chapter 4 SUPPLEMENTARY EXERCISES
1. a. True. This set is Span{v1 , ... v p } , and every subspace is itself a vector space.
b. True. Any linear combination of v1 , …, v p−1 is also a linear combination of v1 , …, v p−1 , v p
using the zero weight on v p .
c. False. Counterexample: Take v p = 2 v1 . Then {v1 , . .. v p } is linearly dependent.
d. False. Counterexample: Let {e1 , e 2 , e3 } be the standard basis for 3. Then {e1 , e 2 } is a linearly
independent set but is not a basis for 3.
e. True. See the Spanning Set Theorem (Section 4.3).
f. True. By the Basis Theorem, S is a basis for V because S spans V and has exactly p elements. So
S must be linearly independent.
g. False. The plane must pass through the origin to be a subspace.
ª2
«
h. False. Counterexample: « 0
«¬ 0
5
−2
0
7
0
0
0º
3»» .
0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
268
CHAPTER 4
• Vector Spaces
i. True. This statement appears before Theorem 13 in Section 4.6.
j. False. Row operations on A do not change the solutions of Ax = 0.
ª1 2 º
k. False. Counterexample: A = «
» ; A has two nonzero rows but the rank of A is 1.
¬3 6 ¼
l. False. If U has k nonzero rows, then rank A = k and dimNul A = n – k by the Rank Theorem.
m. True. Row equivalent matrices have the same number of pivot columns.
n. False. The nonzero rows of A span Row A but they may not be linearly independent.
o. True. The nonzero rows of the reduced echelon form E form a basis for the row space of each
matrix that is row equivalent to E.
p. True. If H is the zero subspace, let A be the 3 × 3 zero matrix. If dim H = 1, let {v} be a basis
for H and set A = [ v v v ] . If dim H = 2, let {u,v} be a basis for H and set A = [u v v ] ,
for example. If dim H = 3, then H = 3, so A can be any 3 × 3 invertible matrix. Or, let {u, v,
w} be a basis for H and set A = [u v w ] .
ª1 0 0 º
q. False. Counterexample: A = «
» . If rank A = n (the number of columns in A), then the
¬0 1 0¼
transformation x6Ax is one-to-one.
r. True. If x6Ax is onto, then Col A =
m
and rank A = m. See Theorem 12(a) in Section 1.9.
s. True. See the second paragraph after Theorem 15 in Section 4.7.
t. False. The j th column of P is ª¬b j º¼ .
C
C ←B
°
°
2. The set is SpanS, where S = ®
°
°
¯
ª −2 º
« »
« 5» ,
« −4 »
« »
¬« 1¼»
½
°
°
¾ . Note that S is a linearly dependent set, but each
°
°
¿
ª 1º ª −2 º
« 2» « 5»
pair of vectors in S forms a linearly independent set. Thus any two of the three vectors « » , « » ,
«−1» « −4 »
« » « »
¬« 3¼» ¬« 1¼»
ª 1º
« »
« 2» ,
« −1»
« »
¬« 3¼»
ª 5º
« »
« −8»
« 7»
« »
¬« 1¼»
ª 5º
« −8 »
« » will be a basis for SpanS.
« 7»
« »
¬« 1¼»
3. The vector b will be in W = Span{u1 , u 2 } if and only if there exist constants c1 and c2 with
c1u1 + c2 u 2 = b. Row reducing the augmented matrix gives
ª −2
« 4
«
¬« −6
1
2
−5
b1 º ª −2
b2 »» ∼ «« 0
b3 ¼» ¬« 0
1
4
0
º
2b1 + b2 »»
b1 + 2b2 + b3 ¼»
b1
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 4
• Supplementary Exercises
269
so W = Span{u1 , u 2 } is the set of all (b1 , b2 , b3 ) satisfying b1 + 2b2 + b3 = 0.
4. The vector g is not a scalar multiple of the vector f, and f is not a scalar multiple of g, so the set
{f, g} is linearly independent. Even though the number g(t) is a scalar multiple of f(t) for each t, the
scalar depends on t.
5. The vector p1 is not zero, and p 2 is not a multiple of p1 . However, p3 is 2p1 + 2p 2 , so p3 is
discarded. The vector p 4 cannot be a linear combination of p1 and p 2 since p 4 involves t 2 but p1
and p 2 do not involve t 2 . The vector p5 is (3/ 2)p1 − (1/ 2)p 2 + p 4 (which may not be so easy to see
at first.) Thus p5 is a linear combination of p1 , p 2 , and p 4 , so p5 is discarded. So the resulting
basis is {p1 , p 2 , p 4 }.
6. Find two polynomials from the set {p1 , . . . , p 4 } that are not multiples of one another. This is easy,
because one compares only two polynomials at a time. Since these two polynomials form a linearly
independent set in a two-dimensional space, they form a basis for H by the Basis Theorem.
7. You would have to know that the solution set of the homogeneous system is spanned by two
solutions. In this case, the null space of the 18 × 20 coefficient matrix A is at most two-dimensional.
By the Rank Theorem, dimCol A = 20 – dimNul A ≥ 20 – 2 = 18. Since Col A is a subspace of 18,
Col A = 18. Thus Ax = b has a solution for every b in 18.
8. If n = 0, then H and V are both the zero subspace, and H = V. If n > 0, then a basis for H consists of n
linearly independent vectors u1 , . . . , u n . These vectors are also linearly independent as elements of V.
But since dimV = n, any set of n linearly independent vectors in V must be a basis for V by the Basis
Theorem. So u1 , . . . , u n span V, and H = Span{u1 , . . . , u n } = V .
9. Let T: n → m be a linear transformation, and let A be the m × n standard matrix of T.
a. If T is one-to-one, then the columns of A are linearly independent by Theorem 12 in Section 1.9,
so dimNul A = 0. By the Rank Theorem, dimCol A = n – 0 = n, which is the number of columns
of A. As noted in Section 4.2, the range of T is Col A, so the dimension of the range of T is n.
b. If T maps n onto m, then the columns of A span m by Theoerm 12 in Section 1.9, so dimCol A
= m. By the Rank Theorem, dimNul A = n – m. As noted in Section 4.2, the kernel of T is Nul A,
so the dimension of the kernel of T is n – m. Note that n – m must be nonnegative in this case:
since A must have a pivot in each row, n ≥ m.
10. Let S = {v1 , . .. , v p }. If S were linearly independent and not a basis for V, then S would not span V.
In this case, there would be a vector v p+1 in V that is not in Span{v1 , ... , v p }. Let
S ′ = {v1 , ... , v p , v p +1}. Then S ′ is linearly independent since none of the vectors in S ′ is a linear
combination of vectors that precede it. Since S ′ has more elements than S, this would contradict the
maximality of S. Hence S must be a basis for V.
11. If S is a finite spanning set for V, then a subset of S is a basis for V. Denote this subset of S by S ′.
Since S ′ is a basis for V, S ′ must span V. Since S is a minimal spanning set, S ′ cannot be a proper
subset of S. Thus S ′ = S, and S is a basis for V.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
270
CHAPTER 4
• Vector Spaces
12. a. Let y be in Col AB. Then y = ABx for some x. But ABx = A(Bx), so y = A(Bx), and y is in Col A.
Thus Col AB is a subspace of Col A, so rank AB = dimCol AB ≤ dimCol A = rank A by Theorem
11 in Section 4.5.
b. By the Rank Theorem and part a.:
rank AB = rank( AB)T = rank BT AT ≤ rank BT = rank B
13. By Exercise 12, rank PA ≤ rank A, and rank A = rank( P −1 P ) A = rank P −1 ( PA) ≤ rank PA , so
rank PA = rank A.
14. Note that ( AQ )T = QT AT . Since QT is invertible, we can use Exercise 13 to conclude that
rank( AQ)T = rank QT AT = rank AT. Since the ranks of a matrix and its transpose are equal (by the
Rank Theorem), rank AQ = rank A.
15. The equation AB = O shows that each column of B is in Nul A. Since Nul A is a subspace of n, all
linear combinations of the columns of B are in Nul A. That is, Col B is a subspace of Nul A. By
Theorem 11 in Section 4.5, rank B = dimCol B ≤ dimNul A. By this inequality and the Rank
Theorem applied to A,
n = rank A + dimNul A ≥ rank A + rank B
16. Suppose that rank A = r1 and rank B = r2 . Then there are rank factorizations A = C1 R1 and B = C2 R2
of A and B, where C1 is m × r1 with rank r1 , C2 is m × r2 with rank r2 , R1 is r1 × n with rank r1 , and
R2 is r2 × n with rank r2 . Create an m × (r1 + r2 ) matrix C = [C1
C2 ] and an ( r1 + r2 ) × n matrix R
by stacking R1 over R2 . Then
ªR º
C2 ] « 1 » = CR
¬ R2 ¼
Since the matrix CR is a product, its rank cannot exceed the rank of either of its factors by Exercise
12. Since C has r1 + r2 columns, the rank of C cannot exceed r1 + r2 . Likewise R has r1 + r2 rows, so
the rank of R cannot exceed r1 + r2 . Thus the rank of A + B cannot exceed r1 + r2 = rank A + rank B,
or rank (A + B) ≤ rank A + rank B.
A + B = C1 R1 + C2 R2 = [C1
17. Let A be an m × n matrix with rank r.
(a) Let A1 consist of the r pivot columns of A. The columns of A1 are linearly independent, so A1
is an m × r matrix with rank r.
(b) By the Rank Theorem applied to A1, the dimension of RowA1 is r, so A1 has r linearly
independent rows. Let A2 consist of the r linearly independent rows of A1. Then A2 is an r × r
matrix with linearly independent rows. By the Invertible Matrix Theorem, A2 is invertible.
18. Let A be a 4 × 4 matrix and B be a 4 × 2 matrix, and let u 0 , . . . , u 3 be a sequence of input vectors in
2
.
a. Use the equation x k +1 = Ax k + Bu k for k = 0, . . . , 4, with x 0 = 0.
x1 = Ax 0 + Bu 0 = Bu 0
x 2 = Ax1 + Bu1 = ABu 0 + Bu1
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 4
• Supplementary Exercises
271
x3 = Ax 2 + Bu 2 = A( ABu 0 + Bu1 ) + Bu 2 = A2 Bu 0 + ABu1 + Bu 2
x 4 = Ax3 + Bu3 = A( A2 Bu 0 + ABu1 + Bu 2 ) + Bu3
= A3 Bu 0 + A2 Bu1 + ABu 2 + Bu 3
= ª¬ B
AB
ªu3 º
«u »
A3 B º¼ « 2 » = M u
« u1 »
« »
¬«u0 ¼»
A2 B
Note that M has 4 rows because B does, and that M has 8 columns because B and each of the
matrices Ak B have 2 columns. The vector u in the final equation is in 8, because each u k is in
2
.
b. If (A, B) is controllable, then the controllability matrix has rank 4, with a pivot in each row, and
the columns of M span 4. Therefore, for any vector v in 4, there is a vector u in 8 such that
v = Mu. However, from part a. we know that x 4 = M u when u is partitioned into a control
sequence u 0 ,…, u 3 . This particular control sequence makes x 4 = v.
19. To determine if the matrix pair (A, B) is controllable, we compute the rank of the matrix
ª B AB A2 B º . To find the rank, we row reduce:
¬
¼
ªB
¬
AB
ª0
A B º¼ = «« 1
«¬ 1
2
1
−.9
.5
0º ª 1
.81»» ∼ ««0
.25»¼ «¬0
0
1
0
0º
0 »» .
1»¼
The rank of the matrix is 3, and the pair (A, B) is controllable.
20. To determine if the matrix pair (A, B) is controllable, we compute the rank of the matrix
ª B AB A2 B º . To find the rank, we note that :
¬
¼
ªB
¬
AB
ª 1 .5
A B º¼ = «« 1 .7
«¬0 0
2
.19 º
.45»» .
0 »¼
The rank of the matrix must be less than 3, and the pair (A, B) is not controllable.
21. [M] To determine if the matrix pair (A, B) is controllable, we compute the rank of the matrix
ª B AB A2 B A3 B º . To find the rank, we row reduce:
¬
¼
ªB
¬
AB
A2 B
0
ª 1
« 0
0
A3 B º¼ = «
« 0 −1
«
¬« −1 1.6
0
−1
1.6
−.96
−1º ª 1
1.6»» «« 0
∼
−.96 » «0
» «
−.024¼» ¬« 0
0
1
0
0
0
0
1
0
−1º
−1.6»»
.
−1.6»
»
0¼»
The rank of the matrix is 3, and the pair (A, B) is not controllable.
22. [M] To determine if the matrix pair (A, B) is controllable, we compute the rank of the matrix
ª B AB A2 B A3 B º . To find the rank, we row reduce:
¬
¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
272
CHAPTER 4
ªB
¬
• Vector Spaces
AB
A2 B
ª 1
« 0
3
«
º
A B¼ =
« 0
«
¬« −1
0
0
0
−1
.5
−1
.5 11.45
−1º ª1
.5»» ««0
∼
11.45» «0
» «
−10.275¼» ¬«0
0
1
0
0
0
0
1
0
0º
0»»
.
0»
»
1 ¼»
The rank of the matrix is 4, and the pair (A, B) is controllable.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.1
SOLUTIONS
Notes: Exercises 1–6 reinforce the deefinitions of eigenvalues and eigenvectors. The suubsection on
eigenvectors and difference equations, aalong with Exercises 33 and 34, refers to the chapter introductory
example and anticipates discussions of ddynamical systems in Sections 5.2 and 5.6.
1. The number 2 is an eigenvalue of A if and only if the equation Ax = 2x has a nontrivial solution.
s
This equation is equivalent to ( A − 2 I )x = 0. Compute
ª3 2 º ª 2 0 º ª 1 2 º
A − 2I = «
»−«
»=«
»
¬3 8 ¼ ¬ 0 2 ¼ ¬ 3 6 ¼
The columns of A are obviously lineearly dependent, so ( A − 2 I )x = 0 has a nontrivial solu
ution, and
so 2 is an eigenvalue of A.
2. The number − 3 is an eigenvalue off A if and only if the equation Ax = −3x has a nontriviial solution.
This equation is equivalent to ( A + 3I )x = 0. Compute
ª −1 4º ª 3 0º ª 2 4º
A + 3I = «
»+«
»=«
»
¬ 6 9¼ ¬0 3¼ ¬ 6 12¼
The columns of A are obviously lineearly dependent, so ( A + 3I )x = 0 has a nontrivial solu
ution, and
so −3 is an eigenvalue of A.
ª1
3. Is Ax a multiple of x? Compute «
¬6
eigenvalue −2 .
−1º ª 1º ª −2º
ª 1º
ª 1º
= « » = −2 « » . So « » is an eigenvector of A with
»
«
»
−4¼ ¬3¼ ¬ −6¼
¬3¼
¬3¼
ª5
4. Is Ax a multiple of x? Compute «
¬3
eigenvalue 3 .
2º ª −1º ª« −3º»
ª −1º
ª−1º
= « » = 3 « » So « » is an eigenvectorr of A with
»
«
»
6¼ ¬ 1¼ ¬« 3¼»
¬ 1¼
¬ 1¼
Copyright © 2012 Pearrson Education, Inc. Publishing as Addison-Wesley.
273
274
CHAPTER 5
• Eigenvalues and Eigenvectors
ª −4
5. Is Ax a multiple of x? Compute «« 2
«¬ −1
of A for the eigenvalue −5 .
ª3
6. Is Ax a multiple of x? Compute ««3
«¬5
3
−3
0
3º ª 3º ª −15º
ª 3º
ª 3º
»
«
»
«
»
«
»
−2 » « −2 » = « 10 » = −5 « −2 » . So «« −2 »» is an eigenvector
«¬ 1»¼
«¬ 1»¼
−2 »¼ «¬ 1»¼ «¬ −5»¼
7 º ª 1º ª 5º
ª 1º
ª 1º
»
«
»
«
»
«
»
7 » « −2» = «13» ≠ λ « −2 » So «« −2 »» is not
«¬ 2 »¼
«¬ 2 »¼
4 »¼ «¬ 2»¼ «¬ 1»¼
6
2
6
an eigenvector of A.
7. To determine if 4 is an eigenvalue of A, decide if the matrix A − 4 I is invertible.
ª 3
A − 4 I = «« 2
«¬ −3
0
3
4
−1º ª 4
1»» − «« 0
5»¼ «¬ 0
0
4
0
0 º ª −1
0 »» = «« 2
4 »¼ «¬ −3
−1º
1»»
1»¼
0
−1
4
Invertibility can be checked in several ways, but since an eigenvector is needed in the event that one
exists, the best strategy is to row reduce the augmented matrix for ( A − 4 I )x = 0 :
ª −1
« 2
«
«¬ −3
0
−1
−1
4
1
1
0 º ª −1
0 »» «« 0
0 »¼ «¬ 0
0
−1
−1
4
−1
4
0º ª 1
0 »» ««0
0»¼ «¬0
0
1
−1
0
−1
0
0º
0»»
0»¼
The equation ( A − 4 I )x = 0 has a nontrivial solution, so 4 is an eigenvalue. Any nonzero solution of
( A − 4 I )x = 0 is a corresponding eigenvector. The entries in a solution satisfy x1 + x3 = 0 and
− x2 − x3 = 0, with x3 free. The general solution is not requested, so to save time, simply take any
nonzero value for x3 to produce an eigenvector. If x3 = 1, then x = (−1, − 1, 1).
Note: The answer in the text is (1, 1, − 1), written in this form to make the students wonder whether the
more common answer given above is also correct. This may initiate a class discussion of what answers
are “correct.”
8. To determine if 1 is an eigenvalue of A, decide if the matrix A − I is invertible.
ª 4
A − I = «« 0
«¬ −1
−2
−1
2
3º ª 1
3»» − «« 0
−2 »¼ «¬ 0
0
1
0
0º ª 3
0»» = «« 0
1»¼ «¬ −1
−2
−2
2
3º
3»»
−3»¼
Row reducing the augmented matrix [(A − I ) 0] yields:
ª 3
« 0
«
«¬ −1
−2
3
−2
2
3
−3
0º ª 1
0 »» «« 0
0 »¼ «¬ 0
−2
3
−2
4
3
−6
0º ª 1
0 »» «« 0
0 »¼ «¬ 0
0
0
2
0
−3
0
0º ª 1
0 »» «« 0
0 »¼ «¬ 0
0
0
1
0
−3 / 2
0
0º
0 »»
0 »¼
The equation ( A − I )x = 0 has a nontrivial solution, so 1 is an eigenvalue. Any nonzero solution of
( A − I )x = 0 is a corresponding eigenvector. The entries in a solution satisfy x1 = 0 and
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.1
• Solutions
275
x2 − (3 / 2) x3 = 0, with x3 free. The general solution is not requested, so to save time, simply take
ª0º
any nonzero value for x3 to produce an eigenvector. If x3 = 2, then x = «« 3»» .
«¬ 2»¼
ª3
9. For λ = 1 : A − 1I = «
¬2
0º ª 1
−
1»¼ «¬0
0º ª 2
=
1»¼ «¬ 2
ª2
The augmented matrix for ( A − I )x = 0 is «
¬2
0º
0»¼
0
0
0º
. Thus x1 = 0 and x2 is free. The general
0»¼
ª0 º
solution of ( A − I )x = 0 is x2 e 2 , where e2 = « » , and so e 2 is a basis for the eigenspace
¬1 ¼
corresponding to the eigenvalue 1.
0º
ª 3 0º ª3 0º ª 0
−«
=«
For λ = 3 : A − 3I = «
»
»
»
¬ 2 1 ¼ ¬0 3¼ ¬ 2 −2 ¼
The equation ( A − 3I )x = 0 leads to 2 x1 − 2 x2 = 0, so that x1 = x2 and x2 is free. The general
solution is
ªx º
« 1»
« »
«¬ x2 »¼
ªx º
2»
»
«¬ x2 »¼
= ««
ª1º
ª1º
= x2 « » . So « » is a basis for the eigenspace.
¬1¼
¬1¼
Note: For simplicity, the answer omits the set brackets when listing a basis. I permit my students to
list a basis without the set brackets. Some instructors may prefer to include brackets.
ª −4
10. For λ = −5 : A + 5I = «
¬3
2º
ª1
+ 5«
»
1¼
¬0
0º ª1
=
1 »¼ «¬3
2º
.
6 »¼
ª1 2 0º ª1 2 0º
The augmented matrix for ( A + 5I )x = 0 is «
»«
» . Thus x1 = −2 x2 and x2 is
¬3 6 0 ¼ ¬ 0 0 0 ¼
ªx º
ª −2 x º
ª −2 º
1
2»
free. The general solution is «« »» = ««
» = x2 «
» . A basis for the eigenspace corresponding to −5
«¬ x2 »¼
«¬ x2 »¼
¬1¼
ª−2º
is « » .
¬1¼
ª 1
11. For λ = −1 : A + I = «
¬ −4
−3º ª1
+
5»¼ «¬0
0º ª 2
=
1 »¼ «¬ −4
−3º
6»¼
ª 2 −3 0º ª1 −3 / 2 0º
. Thus x1 = (3 / 2) x2 and
The augmented matrix for ( A + I )x = 0 is «
6 0»¼ «¬0
0
0»¼
¬ −4
ª x1 º ª (3 / 2) x2 º»
ª3 / 2 º
x2 is free. The general solution is « » = ««
» = x2 «
» . A basis for the eigenspace
x2 »¼
¬ 1¼
¬ x2 ¼ «¬
ª3 / 2 º
corresponding to −1 is «
» . Another choice is
¬ 1 ¼
ª 3º
« 2» .
¬ ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
276
CHAPTER 5
• Eigenvalues and Eigenvectors
ª 1
For λ = 7 : A − 7 I = «
¬ −4
−3º ª7
−
5»¼ «¬ 0
0 º ª −6
=
7 »¼ «¬ −4
−3º
−2»¼
ª −6 −3 0º ª1 1 / 2 0º
. Thus x1 = ( −1 / 2) x2 and x2
The augmented matrix for ( A − 7 I )x = 0 is «
»«
0
0»¼
¬ −4 −2 0¼ ¬0
ª x1 º ª (−1 / 2) x2 º»
ª −1 / 2 º
is free. The general solution is « » = ««
. A basis for the eigenspace
» = x2 «
x2 ¼»
1»¼
¬
¬ x2 ¼ ¬«
ª −1 / 2 º
corresponding to 7 is «
» . Another choice is
¬ 1 ¼
ª4
12. For λ = 3 : A − 3I = «
¬3
1º ª3
−
6»¼ «¬0
0º ª 1
=
3»¼ «¬3
ª −1º
« 2» .
¬ ¼
1º
3»¼
ª1
The augmented matrix for ( A − 3I )x = 0 is «
¬3
0 º ª1
3 0»¼ «¬0
ª −1º
A basis for the eigenspace corresponding to 1 is « » .
¬ 1¼
ª4
For λ = 7 : A − 7 I = «
¬3
0 º ª −3
=
7 »¼ «¬ 3
1º ª7
−
6»¼ «¬ 0
1
1
0
0º
. Thus x1 = − x2 and x2 is free.
0»¼
1º
.
−1»¼
1 0º ª 1 −1 / 3 0º
ª −3
. Thus x1 = (1 / 3) x2 and
The augmented matrix for ( A − 7 I )x = 0 is «
»«
0 0»¼
¬ 3 −1 0¼ ¬0
ªx º
ª (1 / 3) x º
ª1 / 3º
1
2»
x2 is free. The general solution is «« »» = ««
» = x2 «
» . A basis for the eigenspace
x2 »¼
«¬ x2 »¼
«¬
¬ 1¼
ª1 / 3º
ª 1º
. Another choice is « » .
corresponding to 7 is «
»
¬ 1¼
¬3¼
13. For λ = 1:
ª 4
A − 1I = «« −2
«¬ −2
0
1
0
1º ª1
0 »» − «« 0
1»¼ «¬ 0
0
1
0
0º ª 3
0 »» = «« −2
1 »¼ «¬ −2
0
0
0
1º
0»»
0 »¼
3 x1 + x3 = 0 ½
The equations for ( A − I )x = 0 are easy to solve: ®
¾
= 0¿
¯−2 x1
Row operations hardly seem necessary. Obviously x1 is zero, and hence x3 is also zero. There are
ª0 º
three-variables, so x2 is free. The general solution of ( A − I )x = 0 is x2 e 2 , where e 2 = ««1 »» , and
«¬0 »¼
so e 2 provides a basis for the eigenspace.
For λ = 2:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.1
ª 4
A − 2 I = «« −2
«¬ −2
1º ª 2
0 »» − «« 0
1»¼ «¬ 0
0
1
0
ª 2
[( A − 2 I ) 0] = «« −2
«¬ −2
0
2
0
0
1
−1
0
0
−1
0º ª 2
0 »» = «« −2
2 »¼ «¬ −2
0º ª 2
0 »» «« 0
0 »¼ «¬ 0
• Solutions
277
1º
0 »»
−1»¼
0
−1
0
0
1
−1
0
1
0
0 º ª1
0 »» ««0
0 »¼ «¬0
0
1/ 2
1
0
−1
0
0º
0 »»
0 »¼
ª −1/ 2 º
So x1 = − (1/ 2) x3 , x2 = x3 , with x3 free. The general solution of ( A − 2 I )x = 0 is x3 «« 1 »» . A nice
¬« 1 ¼»
ª −1º
basis vector for the eigenspace is «« 2 »» .
«¬ 2 »¼
For λ = 3:
ª 4
A − 3I = «« −2
«¬ −2
0
1
0
ª 1
[( A − 3I ) 0] = «« −2
«¬ −2
1º ª 3
0 »» − «« 0
1»¼ «¬0
0
3
0
0
1
−2
0
0
−2
0º ª 1
0 »» = «« −2
3 »¼ «¬ −2
0º ª 1
0 »» «« 0
0 »¼ «¬ 0
1º
0»»
−2 »¼
0
−2
0
0
1
−2
0
2
0
0º ª 1
0 »» «« 0
0 »¼ «¬ 0
0
1
1
0
−1
0
0º
0 »»
0 »¼
ª −1º
So x1 = − x3 , x2 = x3 , with x3 free. A basis vector for the eigenspace is «« 1»» .
«¬ 1»¼
ª4
14. For λ = 3 : A − (3I ) = A − 3I = «« 3
«¬ 2
0
0
−2
−1º ª 3
3»» − ««0
5»¼ «¬0
0
3
0
0º ª 1
0 »» = «« 3
3»¼ «¬ 2
0
−3
−2
−1º
3»» .
2 »¼
The augmented matrix for ( A − 3I )x = 0 is
ª1
[( A − 3I ) 0] = «« 3
«¬ 2
0
−1
−3
−2
3
2
0º ª 1
0 »» «« 0
0 »¼ «¬0
0
−1
−3
0
6
0
0º ª 1
0 »» «« 0
0 »¼ «¬ 0
0
−1
1
0
−2
0
0º
0»»
0 »¼
ª1 º
Thus x1 = x3 , x2 = 2 x3 , with x3 free. The general solution of ( A − 3I )x = 0 is x3 «« 2 »» .
«¬ 1 »¼
ª1 º
A basis for the eigenspace corresponding to 2 is «« 2 »» .
«¬ 1 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
278
CHAPTER 5
• Eigenvalues and Eigenvectors
ª1
15. For λ = −5 : [( A + 5 I ) 0] = «« 2
«¬ 3
0º
2 2
0 0 0 »» . Thus x1 + x2 + x3 = 0, with x2 and
3 3
0 0 0 »¼
ª− x − x º
ª −1º
ª −1º
2
3»
«
«
»
«
»
x3 free. The general solution of ( A + 5 I )x = 0 is x = « x2 » = x2 1 + x3 « 0 » . A basis for the
« »
« »
«
»
«
«¬ 0 »¼
«¬ 1»¼
x3 »»¼
«¬
1
°
eigenspace corresponding to −5 is ®
°
¯
0º ª 1
0 »» ««0
0 »¼ «¬0
1
ª −1º ª −1º
« 1» , « 0 »
« » « »
«¬ 0 »¼ «¬ 1»¼
1
1
½
°
¾
°
¿
Note: For simplicity, the text answer omits the set brackets. I permit my students to list a basis without
the set brackets. Some instructors may prefer to include brackets.
16. For λ = 4 :
[( A − 4 I )
ª1
«1
0] = «
«2
«
¬4
0
−1
−1
0
−1
−2
−1
−2
0 º ª1
0 »» «« 0
0» «0
» «
0¼ ¬0
0
−1
−1
1
−1
−2
1
2
0º ª1
0 »» «« 0
=
0» « 0
» «
0¼ ¬ 0
0
−1
1
−1
0
0
0
0
0º
0»»
.
0»
»
0¼
So x1 = x3 , x2 = x3 , with x3 and x4
free variables. The general solution of ( A − 4 I )x = 0 is
x=
ªx º
« 1»
« »
« x2 »
« »
«x »
« 3»
« »
«¬ x4 »¼
=
ªx º
« 3»
«
»
« x3 »
«
»
«x »
« 3»
«
»
«¬ x4 »¼
ª1 º
ª0º
°
«1 »
«0»
°
= x3 « » + x4 « » . Basis for the eigenspace : ®
«1 »
«0»
°
« »
« »
°
0
1
¬ ¼
¬ ¼
¯
ª1 º ª 0º
«1 » « 0»
« »,« »
«1 » « 0»
« » « »
¬ 0¼ ¬1 ¼
½
°
°
¾
°
°
¿
Note: I urge my students always to include the extra column of zeros when solving a homogeneous
system. Exercise 16 provides a situation in which failing to add the column is likely to create problems
for a student, because the matrix A − 4 I itself has a column of zeros.
ª0
17. The eigenvalues of ««0
«¬ 0
1.
ª 5
18. The eigenvalues of «« 0
«¬ −1
ª1
19. The matrix ««1
«¬1
0
3
0
0
0
0
0º
4 »» are 0, 3, and −2 , the numbers on the main diagonal, by Theorem
−2 »¼
0º
0 »» are 5, 0, and 3, the numbers on the main diagonal, by Theorem 1.
3»¼
3º
2 3»» is not invertible because its columns are linearly dependent. So the number 0
2 3»¼
is an eigenvalue of the matrix. See the discussion following Example 5.
2
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.1
ª2
20. The matrix A = «« 2
«¬ 2
• Solutions
279
2º
2 2 »» is not invertible because its columns are linearly dependent. So the
2 2 »¼
number 0 is an eigenvalue of A. Eigenvectors for the eigenvalue 0 are solutions of Ax = 0 and
therefore have entries that produce a linear dependence relation among the columns of A. Any
nonzero vector (in R3 ) whose entries sum to 0 will work. Find any two such vectors that are not
multiples; for instance,
2
ª1º
ª1º
« 1 » and « −1» .
« »
« »
«¬ 0 »¼
«¬ −2 »¼
21. a. False. The equation Ax = λx must have a nontrivial solution.
b. True. See the paragraph after Example 5.
c. True. See the discussion of equation (3).
d. True. See Example 2 and the paragraph preceding it. Also, see the Numerical Note.
e. False. See the warning after Example 3.
22. a. False. The vector x in Ax = λx must be nonzero.
b. False. See Example 4 for a two-dimensional eigenspace, which contains two linearly independent
eigenvectors corresponding to the same eigenvalue. The statement given is not at all the same as
Theorem 2. In fact, it is the converse of Theorem 2 (for the case r = 2 ).
c. True. See the paragraph after Example 1.
d. False. Theorem 1 concerns a triangular matrix. See Examples 3 and 4 for counterexamples.
e. True. See the paragraph following Example 3. The eigenspace of A corresponding to λ is the null
space of the matrix A − λ I .
23. If a 2 × 2 matrix A were to have three distinct eigenvalues, then by Theorem 2 there would
correspond three linearly independent eigenvectors (one for each eigenvalue). This is impossible
because the vectors all belong to a two-dimensional vector space, in which any set of three vectors is
linearly dependent. See Theorem 8 in Section 1.7. In general, if an n × n matrix has p distinct
eigenvalues, then by Theorem 2 there would be a linearly independent set of p eigenvectors (one for
each eigenvalue). Since these vectors belong to an n-dimensional vector space, p cannot exceed n.
24. A simple example of a 2 × 2 matrix with only one distinct eigenvalue is a triangular matrix with the
same number on the diagonal. By experimentation, one finds that if such a matrix is actually a
diagonal matrix then the eigenspace is two dimensional, and otherwise the eigenspace is only one
dimensional.
ª4
Examples: «
¬0
1º
ª4
and «
»
4¼
¬0
5º
.
4»¼
25. If is an eigenvalue of A, then there is a nonzero vector x such that Ax = x. Since A is invertible,
A−1 Ax = A−1 ( x), and so x = ( A−1x). Since x ≠ 0 (and since A is invertible), cannot be zero.
Then
−1
x = A−1x, which shows that
−1
is an eigenvalue of A−1.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
280
CHAPTER 5
• Eigenvalues and Eigenvectors
Note: The Study Guide points out here that the relation between the eigenvalues of A and A−1 is
important in the so-called inverse power method for estimating an eigenvalue of a matrix. See Section 5.8.
26. Suppose that A2 is the zero matrix. If Ax = x for some x ≠ 0, then
A2 x = A( Ax) = A( x) = Ax =
zero.
2
x. Since x is nonzero,
must be zero. Thus each eigenvalue of A is
27. Use the Hint in the text to write, for any , ( A − I )T = AT − ( I )T = AT − I . Since ( A − I )T is
invertible if and only if A − I is invertible (by Theorem 6(c) in Section 2.2), it follows that AT − I
is not invertible if and only if A − I is not invertible. That is, is an eigenvalue of AT if and only
if is an eigenvalue of A.
Note: If you discuss Exercise 27, you might ask students on a test to show that A and AT have the same
characteristic polynomial (discussed in Section 5.2). Since det A = det AT , for any square matrix A,
det( A − I ) = det( A − I )T = det( AT − ( I )T ) = det( A − I ).
28. If A is lower triangular, then AT is upper triangular and has the same diagonal entries as A. Hence,
by the part of Theorem 1 already proved in the text, these diagonal entries are eigenvalues of AT . By
Exercise 27, they are also eigenvalues of A.
29. Let v be the vector in R n whose entries are all ones. Then Av = sv.
30. Suppose the column sums of an n × n matrix A all equal the same number s. By Exercise 29 applied
to AT in place of A, the number s is an eigenvalue of AT . By Exercise 27, s is an eigenvalue of A.
31. Suppose T reflects points across (or through) a line that passes through the origin. That line consists
of all multiples of some nonzero vector v. The points on this line do not move under the action of A.
So T ( v) = v. If A is the standard matrix of T, then Av = v. Thus v is an eigenvector of A
corresponding to the eigenvalue 1. The eigenspace is Span {v}. Another eigenspace is generated by
any nonzero vector u that is perpendicular to the given line. (Perpendicularity in R 2 should be a
familiar concept even though orthogonality in R n has not been discussed yet.) Each vector x on the
line through u is transformed into the vector −x. The eigenvalue is −1.
32. Since T rotates points around a given line, the points on the line are not moved at all. Hence 1 is an
eigenvalue of the standard matrix A of T, and the corresponding eigenspace is the line the points are
being rotated around.
33. (The solution is given in the text.)
a. Replace k by k + 1 in the definition of x k , and obtain x k +1 = c1λ k +1u + c2 µ k +1v.
b. Ax k = A(c1λ k u + c2 µ k v)
= c1λ k Au + c2 µ k Av by linearity
= c1λ k λ u + c2 µ k µ v since u and v are eigenvectors
= x k +1
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.1
• Solutions
281
34. You could try to write x0 as linear combination of eigenvectors, v1 , …, v p . If λ1 , …,λ p are
corresponding eigenvalues, and if x 0 = c1v1 + " + c p v p , then you could define
x k = c1λ1k v1 + " + c p λkp v p
In this case, for k = 0, 1, 2, …,
Ax k = A(c1λ1k v1 + " + c p λkp v p )
= c1λ1k Av1 + " + c p λkp Av p
Linearity
= c1λ1k +1 v1 + " + c p λkp +1 v p
The v i are eigenvectors.
= x k +1
35. Using the figure in the exercise, plot T (u) as 2u, because u is an eigenvector for the eigenvalue 2 of
the standard matrix A. Likewise, plot T ( v) as 3v, because v is an eigenvector for the eigenvalue 3.
Since T is linear, the image of w is T (w ) = T (u + v) = T (u) + T ( v).
36. As in Exercise 35, T (u) = −u and T ( v) = 3v because u and v are eigenvectors for the eigenvalues
−1 and 3, respectively, of the standard matrix A. Since T is linear, the image of w is
T (w ) = T (u + v) = T (u) + T ( v).
Note: The matrix programs supported by this text all have an eigenvalue command. In some cases, such
as MATLAB, the command can be structured so it provides eigenvectors as well as a list of the
eigenvalues. At this point in the course, students should not use the extra power that produces
eigenvectors. Students need to be reminded frequently that eigenvectors of A are null vectors of a
translate of A. That is why the instructions for Exercises 35–38 tell students to use the method of Example
4.
It is my experience that nearly all students need manual practice finding eigenvectors by the method
of Example 4, at least in this section if not also in Sections 5.2 and 5.3. However, [M] exercises do create
a burden if eigenvectors must be found manually. For this reason, the data files for the text include a
special command, nulbasis for each matrix program (MATLAB, Maple, etc.). The output of
nulbasis (A) is a matrix whose columns provide a basis for the null space of A, and these columns
are identical to the ones a student would find by row reducing the augmented matrix [ A 0]. With
nulbasis, student answers will be the same (up to multiples) as those in the text. I encourage my students
to use technology to speed up all numerical homework here, not just the [M] exercises,
37. [M] Let A be the given matrix. Use the MATLAB commands eig and nulbasis (or equivalent
commands). The command ev = eig(A) computes the three eigenvalues of A and stores them in a
vector ev. In this exercise, ev = (10, 15, 5). The eigenspace for the eigenvalue 10 is the null space
of A − 10 I . Use nulbasis to produce a basis for each null space:
ª −3º
nulbasis(A -ev(1)*eye(3))= «« 2 »» .
«¬ 1 »¼
°
Basis for the eigenspace for λ = 10 : ®
°
¯
ª −3 º
« 2»
« »
«¬ 1»¼
½
°
¾
°
¿
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
282
CHAPTER 5
• Eigenvalues and Eigenvectors
ª 2º
For the next eigenvalue, 15, compute nulbasis(A -ev(2)*eye(3))= «« 2 »» .
«¬ 1»¼
°
Basis for the eigenspace for λ = 15 : ®
°
¯
ª2º
«2»
« »
«¬ 1»¼
½
°
¾.
°
¿
ª −1 / 2 º
For the next eigenvalue, 5, compute nulbasis(A -ev(3)*eye(3))= «« − 1 / 2 »» .
«¬
1»¼
Scaling this vector by 2 to eliminate the fractions provides a basis for the eigenspace for
ª −1º ½
°
°
λ = 5 : ® «« −1»» ¾ .
° « 2» °
¯¬ ¼¿
38. [M] ev = eig(A)= (2, − 2, − 1, 1).
ª 2º
« 2»
For λ = 2 : nulbasis (A -ev(1)*eye(4))= « » . Basis:
«1 »
« »
¬1 ¼
°
°
®
°
°
¯
ª 2º ½
« 2» °
« » °¾
« 1» °
« »°
¬ 1¼ ¿
ª0 º
ª0º
°« »
«1 »
° 1
«
»
For λ = −2 : nulbasis(A -ev(2)*eye(4))=
. Basis: ® « »
«1 »
° «1 »
« »
°« »
¬0¼
¯ ¬0 ¼
½
°
°
¾
°
°
¿
ª1 º
ª1 º
°« »
«1 »
° 1
«
»
For λ = −1 : nulbasis(A -ev(3)*eye(4))=
. Basis: ® « »
«0»
° «0 »
« »
°« »
¬1 ¼
¯ ¬1 ¼
½
°
°
¾
°
°
¿
ª2º
ª 2º
°« »
« 2»
° 2
«
»
For λ = 1 : nulbasis(A -ev(4)*eye(4))=
. Basis: ® « »
«0»
° «0 »
« »
°« »
¬1 ¼
¯ ¬1 ¼
°
°
°
39. [M] For λ = −4 , basis: ®
°
°
°¯
ª 0º
« 1»
« »
« 2»
« »
« 0»
«¬ 1»¼
ª 0º ª2º
½
°« » « »
°
° « 0 » « 1»
°
°
°
¾ . For λ = 12, basis: ® « −1» , « 2 »
° « 1» « 0 »
°
°« » « »
°
« » « »
°¿
¯° ¬ 0 ¼ ¬ 1¼
½
°
°
¾
°
°
¿
½
°
°
°
¾ . For λ = −8 , basis:
°
°
¿°
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
°
°
°
®
°
°
°¯
ª 6º ª 0º
« 3» « 0 »
« » « »
« 3» , « −1»
« » « »
« 2» « 0»
«¬ 0 »¼ «¬ 1»¼
½
°
°
°
¾.
°
°
°¿
5.2
°
°
°
40. [M] For λ = −14 , basis: ®
°
°
°¯
ª 0 º ª −5 º
ª −1º ª10 º ½
°« » « »
« 0» « 6 » °
° « 1» « −1»
« » « » °°
°
« 1» , « 0 » ¾ . For λ = 42, basis: ® « −2 » , « −3»
« » « »°
° « 5» « 0 »
« 0» « 5 » °
°« » « »
«¬ 0 »¼ «¬ 1 »¼ °
°¯ «¬ 0 »¼ «¬ 5»¼
¿
• Solutions
283
½
°
°
°
¾.
°
°
¿°
Note: Since so many eigenvalues in text problems are small integers, it is easy for students to form a
habit of entering a value for in nulbasis (A - λI) based on a visual examination of the eigenvalues
are displayed. Using ev = eig(A) and:
produced by eig(A)when only a few decimal places for
nulbasis(A -ev(j)*eye(n)) help avoid problems caused by roundoff errors.
5.2
SOLUTIONS
Notes: Exercises 9–14 can be omitted, unless you want your students to have some facility with
determinants of 3 × 3 matrices. In later sections, the text will provide eigenvalues when they are needed
for matrices larger than 2 × 2. If you discussed partitioned matrices in Section 2.4, you might wish to
bring in Supplementary Exercises 12–14 in Chapter 5. (Also, see Exercise 14 of Section 2.4.)
Exercises 25 and 27 support the subsection on dynamical systems. The calculations in these exercises
and Example 5 prepare for the discussion in Section 5.6 about eigenvector decompositions.
ª2
1. A = «
¬7
7º
ª2
, A − λI = «
»
2¼
¬7
7 º ªλ
−
2»¼ «¬ 0
0 º ª2 − λ
=
λ »¼ «¬ 7
7 º
. The characteristic polynomial is
2 − λ »¼
det( A − λI ) = (2 − λ)2 − 7 2 = 4 − 4λ + λ 2 − 49 = λ 2 − 4λ − 45
In factored form, the characteristic equation is (λ − 9)(λ + 5) = 0, so the eigenvalues of A are 9 and
−5.
ª −4
2. A = «
¬6
−1º
ª −4 − λ
, A − λI = «
»
1¼
¬ 6
−1 º
. The characteristic polynomial is
1 − λ »¼
det( A − λ I ) = (−4 − λ )(1 − λ ) − 6 ⋅ (−1) = λ 2 + 3λ + 2
Since λ 2 + 3λ + 2 = (λ + 2)(λ + 1), the eigenvalues of A are −2 and −1 .
ª −4
3. A = «
¬6
2º
ª −4 − λ
, A − λI = «
»
7¼
¬ 6
2 º
. The characteristic polynomial is
7 − λ »¼
det( A − λI ) = ( −4 − λ)(7 − λ) − (2)(6) = λ 2 − 3λ − 40
Use the quadratic formula to solve the characteristic equation and find the eigenvalues:
λ=
−b ± b 2 − 4ac 3 ± 9 + 160 3 ± 13
=
=
= 8, −5
2a
2
2
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
284
CHAPTER 5
ª8
4. A = «
¬3
• Eigenvalues and Eigenvectors
2º
ª8 − λ
, A − λI = «
»
3¼
¬ 3
2 º
. The characteristic polynomial of A is
3 − λ »¼
det( A − λ I ) = (8 − λ )(3 − λ ) − (3)(2) = λ 2 − 11λ + 18 = (λ − 9)(λ − 2). The eigenvalues of A are
9 and 2.
ª8
5. A = «
¬4
4º
ª8 − λ
, A − λI = «
»
8¼
¬ 4
4 º
. The characteristic polynomial of A is
8 − λ »¼
det( A − λI ) = (8 − λ)(8 − λ) − (4)(4) = λ 2 − 16λ + 48 = (λ − 4)(λ − 12)
Thus, A has eigenvalues 4 and 12.
ª9
6. A = «
¬2
−2º
ª9 − λ
, A − λI = «
»
5¼
¬ 2
−2 º
. The characteristic polynomial is
5 − λ »¼
det( A − λ I ) = (9 − λ )(5 − λ ) − (−2)(2) = λ 2 − 14λ + 49 = (λ − 7)(λ − 7)
Thus A has only one eigenvalue, 7, with multiplicity 2.
ª 5
7. A = «
¬ −4
3º
ª5 − λ
, A − λI = «
»
4¼
¬ −4
3 º
. The characteristic polynomial is
4 − λ »¼
det( A − λI ) = (5 − λ)(4 − λ) − (3)(−4) = λ2 − 9λ + 32
Use the quadratic formula to solve det ( A − λI ) = 0 :
9 ± 81 − 4(32) 9 ± −47
=
2
2
These values are complex numbers, not real numbers, so A has no real eigenvalues. There is no
nonzero vector x in R 2 such that Ax = λ x, because a real vector Ax cannot equal a complex
multiple of x.
λ=
ª −4
8. A = «
¬ 2
3º
ª −4 − λ
, A − λI = «
»
1¼
¬ 2
3 º
. The characteristic polynomial is
1 − λ »¼
det( A − λ I ) = (−4 − λ )(1 − λ ) − (2)(3) = λ 2 + 3λ − 10 = (λ + 5)(λ − 2)
Thus, A has eigenvalues −5 and 2.
0
ª4 − λ
«
4−λ
9. det( A − λ I ) = det « 0
«¬ 1
0
characteristic polynomial is
−1 º
−1 »» . Using cofactor expansion down column 1, the
2 − λ »¼
ª(4 − λ )
det( A − λ I ) = (4 − λ ) ⋅ det «
¬ 0
−1 º
ª 0
+ 0 + 1 ⋅ det «
»
(2 − λ ) ¼
¬(4 − λ )
−1º
−1»¼
= (4 − λ )(λ 2 − 6λ + 8) + (4 − λ )
= (4 − λ )(λ 2 − 6λ + 9) = (4 − λ )(3 − λ )(3 − λ )
= −λ 3 + 10λ 2 − 33λ + 36
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.2
• Solutions
285
This matrix has eigenvalue 4 with multiplicity 1, and eigenvalue 3 with multiplicity 2.
1
1 º
ª3 − λ
«
5−λ
0 »» . Using cofactor expansion down the first column, the
10. det( A − λ I ) = det « 0
«¬ −2
0
7 − λ »¼
characteristic polynomial is
ª(5 − λ )
det( A − λ I ) = (3 − λ ) ⋅ det «
¬ 0
º
ª 1
+ 0 + (−2) det «
»
(7 − λ ) ¼
¬(5 − λ )
0
1º
0 »¼
= (5 − λ )(λ 2 − 10λ + 21) + 2(5 − λ )
= (5 − λ )(λ 2 − 10λ + 23)
= −λ 3 + 15λ 2 − 73λ + 115
11. The special arrangements of zeros in A makes a cofactor expansion along the first row highly
effective.
ª3 − λ
«
det( A − λ I ) = det « 2
«¬ 1
0
1− λ
0
0 º
ª1 − λ
4 »» = (3 − λ ) det «
¬ 0
4 − λ »¼
4 º
4 − λ »¼
= (3 − λ )(1 − λ)(4 − λ ) = (3 − λ )(λ 2 − 5λ + 4) = −λ 3 + 8λ 2 − 19λ + 12
If only the eigenvalues were required, there would be no need here to write the characteristic
polynomial in expanded form.
12. Make a cofactor expansion along the third row:
ª −1 − λ
«
det( A − λ I ) = det « 3
«¬ 0
2 º
ª −1 − λ
0 »» = −1 ⋅ det «
¬ 3
2 − λ »¼
0
1− λ
1
2º
ª −1 − λ
+ (2 − λ ) ⋅ det «
»
0¼
¬ 3
= 6 + (2 − λ )(−1 − λ )(1 − λ ) = −λ 3 + 2λ 2 + λ + 4
13. Make a cofactor expansion down the third column:
ª6 − λ
«
det( A − λ I ) = det « −2
«¬ 5
−2
0 º
ª6 − λ
0 »» = (3 − λ ) ⋅ det «
¬ −2
3 − λ »¼
9−λ
8
−2 º
9 − λ »¼
= (3 − λ )[(6 − λ )(9 − λ ) − ( −2)( −2)] = (3 − λ )(λ 2 − 15λ + 50)
= −λ 3 + 18λ 2 − 95λ + 150 or (3 − λ )(λ − 5)(λ − 10)
14. Make a cofactor expansion along the second column:
ª4 − λ
det( A − λ I ) = det «« −1
«¬ 0
0
−λ
2
−1 º
ª4 − λ
4 »» = (−λ ) ⋅ det «
¬ 0
3 − λ »¼
−1 º
ª4 − λ
− 2 ⋅ det «
»
3− λ¼
¬ −1
−1º
4 »¼
= (−λ ) ⋅ [(4 − λ )(3 − λ )] − 2[(4 − λ )4 − 1] = −λ 3 + 7λ 2 − 12λ + 8λ − 30
= −λ 3 + 7λ 2 − 4λ − 30
15. Use the fact that the determinant of a triangular matrix is the product of the diagonal entries:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0 º
1 − λ »¼
286
CHAPTER 5
• Eigenvalues and Eigenvectors
ª5 − λ
« 0
det( A − λ I ) = det «
« 0
«
¬ 0
5
0
2−λ
−3
0
0
3−λ
0
2 º
6 »»
= (5 − λ ) 2 (2 − λ )(3 − λ )
»
−2
»
5 − λ¼
The eigenvalues are 5, 5, 2, and 3.
16. The determinant of a triangular matrix is the product of its diagonal entries:
ª3 − λ
« 6
det( A − λ I ) = det «
« 0
«
¬ 2
0
0
2−λ
0
3
3
6−λ
3
º
0 »»
= (3 − λ )(2 − λ )(6 − λ )(−5 − λ )
0 »
»
−5 − λ ¼
0
The eigenvalues are 3, 2, 6, and −5.
17. The determinant of a triangular matrix is the product of its diagonal entries:
ª3 − λ
« −5
«
« 3
«
« 0
¬« −4
0
0
0
1− λ
8
0
0−λ
0
0
−7
1
2
9
1− λ
−2
0 º
0 »»
0 » = (3 − λ )2 (1 − λ )2 (−λ )
»
0 »
3 − λ ¼»
The eigenvalues are 3, 3, 1, 1, and 0.
18. Row reduce the augmented matrix for the equation ( A − 4 I )x = 0 :
ª0
«0
«
«0
«
¬0
2
−2
0
0
3
h
0
0
3
3
14
−2
0º ª0
0 »» ««0
0» «0
» «
0¼ ¬0
2
0
0
0
3
h+3
0
0
3
6
1
0
0º ª0
0 »» «« 0
0 » «0
» «
0 ¼ ¬0
2
0
0
0
3
h+3
0
0
0
0
1
0
0º
0 »»
0»
»
0¼
For a two-dimensional eigenspace, the system above needs two free variables. This happens if and
only if h = −3.
19. Since the equation det( A − λ I ) = (λ1 − λ )(λ 2 − λ ) " (λ n − λ ) holds for all λ , set λ = 0 and conclude
that det A = λ1λ 2 " λ n .
20. det( AT − λI ) = det( AT − λ I T )
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.2
= det( A − λ I )T
Transpose property
= det( A − λ I )
Theorem 3(c)
21. a.
b.
c.
d.
False. See Example 1.
False. See Theorem 3.
True. See Theorem 3.
False. See the solution of Example 4.
22. a.
b.
c.
d.
False. See the paragraph before Theorem 3.
False. See Theorem 3.
True. See the paragraph before Example 4.
False. See the warning after Theorem 4.
• Solutions
287
23. If A = QR, with Q invertible, and if A1 = RQ, then write A1 = Q −1QRQ = Q −1 AQ, which shows that
A1 is similar to A.
24. First, observe that if P is invertible, then Theorem 3(b) shows that
1 = det I = det( PP −1 ) = (det P )(det P −1 )
Use Theorem 3(b) again when A = PBP −1 ,
det A = det( PBP −1 ) = (det P )(det B )(det P −1 ) = (det B)(det P )(det P −1 ) = det B
25. Example 5 of Section 4.9 showed that Av1 = v1 , which means that v1 is an eigenvector of A
corresponding to the eigenvalue 1.
a. Since A is a 2 × 2 matrix, the eigenvalues are easy to find, and factoring the characteristic
polynomial is easy when one of the two factors is known.
ª.6 − λ
det «
¬ .4
.3 º
= (.6 − λ)(.7 − λ) − (.3)(.4) = λ2 − 1.3λ + .3 = (λ − 1)(λ − .3)
»
.7 − λ¼
The eigenvalues are 1 and .3. For the eigenvalue .3, solve ( A − .3I )x = 0 :
0 º ª.3 .3 0 º ª1 1 0º
.3
ª.6 − .3
=
« .4
.7 − .3 0 »¼ «¬.4 .4 0 »¼ «¬0 0 0»¼
¬
Here x1 + x2 = 0, with x2 free. The general solution is not needed. Set x2 = 1 to find an
ª −1º
eigenvector v 2 = « » . A suitable basis for R 2 is {v1 , v 2 }.
¬1¼
ª1/ 2 º ª 3/ 7 º
ª −1º
b. Write x 0 = v1 + cv 2 : « » = « » + c « » . By inspection, c is −1/14. (The value of c depends
¬1/ 2 ¼ ¬ 4/ 7 ¼
¬ 1¼
on how v 2 is scaled.)
c. For k = 1, 2, …, define x k = Ak x0 . Then x1 = A( v1 + cv 2 ) = Av1 + cAv 2 = v1 + c (.3) v 2 , because
v1 and v 2 are eigenvectors. Again
x 2 = Ax1 = A( v1 + c (.3) v 2 ) = Av1 + c (.3) Av 2 = v1 + c (.3)(.3) v 2 .
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
288
CHAPTER 5
• Eigenvalues and Eigenvectors
Continuing, the general pattern is x k = v1 + c(.3) k v 2 . As k increases, the second term tends to 0
and so x k tends to v1 .
b
º
ªa b º ªa
«
= U , and det A = (a)(d − ca −1b) = ad − bc. If a = 0,
26. If a ≠ 0, then A = «
»
−1 »
¬ c d ¼ ¬ 0 d − ca b ¼
ª0 b º ª c d º
«
= U (with one interchange), so det A = (−1)1 (cb) = 0 − bc = ad − bc.
then A = «
»
»
¬ c d ¼ ¬0 b ¼
27. a. Av1 = v1 , Av 2 = .5 v 2 , Av 3 = .2 v 3 .
b. The set {v1 , v 2 , v 3 } is linearly independent because the eigenvectors correspond to different
eigenvalues (Theorem 2). Since there are three vectors in the set, the set is a basis for 3. So there
exist unique constants such that x 0 = c1 v1 + c2 v 2 + c3 v 3 , and w T x 0 = c1wT v1 + c2 w T v 2 + c3 w T v 3 .
Since x0 and v1 are probability vectors and since the entries in v 2 and v 3 sum to 0, the above
equation shows that c1 = 1.
c. By (b), x 0 = c1 v1 + c2 v 2 + c3 v 3 . Using (a),
x k = Ak x0 = c1 Ak v1 + c2 Ak v 2 + c3 Ak v 3 = v1 + c2 (.5) k v 2 + c3 (.2) k v 3 → v1 as k → ∞
28. [M] Answers will vary, but should show that the eigenvectors of A are not the same as the
eigenvectors of AT , unless, of course, AT = A.
29. [M]
Answers will vary. The product of the eigenvalues of A should equal det A.
30. [M]
The characteristic polynomials and the eigenvalues for the various values of a are given in the
following table:
a
Characteristic Polynomial
31.8
−.4 − 2.6t + 4t 2 − t 3
Eigenvalues
3.1279, 1, − .1279
31.9
.8 − 3.8t + 4t 2 − t 3
2.7042, 1, .2958
32.0
2 − 5t + 4t 2 − t 3
2, 1, 1
32.1
3.2 − 6.2t + 4t 2 − t 3
1.5 ± .9747i, 1
32.2
4.4 − 7.4t + 4t 2 − t 3
1.5 ± 1.4663i, 1
The graphs of the characteristic polynomials are:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.3
• Solu
utions
289
Notes: An appendix in Section 5.3 of thhe Study Guide gives an example of factoring a cubicc polynomial
with integer coefficients, in case you w
want your students to find integer eigenvalues of sim
mple 3 × 3 or
perhaps 4 × 4 matrices.
The MATLAB box for Section 5.3 introduces the command poly (A), which lists thee coefficients
matrix A, and it gives MATLAB code that will producce a graph of
of the characteristic polynomial of the m
the characteristic polynomial. (This is nneeded for Exercise 30.) The Maple and Mathematicaa appendices
have corresponding information. The apppendices for the TI calculators contain only the com
mmands that
list the coefficients of the characteristic ppolynomial.
5.3
SOLUTIONS
ª5 7 º
ª 2 0º
,D = «
, A = PDP −1 , and A4 = PD 4 P −1 .
1. P = «
»
»
¬ 2 3¼
¬0 1¼
ª 3 −7 º 4 ª16 0º
ª 5 7 º ª16
P −1 = «
,D =«
, and A4 = «
»
»
»«
5¼
¬ −2
¬ 0 1¼
¬ 2 3¼ ¬ 0
We compute
0º ª 3
1 »¼ «¬ −2
−7 º ª 226
=
5»¼ «¬ 90
ª 1 2º
ª1 0 º
,D = «
, A = PDP −1 , and A4 = PD 4 P −1 . We compute
2. P = «
»
»
¬ 2 3¼
¬0 3¼
ª −3 2 º 4 ª1 0 º
ª 1 2º ª1 0 º ª −3 2 º ª 321
P −1 = «
,D = «
, annd A4 = «
»
»
»«
»«
»=«
¬ 2 −1¼
¬0 81¼
¬ 2 3¼ ¬0 81¼ ¬ 2 −1¼ ¬ 480
ª1
3. Ak = PD k P −1 = «
¬2
0 º ª« a k
«
1 »¼ «¬ 0
0 º» ª 1
»«
b k »¼ ¬ −2
ª3
4. Ak = PD k P −1 = «
¬2
−2 º ª« (−3)k
«
−1»¼ «¬ 0
0 º ª«
ak
=«
1»¼ «¬ 2 a k − 2b k
º
» ª −1
»«
( −2) k »¼ ¬ −2
0
−525º
−209»¼
−160º
−239»¼
0 º»
».
bk »¼
2 º ª −3 ⋅ ( −3)k + 4 ⋅ (−2)k
=«
3»¼ « −2 ⋅ (−3) k + 2 ⋅ (−2) k
¬
6 ⋅ (−3)k − 6 ⋅ (−2)k º
».
k
k
4 ⋅ (−3) − 3 ⋅ (−2) »¼
5. By the Diagonalization Theorem, eiigenvectors form the columns of the left factor, and th
hey
correspond respectively to the eigennvalues on the diagonal of the middle factor.
Copyright © 2012 Pearrson Education, Inc. Publishing as Addison-Wesley.
290
CHAPTER 5
• Eigenvalues and Eigenvectors
ª −1º
ª 1º ª 0º
«
»
= 2 : « 1 » ; = 3 : «« −1»» , «« −1»»
«¬ −1»¼
«¬ 0 »¼ «¬ 1»¼
6. As in Exercise 5, inspection of the factorization gives:
ª0 º
ª 3º ª −1º
«
»
= 4 : « 1» ; = 3 : ««0 »» , «« −3»»
«¬0 »¼
«¬ 1»¼ «¬ 0 »¼
7. Since A is triangular, its eigenvalues are obviously ±1.
ª0
For λ = 1: A − 1I = «
¬6
0º
. The equation ( A − 1I )x = 0 amounts to 6 x1 − 2 x2 = 0, so x1 = (1/ 3) x2
−2»¼
ª1/3º
ª1º
with x2 free. The general solution is x2 « » , and a nice basis vector for the eigenspace is v1 = « » .
¬ 1¼
¬3 ¼
ª2
For λ = −1: A + 1I = «
¬6
0º
. The equation ( A + 1I )x = 0 amounts to 2 x1 = 0, so x1 = 0 with x2
0»¼
ª0 º
ª0 º
free. The general solution is x2 « » , and a basis vector for the eigenspace is v 2 = « » .
¬1 ¼
¬1 ¼
ª1
v2 º¼» = «
¬3
in D correspond to v1 and v 2 respectively.
From v1 and v 2 construct P = ª¬« v1
0º
ª1
. Then set D = «
»
1¼
¬0
0º
, where the eigenvalues
−1»¼
8. Since A is triangular, its only eigenvalue is obviously 3.
ª0
For λ = 3: A − 3I = «
¬0
2º
. The equation ( A − 3I )x = 0 amounts to x2 = 0, so x2 = 0 with x1 free.
0 »¼
ª1 º
The general solution is x1 « » . Since we cannot generate an eigenvector basis for 2, A is not
¬0 ¼
diagonalizable.
9. To find the eigenvalues of A, compute its characteristic polynomial:
−1 º
ª2 −
= (2 − )(4 − ) − ( −1)(1) =
det( A − I ) = det «
4 − »¼
¬ 1
Thus the only eigenvalue of A is 3.
2
− 6 + 9 = ( − 3) 2
ª −1
For λ = 3: A − 4 I = «
¬ 1
−1º
. The equation ( A − 3I )x = 0 amounts to x1 + x2 = 0, so x1 = − x2 with
1»¼
ª−1º
x2 free. The general solution is x2 « » . Since we cannot generate an eigenvector basis for 2, A is
¬ 1¼
not diagonalizable.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.3
• Solutions
291
10. To find the eigenvalues of A, compute its characteristic polynomial:
3 º
ª1 −
= (1 − )(2 − ) − (3)(4) =
det( A − I ) = det «
2 − »¼
¬ 4
Thus the eigenvalues of A are 5 and −2 .
2
− 3 − 10 = ( − 5)( + 2)
ª3
For λ = −2: A + 2 I = «
¬4
3º
. The equation ( A + 1I )x = 0 amounts to x1 + x2 = 0, so x1 = − x2 with
4»¼
ª −1º
ª −1º
x2 free. The general solution is x2 « » , and a nice basis vector for the eigenspace is v1 = « » .
¬ 1¼
¬ 1¼
ª −4
For λ = 5: A − 5I = «
¬ 4
3º
. The equation ( A − 3I )x = 0 amounts to −4 x1 + 3 x2 = 0, so
−3»¼
ª3 / 4 º
x1 = (3 / 4) x2 with x2 free. The general solution is x2 «
» , and a basis vector for the eigenspace is
¬ 1 ¼
ª 3º
v2 = « » .
¬ 4¼
ª −1
v2 º»¼ = «
¬ 1
in D correspond to v1 and v 2 respectively.
From v1 and v 2 construct P = ª«¬ v1
3º
ª −2
. Then set D = «
»
4¼
¬ 0
0º
, where the eigenvalues
5»¼
11. The eigenvalues of A are given to be -1, and 5.
ª1
For λ = -1: A + I = «« 2
¬« 3
1
2
3
1º
2 »» , and row reducing [ A + I
3¼»
ª1
0] yields ««0
¬«0
1
1
0
0
0
0
0º
0 »» . The general
0 ¼»
ª −1º ª −1º ½
ª −1º
ª −1º
°
°
«
»
«
»
solution is x2 « 1 » + x3 « 0 » , and a nice basis for the eigenspace is { v1 , v 2 } = ® «« 1 »» , «« 0 »» ¾ .
°« 0 » « 1 » °
«¬ 0 »¼
«¬ 1»¼
¯¬ ¼ ¬ ¼ ¿
1
1º
ª −5
ª 1 0 −1 / 3
«
»
For λ = 5: A − 5I = « 2 −4
2 » , and row reducing [ A − 5I 0] yields ««0 1 −2 / 3
«¬0 0
«¬ 3
3 −3»¼
0
ª 1 / 3º
ª1 º
«
»
general solution is x3 « 2 / 3» , and a nice basis vector for the eigenspace is v 3 = «« 2 »» .
«¬ 1»¼
«¬ 3 »¼
From v1 , v 2 and v 3 construct P = ª«¬ v1
v2
ª −1
v3 º»¼ = «« 1
¬« 0
−1
0
1
1º
2 »» . Then set D =
3 ¼»
ª −1
«0
«
¬« 0
0
−1
0
0º
0 »» . The
0 »¼
0º
0 »» ,
5 ¼»
where the eigenvalues in D correspond to v1 , v 2 and v 3 respectively. Note that if changing the
order of the vectors in P also changes the order of the diagonal elements in D, and results in the
answer given in the text.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
292
CHAPTER 5
• Eigenvalues and Eigenvectors
12. The eigenvalues of A are given to be 2 and 5.
ª1 1 1º
ª 1 1 1 0º
«
»
For λ = 2: A − 2 I = «1 1 1» , and row reducing [ A − 2 I 0] yields ««0 0 0 0 »» . The general
«¬0 0 0 0 »¼
«¬1 1 1»¼
ª −1º ª −1º ½
ª −1º
ª −1º
°
°
«
»
«
»
solution is x2 « 1 » + x3 « 0 » , and a nice basis for the eigenspace is {v1 , v 2 } = ® «« 1 »» , «« 0 »» ¾ . .
°« 0 » « 1 » °
«¬ 0 »¼
«¬ 1»¼
¯¬ ¼ ¬ ¼ ¿
ª −2
For λ = 5: A − 5 I = «« 1
«¬ 1
ª1º
general solution is x3 ««1»» ,
«¬1»¼
0
−1
1
0
−1
0
−1 1º
ª2
»
From v1 , v 2 and v 3 construct P = ª¬« v1 v 2
0 1» . Then set D = «« 0
«¬ 0
1 1»¼
the eigenvalues in D correspond to v1 , v 2 and v 3 respectively.
0
ª1
−2
0] yields ««0
«¬0
1
ª1º
and a basis for the eigenspace is v 3 = ««1»» .
«¬1»¼
1
1º
1 »» , and row reducing [ A − 5I
−2 »¼
ª −1
v3 º¼» = «« 1
«¬ 0
2
0
0º
0 »» . The
0 »¼
0º
0 »» , where
5 »¼
13. The eigenvalues of A are given to be 5 and 1.
ª −3
For λ = 5: A − 5I = «« 1
¬« −1
2
−2
−2
−1º
−1»» , and row reducing [ A − 5I
−3¼»
ª1
0] yields ««0
¬«0
0
1
1
0
1
0
0º
0 »» . The
0 ¼»
ª −1º
ª −1º
«
»
general solution is x3 « −1» , and a basis for the eigenspace is v1 = «« −1»» .
«¬ 1»¼
«¬ 1»¼
2 −1º
ª 1
ª1 2
«
»
For λ = 1: A − 1I = « 1
2 −1» , and row reducing [ A − I 0] yields ««0 0
«¬0 0
«¬ −1 −2
1»¼
ª −2 º
ª1 º
°
«
»
«
»
general solution is x2 « 1» + x3 «0 » , and a basis for the eigenspace is {v 2 , v 3 } = ®
°
«¬ 0 »¼
«¬1 »¼
¯
From v1 , v 2 and v 3 construct P = ª«¬ v1
v2
ª −1
v3 º»¼ = «« −1
¬« 1
−2
1
0
−1
0º
0 0 »» . The
0 0 »¼
ª −2 º ª 1 º ½
« 1» , «0 » ° .
« » « »¾
«¬ 0 »¼ «¬1 »¼ °¿
1º
ª5
»
0 » . Then set D = ««0
1¼»
¬«0
where the eigenvalues in D correspond to v1 , v 2 and v 3 respectively.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0
1
0
0º
0 »» ,
1 ¼»
5.3
• Solutions
1
0
0
0
1
0
293
14. The eigenvalues of A are given to be 2 and 3.
ª0
For λ = 2: A − 2 I = «« 1
«¬0
0
1
0
−2 º
2 »» , and row reducing [ A − 2 I
1»¼
ª1
0] yields ««0
«¬0
0º
0 »» . The
0 »¼
ª −1º
ª −1º
«
»
general solution is x2 « 1 » , and a basis for the eigenspace is v1 = «« 1 »» .
«¬ 0 »¼
«¬ 0 »¼
ª −1 0 −2 º
ª1 0 2
«
»
For λ = 3: A − 3I = « 1 0
2 » , and row reducing [ A − 3I 0] yields ««0 0 0
«¬0 0 0
«¬ 0 0
0 »¼
ª0º
ª −2 º
°
«
»
«
»
general solution is x2 «1 » + x3 « 0 » , and a nice basis for the eigenspace is {v 2 , v 3 } = ®
°
«¬0 »¼
«¬ 1»¼
¯
ª −1
v3 º¼» = «« 1
«¬ 0
−2 º
ª2
»
From v1 , v 2 and v 3 construct P = ª¬« v1 v 2
1
0 » . Then set D = «« 0
«¬ 0
0
1»¼
where the eigenvalues in D correspond to v1 , v 2 and v 3 respectively.
0
0º
0 »» . The
0 »¼
ª 0 º ª −2 º
« 1» , « 0 »
« » « »
«¬0 »¼ «¬ 1 »¼
0
3
0
½
°
¾.
°
¿
0º
0 »» ,
3 »¼
15. The eigenvalues of A are given to be 0 and 1.
ª 0
For λ = 0: A − 0 I = «« 1
¬« −1
−1
2
−1
−1º
1»» , and row reducing [ A − 0 I
0 ¼»
ª1
0] yields ««0
¬«0
0
−1
1
0
1
0
0º
0 »» . The
0 ¼»
ª 1º
ª 1º
«
»
general solution is x3 « −1» , and a basis for the eigenspace is v1 = «« −1»»
«¬ 1»¼
«¬ 1»¼
ª −1
For λ = 1: A − I = «« 1
«¬ −1
−1
−1º
ª1 1 1 0 º
»
1
1» , and row reducing [ A − I 0] yields ««0 0 0 0 »» . The
«¬0 0 0 0 »¼
−1 −1»¼
ª −1º ª −1º ½
ª −1º
ª −1º
°
°
«
»
«
»
general solution is x2 « 1 » + x3 « 0 » , and a basis for the eigenspace is {v 2 , v 3 } = ® «« 1»» , «« 0 »» ¾ .
° « 0» « 1 » °
«¬ 0 »¼
«¬ 1»¼
¯¬ ¼ ¬ ¼¿
From v1 , v 2 and v 3 construct P = ª«¬ v1
v2
ª 1
v3 º»¼ = «« −1
¬« 1
−1
1
0
−1º
ª0
»
0 » . Then set D = ««0
1¼»
¬«0
0
1
0
0º
0 »» ,
1 ¼»
where the eigenvalues in D correspond to v1 , v 2 and v 3 respectively. Note that the answer for P
given in the text has the first column scaled by −1 , which is also a correct answer, since any nonzero
multiple of an eiegnvector is an eigenvector.
16. The only eigenvalue of A given is 0.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
294
CHAPTER 5
• Eigenvalues and Eigenvectors
ª1
For λ = 0: A − 0 I = «« 2
«¬ 1
2
5
3
−3º
−2 »» , and row reducing [ A − 0 I
1»¼
ª1
0] yields ««0
«¬0
0
−11
1
0
4
0
0º
0 »» . The
0 »¼
ª 11º
ª 11º
«
»
general solution is x2 « −4 » , and a basis for the eigenspace is v1 = «« −4 »» .
«¬ 1»¼
«¬ 1»¼
Since = 0 has only a one-dimensional eigenspace, we can find at most one linearly independent
eigenvector for A, so A is not diagonalizable over the real numbers. The remaining eigenvalues are
complex, and this situation is dealt with in Section 5.
17. Since A is triangular, its eigenvalue is obviously 2.
ª0 0 0 º
For λ = 2: A − 2 I = «« 2 0 0 »» , and row reducing [ A − 2 I
«¬ 2 2 0 »¼
ª1
0] yields ««0
«¬0
0
0
1
0
0
0
0º
0 »» . The general
0 »¼
ª0º
ª0º
«
»
solution is x3 « 0 » , and a basis for the eigenspace is v1 = ««0 »» .
«¬1 »¼
«¬1 »¼
Since = 2 has only a one-dimensional eigenspace, we can find at most one linearly independent
eigenvector for A, so A is not diagonalizable.
18. The eigenvalues of A are given to be -2, -1 and 0.
ª4
For λ = -2: A + 2 I = «« 3
¬« 2
−2
−1
−2
−2 º
−2 »» , and row reducing [ A + 2 I
0 ¼»
ª1
0] yields ««0
¬«0
0
−1
1
0
−1
0
0º
0 »» . The
0 ¼»
ª1º
ª1º
«
»
general solution is x3 «1» , and a basis for the eigenspace is v1 = ««1»» .
«¬1»¼
«¬1»¼
ª 3 −2
For λ = -1: A + I = «« 3 −2
«¬ 2 −2
ª 1 º
general solution is x3 ««1 / 2 »» ,
«¬ 1 »¼
−2 º
−2 »» , and row reducing [ A + I
−1»¼
ª1
0] yields ««0
«¬0
ª 2º
and a basis for the eigenspace is v 2 = ««1 »» .
«¬ 2»¼
0
−1
1
0
−1 / 2
0
ª 2 −2 −2 º
ª 1 −1
«
»
For λ = 0: A − 0 I = « 3 −3 −2 » , and row reducing [ A − 0 I 0] yields ««0
0
«¬0
«¬ 2 −2 −2 »¼
0
ª1 º
ª 1º
«
»
general solution is x2 «1 » , and a nice basis vector for the eigenspace is v 3 = «« 1»» .
«¬0 »¼
«¬0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0
1
0
0º
0 »» . The
0 »¼
0º
0 »» . The
0 »¼
5.3
• Solutions
1º
ª −2
»
From v1 , v 2 and v 3 construct P = ª¬« v1 v2
1 1» . Then set D = «« 0
«¬ 0
2 0 »¼
the eigenvalues in D correspond to v1 , v 2 and v 3 respectively.
ª1
v3 º¼» = ««1
«¬1
2
295
0º
0 »» , where
0 »¼
0
−1
0
19. Since A is triangular, its eigenvalues are 2, 3, and 5.
ª3
«0
For = 2: A − 2 I = «
«0
«
«¬0
−3
0
ª2
«0
For λ = 3: A − 3I = «
«0
«
¬« 0
−3
0
0
0
0
1
−1
0
9º
ª1
«0
»
1 1 −2 »
, and row reducing [ A − 2 I 0] yields «
«0
0 0
0»
«
»
0 0
0 »¼
«¬ 0
ª −1º
ª −1º
« −1»
« 2»
«
»
The general solution is x3
+ x « » , and a nice basis for the eigenspace is
« 1» 4 « 0 »
« »
« »
«¬ 0 »¼
«¬ 1»¼
ª −1º ª −1º ½
°« » « »°
2 °
° −1
{v1 , v 2 } = ® « » , « » ¾.
° « 1» « 0 » °
° ««¬ 0 »»¼ ««¬ 1»»¼ °
¯
¿
ª1
«0
«
«0
«
¬«0
−3/ 2
0
0
0
0
1
0
0
0
0
1
0
0
1
0
0
1
1
0
0
1
−2
0
0
0º
0 »»
.
0»
»
0 »¼
9º
−2 »»
, and row reducing [ A − 3I 0] yields
0»
»
−1¼»
0º
ª 3/ 2 º
« 1»
»
0»
. The general solution is x2 « » , and a nice basis for the eigenspace is
« 0»
0»
« »
»
0 ¼»
¬« 0 »¼
ª3º
«2»
v3 = « » .
«0 »
« »
«¬ 0 »¼
ª0
«0
For λ = 5: A − 5 I = «
«0
«
¬«0
−3
−2
0
0
0
1
−3
0
9º
ª0
»
«0
−2 »
, and row reducing [ A − 5 I 0] yields «
«0
0»
»
«
−3¼»
¬«0
ª1 º
ª1 º
«0 »
«0 »
The general solution is x1 « » , and a basis for the eigenspace is v 4 = « » .
«0 »
«0 »
« »
« »
«¬0 »¼
«¬ 0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1
0
0
0
0
1
0
0
0
0
1
0
0º
0 »»
.
0»
»
0 ¼»
296
CHAPTER 5
• Eigenvalues and Eigenvectors
ª −1
« −1
From v1 , v 2 , v 3 and v 4 construct P = ª¬ v1 v 2 v3 v 4 º¼ = «
« 1
«
¬« 0
ª2
«0
D=«
«0
«
«¬ 0
−1
2
0
1
3
2
0
0
1º
0 »»
. Then set
0»
»
0 ¼»
0º
0 »»
, where the eigenvalues in D correspond to v1 , v 2 , v 3 and v 4 respectively.
0»
»
5 »¼
Note that this answer differs from the text. There, P = [ v 4 v 3 v1 v 2 ] and the entries in D are
rearranged to match the new order of the eigenvectors. According to the Diagonalization Theorem,
both answers are correct.
0
2
0
0
0
0
3
0
20. Since A is triangular, its eigenvalues are 2 and 3.
ª1
«0
For λ = 2: A − 2 I = «
«0
«
¬1
0
0
0
0
0
0
0
0
0º
ª1
«0
»
0»
, and row reducing [ A − 2 I 0] yields «
«0
0»
«
»
1¼
¬0
ª0 º
ª0º
°
«1 »
«0»
°
general solution is x2 « » + x3 « » , and a basis for the eigenspace is {v1 , v 2 } = ®
«0 »
«1 »
°
« »
« »
°
0
0
¬ ¼
¬ ¼
¯
0
0
0
0
0
0
0
0
0
1
0
0
ª0º ª0º
«1 » « 0 »
« »,« »
« 0 » «1 »
« » « »
¬0¼ ¬0¼
0º
0 »»
. The
0»
»
0¼
½
°
°
¾.
°
°
¿
0º
ª1 0 0 0 0 º
«0 1 0 0 0 »
»
0»
».
, and row reducing [ A − 3I 0] yields «
«0 0 1 0 0 »
0»
»
«
»
0¼
¬0 0 0 0 0 ¼
ª0 º
ª0º
«0»
«0 »
The general solution is x4 « » , and a basis for the eigenspace is v 3 = « » .
«0»
«0 »
« »
« »
¬1 ¼
¬1 ¼
Since = 3 has only a one-dimensional eigenspace, we can find at most three linearly independent
eigenvectors for A, so A is not diagonalizable.
ª0
«0
For λ = 3: A − 3I = «
«0
«
¬1
0
−1
0
0
0
0
−1
0
21. a. False. The symbol D does not automatically denote a diagonal matrix.
b. True. See the remark after the statement of the Diagonalization Theorem.
c. False. The 3 × 3 matrix in Example 4 has 3 eigenvalues, counting multiplicities, but it is not
diagonalizable.
d. False. Invertibility depends on 0 not being an eigenvalue. (See the Invertible Matrix Theorem.)
A diagonalizable matrix may or may not have 0 as an eigenvalue. See Examples 3 and 5 for both
possibilities.
22. a. False. The n eigenvectors must be linearly independent. See the Diagonalization Theorem.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.3
• Solutions
297
b. False. The matrix in Example 3 is diagonalizable, but it has only 2 distinct eigenvalues. (The
statement given is the converse of Theorem 6.)
c. True. This follows from AP = PD and formulas (1) and (2) in the proof of the Diagonalization
Theorem.
d. False. See Example 4. The matrix there is invertible because 0 is not an eigenvalue, but the matrix
is not diagonalizable.
23. A is diagonalizable because you know that five linearly independent eigenvectors exist: three in the
three-dimensional eigenspace and two in the two-dimensional eigenspace. Theorem 7 guarantees that
the set of all five eigenvectors is linearly independent.
24. No, by Theorem 7(b). Here is an explanation that does not appeal to Theorem 7: Let v1 and v 2 be
eigenvectors that span the two one-dimensional eigenspaces. If v is any other eigenvector, then it
belongs to one of the eigenspaces and hence is a multiple of either v1 or v 2 . So there cannot exist
three linearly independent eigenvectors. By the Diagonalization Theorem, A cannot be
diagonalizable.
25. Let {v1} be a basis for the one-dimensional eigenspace, let v 2 and v 3 form a basis for the twodimensional eigenspace, and let v 4 be any eigenvector in the remaining eigenspace. By Theorem 7,
{v1 , v 2 , v 3 , v 4 } is linearly independent. Since A is 4 × 4, the Diagonalization Theorem shows that
A is diagonalizable.
26. Yes, if the third eigenspace is only one-dimensional. In this case, the sum of the dimensions of the
eigenspaces will be six, whereas the matrix is 7 × 7. See Theorem 7(b). An argument similar to that
for Exercise 24 can also be given.
27. If A is diagonalizable, then A = PDP −1 for some invertible P and diagonal D. Since A is invertible, 0
is not an eigenvalue of A. So the diagonal entries in D (which are eigenvalues of A) are not zero, and
D is invertible. By the theorem on the inverse of a product,
A−1 = ( PDP −1 ) −1 = ( P −1 ) −1 D −1 P −1 = PD −1 P −1
Since D −1 is obviously diagonal, A−1 is diagonalizable.
28. If A has n linearly independent eigenvectors, then by the Diagonalization Theorem, A = PDP −1 for
some invertible P and diagonal D. Using properties of transposes,
AT = ( PDP −1 )T = ( P −1 )T DT PT
= ( PT )−1 DPT = QDQ −1
where Q = ( PT )−1. Thus AT is diagonalizable. By the Diagonalization Theorem, the columns of Q
are n linearly independent eigenvectors of AT .
29. The diagonal entries in D1 are reversed from those in D. So interchange the (eigenvector) columns of
P to make them correspond properly to the eigenvalues in D1 . In this case,
ª 1
P1 = «
¬ −2
1º
ª3
and D1 = «
»
−1¼
¬0
0º
5»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
298
CHAPTER 5
• Eigenvalues and Eigenvectors
Although the first column of P must be an eigenvector corresponding to the eigenvalue 3, there is
1º
ª 1º
ª−3º
ª −3
nothing to prevent us from selecting some multiple of « » , say « » , and letting P2 = «
».
¬ −2¼
¬ 6¼
¬ 6 −1¼
We now have three different factorizations or “diagonalizations” of A:
A = PDP −1 = P1 D1P1−1 = P2 D1 P2−1
30. A nonzero multiple of an eigenvector is another eigenvector. To produce P2 , simply multiply one or
both columns of P by a nonzero scalar other than 1.
31. For a 2 × 2 matrix A to be invertible, its eigenvalues must be nonzero. A first attempt at a
ª 2 3º
construction might be something such as «
» , whose eigenvalues are 2 and 4. Unfortunately, a
¬0 4¼
2 × 2 matrix with two distinct eigenvalues is diagonalizable (Theorem 6). So, adjust the construction
ªa b º
ª 2 3º
, which works. In fact, any matrix of the form «
to «
» has the desired properties when a
»
¬0 a¼
¬0 2¼
and b are nonzero. The eigenspace for the eigenvalue a is one-dimensional, as a simple calculation
shows, and there is no other eigenvalue to produce a second eigenvector.
32. Any 2 × 2 matrix with two distinct eigenvalues is diagonalizable, by Theorem 6. If one of those
ªa b º
eigenvalues is zero, then the matrix will not be invertible. Any matrix of the form «
» has the
¬0 0¼
desired properties when a and b are nonzero. The number a must be nonzero to make the matrix
ª0 0º
diagonalizable; b must be nonzero to make the matrix not diagonal. Other solutions are «
»
¬a b ¼
ª0
and «
¬0
ª 9
« −56
33. A = «
« −14
«
¬ 42
aº
.
b »¼
−4
−2
32
−28
−14
−33
6
21
−4 º
44 »»
, ev = eig(A)=(13,-12,-12, 13),
−14 »
»
−45¼
ª −0.5000
«
0
nulbasis(A-ev(1)*eye(4)) = «
« 1.0000
«
0
¬
0.3333º
−1.3333»»
.
0»
»
1.0000 ¼
ª −1º ª 1 º ½
°« » « » °
−4 °
° 0
A basis for the eigenspace of λ = 13 is ® « » , « » ¾.
°« 2 » « 0 » °
° «¬ 0 »¼ «¬ 3 »¼ °
¯
¿
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.3
ª0.2857
« 1.0000
nulbasis(A-ev(2)*eye(4)) «
« 1.0000
«
0
¬
• Solutions
0º
−1.0000»»
,
0»
»
1.0000 ¼
ª 2º ª 0º ½
1 2 0º
ª −1
°« » « » °
«
0 −4 7 1»»
° 7 −1 °
A basis for the eigenspace of λ = −12 is ® « » , « » ¾. Thus we construct P = «
,
« 2
0 7 0»
° «7 » « 0» °
«
»
°« » « » °
3 0 1¼
¬ 0
¯ ¬ 0 ¼ ¬ 1¼ ¿
0
0º
ª13 0
« 0 13
0
0 »»
and D = «
. Notice that the anwer in the text lists the eigenvector in an different
« 0 0 −12
0»
«
»
0 −12 ¼
¬0 0
order in P, and hence the eigenvalues are listed in a different order in D. Both answers are correct.
ª 4
« −7
«
34. A = « 5
«
« −2
«¬ −3
−9
−7
8
−9
10
0
5
7
−5
3
−13
7
−7
0
10
2º
14 »»
−10 » , ev = eig(A)=(5,-2,-2,5,5),
»
4»
11»¼
ª 2.0000
« −1.0000
«
nulbasis(A-ev(1)*eye(5)) = « 1.0000
«
0
«
«¬
0
−1.0000
1.0000
0
1.0000
0
2.0000 º
0 »»
0» ,
»
0»
1.0000 »¼
ª 2 º ª −1º ª 2 º ½
°« » « » « » °
° « −1» « 1» « 0 » °
°
°
A basis for the eigenspace of λ = 5 is ® « 1» , « 0 » , « 0 » ¾ .
° « 0 » « 1» « 0 » °
°« » « » « » °
« » « » « »
¯° ¬ 0 ¼ ¬ 0 ¼ ¬ 1¼ ¿°
ª −0.4000
« 1.4000
«
nulbasis(A-ev(2)*eye(5)) = « −1.0000
«
« 1.0000
«¬ 0
0.6000 º
1.4000 »»
−1.0000 » ,
»
0 »
1.0000 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
299
300
CHAPTER 5
• Eigenvalues and Eigenvectors
ª −2 º ª 3 º ½
°« » « » °
°« 7 » « 7 » °
°
°
A basis for the eigenspace of λ = −2 is ® « −5» , « −5» ¾ .
°« 5 » « 0 » °
°« » « » °
« » « »
¯° ¬ 0 ¼ ¬ 5 ¼ ¿°
ª 2
« −1
«
Thus we construct P = « 1
«
« 0
«¬ 0
ª 13
« 6
«
35. A = « 6
«
« 6
¬« −6
−12
9
−15
−5
−12
9
−5
−15
6
−12
12
9
12
−8
−6
−1
2
−2
1
0
7
0
0
−5
1
0
0
1
5
0
3º
ª5
»
«0
7»
«
−5» and D = «0
»
«
0»
«0
»
«¬0
5¼
0
0
0
5
0
0
0
5
0
0
0
0
0
−2
0
0º
0 »»
0» .
»
0»
−2 »¼
9º
9 »»
9 » , ev = eig(A) =(7,− 14,-14,7,7),
»
9»
−2 ¼»
ª 2.0000
« 1.0000
«
nulbasis(A-ev(1)*eye(5)) = «
0
«
0
«
«¬
0
1.0000
0
1.0000
1.0000
0
−1.5000 º
0 »»
0» ,
»
0»
1.0000 »¼
ª 2 º ª 1º ª −3º ½
°« » « » « » °
° « 1» « 0 » « 0 » °
°
°
A basis for the eigenspace of λ = 7 is ® « 0 » , « 1» , « 0 » ¾ .
° « 0 » « 1» « 0 » °
°« » « » « » °
°¯ «¬ 0 »¼ «¬0 »¼ «¬ 2 »¼ °¿
ª1
«1
«
nulbasis(A-ev(2)*eye(5)) = « 0
«
«1
«¬ 0
0º
0 »»
−1» ,
»
0»
1»¼
ª 1º ª 0 º ½
°« » « » °
° « 1» « 0 » °
°
°
A basis for the eigenspace of λ = −14 is ® « 0 » , « −1» ¾ .
° « 1» « 0 » °
°« » « » °
°¯ «¬ 0 »¼ «¬ 1»¼ °¿
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.3
ª2
«1
«
Thus we construct P = « 0
«
«0
«¬ 0
ª 24
«
« 72
36. A = « 0
«
« 72
«¬ 0
−6
2
6
51
9
−63
15
−99
63
15
63
9
21
−63
−63
1
−3
1
0
0
1
1
0
0
1
0
0
2
1
0
0º
ª7
»
«0
0»
«
−1» , and D = « 0
»
«
0»
«0
»
«¬ 0
1¼
0
0
0
7
0
0
0
7
0
0
0
0
0
−14
0
0 º
0 »»
0 ».
»
0 »
−14 »¼
2º
9 »»
63» , ev = eig(A)=(24,-48,36,-48,36),
»
9»
−27 »¼
ª 1º
« 1»
« »
nulbasis(A-ev(1)*eye(5)) = «0 » ,
« »
« 1»
«¬0 »¼
ª 1º
« 1»
« »
A basis for the eigenspace of λ = 24 is «0 » .
« »
« 1»
«¬0 »¼
nulbasis(A-ev(2)*eye(5)),
ª0º ª 0º ½
°« » « » °
° « 1» « 0 » °
°
°
A basis for the eigenspace of λ = −48 is ® « 0 » , « −1» ¾ .
° « 1» « 0 » °
°« » « » °
°¯ «¬ 0 »¼ «¬ 1»¼ °¿
ª1.0000
« 0.0000
«
nulbasis(A-ev(3)*eye(5)) = «3.0000
«
«1.0000
«¬ 0
−0.3333º
1.0000 »»
0.0000 » ,
»
0 »
1.0000 »¼
ª1 º ª −1º ½
°« » « » °
° «0» « 3 » °
°
°
A basis for the eigenspace of λ = 36 is ® « 3» , « 0 » ¾ .
° «1 » « 0 » °
°« » « » °
°¯ «¬ 0 »¼ «¬ 3 »¼ °¿
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
• Solutions
301
302
CHAPTER 5
• Eigenvalues and Eigenvectors
−1º
0
0
0
0º
ª 24
»
«
0
0
0 »»
1
0 0
3»
« 0 −48
0
0 » . Notice that
0 −1 3
0 » and D = « 0
−48 0
»
«
»
0
0
36 0 »
1
0 1
0»
«0
«¬ 0
0
0
0 36 »¼.
0
1 0
3»¼
the anwer in the text lists the eigenvector in an different order in P, and hence the eigenvalues are listed
in a different order in D. Both answers are correct.
ª1
«1
«
Thus we construct P = « 0
«
«1
«¬ 0
0
0
1
Notes: For your use, here is another matrix with five distinct real eigenvalues. To four decimal places,
they are 11.0654, 9.8785, 3.8238, −3.7332, and −6.0345.
ª 6
« −7
«
« −3
«
« 0
«¬ −5
−8
5
−3
3
−7
−5
5
3
−3
−4
−3
1
−2
−7
0
0º
0 »»
5»
»
5»
8»¼
The MATLAB box in the Study Guide encourages students to use eig (A) and nulbasis to
practice the diagonalization procedure in this section. It also remarks that in later work, a student may
automate the process, using the command [P D] = eig (A). You may wish to permit students to use
the full power of eig in some problems in Sections 5.5 and 5.7.
5.4
SOLUTIONS
ª −1º
ª 3º
1. Since T (b1 ) = 3d1 − 5d 2 ,[T (b1 )]D = « » . Likewise T (b 2 ) = −d1 + 6d 2 implies that [T (b 2 )]D = « »
¬ 6¼
¬ −5¼
ª0 º
and T (b 3 ) = 4d 2 implies that [T (b3 )]D = « » . Thus the matrix for T relative to B and
¬4¼
ª 3
D is ª¬[T (b1 )]D [T (b 2 )]D [T (b3 )]D º¼ = «
¬ −5
−1
6
0º
.
4»¼
ª 3º
ª −2º
2.Since T (d1 ) = 3b1 − 3b 2 , [T (d1 )]B = « » . Likewise T (d 2 ) = −2b1 + 5b 2 implies that [T (d 2 )]B = « » .
¬ −3¼
¬5¼
ª 3 −2º
.
Thus the matrix for T relative to D and B is ª¬[T (d1 )]B [T (d 2 )]B º¼ = «
5»¼
¬ −3
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.4
• Solutions
303
3. a. T (e1 ) = 0b1 − 0b 2 + b 3 , T (e 2 ) = −b1 − 2b 2 + 0b 3 , T (e3 ) = 2b1 + 0b 2 + 3b 3
ª0 º
ª −1º
ª 2º
«
»
«
»
b. [T (e1 )]B = «0 » , [T (e 2 )]B = « −2 » , [T (e3 )]B = «« 0»» .
«¬ 0 »¼
¬« 1¼»
¬« 3¼»
c. The matrix for T relative to ! and B is [ [T (e1 )]B [T (e 2 )]B
4. Let ! = {e1 , e 2 } be the standard basis for
ª0
[T (e3 )]B ] = ««0
«¬ 1
−1
−2
0
2º
0 »» .
3»¼
ª2º
ª −3º
. Since [T (b1 )]! = T (b1 ) = « » , [T (b 2 )]! = T (b 2 ) = « » ,
¬ −2¼
¬0¼
ª1 º
and [T (b3 )]! = T (b3 ) = « » , the matrix for T relative to B and
¬5¼
ª 2 −3 1º
! is [[T (b1 )]! [T (b 2 )]! [T (b 3 )]! ] = «
».
¬ −2 0 5¼
5. a. T (p) = (t + 3)(3 − 2t + t 2 ) = 9 − 3t + t 2 + t 3
b. Let p and q be polynomials in
2,
and let c be any scalar. Then
T (p(t ) + q(t )) = (t + 3)[p(t ) + q(t )] = (t + 3)p(t ) + (t + 3)q(t )
= T (p(t )) + T (q(t ))
T (c ⋅ p(t )) = (t + 3)[c ⋅ p(t )] = c ⋅ (t + 3)p (t )
= c ⋅ T [p(t )]
and T is a linear transformation.
ª3º
«1 »
c. Let B = {1, t, t 2 } and C = {1, t, t 2, t 3 } . Since T (b1 ) = T (1) = (t + 3)(1) = t + 3, [T (b1 )]C = « » .
«0»
« »
¬0¼
ª0º
« 3»
Likewise since T (b 2 ) = T (t ) = (t + 3)(t ) = t 2 + 3t , [T (b 2 )]C = « » , and since
«1 »
« »
¬0¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
304
CHAPTER 5
• Eigenvalues and Eigenvectors
ª0º
«0»
T (b3 ) = T (t 2 ) = (t + 3)(t 2 ) = t 3 + 3t 2 , [T (b3 )]C = « » . Thus the matrix for T relative to B and
« 3»
« »
¬1 ¼
C is [ [T (b1 )]C [T (b 2 )]C
ª3
«1
[T (b3 )]C ] = «
«0
«
¬0
0
3
1
0
0º
0 »»
.
3»
»
1¼
6. a. T (p) = (3 − 2t + t 2 ) + 2t 2 (3 − 2t + t 2 ) = 3 − 2t + 7t 2 − 4t 3 + 2t 4
b. Let p and q be polynomials in
2,
and let c be any scalar. Then
T (p(t ) + q(t )) = [p(t ) + q(t )] + 2t 2 [p(t ) + q(t )]
= [p(t ) + 2t 2p(t )] + [q (t ) + 2t 2q (t )]
= T (p(t )) + T (q(t ))
T (c ⋅ p(t )) = [c ⋅ p (t )] + 2t 2 [c ⋅ p(t )]
= c ⋅ [p(t ) + 2t 2p(t )]
= c ⋅ T [p(t )]
and T is a linear transformation.
ª1 º
«0 »
« »
c. Let B = {1, t, t 2 } and C = {1, t, t 2, t 3, t 4 } . Since T (b1 ) = T (1) = 1 + 2t 2 (1) = 1 + 2t 2 , [T (b1 )]C = « 2 » .
« »
«0 »
«¬0 »¼
ª0º
«1 »
« »
Likewise since T (b 2 ) = T (t ) = t + (2t 2 )(t ) = t + 2t 3 , [T (b 2 )]C = « 0 », and
« »
« 2»
«¬ 0 »¼
ª0 º
«0 »
« »
since T (b3 ) = T (t 2 ) = t 2 + (2t 2 )(t 2 ) = t 2 + 2t 4 , [T (b3 )]C = «1 » . Thus the matrix for T relative to
« »
«0 »
«¬ 2 »¼
ª1 0 0 º
«0 1 0 »
«
»
B and C is [ [T (b1 )]C [T (b 2 )]C [T (b3 )]C ] = « 2 0 1 » .
«
»
«0 2 0 »
«¬0 0 2 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.4
• Solutions
ª3º
7. Since T (b1 ) = T (1) = 3 + 5t, [T (b1 )]B = ««5 »» . Likewise since
«¬0 »¼
ª 0º
ª0º
«
»
2
2
2
T (b 2 ) = T (t ) = −2t + 4t , [T (b 2 )]B = « −2 » , and since T (b3 ) = T (t ) = t , [T (b3 )]B = «« 0 »» . Thus the
«¬ 4 »¼
«¬1 »¼
matrix representation of T relative to the basis
0 0º
ª3
«
B is ª¬[T (b1 )]B [T (b 2 )]B [T (b3 )]B º¼ = « 5 −2 0 »» . Perhaps a faster way is to realize that the
«¬ 0
4 1»¼
information given provides the general form of T (p) as shown in the figure below:
T
a0 + a1t + a2t 2 → 3a0 + (5a0 − 2a1 )t + (4a1 + a2 )t 2
coordinate
mapping
coordinate
mapping
ªa º
« 0»
«
»
« a1 »
«
»
«a »
¬« 2 ¼»
ª 3a
º
0
«
»
«
»
→ « 5a0 −2a1 »
«
»
by[T ]B
« 4a + a »
1
2 ¼»
¬«
multiplication
The matrix that implements the multiplication along the bottom of the figure is easily filled in by
inspection:
ª?
«?
«
«¬?
?
?
?
? º ª« a0 º» ª« 3a0 º»
ª3
«
»
«
»
»
? » « a1 » = «5a0 −2a1 » implies that [T ]B = «« 5
«
»
«
»
«¬ 0
? »¼ «¬« a2 »¼» «¬« 4a1 + a2 »¼»
0
−2
4
0º
0 »»
1»¼
ª 4º
ª0
«
»
8. Since [4b1 − 3b 2 ]B = « −3» , [T (4b1 − 3b 2 )]B = [T ]B [4b1 − 3b 2 ]B = «« 2
«¬ 0 »¼
«¬ 1
0
1
3
1º ª 4 º ª 0 º
−2 »» «« −3»» = «« 5»»
1»¼ «¬ 0 »¼ «¬ −5»¼
and T (4b1 − 3b 2 ) = 5b 2 − 5b 3 .
ª5 + 3(−1) º ª 2 º
9. a. T (p) = «« 5 + 3(0) »» = ««5 »»
«¬ 5 + 3(1) »¼ «¬8 »¼
b. Let p and q be polynomials in
2,
and let c be any scalar. Then
ª(p + q)(−1) º ªp(−1) + q(−1) º ªp( −1) º ªq(−1) º
T (p + q) = «« (p + q)(0) »» = «« p(0) + q(0) »» = «« p(0) »» + «« q(0) »» = T (p) + T (q)
«¬ (p + q)(1) »¼ «¬ p(1) + q(1) »¼ «¬ p(1) »¼ «¬ q (1) »¼
ªp (−1) º
ª(c ⋅ p)(−1) º ªc ⋅ (p(−1)) º
«
»
«
»
T (c ⋅ p ) = (c ⋅ p )(0) = « c ⋅ (p(0)) » = c ⋅ «« p(0) »» = c ⋅ T (p)
«
»
«¬ p(1) »¼
«¬ (c ⋅ p)(1) »¼ «¬ c ⋅ (p(1)) »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
305
306
CHAPTER 5
• Eigenvalues and Eigenvectors
and T is a linear transformation.
c. Let B = {1, t, t 2 } and ! = {e1 , e 2 , e3 } be the standard basis for 3. Since
ª1º
ª −1º
«
»
[T (b1 )]! = T (b1 ) = T (1) = «1» , [T (b 2 )]! = T (b 2 ) = T (t ) = «« 0 »» , and
«¬1»¼
«¬ 1»¼
ª1 º
[T (b3 )]! = T (b 3 ) = T (t ) = «« 0 »» , the matrix for T relative to B and ! is
«¬1 »¼
ª1 −1 1º
«
ª[T (b )]
º
0 0 »» .
1 ! [T (b 2 )]! [T (b 3 )]! ¼ = «1
¬
1 1¼»
¬«1
2
10. a. Let p and q be polynomials in
3,
and let c be any scalar. Then
ª(p + q )(−2) º ªp(−2) + q (−2) º ªp(−2) º ªq(−2) º
« (p + q )(3) » « p(3) + q(3) » « p(3) » « q(3) »
»=«
» =«
»+«
» = T (p) + T (q)
T (p + q) = «
« (p + q)(1) » « p(1) + q(1) » « p(1) » « q(1) »
» «
» «
»
«
» «
¬ (p + q)(0) ¼ ¬ p (0) + q (0) ¼ ¬ p(0) ¼ ¬ q(0) ¼
ª(c ⋅ p)(−2) º ªc ⋅ (p(−2)) º
ªp(−2) º
« (c ⋅ p)(3) » « c ⋅ (p(3)) »
«
»
»=«
» = c ⋅ « p(3) » = c ⋅ T (p)
T (c ⋅ p ) = «
« (c ⋅ p)(1) » « c ⋅ (p(1)) »
« p(1) »
«
» «
»
«
»
¬ (c ⋅ p)(0) ¼ ¬ c ⋅ (p(0)) ¼
¬ p(0) ¼
and T is a linear transformation.
b. Let B = {1, t , t 2 , t 3 } and ! = {e1 , e 2 , e3 , e 4 } be the standard basis for 4. Since
ª1º
ª −2 º
ª 4º
«1»
« 3»
«9»
[T (b1 )]! = T (b1 ) = T (1) = « » , [T (b 2 )]! = T (b 2 ) = T (t ) = « » , [T (b3 )]! = T (b3 ) = T (t 2 ) = « » ,
«1»
« 1»
«1 »
«»
« »
« »
¬1¼
¬ 0¼
¬0¼
ª −8º
« 27 »
3
and [T (b 4 )]! = T (b 4 ) = T (t ) = « » , the matrix for T relative to B and ! is
« 1»
« »
¬ 0¼
ª1 −2 4 −8º
«1
3 9 27 »»
ª[T (b )]
º=«
T
T
T
[
(
)]
[
(
)]
[
(
)]
.
b
b
b
1 !
2 !
3 !
4 !¼
¬
«1
1 1
1»
«
»
0 0
0¼
¬1
11. Following Example 4, if P = ª«¬b1
ª −1
b2 º»¼ = «
¬ 2
−1º
, then the B-matrix is
1»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.4
ª 1
P −1 AP = «
¬ −2
1º ª −4
−1»¼ «¬ 6
−1º ª −1
1»¼ «¬ 2
12. Following Example 4, if P = ª«¬b1
ª 2
P −1 AP = «
¬ −1
1º ª −6
0»¼ «¬ 4
−1º ª −2
=
1»¼ «¬ 0
ª0
b2 º»¼ = «
¬1
−2º ª0
0»¼ «¬ 1
• Solutions
307
−2º
−1»¼
−1º
, then the B-matrix is
2»¼
−1º ª −4
=
2»¼ «¬ 2
0º
−2 »¼
13. Start by diagonalizing A. The characteristic polynomial is
eigenvalues of A are 1 and 3.
2
− 4 + 3 = ( − 1)( − 3), so the
ª −1
For λ = 1: A − I = «
¬ −3
1º
. The equation ( A − I )x = 0 amounts to − x1 + x2 = 0, so x1 = x2 with x2
3»¼
ª1º
free. A basis vector for the eigenspace is thus v1 = « » .
¬1¼
ª −3 1º
For λ = 3: A − 3I = «
» . The equation ( A − 3I )x = 0 amounts to −3 x1 + x2 = 0, so x1 = (1/ 3) x2
¬ −3 1¼
ª1º
with x2 free. A nice basis vector for the eigenspace is thus v 2 = « » .
¬3 ¼
ª1 1º
v2 º»¼ = «
» which diagonalizes A. By Theorem 8, the
¬1 3¼
basis B = {v1 , v 2 } has the property that the B-matrix of the transformation x 6 Ax is a diagonal
matrix.
From v1 and v 2 we may construct P = ª«¬ v1
14. Start by diagonalizing A. The characteristic polynomial is
eigenvalues of A are 5 and −1.
2
− 4 − 5 = ( − 5)( + 1), so the
ª −3
For λ = 5: A − 5I = «
¬ 3
3º
. The equation ( A − 5 I )x = 0 amounts to x1 − x2 = 0, so x1 = x2 with
−3»¼
ª1º
x2 free. A basis vector for the eigenspace is thus v1 = « » .
¬1¼
ª3
For λ = −1 : A + I = «
¬3
3º
. The equation ( A + I )x = 0 amounts to x1 + x2 = 0, so x1 = − x2 with x2
3»¼
ª −1º
free. A nice basis vector for the eigenspace is thus v 2 = « » .
¬1¼
ª1 −1º
v2 º»¼ = «
which diagonalizes A. By Theorem 8,
1»¼
¬1
the basis B = {v1 , v 2 } has the property that the B-matrix of the transformation x 6 Ax is a diagonal
matrix.
From v1 and v 2 we may construct P = ª«¬ v1
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
308
CHAPTER 5
• Eigenvalues and Eigenvectors
15. Start by diagonalizing A. The characteristic polynomial is
eigenvalues of A are − 5 and 2.
2
+ 3 − 10 = ( +5)( − 2), so the
ª6
For λ = 5: A + 5I = «
¬3
2º
. The equation ( A + 5 I )x = 0 amounts to 3 x1 + x2 = 0, so x1 = ( −1 / 3) x2
1»¼
ª −1º
with x2 free. A basis vector for the eigenspace is thus v1 = « » .
¬ 3¼
ª −1
For λ = 2: A − 2 I = «
¬ 3
2º
. The equation ( A − 2 I )x = 0 amounts to x1 − 2 x2 = 0, so x1 = 2 x2
−6 »¼
ª 2º
with x2 free. A basis vector for the eigenspace is thus v 2 = « » .
¬1 ¼
ª −1 2º
v2 º¼» = «
» which diagonalizes A. By Theorem 8,
¬ 3 1¼
the basis B = {v1 , v 2 } has the property that the B-matrix of the transformation x 6 Ax is a diagonal
matrix. Note that the solution in the text lists the vectors in the reverse order, which is also correct.
From v1 and v 2 we may construct P = ª¬« v1
16. Start by diagonalizing A. The characteristic polynomial is
eigenvalues of A are 3 and 6.
2
− 9 +18 = ( − 3)( − 6), so the
−2º
. The equation ( A − 3I )x = 0 amounts to x1 − 2 x2 = 0, so x1 = 2 x2
2 »¼
ª 2º
with x2 free. A basis vector for the eigenspace is thus v1 = « » .
¬ 1¼
ª1
For λ = 3: A − 3I = «
¬ −1
ª −2
For λ = 6: A − 6 I = «
¬ −1
−2º
. The equation ( A − 6 I )x = 0 amounts to x1 + x2 = 0, so x1 = − x2 with
−1»¼
ª −1º
x2 free. A basis vector for the eigenspace is thus v 2 = « » .
¬ 1¼
ª 2 −1º
v2 º¼» = «
which diagonalizes A. By Theorem 8,
1»¼
¬1
the basis B = {v1 , v 2 } has the property that the B-matrix of the transformation x 6 Ax is a diagonal
matrix.
From v1 and v 2 we may construct P = ª¬« v1
17. a. We compute that
ª 4 1º ª 1 º ª 3 º
Ab1 = «
» « » = « » = 3b1
¬ −1 2 ¼ ¬ −1¼ ¬ −3¼
so b1 is an eigenvector of A corresponding to the eigenvalue 3. The characteristic polynomial of
1º
ª1
− 6 + 9 = ( − 3) 2 , so 3 is the only eigenvalue for A. Now A − 3I = «
» , which
¬ −1 −1¼
implies that the eigenspace corresponding to the eigenvalue 3 is one-dimensional. Thus the matrix
A is not diagonalizable.
A is
2
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.4
b. Following Example 4, if P = ª«¬b1
ª 2 1º ª 4
P −1 AP = «
»«
¬ 1 1¼ ¬ −1
• Solutions
309
b2 º»¼ , then the B-matrix for T is
1º ª 1
2»¼ «¬ −1
−1º ª 3
=
2»¼ «¬0
1º
3»¼
18. If there is a basis B such that [T ]B is diagonal, then A is similar to a diagonal matrix, by the second
paragraph following Example 3. In this case, A would have three linearly independent eigenvectors.
However, this is not necessarily the case, because A has only two distinct eigenvalues.
19. If A is similar to B, then there exists an invertible matrix P such that P −1 AP = B. Thus B is invertible
because it is the product of invertible matrices. By a theorem about inverses of products,
B −1 = P −1 A−1 ( P −1 )−1 = P −1 A−1 P, which shows that A−1 is similar to B −1.
20. If A = PBP −1 , then A2 = ( PBP −1 )( PBP −1 ) = PB ( P −1P ) BP −1 = PB ⋅ I ⋅ BP −1 = PB 2 P −1. So A2 is
similar to B 2.
21. By hypothesis, there exist invertible P and Q such that P −1 BP = A and Q −1CQ = A. Then
P −1 BP = Q −1CQ. Left-multiply by Q and right-multiply by Q −1 to obtain
QP −1 BPQ −1 = QQ −1CQQ −1.
So C = QP −1 BPQ −1 = ( PQ −1 )−1 B ( PQ −1 ), which shows that B is similar to C.
22. If A is diagonalizable, then A = PDP −1 for some P. Also, if B is similar to A, then B = QAQ −1
for some Q. Then B = Q( PDP −1 )Q −1 = (QP ) D( P −1Q −1 ) = (QP ) D(QP) −1
So B is diagonalizable.
23. If Ax = x, x ≠ 0, then P −1 Ax = P −1x. If B = P −1 AP, then
B ( P −1x) = P −1 AP ( P −1x) = P −1 Ax = P −1x
(*)
by the first calculation. Note that P −1x ≠ 0, because x ≠ 0 and P −1 is invertible. Hence (*) shows
that P −1x is an eigenvector of B corresponding to . (Of course,
because the matrices are similar, by Theorem 4 in Section 5.2.)
is an eigenvalue of both A and B
24. If A = PBP −1 , then rank A = rank P ( BP −1 ) = rank BP −1 , by Supplementary Exercise 13 in Chapter
4. Also, rank BP −1 = rank B, by Supplementary Exercise 14 in Chapter 4, since P −1 is invertible.
Thus rank A = rank B.
25. If A = PBP −1 , then
tr( A) = tr(( PB) P −1 ) = tr( P −1 ( PB))
By the trace property
= tr( P −1 PB ) = tr( IB ) = tr( B )
If B is diagonal, then the diagonal entries of B must be the eigenvalues of A, by the Diagonalization
Theorem (Theorem 5 in Section 5.3). So tr A = tr B = {sum of the eigenvalues of A}.
26. If A = PDP −1 for some P, then the general trace property from Exercise 25 shows that
tr A = tr [( PD ) P −1 ] = tr [ P −1 PD ] = tr D. (Or, one can use the result of Exercise 25 that since A is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
310
CHAPTER 5
• Eigenvalues and Eigenvectors
similar to D, tr A = tr D. ) Since the eigenvalues of A are on the main diagonal of D, tr D is the sum
of the eigenvalues of A.
27. For each j, I (b j ) = b j . Since the standard coordinate vector of any vector in
n
is just the vector
itself, [ I (b j )]ε = b j . Thus the matrix for I relative to B and the standard basis ! is simply
ªb
«¬ 1
b2
… b n º»¼ . This matrix is precisely the change-of-coordinates matrix PB defined in Section
4.4.
28. For each j, I (b j ) = b j , and [ I (b j )]C = [b j ]C . By formula (4), the matrix for I relative to the bases
B and C is M = ª¬«[b1 ]C
… [b n ]C º¼» In Theorem 15 of Section 4.7, this matrix was
.
denoted by P and was called the change-of-coordinates matrix from B to C.
[b2 ]C
C ←B
29. If B = {b1 , …, b n }, then the B-coordinate vector of b j is e j , the standard basis vector for
n
. For
instance,
b1 = 1 ⋅ b1 + 0 ⋅ b 2 + … + 0 ⋅ b n
Thus [ I (b j )]B = [b j ]B = e j , and
[ I ]B = ª¬[ I (b1 )]B " [ I (b n )]B º¼ = [e1 " e n ] = I
30. [M] If P is the matrix whose columns come from B, then the B-matrix of the transformation
x 6 Ax is D = P −1 AP. From the data in the text,
−2
ª6
A = «« 3
«¬ 2
−2 º
−2 »» , P = ª¬«b1
2 »¼
1
−2
ª −3
D = «« 1
«¬ −2
3
−1
1
1º ª 6
0 »» «« 3
1»¼ «¬ 2
−2
1
−2
b2
b3 º¼»
−2 º ª1
−2 »» ««1
2 »¼ «¬1
ª1
= ««1
«¬1
2
1
3
2
1
3
−1º
−1»» ,
0 »¼
−1º ª 2
−1»» = «« 0
0 »¼ «¬ 0
−1
3
1
0º
0 »»
4 »¼
31. [M] If P is the matrix whose columns come from B, then the B-matrix of the transformation
x 6 Ax is D = P −1 AP. From the data in the text,
ª −7
A = «« 1
«¬ −3
ª −1
D = «« 1
«¬ 0
−48
14
−45
−3
3
−1
ª −6
« −3
32. [M] A = «
« −1
«
¬ −4
−16 º
6 »» , P = ª¬«b1
−19 »¼
−1/ 3º ª −7
0 »» «« 1
−1/ 3»¼ «¬ −3
4
0
0
1
−2
4
1
0
3º
ª −3 −2
«
1 −1»» ,
b2
=« 1
«¬ −3 −3
0»¼
−48 −16 º ª −3 −2
3º ª −7
»
«
14
6» « 1
1 −1»» = «« 0
0 »¼ «¬ 0
−45 −19 »¼ «¬ −3 −3
b3 º¼»
−2
−4
0
−6 º
−6 »»
−1»¼
9º
6 »»
,
0»
»
7¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.4
• Solutions
311
ev = eig(A) = (5, 1, -2, -2)
ª 1.0000 º
« 0.5000 »
»
nulbasis(A-ev(1)*eye(4)) = «
« −0.5000 »
«
»
¬ 1.0000 ¼
A basis for the eigenspace of
ª 2º
« 1»
= 5 is b1 = « » .
« −1»
« »
¬ 2¼
ª 1.0000 º
« −0.5000 »
».
nulbasis(A-ev(2)*eye(4)) = «
« −3.5000 »
«
»
¬ 1.0000 ¼
A basis for the eigenspace of
ª 2º
« −1»
= 1 is b 2 = « » .
« −7 »
« »
¬ 2¼
ª1.0000
«1.0000
nulbasis(A-ev(3)*eye(4)) = «
«1.0000
«
0
¬
A basis for the eigenspace of
1.5000 º
−0.7500 »»
0»
»
1.0000 ¼
ª 1º ª 6 º ½
°« » « » °
° 1 −3 °
= −2 is {b3 , b 4 } = ® « » , « » ¾ .
° « 1» « 0 » °
°« » « » °
¯ ¬0¼ ¬ 4¼ ¿
The basis B = {b1 , b 2 , b 3 , b 4 } is a basis for
4
with the property that [T ]B is diagonal.
Note: The Study Guide comments on Exercise 26 and tells students that the trace of any square matrix A
equals the sum of the eigenvalues of A, counted according to multiplicities. This provides a quick check
on the accuracy of an eigenvalue calculation. You could also refer students to the property of the
determinant described in Exercise 19 of Section 5.2.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
312
CHAPTER 5
• Eigenvalues and Eigenvectors
5.5
SOLUTIONS_____________________________________________
ª1
1. A = «
¬1
−2º
ª1 − λ
, A − λI = «
»
3¼
¬ 1
−2 º
3 − λ»¼
det( A − I ) = (1 − )(3 − ) − (−2) =
2
−4 +5
Use the quadratic formula to find the eigenvalues: λ = 4 ± 16 − 20 = 2 ± i. Example 2 gives a
2
shortcut for finding one eigenvector, and Example 5 shows how to write the other eigenvector with
no effort.
ª −1 − i
For λ = 2 + i: A − (2 + i ) I = «
¬ 1
−2 º
. The equation ( A − I )x = 0 gives
1 − i »¼
(−1 − i ) x1 − 2 x2 = 0
x1 + (1 − i ) x2 = 0
As in Example 2, the two equations are equivalent—each determines the same relation between x1
and x2 . So use the second equation to obtain x1 = − (1 − i ) x2 , with x2 free. The general solution is
ª −1 + i º
ª −1 + i º
x2 «
, and the vector v1 = «
» provides a basis for the eigenspace.
»
¬ 1 ¼
¬ 1 ¼
ª −1 − i º
For λ = 2 – i: Let v 2 = v1 = «
» . The remark prior to Example 5 shows that v 2 is
¬ 1 ¼
automatically an eigenvector for 2 + i. In fact, calculations similar to those above would show that
{v 2 } is a basis for the eigenspace. (In general, for a real matrix A, it can be shown that the set of
complex conjugates of the vectors in a basis of the eigenspace for λ is a basis of the eigenspace for
λ .)
ª3
2. A = «
¬3
−3º
. The characteristic polynomial is
3»¼
2
− 6 + 18, so the eigenvalues of A are
= 6 ± 36 − 72 = 3 ± 3i.
2
ª −3i
For λ = 3 + 3i: A − (3 + 3i) I = «
¬ 3
−3 º
. The equation ( A − (3 + 3i) I )x = 0 amounts to
−3i »¼
ªi º
x1 − ix2 = 0, so x1 = ix2 with x2 free. A basis vector for the eigenspace is thus v1 = « » .
¬1¼
ª −i º
For λ = 3 – 3i: A basis vector for the eigenspace is v 2 = v1 = « » .
¬1¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.5
ª 5
3. A = «
¬ −8
1º
. The characteristic polynomial is
1»¼
2
• Solutions
313
− 6 + 13, so the eigenvalues of A are
= 6 ± 36 − 52 = 3 ± 2i.
2
1 º
ª 2 − 2i
. The equation ( A − (3 + 2i ) I )x = 0 amounts to
For λ = 3 + 2i: A − (3 + 2i ) I = «
−2 − 2i »¼
¬ −8
−1 − i
x2 with x2 free. A nice basis vector for the eigenspace is thus
−8 x1 + ( −2 − 2i ) x2 = 0, so x1 =
4
ª −1 − i º
v1 = «
».
¬ 4 ¼
ª −1 + i º
For λ = 3 – 2i: A basis vector for the eigenspace is v 2 = v1 = «
».
¬ 4 ¼
ª1 −2 º
. The characteristic polynomial is
4. A = «
3»¼
¬1
= 4 ± −4 = 2 ± i.
2
2
− 4 + 5, so the eigenvalues of A are
ª −1 − i −2 º
. The equation ( A − (2 + i) I )x = 0 amounts to
For λ = 2 + i: A − (2 + i ) I = «
1 − i »¼
¬ 1
x1 + (1 − i ) x2 = 0, so x1 = −(1 − i ) x2 with x2 free. A basis vector for the eigenspace is thus
ª −1 + i º
v1 = «
».
¬ 1 ¼
ª −1 − i º
For λ = 2 – i: A basis vector for the eigenspace is v 2 = v1 = «
».
¬ 1 ¼
ª 3 1º
5. A = «
» . The characteristic polynomial is
¬ −2 5¼
= 8 ± −4 = 4 ± i.
2
2
− 8 + 17, so the eigenvalues of A are
1 º
ª −1 − i
. The equation ( A − (4 + i) I )x = 0 amounts to
For λ = 4 + i: A − (4 + i ) I = «
1 − i »¼
¬ −2
(1 − i)
x2 with x1 free. A basis vector for the eigenspace is thus
−2 x1 + (1 − i ) x2 = 0, so x1 =
2
ª1 − i º
v1 = «
».
¬ 2 ¼
ª1 + i º
For λ = 4 – i: A basis vector for the eigenspace is v 2 = v1 = «
».
¬ 2 ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
314
CHAPTER 5
• Eigenvalues and Eigenvvectors
ª7 −5º
. The characteristic poolynomial is
6. A = «
3»¼
¬1
= 10 ± −4 = 5 ± i.
2
2
− 10 + 26, so the eigenvalues of A arre
−5º
ª2 − i
For λ = 5 + i: A − (5 + i ) I = «
» . The equation ( A − (5 + i) I )x = 0 amounts to
¬ 1 −2 − i ¼
x1 − (2 + i ) x2 = 0, so x1 = (2 + i ) x2 with x2 free. A basis vector for the eigenspace is thuss
ª2 + i º
v1 = «
».
¬ 1 ¼
ª2 − i º
For λ = 5 – i: A basis vector for thhe eigenspace is v 2 = v1 = «
».
¬ 1 ¼
ª
«
7. A = «
«
«¬
−1 º»
3
».
3 »»¼
1
From Example 6, thhe eigenvalues are
3 ± i. The scale factor for the tran
nsformation
2. For the angle of rotation, plot the point (a, b) = ( 3,1) in the
x 6 Ax is r =| |= ( 3) 2 + 12 = 2
xy-plane and use trigonometry:
ϕ = arctan (b/a ) = arctan (1// 3) = π/ 6 radians.
Note: Your students will want to know
w whether you permit them on an exam to omit calcu
ulations for a
−b º
and simplyy write the eigenvalues a ± bi. A similar question may
y arise about
a »¼
ªa
matrix of the form «
¬b
ª1º
the corresponding eigenvectors, « » aand
¬ −i ¼
ª1º
«i » , which are announced in the Practice Probleem. Students
¬¼
may have trouble keeping track of the coorrespondence between eigenvalues and eigenvectors.
8. A =
ª
« 3
«
«
¬« −3
3 3 º»
3
3
».
»
¼»
The eigenvaluees are 3 ± (3 3)i. The scale factor for the transformatio
on x 6 Ax
is r = | | = (3) 2 + (3 3) 2 = 6. Froom trigonometry, the angle of rotation ϕ is arctan (b/a ) = arctan
(−3 3 / 3) = −π / 3 radians.
Copyright © 2012 Pearrson Education, Inc. Publishing as Addison-Wesley.
5.5
ª 0
9. A = «
¬ −2
• Solutions
315
2º
. The eigenvalues are ±2i. The scale factor for the transformation x 6 Ax is
0 »¼
r = | | = (0)2 + (2) 2 = 2. From trigonometry, the angle of rotation ϕ is arctan (b/a ) = arctan
(−∞) = −π / 2 radians.
ª 0
10. A = «
¬ −.5
.5º
. The eigenvalues are ±.5i. The scale factor for the transformation x 6 Ax is
0»¼
r = | | = (0) 2 + .52 = .5. From trigonometry, the angle of rotation ϕ is
arctan (b/a ) = arctan (−∞) = −π/ 2 radians.
ª− 3
11. A = «
¬« −1
1º
» . The eigenvalues are − 3 ± i. The scale factor for the transformation x 6 Ax is
− 3 ¼»
r = | | = ( 3) 2 + (1) 2 = 2. From trigonometry, the angle of rotation ϕ is
arctan (b/a ) = arctan (1 / 3) = −5π / 6 radians.
ª 3
12. A = «
«¬ 3
− 3º
» . The eigenvalues are 3 ± ( 3)i. The scale factor for the transformation x 6 Ax is
3»¼
r =| |= 32 + ( 3)2 = 2 3. From trigonometry, the angle of rotation ϕ is
arctan (b/a ) = arctan ( 3 / 3) = π / 6 radians.
ª −1 − i º
= 2 ± i, and the eigenvector v = «
» corresponds to
¬ 1 ¼
ª −1 º
ª −1 −1º
and Im v = « » , take P = «
» . Then compute
¬ 0¼
¬ 1 0¼
13. From Exercise 1,
ª −1º
v=« »
¬ 1¼
= 2 − i. Since Re
1º ª1 −2º ª −1 −1º ª 0
1º ª −3 −1º ª 2 −1º
ª 0
C = P −1 AP = «
=«
=
»
«
»
«
»
3¼ ¬ 1 0¼ ¬ −1 −1¼» ¬« 2 −1¼» ¬« 1 2¼»
¬ −1 −1¼ ¬1
Actually, Theorem 9 gives the formula for C. Note that the eigenvector v corresponds to a − bi
ª 2 1º
instead of a + bi. If, for instance, you use the eigenvector for 2 + i, your C will be «
».
¬ −1 2 ¼
Notes: The Study Guide points out that the matrix C is described in Theorem 9 and the first column of C
is the real part of the eigenvector corresponding to a − bi, not a + bi, as one might expect. Since students
may forget this, they are encouraged to compute C from the formula C = P −1 AP, as in the solution above.
The Study Guide also comments that because there are two possibilities for C in the factorization of a
2 × 2 matrix as in Exercise 13, the measure of rotation of the angle associated with the transformation
x 6 Ax is determined only up to a change of sign. The “orientation” of the angle is determined by the
change of variable x = Pu. See Figure 4 in the text.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
316
CHAPTER 5
−3º
. The eigenvalues of A are
1»¼
ª 2 + iº
= 2 ± ( 2)i, and the eigenvector v = «
» corresponds
¬ i ¼
ª 2 1º
= 2 − ( 2)i. By Theorem 9, P = [Re v Im v ] = «
» and
¬ 0 1¼
ª3
14. A = «
¬1
to
• Eigenvalues and Eigenvectors
ª 2 / 2 − 2 / 2 º ª3 −3º ª 2 1º ª 2 − 2 º
C = P −1 AP = «
» There are two choices for whice
»«
«
»=«
1»¼ ¬ 0 1¼ ¬« 2
0
1¼ ¬ 1
2 ¼»
¬
eigenvalues to use and many choices for the muliple of the eigenvector to use,. The solutions can
look quite different. For example, for λ = 2 + 2i an eigenvector is
ª− 2 + iº
ª1 + ( 2)i º
−v = «
» =i⋅«
» = i ⋅ w Using w in Theorem 9, results in
1
¬ i ¼
¬
¼
ª1
2º
P = [Re w Im w ] = «
» and
¬1 0 ¼
0
1º ª3
ª
C = P −1 AP = «
»«
¬ 2 / 2 − 2 / 2¼ ¬ 1
−3º ª1
«
1»¼ ¬1
2º ª 2
»=«
0 ¼ «¬ − 2
2º
».
2 »¼
ª3 + i º
= 1 ± 3i, and the eigenvector v = «
» corresponds to
¬ 2i ¼
ª3 1 º
= 1 + 3i. By Theorem 9, P = [Re v Im v ] = «
» and
¬0 2 ¼
ª 0
15. A = «
¬ −2
5º
. The eigenvalues of A are
2 »¼
1 ª 2 −1º ª 0 5º ª3 1 º ª 1 3º
C = P −1 AP = «
=
. There are many choices for the muliple of the
3»¼ «¬ −2 2»¼ «¬0 2»¼ «¬ −3 1»¼
6 ¬0
eigenvector to use, and the solutions can look quite different. For example,
ª1 −3º
ª3 + i º
ª1 − 3i º
v=«
=i⋅«
= i ⋅ w . Using w in Theorem 9, results in P = [Re w Im w ] = «
» and
»
»
¬2 0 ¼
¬ 2i ¼
¬ 2 ¼
1ª 0
C = P −1 AP = «
6 ¬ −2
ª4
16. A = «
¬1
3º ª 0
1»¼ «¬ −2
5º ª 1
2»¼ «¬ 2
−3º ª 1
=
0 »¼ «¬ −3
−2 º
. The eigenvalues of A are
6 »¼
= 5 + i. By Theorem 9, P = [ Re v
3º
.
1»¼
ª1 + i º
= 5 ± i, and the eigenvector v = «
» corresponds to
¬ −i ¼
ª1
Im v ] = «
¬0
1º
and
−1»¼
1º ª 4 −2º ª 1
1º ª 5 1º
ª1
C = P −1 AP = «
=
. There are many choices for the muliple of the
»
«
»
«
6¼ ¬0 −1»¼ «¬ −1 5»¼
¬0 −1¼ ¬ 1
eigenvector to use, and the solutions can look quite different. For example,
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.5
• Solutions
ª −1
ª1 + i º
ª −1 + i º
v=«
= −i ⋅ «
= −i ⋅ w . Using w in Theorem 9, results in P = [Re w Im w ] = «
»
»
¬1
¬ −i ¼
¬ 1 ¼
ª0 1º ª 4 −2º ª −1 1º ª 5 1º
C = P −1 AP = «
=
.
»«
6»¼ «¬ 1 0»¼ «¬ −1 5»¼
¬ 1 1¼ ¬ 1
ª −11 −4º
. The eigenvalues of A are = −3 ± 4i, and the eigenvector
17. A = «
5»¼
¬ 20
ª1 + 2i º
ª1
v=«
corresponds to = −3 + 4i. By Theorem 9, P = [ Re v Im v ] = «
»
¬ −5i ¼
¬0
317
1º
and
0»¼
2º
and
−5»¼
2º ª −11 −4º ª 1
2º ª −3
4º
1 ª5
C = P −1 AP = «
=«
. There are many choices for the muliple
»
«
»
«
»
5¼ ¬0 −5¼ ¬ −4 −3»¼
5 ¬0 −1¼ ¬ 20
of the eigenvector to use, and the solutions can look quite different. For example,
ª −2 1 º
ª1 + 2i º
ª −2 + i º
v=«
= −i ⋅ «
= −i ⋅ w . Using w in Theorem 9, results in P = [Re w Im w ] = «
»
»
»
¬ 5 0¼
¬ −5i ¼
¬ 5 ¼
1 ª0
and C = P −1 AP = «
5 ¬5
ª3
18. A = «
¬2
1º ª −11
2»¼ «¬ 20
−4º ª −2
5»¼ «¬ 5
1º ª −3
=
0»¼ «¬ −4
4º
.
−3»¼
−5º
. The eigenvalues of A are
5»¼
= 4 + 3i. By Theorem 9, P = [ Re v
ª3 + i º
= 4 ± 3i, and the eigenvector v = «
» corresponds to
¬ −2i ¼
1º
ª3
Im v ] = «
» and
¬0 −2¼
1º ª 3 −5º ª 3
1º ª 4 3º
1 ª2
C = P −1 AP = «
=
. There are many choices for the muliple of
»
«
»
«
5¼ ¬0 −2»¼ «¬ −3 4»¼
6 ¬ 0 −3¼ ¬ 2
the eigenvector to use, and the solutions can look quite different. For example,
ª3 + i º
ª −1 + 3i º
v == «
= −i ⋅ «
»
» = −i ⋅ w . Using w in Theorem 9, results in
¬ −2i ¼
¬ 2 ¼
ª −1
P = [Re w Im w ] = «
¬2
3º
1 ª0
and C = P −1 AP = «
»
0¼
6 ¬2
3º ª 3
1»¼ «¬ 2
−5º ª −1
5»¼ «¬ 2
3º ª 4
=
0»¼ «¬ −3
3º
.
4»¼
ª1.52 −.7 º
. The characteristic polynomial is 2 − 1.92 + 1, so the eigenvalues of A are
19. A = «
»
.4¼
¬ .56
= .96 ± .28i. To find an eigenvector corresponding to .96 − .28i, we compute
ª.56 + .28i
A − (.96 − .28i) I = «
¬ .56
−.7
º
»
−.56 + .28i ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
318
CHAPTER 5
• Eigenvalues and Eigenvectors
The equation ( A − (.96 − .28i ) I )x = 0 amounts to .56 x1 + ( −.56 + .28i ) x2 = 0, so x1 = ((2 − i ) / 2) x2
ª2 − i º
with x2 free. A nice eigenvector corresponding to .96 − .28i is thus v = «
» . By Theorem 9,
¬ 2 ¼
ª 2 −1º
1 ª 0 1º ª1.52 −.7 º ª 2 −1º ª.96 −.28º
P = [ Re v Im v ] = «
=
and C = P −1 AP = «
»
.4»¼ «¬ 2
.96»¼
0¼
0»¼ «¬.28
2 ¬ −2 2»¼ «¬ .56
¬2
ª −3
20. A = «
¬ 4
−8º
. The eigenvalues of A are
5»¼
ª1 + i º
= 1 ± 4i, and the eigenvector v = «
» corresponds to
¬ −i ¼
1º
ª1
= 1 + 4i. By Theorem 9,By Theorem 9, P = [ Re v Im v ] = «
» and
¬0 −1¼
1º ª −3 −8º ª 1
1º ª 1 4º
ª1
C = P −1 AP = «
=
. There are many choices for the muliple
»
«
»
«
5¼ ¬0 −1»¼ «¬ −4 1»¼
¬0 −1¼ ¬ 4
of the eigenvector to use, and the solutions can look quite different. For example,
ª1 + i º
ª −1 + i º
v=«
= −i ⋅ «
»
» = −i ⋅ w . Using w in Theorem 9, results in
¬ −i ¼
¬ 1 ¼
ª −1
P = [Re w Im w ] = «
¬1
1º
ª0 1º ª −3
and C = P −1 AP = «
»
»«
0¼
¬ 1 1¼ ¬ 4
−8º ª −1
5»¼ «¬ 1
1º ª 1
=
0»¼ «¬ −4
4º
.
1»¼
21. The first equation in (2) is ( −.3 + .6i ) x1 − .6 x2 = 0. We solve this for x2 to find that
ª 2 º
x2 = (( −.3 + .6i ) /.6) x1 = (( −1 + 2i ) / 2) x1 . Letting x1 = 2, we find that y = «
» is an eigenvector
¬ −1 + 2i ¼
ª 2 º −1 + 2i ª −2 − 4i º −1 + 2i
for the matrix A. Since y = «
» = 5 « 5 » = 5 v1 the vector y is a complex
¬ −1 + 2i ¼
¬
¼
multiple of the vector v1 used in Example 2.
22. Since A( µ x) = µ ( Ax) = µ ( x) = ( µ x), µ x is an eigenvector of A.
23. (a) properties of conjugates and the fact that xT = xT
(b) Ax = Ax and A is real
(c) xT Ax is a scalar and hence may be viewed as a 1 × 1 matrix
(d) properties of transposes
(e) AT = A and the definition of q
24. xT Ax = xT (λx) = λ ⋅ xT x because x is an eigenvector. It is easy to see that xT x is real (and positive)
because zz is nonnegative for every complex number z. Since xT Ax is real, by Exercise 23, so is .
Next, write x = u + iv , where u and v are real vectors. Then
Ax = A(u + iv) = Au + iAv and
x= u+i v
The real part of Ax is Au because the entries in A, u, and v are all real. The real part of x is u
because and the entries in u and v are real. Since Ax and x are equal, their real parts are equal,
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.5
• Solutions
319
too. (Apply the corresponding statement about complex numbers to each entry of Ax.) Thus
Au = u, which shows that the real part of x is an eigenvector of A.
25. Write x = Re x + i (Im x), so that Ax = A(Re x) + iA(Im x). Since A is real, so are A(Re x) and
A(Im x). Thus A(Re x) is the real part of Ax and A(Im x) is the imaginary part of Ax.
26. a. If
= a − bi, then
Av = v = (a − bi )(Re v + i Im v )
= (a Re v + b Im v ) + i (a Im v − b Re v )
Re Av
Im Av
By Exercise 25,
A(Re v ) = Re Av = a Re v + b Im v
A(Im v ) = Im Av = −b Re v + a Im v
b. Let P = [ Re v
Im v ]. By (a),
ªa º
ª −b º
A(Re v) = P « » , A(Im v) = P « »
¬b ¼
¬ a¼
So
AP = [ A(Re v )
A(Im v) ]
ª ª a º ª −b º º
ªa
= «P « » P « »» = P «
¬b
¬ ¬b ¼ ¬ a ¼ ¼
ª 26
« −6
27. [M ] A = «
« −14
«
¬ −20
33
23
−8
−1
−19
−20
−16
−20
−b º
= PC
a »¼
20 º
−13»»
3»
»
−14 ¼
The MATLAB command [V D] = eig(A) returns
V = -0.5709 - 0.4172i
-0.5709 + 0.4172i
-0.3708 + 0.0732i
-0.3708 - 0.0732i
0.4941 - 0.0769i
0.4941 + 0.0769i
0.4440 + 0.2976i
0.4440 - 0.2976i
0.0769 + 0.4941i
0.0769 - 0.4941i
-0.4440 - 0.2976i
-0.4440 + 0.2976i
-0.0000 - 0.0000i
-0.0000 + 0.0000i
0.2976 - 0.4440i
0.2976 + 0.4440i
D = -2.0000 + 5.0000i
0
0
0
0
-2.0000 - 5.0000i
0
0
0
0
-4.0000 +10.0000i
0
0
0
0
-4.0000 -10.0000i
The eigenvalues of A are the elements listed on the diagonal of D. The corresponding
eigenvectors are listed in the corresponding columns of V. To get a nice eigenvector for
ª −1 + i º
« −i »
»
λ = −2 + 5i take v1=V(1:4,1)/V(3,1), resulting in v1 = «
« 1 »
«
»
¬ 0 ¼
For λ = −4 + 10i, take v2=V(1:4,3)/V(4,3), and then multiply the vector by 2 resulting in
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
320
CHAPTER 5
• Eigenvalues and Eigenvectors
ª −1 − i º
« 2i »
».
v2 = «
« −2i »
«
»
¬ 2 ¼
Hence by Theorem 9, P = ª¬« Re v1
ª −2
« −5
C=«
« 0
«
¬ 0
5
0
−2
0
0
0
−4
−10
ª 7
« −20
28. [M ] A = «
« 0
«
¬ 10
Im v1
Re v2
ª −1
« 0
Im v 2 º¼» = «
« 1
«
¬ 0
1
−1
−1
0
0
0
0
2
−1º
2 »»
and
−2 »
»
0¼
0º
0 »»
. Other choices are possible, but C must equal P −1 AP.
10 »
»
−4 ¼
11
20
−40
−86
−5
28
−10
60
17 º
−74 »»
−10 »
»
53¼
The matlab command [V D] = eig(A) returns
V= -0.2132 + 0.2132i
0 - 0.8528i
0
0 + 0.4264i
-0.2132 - 0.2132i
0 + 0.8528i
0
0 - 0.4264i
0.1085 - 0.3254i
0.3254 + 0.1085i
0 - 0.5423i
-0.2169 + 0.6508i
0.1085 + 0.3254i
0.3254 - 0.1085i
0 + 0.5423i
-0.2169 - 0.6508i
D = 2.0000 + 5.0000i
0
0
0
0
2.0000 - 5.0000i 0
0
0
0
3.0000 + 1.0000i 0
0
0
0
3.0000 - 1.0000
The eigenvalues of A are the elements listed on the diagonal of D. The corresponding
eigenvectors are listed in the corresponding columns of V. To get a nice eigenvector for
ª1 + i º
« −4 »
λ = 2 + 5i take v1=V(1:4,1)/V(4,1), and then multiply the vector by 2 resulting in v1 = « »
« 0»
«
»
¬ 2¼
For
ª −2 º
« −2i »
»
= 3 + i, take v2=V(1:4,3)/V(4,3), and then multiply the vector by 4 resulting in v 2 = «
« −3 + i »
«
»
4¼
¬
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.6
Hence by Theorem 9, P = ª¬« Re v1
ª 2
« −5
C=«
« 0
«
¬ 0
5.6
5
0
2
0
0
0
3
−1
Im v1
Re v2
ª 1
« −4
Im v2 º¼» = «
« 0
«
¬ 2
1
−2
0
0
0
0
−3
4
• Solutions
321
0º
−2 »»
and
1»
»
0¼
0º
0 »»
. Other choices are possible, but C must equal P −1 AP.
1»
»
3¼
SOLUTIONS
1. The exercise does not specify the matrix A, but only lists the eigenvalues 3 and 1/3, and the
ª1º
ª −1º
ª9 º
corresponding eigenvectors v1 = « » and v 2 = « » . Also, x0 = « » .
¬1¼
¬ 1¼
¬1 ¼
a. To find the action of A on x 0 , express x0 in terms of v1 and v 2 . That is, find c1 and c2 such
that x 0 = c1 v1 + c2 v 2 . This is certainly possible because the eigenvectors v1 and v 2 are linearly
independent (by inspection and also because they correspond to distinct eigenvalues) and hence
form a basis for R 2 . (Two linearly independent vectors in R 2 automatically span R 2 . ) The row
5º
ª1 −1 9º ª 1 0
«
reduction ª¬« v1 v2 x0 º¼» = «
shows that x 0 = 5 v1 − 4 v 2 . Since v1 and
»
1 1¼ ¬0 1 −4»¼
¬1
v 2 are eigenvectors (for the eigenvalues 3 and 1/3):
ª15º ª −4/ 3º ª 49/3º
x1 = Ax0 = 5 Av1 − 4 Av 2 = 5 ⋅ 3v1 − 4 ⋅ (1/ 3) v 2 = « » − «
»=«
»
¬15¼ ¬ 4/ 3¼ ¬ 41/ 3 ¼
b. Each time A acts on a linear combination of v1 and v 2 , the v1 term is multiplied by the
eigenvalue 3 and the v 2 term is multiplied by the eigenvalue 1/3:
x 2 = Ax1 = A[5 ⋅ 3v1 − 4(1/ 3) v 2 ] = 5(3) 2 v1 − 4(1/ 3) 2 v 2
In general, x k = 5(3) k v1 − 4(1/ 3) k v 2 , for k ≥ 0.
ª 1º
ª 2º
ª −3º
«
»
«
»
2. The vectors v1 = « 0 » , v 2 = « 1» , v 3 = «« −3»» are eigenvectors of a 3 × 3 matrix A, corresponding to
«¬ −3»¼
«¬ −5»¼
«¬ 7 »¼
ª −2 º
eigenvalues 3, 4/5, and 3/5, respectively. Also, x0 = «« −5»» . To describe the solution of the equation
«¬ 3»¼
x k +1 = Ax k ( k = 1, 2, …), first write x0 in terms of the eigenvectors.
ªv
¬« 1
v2
v3
x0 º¼»
ª 1
= «« 0
¬« −3
2
−3
1
−5
−3
7
−2 º ª1
−5»» ««0
3¼» ¬«0
0
0
1
0
0
0
2º
1 »» x0 = 2 v1 + v 2 + 2 v 3
2¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
322
CHAPTER 5
• Eigenvalues and Eigenvectors
Then, x1 = A(2 v1 + v 2 + 2 v 3 ) = 2 Av1 + Av 2 + 2 Av 3 = 2 ⋅ 3v1 + (4 / 5) v 2 + 2 ⋅ (3/ 5) v 3 . In general,
x k = 2 ⋅ 3k v1 + (4/ 5) k v 2 + 2 ⋅ (3/5)k v3 . For all k sufficiently large,
ª 1º
x k ≈ 2 ⋅ 3 v1 = 2 ⋅ 3 «« 0 »»
«¬ −3»¼
k
k
ª .5 .4º
2
3. A = «
» , det( A − λ I ) = (.5 − λ )(1.1 − λ ) + .08 = λ − 1.6λ + .63. This characteristic polynomial
−.
2
1
.
1
¬
¼
factors as (λ − .9)(λ − .7), so the eigenvalues are .9 and .7. If v1 and v 2 denote corresponding
eigenvectors, and if x 0 = c1 v1 + c2 v 2 , then
x1 = A(c1 v1 + c2 v 2 ) = c1 Av1 + c2 Av 2 = c1 (.9) v1 + c2 (.7) v 2
and for k ≥ 1,
x k = c1 (.9) k v1 + c2 (.7) k v 2
For any choices of c1 and c2 , both the owl and wood rat populations decline over time.
.5 .4 º
ª
, det( A − λ I ) = (.5 − λ )(1.1 − λ ) − (.4)(.125) = λ 2 − 1.6λ + .6. This characteristic
4. A = «
»
¬ −.125 1.1¼
polynomial factors as (λ − 1)(λ − .6), so the eigenvalues are 1 and .6. For the eigenvalue 1, solve
ª −.5 .4 0º ª −5 4 0º
ª 4º
( A − I )x = 0 : «
«
. A basis for the eigenspace is v1 = « » . Let v 2 be an
»
»
¬ −.125 .1 0¼ ¬ 0 0 0¼
¬5 ¼
eigenvector for the eigenvalue .6. (The entries in v 2 are not important for the long-term behavior of
the system.) If x 0 = c1 v1 + c2 v 2 , then x1 = c1 Av1 + c2 Av 2 = c1 v1 + c2 (.6) v 2 , and for k sufficiently
large,
ª 4º
ª 4º
xk = c1 « » + c2 (.6) k v 2 ≈ c1 « »
¬5 ¼
¬5¼
Provided that c1 ≠ 0, the owl and wood rat populations each stabilize in size, and eventually the
populations are in the ratio of 4 owls for each 5 thousand rats. If some aspect of the model were to
change slightly, the characteristic equation would change slightly and the perturbed matrix A might
not have 1 as an eigenvalue. If the eigenvalue becomes slightly larger than 1, the two populations
will grow; if the eigenvalue becomes slightly less than 1, both populations will decline.
.3º
ª .4
, det( A − λ I ) = λ 2 − 1.6λ + .5775. The quadratic formula provides the roots of the
5. A = «
»
¬ −.325 1.2¼
characteristic equation:
1.6 ± 1.62 − 4(.5775) 1.6 ± .25
=
= 1.05 and .55
2
2
Because one eigenvalue is larger than one, both populations grow in size. Their relative sizes are
determined eventually by the entries in the eigenvector corresponding to 1.05. Solve
( A − 1.05I ) x = 0 :
λ=
ª −.65
« −.325
¬
.3
.15
0º ª −13
0»¼ «¬ 0
6
0
0º
ª 6º
. An eigenvector is v1 = « » .
»
0¼
¬13¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.6
• Solutions
323
Eventually, there will be about 6 spotted owls for every 13 (thousand) flying squirrels.
.3º
ª .4
, and det( A − λ I ) = λ 2 − 1.6λ + .63 = (λ − .9)(λ − .7).
6. When p = .5, A = «
»
¬ −.5 1.2¼
The eigenvalues of A are .9 and .7, both less than 1 in magnitude. The origin is an attractor for the
dynamical system and each trajectory tends toward 0. So both populations of owls and squirrels
eventually perish.
The calculations in Exercise 4 (as well as those in Exercises 27 and Exercise 33 in Section 5.2) show
that if the largest eigenvalue of A is 1, then in most cases the population vector x k will tend toward a
multiple of the eigenvector corresponding to the eigenvalue 1. [If v1 and v 2 are eigenvectors, with
v1 corresponding to λ = 1, and if x 0 = c1 v1 + c2 v 2 , then x k tends toward c1 v1 , provided c1 is not
zero.] So the problem here is to determine the value of the predation parameter p such that the largest
eigenvalue of A is 1. Compute the characteristic polynomial:
.3 º
ª.4 − λ
det «
= (.4 − λ)(1.2 − λ) + .3 p = λ2 − 1.6λ + (.48 + .3 p)
»
¬ − p 1.2 − λ ¼
By the quadratic formula,
1.6 ± 1.62 − 4(.48 + .3 p)
λ=
2
The larger eigenvalue is 1 when
1.6 + 1.62 − 4(.48 + .3 p ) = 2 and
2.56 − 1.92 − 1.2 p = .4
In this case, .64 − 1.2 p = .16, and p = .4.
7. a. The matrix A in Exercise 1 has eigenvalues 3 and 1/3. Since | 3 | > 1 and | 1/ 3 | < 1, the origin is a
saddle point.
ª −1º
b. The direction of greatest attraction is determined by v 2 = « » , the eigenvector corresponding to
¬ 1¼
the eigenvalue with absolute value less than 1. The direction of greatest repulsion is determined
ª1º
by v1 = « » , the eigenvector corresponding to the eigenvalue greater than 1.
¬1¼
c. The drawing below shows: (1) lines through the eigenvectors and the origin, (2) arrows toward
the origin (showing attraction) on the line through v 2 and arrows away from the origin (showing
repulsion) on the line through v1 , (3) several typical trajectories (with arrows) that show the
general flow of points. No specific points other than v1 and v 2 were computed. This type of
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
324
CHAPTER 5
• Eigenvalues and Eigenvvectors
drawing is about all that one cann make without using a computer to plot points.
Note: If you wish your class to sketch ttrajectories for anything except saddle points, you willl need to go
beyond the discussion in the text. The foollowing remarks from the Study Guide are relevant.
Sketching trajectories for a dynamiccal system in which the origin is an attractor or a repeellor is more
difficult than the sketch in Exercise 77. There has been no discussion of the direction in
n which the
trajectories “bend” as they move towardd or away from the origin. For instance, if you rotatee Figure 1 of
Section 5.6 through a quarter-turn and rrelabel the axes so that x1 is on the horizontal axis, then
t
the new
figure corresponds to the matrix A with the diagonal entries .8 and .64 interchanged. In geneeral, if A is a
diagonal matrix, with positive diagonal entries a and d, unequal to 1, then the trajectories liee on the axes
pends on the
or on curves whose equations have thee form x2 = r ( x1 ) s , where s = (ln d ) / (ln a) and r dep
initial point x0 . (See Encounters with Ch
Chaos, by Denny Gulick, New York: McGraw-Hill, 1992, pp. 147–
150.)
8. The matrix from Exercise 2 has eigeenvalues 3, 4/5, and 3/5. Since one eigenvalue is greatter than 1
and the others are less than one in m
magnitude, the origin is a saddle point. The direction of
o greatest
repulsion is the line through the origgin and the eigenvector (1, 0,−3) for the eigenvalue 3. The
direction of greatest attraction is thee line through the origin and the eigenvector (−3, −3, 7)) for the
smallest eigenvalue 3/5.
ª 1.7
9. A = «
¬ −1.2
−.3º
, det( A − λI ) =
.8¼»
2
− 2.5 + 1 = 0
2.5 ± 2.52 − 4(1) 2.5 ± 2.25 2.5 ± 1.5
=
= 2 and .5
=
2
2
2
The origin is a saddle point becausee one eigenvalue is greater than 1 and the other eigenvalue is less
than 1 in magnitude. The direction oof greatest repulsion is through the origin and the eigeenvector v1
=
−.3 0 º ª1 1 0º
ª −.3
found below. Solve ( A − 2 I )x = 0 : «
»«
» , so x1 = –x2, and x2 iss free. Take
¬ −1.2 −1.2 0 ¼ ¬0 0 0¼
ª −1º
v1 = « » . The direction of greatest attraction is through the origin and the eigenvector v 2 found
¬ 1¼
Copyright © 2012 Pearrson Education, Inc. Publishing as Addison-Wesley.
5.6
ª 1.2
below. Solve ( A − .5I )x = 0 : «
¬ −1.2
ª1 º
v2 = « » .
¬ 4¼
ª .3 .4º
10. A = «
» , det( A − I ) =
¬ −.3 1.1¼
2
−.3
−.25
0º ª 1
0¼» ¬«0
.3
0
• Solutions
325
0º
, so x1 = .25 x2 , and x2 is free. Take
0¼»
− 1.4 + .45 = 0
1.4 ± 1.42 − 4(.45) 1.4 ± .16 1.4 ± .4
=
=
= .5 and .9
2
2
2
The origin is an attractor because both eigenvalues are less than 1 in magnitude. The direction of
greatest attraction is through the origin and the eigenvector v1 found below. Solve
=
ª −.2
( A − .5I )x = 0 : «
¬ −.3
ª .4
11. A = «
¬ −.4
.4
.6
0º ª 1
0»¼ «¬0
.5º
, det( A − I ) =
1.3»¼
2
−2
0
0º
ª 2º
, so x1 = 2 x2 , and x2 is free. Take v1 = « » .
»
0¼
¬1 ¼
− 1.7 + .72 = 0
1.7 ± 1.7 2 − 4(.72) 1.7 ± .01 1.7 ± .1
=
=
=
= .8 and .9
2
2
2
The origin is an attractor because both eigenvalues are less than 1 in magnitude. The direction of
greatest attraction is through the origin and the eigenvector v1 found below. Solve
ª −.4
( A − .8I )x = 0 : «
¬ −.4
.5
.5
0º ª 1
0 »¼ «¬0
ª .5 .6º
12. A = «
» , det( A − I ) =
¬ −.3 1.4¼
2
−1.25
0
0º
ª5 º
, so x1 = 1.25 x2 , and x2 is free. Take v1 = « » .
»
0¼
¬ 4¼
− 1.9 + .88 = 0
1.9 ± 1.92 − 4(.88) 1.9 ± .09 1.9 ± .3
=
=
=
= .8 and 1.1
2
2
2
The origin is a saddle point because one eigenvalue is greater than 1 and the other eigenvalue is less
than 1 in magnitude. The direction of greatest repulsion is through the origin and the eigenvector v1
ª −.6 .6 0º ª 1 −1 0º
, so x1 = x2 , and x2 is free. Take
found below. Solve ( A − 1.1I )x = 0 : «
»«
0 0»¼
¬ −.3 .3 0¼ ¬0
ª1º
v1 = « » . The direction of greatest attraction is through the origin and the eigenvector v 2 found
¬1¼
ª −.3
below. Solve ( A − .8I )x = 0 : «
¬ −.3
ª 2º
v2 = « » .
¬1 ¼
.6
.6
0º ª 1
0»¼ «¬0
−2
0
0º
, so x1 = 2 x2 , and x2 is free. Take
0»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
326
CHAPTER 5
ª .8
13. A = «
¬ −.4
• Eigenvalues and Eigenvectors
.3º
, det( A − I ) =
1.5»¼
2
− 2.3 + 1.32 = 0
2.3 ± 2.32 − 4(1.32) 2.3 ± .01 2.3 ± .1
=
=
= 1.1 and 1.2
2
2
2
The origin is a repellor because both eigenvalues are greater than 1 in magnitude. The direction of
greatest repulsion is through the origin and the eigenvector v1 found below. Solve
=
ª −.4
( A − 1.2 I )x = 0 : «
¬ −.4
ª1.7
14. A = «
¬ −.4
.3
0º ª 1
0 »¼ «¬0
.3
.6 º
, det( A − I ) =
.7 »¼
2
−.75
0
0º
ª 3º
, so x1 = .75 x2 , and x2 is free. Take v1 = « » .
»
0¼
¬ 4¼
− 2.4 + 1.43 = 0
2.4 ± 2.42 − 4(1.43) 2.4 ± .04 2.4 ± .2
=
=
= 1.1 and 1.3
2
2
2
The origin is a repellor because both eigenvalues are greater than 1 in magnitude. The direction of
greatest repulsion is through the origin and the eigenvector v1 found below. Solve
=
ª .4
( A − 1.3I )x = 0 : «
¬ −.4
ª.4
15. A = ««.3
¬«.3
.6
−.6
0º ª 1 1.5
0»¼ «¬0
0
0º
ª −3º
, so x1 = −1.5 x2 , and x2 is free. Take v1 = « » .
»
0¼
¬ 2¼
.2 º
ª . 1º
»
.3» . Given eigenvector v1 = ««.6 »» and eigenvalues .5 and .2. To find the eigenvalue for
.5 ¼»
¬«.3¼»
0
.8
.2
v1 , compute
ª.4
Av1 = ««.3
«¬.3
.8
.2
ª −.1
«
For λ = .5 : « .3
«¬ .3
ª.2
For λ = .2 : ««.3
¬«.3
.2 º ª .1º ª .1º
.3»» ««.6 »» = ««.6»» = 1 ⋅ v1 Thus v1 is an eigenvector for λ = 1.
.5»¼ «¬.3»¼ «¬.3»¼
0
0
.2
.3
.2
.3
0
0
.2
.6
.2
.3
.3
0º ª 1
0»» ««0
0 »¼ «¬0
0 º ª1
0 »» «« 0
0 ¼» ¬« 0
0
−2
1
0
3
0
0
1
1
0
0
0
0 º x1 = 2 x3
ª 2º
»
0» , x2 = −3x3 . Set v 2 = «« −3»» .
«¬ 1»¼
0»¼ x3 is free
0 º x1 = − x3
ª −1º
»
0 » , x2 = 0 . Set v 3 = «« 0»»
«¬ 1»¼
0 »¼ x3 is free
Given x 0 = (0, .3, .7), find weights such that x 0 = c1 v1 + cv 2 + c3 v 3 .
ªv
¬« 1
v2
v3
x0 º¼»
ª .1
= ««.6
«¬.3
2
−3
1
−1
0º ª 1
0 .3»» «« 0
1 .7 »¼ «¬ 0
0
0
1
0
0
0
1º
.1»» .
.3»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.6
• Solutions
327
x 0 = v1 + .1v 2 + .3v 3
x1 = Av1 + .1Av 2 + .3 Av 3 = v1 + .1(.5) v 2 + .3(.2) v 3 , and
xk = v1 +.1(.5)k v2 +.3(.2)k v3 . As k increases, xk approaches v1.
16. [M]
ª.90
A = «« .01
«¬.09
.01
.09 º
ª 1.0000 º
»
.90 .01» ⋅ ev = eig(A)= ««0.8900 »» . To four decimal places,
«¬ .8100 »¼
.09 .90 »¼
ª 0.9192 º
ª91/ 99 º
«
»
v1 = nulbasis(A -eye(3))= « 0.1919 » . Exact : ««19 / 99 »»
«¬1.0000 »¼
«¬ 1 »¼
ª −1º
v2 = nulbasis(A -ev(2)*eye(3))= «« 1»»
«¬ 0 »¼
ª −1º
v3 = nulbasis(A -ev(3)*eye(3))= «« 0 »»
«¬ 1»¼
The general solution of the dynamical system is x k = c1 v1 + c2 (.89) k v 2 + c3 (.81) k v 3 .
Note: When working with stochastic matrices and starting with a probability vector (having nonnegative
entries whose sum is 1), it helps to scale v1 to make its entries sum to 1. If
v1 = (91/ 209, 19 / 209, 99 / 209), or (.435, .091, .474) to three decimal places, then the weight c1 above
turns out to be 1. See the text’s discussion of Exercise 27 in Section 5.2.
ª0
17. a. A = «
¬.3
ª −λ
b. det «
¬ .3
1.6 º
.8»¼
1.6 º
= λ 2 − .8λ − .48 = 0. The eigenvalues of A are given by
»
.8 − λ ¼
.8 ± (−.8) 2 − 4(−.48) .8 ± 2.56 .8 ± 1.6
λ=
=
=
= 1.2 and − .4
2
2
2
The numbers of juveniles and adults are increasing because the largest eigenvalue is greater than
1. The eventual growth rate of each age class is 1.2, which is 20% per year.
To find the eventual relative population sizes, solve ( A − 1.2 I ) x = 0 :
ª −1.2 1.6 0º ª1 −4/3 0º x1 = (4/ 3) x2
ª4º
.
. Set v1 = « » .
« .3 −.4 0» «0
»
0
0¼ x2 is free
¬
¼ ¬
¬3¼
Eventually, there will be about 4 juveniles for every 3 adults.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
328
CHAPTER 5
• Eigenvalues and Eigenvectors
c. [M] Suppose that the initial populations are given by x 0 = (15, 10). The Study Guide describes
how to generate the trajectory for as many years as desired and then to plot the values for each
population. Let x k = ( jk , a k ). Then we need to plot the sequences {jk }, {a k }, {jk + a k }, and
{jk /a k }. Adjacent points in a sequence can be connected with a line segment. When a sequence is
plotted, the resulting graph can be captured on the screen and printed (if done on a computer) or
copied by hand onto paper (if working with a graphics calculator).
ª0
18. a. A = ««.6
«¬ 0
0
0
.75
.42 º
0 »»
.95 »¼
ª 0.0774 + 0.4063i º
b. ev = eig(A)= «« 0.0774 − 0.4063i »»
«¬1.1048
»¼
The long-term growth rate is 1.105, about 10.5 % per year.
ª 0.3801º
v = nulbasis(A -ev(3)*eye(3))= ««0.2064 »»
«¬1.0000 »¼
For each 100 adults, there will be approximately 38 calves and 21 yearlings.
Note: The MATLAB box in the Study Guide and the various technology appendices all give directions
for generating the sequence of points in a trajectory of a dynamical system. Details for producing a
graphical representation of a trajectory are also given, with several options available in MATLAB, Maple,
and Mathematica.
5.7
SOLUTIONS
1. From the “eigendata” (eigenvalues and corresponding eigenvectors) given, the eigenfunctions for the
differential equation x′ = Ax are v1e 4t and v 2 e 2t . The general solution of x′ = Ax has the form
ª −3º
ª −1º
c1 « » e4t + c2 « » e2t
¬ 1¼
¬ 1¼
ª −6º
The initial condition x(0) = « » determines c1 and c2 :
¬ 1¼
ª −3º
ª −1º
ª −6 º
c1 « » e 4(0) + c2 « » e2(0) = « »
¬ 1¼
¬ 1¼
¬ 1¼
5/ 2 º
ª −3 −1 −6 º ª 1 0
«
« 1
»
1
1¼ ¬ 0 1 −3/ 2 ¼»
¬
5 ª −3º
3 ª −1º
Thus c1 = 5/ 2, c2 = −3/ 2, and x(t ) = « » e4t − « » e2t .
2 ¬ 1¼
2 ¬ 1¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.7
• Solutions
329
2. From the eigendata given, the eigenfunctions for the differential equation x′ = Ax are v1e−3t and
v 2 e −1t . The general solution of x′ = Ax has the form
ª −1º
ª1º
c1 « » e−3t + c2 « » e−1t
¬ 1¼
¬1¼
ª 2º
The initial condition x(0) = « » determines c1 and c2 :
¬ 3¼
ª −1º
ª1º
ª 2º
c1 « » e −3(0) + c2 « » e −1(0) = « »
¬ 1¼
¬1¼
¬ 3¼
ª −1 1 2 º ª1 0 1/ 2 º
« 1 1 3» « 0 1 5/ 2 »
¬
¼ ¬
¼
1 ª −1º
5 ª1º
Thus c1 = 1/ 2, c2 = 5 / 2, and x(t ) = « » e−3t + « » e−t .
2 ¬ 1¼
2 ¬1¼
ª 2
3. A = «
¬ −1
3º
, det( A − I ) =
−2»¼
ª 1
For λ = 1: «
¬ −1
ª 3
For λ = –1: «
¬ −1
2
0º ª1
0»¼ «¬0
3
−3
3
−1
− 1 = ( − 1)( + 1) = 0. Eigenvalues: 1 and −1.
0º
ª −3º
, so x1 = −3 x2 with x2 free. Take x2 = 1 and v1 = « » .
»
0¼
¬ 1¼
3
0
0 º ª1
0»¼ «¬0
1
0
0º
ª −1º
, so x1 = − x2 with x2 free. Take x2 = 1 and v 2 = « » .
»
0¼
¬ 1¼
ª 3º
For the initial condition x(0) = « » , find c1 and c2 such that c1 v1 + c2 v 2 = x(0) :
¬ 2¼
ª¬ v1
ª −3
x(0) º¼ = «
¬ 1
v2
−1
3º ª 1
2»¼ «¬0
1
0
1
−5/ 2º
9/ 2»¼
ª −3º
ª −1º
Thus c1 = −5/ 2, c2 = 9 / 2, and x(t ) = − 5 « » et + 9 « » e−t .
2 ¬ 1¼
2 ¬ 1¼
Since one eigenvalue is positive and the other is negative, the origin is a saddle point of the
dynamical system described by x′ = Ax. The direction of greatest attraction is the line through v 2
and the origin. The direction of greatest repulsion is the line through v1 and the origin.
ª −2
4. A = «
¬ 1
−5º
, det( A − I ) =
4»¼
ª −5
For λ = 3: «
¬ 1
ª −1
For λ = –1: «
¬ 1
−5
1
−5
5
2
0 º ª1
0 »¼ «¬0
0 º ª1
0»¼ «¬0
− 2 − 3 = ( + 1)( − 3) = 0. Eigenvalues: −1 and 3.
1
0
5
0
0º
ª −1º
, so x1 = − x2 with x2 free. Take x2 = 1 and v1 = « » .
»
0¼
¬ 1¼
0º
ª −5º
, so x1 = −5 x2 with x2 free. Take x2 = 1 and v 2 = « » .
»
0¼
¬ 1¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
330
CHAPTER 5
• Eigenvalues and Eigenvectors
ª 3º
For the initial condition x(0) = « » , find c1 and c2 such that c1 v1 + c2 v 2 = x (0) :
¬ 2¼
ª¬ v1
ª −1
x(0) º¼ = «
¬ 1
v2
−5
3º ª1
2»¼ «¬0
1
0
1
13/ 4 º
−5/ 4»¼
13 ª −1º 3t 5 ª −5º −t
e − « »e .
4 «¬ 1»¼
4 ¬ 1¼
Since one eigenvalue is positive and the other is negative, the origin is a saddle point of the
dynamical system described by x′ = Ax. The direction of greatest attraction is the line through v 2
and the origin. The direction of greatest repulsion is the line through v1 and the origin.
Thus c1 = 13/ 4, c2 = −5/ 4, and x(t ) =
ª7
5. A = «
¬3
−1º
, det ( A − I ) =
3»¼
ª3
For λ = 4: «
¬3
−1
ª1
For λ = 6: «
¬3
−1
−1
−3
2
− 10 + 24 = ( − 4)( − 6) = 0. Eigenvalues: 4 and 6.
0º ª 1
0»¼ «¬0
−1/ 3
0º ª 1
0»¼ «¬0
−1
0º
ª1º
, so x1 = (1/ 3) x2 with x2 free. Take x2 = 3 and v1 = « » .
»
0¼
¬3 ¼
0
0
0º
ª1º
, so x1 = x2 with x2 free. Take x2 = 1 and v 2 = « » .
»
0¼
¬1¼
ª 3º
For the initial condition x(0) = « » , find c1 and c2 such that c1 v1 + c2 v 2 = x(0) :
¬ 2¼
ª¬ v1
ª1 1
x(0) º¼ = «
¬3 1
v2
3º ª 1
2¼» ¬«0
0
1
−1/ 2º
7 / 2»¼
1 ª1º
7 ª1º
Thus c1 = −1/ 2, c2 = 7 / 2, and x(t ) = − « » e4t + « » e6t .
2 ¬3¼
2 ¬1¼
Since both eigenvalues are positive, the origin is a repellor of the dynamical system described by
x′ = Ax. The direction of greatest repulsion is the line through v 2 and the origin.
ª1
6. A = «
¬3
−2 º
, det ( A − I ) =
−4 »¼
ª3
For λ = –2: «
¬3
−2
ª2
For λ = –1: «
¬3
−2
−2
−3
2
+ 3 + 2 = ( + 1)( + 2) = 0. Eigenvalues: −1 and −2.
0º ª 1
0»¼ «¬0
−2/ 3
0º ª 1
0»¼ «¬0
−1
0
0
0º
ª 2º
, so x1 = (2 / 3) x2 with x2 free. Take x2 = 3 and v1 = « » .
»
0¼
¬ 3¼
0º
ª1º
, so x1 = x2 with x2 free. Take x2 = 1 and v 2 = « » .
»
0¼
¬1¼
ª 3º
For the initial condition x(0) = « » , find c1 and c2 such that c1 v1 + c2 v 2 = x (0) :
¬ 2¼
[ v1
v2
ª2
x(0)] = «
¬3
1
1
3º ª 1
2»¼ «¬0
0
1
−1º
5»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.7
• Solutions
331
ª 2º
ª1º
Thus c1 = −1, c2 = 5, and x(t ) = − « » e−2t + 5 « » e−t .
¬3¼
¬1¼
Since both eigenvalues are negative, the origin is an attractor of the dynamical system described by
x′ = Ax. The direction of greatest attraction is the line through v1 and the origin.
−1º
ª1º
ª1º
, with eigenvectors v1 = « » and v 2 = « » corresponding to
»
3¼
¬3 ¼
¬1¼
ª1 1º
eigenvalues 4 and 6 respectively. To decouple the equation x′ = Ax, set P = [ v1 v 2 ] = «
» and let
¬3 1¼
ª 4 0º
−1
−1
D=«
» , so that A = PDP and D = P AP. Substituting x(t ) = Py (t ) into x′ = Ax we have
0
6
¬
¼
ª7
7. From Exercise 5, A = «
¬3
d
( P y ) = A( P y ) = PDP −1 ( P y ) = PD y
dt
Since P has constant entries,
d
dt
( P y ) = P ( dtd (y )), so that left-multiplying the equality
P ( dtd (y )) = PD y by P −1 yields y′ = D y , or
ª y1′ (t ) º ª 4
« y′ (t ) » = « 0
¬ 2 ¼ ¬
0º ª y1 (t ) º
6»¼ «¬ y2 (t ) »¼
ª1
8. From Exercise 6, A = «
¬3
−2 º
ª 2º
ª1º
, with eigenvectors v1 = « » and v 2 = « » corresponding to
»
−4 ¼
¬ 3¼
¬1¼
eigenvalues −2 and −1 respectively. To decouple the equation x′ = Ax, set P = ª¬« v1
ª −2
and let D = «
¬ 0
we have
ª2
v2 º¼» = «
¬3
1º
1»¼
0º
, so that A = PDP −1 and D = P −1 AP. Substituting x(t ) = Py (t ) into x′ = Ax
−1»¼
d
( Py ) = A( Py ) = PDP −1 ( Py ) = PDy
dt
Since P has constant entries,
d
dt
( Py ) = P ( dtd (y ) ) , so that left-multiplying the equality
P ( dtd (y ) ) = PDy by P −1 yields y′ = Dy , or
ª y1′ (t ) º ª −2
« y′ (t ) » = « 0
¬ 2 ¼ ¬
0º ª y1 (t ) º
−1»¼ «¬ y2 (t ) »¼
2º
ª −3
ª1 − i º
. An eigenvalue of A is −2 + i with corresponding eigenvector v = «
9. A = «
»
» . The
¬ −1 −1¼
¬ 1 ¼
complex eigenfunctions ve λt and ve λt form a basis for the set of all complex solutions to x′ = Ax.
The general complex solution is
ª1 − i º ( −2+i ) t
ª1 + i º ( −2 −i )t
c1 «
e
+ c2 «
»
»e
¬ 1 ¼
¬ 1 ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
332
CHAPTER 5
• Eigenvalues and Eigenvectors
where c1 and c2 are arbitrary complex numbers. To build the general real solution, rewrite ve( −2 + i )t
as:
ª1 − i º −2t it ª1 − i º −2t
ve( −2+ i )t = «
» e e = « 1 » e (cos t + i sin t )
¬ 1 ¼
¬
¼
ª cos t − i cos t + i sin t − i 2 sin t º −2t
=«
»e
cos t + i sin t
¬
¼
ª cos t + sin t º −2t ªsin t − cos t º −2t
=«
» e +i « sin t » e
¬ cos t ¼
¬
¼
The general real solution has the form
ªcos t + sin t º −2t
ªsin t − cos t º −2t
c1 «
e + c2 «
»
»e
¬ cos t ¼
¬ sin t ¼
where c1 and c2 now are real numbers. The trajectories are spirals because the eigenvalues are
complex. The spirals tend toward the origin because the real parts of the eigenvalues are negative.
ª3
10. A = «
¬ −2
1º
ª1 + i º
. An eigenvalue of A is 2 + i with corresponding eigenvector v = «
»
» . The complex
1¼
¬ −2 ¼
eigenfunctions ve t and ve
general complex solution is
t
form a basis for the set of all complex solutions to x′ = Ax. The
ª1 + i º (2+i )t
ª1 − i º (2−i )t
c1 «
e
+ c2 «
»
»e
¬ −2 ¼
¬ −2 ¼
where c1 and c2 are arbitrary complex numbers. To build the general real solution, rewrite ve (2 + i )t
as:
ª1 + i º 2t it ª1 + i º 2t
ve(2+i )t = «
» e e = « −2 » e (cos t + i sin t )
¬ −2 ¼
¬
¼
ªcos t + i cos t + i sin t + i 2 sin t º 2t
=«
»e
−2cos t − 2i sin t
¬
¼
ªcos t − sin t º 2t ªsin t + cos t º 2t
=«
» e + i « −2sin t » e
¬ −2cos t ¼
¬
¼
The general real solution has the form
ªcos t − sin t º 2t
ªsin t + cos t º 2t
c1 «
e + c2 «
»
»e
¬ −2cos t ¼
¬ −2sin t ¼
where c1 and c2 now are real numbers. The trajectories are spirals because the eigenvalues are
complex. The spirals tend away from the origin because the real parts of the eigenvalues are positive.
ª −3
11. A = «
¬ 2
−9º
ª −3 + 3i º
. An eigenvalue of A is 3i with corresponding eigenvector v = «
»
» . The
3¼
¬ 2 ¼
complex eigenfunctions ve t and ve
The general complex solution is
t
form a basis for the set of all complex solutions to x′ = Ax.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.7
• Solutions
333
ª −3 + 3i º (3i )t
ª −3 − 3i º ( −3i )t
c1 «
e
+ c2 «
»
»e
¬ 2 ¼
¬ 2 ¼
where c1 and c2 are arbitrary complex numbers. To build the general real solution, rewrite ve (3i ) t
as:
ª −3 + 3i º
ve(3i )t = «
» (cos3t + i sin 3t )
¬ 2 ¼
ª −3cos3t − 3sin 3t º ª −3sin 3t + 3cos3t º
=«
» +i«
»
2cos3t
2sin 3t
¬
¼ ¬
¼
The general real solution has the form
ª −3cos3t − 3sin 3t º
ª −3sin 3t + 3cos3t º
c1 «
+ c2 «
»
»
2cos3t
2sin 3t
¬
¼
¬
¼
where c1 and c2 now are real numbers. The trajectories are ellipses about the origin because the real
parts of the eigenvalues are zero.
ª −7
12. A = «
¬ −4
10º
ª3 − i º
. An eigenvalue of A is −1 + 2i with corresponding eigenvector v = «
»
» . The
5¼
¬ 2 ¼
complex eigenfunctions ve t and ve
The general complex solution is
t
form a basis for the set of all complex solutions to x′ = Ax.
ª3 − i º ( −1+ 2i )t
ª3 + i º ( −1− 2i )t
c1 «
e
+ c2 «
»
»e
¬ 2 ¼
¬ 2 ¼
where c1 and c2 are arbitrary complex numbers. To build the general real solution, rewrite ve( −1+ 2i )t
as:
ª3 − i º − t
ve( −1+ 2i )t = «
» e (cos 2t + i sin 2t )
¬ 2 ¼
ª3cos 2t + sin 2t º −t ª3sin 2t − cos 2t º −t
=«
»e + i«
»e
2cos 2t
2sin 2t
¬
¼
¬
¼
The general real solution has the form
ª3cos 2t + sin 2t º −t
ª3sin 2t − cos 2t º −t
c1 «
e + c2 «
»
»e
2cos 2t
2sin 2t
¬
¼
¬
¼
where c1 and c2 now are real numbers. The trajectories are spirals because the eigenvalues are
complex. The spirals tend toward the origin because the real parts of the eigenvalues are negative.
ª4
13. A = «
¬6
−3º
ª1 + i º
. An eigenvalue of A is 1 + 3i with corresponding eigenvector v = «
»
» . The
−2 ¼
¬ 2 ¼
complex eigenfunctions ve t and ve
The general complex solution is
t
form a basis for the set of all complex solutions to x′ = Ax.
ª1 + i º (1+ 3i )t
ª1 − i º (1−3i )t
c1 «
e
+ c2 «
»
»e
¬ 2 ¼
¬ 2 ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
334
CHAPTER 5
• Eigenvalues and Eigenvectors
where c1 and c2 are arbitrary complex numbers. To build the general real solution, rewrite ve (1+ 3i )t
as:
ª1 + i º t
ve(1+3i )t = «
» e (cos 3t + i sin 3t )
¬ 2 ¼
ªcos3t − sin 3t º t ªsin 3t + cos 3t º t
=«
» e + i « 2sin 3t » e
¬ 2cos3t ¼
¬
¼
The general real solution has the form
ªcos3t − sin 3t º t
ªsin 3t + cos3t º t
c1 «
e + c2 «
»
»e
¬ 2cos3t ¼
¬ 2sin 3t ¼
where c1 and c2 now are real numbers. The trajectories are spirals because the eigenvalues are
complex. The spirals tend away from the origin because the real parts of the eigenvalues are positive.
ª −2
14. A = «
¬ −8
1º
ª1 − i º
. An eigenvalue of A is 2i with corresponding eigenvector v = «
»
» . The complex
2¼
¬ 4 ¼
eigenfunctions ve t and ve
general complex solution is
t
form a basis for the set of all complex solutions to x′ = Ax. The
ª1 − i º (2i )t
ª1 + i º ( −2i ) t
c1 «
e
+ c2 «
»
»e
¬ 4 ¼
¬ 4 ¼
where c1 and c2 are arbitrary complex numbers. To build the general real solution, rewrite ve (2 i )t
as:
ª1 − i º
ve(2i )t = «
» (cos 2t + i sin 2t )
¬ 4 ¼
ª cos 2t + sin 2t º ªsin 2t − cos 2t º
=«
» +i«
»
¬ 4cos 2t ¼ ¬ 4sin 2t ¼
The general real solution has the form
ªcos 2t + sin 2t º
ªsin 2t − cos 2t º
c1 «
+ c2 «
»
»
¬ 4cos 2t ¼
¬ 4sin 2t ¼
where c1 and c2 now are real numbers. The trajectories are ellipses about the origin because the real
parts of the eigenvalues are zero.
ª −8
15. [M] A = «« 2
«¬ 7
−12
1
12
−6 º
2 »» . The eigenvalues of A are:
5»¼
ev = eig(A)=
1.0000
-1.0000
-2.0000
nulbasis(A-ev(1)*eye(3)) =
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.7
• Solutions
335
-1.0000
0.2500
1.0000
ª− 4 º
so that v1 = «« 1»»
«¬ 4 »¼
nulbasis(A-ev(2)*eye(3)) =
-1.2000
0.2000
1.0000
ª −6 º
so that v 2 = «« 1»»
«¬ 5»¼
nulbasis (A-ev(3)*eye(3)) =
-1.0000
0.0000
1.0000
ª −1º
so that v 3 = «« 0 »»
«¬ 1»¼
ª −4 º
ª −6 º
ª −1º
«
»
«
»
t
−t
Hence the general solution is x(t ) = c1 « 1» e + c2 « 1» e + c3 «« 0»» e−2t . The origin is a saddle point.
«¬ 4 »¼
«¬ 5»¼
«¬ 1»¼
A solution with c1 = 0 is attracted to the origin while a solution with c2 = c3 = 0 is repelled.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
336
CHAPTER 5
ª −6
16. [M] A = «« 2
«¬ −4
• Eigenvalues and Eigenvectors
−11
5
−5
16 º
−4 »» . The eigenvalues of A are:
10 »¼
ev = eig(A)=
4.0000
3.0000
2.0000
nulbasis(A-ev(1)*eye(3)) =
2.3333
-0.6667
1.0000
ª 7º
so that v1 = «« −2 »»
¬« 3¼»
nulbasis(A-ev(2)*eye(3)) =
3.0000
-1.0000
1.0000
ª 3º
so that v 2 = «« −1»»
«¬ 1»¼
nulbasis(A-ev(3)*eye(3)) =
2.0000
0.0000
1.0000
ª2º
so that v3 = ««0 »»
¬«1 ¼»
ª 7º
ª 3º
ª 2º
«
»
«
»
4t
3t
Hence the general solution is x(t ) = c1 « −2 » e + c2 « −1» e + c3 «« 0 »» e2t . The origin is a repellor,
«¬ 3»¼
«¬ 1»¼
«¬ 1 »¼
because all eigenvalues are positive. All trajectories tend away from the origin.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.7
ª 30
17. [M] A = «« −11
«¬ 6
64
−23
15
• Solutions
23º
−9 »» . The eigenvalues of A are:
4 »¼
ev = eig(A)=
5.0000 + 2.0000i
5.0000 - 2.0000i
1.0000
nulbasis(A-ev(1)*eye(3)) =
7.6667 - 11.3333i
-3.0000 + 4.6667i
1.0000
ª 23 − 34i º
so that v1 = «« −9 + 14i »»
«¬ 3 »¼
nulbasis (A-ev(2)*eye(3)) =
7.6667 + 11.3333i
-3.0000 - 4.6667i
1.0000
ª 23 + 34i º
so that v 2 = «« −9 − 14i »»
«¬ 3 »¼
nulbasis (A-ev(3)*eye(3)) =
-3.0000
1.0000
1.0000
ª −3º
so that v 3 = «« 1»»
«¬ 1»¼
Hence the general complex solution is
ª 23 − 34i º
ª 23 + 34i º
ª −3º
«
»
«
»
(5 + 2i ) t
(5− 2 i ) t
x(t ) = c1 « −9 + 14i » e
+ c2 « −9 − 14i » e
+ c3 «« 1»» et
«¬ 3 »¼
«¬ 3 »¼
«¬ 1»¼
Rewriting the first eigenfunction yields
ª 23 − 34i º
ª 23cos 2t + 34sin 2t º
ª 23sin 2t − 34cos 2t º
« −9 + 14i » e5t (cos 2t + i sin 2t ) = « −9cos 2t − 14sin 2t » e5t + i « −9sin 2t + 14cos 2t » e5t
«
»
«
»
«
»
«¬ 3 »¼
«¬
»¼
«¬
»¼
3cos 2t
3sin 2t
Hence the general real solution is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
337
338
CHAPTER 5
• Eigenvalues and Eigenvectors
ª 23cos 2t + 34sin 2t º
ª 23sin 2t − 34cos 2t º
ª −3º
«
»
«
»
5t
5t
x(t ) = c1 « −9cos 2t − 14sin 2t » e + c2 « −9sin 2t + 14cos 2t » e + c3 «« 1»» et
«¬
»¼
«¬
»¼
«¬ 1»¼
3cos 2t
3sin 2t
where c1 , c2 , and c3 are real. The origin is a repellor, because the real parts of all eigenvalues are
positive. All trajectories spiral away from the origin.
ª 53
18. [M] A = «« 90
«¬ 20
−30
−52
−10
−2 º
−3»» . The eigenvalues of A are:
2 »¼
ev = eig(A)=
-7.0000
5.0000 + 1.0000i
5.0000 - 1.0000i
nulbasis(A-ev(1)*eye(3)) =
0.5000
1.0000
0.0000
ª1 º
so that v1 = «« 2 »»
«¬ 0 »¼
nulbasis(A-ev(2)*eye(3)) =
0.6000 + 0.2000i
0.9000 + 0.3000i
1.0000
ª6 + 2i º
so that v 2 = «« 9 + 3i »»
«¬ 10 »¼
nulbasis(A-ev(3)*eye(3)) =
0.6000 - 0.2000i
0.9000 - 0.3000i
1.0000
ª6 − 2i º
so that v 3 = «« 9 − 3i »»
«¬ 10 »¼
Hence the general complex solution is
ª1 º
ª 6 + 2i º
ª 6 − 2i º
«
»
«
»
−7 t
(5+ i ) t
x(t ) = c1 « 2 » e + c2 « 9 + 3i » e
+ c3 «« 9 − 3i »» e(5−i ) t
«¬0 »¼
«¬ 10 »¼
«¬ 10 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.7
• Solutions
339
Rewriting the second eigenfunction yields
ª6 + 2i º
ª 6cos t − 2sin t º
ª 6sin t + 2cos t º
« 9 + 3i » e5t (cos t + i sin t ) = « 9cos t − 3sin t » e5t + i « 9sin t + 3cos t » e5t
«
»
«
»
«
»
«¬ 10 »¼
«¬ 10cos t »¼
«¬ 10sin t
»¼
Hence the general real solution is
ª1 º
ª6cos t − 2sin t º
ª 6sin t + 2cos t º
«
»
«
»
−7 t
5t
x(t ) = c1 « 2 » e + c2 « 9cos t − 3sin t » e + c3 «« 9sin t + 3cos t »» e5t
«¬ 0 »¼
«¬ 10cos t »¼
«¬ 10sin t
»¼
where c1 , c2 , and c3 are real. When c2 = c3 = 0 the trajectories tend toward the origin, and in other
cases the trajectories spiral away from the origin.
19. [M] Substitute R1 = 1/ 5, R2 = 1/ 3, C1 = 4, and C 2 = 3 into the formula for A given in Example 1, and
use a matrix program to find the eigenvalues and eigenvectors:
ª −2
A=«
¬ 1
3/ 4º
,
−1¼»
1
ª1 º
= −.5 : v1 = « » ,
¬ 2¼
2
ª −3º
= −2.5 : v1 = « »
¬ 2¼
ª1 º
ª −3º
ª 4º
The general solution is thus x(t ) = c1 « » e−.5t + c2 « » e−2.5t. The condition x(0) = « » implies that
¬ 2¼
¬ 2¼
¬ 4¼
ª
º
ª 1 −3º « c1 » ª 4 º
. By a matrix program, c1 = 5 / 2 and c2 = −1/ 2, so that
« »=
«2
2 »¼ «¬ c2 »¼ «¬ 4 »¼
¬
ª v1 (t ) º
5 ª1 º −.5t 1 ª −3º −2.5t
«v (t ) » = x(t ) = « 2» e − « 2» e
2¬ ¼
2¬ ¼
¬ 2 ¼
20. [M] Substitute R1 = 1 / 15, R2 = 1 / 3, C1 = 9, and C2 = 2 into the formula for A given in Example 1,
and use a matrix program to find the eigenvalues and eigenvectors:
ª −2
A=«
¬3/ 2
1/ 3º
,
−3/ 2¼»
1
ª1º
= −1 : v1 = « » ,
¬3¼
2
ª −2º
= −2.5 : v 2 = « »
¬ 3¼
ª1º
ª −2 º
ª 3º
The general solution is thus x(t ) = c1 « » e−t + c2 « » e−2.5t. The condition x(0) = « » implies
¬3¼
¬ 3¼
¬ 3¼
ª 1 −2 º ª« c1 º» ª3º
. By a matrix program, c1 = 5 / 3 and c2 = −2 / 3, so that
that «
« »=
3»¼ «¬ c2 »¼ «¬3»¼
¬3
ª v1 (t ) º
5 ª1º −t 2 ª −2 º −2.5t
«v (t ) » = x(t ) = «3» e − « 3» e
3¬ ¼
3¬ ¼
¬ 2 ¼
ª −1
21. [M] A = «
¬ 5
−8º
. Using a matrix program we find that an eigenvalue of A is −3 + 6i with
−5»¼
ª 2 + 6i º
corresponding eigenvector v = «
» . The conjugates of these form the second
¬ 5 ¼
eigenvalue-eigenvector pair. The general complex solution is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
340
CHAPTER 5
• Eigenvalues and Eigenvectors
ª 2 + 6i º ( −3+6i )t
ª 2 − 6i º ( −3−6i ) t
x(t ) = c1 «
e
+ c2 «
»
»e
¬ 5 ¼
¬ 5 ¼
where c1 and c2 are arbitrary complex numbers. Rewriting the first eigenfunction and taking its real
and imaginary parts, we have
ª 2 + 6i º −3t
ve( −3+ 6i )t = «
» e (cos 6t + i sin 6t )
¬ 5 ¼
ª 2cos 6t − 6sin 6t º −3t ª 2sin 6t + 6cos 6t º −3t
=«
»e + i«
»e
5cos 6t
5sin 6t
¬
¼
¬
¼
The general real solution has the form
ª 2cos 6t − 6sin 6t º −3t
ª 2sin 6t + 6cos6t º −3t
x(t ) = c1 «
e + c2 «
»
»e
5cos 6t
5sin 6t
¬
¼
¬
¼
ª 0º
where c1 and c2 now are real numbers. To satisfy the initial condition x(0) = « » , we solve
¬15¼
ª2º
ª6º ª 0 º
c1 « » + c2 « » = « » to get c1 = 3, c2 = −1. We now have
¬5 ¼
¬0¼ ¬15¼
−20sin 6t
ª iL (t ) º
ª 2cos 6t − 6sin 6t º −3t ª 2sin 6t + 6cos 6t º −3t ª
º −3t
e −«
e =«
«v (t ) » = x(t ) = 3 «
»
»
»e
5cos 6t
5sin 6t
¬
¼
¬
¼
¬15cos 6t − 5sin 6t ¼
¬ C ¼
ª 0
22. [M] A = «
¬ −.4
2º
. Using a matrix program we find that an eigenvalue of A is −.4 + .8i with
−.8»¼
ª −1 − 2i º
corresponding eigenvector v = «
» . The conjugates of these form the second eigenvalue¬ 1 ¼
eigenvector pair. The general complex solution is
ª −1 − 2i º ( −.4+.8i )t
ª −1 + 2i º ( −.4 −.8i )t
x(t ) = c1 «
e
+ c2 «
»
»e
¬ 1 ¼
¬ 1 ¼
where c1 and c2 are arbitrary complex numbers. Rewriting the first eigenfunction and taking its real
and imaginary parts, we have
ª −1 − 2i º −.4t
ve( −.4+.8i ) t = «
» e (cos .8t + i sin .8t )
¬ 1 ¼
ª − cos .8t + 2sin .8t º −.4t ª − sin .8t − 2cos .8t º −.4t
=«
»e + i«
»e
cos .8t
sin .8t
¬
¼
¬
¼
The general real solution has the form
ª − cos .8t + 2sin .8t º −.4t
ª − sin .8t − 2cos .8t º −.4t
x(t ) = c1 «
e + c2 «
»
»e
cos .8t
sin .8t
¬
¼
¬
¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.8
• Solutions
341
ª 0º
where c1 and c2 now are real numbers. To satisfy the initial condition x(0) = « » , we solve
¬12¼
ª −1º
ª −2º ª 0 º
c1 « » + c2 « » = « » to get c1 = 12, c2 = −6. We now have
¬ 1¼
¬ 0¼ ¬12¼
30sin .8t
ª iL (t ) º
ª − cos .8t + 2sin .8t º −.4t
ª − sin .8t − 2cos .8t º −.4t ª
º −.4t
e − 6«
e =«
«v (t ) » = x(t ) = 12 «
»
»
»e
cos .8t
sin .8t
¬
¼
¬
¼
¬12cos .8t − 6sin .8t ¼
¬ C ¼
5.8
SOLUTIONS
1. The vectors in the given sequence approach an eigenvector v1 . The last vector in the sequence,
ª 1 º
x4 = «
» , is probably the best estimate for v1 . To compute an estimate for λ1 , examine
¬.3326¼
ª 4.9978º
Ax4 = «
» . This vector is approximately λ1 v1. From the first entry in this vector, an estimate
¬1.6652 ¼
of λ1 is 4.9978.
2. The vectors in the given sequence approach an eigenvector v1 . The last vector in the sequence,
ª −.2520º
x4 = «
» , is probably the best estimate for v1 . To compute an estimate for λ1 , examine
¬ 1 ¼
ª −1.2536º
Ax4 = «
» . This vector is approximately λ1 v1. From the second entry in this vector, an
¬ 5.0064 ¼
estimate of λ1 is 5.0064.
3. The vectors in the given sequence approach an eigenvector v1 . The last vector in the sequence,
ª.5188º
x4 = «
» , is probably the best estimate for v1 . To compute an estimate for λ1 , examine
¬ 1 ¼
ª.4594º
Ax4 = «
» . This vector is approximately λ1 v1. From the second entry in this vector, an estimate
¬.9075¼
of λ1 is .9075.
4. The vectors in the given sequence approach an eigenvector v1 . The last vector in the sequence,
ª 1 º
x4 = «
» , is probably the best estimate for v1 . To compute an estimate for λ1 , examine
¬.7502¼
ª −.4012 º
Ax4 = «
» . This vector is approximately λ1 v1. From the first entry in this vector, an estimate
¬ −.3009 ¼
of λ1 is −.4012.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
342
CHAPTER 5
• Eigenvalues and Eigenvectors
ª 24991º
5. Since A5 x = «
» is an estimate for an eigenvector, the vector
¬ −31241¼
1 ª 24991º ª −.7999º
v=−
=
is a vector with a 1 in its second entry that is close to an
1»¼
31241 «¬ −31241»¼ «¬
ª 4.0015º
eigenvector of A. To estimate the dominant eigenvalue λ1 of A, compute Av = «
» . From the
¬ −5.0020 ¼
second entry in this vector, an estimate of λ1 is −5.0020.
ª −2045º
1 ª −2045º ª −.4996º
=
6. Since A5 x = «
is an estimate for an eigenvector, the vector v =
is
»
1»¼
4093 «¬ 4093»¼ «¬
¬ 4093¼
a vector with a 1 in its second entry that is close to an eigenvector of A. To estimate the dominant
ª −2.0008º
eigenvalue λ1 of A, compute Av = «
» . From the second entry in this vector, an estimate of
¬ 4.0024 ¼
λ1 is 4.0024.
ª6 7 º
ª1 º
, x0 = « » . The data in the table below was calculated using Mathematica, which
7. [M] A = «
»
¬8 5 ¼
¬0¼
carried more digits than shown here.
k
0
1
2
3
4
5
xk
ª1 º
«0»
¬ ¼
ª.75º
«1»
¬ ¼
ª 1 º
«.9565»
¬
¼
ª.9932º
« 1 »
¬
¼
ª 1 º
«.9990»
¬
¼
ª.9998º
« 1 »
¬
¼
Ax k
ª6º
«8 »
¬ ¼
ª11.5º
«11.0 »
¬
¼
ª12.6957 º
«12.7826 »
¬
¼
ª12.9592 º
«12.9456 »
¬
¼
ª12.9927 º
«12.9948 »
¬
¼
ª12.9990 º
«12.9987 »
¬
¼
µk
8
11.5
12.7826
12.9592
12.9948
12.9990
The actual eigenvalue is 13.
ª 2 1º
ª1 º
, x0 = « » . The data in the table below was calculated using Mathematica, which
8. [M] A = «
»
¬ 4 5¼
¬0¼
carried more digits than shown here.
k
0
1
2
3
4
5
xk
ª1 º
«0»
¬ ¼
ª.5º
«1»
¬ ¼
ª.2857 º
« 1 »
¬
¼
ª.2558º
« 1 »
¬
¼
ª.2510º
« 1 »
¬
¼
ª.2502º
« 1 »
¬
¼
Ax k
ª2º
«4»
¬ ¼
ª2º
«7 »
¬ ¼
ª1.5714 º
«6.1429 »
¬
¼
ª1.5116 º
«6.0233»
¬
¼
ª1.5019 º
«6.0039 »
¬
¼
ª1.5003 º
«6.0006 »
¬
¼
µk
4
7
6.1429
6.0233
6.0039
6.0006
The actual eigenvalue is 6.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.8
ª8
9. [M] A = «« 1
«¬0
• Solutions
12 º
ª1 º
»
−2
1» , x0 = «« 0 »» . The data in the table below was calculated using Mathematica,
«¬0 »¼
3 0 »¼
which carried more digits than shown here.
0
k
0
1
2
3
4
5
6
xk
ª1 º
«0»
« »
«¬ 0 »¼
ª 1 º
«.125»
«
»
«¬ 0 »¼
ª 1 º
«.0938»
«
»
«¬.0469»¼
ª 1 º
«.1004»
«
»
«¬.0328»¼
ª 1 º
«.0991»
«
»
¬«.0359¼»
ª 1 º
«.0994»
«
»
«¬.0353»¼
ª 1 º
«.0993»
«
»
«¬.0354»¼
Ax k
ª8 º
«1 »
« »
«¬ 0 »¼
ª 8º
« .75»
«
»
«¬.375»¼
ª8.5625º
« .8594 »
«
»
«¬ .2812 »¼
ª8.3942 º
« .8321»
«
»
«¬ .3011»¼
ª8.4304 º
« .8376 »
«
»
«¬ .2974 »¼
ª8.4233º
« .8366 »
«
»
«¬ .2981»¼
ª8.4246 º
« .8368»
«
»
«¬ .2979 »¼
µk
8
8
8.5625
8.3942
8.4304
8.4233
8.4246
Thus µ5 = 8.4233 and µ 6 = 8.4246. The actual eigenvalue is (7 + 97 ) / 2, or 8.42443 to five
decimal places.
ª 1 2 −2 º
ª1 º
«
»
10. [M] A = « 1 1
9 » , x0 = «« 0 »» . The data in the table below was calculated using Mathematica,
«¬0 1
«¬ 0 »¼
9 »¼
which carried more digits than shown here.
k
0
1
2
3
4
5
6
xk
ª1 º
«0»
« »
«¬ 0 »¼
ª1 º
«1 »
« »
«¬ 0 »¼
ª 1 º
«.6667 »
«
»
«¬.3333»¼
ª .3571º
« 1 »
«
»
«¬.7857 »¼
ª.0932º
« 1 »
«
»
«¬.9576»¼
ª.0183º
« 1 »
«
»
«¬.9904»¼
ª.0038º
« 1 »
«
»
«¬.9982»¼
Ax k
ª1 º
«1 »
« »
«¬ 0 »¼
ª3º
«2»
« »
«¬ 1 »¼
ª1.6667 º
« 4.6667 »
«
»
«¬3.6667 »¼
ª .7857 º
«8.4286 »
«
»
«¬8.0714 »¼
ª .1780 º
«9.7119 »
«
»
«¬9.6186 »¼
ª .0375 º
«9.9319 »
«
»
«¬9.9136 »¼
ª .0075 º
«9.9872 »
«
»
«¬9.9834 »¼
µk
1
3
4.6667
8.4286
9.7119
9.9319
9.9872
Thus µ 5 = 9.9319 and µ 6 = 9.9872. The actual eigenvalue is 10.
ª5 2º
ª1 º
, x0 = « » . The data in the table below was calculated using Mathematica, which
11. [M] A = «
»
¬ 2 2¼
¬0¼
carried more digits than shown here.
k
0
1
2
3
4
xk
ª1 º
«0»
¬ ¼
ª1º
«.4»
¬ ¼
ª 1 º
«.4828»
¬
¼
ª 1 º
«.4971»
¬
¼
ª 1 º
«.4995»
¬
¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
343
344
CHAPTER 5
• Eigenvalues and Eigenvectors
Ax k
ª5º
«2»
¬ ¼
ª5.8 º
« 2.8»
¬ ¼
ª5.9655 º
« 2.9655»
¬
¼
ª5.9942 º
« 2.9942»
¬
¼
ª5.9990 º
« 2.9990»
¬
¼
µk
5
5.8
5.9655
5.9942
5.9990
R (x k )
5
5.9655
5.9990
5.99997
5.9999993
The actual eigenvalue is 6. The bottom two columns of the table show that R ( x k ) estimates the
eigenvalue more accurately than µ k .
ª −3 2º
ª1 º
, x0 = « » . The data in the table below was calculated using Mathematica,
12. [M] A = «
»
¬ 2 2¼
¬0¼
which carried more digits than shown here.
k
0
1
2
3
4
xk
ª1 º
«0»
¬ ¼
ª 1 º
«
»
¬ −.6667 ¼
ª 1 º
« −.4615»
¬
¼
ª 1 º
« −.5098»
¬
¼
ª 1 º
«
»
¬ −.4976 ¼
Ax k
ª −3º
« 2»
¬ ¼
ª −4.3333º
« 2.0000»
¬
¼
ª −3.9231º
« 2.0000»
¬
¼
ª −4.0196º
« 2.0000»
¬
¼
ª −3.9951º
« 2.0000»
¬
¼
µk
−3
−4.3333
−3.9231
−4.0196
−3.9951
R (x k )
−3
−3.9231
−3.9951
−3.9997
−3.99998
The actual eigenvalue is −4. The bottom two columns of the table show that R ( x k ) estimates the
eigenvalue more accurately than µ k .
13. If the eigenvalues close to 4 and −4 have different absolute values, then one of these is a strictly
dominant eigenvalue, so the power method will work. But the power method depends on powers of
the quotients λ 2 /λ1 and λ 3 /λ1 going to zero. If | λ 2 /λ1 | is close to 1, its powers will go to zero
slowly, and the power method will converge slowly.
14. If the eigenvalues close to 4 and −4 have the same absolute value, then neither of these is a strictly
dominant eigenvalue, so the power method will not work. However, the inverse power method may
still be used. If the initial estimate is chosen near the eigenvalue close to 4, then the inverse power
method should produce a sequence that estimates the eigenvalue close to 4.
15. Suppose Ax = λx, with x ≠ 0. For any α , Ax − α Ix = (λ − α )x. If α is not an eigenvalue of A, then
A − αI is invertible and λ − α is not 0; hence
x = ( A − α I ) −1 (λ − α )x and (λ − α )−1 x = ( A − α I ) −1 x
This last equation shows that x is an eigenvector of ( A − αI ) −1 corresponding to the eigenvalue
(λ − α ) −1.
16. Suppose that µ is an eigenvalue of ( A − αI ) −1 with corresponding eigenvector x. Since
( A − α I )−1 x = µ x,
x = ( A − α I )( µ x) = A( µ x) − (α I )( µ x) = µ ( Ax) − αµ x
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.8
• Solutions
Solving this equation for Ax, we find that
§1·
§
1·
Ax = ¨ ¸ (αµ x + x) = ¨ α + ¸ x
µ¹
©µ¹
©
Thus λ = α + (1/µ ) is an eigenvalue of A with corresponding eigenvector x.
ª 10 −8 −4 º
ª1 º
«
»
17. [M] A = « −8 13
4 » , x 0 = «« 0 »» ,α = 3.3. The data in the table below was calculated using
«¬ −4
«¬ 0 »¼
5
4 »¼
Mathematica, which carried more digits than shown here.
k
0
1
2
xk
ª1 º
«0»
« »
«¬ 0 »¼
ª 1 º
«.7873»
«
»
«¬.0908»¼
ª 1 º
«.7870 »
«
»
«¬.0957 »¼
yk
ª 26.0552º
« 20.5128»
«
»
«¬ 2.3669»¼
ª 47.1975º
«37.1436 »
«
»
«¬ 4.5187 »¼
ª 47.1233º
«37.0866 »
«
»
«¬ 4.5083»¼
µk
26.0552
47.1975
47.1233
νk
3.3384
3.32119
3.3212209
Thus an estimate for the eigenvalue to four decimal places is 3.3212. The actual eigenvalue is
(25 − 337) / 2, or 3.3212201 to seven decimal places.
0 12 º
ª8
ª1 º
«
»
18. [M] A = « 1 −2
1» , x 0 = ««0 »» ,α = −1.4. The data in the table below was calculated using
«¬0
«¬0 »¼
3 0 »¼
Mathematica, which carried more digits than shown here.
k
0
1
2
3
4
xk
ª1 º
«0»
« »
«¬ 0 »¼
ª 1 º
« .3646 »
«
»
«¬ −.7813»¼
ª 1 º
« .3734 »
«
»
«¬ −.7855»¼
ª 1 º
« .3729 »
«
»
«¬ −.7854 »¼
ª 1 º
« .3729 »
«
»
«¬ −.7854 »¼
yk
40 º
ª
«14.5833»
«
»
«¬ −31.25»¼
ª −38.125º
«−14.2361»
«
»
«¬ 29.9479»¼
ª−41.1134º
«−15.3300»
«
»
«¬ 32.2888»¼
ª −40.9243º
«−15.2608»
«
»
«¬ 32.1407 »¼
ª −40.9358º
«−15.2650»
«
»
«¬ 32.1497 »¼
µk
40
−38.125
−41.1134
−40.9243
−40.9358
νk
−1.375
−1.42623
−1.42432
−1.42444
−1.42443
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
345
346
CHAPTER 5
• Eigenvalues and Eigenvectors
Thus an estimate for the eigenvalue to four decimal places is −1.4244. The actual eigenvalue is
(7 − 97) / 2, or −1.424429 to six decimal places.
ª10
«7
19. [M] A = «
«8
«
«¬ 7
7
8
5
6
6
5
10
9
7º
ª1 º
»
«0 »
5»
, x0 = « » .
«0 »
9»
»
« »
10 »¼
«¬0 »¼
(a) The data in the table below was calculated using Mathematica, which carried more digits than
shown here.
k
0
1
2
3
xk
ª1 º
«0»
« »
«0»
« »
«¬ 0 »¼
ª1º
«.7 »
« »
«.8 »
« »
«¬.7 »¼
ª.988679 º
«.709434 »
«
»
« 1 »
«
»
«¬.932075»¼
ª.961467 º
« .691491»
«
»
« 1 »
«
»
«¬ .942201»¼
Ax k
ª10º
« 7»
« »
« 8»
« »
¬« 7 ¼»
ª 26.2 º
«18.8 »
«
»
« 26.5 »
«
»
¬« 24.7 ¼»
ª 29.3774 º
« 21.1283»
«
»
« 30.5547 »
«
»
¬« 28.7887 ¼»
ª 29.0505 º
« 20.8987 »
«
»
« 30.3205 »
«
»
¬« 28.6097 ¼»
µk
10
26.5
30.5547
30.3205
k
4
5
6
7
xk
ª.958115º
«.689261»
«
»
« 1 »
«
»
«¬.943578»¼
ª.957691º
«.688978»
«
»
« 1 »
«
»
«¬.943755»¼
ª.957637 º
«.688942 »
«
»
« 1 »
«
»
«¬.943778 »¼
ª.957630 º
«.688938»
«
»
« 1 »
«
»
«¬ .943781»¼
Ax k
ª 29.0110º
« 20.8710»
«
»
«30.2927 »
«
»
¬« 28.5889¼»
ª 29.0060º
« 20.8675»
«
»
«30.2892 »
«
»
¬« 28.5863¼»
ª 29.0054º
« 20.8671»
«
»
«30.2887 »
«
»
¬« 28.5859¼»
ª 29.0053º
« 20.8670»
«
»
«30.2887 »
«
»
¬« 28.5859¼»
µk
30.2927
30.2892
30.2887
30.2887
Thus an estimate for the eigenvalue to four decimal places is 30.2887. The actual eigenvalue is
30.2886853 to seven decimal places. An estimate for the corresponding eigenvector is
ª.957630º
«.688938»
«
».
« 1 »
«
»
«¬.943781»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
5.8
• Solutions
347
(b) The data in the table below was calculated using Mathematica, which carried more digits than
shown here.
k
0
1
2
3
4
xk
ª1 º
«0»
« »
«0»
« »
«¬ 0 »¼
ª −.609756º
«
»
1
«
»
« −.243902»
«
»
«¬ .146341 »¼
ª −.604007 º
«
»
1
«
»
« −.251051»
«
»
«¬ .148899 »¼
ª −.603973º
«
»
1
«
»
« −.251134»
«
»
«¬ .148953 »¼
ª −.603972º
«
»
1
«
»
« −.251135»
«
»
«¬ .148953 »¼
yk
ª 25º
« −41»
«
»
« 10 »
«
»
¬« −6 ¼»
ª −59.5610º
« 98.6098»
«
»
« −24.7561»
«
»
¬« 14.6829¼»
ª −59.5041º
« 98.5211»
«
»
« −24.7420»
«
»
¬« 14.6750¼»
ª −59.5044º
« 98.5217 »
«
»
« −24.7423»
«
»
¬« 14.6751¼»
ª −59.5044º
« 98.5217 »
«
»
« −24.7423»
«
»
¬« 14.6751¼»
µk
−41
98.6098
98.5211
98.5217
98.5217
νk
−.0243902
.0101410
.0101501
.0101500
.0101500
Thus an estimate for the eigenvalue to five decimal places is .01015. The actual eigenvalue is
.01015005 to eight decimal places. An estimate for the corresponding eigenvector is
ª −.603972 º
«
»
1
«
».
« −.251135»
«
»
¬« .148953 ¼»
ª 1 2
« 2 12
20. [M] A = «
« −2
3
«
5
¬« 4
3
13
0
7
2º
ª1 º
»
«0 »
11»
, x0 = « » .
«0 »
2»
»
« »
2 ¼»
¬«0 ¼»
(a) The data in the table below was calculated using Mathematica, which carried more digits than
shown here.
k
0
1
2
3
4
xk
ª1 º
«0»
« »
«0»
« »
¬« 0 ¼»
ª .25 º
« .5 »
« »
« −.5»
« »
¬« 1 ¼»
ª .159091º
« 1 »
«
»
«.272727 »
«
»
¬«.181818 ¼»
ª.187023º
« 1 »
«
»
«.170483»
«
»
¬«.442748¼»
ª.184166 º
« 1 »
«
»
«.180439 »
«
»
¬«.402197 ¼»
Ax k
ª 1º
« 2»
« »
« −2 »
« »
«¬ 4 »¼
ª1.75º
« 11»
«
»
« 3»
«
»
«¬ 2 »¼
ª 3.34091º
«17.8636 »
«
»
«3.04545 »
«
»
«¬7.90909 »¼
ª3.58397 º
«19.4606 »
«
»
« 3.51145 »
«
»
«¬7.82697 »¼
ª3.52988 º
«19.1382 »
«
»
«3.43606 »
«
»
«¬ 7.80413»¼
µk
4
11
17.8636
19.4606
19.1382
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
348
CHAPTER 5
• Eigenvalues and Eigenvectors
k
5
6
7
8
9
xk
ª .184441º
« 1 »
«
»
«.179539 »
«
»
¬«.407778¼»
ª.184414 º
« 1 »
«
»
«.179622 »
«
»
¬« .407021¼»
ª.184417 º
« 1 »
«
»
«.179615 »
«
»
¬« .407121¼»
ª.184416 º
« 1 »
«
»
«.179615»
«
»
¬«.407108¼»
ª.184416 º
« 1 »
«
»
«.179615»
«
»
¬«.407110 ¼»
Ax k
ª 3.53861º
«19.1884 »
«
»
«3.44667 »
«
»
«¬7.81010 »¼
ª3.53732 º
«19.1811 »
«
»
« 3.44521»
«
»
«¬7.80905»¼
ª3.53750 º
«19.1822 »
«
»
« 3.44541»
«
»
«¬ 7.80921»¼
ª3.53748 º
«19.1820 »
«
»
«3.44538 »
«
»
«¬7.80919 »¼
ª3.53748 º
«19.1820 »
«
»
«3.44539 »
«
»
¬7.80919 ¼
µk
19.1884
19.1811
19.1822
19.1820
19.1820
Thus an estimate for the eigenvalue to four decimal places is 19.1820. The actual eigenvalue is
19.1820368 to seven decimal places. An estimate for the corresponding eigenvector is
ª.184416º
« 1 »
«
».
«.179615»
«
»
¬«.407110¼»
(b) The data in the table below was calculated using Mathematica, which carried more digits than
shown here.
k
0
1
2
xk
ª1 º
«0»
« »
«0»
« »
¬« 0 ¼»
1
ª
º
« .226087 »
«
»
« −.921739»
«
»
¬« .660870 ¼»
1
ª
º
« .222577 »
«
»
« −.917970»
«
»
¬« .660496 ¼»
yk
ª 115º
« 26 »
«
»
« −106 »
«
»
«¬ 76 »¼
ª 81.7304º
« 18.1913»
«
»
« −75.0261»
«
»
«¬ 53.9826»¼
ª 81.9314 º
« 18.2387 »
«
»
« −75.2125»
«
»
«¬ 54.1143»¼
µk
115
81.7304
81.9314
νk
.00869565
.0122353
.0122053
Thus an estimate for the eigenvalue to four decimal places is .0122. The actual eigenvalue is
.01220556 to eight decimal places. An estimate for the corresponding eigenvector is
1
ª
º
« .222577 »
«
».
« −.917970 »
«
»
¬« .660496 ¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 5
ª.8
21. (a) A = «
¬0
• Supplementary Exercises
349
0º
ª.5º
, x = « » . Here is the sequence Ak x for k = 1, …5 :
»
.2¼
¬.5¼
ª.4º ª.32 º ª.256º ª.2048º ª.16384º
«.1» , «.02 » , «.004» , «.0008» , «.00016»
¬ ¼ ¬ ¼ ¬
¼ ¬
¼ ¬
¼
Notice that A5 x is approximately .8( A4 x).
Conclusion: If the eigenvalues of A are all less than 1 in magnitude, and if x ≠ 0, then Ak x is
approximately an eigenvector for large k.
ª 1 0º
ª.5º
, x = « » . Here is the sequence Ak x for k = 1, …5 :
(b) A = «
»
¬0 .8¼
¬.5¼
.5º
ª.5º ª .5º ª .5º ª .5º ª
«.4» , «.32 » , «.256» , «.2048» , «.16384»
¬ ¼ ¬ ¼ ¬
¼ ¬
¼ ¬
¼
ª.5º
Notice that Ak x seems to be converging to « » .
¬ 0¼
Conclusion: If the strictly dominant eigenvalue of A is 1, and if x has a component in the
direction of the corresponding eigenvector, then { Ak x} will converge to a multiple of that
eigenvector.
ª8
c. A = «
¬0
0º
ª.5º
, x = « » . Here is the sequence Ak x for k = 1,…5 :
»
2¼
¬.5¼
ª 4º ª32º ª 256º ª 2048º ª16384º
,
«1 » , « 2 » , « 4 » , «
8»¼ «¬ 16»¼
¬ ¼ ¬ ¼ ¬
¼ ¬
Notice that the distance of Ak x from either eigenvector of A is increasing rapidly as k increases.
Conclusion: If the eigenvalues of A are all greater than 1 in magnitude, and if x is not an
eigenvector, then the distance from Ak x to the nearest eigenvector will increase as k → ∞.
Chapter 5 SUPPLEMENTARY EXERCISES
1. a. True. If A is invertible and if Ax = 1 ⋅ x for some nonzero x, then left-multiply by A−1 to obtain
x = A−1x, which may be rewritten as A−1x = 1 ⋅ x. Since x is nonzero, this shows 1 is an
eigenvalue of A−1 .
b. False. If A is row equivalent to the identity matrix, then A is invertible. The matrix in Example 4
of Section 5.3 shows that an invertible matrix need not be diagonalizable. Also, see Exercise 31
in Section 5.3.
c. True. If A contains a row or column of zeros, then A is not row equivalent to the identity matrix
and thus is not invertible. By the Invertible Matrix Theorem (as stated in Section 5.2), 0 is an
eigenvalue of A.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
350
CHAPTER 5
• Eigenvalues and Eigenvectors
d. False. Consider a diagonal matrix D whose eigenvalues are 1 and 3, that is, its diagonal entries
are 1 and 3. Then D2 is a diagonal matrix whose eigenvalues (diagonal entries) are 1 and 9. In
general, the eigenvalues of A2 are the squares of the eigenvalues of A.
e. True. Suppose a nonzero vector x satisfies Ax = λ x, then
A2 x = A( Ax) = A(λx) = λ Ax = λ2 x
This shows that x is also an eigenvector for A2
f. True. Suppose a nonzero vector x satisfies Ax = λ x, then left-multiply by A−1 to obtain
x = A−1 (λ x) = λ A−1x. Since A is invertible, the eigenvalue λ is not zero. So λ −1x = A−1x, which
shows that x is also an eigenvector of A−1 .
g. False. Zero is an eigenvalue of each singular square matrix.
h. True. By definition, an eigenvector must be nonzero.
ª2
¬0
i. False. Let A = «
0º
ª1 º
ª0º
then e1 = « » and e 2 = « » are eigenvectors of A for the eigenvalue 2,
»
2¼
¬0¼
¬1 ¼
and they are linearly independent.
j. True. This follows from Theorem 4 in Section 5.2
k. False. Let A be the 3 × 3 matrix in Example 3 of Section 5.3. Then A is similar to a diagonal
matrix D. The eigenvectors of D are the columns of I 3 , but the eigenvectors of A are entirely
different.
ª1 º
ª0º
ª 2 0º
. Then e1 = « » and e2 = « » are eigenvectors of A, but e1 + e 2 is not.
l. False. Let A = «
»
¬0¼
¬1 ¼
¬ 0 3¼
(Actually, it can be shown that if two eigenvectors of A correspond to distinct eigenvalues, then
their sum cannot be an eigenvector.)
m. False. All the diagonal entries of an upper triangular matrix are the eigenvalues of the matrix
(Theorem 1 in Section 5.1). A diagonal entry may be zero.
n. True. Matrices A and AT have the same characteristic polynomial, because
det( AT − λI ) = det( A − λI )T = det( A − λI ), by the determinant transpose property.
o. False. Counterexample: Let A be the 5 × 5 identity matrix.
p. True. For example, let A be the matrix that rotates vectors through π/ 2 radians about the origin.
Then Ax is not a multiple of x when x is nonzero.
q. False. If A is a diagonal matrix with 0 on the diagonal, then the columns of A are not linearly
independent.
r. True. If Ax = λ1x and Ax = λ2 x, then λ1x = λ2 x and (λ1 − λ2 ) x = 0. If x ≠ 0, then λ1 must equal
λ2 .
s. False. Let A be a singular matrix that is diagonalizable. (For instance, let A be a diagonal matrix
with 0 on the diagonal.) Then, by Theorem 8 in Section 5.4, the transformation x 6 Ax is
represented by a diagonal matrix relative to a coordinate system determined by eigenvectors of
A.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 5
• Supplementary Exercises
351
t. True. By definition of matrix multiplication,
A = AI = A[e1
e2
" en ] = [ Ae1
Ae2
"
Aen ]
If Ae j = d j e j for j = 1, …, n, then A is a diagonal matrix with diagonal entries d1 , …, d n .
u. True. If B = PDP −1 , where D is a diagonal matrix, and if A = QBQ −1 , then
A = Q( PDP −1 )Q −1 = (QP) D(QP) −1 , which shows that A is diagonalizable.
v. True. Since B is invertible, AB is similar to B ( AB ) B −1 , which equals BA.
w. False. Having n linearly independent eigenvectors makes an n × n matrix diagonalizable (by the
Diagonalization Theorem 5 in Section 5.3), but not necessarily invertible. One of the eigenvalues
of the matrix could be zero.
x. True. If A is diagonalizable, then by the Diagonalization Theorem, A has n linearly independent
eigenvectors v1 , …, v n in R n . By the Basis Theorem, {v1 , …, v n } spans R n . This means that
each vector in R n can be written as a linear combination of v1 , …, v n .
2. Suppose Bx ≠ 0 and ABx = λx for some λ . Then A( Bx) = λx. Left-multiply each side by B and
obtain BA( Bx) = B(λx) = λ( Bx). This equation says that Bx is an eigenvector of BA, because
B x ≠ 0.
3. a. Suppose Ax = λx, with x ≠ 0. Then (5 I − A)x = 5x − Ax = 5x − λx = (5 − λ)x. The eigenvalue
is 5 − λ .
b. (5 I − 3 A + A2 )x = 5x − 3 Ax + A( Ax) = 5x − 3(λx) + λ 2 x = (5 − 3λ + λ 2 ) x. The eigenvalue is
5 − 3λ + λ 2.
4. Assume that Ax = λx for some nonzero vector x. The desired statement is true for m = 1, by the
assumption about λ . Suppose that for some k ≥ 1, the statement holds when m = k . That is, suppose
that A k x = λ k x. Then Ak +1x = A( Ak x) = A(λk x) by the induction hypothesis. Continuing,
Ak +1x = λ k Ax = λ k +1x, because x is an eigenvector of A corresponding to A. Since x is nonzero, this
equation shows that λk +1 is an eigenvalue of Ak +1 , with corresponding eigenvector x. Thus the
desired statement is true when m = k + 1. By the principle of induction, the statement is true for each
positive integer m.
5. Suppose Ax = λx, with x ≠ 0. Then
p( A)x = (c0 I + c1 A + c2 A2 + …+ cn An )x
= c0 x + c1 Ax + c2 A2 x + …+ cn An x
= c0 x + c1λx + c2 λ 2 x + …+ cn λ n x = p (λ )x
So p(λ) is an eigenvalue of p( A).
6. a. If A = PDP −1 , then Ak = PDk P −1 , and
B = 5I − 3 A + A2 = 5PIP −1 − 3PDP −1 + PD 2 P −1
= P (5I −3D + D2 ) P−1
Since D is diagonal, so is 5 I − 3 D + D 2 . Thus B is similar to a diagonal matrix.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
352
CHAPTER 5
• Eigenvalues and Eigenvectors
−1
2 −1
n −1
b. p( A) = c0 I + c1PDP + c2 PD P + " + cn PD P
= P (c0 I + c1D + c2 D2 + " + cn Dn ) P−1
= Pp ( D) P−1
This shows that p( A) is diagonalizable, because p( D) is a linear combination of diagonal
matrices and hence is diagonal. In fact, because D is diagonal, it is easy to see that
ª p(2)
p ( D) = «
¬ 0
0 º
p(7) »¼
7. If A = PDP −1 , then p( A) = Pp( D) P −1 , as shown in Exercise 6. If the ( j, j ) entry in D is λ , then the
( j, j ) entry in D k is λ k , and so the ( j, j ) entry in p( D) is p(λ). If p is the characteristic
polynomial of A, then p(λ) = 0 for each diagonal entry of D, because these entries in D are the
eigenvalues of A. Thus p( D) is the zero matrix. Thus p ( A) = P ⋅ 0 ⋅ P −1 = 0.
8. a. If λ is an eigenvalue of an n × n diagonalizable matrix A, then A = PDP −1 for an invertible
matrix P and an n × n diagonal matrix D whose diagonal entries are the eigenvalues of A. If the
multiplicity of λ is n, then λ must appear in every diagonal entry of D. That is, D = λ I . In this
case, A = P (λ I ) P −1 = λ PIP −1 = λ PP −1 = λ I .
ª 3 1º
b. Since the matrix A = «
» is triangular, its eigenvalues are on the diagonal. Thus 3 is an
¬0 3¼
eigenvalue with multiplicity 2. If the 2 × 2 matrix A were diagonalizable, then A would be 3I, by
part (a). This is not the case, so A is not diagonalizable.
9. If I − A were not invertible, then the equation ( I − A)x = 0. would have a nontrivial solution x. Then
x − Ax = 0 and Ax = 1 ⋅ x, which shows that A would have 1 as an eigenvalue. This cannot happen if
all the eigenvalues are less than 1 in magnitude. So I − A must be invertible.
10. To show that Ak tends to the zero matrix, it suffices to show that each column of Ak can be made as
close to the zero vector as desired by taking k sufficiently large. The jth column of A is Ae j , where
e j is the jth column of the identity matrix. Since A is diagonalizable, there is a basis for
n
consisting of eigenvectors v1 ,…, v n , corresponding to eigenvalues λ1 ,…,λ n . So there exist scalars
c1 , …, cn , such that
e j = c1 v1 + …+ cn v n
(an eigenvector decomposition of e j )
Then, for k = 1, 2,…,
Ak e j = c1 (λ1 ) k v1 + " + cn (λ n ) k v n
(∗)
If the eigenvalues are all less than 1 in absolute value, then their kth powers all tend to zero. So (∗)
shows that Ak e j tends to the zero vector, as desired.
11. a. Take x in H. Then x = cu for some scalar c. So Ax = A(cu) = c( Au) = c(λu) = (cλ)u, which
shows that Ax is in H.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 5
• Supplementary Exercises
353
b. Let x be a nonzero vector in K. Since K is one-dimensional, K must be the set of all scalar
multiples of x. If K is invariant under A, then Ax is in K and hence Ax is a multiple of x. Thus x
is an eigenvector of A.
12. Let U and V be echelon forms of A and B, obtained with r and s row interchanges, respectively, and
no scaling. Then det A = (−1) r det U and det B = (−1) s det V
Using first the row operations that reduce A to U, we can reduce G to a matrix of the form
ªU Y º
G′ = «
» . Then, using the row operations that reduce B to V, we can further reduce G′ to
¬ 0 B¼
ªU
G′′ = «
¬0
Yº
. There will be r + s row interchanges, and so
V »¼
ªA X º
ªU Y º
ªU Y º
det G = det «
= (−1)r + s det «
Since «
»
»
» is upper triangular, its determinant
¬0 B ¼
¬0 V¼
¬0 V¼
equals the product of the diagonal entries,
and since U and V are upper triangular, this product also equals (det U ) (det V ). Thus
det G = (−1)r + s (det U )(det V ) = (det A)(det B )
For any scalar λ , the matrix G − λ I has the same partitioned form as G, with A − λ I and B − λ I as
its diagonal blocks. (Here I represents various identity matrices of appropriate sizes.) Hence the
result about det G shows that det(G − λI ) = det( A − λI ) ⋅ det( B − λI )
13. By Exercise 12, the eigenvalues of A are the eigenvalues of the matrix [3] together with the
ª 5 −2 º
. The only eigenvalue of [3] is 3, while the eigenvalues of
eigenvalues of «
3»¼
¬ −4
1 and 7. Thus the eigenvalues of A are 1, 3, and 7.
ª 5
« −4
¬
−2º
are
3»¼
ª1 5 º
14. By Exercise 12, the eigenvalues of A are the eigenvalues of the matrix «
» together with the
¬ 2 4¼
ª1 5 º
ª −7 −4º
. The eigenvalues of «
eigenvalues of «
» are −1 and 6, while the eigenvalues of
»
1¼
¬ 2 4¼
¬ 3
ª −7 −4º
are −5 and −1. Thus the eigenvalues of A are −1, − 5, and 6, and the eigenvalue −1 has
« 3
1»¼
¬
multiplicity 2.
15. Replace a by a − λ in the determinant formula from Exercise 16 in Chapter 3 Supplementary
Exercises.
det( A − λI ) = (a − b − λ) n −1[a − λ + (n − 1)b]
This determinant is zero only if a − b − λ = 0 or a − λ + (n − 1)b = 0. Thus λ is an eigenvalue of A if
and only if λ = a − b or λ = a + (n − 1). From the formula for det( A − λI ) above, the algebraic
multiplicity is n − 1 for a − b and 1 for a + (n − 1)b.
16. The 3 × 3 matrix has eigenvalues 1 − 2 and 1 + (2)(2), that is, −1 and 5. The eigenvalues of the 5 × 5
matrix are 7 − 3 and 7 + (4)(3), that is 4 and 19.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
354
CHAPTER 5
• Eigenvalues and Eigenvectors
17. Note that det( A − λI ) = ( a11 − λ)( a22 − λ) − a12 a21 = λ 2 − (a11 + a22 )λ + (a11a22 − a12 a21 )
= λ 2 − (tr A)λ + det A, and use the quadratic formula to solve the characteristic equation:
tr A ± (tr A)2 − 4det A
2
The eigenvalues are both real if and only if the discriminant is nonnegative, that is,
λ=
2
§ trA ·
(tr A) 2 − 4 det A ≥ 0. This inequality simplifies to (tr A) 2 ≥ 4 det A and ¨
¸ ≥ det A.
© 2 ¹
18. The eigenvalues of A are 1 and .6. Use this to factor A and Ak .
ª −1
A= «
¬ 2
−3º ª1
2 »¼ «¬0
ª −1
A =«
¬ 2
−3º ª«1k
«
2 »¼ «¬« 0
k
=
1 ª −1
4 «¬ 2
=
1 « −2 + 6(.6)
«
4 ««¬ 4−4(.6)k
º
»
»⋅
k»
.6 ¼»
0
k
3º
−1»¼
1ª 2
4 «¬ −2
2
−3º ª«
«
»
2 ¼ «¬ −2 ⋅ (.6) k
ª
1 ª −2
→ «
4¬ 4
0º 1 ª 2
⋅
.6 »¼ 4 «¬ −2
3º
−1»¼
º
»
»
−(.6)k »¼
3
−3+3(.6)k º»
»
6−2(.6)k »»¼
−3º
as k → ∞
6 »¼
ª 0
19. C p = «
¬ −6
1º
; det(C p − λI ) = 6 − 5λ + λ 2 = p(λ)
5»¼
ª 0
«
20. C p = « 0
¬« 24
1
0
−26
0º
1»» ;
9»¼
det(C p − λ I ) = 24 − 26λ + 9λ 2 − λ 3 = p (λ )
21. If p is a polynomial of order 2, then a calculation such as in Exercise 19 shows that the characteristic
polynomial of C p is p (λ ) = (−1) 2 p(λ ), so the result is true for n = 2. Suppose the result is true for
n = k for some k ≥ 2, and consider a polynomial p of degree k + 1. Then expanding det(C p − λI )
by cofactors down the first column, the determinant of C p − λI equals
ª −λ
« #
(−λ ) det «
« 0
«
«¬ − a1
1
"
− a2
"
º
»
» + ( −1) k +1 a
0
»
»
− ak − λ »¼
0
#
1
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 5
• Supplementary Exercises
355
The k × k matrix shown is Cq − λI , where q (t ) = a1 + a2t + " + ak t k −1 + t k . By the induction
assumption, the determinant of Cq − λI is (−1)k q(λ ). Thus
det(C p − λI ) = (−1) k +1 a0 + (−λ )(−1)k q (λ )
= (−1) k +1[ a0 + λ (a1 + " + ak λ k −1 + λ k )]
= (−1) k +1 p (λ )
So the formula holds for n = k + 1 when it holds for n = k . By the principle of induction, the formula
for det(C p − λI ) is true for all n ≥ 2.
22. a. C p =
ª 0
«
«
« 0
«
« −a
0
¬«
1
0
º
»
»
»
»
−a2 »¼»
0
1
−a1
b. Since λ is a zero of p, a0 + a1λ + a2 λ 2 + λ 3 = 0 and −a0 − a1λ − a2 λ 2 = λ3 . Thus
ª
º
« 1 »
«
»
Cp « λ »
«
»
« 2»
«¬ λ »¼
=
ª
«
«
«
«
«
«¬ − a0
λ
λ2
− a1λ −
º
»
»
»
»
2
a2λ »»¼
=
ª λ º
«
»
« 2»
«λ »
«
»
« 3»
«¬ λ »¼
That is, C p (1,λ,λ 2 ) = λ (1,λ,λ 2 ), which shows that (1,λ,λ 2 ) is an eigenvector of C p
corresponding to the eigenvalue λ .
23. From Exercise 22, the columns of the Vandermonde matrix V are eigenvectors of C p , corresponding
to the eigenvalues λ1 ,λ 2 ,λ 3 (the roots of the polynomial p). Since these eigenvalues are distinct, the
eigenvectors from a linearly independent set, by Theorem 2 in Section 5.1. Thus V has linearly
independent columns and hence is invertible, by the Invertible Matrix Theorem. Finally, since the
columns of V are eigenvectors of C p , the Diagonalization Theorem (Theorem 5 in Section 5.3)
shows that V −1C pV is diagonal.
24. [M] The MATLAB command roots(p) requires as input a row vector p whose entries are the
coefficients of a polynomial, with the highest order coefficient listed first. MATLAB constructs a
companion matrix C p whose characteristic polynomial is p, so the roots of p are the eigenvalues of
C p . The numerical values of the eigenvalues (roots) are found by the same QR algorithm used by
the command eig(A).
25. [M] The MATLAB command [P
D]= eig(A) produces a matrix P, whose condition number is
1.6 ×108 , and a diagonal matrix D, whose entries are almost 2, 2, 1. However, the exact eigenvalues
of A are 2, 2, 1, and A is not diagonalizable.
26. [M] This matrix may cause the same sort of trouble as the matrix in Exercise 25. A matrix program
that computes eigenvalues by an interative process may indicate that A has four distinct eigenvalues,
all close to zero. However, the only eigenvalue is 0, with multiplicity 4, because A4 = 0.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.1
SOLUTIONS
Notes: The first half of this section is ccomputational and is easily learned. The second half concerns
c
the
concepts of orthogonality and orthogonaal complements, which are essential for later work. Theorem
T
3 is
an important general fact, but is needed only for Supplementary Exercise 13 at the end of thee chapter and
in Section 7.4. The optional material on angles is not used later. Exercises 27–31 concern factss used later.
ª −1º
v ⋅u 8
ª4º
= .
1. Since u = « » and v = « » , u ⋅ u = (−1)2 + 22 = 5 , v ⋅ u = 4(–1) + 6(2) = 8, and
u⋅u 5
¬ 2¼
¬6¼
ª 3º
ª 6º
«
»
2. Since w = « −1» and x = «« −2 »» , w ⋅ w = 32 + (−1)2 + (−5)2 = 35 , x ⋅ w = 6(3) + (–2)(–1) + 3(–5) = 5,
«¬ −5»¼
«¬ 3»¼
x⋅w 5 1
= = .
and
w ⋅ w 35 7
ª 3º
ª 3/ 35º
1
«
»
2
2
2
w = «« −1/ 35»» .
3. Since w = « −1» , w ⋅ w = 3 + (−1) + (−5) = 35 , and
w⋅w
«¬ −1/ 7 »¼
«¬ −5»¼
ª −1º
ª −1/ 5º
1
u=«
4. Since u = « » , u ⋅ u = (−1)2 + 22 = 5 and
».
u⋅u
¬ 2¼
¬ 2 / 5¼
ª −1º
ª4º
5. Since u = « » and v = « » , u ⋅ v = (–1)(4) + 2(6) = 8, v ⋅ v = 42 + 62 = 52, and
¬ 2¼
¬6¼
2 ª 4º ª 8 /13º
§ u⋅v ·
¨
¸ v = «6 » = «12 /13» .
13 ¬ ¼ ¬
© v⋅v¹
¼
Copyright © 2012 Pearrson Education, Inc. Publishing as Addison-Wesley.
357
358
CHAPTER 6
• Orthogonality and Least Squares
ª 6º
ª 3º
«
»
6. Since x = « −2» and w = «« −1»» , x ⋅ w = 6(3) + (–2)(–1) + 3(–5) = 5, x ⋅ x = 62 + (−2)2 + 32 = 49, and
«¬ 3»¼
«¬ −5»¼
ª 6 º ª 30 / 49º
5 « » «
§ x⋅w ·
»
¨
¸ x = « −2 » = « −10 / 49» .
49
© x⋅x ¹
¬« 3¼» ¬« 15 / 49¼»
ª 3º
7. Since w = «« −1»» , || w ||= w ⋅ w = 32 + ( −1) 2 + ( −5) 2 = 35.
«¬ −5»¼
ª 6º
8. Since x = «« −2 »» , || x ||= x ⋅ x = 6 2 + ( −2) 2 + 32 = 49 = 7.
«¬ 3»¼
9. A unit vector in the direction of the given vector is
ª −30º 1 ª −30 º ª −3/ 5º
«
»= «
»=«
»
(−30) + 40 ¬ 40¼ 50 ¬ 40 ¼ ¬ 4 / 5¼
1
2
2
10. A unit vector in the direction of the given vector is
ª −6 º
ª −6 º ª −6 / 61 º
1
« 4» =
« 4 » = « 4 / 61 »
«
»
« »
61 « » «
»
( −6) 2 + 4 2 + ( −3) 2 « −3»
«¬ −3»¼
¬ ¼
«¬ −3 61 »¼
1
11. A unit vector in the direction of the given vector is
ª7 / 4 º
ª 7 / 4 º ª 7 / 69 º
1
1
« 1/ 2 » =
« 1/ 2 » = « 2 / 69 »
»
«
»
» «
69 /16 «
«
(7 / 4) 2 + (1/ 2) 2 + 12 « 1»
«¬ 1»¼ 4 / 69 »
¬
¼
¬«
¼»
12. A unit vector in the direction of the given vector is
ª8 / 3º
ª8 / 3º ª 4 / 5º
1
« 2» =
« »=«
»
100 / 9 ¬ 2¼ ¬ 3/ 5 ¼
(8 / 3) + 2 ¬ ¼
1
2
2
ª −1º
ª 10 º
13. Since x = « » and y = « » , || x − y ||2 = [10 − (−1)]2 + [−3 − (−5)]2 = 125 and
¬ −5¼
¬ −3¼
dist (x, y) = 125 = 5 5.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.1
• Solutions
359
ª 0º
ª −4 º
«
»
14. Since u = « −5» and z = «« −1»» , || u − z ||2 = [0 − (−4)]2 + [−5 − (−1)]2 + [2 − 8]2 = 68 and
«¬ 2»¼
«¬ 8»¼
dist (u, z) = 68 = 2 17.
15. Since a ⋅ b = 8(–2) + (–5)( –3) = –1 ≠ 0, a and b are not orthogonal.
16. Since u ⋅ v = 12(2) + (3)( –3) + (–5)(3) = 0, u and v are orthogonal.
17. Since u ⋅ v = 3(–4) + 2(1) + (–5)( –2) + 0(6) = 0, u and v are orthogonal.
18. Since y ⋅ z = (–3)(1) + 7(–8) + 4(15) + 0(–7) = 1 ≠ 0, y and z are not orthogonal.
19. a. True. See the definition of || v ||.
b. True. See Theorem 1(c).
c. True. See the discussion of Figure 5.
ª 1 1º
d. False. Counterexample: «
».
¬0 0 ¼
e. True. See the box following Example 6.
20. a.
b.
c.
d.
e.
True. See Example 1 and Theorem 1(a).
False. The absolute value sign is missing. See the box before Example 2.
True. See the defintion of orthogonal complement.
True. See the Pythagorean Theorem.
True. See Theorem 3.
21. Theorem 1(b):
(u + v) ⋅ w = (u + v)T w = (uT + vT )w = uT w + vT w = u ⋅ w + v ⋅ w
The second and third equalities used Theorems 3(b) and 2(c), respectively, from Section 2.1.
Theorem 1(c):
(cu) ⋅ v = (cu)T v = c(uT v) = c(u ⋅ v)
The second equality used Theorems 3(c) and 2(d), respectively, from Section 2.1.
22. Since u ⋅ u is the sum of the squares of the entries in u, u ⋅ u ≥ 0. The sum of squares of numbers is
zero if and only if all the numbers are themselves zero.
23. One computes that u ⋅ v = 2(–7) + (–5)( –4) + (–1)6 = 0, || u ||2 = u ⋅ u = 22 + (−5)2 + (−1)2 = 30,
|| v ||2 = v ⋅ v = (−7)2 + (−4)2 + 62 = 101, and || u + v ||2 = (u + v) ⋅ (u + v) =
(2 + (−7))2 + (−5 + (−4))2 + (−1 + 6)2 = 131.
24. One computes that
|| u + v ||2 = (u + v) ⋅ (u + v) = u ⋅ u + 2u ⋅ v + v ⋅ v =|| u ||2 +2u ⋅ v+ || v ||2
and
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
360
CHAPTER 6
• Orthogonality and Least Squares
|| u − v ||2 = (u − v) ⋅ (u − v) = u ⋅ u − 2u ⋅ v + v ⋅ v =|| u ||2 −2u ⋅ v + || v ||2
so
|| u + v ||2 + || u − v ||2 =|| u ||2 +2u ⋅ v + || v ||2 + || u ||2 −2u ⋅ v+ || v ||2 = 2 || u ||2 +2 || v ||2
ªaº
ª xº
25. When v = « » , the set H of all vectors « » that are orthogonal to v is the subspace of vectors whose
¬ y¼
¬b ¼
entries satisfy ax + by = 0. If a ≠ 0, then x = – (b/a)y with y a free variable, and H is a line through
° ª −b º ½°
the origin. A natural choice for a basis for H in this case is ® « » ¾ . If a = 0 and b ≠ 0, then by = 0.
¯° ¬ a ¼ ¿°
Since b ≠ 0, y = 0 and x is a free variable. The subspace H is again a line through the origin. A
° ª 1 º °½
° ª −b º °½
natural choice for a basis for H in this case is ® « » ¾ , but ® « » ¾ is still a basis for H since a = 0
°¯ ¬ a ¼ ¿°
¯° ¬ 0 ¼ ¿°
and b ≠ 0. If a = 0 and b = 0, then H =
y.
2
since the equation 0x + 0y = 0 places no restrictions on x or
26. Theorem 2 in Chapter 4 may be used to show that W is a subspace of 3, because W is the null space
of the 1 × 3 matrix u T . Geometrically, W is a plane through the origin.
27. If y is orthogonal to u and v, then y ⋅ u = y ⋅ v = 0, and hence by a property of the inner product,
y ⋅ (u + v) = y ⋅ u + y ⋅ v = 0 + 0 = 0. Thus y is orthogonal to u + v.
28. An arbitrary w in Span{u, v} has the form w = c1u + c2 v . If y is orthogonal to u and v, then
u ⋅ y = v ⋅ y = 0. By Theorem 1(b) and 1(c),
w ⋅ y = (c1u + c2 v ) ⋅ y = c1 (u ⋅ y ) + c2 ( v ⋅ y ) = 0 + 0 = 0
29. A typical vector in W has the form w = c1v1 +…+ c p v p . If x is orthogonal to each vj , then by
Theorems 1(b) and 1(c),
w ⋅ x = (c1v1 +…+ c p v p ) ⋅ x = c1 ( v1 ⋅ x) +…+ c p ( v p ⋅ x) = 0
So x is orthogonal to each w in W.
30. a. If z is in W ⊥ , u is in W, and c is any scalar, then (cz) ⋅ u = c(z ⋅ u) = c 0 = 0. Since u is any
element of W, c z is in W ⊥ .
b. Let z 1 and z 2 be in W ⊥ . Then for any u in W, ( z1 + z 2 ) ⋅ u = z1 ⋅ u + z 2 ⋅ u = 0 + 0 = 0. Thus
z1 + z 2 is in W ⊥ .
c. Since 0 is orthogonal to every vector, 0 is in W ⊥ . Thus W ⊥ is a subspace.
31. Suppose that x is in W and W ⊥ . Since x is in W ⊥ , x is orthogonal to every vector in W, including x
itself. So x ⋅ x = 0, which happens only when x = 0.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.2
• Solutions
361
32. [M]
a. One computes that || a1 || = || a 2 || = || a3 || = || a 4 || = 1 and that ai ⋅ a j = 0 for i ≠ j.
b. Answers will vary, but it should be that || Au || = || u || and || Av || = || v ||.
c. Answers will again vary, but the cosines should be equal.
d. A conjecture is that multiplying by A does not change the lengths of vectors or the angles
between vectors.
33. [M] Answers to the calculations will vary, but will demonstrate that the mapping
§ x⋅v ·
x 6 T (x) = ¨
¸ v (for v ≠ 0) is a linear transformation. To confirm this, let x and y be in
© v⋅v¹
let c be any scalar. Then
n
, and
§ (x + y ) ⋅ v ·
§ (x ⋅ v ) + (y ⋅ v) ·
§ x⋅v ·
§ y⋅v·
T (x + y ) = ¨
¸v = ¨
¸v =¨
¸v +¨
¸ v = T ( x) + T (y )
v⋅v
© v⋅v ¹
©
¹
© v⋅v¹
© v⋅v¹
and
§ ( cx ) ⋅ v ·
§ c(x ⋅ v ) ·
§ x⋅v ·
T ( cx ) = ¨
¸v = ¨
¸ v = c¨
¸ v = cT ( x )
v
⋅
v
v
⋅
v
©
¹
©
¹
© v⋅v¹
34. [M] One finds that
ª −5
«
« −1
N =« 1
«
« 0
«¬ 0
1º
4»»
ª1
0» , R = ««0
»
«¬0
−1»
3»¼
0
1
5
1
0
0
0
0
1
−1/ 3º
−4 / 3»»
1/ 3»¼
The row-column rule for computing RN produces the 3 × 2 zero matrix, which shows that the rows of
R are orthogonal to the columns of N. This is expected by Theorem 3 since each row of R is in Row
A and each column of N is in Nul A.
6.2
SOLUTIONS
Notes: The nonsquare matrices in Theorems 6 and 7 are needed for the QR factorization in Section 6.4. It
is important to emphasize that the term orthogonal matrix applies only to certain square matrices. The
subsection on orthogonal projections not only sets the stage for the general case in Section 6.3, it also
provides what is needed for the orthogonal diagonalization exercises in Section 7.1, because none of the
eigenspaces there have dimension greater than 2. For this reason, the Gram-Schmidt process (Section 6.4)
is not really needed in Chapter 7. Exercises 13 and 14 are good preparation for Section 6.3.
ª −1º ª 3º
« » « »
1. Since « 4» ⋅ « −4» = 2 ≠ 0, the set is not orthogonal.
«¬−3»¼ «¬ −7 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
362
CHAPTER 6
• Orthogonality and Least Squares
ª 1º ª 0 º ª 1º ª −5º ª 0º ª −5º
« » « » « » « » « » « »
2. Since « −2 » ⋅ «1 » = « −2» ⋅ « −2 » = « 1» ⋅ « −2» = 0, the set is orthogonal.
«¬ 1»¼ «¬ 2 »¼ «¬ 1»¼ «¬ 1»¼ «¬ 2»¼ «¬ 1»¼
ª −6º ª 3º
« » « »
3. Since « −3» ⋅ « 1» = −30 ≠ 0, the set is not orthogonal.
«¬ 9»¼ «¬ −1»¼
ª 2º ª0º ª 2º ª 4º ª 0º ª 4 º
4. Since «« −5»» ⋅ ««0»» = «« −5»» ⋅ «« −2»» = «« 0»» ⋅ «« −2 »» = 0, the set is orthogonal.
«¬ −3»¼ «¬0»¼ «¬ −3»¼ «¬ 6»¼ «¬ 0»¼ «¬ 6 »¼
ª 3º ª −1º ª 3º ª 3º ª −1º ª 3º
« −2 » « 3» « −2 » « 8» « 3» « 8»
5. Since « » ⋅ « » = « » ⋅ « » = « » ⋅ « » = 0, the set is orthogonal.
« 1» « −3» « 1» «7 » « −3» « 7 »
« » « » « » « » « » « »
«¬ 3»¼ «¬ 4»¼ «¬ 3»¼ «¬ 0»¼ «¬ 4»¼ «¬ 0»¼
ª −4º ª 3º
« 1» « 3»
6. Since « » ⋅ « » = −32 ≠ 0, the set is not orthogonal.
« −3» « 5»
« » « »
¬« 8¼» ¬« −1¼»
7. Since u1 ⋅ u 2 = 12 − 12 = 0, {u1 , u 2 } is an orthogonal set. Since the vectors are non-zero, u1 and u 2
are linearly independent by Theorem 4. Two such vectors in 2 automatically form a basis for 2. So
{u1 , u 2 } is an orthogonal basis for 2. By Theorem 5,
x=
x ⋅ u1
x ⋅ u2
1
u1 +
u 2 = 3u1 + u 2
u1 ⋅ u1
u2 ⋅ u2
2
8. Since u1 ⋅ u 2 = −6 + 6 = 0, {u1 , u 2 } is an orthogonal set. Since the vectors are non-zero, u1 and u 2
are linearly independent by Theorem 4. Two such vectors in 2 automatically form a basis for 2. So
{u1 , u 2 } is an orthogonal basis for 2. By Theorem 5,
x=
x ⋅ u1
x ⋅ u2
3
3
u1 +
u 2 = − u1 + u 2
u1 ⋅ u1
u2 ⋅ u2
2
4
9. Since u1 ⋅ u 2 = u1 ⋅ u 3 = u 2 ⋅ u 3 = 0, {u1 , u 2 , u 3 } is an orthogonal set. Since the vectors are non-zero,
u1 , u 2 , and u 3 are linearly independent by Theorem 4. Three such vectors in 3 automatically form
a basis for 3. So {u1 , u 2 , u 3 } is an orthogonal basis for 3. By Theorem 5,
x ⋅ u1
x ⋅ u2
x ⋅ u3
5
3
x=
u1 +
u2 +
u 3 = u1 − u 2 + 2 u 3
u1 ⋅ u1
u2 ⋅ u2
u3 ⋅ u3
2
2
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.2
• Solutions
363
10. Since u1 ⋅ u 2 = u1 ⋅ u 3 = u 2 ⋅ u 3 = 0, {u1 , u 2 , u 3 } is an orthogonal set. Since the vectors are non-zero,
u1 , u 2 , and u 3 are linearly independent by Theorem 4. Three such vectors in 3 automatically form
a basis for 3. So {u1 , u 2 , u 3 } is an orthogonal basis for 3. By Theorem 5,
x=
x ⋅ u1
x ⋅ u2
x ⋅ u3
4
1
1
u1 +
u2 +
u 3 = u1 + u 2 + u 3
u1 ⋅ u1
u2 ⋅ u2
u3 ⋅ u3
3
3
3
ª −4 º
ª 1º
11. Let y = « » and u = « » . The orthogonal projection of y onto the line through u and the origin is
¬ 2¼
¬7 ¼
the orthogonal projection of y onto u, and this vector is
yˆ =
ª −2º
y ⋅u
1
u= u=« »
u⋅u
2
¬ 1¼
ª −1º
ª 1º
12. Let y = « » and u = « » . The orthogonal projection of y onto the line through u and the origin is
¬ 3¼
¬ −1¼
the orthogonal projection of y onto u, and this vector is
yˆ =
ª 2 / 5º
y ⋅u
2
u=− u=«
»
u⋅u
5
¬ −6 / 5¼
13. The orthogonal projection of y onto u is
yˆ =
ª −4 / 5º
y ⋅u
13
u=− u=«
»
u ⋅u
65
¬ 7 / 5¼
The component of y orthogonal to u is
ª14/5º
y − yˆ = «
»
¬ 8/5¼
ª −4/5º ª14/5º
Thus y = yˆ + ( y − yˆ ) = «
»+«
».
¬ 7/5¼ ¬ 8/5¼
14. The orthogonal projection of y onto u is
yˆ =
ª14 / 5º
y ⋅u
2
u= u=«
»
u ⋅u
5
¬ 2 / 5¼
The component of y orthogonal to u is
ª −4/5º
y − yˆ = «
»
¬ 28/5¼
ª14/5º ª −4/5º
Thus y = yˆ + ( y − yˆ ) = «
»+«
».
¬ 2/5¼ ¬ 28/5¼
15. The distance from y to the line through u and the origin is ||y – ŷ ||. One computes that
y − yˆ = y −
ª3º 3 ª 8º ª 3/ 5º
y ⋅u
u=« »− « »=«
»
u ⋅u
¬ 1¼ 10 ¬6¼ ¬ −4 / 5¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
364
CHAPTER 6
• Orthogonality and Least Squares
so || y − yˆ ||= 9/25 +16/25 =1 is the desired distance.
16. The distance from y to the line through u and the origin is ||y – ŷ ||. One computes that
y − yˆ = y −
ª −3º ª 1º ª −6º
y ⋅u
u = « » − 3« » = « »
u ⋅u
¬ 9 ¼ ¬ 2 ¼ ¬ 3¼
so || y − yˆ ||= 36 + 9 = 3 5 is the desired distance.
ª1/ 3º
ª −1/ 2 º
«
»
«
0 »» . Since u ⋅ v = 0, {u, v} is an orthogonal set. However, || u ||2 = u ⋅ u = 1/ 3
17. Let u = «1/ 3» , v = «
¬«1/ 3¼»
¬« 1/ 2 ¼»
and || v ||2 = v ⋅ v = 1/ 2, so {u, v} is not an orthonormal set. The vectors u and v may be normalized to
form the orthonormal set
ª 3 / 3º ª − 2 / 2 º ½
» «
» °°
u
v ½ °° «
,
0» ¾
®
¾ = ® « 3 / 3» , «
¯ || u || || v || ¿ ° «
» «
»°
°¯ ¬« 3 / 3¼» ¬ 2 / 2¼ °¿
ª0º
ª 0º
«
»
18. Let u = « 1» , v = ««−1»» . Since u ⋅ v = –1 ≠ 0, {u, v} is not an orthogonal set.
«¬0 »¼
«¬ 0 »¼
ª −.6 º
ª.8º
19. Let u = « » , v = « » . Since u ⋅ v = 0, {u, v} is an orthogonal set. Also, || u ||2 = u ⋅ u = 1 and
¬.6 ¼
¬ .8¼
|| v ||2 = v ⋅ v = 1, so {u, v} is an orthonormal set.
ª −2 / 3º
ª 1/ 3º
«
»
«
»
20. Let u = « 1/ 3» , v = « 2 / 3» . Since u ⋅ v = 0, {u, v} is an orthogonal set. However, || u ||2 = u ⋅ u = 1
«¬ 0 »¼
«¬ 2 / 3»¼
and || v ||2 = v ⋅ v = 5/ 9, so {u, v} is not an orthonormal set. The vectors u and v may be normalized
to form the orthonormal set
−2 / 3 ª 1/ 5 º ½
ª
º
» °°
u
v ½ °° «
» «
,
®
¾ = ® « 1/ 3» , « 2 / 5 » ¾
¯ || u || || v || ¿ ° «
«
0»» °
¬ 2 / 3»¼ ¬«
¼ ¿°
¯°
ª 1/ 10 º
ª 3/ 10 º
0º
ª
«
»
«
»
«
»
21. Let u = «3/ 20 » , v = « −1/ 20 » , and w = « −1/ 2 » . Since u ⋅ v = u ⋅ w = v ⋅ w = 0, {u, v, w} is an
«
»
«
»
« 1/ 2 »
«¬ −1/ 20 »¼
«¬3/ 20 »¼
¬
¼
orthogonal set. Also, || u ||2 = u ⋅ u = 1, || v ||2 = v ⋅ v = 1, and || w ||2 = w ⋅ w = 1, so {u, v, w} is an
orthonormal set.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.2
• Solutions
365
ª 1/ 18 º
ª 1/ 2 º
ª −2 / 3º
«
»
«
»
0» , and w = «« 1/ 3»» . Since u ⋅ v = u ⋅ w = v ⋅ w = 0, {u, v, w} is an
22. Let u = « 4 / 18 » , v = «
«
»
« −1/ 2 »
«¬ −2 / 3»¼
«¬ 1/ 18 »¼
¬
¼
orthogonal set. Also, || u ||2 = u ⋅ u = 1, || v ||2 = v ⋅ v = 1, and || w ||2 = w ⋅ w = 1, so {u, v, w} is an
orthonormal set.
23. a.
b.
c.
d.
e.
True. For example, the vectors u and y in Example 3 are linearly independent but not orthogonal.
True. The formulas for the weights are given in Theorem 5.
False. See the paragraph following Example 5.
False. The matrix must also be square. See the paragraph before Example 7.
False. See Example 4. The distance is ||y – ŷ ||.
24. a. True. But every orthogonal set of nonzero vectors is linearly independent. See Theorem 4.
b. False. To be orthonormal, the vectors is S must be unit vectors as well as being orthogonal to each
other.
c. True. See Theorem 7(a).
d. True. See the paragraph before Example 3.
e. True. See the paragraph before Example 7.
25. To prove part (b), note that
(Ux) ⋅ (U y) = (Ux)T (U y) = xTU T U y = xT y = x ⋅ y
because U T U = I . If y = x in part (b), (Ux) ⋅ (Ux) = x ⋅ x, which implies part (a). Part (c) of the
Theorem follows immediately fom part (b).
26. A set of n nonzero orthogonal vectors must be linearly independent by Theorem 4, so if such a set
spans W it is a basis for W. Thus W is an n-dimensional subspace of n, and W = n.
27. If U has orthonormal columns, then U T U = I by Theorem 6. If U is also a square matrix, then the
equation U T U = I implies that U is invertible by the Invertible Matrix Theorem.
28. If U is an n × n orthogonal matrix, then I = UU −1 = UU T . Since U is the transpose of U T , Theorem
6 applied to U T says that U T has orthogonal columns. In particular, the columns of U T are linearly
independent and hence form a basis for n by the Invertible Matrix Theorem. That is, the rows of U
form a basis (an orthonormal basis) for n.
29. Since U and V are orthogonal, each is invertible. By Theorem 6 in Section 2.2, UV is invertible and
(UV )−1 = V −1U −1 = V TU T = (UV )T , where the final equality holds by Theorem 3 in Section 2.1. Thus
UV is an orthogonal matrix.
30. If U is an orthogonal matrix, its columns are orthonormal. Interchanging the columns does not
change their orthonormality, so the new matrix – say, V – still has orthonormal columns. By
Theorem 6, V T V = I . Since V is square, V T = V −1 by the Invertible Matrix Theorem.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
366
CHAPTER 6
• Orthogonality and Least Squares
31. Suppose that yˆ =
y ⋅u
u . Replacing u by cu with c ≠ 0 gives
u⋅u
y ⋅ (cu)
y ⋅u
c(y ⋅ u )
c 2 ( y ⋅ u)
u=
u = yˆ
(cu) = 2
(c)u = 2
u ⋅u
(cu) ⋅ (cu)
c (u ⋅ u )
c (u ⋅ u )
So ŷ does not depend on the choice of a nonzero u in the line L used in the formula.
32. If v1 ⋅ v 2 = 0 , then by Theorem 1(c) in Section 6.1,
(c1 v1 ) ⋅ (c2 v 2 ) = c1[ v1 ⋅ (c2 v 2 )] = c1c2 ( v1 ⋅ v 2 ) = c1c2 0 = 0
x⋅u
u . For any vectors x and y in
u ⋅u
scalars c and d, the properties of the inner product (Theorem 1) show that
(cx + dy) ⋅ u
T (cx + dy ) =
u
u⋅u
cx ⋅ u + dy ⋅ u
=
u
u⋅u
cx ⋅ u
dy ⋅ u
=
u+
u
u ⋅u
u ⋅u
= cT (x) + dT (y)
33. Let L = Span{u}, where u is nonzero, and let T (x) =
n
and any
Thus T is a linear transformation. Another approach is to view T as the composition of the following
three linear mappings: x6a = x ⋅ v, a 6b = a / v ⋅ v, and b 6bv.
34. Let L = Span{u}, where u is nonzero, and let T ( x) = refl L y = 2projL y − y . By Exercise 33, the
mapping y 6 projL y is linear. Thus for any vectors y and z in n and any scalars c and d,
T (c y + d z ) = 2 projL (c y + d z ) − (c y + d z )
= 2(c projL y + d projL z ) − c y − d z
= 2c projL y − c y + 2d projL z − d z
= c (2 projL y − y ) + d (2 projL z − z )
= cT (y) + dT (z)
Thus T is a linear transformation.
35. [M] One can compute that AT A = 100I 4 . Since the off-diagonal entries in AT A are zero, the columns
of A are orthogonal.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.3
• Solutions
367
36. [M]
a. One computes that U TU = I 4 , while
ª 82
«
« 0
« −20
«
§ 1 ·« 8
T
UU = ¨
¸
© 100 ¹ « 6
«
« 20
« 24
«
«¬ 0
0
42
24
0
−20
6
20
−32
−20
24
58
20
0
32
0
6
8
0
20
82
24
−20
6
0
6
−20
0
24
18
0
−8
20
20
6
32
−20
0
58
0
24
24
20
0
6
−8
0
18
−20
0º
−32 »»
6»
»
0»
20 »
»
24 »
−20 »
»
42 »¼
The matrices U T U and UU T are of different sizes and look nothing like each other.
b. Answers will vary. The vector p = UU T y is in Col U because p = U (U T y) . Since the columns of
U are simply scaled versions of the columns of A, Col U = Col A. Thus each p is in Col A.
c. One computes that U T z = 0 .
d. From (c), z is orthogonal to each column of A. By Exercise 29 in Section 6.1, z must be
orthogonal to every vector in Col A; that is, z is in (Col A)⊥ .
6.3
SOLUTIONS
Notes: Example 1 seems to help students understand Theorem 8. Theorem 8 is needed for the Gram-
Schmidt process (but only for a subspace that itself has an orthogonal basis). Theorems 8 and 9 are
needed for the discussions of least squares in Sections 6.5 and 6.6. Theorem 10 is used with the QR
factorization to provide a good numerical method for solving least squares problems, in Section 6.5.
Exercises 19 and 20 lead naturally into consideration of the Gram-Schmidt process.
1. The vector in Span{u 4 } is
ª 10 º
« −6 »
x ⋅ u4
72
u 4 = u 4 = 2u 4 = « »
« −2 »
u4 ⋅ u4
36
« »
«¬ 2 »¼
Since x = c1u1 + c2 u 2 + c3u 3 +
x ⋅ u4
u 4 , the vector
u4 ⋅ u4
ª 10 º ª 10 º ª 0º
« −8» « −6 » « −2»
x ⋅ u4
x−
u4 = « » − « » = « »
« 2 » « −2 » « 4»
u4 ⋅ u4
« » « » « »
«¬ 0 »¼ «¬ 2 »¼ «¬ −2»¼
is in Span{u1 , u 2 , u 3 }.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
368
CHAPTER 6
• Orthogonality and Least Squares
2. The vector in Span{u1} is
ª2º
«4»
v ⋅ u1
14
u1 = u1 = 2u1 = « »
«2»
u1 ⋅ u1
7
« »
¬« 2 ¼»
Since x =
v ⋅ u1
u1 + c2 u 2 + c3u 3 + c4 u 4 , the vector
u1 ⋅ u1
ª 4º ª2º ª 2º
« 5» « 4 » « 1»
v ⋅ u1
v−
u1 = « » − « » = « »
« −3» « 2 » « −5»
u1 ⋅ u1
« » « » « »
¬« 3¼» ¬« 2 ¼» ¬« 1¼»
is in Span{u 2 , u3 , u 4 }.
3. Since u1 ⋅ u 2 = −1 + 1 + 0 = 0, {u1 , u 2 } is an orthogonal set. The orthogonal projection of y onto
Span{u1 , u 2 } is
ª1 º
ª −1º ª −1º
3
5
3« » 5« » « »
y ⋅ u1
y ⋅ u2
yˆ =
u1 +
u 2 = u1 + u 2 = «1 » + « 1» = « 4»
2
2
2
2
u1 ⋅ u1
u2 ⋅ u2
«¬0 »¼
«¬ 0 »¼ «¬ 0»¼
4. Since u1 ⋅ u 2 = −12 + 12 + 0 = 0, {u1 , u 2 } is an orthogonal set. The orthogonal projection of y onto
Span{u1 , u 2 } is
ª3º
ª −4º ª 6º
30
15
6« » 3« » « »
y ⋅ u1
y ⋅ u2
yˆ =
u1 +
u 2 = u1 − u 2 = « 4 » − « 3» = « 3»
25
25
5
5
u1 ⋅ u1
u2 ⋅ u2
«¬0 »¼
«¬ 0»¼ «¬ 0»¼
5. Since u1 ⋅ u 2 = 3 + 1 − 4 = 0, {u1 , u 2 } is an orthogonal set. The orthogonal projection of y onto
Span{u1 , u 2 } is
ª 3º
ª 1º ª −1º
y ⋅ u1
y ⋅ u2
7
15
1« » 5« » « »
yˆ =
u1 +
u 2 = u1 − u 2 = « −1» − « −1» = « 2»
14
6
2
2
u1 ⋅ u1
u2 ⋅ u2
«¬ −2 »¼
«¬ 2 »¼ «¬ 6»¼
6. Since u1 ⋅ u 2 = 0 − 1 + 1 = 0, {u1 , u 2 } is an orthogonal set. The orthogonal projection of y onto
Span{u1 , u 2 } is
ª −4 º
ª0 º ª 6 º
27
5
3« » 5« » « »
y ⋅ u1
y ⋅ u2
yˆ =
u1 +
u 2 = − u1 + u 2 = − « −1» + «1 » = « 4 »
18
2
2
2
u1 ⋅ u1
u2 ⋅ u2
«¬ 1»¼
«¬1 »¼ «¬1 »¼
7. Since u1 ⋅ u 2 = 5 + 3 − 8 = 0, {u1 , u 2 } is an orthogonal set. By the Orthogonal Decomposition
Theorem,
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.3
• Solutions
ª10 / 3º
ª −7 / 3º
2
y ⋅ u1
y ⋅ u2
«
»
yˆ =
u1 +
u 2 = 0u1 + u 2 = « 2 / 3» , z = y − yˆ = «« 7 / 3»»
3
u1 ⋅ u1
u2 ⋅ u2
«¬ 8 / 3»¼
«¬ 7 / 3»¼
and y = ŷ + z, where ŷ is in W and z is in W ⊥ .
8. Since u1 ⋅ u 2 = −1 + 3 − 2 = 0, {u1 , u 2 } is an orthogonal set. By the Orthogonal Decomposition
Theorem,
ª 3/ 2 º
ª −5/ 2 º
1
y ⋅ u1
y ⋅ u2
«
»
yˆ =
u1 +
u 2 = 2u1 + u 2 = «7 / 2 » , z = y − yˆ = «« 1/ 2 »»
2
u1 ⋅ u1
u2 ⋅ u2
«¬ 1»¼
«¬
2 »¼
and y = ŷ + z, where ŷ is in W and z is in W ⊥ .
9. Since u1 ⋅ u 2 = u1 ⋅ u3 = u 2 ⋅ u 3 = 0, {u1 , u 2 , u 3 } is an orthogonal set. By the Orthogonal
Decomposition Theorem,
ª 2º
ª 2º
« 4»
« −1»
y ⋅ u3
y ⋅ u1
y ⋅ u2
2
2
«
»
yˆ =
u1 +
u2 +
u3 = 2u1 + u 2 − u3 =
, z = y − yˆ = « »
«0»
« 3»
u1 ⋅ u1
u2 ⋅ u2
u3 ⋅ u3
3
3
« »
« »
«¬ 0 »¼
«¬ −1»¼
and y = ŷ + z, where ŷ is in W and z is in W ⊥ .
10. Since u1 ⋅ u 2 = u1 ⋅ u3 = u 2 ⋅ u 3 = 0, {u1 , u 2 , u 3 } is an orthogonal set. By the Orthogonal
Decomposition Theorem,
ª5 º
ª −2 º
«2»
« 2»
y ⋅ u3
y ⋅ u1
y ⋅ u2
1
14
5
yˆ =
u1 +
u2 +
u3 = u1 + u 2 − u3 = « » , z = y − yˆ = « »
«3»
« 2»
u1 ⋅ u1
u2 ⋅ u2
u3 ⋅ u 3
3
3
3
« »
« »
¬«6 ¼»
¬« 0 ¼»
and y = ŷ + z, where ŷ is in W and z is in W ⊥ .
11. Note that v1 and v 2 are orthogonal. The Best Approximation Theorem says that ŷ , which is the
orthogonal projection of y onto W = Span{v1 , v 2 }, is the closest point to y in W. This vector is
ª 3º
« −1»
y ⋅ v1
y ⋅ v2
1
3
yˆ =
v1 +
v 2 = v1 + v 2 = « »
« 1»
v1 ⋅ v1
v2 ⋅ v2
2
2
« »
¬« −1¼»
12. Note that v1 and v 2 are orthogonal. The Best Approximation Theorem says that ŷ , which is the
orthogonal projection of y onto W = Span{v1 , v 2 }, is the closest point to y in W. This vector is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
369
370
CHAPTER 6
• Orthogonality and Least Squares
ª −1º
« −5»
y ⋅ v1
y ⋅ v2
ˆy =
v1 +
v 2 = 3v1 + 1v 2 = « »
« −3»
v1 ⋅ v1
v2 ⋅ v2
« »
¬« 9 ¼»
13. Note that v1 and v 2 are orthogonal. By the Best Approximation Theorem, the closest point in
Span{v1 , v 2 } to z is
ª −1º
« −3»
z ⋅ v1
z ⋅ v2
2
7
ˆz =
v1 +
v 2 = v1 − v 2 = « »
« −2 »
v1 ⋅ v1
v2 ⋅ v2
3
3
« »
¬« 3¼»
14. Note that v1 and v 2 are orthogonal. By the Best Approximation Theorem, the closest point in
Span{v1 , v 2 } to z is
1º
ª
«
0 »»
z ⋅ v1
z ⋅ v2
1
«
ˆz =
v1 +
v 2 = v1 + 0 v 2 =
« −1/ 2 »
v1 ⋅ v1
v2 ⋅ v2
2
«
»
¬« −3/ 2 ¼»
15. The distance from the point y in 3 to a subspace W is defined as the distance from y to the closest
point in W. Since the closest point in W to y is yˆ = projW y , the desired distance is || y – ŷ ||. One
ª 3º
ª2º
«
»
computes that yˆ = « −9 » , y − yˆ = «« 0 »» , and || y − yˆ || = 40 = 2 10.
«¬ −1»¼
«¬ 6 »¼
16. The distance from the point y in 4 to a subspace W is defined as the distance from y to the closest
point in W. Since the closest point in W to y is yˆ = projW y , the desired distance is || y – ŷ ||. One
ª −1º
ª4º
« −5»
«4»
computes that yˆ = « » , y − yˆ = « » , and || y – ŷ || = 8.
« −3»
«4»
« »
« »
¬« 9 ¼»
¬« 4 ¼»
ª1
17. a. U U = «
¬0
T
ª 8/9
0º
T
,UU = «« −2 / 9
»
1¼
«¬ 2 / 9
−2 / 9
5/9
4/9
2 / 9º
4 / 9 »»
5 / 9 »¼
b. Since U TU = I 2 , the columns of U form an orthonormal basis for W, and by Theorem 10
ª 8/9
projW y = UU y = «« −2 / 9
«¬ 2 / 9
T
−2 / 9
5/9
4/9
2 / 9 º ª 4º ª 2º
4 / 9 »» ««8 »» = «« 4 »» .
5 / 9 »¼ «¬ 1 »¼ «¬ 5 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.3
ª 1/10
18. a. U T U = [1] = 1, UU T = «
¬ −3/10
• Solutions
371
−3/10º
9 /10»¼
b. Since U T U = 1, {u1} forms an orthonormal basis for W, and by Theorem 10
ª 1/10
projW y = UU T y = «
¬ −3/10
−3/10º ª7 º ª −2 º
.
=
9 /10»¼ «¬9 »¼ «¬ 6 »¼
19. By the Orthogonal Decomposition Theorem, u 3 is the sum of a vector in W = Span{u1 , u 2 } and a
vector v orthogonal to W. This exercise asks for the vector v:
0º ª 0º
ª0º ª
1 · « » «
§ 1
v = u 3 − projW u3 = u3 − ¨ − u1 + u 2 ¸ = «0 » − « −2 / 5»» = «« 2 / 5»»
15 ¹
© 3
«¬1 »¼ «¬ 4 / 5»¼ «¬ 1/ 5»¼
Any multiple of the vector v will also be in W ⊥ .
20. By the Orthogonal Decomposition Theorem, u 4 is the sum of a vector in W = Span{u1 , u 2 } and a
vector v orthogonal to W. This exercise asks for the vector v:
0º ª 0º
ª0º ª
1
§1
· « » «
v = u 4 − projW u 4 = u 4 − ¨ u1 − u 2 ¸ = «1 » − « 1/ 5»» = «« 4 / 5»»
30 ¹
©6
«¬0 »¼ «¬ −2 / 5»¼ «¬ 2 / 5»¼
Any multiple of the vector v will also be in W ⊥ .
21. a. True. See the calculations for z 2 in Example 1 or the box after Example 6 in Section 6.1.
b. True. See the Orthogonal Decomposition Theorem.
c. False. See the last paragraph in the proof of Theorem 8, or see the second paragraph after the
statement of Theorem 9.
d. True. See the box before the Best Approximation Theorem.
e. True. Theorem 10 applies to the column space W of U because the columns of U are linearly
independent and hence form a basis for W.
22. a.
b.
c.
d.
True. See the proof of the Orthogonal Decomposition Theorem.
True. See the subsection “A Geometric Interpretation of the Orthogonal Projection.”
True. The orthgonal decomposition in Theorem 8 is unique.
False. The Best Approximation Theorem says that the best approximation to y is projW y.
e. False. This statement is only true if x is in the column space of U. If n > p, then the column space
of U will not be all of n, so the statement cannot be true for all x in n.
23. By the Orthogonal Decomposition Theorem, each x in n can be written uniquely as x = p + u, with
p in Row A and u in (Row A)⊥ . By Theorem 3 in Section 6.1, (Row A)⊥ = Nul A, so u is in Nul A.
Next, suppose Ax = b is consistent. Let x be a solution and write x = p + u as above. Then
Ap = A(x – u) = Ax – Au = b – 0 = b, so the equation Ax = b has at least one solution p in Row A.
Finally, suppose that p and p1 are both in Row A and both satisfy Ax = b. Then p − p1 is in
Nul A = (Row A)⊥ , since A(p − p1 ) = Ap − Ap1 = b − b = 0 . The equations p = p1 + (p − p1 ) and
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
372
CHAPTER 6
• Orthogonality and Least Squares
p = p + 0 both then decompose p as the sum of a vector in Row A and a vector in (Row A)⊥ . By
the uniqueness of the orthogonal decomposition (Theorem 8), p = p1 , and p is unique.
24. a. By hypothesis, the vectors w 1 , …, w p are pairwise orthogonal, and the vectors v1 , …, vq are
pairwise orthogonal. Since w i is in W for any i and v j is in W ⊥ for any j, wi ⋅ v j = 0 for any i
and j. Thus {w1 ,…, w p , v1 ,…, vq } forms an orthogonal set.
b. For any y in
n
, write y = ŷ + z as in the Orthogonal Decomposition Theorem, with ŷ in
W and z in W . Then there exist scalars c1 ,…, c p and d1 ,…, dq such that y = yˆ + z =
⊥
c1w1 +…+ c p w p + d1v1 +…+ dq v q . Thus the set {w1 ,…, w p , v1 ,…, vq } spans
c. The set {w1 ,…, w p , v1 ,…, vq } is linearly independent by (a) and spans
basis for
. Hence dimW + dimW ⊥ = p + q = dim
n
n
n
.
by (b), and is thus a
n
.
25. [M] Since U T U = I 4 , U has orthonormal columns by Theorem 6 in Section 6.2. The closest point to
y in Col U is the orthogonal projection ŷ of y onto Col U. From Theorem 10,
ª1.2 º
« .4 »
« »
«1.2 »
« »
1.2
Τ
yˆ = UU y = « »
« .4 »
« »
«1.2 »
« .4 »
« »
¬« .4 ¼»
26. [M] The distance from b to Col U is || b – b̂ ||, where bˆ = UU Τb. One computes that
ª .2 º
ª .8º
« .92 »
« .08»
«
»
«
»
« .44 »
« .56 »
«
»
«
»
1»
0»
112
bˆ = UU Τb = «
, b − bˆ = «
, || b − bˆ ||=
« −.2 »
« −.8»
5
«
»
«
»
« −.44 »
« −.56 »
« .6 »
« −1.6 »
«
»
«
»
¬« −.92 ¼»
¬« −.08¼»
which is 2.1166 to four decimal places.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.4
6.4
• Solutions
373
SOLUTIONS
Notes: The QR factorization encapsulates the essential outcome of the Gram-Schmidt process, just as the
LU factorization describes the result of a row reduction process. For practical use of linear algebra, the
factorizations are more important than the algorithms that produce them. In fact, the Gram-Schmidt
process is not the appropriate way to compute the QR factorization. For that reason, one should consider
deemphasizing the hand calculation of the Gram-Schmidt process, even though it provides easy exam
questions.
The Gram-Schmidt process is used in Sections 6.7 and 6.8, in connection with various sets of
orthogonal polynomials. The process is mentioned in Sections 7.1 and 7.4, but the one-dimensional
projection constructed in Section 6.2 will suffice. The QR factorization is used in an optional subsection
of Section 6.5, and it is needed in Supplementary Exercise 7 of Chapter 7 to produce the Cholesky
factorization of a positive definite matrix.
ª −1º
x 2 ⋅ v1
v1 = x 2 − 3v1 = «« 5»» . Thus an orthogonal basis for W
1. Set v1 = x1 and compute that v 2 = x 2 −
v1 ⋅ v1
«¬ −3»¼
ª 3º ª −1º ½
°« » « »°
is ® « 0» , « 5» ¾ .
° « −1» « −3» °
¯¬ ¼ ¬ ¼¿
ª 5º
1
x 2 ⋅ v1
v1 = x 2 − v1 = «« 4 »» . Thus an orthogonal basis for W
2. Set v1 = x1 and compute that v 2 = x 2 −
2
v1 ⋅ v1
«¬ −8»¼
ª 0 º ª 5º ½
°
°
is ® «« 4 »» , «« 4 »» ¾ .
° « 2 » « −8» °
¯¬ ¼ ¬ ¼¿
ª 3º
1
x 2 ⋅ v1
v1 = x 2 − v1 = ««3/ 2 »» . Thus an orthogonal basis for
3. Set v1 = x1 and compute that v 2 = x 2 −
2
v1 ⋅ v1
«¬3/ 2 »¼
ª 2 º ª 3º ½
°
°
W is ® «« −5»» , ««3/ 2»» ¾ .
° « 1» «3/ 2» °
¼¿
¯¬ ¼ ¬
ª3º
x 2 ⋅ v1
v1 = x 2 − (−2) v1 = ««6 »» . Thus an orthogonal basis for
4. Set v1 = x1 and compute that v 2 = x 2 −
v1 ⋅ v1
«¬3 »¼
ª 3º ª 3 º ½
°
°
W is ® «« −4 »» , ««6»» ¾ .
° « 5» «3» °
¯¬ ¼ ¬ ¼¿
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
374
CHAPTER 6
• Orthogonality and Least Squares
ª 5º
« 1»
x 2 ⋅ v1
v1 = x 2 − 2 v1 = « » . Thus an orthogonal basis for W
5. Set v1 = x1 and compute that v 2 = x 2 −
« −4 »
v1 ⋅ v1
« »
¬« −1¼»
°
°
is ®
°
°
¯
ª 1º ª 5º
« −4 » « 1»
« », « »
« 0 » « −4 »
« » « »
¬« 1¼» ¬« −1¼»
½
°
°
¾.
°
°
¿
ª 4º
« 6»
x ⋅v
6. Set v1 = x1 and compute that v 2 = x 2 − 2 1 v1 = x 2 − (−3) v1 = « » . Thus an orthogonal basis for
« −3»
v1 ⋅ v1
« »
¬« 0 ¼»
°
°
W is ®
°
°
¯
ª 3º ª 4 º
« −1» « 6 »
« », « »
« 2 » « −3»
« » « »
¬« −1¼» ¬« 0 ¼»
½
°
°
¾.
°
°
¿
7. Since || v1 ||= 30 and || v2 ||= 27 / 2 = 3 6 / 2, an orthonormal basis for W is
v1
v 2 ½ °°
,
®
¾=®
¯ || v1 || || v 2 || ¿ °
¯°
ª 2 / 30 º ª 2 / 6 º ½
«
» «
» °°
« −5 / 30 » , « 1/ 6 » ¾ .
«
» «
»°
¬« 1/ 30 »¼ «¬ 1/ 6 ¼» °¿
8. Since || v1 ||= 50 and || v 2 ||= 54 = 3 6, an orthonormal basis for W is
v1
v 2 ½ °°
,
®
¾=®
¯ || v1 || || v 2 || ¿ °
¯°
ª 3/ 50 º ª 1/ 6 º ½
«
» «
» °°
« −4 / 50 » , « 2 / 6 » ¾.
«
» «
»°
«¬ 5 / 50 »¼ «¬ 1/ 6 »¼ °¿
9. Call the columns of the matrix x1 , x2 , and x3 and perform the Gram-Schmidt process on these
vectors:
v1 = x1
ª 1º
« 3»
x ⋅v
v 2 = x 2 − 2 1 v1 = x 2 − (−2) v1 = « »
« 3»
v1 ⋅ v1
« »
¬« −1¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.4
• Solutions
ª −3º
« 1»
x3 ⋅ v1
x3 ⋅ v 2
3
§ 1·
v 3 = x3 −
v1 −
v 2 = x3 − v1 − ¨ − ¸ v 2 = « »
« 1»
v1 ⋅ v1
v2 ⋅ v2
2
© 2¹
« »
¬« 3¼»
°
°
Thus an orthogonal basis for W is ®
°
°
¯
ª 3º ª 1º ª −3º
« 1» « 3» « 1»
« », « », « »
« −1» « 3» « 1»
« » « » « »
¬« 3¼» ¬« −1¼» ¬« 3¼»
½
°
°
¾.
°
°
¿
10. Call the columns of the matrix x1 , x2 , and x3 and perform the Gram-Schmidt process on these
vectors:
v1 = x1
ª 3º
« 1»
x 2 ⋅ v1
v2 = x2 −
v1 = x 2 − (−3) v1 = « »
« 1»
v1 ⋅ v1
« »
¬« −1¼»
ª −1º
« −1»
x ⋅v
x ⋅v
1
5
v 3 = x3 − 3 1 v1 − 3 2 v 2 = x3 − v1 − v 2 = « »
« 3»
v1 ⋅ v1
v2 ⋅ v2
2
2
« »
¬« −1¼»
°
°
Thus an orthogonal basis for W is ®
°
°
¯
ª −1º ª 3º ª −1º
« 3» « 1» « −1»
« », « », « »
« 1» « 1» « 3»
« » « » « »
«¬ 1»¼ «¬ −1»¼ «¬ −1»¼
½
°
°
¾.
°
°
¿
11. Call the columns of the matrix x1 , x2 , and x3 and perform the Gram-Schmidt process on these
vectors:
v1 = x1
ª 3º
« 0»
« »
x ⋅v
v 2 = x 2 − 2 1 v1 = x 2 − (−1) v1 = « 3»
v1 ⋅ v1
« »
« −3»
«¬ 3»¼
ª 2º
« 0»
« »
x ⋅v
x ⋅v
§ 1·
v 3 = x3 − 3 1 v1 − 3 2 v 2 = x3 − 4 v1 − ¨ − ¸ v 2 = « 2 »
v1 ⋅ v1
v2 ⋅ v2
© 3¹
« »
« 2»
«¬ −2 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
375
376
CHAPTER 6
• Orthogonality and Least Squares
°
°
°
Thus an orthogonal basis for W is ®
°
°
°¯
½
°
°
°
¾.
°
°
°¿
ª 1º ª 3º ª 2 º
« −1» « 0 » « 0 »
« » « » « »
« −1» , « 3» , « 2 »
« » « » « »
« 1» « −3» « 2 »
«¬ 1»¼ «¬ 3»¼ «¬ −2 »¼
12. Call the columns of the matrix x1 , x2 , and x3 and perform the Gram-Schmidt process on these
vectors:
v1 = x1
ª −1º
« 1»
« »
x2 ⋅ v1
v2 = x2 −
v1 = x2 − 4 v1 = « 2 »
v1 ⋅ v1
« »
« 1»
«¬ 1»¼
ª 3º
« 3»
« »
x3 ⋅ v1
x3 ⋅ v 2
7
3
v 3 = x3 −
v1 −
v 2 = x3 − v1 − v 2 = « 0 »
2
2
v1 ⋅ v1
v2 ⋅ v2
« »
« −3»
«¬ 3»¼
°
°
°
Thus an orthogonal basis for W is ®
°
°
°¯
ª 1º ª −1º ª 3º
« −1» « 1» « 3»
« » « » « »
« 0 » , « 2» , « 0»
« » « » « »
« 1» « 1» « −3»
«¬ 1»¼ «¬ 1»¼ «¬ 3»¼
½
°
°
°
¾.
°
°
°¿
13. Since A and Q are given,
ª 5/6
R = QT A = «
¬ −1/ 6
1/ 6
−3/ 6
5/ 6
1/ 6
ª 5
1/ 6º «« 1
3/ 6»¼ « −3
«
¬« 1
9º
7 »» ª 6
=
−5» «¬ 0
»
5¼»
12º
6»¼
14. Since A and Q are given,
ª −2 / 7
R = QT A = «
¬ 5/ 7
5/7
2/7
2/7
−4 / 7
ª −2
4 / 7 º «« 5
2 / 7 »¼ « 2
«
¬« 4
3º
7 »» ª 7
=
−2» «¬ 0
»
6 ¼»
7º
7 »¼
15. The columns of Q will be normalized versions of the vectors v1 , v 2 , and v3 found in Exercise 11.
Thus
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.4
ª 1/
«
« −1/
«
Q = « −1/
«
« 1/
« 1/
¬
5
1/ 2
5
0
5
1/ 2
5
−1/ 2
5
1/ 2
1/ 2 º
»
ª 5
0»
«
»
T
1/ 2 » , R = Q A = « 0
« 0
»
1/ 2 »
¬
−1/ 2 »¼
• Solutions
377
4 5º
»
−2 »
4 »¼
− 5
6
0
16. The columns of Q will be normalized versions of the vectors v1 , v 2 , and v3 found in Exercise 12.
Thus
ª 1/ 2
«
« −1 / 2
«
Q=«
0
«
« 1/ 2
« 1/ 2
¬
−1 / (2 2)
1 / (2 2)
1/ 2
1 / (2 2)
1 / (2 2)
1 / 2º
»
ª2
1 / 2»
«
»
T
0» , R = Q A = «0
«0
»
−1 / 2 »
¬
»
1 / 2¼
8
2 2
0
7º
»
3 2»
6 »¼
17. a. False. Scaling was used in Example 2, but the scale factor was nonzero.
b. True. See (1) in the statement of Theorem 11.
c. True. See the solution of Example 4.
18. a. False. The three orthogonal vectors must be nonzero to be a basis for a three-dimensional
subspace. (This was the case in Step 3 of the solution of Example 2.)
b. True. If x is not in a subspace w, then x cannot equal projW x , because projW x is in W. This idea
was used for v k +1 in the proof of Theorem 11.
c. True. See Theorem 12.
19. Suppose that x satisfies Rx = 0; then Q Rx = Q0 = 0, and Ax = 0. Since the columns of A are linearly
independent, x must be 0. This fact, in turn, shows that the columns of R are linearly indepedent.
Since R is square, it is invertible by the Invertible Matrix Theorem.
20. If y is in Col A, then y = Ax for some x. Then y = QRx = Q(Rx), which shows that y is a linear
combination of the columns of Q using the entries in Rx as weights. Conversly, suppose that y = Qx
for some x. Since R is invertible, the equation A = QR implies that Q = AR−1 . So
y = AR−1x = A( R−1x), which shows that y is in Col A.
21. Denote the columns of Q by {q1 ,…, q n } . Note that n ≤ m, because A is m × n and has linearly
independent columns. The columns of Q can be extended to an orthonormal basis for m as follows.
Let f1 be the first vector in the standard basis for m that is not in Wn = Span{q1 ,…, q n }, let
u1 = f1 − projWn f1 , and let q n+1 = u1 / || u1 || . Then {q1 ,…, q n , q n+1} is an orthonormal basis for
Wn +1 = Span{q1 ,…, q n , q n +1}. Next let f 2 be the first vector in the standard basis for
m
that is
not in Wn +1 , let u 2 = f2 − projWn+1 f2 , and let q n+ 2 = u 2 / || u 2 || . Then {q1 ,…, q n , q n +1 , q n + 2 } is an
orthogonal basis for Wn + 2 = Span{q1 ,…, q n , q n +1 , q n + 2 }. This process will continue until m – n vectors
have been added to the original n vectors, and {q1 ,…, q n , q n +1 ,…, q m } is an orthonormal basis for m.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
378
CHAPTER 6
• Orthogonality and Least Squares
Let Q0 = [qn+1
… qm ] and Q1 = [Q
Q0 ] . Then, using partitioned matrix multiplication,
ªRº
Q1 « » = QR = A .
¬O ¼
22. We may assume that {u1 ,…, u p } is an orthonormal basis for W, by normalizing the vectors in the
original basis given for W, if necessary. Let U be the matrix whose columns are u1 ,…, u p . Then, by
Theorem 10 in Section 6.3, T (x) = projW x = (UU T )x for x in n. Thus T is a matrix transformation
and hence is a linear transformation, as was shown in Section 1.8.
23. Given A = QR, partition A = [ A1
A2 ] , where A1 has p columns. Partition Q as Q = [Q1
ª R11
where Q1 has p columns, and partition R as R = «
¬O
Q2 ]
R12 º
, where R11 is a p × p matrix. Then
R22 »¼
R12 º
ªR
Q2 ] « 11
» = [Q1 R11 Q1R12 + Q2 R22 ]
¬ O R22 ¼
Thus A1 = Q1 R11 . The matrix Q1 has orthonormal columns because its columns come from Q. The
matrix R11 is square and upper triangular due to its position within the upper triangular matrix R. The
diagonal entries of R11 are positive because they are diagonal entries of R. Thus Q1 R11 is a QR
factorization of A1 .
24. [M] Call the columns of the matrix x1 , x2 , x3 , and x4 and perform the Gram-Schmidt process on
these vectors:
v1 = x1
A = [ A1
A2 ] = QR = [Q1
ª 3º
« 3»
« »
x ⋅v
v 2 = x 2 − 2 1 v1 = x 2 − (−1) v1 = « −3»
v1 ⋅ v1
« »
« 0»
«¬ 3»¼
ª6º
«0»
« »
x3 ⋅ v1
x3 ⋅ v 2
§ 1·
§ 4·
v 3 = x3 −
v1 −
v 2 = x3 − ¨ − ¸ v1 − ¨ − ¸ v 2 = «6 »
v1 ⋅ v1
v2 ⋅ v2
© 2¹
© 3¹
« »
«6»
«¬0 »¼
ª 0º
« 5»
« »
⋅
x
v
1
x ⋅v
x ⋅v
§ 1·
v 4 = x 4 − 4 1 v1 − 4 2 v 2 − 4 3 v 3 = x 4 − v1 − (−1) v 2 − ¨ − ¸ v 3 = « 0 »
v1 ⋅ v1
v2 ⋅ v2
v3 ⋅ v3
2
« »
© 2¹
« 0»
«¬ −5»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.5
°
°°
Thus an orthogonal basis for W is ®
°
°
°¯
ª −10 º ª 3º ª 6º ª 0º
« 2 » « 3» « 0 » « 5 »
«
» « » « » « »
« −6 » , « −3» , « 6» , « 0 »
«
» « » « » « »
« 16 » « 0 » « 6» « 0»
«¬ 2 »¼ «¬ 3»¼ «¬ 0»¼ «¬ −5»¼
• Solutions
379
½
°
°°
¾.
°
°
°¿
25. [M] The columns of Q will be normalized versions of the vectors v1 , v 2 , and v3 found in Exercise
24. Thus
ª −1/ 2
«
« 1/10
«
Q = « −3/10
«
« 4/5
« 1/10
¬
1/ 2
1/ 3
1/ 2
0
−1/ 2
1/ 3
0
1/ 3
1/ 2
0
0º
»
ª 20
1/ 2 »
« 0
»
«
T
,
R
Q
A
=
=
0»
« 0
»
«
0»
«¬ 0
−1/ 2 »¼
−20
−10
6
−8
0
6 3
0
0
10 º
−6 »»
−3 3 »
»
5 2 »¼
26. [M] In MATLAB, when A has n columns, suitable commands are
Q = A(:,1)/norm(A(:,1))
% The first column of Q
for j=2: n
v=A(:,j) – Q*(Q’*A(:,j))
Q(:,j)=v/norm(v)
% Add a new column to Q
end
6.5
SOLUTIONS
Notes: This is a core section – the basic geometric principles in this section provide the foundation
for all the applications in Sections 6.6–6.8. Yet this section need not take a full day. Each example
provides a stopping place. Theorem 13 and Example 1 are all that is needed for Section 6.6. Theorem 15,
however, gives an illustration of why the QR factorization is important. Example 4 is related to Exercise
17 in Section 6.6.
1. To find the normal equations and to find x̂ , compute
ª −1
A A=«
¬ 2
T
ª −1
A b=«
¬ 2
T
2
−3
2
−3
ª −1
−1º «
2
3»¼ «
«¬ −1
−1º
3¼»
2º
ª 6
−3»» = «
−11
3»¼ ¬
−11º
22 »¼
ª4º
«1 » = ª −4º
« » « 11 »
«¬ 2»¼ ¬ ¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
380
CHAPTER 6
• Orthogonality and Least Squares
−11º ª x1 º ª −4º
.
=
22»¼ «¬ x2 »¼ «¬ 11»¼
ª 6
a. The normal equations are ( AT A)x = AT b : «
¬ −11
b. Compute
−1
−11º ª −4 º 1 ª 22 11º ª −4º
=
22 »¼ «¬ 11»¼ 11 «¬ 11 6»¼ «¬ 11»¼
ª 6
x̂ = ( AT A) −1 AT b = «
¬ −11
=
1 ª 33º ª 3º
=
11 «¬ 22»¼ «¬ 2»¼
2. To find the normal equations and to find xˆ , compute
ª2
A A=«
¬1
−2
ª2
A b=«
¬1
−2
T
0
T
ª 2
2º «
−2
3»¼ «
¬« 2
1º
8º
ª12
0 »» = «
8 10»¼
¬
3¼»
ª −5º
2 º « » ª −24º
8 =
3»¼ « » «¬ −2»¼
«¬ 1»¼
0
ª12
a. The normal equations are ( AT A)x = AT b : «
¬8
b. Compute
ª12
x̂ = ( A A) A b = «
¬ 8
T
−1
T
=
8 º ª x1 º ª −24 º
.
=
10 »¼ «¬ x2 »¼ «¬ −2 »¼
−1
8º ª −24º 1 ª 10
=
10 »¼ «¬ −2 »¼ 56 «¬ −8
−8º ª −24 º
12»¼ «¬ −2 »¼
1 ª −224º ª −4º
=
56 «¬ 168 »¼ «¬ 3 »¼
3. To find the normal equations and to find x̂ , compute
ª 1
AT A = «
¬ −2
ª 1
AT b = «
¬ −2
−1
0
2
3
−1
0
2
3
ª 1
2 º «« −1
5»¼ « 0
«
¬« 2
−2 º
2 »» ª 6
=
3» «¬ 6
»
5¼»
6º
42»¼
ª 3º
2 º «« 1»» ª 6 º
=
5»¼ « −4 » «¬ −6 »¼
« »
«¬ 2 »¼
ª6
a. The normal equations are ( AT A)x = AT b : «
¬6
b. Compute
6º ª x1 º ª 6º
=
42»¼ «¬ x2 »¼ «¬ −6»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.5
−1
ª6
xˆ = (Α Α) Α b = «
¬6
T
−1
6º ª 6º
1 ª 42
=
»
«
»
42 ¼ ¬ −6¼ 216 «¬ −6
T
=
• Solutions
381
−6º ª 6º
6»¼ «¬ −6»¼
1 ª 288º ª 4 / 3º
=
216 «¬ −72»¼ «¬ −1/ 3»¼
4. To find the normal equations and to find x̂ , compute
ª1
A A=«
¬3
ª1
1 1º «
1
−1 1»¼ «
«¬1
ª1
A b=«
¬3
ª5 º
1 1º « » ª 6 º
1 =
−1 1»¼ « » «¬14 »¼
«¬0 »¼
T
T
3º
ª 3 3º
−1»» = «
3 11»¼
¬
»
1¼
ª3 3º ª x1 º ª 6º
a. The normal equations are ( AT A)x = AT b : «
»« » = « »
¬3 11¼ ¬ x2 ¼ ¬14¼
b. Compute
−1
ª3
xˆ = (Α Α) Α b = «
¬3
T
−1
3 º ª 6 º 1 ª 11
=
11»¼ «¬14 »¼ 24 «¬ −3
T
=
−3º ª 6 º
3»¼ «¬14»¼
1 ª 24 º ª1º
=
24 «¬ 24 »¼ «¬1»¼
5. To find the least squares solutions to Ax = b, compute and row reduce the augmented matrix for the
system AT Ax = AT b :
ª
¬A A
T
ª4
A b ¼º = «« 2
«¬ 2
T
2
2
2
0
0
2
14 º ª 1
4 »» ∼ ««0
10 »¼ «¬0
0
1
1
−1
0
0
5º
−3»»
0 »¼
ª 5º
ª −1º
«
»
so all vectors of the form xˆ = « −3» + x3 «« 1»» are the least-squares solutions of Ax = b.
¬« 0 ¼»
¬« 1¼»
6. To find the least squares solutions to Ax = b, compute and row reduce the augmented matrix for the
system AT Ax = AT b :
ª AT A
¬
ª6
º
A b ¼ = «« 3
«¬ 3
T
3
3
3
0
0
3
27 º ª 1
12 »» ∼ ««0
15»¼ «¬0
0
1
1
−1
0
0
5º
−1»»
0»¼
ª 5º
ª −1º
«
»
so all vectors of the form xˆ = « −1» + x3 «« 1»» are the least-squares solutions of Ax = b.
«¬ 0 »¼
«¬ 1»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
382
CHAPTER 6
• Orthogonality and Least Squares
ª 1
« −1
7. From Exercise 3, A = «
« 0
«
¬« 2
ª 1
« −1
Axˆ − b = «
« 0
«
¬« 2
−2 º
ª 3º
»
« 1»
2»
ª 4/3º
, b = « » , and xˆ = «
» . Since
« −4 »
3»
¬ −1/3¼
»
« »
5¼»
¬« 2 ¼»
−2 º
ª 3º ª 2º ª 3º ª −1º
»
2» ª 4/3º «« 1»» «« −2»» «« 1»» «« −3»»
−
=
−
=
3» ¬« −1/3¼» « −4 » « −1» « −4 » « 3»
»
« » « » « » « »
5¼»
¬« 2¼» ¬« 1¼» ¬« 2¼» ¬« −1¼»
the least squares error is || Axˆ − b ||= 20 = 2 5.
ª1
8. From Exercise 4, A = ««1
«¬1
ª1
Axˆ − b = ««1
«¬1
3º
ª5º
ª1º
»
−1» , b = ««1 »» , and xˆ = « » . Since
¬1¼
«¬0 »¼
1»¼
3º
ª5º ª 4º ª 5º ª −1º
ª1º « » « » « » « »
»
−1» « » − «1 » = « 0 » − «1 » = « −1»
1
1»¼ ¬ ¼ «¬ 0»¼ «¬ 2»¼ «¬ 0»¼ «¬ 2»¼
the least squares error is || Axˆ − b ||= 6.
9. (a) Because the columns a1 and a 2 of A are orthogonal, the method of Example 4 may be used to
find b̂ , the orthogonal projection of b onto Col A:
ª 1º
ª 5 º ª1 º
ˆb = b ⋅ a1 a + b ⋅ a 2 a = 2 a + 1 a = 2 « 3» + 1 «1 » = «1 »
1
2
1
2
7
7
7« » 7« » « »
a1 ⋅ a1
a2 ⋅ a2
«¬ −2 »¼
«¬ 4 »¼ «¬ 0»¼
(b) The vector x̂ contains the weights which must be placed on a1 and a 2 to produce b̂ . These
ª 2/7º
weights are easily read from the above equation, so xˆ = « » .
¬1/7 ¼
10. (a) Because the columns a1 and a 2 of A are orthogonal, the method of Example 4 may be used to
find b̂ , the orthogonal projection of b onto Col A:
ª 1º
ª 2º ª 4º
ˆb = b ⋅ a1 a + b ⋅ a 2 a = 3a + 1 a = 3 « −1» + 1 « 4 » = « −1»
1
2
1
2
« » 2« » « »
2
a1 ⋅ a1
a2 ⋅ a 2
«¬ 1»¼
«¬ 2 »¼ «¬ 4»¼
(b) The vector x̂ contains the weights which must be placed on a1 and a 2 to produce b̂ . These
ª 3º
weights are easily read from the above equation, so xˆ = « » .
¬1/ 2 ¼
11. (a) Because the columns a1 , a 2 and a3 of A are orthogonal, the method of Example 4 may be used
to find b̂ , the orthogonal projection of b onto Col A:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.5
• Solutions
383
b ⋅ a3
2
1
b ⋅ a1
b ⋅ a2
bˆ =
a1 +
a2 +
a3 = a1 + 0a2 + a3
3
3
a1 ⋅ a1
a2 ⋅ a2
a3 ⋅ a3
ª 4º
ª 0º
ª 1º ª 3º
«1 »
« −5»
« » « »
2« »
1 « 1» « 1»
«
»
=
+0
+
=
« 1» 3 « 0» « 4»
3 «6 »
« »
« »
« » « »
¬«1 ¼»
¬« −1¼»
¬« −5¼» ¬« −1¼»
(b) The vector x̂ contains the weights which must be placed on a1 , a 2 , and a3 to produce b̂ . These
ª 2 / 3º
weights are easily read from the above equation, so xˆ = «« 0 »» .
«¬ 1/ 3»¼
12. (a) Because the columns a1 , a 2 and a3 of A are orthogonal, the method of Example 4 may be used
to find b̂ , the orthogonal projection of b onto Col A:
b ⋅ a3
1
14
b ⋅ a1
b ⋅ a2
§ 5·
bˆ =
a1 +
a2 +
a3 = a1 + a 2 + ¨ − ¸ a3
a1 ⋅ a1
a2 ⋅ a2
a3 ⋅ a 3
3
3
© 3¹
ª1º
ª1 º
ª 0 º ª5º
«1»
«0»
« » « »
1 « » 14 « » 5 « −1» « 2»
=
+
−
=
3 « 0 » 3 «1 » 3 « 1 » « 3 »
« »
« »
« » « »
«¬ −1»¼
«¬1 »¼
«¬ −1»¼ «¬ 6 »¼
(b) The vector x̂ contains the weights which must be placed on a1 , a 2 , and a3 to produce b̂ . These
ª 1/3º
weights are easily read from the above equation, so xˆ = «« 14/3»» .
«¬ −5/3»¼
13. One computes that
ª 11º
ª 0º
«
»
Au = « −11» , b − Au = «« 2»» , || b − Au ||= 40
«¬ 11»¼
«¬ −6»¼
ª 7º
ª 4º
«
»
Av = « −12 » , b − Av = «« 3»» , || b − Av ||= 29
«¬ 7 »¼
«¬ −2»¼
Since Av is closer to b than Au is, Au is not the closest point in Col A to b. Thus u cannot be a leastsquares solution of Ax = b.
14. One computes that
ª3º
ª 2º
«
»
Au = «8 » , b − Au = «« −4 »» , || b − Au ||= 24
«¬ 2 »¼
«¬ 2 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
384
CHAPTER 6
• Orthogonality and Least Squares
ª7 º
ª −2 º
«
»
Av = « 2 » , b − Av = «« 2 »» , || b − Av ||= 24
«¬8 »¼
«¬ −4 »¼
Since Au and Au are equally close to b, and the orthogonal projection is the unique closest point in
Col A to b, neither Au nor Av can be the closest point in Col A to b. Thus neither u nor v can be a
least-squares solution of Ax = b.
ª3
15. The least squares solution satisfies Rxˆ = QT b. Since R = «
¬0
matrix for the system may be row reduced to find
ªR
¬
ª3
QT b º¼ = «
¬0
5
1
7º ª 1
∼
−1¼» ¬«0
5º
ª 7º
and QT b = « » , the augmented
»
1¼
¬ −1¼
4º
−1¼»
0
1
ª 4º
and so xˆ = « » is the least squares solution of Ax = b.
¬ −1¼
ª2
16. The least squares solution satisfies Rxˆ = QT b. Since R = «
¬0
matrix for the system may be row reduced to find
ªR
¬
ª2
QT b º¼ = «
¬0
3 17 / 2º ª 1
∼
5 9 / 2»¼ «¬0
0
1
3º
ª17 / 2 º
and QT b = «
»
» , the augmented
5¼
¬ 9 / 2¼
2.9º
.9»¼
ª 2.9º
and so xˆ = « » is the least squares solution of Ax = b.
¬ .9¼
17. a.
b.
c.
d.
e.
True. See the beginning of the section. The distance from Ax to b is || Ax – b ||.
True. See the comments about equation (1).
False. The inequality points in the wrong direction. See the definition of a least-squares solution.
True. See Theorem 13.
True. See Theorem 14.
18. a. True. See the paragraph following the definition of a least-squares solution.
b. False. If x̂ is the least-squares solution, then A x̂ is the point in the column space of A closest to
b. See Figure 1 and the paragraph preceding it.
c. True. See the discussion following equation (1).
d. False. The formula applies only when the columns of A are linearly independent. See Theorem
14.
e. False. See the comments after Example 4.
f. False. See the Numerical Note.
19. a. If Ax = 0, then AT Ax = AT 0 = 0. This shows that Nul A is contained in Nul AT A.
b. If AT Ax = 0, then xT AT Ax = xT 0 = 0. So ( Ax)T ( Ax) = 0, which means that || Ax ||2 = 0, and
hence Ax = 0. This shows that Nul AT A is contained in Nul A.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.6
• Solutions
385
20. Suppose that Ax = 0. Then AT Ax = AT 0 = 0. Since AT A is invertible, x must be 0. Hence the
columns of A are linearly independent.
21. a. If A has linearly independent columns, then the equation Ax = 0 has only the trivial solution. By
Exercise 19, the equation AT Ax = 0 also has only the trivial solution. Since AT A is a square
matrix, it must be invertible by the Invertible Matrix Theorem.
b. Since the n linearly independent columns of A belong to m, m could not be less than n.
c. The n linearly independent columns of A form a basis for Col A, so the rank of A is n.
22. Note that AT A has n columns because A does. Then by the Rank Theorem and Exercise 19,
rank AT A = n − dim Nul AT A = n − dim Nul A = rank A
23. By Theorem 14, bˆ = Axˆ = A( AT A)−1 AT b. The matrix A( AT A)−1 AT is sometimes called the hatmatrix in statistics.
24. Since in this case AT A = I , the normal equations give xˆ = AT b.
ª 2 2º ª x º ª6º
25. The normal equations are «
» « » = « » , whose solution is the set of all (x, y) such that x + y =
¬ 2 2¼ ¬ y ¼ ¬6¼
3. The solutions correspond to the points on the line midway between the lines x + y = 2 and x + y =
4.
26. [M] Using .7 as an approximation for
approximation for
6.6
2 / 2, a0 = a2 ≈ .353535 and a1 = .5. Using .707 as an
2 / 2 , a0 = a2 ≈ .35355339 , a1 = .5.
SOLUTIONS
Notes: This section is a valuable reference for any person who works with data that requires statistical
analysis. Many graduate fields require such work. Science students in particular will benefit from
Example 1. The general linear model and the subsequent examples are aimed at students who may take a
multivariate statistics course. That may include more students than one might expect.
1. The design matrix X and the observation vector y are
ª1
«1
X =«
«1
«
¬«1
0º
ª1 º
»
«1 »
1»
, y = « »,
« 2»
2»
»
« »
3 ¼»
¬« 2¼»
and one can compute
ª 4 6º T
ª 6º
ª.9 º
, X y = « » , βˆ = ( X T X ) −1 X T y = « »
XT X = «
»
¬ 6 14¼
¬11¼
¬.4¼
The least-squares line y = β 0 + β1 x is thus y = .9 + .4x.
2. The design matrix X and the observation vector y are
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
386
CHAPTER 6
ª1
«1
X =«
«1
«
¬«1
• Orthogonality and Least Squares
1º
ª0 º
»
«1 »
2»
, y = « »,
« 2»
4»
»
« »
5¼»
¬« 3¼»
and one can compute
ª 4 12 º T
ª 6º
ª −.6 º
, X y = « » , βˆ = ( X T X ) −1 X T y = « »
XT X = «
»
¬12 46 ¼
¬ 25¼
¬ .7 ¼
The least-squares line y = β 0 + β1 x is thus y = –.6 + .7x.
3. The design matrix X and the observation vector y are
ª1
«1
X =«
«1
«
«¬1
−1º
ª0º
»
«1 »
0»
, y = « »,
« 2»
1»
»
« »
2 »¼
«¬ 4»¼
and one can compute
ª 4 2º T
ª 7º
ª1.1º
, X y = « » , βˆ = ( X T X ) −1 X T y = « »
XT X = «
»
¬2 6¼
¬10¼
¬1.3¼
The least-squares line y = β 0 + β1 x is thus y = 1.1 + 1.3x.
4. The design matrix X and the observation vector y are
ª1
«1
X =«
«1
«
¬«1
2º
ª 3º
»
« 2»
3»
, y = « »,
«1 »
5»
»
« »
6 ¼»
¬«0 ¼»
and one can compute
ª 4 16º T
ª 6º
ª 4.3º
, X y = « » , βˆ = ( X T X )−1 X T y = « »
XT X = «
»
¬16 74¼
¬17 ¼
¬ −.7 ¼
The least-squares line y = β 0 + β1 x is thus y = 4.3 – .7x.
5. If two data points have different x-coordinates, then the two columns of the design matrix X cannot
be multiples of each other and hence are linearly independent. By Theorem 14 in Section 6.5, the
normal equations have a unique solution.
6. If the columns of X were linearly dependent, then the same dependence relation would hold for the
vectors in 3 formed from the top three entries in each column. That is, the columns of the matrix
ª1 x1 x12 º
«
»
2
«1 x2 x2 » would also be linearly dependent, and so this matrix (called a Vandermonde matrix)
«
2»
¬«1 x3 x3 ¼»
would be noninvertible. Note that the determinant of this matrix is ( x2 − x1 )( x3 − x1 )( x3 − x2 ) ≠ 0
since x1 , x2 , and x3 are distinct. Thus this matrix is invertible, which means that the columns of X
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.6
• Solutions
387
are in fact linearly independent. By Theorem 14 in Section 6.5, the normal equations have a unique
solution.
7. a. The model that produces the correct least-squares fit is y = Xβ + where
ª1
«2
«
X = «3
«
«4
«¬ 5
1º
ª 1 º
ª 1.8º
« »
»
«
»
4»
« 2»
« 2.7 »
ª β1 º
9 » , y = « 3.4 » , β = « » , and = «3 »
« »
»
« »
¬ β2 ¼
16 »
«4 »
« 3.8»
«¬5 »¼
«¬ 3.9 »¼
25»¼
ª 1.76º
b. [M] One computes that (to two decimal places) βˆ = «
» , so the desired least-squares equation
¬ −.20¼
is y = 1.76 x − .20 x2 .
8. a. The model that produces the correct least-squares fit is y = Xβ + where
ª x1
«
X =« #
«x
¬ n
x12
#
xn2
x13 º
ª y1 º
ª β1 º
ª 1 º
»
«
»
«
»
# » , y = « # » , β = « β 2 » , and = «« # »»
«¬ yn »¼
«¬ β 3 »¼
«¬n »¼
xn3 »¼
b. [M] For the given data,
ª 4
« 6
«
« 8
«
10
X =«
«12
«
«14
«16
«
¬«18
16
36
64
100
144
196
256
324
64 º
ª 1.58º
»
« 2.08»
216 »
«
»
« 2.5»
512 »
»
«
»
1000 »
2.8»
«
and y =
« 3.1»
1728»
»
«
»
2744 »
« 3.4 »
« 3.8»
4096 »
»
«
»
5832 ¼»
¬« 4.32 ¼»
ª .5132º
T
−1 T
ˆ
so β = ( X X ) X y = «« −.03348»» , and the least-squares curve is
«¬ .001016»¼
y = .5132 x − .03348x2 + .001016 x3 .
9. The model that produces the correct least-squares fit is y = Xβ + where
ª 1 º
ª cos 1 sin 1º
ª 7.9º
ª Aº
«
»
«
»
X = «cos 2 sin 2» , y = « 5.4» , β = « » , and = ««2 »»
¬B¼
«¬ cos 3 sin 3»¼
«¬ −.9»¼
«¬3 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
388
CHAPTER 6
• Orthogonality and Least Squares
10. a. The model that produces the correct least-squares fit is y = Xβ + where
ª e −.02(10)
« −.02(11)
«e
X = «« e −.02(12)
« e −.02(14)
«
«¬ e −.02(15)
e −.07(10) º
ª 1 º
ª 21.34 º
»
« »
« 20.68»
e −.07(11) »
« 2»
«
»
ªM A º
−.07(12) »
« »
«
»
e
» , y = « 20.05» , β = « M » , and = «3 » ,
¬ B¼
−.07(14) »
e
«4 »
« 18.87 »
»
«¬5 »¼
«
»
¬ 18.30 ¼
e −.07(15) »¼
ª19.94º
b. [M] One computes that (to two decimal places) βˆ = «
» , so the desired least-squares
¬10.10¼
equation is y = 19.94e−.02t + 10.10e−.07t .
11. [M] The model that produces the correct least-squares fit is y = Xβ + where
3 cos .88º
ª 1 º
ª1
ª 3º
« »
«1
»
«
»
2.3 cos 1.1»
« 2»
«
« 2.3»
ªβ º
X = «1 1.65 cos 1.42 » , y = «1.65» , β = « » , and = «3 »
« »
«
»
«
»
¬e¼
«4 »
«1 1.25 cos 1.77 »
«1.25»
«¬5 »¼
«¬1 1.01 cos 2.14 »¼
«¬1.01»¼
ª1.45º
One computes that (to two decimal places) βˆ = «
» . Since e = .811 < 1 the orbit is an ellipse. The
¬.811¼
equation r = β / (1 – e cos ϑ) produces r = 1.33 when ϑ = 4.6.
12. [M] The model that produces the correct least-squares fit is y = Xβ + , where
ª1
«1
«
X = «1
«
«1
«¬1
3.78º
ª 1 º
ª 91º
« »
»
«
»
4.11»
« 2»
« 98»
ª β0 º
4.39» , y = «103» , β = « » , and = «3 »
« »
»
« »
¬ β1 ¼
4.73»
«4 »
«110»
«¬5 »¼
»
«
»
4.88¼
¬112¼
ª18.56º
One computes that (to two decimal places) βˆ = «
» , so the desired least-squares equation is
¬19.24¼
p = 18.56 + 19.24 ln w. When w = 100, p ≈ 107 millimeters of mercury.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.6
• Solutions
13. [M]
a. The model that produces the correct least-squares fit is y = Xβ + where
0
0º
ª1 0
«1
ª 0 º
ª 0 º
1
1
1»»
«
«»
« 8.8 »
«1 2
« 1»
»
«
22
23 »
«
»
« 2 »
»
«
29.9
«1 3
32
33 »
« »
»
«
«
« 3 »
« 62.0 »
2
3»
4
4 »
«1 4
« 4 »
«104.7 »
«
β
« »
»
«
2
3»
ª
º
0
5
5 »
«1 5
« 5 »
« 159.1 »
«β »
2
3»
1»
«
«
»
«
6
6 , y = 222.0 , β =
, and = « 6 »
X= 1 6
«
»
« »
«
»
«
»
β
2
«1 7
« 7 »
« 294.5»
72
73 »
« »
β3 ¼
«
»
« »
¬
«
»
82
83 »
«1 8
« 8»
« 380.4 »
«
« 9 »
« 471.1»
2
3»
9
9 »
«1 9
« »
«
»
«1 10 102 103 »
«10 »
«571.7 »
«
»
« 11 »
« 686.8 »
«1 11 112 113 »
« »
«
»
«
»
¬«809.2 »¼
¬«12 ¼»
2
3
12 »¼
¬«1 12 12
ª −.8558º
« 4.7025»
ˆ
» , so the desired least-squares
One computes that (to four decimal places) β = «
« 5.5554 »
«
»
«¬ −.0274 »¼
polynomial is y(t ) = −.8558 + 4.7025t + 5.5554t 2 − .0274t 3 .
b. The velocity v(t) is the derivative of the position function y(t), so
v(t ) = 4.7025 + 11.1108t − .0822t 2 , and v(4.5) = 53.0 ft/sec.
14. Write the design matrix as [1
x]. Since the residual vector = y – X βˆ is orthogonal to Col X,
0 = 1 ⋅ = 1 ⋅ (y − X βˆ ) = 1T y − (1T X )βˆ
= ( y1 + … + yn ) − ª¬ n
ª βˆ0 º
º
x
¦ ¼ « ˆ » = ¦ y − nβˆ0 − βˆ1 ¦ x = ny − n βˆ0 − nβˆ1 x
¬« β1 ¼»
This equation may be solved for y to find y = βˆ0 + βˆ1 x .
15. From equation (1) on page 369,
ª1
X X =«
¬ x1
T
ª1
X y=«
¬ x1
T
…
…
…
…
ª1
1 º«
#
xn »¼ «
«¬1
x1 º
ª n
# »» = «
« x
xn »¼ ¬¦
¦ x º»
2
¦ x »¼
ªy º
1 º« 1» ª ¦y º
# =«
»
xn ¼» « » ¬¦ xy ¼
«¬ yn »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
389
390
CHAPTER 6
• Orthogonality and Least Squares
The equations (7) in the text follow immediately from the normal equations X T X β = X T y.
16. The determinant of the coefficient matrix of the equations in (7) is n¦ x 2 − (¦ x)2 . Using the 2 × 2
formula for the inverse of the coefficient matrix,
ª βˆ0 º
1
« »=
2
2
ˆ
¬« β1 ¼» n¦ x − (¦ x)
ª ¦ x2
«
¬« −¦ x
−¦ x º ª ¦ y º
»«
»
n ¼» ¬ ¦ xy ¼
Hence
(¦ x 2 )(¦ y ) − (¦ x)(¦ xy ) ˆ n¦ xy − (¦ x)(¦ y )
ˆ
, β1 =
β0 =
n ¦ x 2 − (¦ x) 2
n ¦ x 2 − (¦ x ) 2
Note: A simple algebraic calculation shows that
¦ y − (¦ x) βˆ1 = nβˆ0 , which provides a simple
formula for β̂0 once β̂1 is known
The mean of the data in Example 1 is x = 5.5, so the data in mean-deviation form are (–3.5, 1),
ª1 −3.5º
«1
−.5»»
«
. The columns of X are
(–.5, 2), (1.5, 3), (2.5, 3), and the associated design matrix is X =
«1
1.5»
«
»
2.5»¼
«¬1
orthogonal because the entries in the second column sum to 0.
17. a.
0º ª β 0 º ª 9º
ª4
b. The normal equations are X T X β = X T y, or «
» « » = « » . One computes that
¬ 0 21¼ ¬ β1 ¼ ¬7.5¼
ª 9 / 4º
*
βˆ = «
» , so the desired least-squares line is y = (9/ 4) + (5/14) x = (9/ 4) + (5/14)( x − 5.5).
5/14
¬
¼
18. Since
ª1
X X =«
¬ x1
T
…
…
ª1
1 º«
#
xn »¼ «
«¬1
X T X is a diagonal matrix when
x1 º
ª n
# »» = «
« x
xn »¼ ¬¦
¦ x º»
2
¦ x »¼
¦ x = 0.
19. The residual vector = y – X βˆ is orthogonal to Col X, while ŷ =X βˆ is in Col X. Since and ŷ are
thus orthogonal, apply the Pythagorean Theorem to these vectors to obtain
SS(T) = || y ||2 = || yˆ + ||2 = || yˆ ||2 + || ||2 = || X βˆ ||2 + || y − X βˆ ||2 = SS(R) + SS(E)
20. Since βˆ satisfies the normal equations, X T X βˆ = X T y, and
|| X βˆ ||2 = ( X βˆ )T ( X βˆ ) = βˆ T X T X βˆ = βˆ T X T y
Since || X βˆ ||2 = SS(R) and yT y = || y ||2 = SS(T) , Exercise 19 shows that
SS(E) = SS(T) − SS(R) = yT y − βˆ T X T y
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.7
6.7
• Solutions
391
SOLUTIONS
Notes: The three types of inner products described here (in Examples 1, 2, and 7) are matched by
examples in Section 6.8. It is possible to spend just one day on selected portions of both sections.
Example 1 matches the weighted least squares in Section 6.8. Examples 2–6 are applied to trend analysis
in Seciton 6.8. This material is aimed at students who have not had much calculus or who intend to take
more than one course in statistics.
For students who have seen some calculus, Example 7 is needed to develop the Fourier series in
Section 6.8. Example 8 is used to motivate the inner product on C[a, b]. The Cauchy-Schwarz and
triangle inequalities are not used here, but they should be part of the training of every mathematics
student.
1. The inner product is ¢ x, y ² = 4 x1 y1 + 5 x2 y2 . Let x = (1, 1), y = (5, –1).
a. Since || x ||2 = ¢ x, x² = 9, || x || = 3. Since || y ||2 = ¢ y, y² = 105, || y ||= 105. Finally,
| ¢ x, y² |2 = 152 = 225.
b. A vector z is orthogonal to y if and only if ¢x, y² = 0, that is, 20 z1 − 5 z2 = 0, or 4 z1 = z 2 . Thus
ª1 º
all multiples of « » are orthogonal to y.
¬ 4¼
2. The inner product is ¢ x, y ² = 4 x1 y1 + 5 x2 y2 . Let x = (3, –2), y = (–2, 1). Compute that
|| x ||2 = ¢ x, x² = 56, || y ||2 = ¢ y, y² = 21, || x ||2 || y ||2 = 56 ⋅ 21 = 1176 , ¢x, y² = –34, and | ¢ x, y² |2 = 1156 .
Thus | ¢ x, y² |2 ≤ || x ||2 || y ||2 , as the Cauchy-Schwarz inequality predicts.
3. The inner product is ¢ p, q² = p(–1)q(–1) + p(0)q(0) + p(1)q(1), so
¢ 4 + t ,5 − 4t 2 ² = 3(1) + 4(5) + 5(1) = 28 .
4. The inner product is ¢ p, q² = p(–1)q(–1) + p(0)q(0) + p(1)q(1), so ¢3t − t 2 , 3 + 2t 2 ² =
(−4)(5) + 0(3) + 2(5) = −10.
5. The inner product is ¢ p, q² = p(–1)q(–1) + p(0)q(0) + p(1)q(1), so
¢ p, p² = ¢ 4 + t ,4 + t ² = 32 + 42 + 52 = 50 and || p ||= ¢ p, p² = 50 = 5 2 . Likewise
¢ q, q² = ¢5 − 4t 2 ,5 − 4t 2 ² = 12 + 52 + 12 = 27 and || q ||= ¢ q, q² = 27 = 3 3 .
6. The inner product is ¢ p, q² = p(–1)q(–1) + p(0)q(0) + p(1)q(1), so ¢ p, p² = ¢3t − t 2 ,3t − t 2 ² =
(−4)2 + 02 + 22 = 20 and || p ||= ¢ p, p² = 20 = 2 5. Likewise ¢ q, q² = ¢3 + 2t 2 ,3 + 2t 2 ² =
52 + 32 + 52 = 59 and || q ||= ¢ q, q² = 59.
7. The orthogonal projection q̂ of q onto the subspace spanned by p is
qˆ =
¢ q, p²
28
56 14
+ t
p=
(4 + t ) =
¢ p, p²
50
25 25
8. The orthogonal projection q̂ of q onto the subspace spanned by p is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
392
CHAPTER 6
qˆ =
• Orthogonality and Least Squares
¢ q, p ²
10
3
1
p = − (3t − t 2 ) = − t + t 2
¢ p, p²
20
2
2
9. The inner product is ¢p, q² = p(–3)q(–3) + p(–1)q(–1) + p(1)q(1) + p(3)q(3).
a. The orthogonal projection p̂2 of p2 onto the subspace spanned by p0 and p1 is
pˆ 2 =
¢ p 2 , p0 ²
¢p ,p ²
20
0
p0 + 2 1 p1 =
t =5
(1) +
¢ p0 , p0 ²
¢ p1 , p1 ²
4
20
b. The vector q = p2 − pˆ 2 = t 2 − 5 will be orthogonal to both p0 and p1 and { p0 , p1 , q} will be an
orthogonal basis for Span{ p0 , p1 , p2 }. The vector of values for q at (–3, –1, 1, 3) is (4, –4, –4, 4),
so scaling by 1/4 yields the new vector q = (1/ 4)(t 2 − 5).
10. The best approximation to p = t 3 by vectors in W = Span{ p0 , p1 , q} will be
pˆ = projW p =
¢ p, p0 ²
¢ p, p1 ²
¢ p, q²
0
164
0 § t 2 − 5 · 41
p0 +
p1 +
q = (1) +
(t ) + ¨
¸= t
4
20
4© 4 ¹ 5
¢ p0 , p0 ²
¢ p1 , p1 ²
¢ q, q²
11. The orthogonal projection of p = t 3 onto W = Span{ p0 , p1 , p2 } will be
pˆ = projW p =
¢ p , p0 ²
¢ p , p1 ²
¢ p , p2 ²
0
34
0
17
p0 +
p1 +
p2 = (1) +
(t ) + (t 2 − 2) = t
¢ p0 , p0 ²
¢ p1 , p1 ²
¢ p2 , p2 ²
5
10
14
5
12. Let W = Span{ p0 , p1 , p2 }. The vector p3 = p − projW p = t 3 − (17 / 5)t will make { p0 , p1 , p2 , p3 }
an orthogonal basis for the subspace 3 of 4. The vector of values for p3 at (–2, –1, 0, 1, 2) is
(–6/5, 12/5, 0, –12/5, 6/5), so scaling by 5/6 yields the new vector p3 = (5/ 6)(t 3 − (17 / 5)t ) =
(5/ 6)t 3 − (17 / 6)t.
13. Suppose that A is invertible and that ¢u, v² = (Au) ⋅ (Av) for u and v in n. Check each axiom in the
definition on page 376, using the properties of the dot product.
i. ¢u, v² = (Au) ⋅ (Av) = (Av) ⋅ (Au) = ¢v, u²
ii. ¢u + v, w² = (A(u + v)) ⋅ (Aw) = (Au + Av) ⋅ (Aw) = (Au) ⋅ (Aw) + (Av) ⋅ (Aw) = ¢u, w² + ¢v, w²
iii. ¢c u, v² = (A( cu)) ⋅ (Av) = (c(Au)) ⋅ (Av) = c((Au) ⋅ (Av)) = c¢u, v²
iv. ¢u, u² = ( Au) ⋅ ( Au) = || Au ||2 ≥ 0, and this quantity is zero if and only if the vector Au is 0. But
Au = 0 if and only u = 0 because A is invertible.
14. Suppose that T is a one-to-one linear transformation from a vector space V into n and that ¢u, v² =
T(u) ⋅ T(v) for u and v in n. Check each axiom in the definition on page 376, using the properties of
the dot product and T. The linearity of T is used often in the following.
i. ¢u, v² = T(u) ⋅ T(v) = T(v) ⋅ T(u) = ¢v, u²
ii. ¢u + v, w² = T(u + v) ⋅ T(w) = (T(u) + T(v)) ⋅ T(w) = T(u) ⋅ T(w) + T(v) ⋅ T(w) = ¢u, w² + ¢v, w²
iii. ¢cu, v² = T(cu) ⋅ T(v) = (cT(u)) ⋅ T(v) = c(T(u) ⋅ T(v)) = c¢u, v²
iv. ¢u, u² = T (u) ⋅ T (u) = || T (u) ||2 ≥ 0, and this quantity is zero if and only if u = 0 since T is a oneto-one transformation.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.7
• Solutions
393
15. Using Axioms 1 and 3, ¢u, c v² = ¢c v, u² = c¢v, u² = c¢u, v².
16. Using Axioms 1, 2 and 3,
|| u − v ||2 = ¢u − v, u − v² = ¢u, u − v² − ¢ v, u − v²
= ¢u, u² − ¢u, v² − ¢ v, u² + ¢ v, v² = ¢u, u² − 2¢u, v² + ¢ v, v²
= || u ||2 −2¢u, v² + || v ||2
Since {u, v} is orthonormal, || u ||2 = || v ||2 = 1 and ¢u, v² = 0. So || u − v ||2 = 2.
17. Following the method in Exercise 16,
|| u + v ||2 = ¢u + v, u + v² = ¢u, u + v² + ¢ v, u + v²
= ¢u, u² + ¢u, v² + ¢ v, u² + ¢ v, v² = ¢u, u² + 2¢u, v² + ¢ v, v²
= || u ||2 + 2¢u, v² + || v ||2
Subtracting these results, one finds that || u + v ||2 − || u − v ||2 = 4¢u, v², and dividing by 4 gives the
desired identity.
18. In Exercises 16 and 17, it has been shown that || u − v ||2 = || u ||2 −2¢u, v² + || v ||2 and || u + v ||2 =
|| u ||2 + 2¢u, v² + || v ||2 . Adding these two results gives || u + v ||2 + || u − v ||2 = 2 || u ||2 + 2 || v ||2 .
ª aº
ª bº
19. let u = « » and v = « » . Then || u ||2 = a + b, || v ||2 = a + b, and ¢u, v² = 2 ab. Since a and b are
«¬ b »¼
«¬ a »¼
nonnegative, || u ||= a + b , || v ||= a + b . Plugging these values into the Cauchy-Schwarz
inequality gives
2 ab = | ¢u, v² | ≤ || u || || v || = a + b a + b = a + b
Dividing both sides of this equation by 2 gives the desired inequality.
20. The Cauchy-Schwarz inequality may be altered by dividing both sides of the inequality by 2 and then
squaring both sides of the inequality. The result is
2
|| u ||2 || v ||2
§ ¢u , v ² ·
¨
¸ ≤
4
© 2 ¹
ªa º
ª1º
Now let u = « » and v = « » . Then || u ||2 = a 2 + b2 , || v ||2 = 2 , and ¢u, v² = a + b. Plugging these
¬b ¼
¬1¼
values into the inequality above yields the desired inequality.
1
21. The inner product is ¢ f , g ² = ³ f (t ) g (t )dt. Let f (t ) = 1 − 3t 2 , g (t ) = t − t 3 . Then
0
1
1
0
0
¢ f , g ² = ³ (1 − 3t 2 )(t − t 3 ) dt = ³ 3t 5 − 4t 3 + t dt = 0
1
22. The inner product is ¢ f , g ² = ³ f (t ) g (t ) dt. Let f (t) = 5t – 3, g (t ) = t 3 − t 2 . Then
0
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
394
CHAPTER 6
• Orthogonality and Least Squares
1
1
0
0
¢ f , g ² = ³ (5t − 3)(t 3 − t 2 ) dt = ³ 5t 4 − 8t 3 + 3t 2 dt = 0
1
1
1
0
0
0
1
1
1
0
0
0
23. The inner product is ¢ f , g ² = ³ f (t ) g (t ) dt , so ¢ f , f ² = ³ (1 − 3t 2 ) 2 dt = ³ 9t 4 − 6t 2 + 1 dt = 4 / 5, and
|| f ||= ¢ f , f ² = 2 / 5.
24. The inner product is ¢ f , g ² = ³ f (t ) g (t ) dt , so ¢ g , g ² = ³ (t 3 − t 2 ) 2 dt = ³ t 6 − 2t 5 + t 4 dt = 1/105, and
|| g ||= ¢ g , g ² = 1/ 105.
1
1
−1
−1
25. The inner product is ¢ f , g ² = ³ f (t ) g (t ) dt. Then 1 and t are orthogonal because ¢1, t ² = ³ t dt = 0.
So 1 and t can be in an orthogonal basis for Span{1, t , t 2 }. By the Gram-Schmidt process, the third
basis element in the orthogonal basis can be
t2 −
¢t 2 ,1²
¢t 2 , t ²
1−
t
¢1,1²
¢t , t ²
1
1
1
−1
−1
−1
Since ¢t 2 ,1² = ³ t 2 dt = 2 / 3, ¢1,1² = ³ 1 dt = 2, and ¢t 2 , t ² = ³ t 3 dt = 0, the third basis element can
be written as t 2 − (1/ 3). This element can be scaled by 3, which gives the orthogonal basis as
{1, t , 3t 2 − 1}.
2
2
−2
−2
26. The inner product is ¢ f , g ² = ³ f (t ) g (t )dt. Then 1 and t are orthogonal because ¢1, t ² = ³ t dt = 0.
So 1 and t can be in an orthogonal basis for Span{1, t , t 2 }. By the Gram-Schmidt process, the third
basis element in the orthogonal basis can be
t2 −
¢t 2 ,1²
¢t 2 , t ²
1−
t
¢1,1²
¢t , t ²
2
2
2
−2
−2
−2
Since ¢t 2 , 1² = ³ t 2 dt = 16 / 3, ¢1, 1² = ³ 1 dt = 4, and ¢t 2 , t ² = ³ t 3 dt = 0, the third basis element can
be written as t 2 − (4/ 3). This element can be scaled by 3, which gives the orthogonal basis as
{1, t , 3t 2 − 4}.
27. [M] The new orthogonal polynomials are multiples of −17t + 5t 3 and 72 − 155t 2 + 35t 4 . These
polynomials may be scaled so that their values at –2, –1, 0, 1, and 2 are small integers.
28. [M] The orthogonal basis is f 0 (t ) = 1, f1 (t ) = cos t , f 2 (t ) = cos 2t − (1/ 2) = (1/ 2)cos 2 t , and
f3 (t ) = cos3t − (3/ 4)cos t = (1/ 4)cos 3t.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.8
6.8
• Solutions
395
SOLUTIONS
Notes: The connections between this section and Section 6.7 are described in the notes for that section.
For my junior-senior class, I spend three days on the following topics: Theorems 13 and 15 in Section 6.5,
plus Examples 1, 3, and 5; Example 1 in Section 6.6; Examples 2 and 3 in Section 6.7, with the
motivation for the definite integral; and Fourier series in Section 6.8.
1. The weighting matrix W, design matrix X, parameter vector β, and observation vector y are:
ª1
«0
«
W = «0
«
«0
«¬0
0
2
0
0
0
0
0
0
2
0
0
2
0
0
0
0º
ª1
»
«1
0»
«
0» , X = «1
»
«
0»
«1
«¬1
1»¼
−2º
ª0º
»
«0»
−1»
« »
ª β0 º
0» , β = « » , y = « 2»
»
« »
¬ β1 ¼
1»
« 4»
«¬ 4»¼
2»¼
The design matrix X and the observation vector y are scaled by W:
ª1
«2
«
WX = « 2
«
«2
«¬ 1
−2º
ª0º
»
«0»
−2»
« »
0» , Wy = « 4»
»
« »
2»
«8 »
»
«¬ 4»¼
2¼
Further compute
ª14
(WX )T WX = «
¬ 0
and find that
0º
ª 28º
, (WX )T Wy = « »
»
16¼
¬ 24¼
ª1/14
¬ 0
0º ª 28º ª 2º
=
1/16»¼ «¬ 24»¼ «¬3/ 2»¼
Thus the weighted least-squares line is y = 2 + (3/2)x.
βˆ = ((WX )T WX ) −1 (WX )T Wy = «
2. Let X be the original design matrix, and let y be the original observation vector. Let W be the
weighting matrix for the first method. Then 2W is the weighting matrix for the second method. The
weighted least-squares by the first method is equivalent to the ordinary least-squares for an equation
whose normal equation is
(WX )T WX βˆ = (WX )T Wy
while the second method is equivalent to the ordinary least-squares for an equation whose normal
equation is
(2WX )T (2W ) X βˆ = (2WX )T (2W )y
(2)
Since equation (2) can be written as 4(WX )T WX βˆ = 4(WX )T Wy, it has the same solutions as
equation (1).
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
(1)
396
CHAPTER 6
• Orthogonality and Least Squares
3. From Example 2 and the statement of the problem, p0 (t ) = 1, p1 (t ) = t , p2 (t ) = t 2 − 2,
p3 (t ) = (5/ 6)t 3 − (17 / 6)t , and g = (3, 5, 5, 4, 3). The cubic trend function for g is the orthogonal
projection p̂ of g onto the subspace spanned by p0 , p1 , p2 , and p3 :
pˆ =
=
¢ g , p0 ²
¢ g , p3 ²
¢ g , p1 ²
¢ g , p2 ²
p0 +
p1 +
p2 +
p3
¢ p0 , p0 ²
¢ p1 , p1 ²
¢ p2 , p 2 ²
¢ p3 , p3 ²
20
−1
−7 2
2 §5
17 ·
(1) + t +
t − 2 + ¨ t3 − t ¸
5
10
14
10 © 6
6 ¹
(
=4−
(
)
)
1
1
1§ 5
17 ·
2
1
1
t − t2 − 2 + ¨ t3 − t ¸ = 5 − t − t2 + t3
10
2
5© 6
6 ¹
3
2
6
This polynomial happens to fit the data exactly.
4. The inner product is ¢ p, q² = p(–5)q(–5) + p(–3)q(–3) + p(–1)q(–1) + p(1)q(1) + p(3)q(3) + p(5)q(5).
a. Begin with the basis {1, t , t 2 } for 2. Since 1 and t are orthogonal, let p0 (t ) = 1 and p1 (t ) = t .
Then the Gram-Schmidt process gives
p2 (t ) = t 2 −
¢t 2 ,1²
¢t 2 , t ²
70 2 35
1−
t = t2 −
=t −
¢1,1²
¢t , t ²
6
3
The vector of values for p2 is (40/3, –8/3, –32/3, –32/3, –8/3, 40/3), so scaling by 3/8 yields the
new function p2 = (3/8)(t 2 − (35/ 3)) = (3/8)t 2 − (35/8).
b. The data vector is g = (1, 1, 4, 4, 6, 8). The quadratic trend function for g is the orthogonal
projection p̂ of g onto the subspace spanned by p0 , p1 and p2 :
pˆ =
¢ g , p0 ²
24
50
6 §3
35 ·
¢ g , p1 ²
¢ g , p2 ²
p0 +
p1 +
p2 = (1) + t + ¨ t 2 − ¸
¢ p0 , p0 ²
¢ p1 , p1 ²
¢ p2 , p2 ²
6
70 84 © 8
8 ¹
=4+
5
1 §3
35 · 59 5
3 2
t + ¨ t2 − ¸ =
+ t+
t
7 14 © 8
8 ¹ 16 7 112
5. The inner product is ¢ f , g ² = ³
2π
0
f (t ) g (t )dt. Let m ≠ n. Then
2π
¢sin mt , sin nt ² = ³ sin mt sin nt dt =
0
1 2π
cos((m − n)t ) − cos((m + n)t )dt = 0
2 ³0
Thus sin mt and sin nt are orthogonal.
6. The inner product is ¢ f , g ² = ³
2π
0
f (t ) g (t ) dt. Let m and n be positive integers. Then
2π
¢sin mt ,cos nt ² = ³ sin mt cos nt dt =
0
1 2π
sin((m + n)t ) + sin((m − n)t )dt = 0
2 ³0
Thus sinmt and cosnt are orthogonal.
7. The inner product is ¢ f , g ² = ³
2π
0
f (t ) g (t ) dt. Let k be a positive integer. Then
2π
|| cos kt ||2 = ¢cos kt ,cos kt ² = ³ cos2 kt dt =
0
1 2π
1 + cos 2kt dt = π
2 ³0
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.8
• Solutions
and
2π
|| sin kt ||2 = ¢sin kt ,sin kt ² = ³ sin 2 kt dt =
0
1 2π
1 − cos 2kt dt = π
2 ³0
8. Let f(t) = t – 1. The Fourier coefficients for f are:
a0 1 1 2π
1 2π
=
f (t ) dt =
t − 1 dt = −1 + π
³
2 2π 0
2π ³0
and for k > 0,
1 2π
1 2π
ak = ³ f (t )cos kt dt = ³ (t − 1)cos kt dt = 0
bk =
π
0
1
2π
π ³0
f (t )sin kt dt =
π
0
1
2π
π ³0
(t − 1)sin kt dt = −
2
k
The third-order Fourier approximation to f is thus
a0
2
+ b1sin t + b2sin 2t + b3sin 3t = −1 + π − 2 sin t − sin 2t − sin 3t
2
3
9. Let f(t) = 2π – t. The Fourier coefficients for f are:
a0 1 1 2π
1 2π
=
f (t ) dt =
2π − t dt = π
³
0
2 2π
2π ³0
and for k > 0,
1 2π
1 2π
ak = ³ f (t ) cos kt dt = ³ (2π − t ) cos kt dt = 0
bk =
π
0
1
2π
π ³0
f (t ) sin kt dt =
π
0
1
2π
π ³0
(2π − t ) sin kt dt =
2
k
The third-order Fourier approximation to f is thus
a0
2
+ b1sin t + b2sin 2t + b3sin 3t = π + 2 sin t + sin 2t + sin 3t
2
3
1
10. Let f (t ) = ®
¯−1
for 0 ≤ t < π
for π ≤ t < 2π
. The Fourier coefficients for f are:
a0 1 1 2π
1 π
1 2π
=
=
−
(
)
f
t
dt
dt
dt = 0
2 2 π ³0
2π ³0
2π ³π
and for k > 0,
1 2π
1 π
1 2π
ak = ³ f (t ) cos kt dt = ³ cos kt dt − ³ cos kt dt = 0
bk =
π
0
1
2π
π
³0
f (t ) sin kt dt =
π
0
1
π
π
³0
π
sin kt dt −
1
π
π
2π
³π
4 /( kπ )
sin kt dt = ®
0
¯
for k odd
for k even
The third-order Fourier approximation to f is thus
b1sin t + b3sin 3t =
4
π
sin t +
4
sin 3t
3π
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
397
398
CHAPTER 6
• Orthogonality and Least Squares
11. The trigonometric identity cos 2t = 1 − 2 sin 2t shows that
1 1
sin 2t = − cos 2t
2 2
The expression on the right is in the subspace spanned by the trigonometric polynomials of order 3 or
less, so this expression is the third-order Fourier approximation to sin 2 t .
12. The trigonometric identity cos 3t = 4 cos3t − 3 cos t shows that
3
1
cos3t = cos t + cos 3t
4
4
The expression on the right is in the subspace spanned by the trigonometric polynomials of order 3 or
less, so this expression is the third-order Fourier approximation to cos3t .
13. Let f and g be in C [0, 2π] and let m be a nonnegative integer. Then the linearity of the inner product
shows that
¢( f + g), cos mt² = ¢ f, cos mt² + ¢g, cos mt², ¢( f + g), sin mt² = ¢ f, sin mt² + ¢ g, sin mt²
Dividing these identities respectively by ¢cos mt, cos mt² and ¢sin mt, sin mt² shows that the Fourier
coefficients am and bm for f + g are the sums of the corresponding Fourier coefficients of f and of g.
14. Note that g and h are both in the subspace H spanned by the trigonometric polynomials of order 2 or
less. Since h is the second-order Fourier approximation to f, it is closer to f than any other function in
the subspace H.
15. [M] The weighting matrix W is the 13 × 13 diagonal matrix with diagonal entries 1, 1, 1, .9, .9, .8, .7,
.6, .5, .4, .3, .2, .1. The design matrix X, parameter vector β, and observation vector y are:
0
0º
ª1 0
«1
ª 0.0 º
1
1
1»»
«
« 8.8»
«1 2
«
»
22
23 »
«
»
«
»
29.9
«1 3
32
33 »
«
»
«
« 62.0 »
2
3»
4
4 »
«1 4
« 104.7 »
«
β
«
»
2
3»
ª
º
0
5
5 »
«1 5
« 159.1»
«β »
62
63 » , β = « 1 » , y = « 222.0 »
X = «1 6
«
»
«
»
«β2 »
«1 7
« 294.5»
72
73 »
« »
«
»
¬« β 3 ¼»
« 380.4 »
82
83 »
«1 8
«
»
«
« 471.1»
2
3»
9
9 »
«1 9
«
»
2
«1 10 10 103 »
« 571.7 »
«
»
« 686.8»
«1 11 112 113 »
«
»
809.2 »¼
«
«
»
¬
2
3
¬«1 12 12 12 ¼»
The design matrix X and the observation vector y are scaled by W:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
6.8
ª1.0
«1.0
«
«1.0
«
« .9
« .9
«
« .8
WX = « .7
«
« .6
« .5
«
« .4
«
« .3
« .2
«
«¬ .1
0.0
1.0
2.0
2.7
3.6
4.0
4.2
4.2
4.0
3.6
3.0
2.2
1.2
0.0
1.0
4.0
8.1
14.4
20.0
25.2
29.4
32.0
32.4
30.0
24.2
14.4
• Solutions
0.0 º
ª 0.00 º
« 8.80 »
»
1.0 »
«
»
« 29.90 »
8.0 »
»
«
»
24.3»
« 55.80 »
« 94.23»
57.6 »
»
«
»
100.0 »
«127.28»
151.2 » , Wy = «155.40 »
»
«
»
205.8»
«176.70 »
«190.20 »
256.0 »»
«
»
«188.44 »
291.6 »
»
«
»
300.0 »
«171.51»
«137.36 »
266.2 »
»
«
»
172.8¼»
¬« 80.92 »¼
Further compute
ª 6.66
« 22.23
(WX )T WX = «
«120.77
«
«¬797.19
22.23
120.77
120.77
797.19
797.19
5956.13
5956.13
48490.23
797.19 º
747.844º
ª
»
«
5956.13»
4815.438»»
, (WX )T Wy = «
« 35420.468»
48490.23»
»
«
»
420477.17 »¼
«¬ 285262.440»¼
and find that
ª −0.2685º
« 3.6095»
T
−1
T
ˆ
»
β = ((WX ) WX ) (WX ) Wy = «
« 5.8576 »
«
»
«¬ −0.0477 »¼
Thus the weighted least-squares cubic is y = g (t ) = −.2685 + 3.6095t + 5.8576t 2 − .0477t 3 . The
velocity at t = 4.5 seconds is g'(4.5) = 53.4 ft./sec. This is about 0.7% faster than the estimate
obtained in Exercise 13 of Section 6.6.
1
16. [M] Let f (t ) = ®
¯−1
for 0 ≤ t < π
for π ≤ t < 2π
. The Fourier coefficients for f have already been found to be
4 /(kπ )
ak = 0 for all k ≥ 0 and bk = ®
¯ 0
for k even
. Thus
4
4
4
4
sin 3t and f5 (t ) = sin t + sin 3t + sin 5t
π
π
3π
3π
5π
A graph of f 4 over the interval [0, 2π] is
f 4 (t ) =
4
for k odd
sin t +
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
399
400
CHAPTER 6
• Orthogonality and Least Squares
1
0.5
1
2
3
4
5
6
2
3
4
5
6
–0.5
–1
A graph of f 5 over the interval [0, 2π] is
1
0.5
1
–0.5
–1
A graph of f 5 over the interval [–2π, 2π] is
1
0.5
–6
–4
–2
2
4
6
–0.5
–1
Chapter 6 SUPPLEMENTARY EXERCISES
1.
a. False. The length of the zero vector is zero.
b. True. By the displayed equation before Example 2 in Section 6.1, with c = –1,
|| –x || = || (–1)x || =| –1 ||| x || = || x ||.
c. True. This is the definition of distance.
d. False. This equation would be true if r|| v || were replaced by | r ||| v ||.
e. False. Orthogonal nonzero vectors are linearly independent.
f. True. If x ⋅ u = 0 and x ⋅ v = 0, then x ⋅ (u – v) = x ⋅ u – x ⋅ v = 0.
g. True. This is the “only if” part of the Pythagorean Theorem in Section 6.1.
h. True. This is the “only if” part of the Pythagorean Theorem in Section 6.1 where v is replaced
by –v, because || −v ||2 is the same as || v ||2 .
i. False. The orthogonal projection of y onto u is a scalar multiple of u, not y (except when y
itself is already a multiple of u).
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 6
• Supplementary Exercises
401
j. True. The orthogonal projection of any vector y onto W is always a vector in W.
k. True. This is a special case of the statement in the box following Example 6 in Section 6.1 (and
proved in Exercise 30 of Section 6.1).
l. False. The zero vector is in both W and W ⊥ .
m. True. See Exercise 32 in Section 6.2. If vi ⋅ v j = 0, then
(ci vi ) ⋅ (c j v j ) = ci c j ( vi ⋅ v j ) = ci c j 0 = 0.
n. False. This statement is true only for a square matrix. See Theorem 10 in Section 6.3.
o. False. An orthogonal matrix is square and has orthonormal columns.
p. True. See Exercises 27 and 28 in Section 6.2. If U has orthonormal columns, then U T U = I . If
U is also square, then the Invertible Matrix Theorem shows that U is invertible and U −1 = U T .
In this case, UU T = I , which shows that the columns of U T are orthonormal; that is, the rows
of U are orthonormal.
q. True. By the Orthogonal Decomposition Theorem, the vectors projW v and v − projW v are
orthogonal, so the stated equality follows from the Pythagorean Theorem.
r. False. A least-squares solution is a vector x̂ (not A x̂ ) such that A x̂ is the closest point to b
in Col A.
s. False. The equation xˆ = ( AΤ A)−1 AΤb describes the solution of the normal equations, not the
matrix form of the normal equations. Furthermore, this equation makes sense only when AT A
is invertible.
2. If {v1 , v 2 } is an orthonormal set and x = c1v1 + c2 v 2 , then the vectors c1 v1 and c2 v 2 are orthogonal
(Exercise 32 in Section 6.2). By the Pythagorean Theorem and properties of the norm
|| x ||2 = || c1v1 + c2 v 2 ||2 = || c1v1 ||2 + || c2 v 2 ||2 = (c1 || v1 ||)2 + (c2 || v 2 ||)2 =| c1 |2 + | c2 |2
So the stated equality holds for p = 2. Now suppose the equality holds for p = k, with k ≥ 2. Let
{v1 ,…, v k +1} be an orthonormal set, and consider x = c1 v1 + … + ck v k + ck +1 v k +1 = u k + ck +1 v k +1 ,
where u k = c1 v1 + … + ck v k . Observe that u k and ck +1 v k +1 are orthogonal because v j ⋅ v k +1 = 0 for j
= 1,…,k. By the Pythagorean Theorem and the assumption that the stated equality holds for k, and
because || ck +1v k +1 ||2 = | ck +1 |2 || v k +1 ||2 = | ck +1 |2 ,
|| x ||2 = || u k + ck +1v k +1 ||2 = || uk ||2 + || ck +1v k +1 ||2 = | c1 |2 +…+ | ck +1 |2
Thus the truth of the equality for p = k implies its truth for p = k + 1. By the principle of induction,
the equality is true for all integers p ≥ 2.
3. Given x and an orthonormal set {v1 ,…, v p } in
n
, let x̂ be the orthogonal projection of x onto the
subspace spanned by v1 ,…, v p . By Theorem 10 in Section 6.3, xˆ = (x ⋅ v1 ) v1 +…+ (x ⋅ v p ) v p . By
Exercise 2, || xˆ ||2 = | x ⋅ v1 |2 +…+ | x ⋅ v p |2 . Bessel’s inequality follows from the fact that
|| xˆ ||2 ≤ || x ||2 , which is noted before the proof of the Cauchy-Schwarz inequality in Section 6.7.
4. By parts (a) and (c) of Theorem 7 in Section 6.2, {Uv1 ,…,Uv k } is an orthonormal set in
there are n vectors in this linearly independent set, the set is a basis for n.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
n
. Since
402
CHAPTER 6
• Orthogonality and Least Squares
5. Suppose that (U x)⋅(U y) = x⋅y for all x, y in
n
, and let e1 ,…, e n be the standard basis for
n
. For
j = 1, …, n, Ue j is the jth column of U. Since || Ue j ||2 = (Ue j ) ⋅ (Ue j ) = e j ⋅ e j = 1, the columns of U
are unit vectors; since (Ue j ) ⋅ (Uek ) = e j ⋅ ek = 0 for j ≠ k, the columns are pairwise orthogonal.
6. If Ux = λx for some x ≠ 0, then by Theorem 7(a) in Section 6.2 and by a property of the norm,
|| x || = || Ux || = || λx || = | λ ||| x ||, which shows that | λ | = 1, because x ≠ 0.
7. Let u be a unit vector, and let Q = I − 2uuT . Since (uuT )T = uTT uT = uuT ,
QT = ( I − 2uuT )T = I − 2(uuT )T = I − 2uuT = Q
Then
QQT = Q2 = ( I − 2uuT )2 = I − 2uuT − 2uuT + 4(uuT )(uuT )
Since u is a unit vector, uT u = u ⋅ u = 1, so (uuT )(uuT ) = u(uT )(u)uT = uuT , and
QQT = I − 2uuT − 2uuT + 4uuT = I
Thus Q is an orthogonal matrix.
8. a. Suppose that x ⋅ y = 0. By the Pythagorean Theorem, || x ||2 + || y ||2 = || x + y ||2 . Since T preserves
lengths and is linear,
|| T (x) ||2 + || T (y) ||2 = || T (x + y) ||2 = || T (x) + T (y) ||2
This equation shows that T(x) and T(y) are orthogonal, because of the Pythagorean Theorem.
Thus T preserves orthogonality.
b. The standard matrix of T is [T (e1 )
… T (en )] , where e1 ,…, e n are the columns of the identity
matrix. Then {T (e1 ),…, T (e n )} is an orthonormal set because T preserves both orthogonality and
lengths (and because the columns of the identity matrix form an orthonormal set). Finally, a
square matrix with orthonormal columns is an orthogonal matrix, as was observed in Section 6.2.
9. Let W = Span{u, v}. Given z in
, let zˆ = projW z . Then ẑ is in Col A, where A = [u
n
v ]. Thus
2
there is a vector, say, x̂ in , with A x̂ = ẑ . So, x̂ is a least-squares solution of Ax = z. The normal
equations may be solved to find x̂ , and then ẑ may be found by computing A xˆ.
10. Use Theorem 14 in Section 6.5. If c ≠ 0, the least-squares solution of Ax = c b is given by
( AT A)−1 AT (cb), which equals c( AT A)−1 AT b, by linearity of matrix multiplication. This solution is
c times the least-squares solution of Ax = b.
ª vT º ª1 −2 5 º
ª 1º
ªxº
ªaº
« »
11. Let x = «« y »» , b = «« b »» , v = «« −2 »» , and A = « vT » = «1 −2 5 » . Then the given set of equations is
«
»
« T » «1 −2 5 »
«¬ 5»¼
«¬ z »¼
«¬ c »¼
v
¼
«¬ »¼ ¬
Ax = b, and the set of all least-squares solutions coincides with the set of solutions of the normal
equations AT Ax = AT b . The column-row expansions of AT A and AT b give
AT A = vvT + vvT + vvT = 3vvT , AT b = av + bv + cv = (a + b + c) v
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 6
• Supplementary Exercises
403
Thus AT Ax = 3( vvT )x = 3v( vT x) = 3( vT x) v since vT x is a scalar, and the normal equations have
become 3( vT x) v = (a + b + c) v, so 3( vT x) = a + b + c, or vT x = (a + b + c) / 3. Computing vT x gives
the equation x – 2y + 5z = (a + b + c)/3 which must be satisfied by all least-squares solutions to Ax =
b.
12. The equation (1) in the exercise has been written as Vλ = b, where V is a single nonzero column
vector v, and b = Av. The least-squares solution λ̂ of Vλ = b is the exact solution of the normal
equations V T V λ = V T b. In the original notation, this equation is vT vλ = vT Av. Since vT v is
nonzero, the least squares solution λ̂ is vT Av /( vT v). This expression is the Rayleigh quotient
discussed in the Exercises for Section 5.8.
13. a. The row-column calculation of Au shows that each row of A is orthogonal to every u in Nul A. So
each row of A is in (Nul A) ⊥ . Since (Nul A)⊥ is a subspace, it must contain all linear
combinations of the rows of A; hence (Nul A)⊥ contains Row A.
b. If rank A = r, then dim Nul A = n – r by the Rank Theorem. By Exercsie 24(c) in Section 6.3,
dimNul A + dim(Nul A)⊥ = n, so dim(Nul A)⊥ must be r. But Row A is an r-dimensional
subspace of (Nul A)⊥ by the Rank Theorem and part (a). Therefore, Row A = (Nul A)⊥ .
c. Replace A by AT in part (b) and conclude that Row AT = (Nul AT )⊥ . Since Row AT = Col A,
Col A = (Nul AT )⊥ .
14. The equation Ax = b has a solution if and only if b is in Col A. By Exercise 13(c), Ax = b has a
solution if and only if b is orthogonal to Nul AT . This happens if and only if b is orthogonal to all
solutions of AT x = 0.
15. If A = URU T with U orthogonal, then A is similar to R (because U is invertible and U T = U −1 ), so A
has the same eigenvalues as R by Theorem 4 in Section 5.2. Since the eigenvalues of R are its n real
diagonal entries, A has n real eigenvalues.
16. a. If U = [u1
u2
… u n ] , then AU = [ λ1u1
…
Au 2
Aun ]. Since u1 is a unit vector and
u 2 ,…, u n are orthogonal to u1 , the first column of U AU is U T (λ1u1 ) = λ1U T u1 = λ1e1.
T
b. From (a),
ªλ1
«0
T
U AU = «
«#
«
«¬ 0
*
*
A1
*
*º
»
»
»
»
»¼
View U T AU as a 2 × 2 block upper triangular matrix, with A1 as the (2, 2)-block. Then from
Supplementary Exercise 12 in Chapter 5,
det(U T AU − λ I n ) = det((λ1 − λ) I1 ) ⋅ det( A1 − λ I n−1 ) = (λ1 − λ) ⋅ det( A1 − λ I n−1 )
This shows that the eigenvalues of U T AU , namely, λ1 ,…, λ n , consist of λ1 and the eigenvalues
of A1 . So the eigenvalues of A1 are λ 2 ,… , λ n .
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley
404
CHAPTER 6
• Orthogonality and Least Squares
17. [M] Compute that || ∆x ||/|| x || = .4618 and cond( A) × (|| ∆b || / || b ||) = 3363 × (1.548 ×10−4 ) = .5206 . In
this case, || ∆x ||/|| x || is almost the same as cond(A) × || ∆b ||/|| b ||.
18. [M] Compute that || ∆x ||/|| x || = .00212 and cond(A) × (|| ∆b ||/|| b ||) = 3363 × (.00212) ≈ 7.130. In
this case, || ∆x ||/|| x || is almost the same as || ∆b ||/|| b ||, even though the large condition number
suggests that || ∆x ||/|| x || could be much larger.
19. [M] Compute that || ∆x || / || x ||= 7.178 × 10−8 and cond( A) × (|| ∆b || / || b ||) = 23683 × (2.832 × 10−4 ) =
6.707. Observe that the relative change in x is much smaller than the relative change in b. In fact the
theoretical bound on the relative change in x is 6.707 (to four significant figures). This exercise
shows that even when a condition number is large, the relative error in the solution need not be as
large as you suspect.
20. [M] Compute that || ∆x ||/|| x || = .2597 and cond( A) × (|| ∆b || / || b ||) = 23683 × (1.097 × 10−5 ) = .2598 .
This calculation shows that the relative change in x, for this particular b and ∆b, should not exceed
.2598. In this case, the theoretical maximum change is almost acheived.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.1
SOLUTIONS
Notes: Students can profit by reviewinng Section 5.3 (focusing on the Diagonalization Theoorem) before
ortant for the
working on this section. Theorems 1 andd 2 and the calculations in Examples 2 and 3 are impo
sections that follow. Note that symmetricc matrix means real symmetric matrix, because all maatrices in the
text have real entries, as mentioned at the beginning of this chapter. The exercises in this section
s
have
Gram-Schmidt process is not needed.
been constructed so that mastery of the G
Theorem 2 is easily proved for the 2 × 2 case:
(
)
1
ªa b º
If A = «
, then λ = a + d ± (a − d ) 2 + 4b2 .
»
2
¬c d ¼
If b = 0 there is nothing to prove. Otherw
wise, there are two distinct eigenvalues, so A must be
ªd − λº
diagonalizable. In each case, an eigenvecctor for λ is «
».
¬ −b ¼
ª3
1. Since A = «
¬5
ª −3
2. Since A = «
¬ −5
5º
= AT , the matriix is symmetric.
−7 »¼
5º
≠ AT , the matriix is not symmetric.
3»¼
ª2
3. Since A = «
¬4
2º
≠ AT , the matrixx is not symmetric.
4 »¼
ª0
4. Since A = « 8
«
«¬ 3
8
0
−2
ª −6
5. Since A = « 0
«
¬« 0
2
−6
0
3º
−2 »» = AT , the matrix is symmetric.
0 »¼
0º
2 »» ≠ AT , thhe matrix is not symmetric.
−6 ¼»
Copyright © 2012 Pearrson Education, Inc. Publishing as Addison-Wesley.
405
406
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
6. Since A is not a square matrix A ≠ AT and the matrix is not symmetric.
ª.6
7. Let P = «
¬.8
.8 º
, and compute that
−.6 »¼
ª.6
PT P = «
¬.8
.8 º ª.6
−.6 »¼ «¬.8
.8º ª 1
=
−.6 »¼ «¬ 0
0º
= I2
1»¼
ª.6
Since P is a square matrix, P is orthogonal and P −1 = PT = «
¬.8
ª1/ 2
8. Let P = «
«¬1/ 2
.8 º
.
−.6 »¼
−1/ 2 º
» , and compute that
1/ 2 »¼
ª 1/ 2
PT P = «
«¬ −1/ 2
1/ 2 º ª1/ 2
»«
1/ 2 »¼ «¬1/ 2
−1/ 2 º ª 1
»=«
1/ 2 »¼ ¬ 0
0º
= I2
1»¼
ª 1/ 2
Since P is a square matrix, P is orthogonal and P −1 = PT = «
«¬ −1/ 2
ª −5
9. Let P = «
¬ 2
1/ 2 º
».
1/ 2 »¼
2º
, and compute that
5»¼
ª −5
PT P = «
¬ 2
2 º ª −5
5»¼ «¬ 2
2 º ª 29
=
5»¼ «¬ 0
0º
≠ I2
29 »¼
Thus P is not orthogonal.
ª −1
10. Let P = « 2
«
«¬ 2
2
−1
2
ª −1
P P = «« 2
«¬ 2
T
2º
2 »» , and compute that
−1»¼
2
−1
2
2 º ª −1
2 »» «« 2
−1»¼ «¬ 2
2
−1
2
2 º ª9
2 »» = ««0
−1»¼ «¬0
0
9
0
0º
0 »» ≠ I 3
9 »¼
Thus P is not orthogonal.
ª 2/3
«
0
11. Let P = «
« 5 /3
¬
2/3
1/ 5
−4 / 45
1/ 3º
»
−2 / 5 » , and compute that
−2 / 45 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.1
ª2 / 3
«
PT P = « 2 / 3
«
«¬ 1/ 3
0
1/ 5
−2 / 5
5 / 3º ª 2 / 3
»«
−4 / 45 » «
0
»«
−2 / 45 »¼ ¬ 5 / 3
2/3
1/ 5
−4 / 45
1/ 3º ª 1
»
−2 / 5 » = «« 0
−2 / 45 »¼ «¬0
ª2 / 3
«
Since P is a square matrix, P is orthogonal and P −1 = PT = « 2 / 3
«
¬« 1/ 3
.5
.5
−.5
−.5
.5
.5
.5
.5
ª .5
« .5
PT P = «
« −.5
«
¬« −.5
−.5
.5
.5
.5
−.5
.5
.5
.5
ª .5
« −.5
12. Let P = «
« .5
«
¬« −.5
0
1/ 5
−2 / 5
0
1
0
• Solutions
407
0º
0 »» = I 3
1»¼
5 / 3º
»
−4 / 45 » .
»
−2 / 45 ¼»
−.5º
.5»»
, and compute that
.5»
»
−.5¼»
−.5º ª .5
.5»» «« −.5
.5» « .5
»«
−.5¼» ¬« −.5
.5
.5
−.5
−.5
.5
.5
.5
.5
−.5º ª1
.5»» ««0
=
.5» « 0
» «
−.5¼» ¬« 0
ª .5
« .5
Since P is a square matrix, P is orthogonal and P −1 = PT = «
« −.5
«
¬« −.5
0
1
0
0
0
0
1
0
0º
0»»
= I4
0»
»
1 ¼»
−.5
.5
.5
.5
−.5
.5
.5
.5
−.5º
.5»»
.
.5»
»
−.5¼»
ª3
13. Let A = «
¬1
1º
. Then the characteristic polynomial of A is (3 − λ)2 − 1 = λ2 − 6λ + 8 = (λ − 4)(λ − 2),
»
3¼
ª1º
so the eigenvalues of A are 4 and 2. For λ = 4, one computes that a basis for the eigenspace is « » ,
¬1¼
ª1/ 2 º
which can be normalized to get u1 = «
» . For λ = 2, one computes that a basis for the eigenspace
¬«1/ 2 ¼»
ª −1/ 2 º
ª −1º
is « » , which can be normalized to get u 2 = «
» . Let
¬ 1¼
«¬ 1/ 2 »¼
P = [u1
ª1/ 2
u2 ] = «
«¬1/ 2
−1/ 2 º
ª4
» and D = «
1/ 2 »¼
¬0
0º
2 »¼
Then P orthogonally diagonalizes A, and A = PDP −1 .
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
408
CHAPTER 7
ª1
14. Let A = «
¬5
• Symmetric Matrices and Quadratic Forms
5º
. Then the characteristic polynomial of A is
1»¼
(1 − λ)2 − 25 = λ 2 − 2λ − 24 = (λ − 6)(λ + 4), so the eigenvalues of A are 6 and –4. For λ = 6, one
ª1/ 2 º
ª1º
computes that a basis for the eigenspace is « » , which can be normalized to get u1 = «
» . For
¬1¼
¬«1/ 2 ¼»
ª −1º
λ = –4, one computes that a basis for the eigenspace is « » , which can be normalized to get
¬ 1¼
ª −1/ 2 º
u2 = «
».
¬« 1/ 2 ¼»
Let
P = [u1
ª1/ 2
u2 ] = «
¬«1/ 2
−1/ 2 º
ª6
» and D = «
1/ 2 ¼»
¬0
0º
−4 »¼
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª 16
15. Let A = «
¬ −4
−4 º
. Then the characteristic polynomial of A is
1»¼
(16 − λ)(1 − λ) − 16 = λ2 − 17λ = (λ − 17)λ , so the eigenvalues of A are 17 and 0. For λ = 17, one
ª −4 / 17 º
ª −4 º
computes that a basis for the eigenspace is « » , which can be normalized to get u1 = «
».
¬ 1¼
«¬ 1/ 17 »¼
ª 1º
For λ = 0, one computes that a basis for the eigenspace is « » , which can be normalized to get
¬4¼
ª 1/ 17 º
u2 = «
» . Let
«¬ 4 / 17 »¼
P = [u1
ª −4 / 17
u2 ] = «
«¬ 1/ 17
1/ 17 º
ª17
» and D = «
4 / 17 »¼
¬ 0
0º
0 »¼
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª −7 24 º
2
16. Let A = «
» . Then the characteristic polynomial of A is (−7 − λ)(7 − λ) − 576 = λ − 625 =
24
7
¬
¼
(λ − 25)(λ + 25) , so the eigenvalues of A are 25 and –25. For λ = 25, one computes that a basis for
ª 3/ 5º
ª 3º
the eigenspace is « » , which can be normalized to get u1 = «
» . For λ = –25, one computes that a
¬ 4 / 5¼
¬4¼
ª −4 / 5 º
ª −4 º
basis for the eigenspace is « » , which can be normalized to get u 2 = «
» . Let
¬ 3 / 5¼
¬ 3¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.1
ª 3/ 5
u2 ] = «
¬4 / 5
P = [u1
−4 / 5 º
ª 25
and D = «
»
3/ 5 ¼
¬ 0
• Solutions
409
0º
−25»¼
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª1
17. Let A = « 1
«
¬«3
1
3
1
3º
1»» . The eigenvalues of A are 5, 2, and –2. For λ = 5, one computes that a basis for
1»¼
ª1/ 3 º
ª1º
«
»
the eigenspace is «1» , which can be normalized to get u1 = «1/ 3 » . For λ = 2, one computes that a
« »
«
»
¬«1¼»
«¬1/ 3 »¼
ª 1/ 6 º
ª 1º
«
»
basis for the eigenspace is « −2 » , which can be normalized to get u 2 = « −2 / 6 » . For λ = –2, one
« »
«
»
¬« 1¼»
«¬ 1/ 6 »¼
ª −1/ 2 º
ª −1º
«
»
«
»
0» . Let
computes that a basis for the eigenspace is 0 , which can be normalized to get u3 = «
« »
« 1/ 2 »
«¬ 1»¼
¬
¼
P = [u1
u2
ª1/ 3
«
u3 ] = «1/ 3
«
¬«1/ 3
1/ 6
−2 / 6
1/ 6
−1/ 2 º
ª5
»
0 » and D = ««0
»
«¬0
1 2 ¼»
0
2
0
0º
0 »»
−2 »¼
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª −2
18. Let A = « −36
«
«¬ 0
−36
−23
0
0º
0 »» . The eigenvalues of A are 25, 3, and –50. For λ = 25, one computes that a
3»¼
ª −4 / 5 º
ª−4º
«
»
basis for the eigenspace is 3 , which can be normalized to get u1 = « 3/ 5 » . For λ = 3, one
«
»
« »
«¬
«¬ 0 »¼
0 »¼
ª0 º
ª0 º
«
»
computes that a basis for the eigenspace is 0 , which is of length 1, so u 2 = « 0 » . For λ = –50, one
« »
« »
«¬ 1»¼
«¬ 1»¼
ª 3º
ª 3 / 5º
«
»
computes that a basis for the eigenspace is 4 , which can be normalized to get u 3 = « 4 / 5 » . Let
« »
«
»
«¬ 0 »¼
«¬ 0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
410
CHAPTER 7
P = [u1
• Symmetric Matrices and Quadratic Forms
u2
ª −4 / 5
u 3 ] = «« 3/ 5
«¬
0
0
0
1
3/ 5 º
ª 25
»
4 / 5 » and D = «« 0
«¬ 0
0 »¼
0
3
0
0º
0 »»
−50 »¼
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª 3
19. Let A = « −2
«
«¬ 4
−2
6
2
°
the eigenspace is ®
°
¯
4º
2 »» . The eigenvalues of A are 7 and –2. For λ = 7, one computes that a basis for
3»¼
ª −1º ª 1º
« 2» , «0»
« » « »
«¬ 0 »¼ «¬ 1»¼
½
°
¾ . This basis may be converted via orthogonal projection to an
°
¿
ª −1º ª 4 º ½
°
°
orthogonal basis for the eigenspace: ® «« 2 »» , «« 2 »» ¾. These vectors can be normalized to get
° « 0 » « 5» °
¯¬ ¼ ¬ ¼¿
ª 4 / 45 º
ª −1/ 5 º
«
»
«
»
u1 = « 2 / 5 » , u 2 = « 2 / 45 » . For λ = –2, one computes that a basis for the eigenspace is
«
»
«
0 »»
«¬ 5 / 45 »¼
«¬
¼
ª −2 / 3 º
which can be normalized to get u 3 = « −1/ 3» . Let
«
»
«¬ 2 / 3»¼
P = [u1
u2
ª −1/ 5
«
u3 ] = « 2 / 5
«
0
«¬
4 / 45
2 / 45
5 / 45
−2 / 3º
ª7
»
−1/ 3» and D = «« 0
»
«¬ 0
2 / 3»¼
0
7
0
ª −2 º
« −1» ,
« »
«¬ 2 »¼
0º
0 »»
−2 »¼
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª 7
20. Let A = « −4
«
«¬ 4
−4
5
0
4º
0»» . The eigenvalues of A are 13, 7, and 1. For λ = 13, one computes that a basis
9 »¼
ª 2º
ª 2 / 3º
«
»
for the eigenspace is −1 , which can be normalized to get u1 = « −1/ 3» . For λ = 7, one computes
« »
«
»
«¬ 2 »¼
«¬ 2 / 3»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.1
• Solutions
411
ª −1º
ª −1/ 3º
«
»
that a basis for the eigenspace is 2 , which can be normalized to get u 2 = « 2 / 3» . For λ = 1, one
« »
«
»
«¬ 2 »¼
«¬ 2 / 3»¼
ª 2º
ª 2 / 3º
«
»
computes that a basis for the eigenspace is 2 , which can be normalized to get u3 = « 2 / 3» . Let
« »
«
»
«¬ −1»¼
«¬ −1/ 3»¼
P = [u1
u2
ª 2/3
u 3 ] = «« −1/ 3
«¬ 2 / 3
−1/ 3
2/3
2/3
2 / 3º
ª13
»
2 / 3» and D = «« 0
«¬ 0
−1/ 3»¼
0
7
0
0º
0 »»
1»¼
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª4
«1
21. Let A = «
«3
«
¬«1
1
4
3
1
1
3
4
1
1º
3 »»
. The eigenvalues of A are 9, 5, and 1. For λ = 9, one computes that a basis
1»
»
4 ¼»
ª1º
ª1/ 2 º
«1»
«1/ 2 »
«
»
« » . For λ = 5, one computes that a
u
=
,
which
can
be
normalized
to
get
for the eigenspace is
1
«1»
«1/ 2 »
« »
« »
¬«1¼»
¬«1/ 2 ¼»
ª −1/ 2 º
ª−1º
« 1/ 2 »
« 1»
» . For λ = 1, one
basis for the eigenspace is « » , which can be normalized to get u 2 = «
« −1/ 2 »
«−1»
«
»
« »
«¬ 1/ 2 ¼»
¬« 1¼»
ª −1º ª 0 º ½
°« » « »°
° 0 −1 °
computes that a basis for the eigenspace is ® « » , « » ¾ . This basis is an orthogonal basis for the
° « 1» « 0 » °
°« » « »°
¯ ¬« 0 ¼» ¬« 1¼» ¿
0º
ª −1/ 2 º
ª
«
»
«
»
−1/ 2 »
0»
«
«
eigenspace, and these vectors can be normalized to get u3 = «
. Let
» , u4 = «
0»
1/
2
«
»
«
»
«¬
«¬ 1/ 2 »¼
0 »¼
P = [u1
u2
u3
ª1/ 2
«
«1/ 2
u4 ] = «
«1/ 2
«
¬1/ 2
−1/ 2
−1/ 2
1/ 2
0
−1/ 2
1/ 2
1/ 2
0
0º
ª9
»
«0
−1/ 2 »
«
and
D
=
»
«0
0»
«
»
«¬ 0
1/ 2 ¼
0
0
5
0
0
1
0
0
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0º
0 »»
0»
»
1 »¼
412
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª2
«0
22. Let A = «
«0
«
«¬ 0
0º
1»»
. The eigenvalues of A are 2 and 0. For λ = 2, one computes that a basis for
0 2 0»
»
1 0 1»¼
ª 1º ª 0 º ª 0 º ½
°« » « » « »°
° 0 1 0 °
the eigenspace is ® « » , « » , « » ¾ . This basis is an orthogonal basis for the eigenspace, and these
° « 0 » « 0 » « 1» °
° ««¬ 0 »»¼ ««¬ 1»»¼ ««¬ 0 »»¼ °
¯
¿
0º
ª
ª 1º
ª0 º
«
»
«0»
«0 »
1/ 2 »
«
«
»
, u =
vectors can be normalized to get u1 =
, and u3 = « » . For λ = 0, one computes
«0» 2 «
« 1»
0»
«
»
« »
« »
«¬0 »¼
«¬0 »¼
«¬1/ 2 »¼
0
1
0
0
ª
ª 0º
«
«−1»
−1/
«
»
, which can be normalized to get u 4 = «
that a basis for the eigenspace is
«
« 0»
«
« »
«¬ 1»¼
«¬ 1/
P = [u1
u2
u3
0
ª1
«
0 1/ 2
u 4 ] = ««
0
0
«
«¬ 0 1/ 2
0
0
1
0
0º
ª2
»
«0
−1/ 2 »
«
D
and
=
«0
0»
»
«
1/ 2 »¼
¬« 0
0
0
2
0
0
2
0
0
0º
»
2»
. Let
0»
»
2 »¼
0º
0 »»
0»
»
0 ¼»
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª3
23. Let A = « 1
«
«¬ 1
1
3
1
1º
ª1º ª3
»
1» . Since each row of A sums to 5, A ««1»» = «« 1
«¬1»¼ «¬ 1
3»¼
1
3
1
1º ª1º ª5º
ª1º
»
«
»
«
»
1» «1» = «5» = 5 ««1»»
«¬1»¼
3»¼ «¬1»¼ «¬5»¼
ª1º
and «1» is an eigenvector of A with corresponding eigenvalue λ = 5. The eigenvector may be
«»
«¬1»¼
ª1/ 3 º
«
»
normalized to get u1 = «1/ 3 » . For λ = 2, one computes that a basis for the eigenspace is
«
»
¬«1/ 3 »¼
ª −1º ª −1º ½
°« » « »°
® « 1» , « 0» ¾ , so λ = 2 is an eigenvalue of A. This basis may be converted via orthogonal
° « 0 » « 1» °
¯¬ ¼ ¬ ¼¿
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.1
• Solutions
413
ª −1º ª −1º ½
°
°
projection to an orthogonal basis ® «« 1»» , «« −1»» ¾ for the eigenspace, and these vectors can be
° « 0 » « 2» °
¯¬ ¼ ¬ ¼¿
ª −1/ 2 º
ª −1/ 6 º
«
»
«
»
normalized to get u 2 = « 1/ 2 » and u3 = « −1/ 6 » . Let
« 0 »
«
»
«¬
»¼
«¬ 2 / 6 »¼
P = [u1
u2
ª1 / 3
«
u3 ] = «1 / 3
«
«¬1 / 3
−1 / 6 º
ª5
»
−1 / 6 » and D = «« 0
»
2 / 6 »¼
¬« 0
−1 / 2
1/ 2
0
0
2
0
0º
0 »»
2 ¼»
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª 5
24. Let A = « −4
«
«¬ −2
−4
5
2
−2 º
ª −2 º ª −20 º
ª −2 º
ª −2 º
«
»
«
»
«
»
»
2 » . One may compute that A « 2 » = « 20 » = 10 « 2 » , so v1 = «« 2 »» is an
«¬ 1»¼ «¬ 10 »¼
«¬ 1»¼
«¬ 1»¼
2 »¼
eigenvector of A with associated eigenvalue λ1 = 10 . Likewise one may compute that
ª 1º
ª 1º ª 1º ª 1º
«
»
«
»
«
»
A « 1» = « 1» = 1 « 1» , so «« 1»» is an eigenvector of A with associated eigenvalue λ 2 = 1 . For λ 2 = 1 , one
¬« 0 ¼»
¬« 0 ¼» ¬« 0 ¼» ¬« 0 ¼»
°
computes that a basis for the eigenspace is ®
°
¯
½
°
¾ . This basis may be converted via orthogonal
°
¿
ª 1º ª 1º ½
°
°
projection to an orthogonal basis for the eigenspace: { v 2 , v 3 } = ® «« 1»» , «« −1»» ¾. The eigenvectors v1 ,
°
°
¯ «¬0 »¼ «¬ 4»¼ ¿
ª 1º ª 1º
« 1» , « 0 »
« » « »
«¬0 »¼ «¬ 2 »¼
ª1/ 2 º
ª 1 / 18 º
ª −2 / 3 º
«
»
«
»
v 2 , and v 3 may be normalized to get the vectors u1 = « 2 / 3» , u 2 = «1/ 2 » , and u3 = « −1 / 18 » .
«
»
«
«
»
«¬ 1/ 3»¼
0 »»
¬«
¼
¬« 4 / 18 ¼»
Let
P = [u1
u2
ª −2 / 3
«
u3 ] = « 2 / 3
«
«¬ 1/ 3
1/ 2
1/ 2
0
1/ 18 º
ª10
»
−1/ 18 » and D = «« 0
»
«¬ 0
4 / 18 »¼
0
1
0
0º
0 »»
1»¼
Then P orthogonally diagonalizes A, and A = PDP −1 .
25. a. True. See Theorem 2 and the paragraph preceding the theorem.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
414
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
b. True. This is a particular case of the statement in Theorem 1, where u and v are nonzero.
c. False. There are n real eigenvalues (Theorem 3), but they need not be distinct (Example 3).
d. False. See the paragraph following formula (2), in which each u is a unit vector.
26. a. True. See Theorem 2.
b. True. See the displayed equation in the paragraph before Theorem 2.
c. False. An orthogonal matrix can be symmetric (and hence orthogonally diagonalizable), but not
every orthogonal matrix is symmetric. See the matrix P in Example 2.
d. True. See Theorem 3(b).
27. Since A is symmetric, ( BT AB)T = BT AT BTT = BT AB , and BT AB is symmetric. Applying this result
with A = I gives B T B is symmetric. Finally, ( BBT )T = BTT BT = BBT , so BBT is symmetric.
28. Let A be an n × n symmetric matrix. Then
( Ax) ⋅ y = ( Ax)T y = xT AT y = xT Ay = x ⋅ ( Ay)
since AT = A .
29. Since A is orthogonally diagonalizable, A = PDP −1 , where P is orthogonal and D is diagonal. Since
A is invertible, A−1 = ( PDP −1 )−1 = PD−1P −1 . Notice that D −1 is a diagonal matrix, so A −1 is
orthogonally diagonalizable.
30. If A and B are orthogonally diagonalizable, then A and B are symmetric by Theorem 2. If AB = BA,
then ( AB)T = ( BA)T = AT BT = AB . So AB is symmetric and hence is orthogonally diagonalizable by
Theorem 2.
31. The Diagonalization Theorem of Section 5.3 says that the columns of P are linearly independent
eigenvectors corresponding to the eigenvalues of A listed on the diagonal of D. So P has exactly k
columns of eigenvectors corresponding to λ. These k columns form a basis for the eigenspace.
32. If A = PRP −1 , then P−1 AP = R . Since P is orthogonal, R = PT AP . Hence
RT = ( PT AP)T = PT AT PTT =
PT AP = R, which shows that R is symmetric. Since R is also upper triangular, its entries above the
diagonal must be zeros to match the zeros below the diagonal. Thus R is a diagonal matrix.
33. It is previously been found that A is orthogonally diagonalized by P, where
ª −1/ 2 −1/ 6
«
P = [u1 u 2 u3 ] = « 1/ 2 −1/ 6
«
0
2/ 6
«¬
Thus the spectral decomposition of A is
A=
T
1u1u1
+
2u 2u 2
T
+
T
3u 3 u 3
1/ 3 º
ª8
»
1/ 3 » and D = ««0
»
1/ 3 »¼
¬«0
T
T
0
6
0
0º
0 »»
3¼»
T
= 8u1u1 + 6u 2u 2 + 3u 3u 3
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.1
−1/ 2
1/ 2
0
ª 1/ 2
= 8 «« −1/ 2
«¬
0
0º
ª 1/ 6
»
0 » + 6 «« 1/ 6
«¬ −2 / 6
0 »¼
−2 / 6 º
ª1/ 3
»
−2 / 6 » + 3 ««1/ 3
«¬1/ 3
4 / 6 »¼
1/ 6
1/ 6
−2 / 6
• Solutions
415
1/ 3º
1/ 3»»
1/ 3»¼
1/ 3
1/ 3
1/ 3
34. It is previously been found that A is orthogonally diagonalized by P, where
ª1/ 2 −1/ 18
«
P = [u1 u 2 u3 ] = «
0
4 / 18
«
1/ 18
«¬1/ 2
Thus the spectral decomposition of A is
A=
T
1u1u1
ª1/ 2
= 7 «« 0
«¬1/ 2
+
2u 2u 2
0
0
0
T
+
T
3u 3 u 3
1/ 2 º
ª 1/18
»
0 » + 7 «« −4 /18
«¬ −1/18
1/ 2 »¼
T
−2 / 3º
ª7
»
−1/ 3» and D = «« 0
»
«¬ 0
2 / 3»¼
T
0
7
0
0º
0 »»
−2 »¼
T
= 7u1u1 + 7u 2 u 2 − 2u 3u 3
−4 /18
16 /18
4 /18
−1/18º
ª 4/9
»
4 /18 » − 2 «« 2 / 9
«¬ −4 / 9
1/18»¼
2/9
1/ 9
−2 / 9
−4 / 9 º
−2 / 9 »»
4 / 9 »¼
35. a. Given x in n, bx = (uuT )x = u(uT x) = (uT x)u, because u T x is a scalar. So Bx = (x ⋅ u)u. Since
u is a unit vector, Bx is the orthogonal projection of x onto u.
b. Since BT = (uuT )T = uTT uT = uuT = B, B is a symmetric matrix. Also,
B2 = (uuT )(uuT ) = u(uT u)uT = uuT = B because u T u = 1.
c. Since u T u = 1 , Bu = (uuT )u = u(uT u) = u(1) = u , so u is an eigenvector of B with corresponding
eigenvalue 1.
36. Given any y in
, let ŷ = By and z = y – ŷ . Suppose that BT = B and B 2 = B . Then
n
B T B = BB = B.
a. Since z ⋅ yˆ = (y − yˆ ) ⋅ ( By ) = y ⋅ ( By ) − yˆ ⋅ ( By) = yT By − ( By)T By = yT By − yT BT By = 0 , z is
orthogonal to yˆ.
b. Any vector in W = Col B has the form Bu for some u. Noting that B is symmetric, Exercise 28
gives
( y – ŷ ) ⋅ (Bu) = [B(y – ŷ )] ⋅ u = [By – BBy] ⋅ u = 0
since B 2 = B. So y – ŷ is in W ⊥ , and the decomposition y = ŷ + (y – ŷ ) expresses y as the sum
of a vector in W and a vector in W ⊥ . By the Orthogonal Decomposition Theorem in Section 6.3,
this decomposition is unique, and so ŷ must be projW y .
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
416
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª 5
« 2
37. [M] Let A = «
« 9
«
¬« −6
2
5
9
−6
−6
9
5
2
−6º
9 »»
. The eigenvalues of A are 18, 10, 4, and –12. For λ = 18, one
2»
»
5¼»
ª−1º
ª −1/ 2 º
« 1»
« 1/ 2 »
» . For
computes that a basis for the eigenspace is « » , which can be normalized to get u1 = «
« −1/ 2 »
«−1»
« »
«
»
¬« 1¼»
¬« 1/ 2 ¼»
ª1º
«1»
λ = 10, one computes that a basis for the eigenspace is « » , which can be normalized to get
«1»
«»
¬«1¼»
ª1/ 2º
«1/ 2»
u 2 = « » . For λ = 4, one computes that a basis for the eigenspace is
«1/ 2»
« »
¬«1/ 2¼»
ª 1º
« 1»
« » , which can be
« −1»
« »
¬« −1¼»
ª 1/ 2 º
ª 1º
« 1/ 2 »
« »
» . For λ = –12, one computes that a basis for the eigenspace is «−1» ,
normalized to get u3 = «
« −1/ 2 »
«−1»
«
»
« »
«¬ −1/ 2 »¼
«¬ 1»¼
ª 1/ 2 º
« −1/ 2 »
» . Let
which can be normalized to get u 4 = «
« −1/ 2 »
«
»
«¬ 1/ 2 »¼
1/ 2
1/ 2º
0º
ª18 0 0
ª −1/ 2 1/ 2
«
« 1/ 2 1/ 2
»
1/ 2 −1/ 2»
0 10 0
0 »»
«
«
P = [u1 u 2 u3 u 4 ] =
. Then P
and D =
« −1/ 2 1/ 2 −1/ 2 −1/ 2»
« 0
0 4
0»
«
»
«
»
1/ 2»¼
0 0 −12 »¼
«¬ 0
«¬ 1/ 2 1/ 2 −1/ 2
orthogonally diagonalizes A, and A = PDP −1 .
−.04 º
.12 »»
. The eigenvalues of A are .25, .30, .55, and .75. For λ =
−.04
.47 −.12»
»
.12 −.12
.41»¼
ª4º
ª.8º
«2»
«.4 »
«
»
, which can be normalized to get u1 = « » .
.25, one computes that a basis for the eigenspace is
«2»
«.4 »
« »
« »
«¬1 »¼
«¬.2 »¼
For
ª .38
« −.18
38. [M] Let A = «
« −.06
«
«¬ −.04
−.18
.59
−.06
−.04
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.1
• Solutions
417
ª −1º
« −2 »
λ = .30, one computes that a basis for the eigenspace is « » , which can be normalized to get
« 2»
« »
¬« 4 ¼»
ª −.2 º
« −.4 »
u 2 = « » . For λ = .55, one computes that a basis for the eigenspace is
« .4 »
« »
¬« .8¼»
ª 2º
« −1»
« » , which can be
« −4 »
« »
¬« 2 ¼»
ª .4 º
« −.2 »
normalized to get u3 = « » . For λ = .75, one computes that a basis for the eigenspace is
« −.8»
« »
¬« .4 ¼»
ª −.4º
« .8»
which can be normalized to get u 4 = « » . Let P = [u1
« −.2»
« »
¬« .4¼»
ª.25
« 0
and D = «
« 0
«
«¬ 0
0
.30
0
0
0
0
.55
0
ª .31
«.58
39. [M] Let A = «
«.08
«
¬«.44
u2
u3
ª.8
«.4
u4 ] = «
«.4
«
¬«.2
−.2
−.4
.4
−.2
.4
.8
−.8
.4
ª −2 º
« 4»
« »,
« −1»
« »
¬« 2 ¼»
−.4 º
.8»»
−.2 »
»
.4 ¼»
0º
0 »»
. Then P orthogonally diagonalizes A, and A = PDP −1 .
0»
»
.75»¼
.58
−.56
.08
.44
.44
−.58
.19
−.08
.44º
−.58»»
. The eigenvalues of A are .75, 0, and –1.25. For λ = .75, one
−.08»
»
.31¼»
ª1 º ª 3 º ½
°« » « »°
° 0 2 °
computes that a basis for the eigenspace is ® « » , « » ¾ . This basis may be converted via orthogonal
° «0 » « 2 » °
° «¬«1 »¼» «¬« 0 »¼» °
¯
¿
ª1 º ª 3 º ½
°« » « »°
4 °
° 0
projection to the orthogonal basis ® « » , « » ¾ . These vectors can be normalized to get
° «0 » « 4 » °
°« » « »°
¯ «¬1 »¼ «¬ −3»¼ ¿
ª 3/ 50 º
ª1/ 2 º
ª −2 º
«
»
«
»
« −1»
0»
« 4 / 50 »
«
« »
u1 = «
, u2 = «
» . For λ = 0, one computes that a basis for the eigenspace is « 4 » ,
0»
« 4 / 50 »
«
»
« »
«
»
«¬ 2 »¼
«¬1/ 2 »¼
3/
50
−
¬
¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
418
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª −.4º
« −.2»
which can be normalized to get u 3 = « » . For λ = –1.25, one computes that a basis for the
« .8»
« »
¬« .4¼»
ª −.4 º
ª −2 º
« .8»
« 4»
eigenspace is « » , which can be normalized to get u 4 = « » .
« −.2 »
« −1»
« »
« »
¬« .4 ¼»
¬« 2 ¼»
Let P = [u1
u2
u3
ª1/
«
«
u4 ] = «
«
«
¬1/
2
3/ 50
−.4
0
4 / 50
−.2
0
4 / 50
.8
2
−3/ 50
.4
−.4 º
ª.75
»
« 0
.8»
«
=
D
and
»
« 0
−.2 »
«
»
«¬ 0
.4 ¼
0
.75
0
0
0
0
0
0
0º
0 »»
.
0»
»
−1.25»¼
Then P orthogonally diagonalizes A, and A = PDP −1 .
ª 10
« 2
«
40. [M] Let A = « 2
«
« −6
«¬ 9
−6
9º
10
2 −6
9 »»
2 10 −6
9 » . The eigenvalues of A are 8, 32, –28, and 17. For λ = 8,
»
9»
−6 −6 26
9
9
9 −19 »¼
ª 1 º ª −1º ½
°« » « »°
°° « −1» « 0 » °°
one computes that a basis for the eigenspace is ® « 0 » , « 1 » ¾ . This basis may be converted via
°«0» «0» °
°« » « »°
« » « »
¯° ¬ 0 ¼ ¬ 0 ¼ ¿°
ª1º ª1º ½
°« » « »°
°° « −1» « 1 » °°
orthogonal projection to the orthogonal basis ® « 0 » , « −2 » ¾ . These vectors can be normalized to get
°«0» «0»°
°« » « »°
« » « »
¯° ¬ 0 ¼ ¬ 0 ¼ ¿°
ª 1/ 6 º
ª 1/ 2 º
ª 1º
«
»
«
»
« 1»
« 1/ 6 »
« −1/ 2 »
« »
» . For λ = 32, one computes that a basis for the eigenspace is « 1» ,
», u = «
u1 = «
0
2
« −2 / 6 »
«
»
« »
«
«
0»
0»
« −3»
«
»
«
»
«¬ 0 »¼
0 ¼»
0 ¼»
¬«
¬«
2
2
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.1
ª 1/
«
« 1/
«
which can be normalized to get u3 = « 1/
«
« −3/
«¬
• Solutions
419
12 º
»
12 »
»
12 » . For λ = –28, one computes that a basis for the
»
12 »
0»¼
ª 1/ 20 º
ª 1º
«
»
« 1»
« 1/ 20 »
« »
«
»
eigenspace is « 1» , which can be normalized to get u 4 = « 1/ 20 » . For λ = 17, one computes that
« »
«
»
« 1»
« 1/ 20 »
«¬ −4 »¼
« −4 / 20 »
¬
¼
ª1/ 5 º
ª1º
«
»
«1»
«1/ 5 »
« »
«
»
a basis for the eigenspace is «1» , which can be normalized to get u5 = «1/ 5 » . Let
« »
«
»
«1»
«1/ 5 »
«¬1»¼
«1/ 5 »
¬
¼
ª 1/ 2
1/ 6
1/ 12
1/ 20 1/ 5 º
«
»
1/ 6
1/ 12
1/ 20 1/ 5 »
« −1/ 2
«
»
P = [u1 u 2 u3 u 4 u5 ] = «
0 −2 / 6
1/ 12
1/ 20 1/ 5 » and
«
»
0
0 −3/ 12
1/ 20 1/ 5 »
«
«
0
0
0 −4 / 20 1/ 5 »¼
¬
0
0
0º
ª8 0
«0 8
0
0
0 »»
«
D = « 0 0 32
0
0 » . Then P orthogonally diagonalizes A, and A = PDP −1 .
«
»
0 −28
0»
«0 0
«¬ 0 0
0
0 17 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
420
7.2
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
SOLUTIONS
Notes: This section can provide a good conclusion to the course, because the mathematics here is widely
used in applications. For instance, Exercises 23 and 24 can be used to develop the second derivative test
for functions of two variables. However, if time permits, some interesting applications still lie ahead.
Theorem 4 is used to prove Theorem 6 in Section 7.3, which in turn is used to develop the singular value
decomposition.
1. a. xT Ax = [ x1
ª 5
x2 ] «
¬1/ 3
1/ 3º ª x1 º
2
= 5 x1 + (2 / 3) x1 x2 + x22
1 »¼ «¬ x2 »¼
ª6 º
b. When x = « » , xT Ax = 5(6)2 + (2/ 3)(6)(1) + (1)2 = 185.
¬1 ¼
ª1 º
c. When x = « » , xT Ax = 5(1)2 + (2/ 3)(1)(3) + (3)2 = 16.
¬ 3¼
2. a. x Ax = [ x1
T
x2
ª4
x3 ] «« 3
«¬ 0
3
2
1
0 º ª x1 º
1 »» «« x2 »» = 4 x12 + 2 x22 + x32 + 6 x1 x2 + 2 x2 x3
1 »¼ «¬ x3 »¼
ª 2º
b. When x = « −1» , xT Ax = 4(2)2 + 2(−1)2 + (5)2 + 6(2)(−1) + 2(−1)(5) = 21.
« »
«¬ 5 »¼
ª1/ 3 º
«
»
c. When x = «1/ 3 » ,
«
»
¬«1/ 3 »¼
xT Ax = 4(1/ 3)2 + 2(1/ 3)2 + (1/ 3)2 + 6(1/ 3)(1/ 3) + 2(1/ 3)(1/ 3) = 5.
ª 10
3. a. The matrix of the quadratic form is «
¬ −3
ª 5
b. The matrix of the quadratic form is «
¬3/ 2
ª 20
4. a. The matrix of the quadratic form is «
¬15 / 2
ª 0
b. The matrix of the quadratic form is «
¬1/ 2
ª 8
5. a. The matrix of the quadratic form is « −3
«
«¬ 2
−3º
.
−3»¼
3/ 2 º
.
0 »¼
15 / 2 º
.
−10 »¼
1/ 2 º
.
0 »¼
−3
7
−1
2º
−1»» .
−3»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.2
ª0
b. The matrix of the quadratic form is « 2
«
«¬ 3
ª 0
b. The matrix of the quadratic form is « −2
«
«¬ 0
421
3º
−4 »» .
0 »¼
2
0
−4
5
ª
«
6. a. The matrix of the quadratic form is 5 / 2
«
«¬ −3 / 2
• Solutions
5/ 2
−1
0
−2
0
2
−3/ 2 º
0 »» .
7 »¼
0º
2 »» .
1»¼
ª 1 5º
7. The matrix of the quadratic form is A = «
» . The eigenvalues of A are 6 and –4. An eigenvector
¬5 1¼
ª1/ 2 º
ª1º
ª −1º
for λ = 6 is « » , which may be normalized to u1 = «
» . An eigenvector for λ = –4 is « » ,
¬1¼
¬ 1¼
¬«1/ 2 ¼»
ª −1/ 2 º
−1
which may be normalized to u 2 = «
» . Then A = PDP , where
«¬ 1/ 2 »¼
ª1/ 2 −1/ 2 º
0º
ª6
P = [u1 u 2 ] = «
» and D = «
» . The desired change of variable is x = Py, and
1/ 2 ¼»
¬0 −4 ¼
¬«1/ 2
the new quadratic form is
xT Ax = ( Py)T A( Py) = yT PT APy = yT Dy = 6 y12 − 4 y22
ª 9 −4 4 º
8. The matrix of the quadratic form is A = « −4
7 0 »» . The eigenvalues of A are 3, 9, and 15. An
«
«¬ 4
0 11»¼
ª −2 / 3 º
ª −2 º
«
»
eigenvector for λ = 3 is −2 , which may be normalized to u1 = « −2 / 3» . An eigenvector for λ = 9
«
»
« »
«¬ 1/ 3»¼
«¬ 1»¼
ª −1/ 3º
ª −1º
«
»
is 2 , which may be normalized to u 2 = « 2 / 3» . An eigenvector for λ = 15 is
«
»
« »
«¬ 2 / 3»¼
«¬ 2 »¼
ª 2 / 3º
be normalized to u3 = « −1/ 3» . Then A = PDP −1 , where
«
»
«¬ 2 / 3»¼
2 / 3º
ª −2 / 3 −1/ 3
ª3
«
»
P = [u1 u 2 u3 ] = « −2 / 3
2 / 3 −1/ 3» and D = «« 0
«¬ 1/ 3
«¬ 0
2/3
2 / 3»¼
is x = Py, and the new quadratic form is
0
9
0
ª 2º
« −1» , which may
« »
«¬ 2 »¼
0º
0 »» . The desired change of variable
15»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
422
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
xT Ax = ( Py )T A( Py) = yT PT APy = yT Dy = 3 y12 + 9 y22 + 15 y32
ª 3
9. The matrix of the quadratic form is A = «
¬ −2
−2 º
. The eigenvalues of A are 7 and 2, so the
6 »¼
ª −1º
quadratic form is positive definite. An eigenvector for λ = 7 is « » , which may be normalized to
¬ 2¼
ª −1/ 5 º
ª2 / 5 º
ª2º
u1 = «
» . An eigenvector for λ = 2 is « » , which may be normalized to u 2 = «
» . Then
¬1 ¼
¬« 2 / 5 ¼»
¬« 1/ 5 ¼»
ª −1/ 5 2 / 5 º
ª7
u2 ] = «
» and D = «
¬0
«¬ 2 / 5 1/ 5 »¼
variable is x = Py, and the new quadratic form is
A = PDP −1 , where P = [u1
0º
. The desired change of
2 »¼
xT Ax = ( Py)T A( Py) = yT PT APy = yT Dy = 7 y12 + 2 y22
ª 9
10. The matrix of the quadratic form is A = «
¬ −4
−4 º
. The eigenvalues of A are 11 and 1, so the
3»¼
ª 2º
quadratic form is positive definite. An eigenvector for λ = 11 is « » , which may be normalized to
¬ −1¼
ª 2/ 5º
ª 1/ 5 º
ª1 º
u1 = «
» . An eigenvector for λ = 1 is « » , which may be normalized to u 2 = «
» . Then
¬2¼
¬« −1/ 5 ¼»
¬« 2 / 5 ¼»
ª 2 / 5 1/ 5 º
ª11
u2 ] = «
» and D = «
¬0
¬« −1/ 5 2 / 5 ¼»
variable is x = Py, and the new quadratic form is
A = PDP −1 , where P = [u1
0º
. The desired change of
1»¼
xT Ax = ( Py)T A( Py) = yT PT APy = yT Dy = 11y12 + y22
ª2
11. The matrix of the quadratic form is A = «
¬5
5º
. The eigenvalues of A are 7 and –3, so the quadratic
2 »¼
ª1/ 2 º
ª1º
form is indefinite. An eigenvector for λ = 7 is « » , which may be normalized to u1 = «
» . An
¬1¼
¬«1/ 2 ¼»
ª −1/ 2 º
ª −1º
−1
eigenvector for λ = –3 is « » , which may be normalized to u 2 = «
» . Then A = PDP ,
¬ 1¼
«¬ 1/ 2 »¼
ª1/ 2 −1/ 2 º
0º
ª7
where P = [u1 u 2 ] = «
» and D = «
» . The desired change of variable is x
1/ 2 ¼»
¬ 0 −3 ¼
¬«1/ 2
= Py, and the new quadratic form is
xT Ax = ( Py)T A( Py) = yT PT APy = yT Dy = 7 y12 − 3 y22
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.2
ª −5
12. The matrix of the quadratic form is A = «
¬ 2
• Solutions
423
2º
. The eigenvalues of A are –1 and –6, so the
−2 »¼
ª1 º
quadratic form is negative definite. An eigenvector for λ = –1 is « » , which may be normalized to
¬2¼
ª 1/ 5 º
ª −2 / 5 º
ª −2 º
u1 = «
» . An eigenvector for λ = –6 is « » , which may be normalized to u 2 = «
» . Then
¬ 1¼
¬« 2 / 5 ¼»
¬« 1/ 5 ¼»
ª 1/ 5 −2 / 5 º
ª −1
u2 ] = «
» and D = «
1/ 5 »¼
¬ 0
«¬ 2 / 5
variable is x = Py, and the new quadratic form is
A = PDP −1 , where P = [u1
0º
. The desired change of
−6»¼
xT Ax = ( Py)T A( Py) = yT PT APy = yT Dy = − y12 − 6 y22
ª 1
13. The matrix of the quadratic form is A = «
¬ −3
−3 º
. The eigenvalues of A are 10 and 0, so the
9 »¼
ª 1º
quadratic form is positive semidefinite. An eigenvector for λ = 10 is « » , which may be
¬ −3 ¼
ª 1/ 10 º
ª 3º
normalized to u1 = «
» . An eigenvector for λ = 0 is « » , which may be normalized to
¬1 ¼
¬« −3 / 10 ¼»
ª 1/ 10 3 / 10 º
ª3 / 10 º
ª10
−1
u2 = «
» and D = «
» . Then A = PDP , where P = [u1 u 2 ] = «
¬ 0
«¬ −3/ 10 1/ 10 »¼
«¬ 1/ 10 »¼
desired change of variable is x = Py, and the new quadratic form is
0º
. The
0 »¼
xT Ax = ( Py )T A( Py ) = yT PT APy = yT Dy = 10 y12
ª8
14. The matrix of the quadratic form is A = «
¬3
3º
. The eigenvalues of A are 9 and –1, so the quadratic
0 »¼
ª3 / 10 º
ª 3º
form is indefinite. An eigenvector for λ = 9 is « » , which may be normalized to u1 = «
» . An
¬1 ¼
¬« 1/ 10 ¼»
ª −1/ 10 º
ª −1º
−1
eigenvector for λ = –1 is « » , which may be normalized to u 2 = «
» . Then A = PDP ,
3
¬ ¼
«¬ 3 / 10 »¼
ª3 / 10 −1/ 10 º
0º
ª9
where P = [u1 u 2 ] = «
» and D = «
» . The desired change of variable is x =
3/ 10 ¼»
¬ 0 −1¼
¬« 1/ 10
Py, and the new quadratic form is
xT Ax = ( Py)T A( Py) = yT PT APy = yT Dy = 9 y12 − y22
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
424
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
2
2
2º
ª −2
« 2 −6
0
0 »»
«
. The eigenvalues of A are 0, –6, –
15. [M] The matrix of the quadratic form is A =
« 2
0 −9
3»
«
»
0
3 −9 ¼»
¬« 2
8, and –12, so the quadratic form is negative semidefinite. The corresponding eigenvectors may be
computed:
ª3º
ª 0º
ª −1º
ª 0º
«1»
« −2»
« 1»
« 0»
«
»
«
»
«
»
, = −6 :
, = −8 :
, = −12 : « »
= 0:
«1»
« 1»
« 1»
« −1»
« »
« »
« »
« »
«¬1»¼
«¬ 1»¼
«¬ 1»¼
«¬ 1»¼
These eigenvectors may be normalized to form the columns of P, and A = PDP −1 , where
ª3/
«
« 1/
P=«
« 1/
«
¬ 1/
12
0
−1/ 2
12
−2 / 6
1/ 2
12
1/ 6
1/ 2
12
1/ 6
1/ 2
0º
ª0
»
«0
0»
«
and
D
=
»
«0
−1/ 2 »
«
»
«¬ 0
1/ 2 ¼
0
0
−6
0
0
−8
0
0
0º
0 »»
0»
»
−12 »¼
The desired change of variable is x = Py, and the new quadratic form is
xT Ax = ( Py )T A( Py) = yT PT APy = yT Dy = −6 y22 − 8 y32 − 12 y42
0
−2 º
ª 4 3/ 2
«3/ 2
4
2
0 »»
. The eigenvalues of A are
16. [M] The matrix of the quadratic form is A = «
« 0
2
4 3/ 2 »
«
»
0 3/ 2
4 »¼
«¬ −2
13/2 and 3/2, so the quadratic form is positive definite. The corresponding eigenvectors may be
computed:
°
°
= 13/ 2 : ®
°
°
¯
ª− 4 º ª 3 º
« 0» «5»
« »,« »
« 3» « 4 »
« » « »
¬« 5¼» ¬« 0 ¼»
½
°
°
°
°
,
=
3/
2
:
¾
®
°
°
°
°
¿
¯
ª 4º ª 3º
« 0» « −5»
« »,« »
« −3» « 4»
« » « »
¬« 5¼» ¬« 0¼»
½
°
°
¾
°
°
¿
Each set of eigenvectors above is already an orthogonal set, so they may be normalized to form the
columns of P, and A = PDP −1 , where
ª 3 / 50
«
« 5 / 50
P=«
« 4 / 50
«
0
¬
−4 / 50
3 / 50
0
−5 / 50
3/ 50
4 / 50
5 / 50
0
4 / 50 º
ª13 / 2
»
« 0
0»
«
and
D
=
»
« 0
−3/ 50 »
«
»
«¬ 0
5 / 50 ¼
0
0
13/ 2
0
0
3/ 2
0
0
The desired change of variable is x = Py, and the new quadratic form is
xT Ax = ( Py )T A( Py ) = yT PT APy = yT Dy =
13 2 13 2 3 2 3 2
y1 + y2 + y3 + y4
2
2
2
2
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
0º
0 »»
0»
»
3/ 2 »¼
7.2
• Solutions
425
0
−6 º
ª 1 9/ 2
«9 / 2
1
6
0 »»
«
. The eigenvalues of A are
17. [M] The matrix of the quadratic form is A =
« 0
6
1 9 / 2»
«
»
0 9/ 2
1¼»
¬« −6
17/2 and –13/2, so the quadratic form is indefinite. The corresponding eigenvectors may be
computed:
°
°
= 17 / 2 : ®
°
°
¯
ª− 4 º ª 3 º
« 0» «5»
« »,« »
« 3» « 4 »
« » « »
«¬ 5»¼ «¬ 0 »¼
½
°
°
°
°
¾ , = −13/ 2 : ®
°
°
°
°
¿
¯
ª 4 º ª 3º
« 0 » « −5»
« »,« »
« −3» « 4»
« » « »
«¬ 5»¼ «¬ 0»¼
½
°
°
¾
°
°
¿
Each set of eigenvectors above is already an orthogonal set, so they may be normalized to form the
columns of P, and A = PDP −1 , where
ª 3/ 50
«
« 5 / 50
P=«
« 4 / 50
«
0
¬
−4 / 50
3/ 50
0
−5 / 50
3 / 50
4 / 50
5 / 50
0
4 / 50 º
ª17 / 2
»
«
0»
0
«
» and D = «
0
−3/ 50 »
«
»
0
¬«
5 / 50 ¼
0
0
17 / 2
0
0
−13 / 2
0
0
0º
0 »»
0»
»
−13/ 2 ¼»
The desired change of variable is x = Py, and the new quadratic form is
xT Ax = ( Py )T A( Py ) = yT PT APy = yT Dy =
17 2 17 2 13 2 13 2
y1 + y2 − y3 − y4
2
2
2
2
ª 11 −6 −6 −6 º
« −6 −1
0
0 »»
. The eigenvalues of A are 17, 1, –
18. [M] The matrix of the quadratic form is A = «
« −6
0
0 −1»
«
»
0 −1
0 ¼»
¬« −6
1, and –7, so the quadratic form is indefinite. The corresponding eigenvectors may be computed:
ª −3º
ª 0º
ª 0º
ª1º
« 1»
« 0»
« −2»
«1»
= 17 : « » , = 1: « » , = −1: « » , = −7 : « »
« 1»
« −1»
« 1»
«1»
« »
« »
« »
«»
¬« 1¼»
¬« 1¼»
¬« 1¼»
¬«1¼»
These eigenvectors may be normalized to form the columns of P, and A = PDP −1 , where
ª −3 /
«
« 1/
P=«
« 1/
«
¬ 1/
12
0
0
12
0
−2 / 6
12
−1 / 2
1/ 6
12
1/ 2
1/ 6
1 / 2º
ª17
»
« 0
1 / 2»
«
and
D
=
»
« 0
1 / 2»
«
»
¬ 0
1 / 2¼
0
0
1
0
0
−1
0
0
0º
0 »»
0»
»
−7 ¼
The desired change of variable is x = Py, and the new quadratic form is
xT Ax = ( Py)T A( Py) = yT PT APy = yT Dy = 17 y12 + y22 − y32 − 7 y42
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
426
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
19. Since 8 is larger than 5, the x22 term should be as large as possible. Since x12 + x22 = 1 , the largest
value that x2 can take is 1, and x1 = 0 when x2 = 1 . Thus the largest value the quadratic form can
take when xT x = 1 is 5(0) + 8(1) = 8.
20. Since 5 is larger in absolute value than –3, the x12 term should be as large as possible. Since
2
2
x1 + x2 = 1 , the largest value that x1 can take is 1, and x 2 = 0 when x1 = 1 . Thus the largest value
the quadratic form can take when xT x = 1 is 5(1) – 3(0) = 5.
21. a. True. See the definition before Example 1, even though a nonsymmetric matrix could be used to
compute values of a quadratic form.
b. True. See the paragraph following Example 3.
c. True. The columns of P in Theorem 4 are eigenvectors of A. See the Diagonalization Theorem in
Section 5.3.
d. False. Q(x) = 0 when x = 0.
e. True. See Theorem 5(a).
f. True. See the Numerical Note after Example 6.
22. a. True. See the paragraph before Example 1.
b. False. The matrix P must be orthogonal and make PT AP diagonal. See the paragraph before
Example 4.
c. False. There are also “degenerate” cases: a single point, two intersecting lines, or no points at all.
See the subsection “A Geometric View of Principal Axes.”
d. False. See the definition before Theorem 5.
e. True. See Theorem 5(b). If xT Ax has only negative values for x ≠ 0, then xT Ax is negative
definite.
23. The characteristic polynomial of A may be written in two ways:
ªa −
det( A − I ) = det «
¬ b
b º
=
d − »¼
2
− ( a + d ) + ad − b 2
and
( − 1 )( −
2)
=
2
−(
1
+
2)
+
1 2
The coefficients in these polynomials may be equated to obtain
1
+
2
= a + d and
1 2
=
2
ad − b = det A .
24. If det A > 0, then by Exercise 23,
1 2
> 0 , so that
1
and
2
have the same sign; also,
2
ad = det A + b > 0 .
a. If det A > 0 and a > 0, then d > 0 also, since ad > 0. By Exercise 23, 1 + 2 = a + d > 0 . Since
and 2 have the same sign, they are both positive. So Q is positive definite by Theorem 5.
b. If det A > 0 and a < 0, then d < 0 also, since ad > 0. By Exercise 23, 1 + 2 = a + d < 0 . Since
and 2 have the same sign, they are both negative. So Q is negative definite by Theorem 5.
c. If det A < 0, then by Exercise 23,
indefinite by Theorem 5.
1 2
< 0 . Thus
1
and
2
have opposite signs. So Q is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
1
1
7.3
• Solutions
427
25. Exercise 27 in Section 7.1 showed that B T B is symmetric. Also xT BT Bx = ( Bx)T Bx = || Bx || ≥ 0 , so
the quadratic form is positive semidefinite, and the matrix B T B is positive semidefinite. Suppose
that B is square and invertible. Then if xT B T Bx = 0, || Bx || = 0 and Bx = 0. Since B is invertible, x =
0. Thus if x ≠ 0, x T B T B x > 0 and B T B is positive definite.
26. Let A = PDPT , where P T = P −1 . The eigenvalues of A are all positive: denote them
1 ,… ,
n.
Let C
be the diagonal matrix with
1 ,…,
n on its diagonal. Then D = C = C C . If B = PCP , then B
is positive definite because its eigenvalues are the positive numbers on the diagonal of C. Also
BT B = ( PCPT )T ( PCPT ) = ( PTT CT PT )( PCPT ) = PCT CPT = PDPT = A
2
T
T
since P T P = I .
27. Since the eigenvalues of A and B are all positive, the quadratic forms xT Ax and xT Bx are positive
definite by Theorem 5. Let x ≠ 0. Then x T A x > 0 and x T B x > 0 , so xT ( A + B)x = xT Ax + xT Bx > 0 ,
and the quadratic form xT ( A + B)x is positive definite. Note that A + B is also a symmetric matrix.
Thus by Theorem 5 all the eigenvalues of A + B must be positive.
28. The eigenvalues of A are all positive by Theorem 5. Since the eigenvalues of A −1 are the reciprocals
of the eigenvalues of A (see Exercise 25 in Section 5.1), the eigenvalues of A −1 are all positive. Note
that A −1 is also a symmetric matrix. By Theorem 5, the quadratic form xT A−1x is positive definite.
7.3
SOLUTIONS
Notes: Theorem 6 is the main result needed in the next two sections. Theorem 7 is mentioned in Example
2 of Section 7.4. Theorem 8 is needed at the very end of Section 7.5. The economic principles in Example
6 may be familiar to students who have had a course in macroeconomics.
2
0º
ª5
«
1. The matrix of the quadratic form on the left is A = 2
6 −2 »» . The equality of the quadratic
«
«¬ 0 −2
7 »¼
forms implies that the eigenvalues of A are 9, 6, and 3. An eigenvector may be calculated for each
eigenvalue and normalized:
ª 1/ 3º
ª 2 / 3º
ª −2 / 3 º
«
»
«
»
= 9 : « 2 / 3» , = 6 : « 1/ 3» , = 3 : «« 2 / 3»»
«¬ −2 / 3»¼
«¬ 1/ 3»¼
«¬ 1/ 3»¼
ª 1/ 3
The desired change of variable is x = Py, where P = « 2 / 3
«
«¬ −2 / 3
2/3
1/ 3
2/3
−2 / 3 º
2 / 3»» .
1/ 3»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
428
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª3 1 1º
2. The matrix of the quadratic form on the left is A = « 1 2 2 » . The equality of the quadratic forms
«
»
«¬ 1 2 2 »¼
implies that the eigenvalues of A are 5, 2, and 0. An eigenvector may be calculated for each
eigenvalue and normalized:
ª1/ 3 º
ª −2 / 6 º
0º
ª
«
»
«
»
«
»
= 5 : «1/ 3 » , = 2 : « 1/ 6 » , = 0 : « −1/ 2 »
«
»
«
»
« 1/ 2 »
«¬1/ 3 »¼
«¬ 1/ 6 »¼
¬
¼
ª1/ 3
«
The desired change of variable is x = Py, where P = «1/ 3
«
¬«1/ 3
−2 / 6
1/ 6
1/ 6
0º
»
−1/ 2 » .
»
1/ 2 ¼»
3. (a) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 is the greatest
eigenvalue 1 of A. By Exercise 1, 1 = 9.
(b) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 occurs at a unit
ª 1/ 3º
eigenvector u corresponding to the greatest eigenvalue 1 of A. By Exercise 1, u = ± « 2 / 3» .
«
»
«¬ −2 / 3»¼
(c) By Theorem 7, the maximum value of xT Ax subject to the constraints xT x = 1 and x T u = 0 is
the second greatest eigenvalue 2 of A. By Exercise 1, 2 = 6.
4. (a) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 is the greatest
eigenvalue 1 of A. By Exercise 2, 1 = 5.
(b) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 occurs at a unit
ª1/ 3 º
«
»
eigenvector u corresponding to the greatest eigenvalue 1 of A. By Exercise 2, u = ± «1/ 3 » .
«
»
¬«1/ 3 ¼»
(c) By Theorem 7, the maximum value of xT Ax subject to the constraints xT x = 1 and x T u = 0 is
the second greatest eigenvalue 2 of A. By Exercise 2, 2 = 2.
ª 5
5. The matrix of the quadratic form is A = «
¬ −2
−2 º
. The eigenvalues of A are
5»¼
1
= 7 and
2
= 3.
(a) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 is the greatest
eigenvalue 1 of A, which is 7.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.3
• Solutions
429
(b) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 occurs at a unit
ª −1º
eigenvector u corresponding to the greatest eigenvalue 1 of A. One may compute that « » is
¬ 1¼
an eigenvector corresponding to
1
ª −1/ 2 º
= 7, so u = ± «
».
¬« 1/ 2 »¼
(c) By Theorem 7, the maximum value of xT Ax subject to the constraints xT x = 1 and x T u = 0 is
the second greatest eigenvalue 2 of A, which is 3.
ª 7
6. The matrix of the quadratic form is A = «
¬3/ 2
2
3/ 2 º
. The eigenvalues of A are
3»¼
1
= 15 / 2 and
= 5 / 2.
(a) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 is the greatest
eigenvalue 1 of A, which is 15/2.
(b) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 occurs at a unit
ª 3º
eigenvector u corresponding to the greatest eigenvalue 1 of A. One may compute that « » is an
¬1 ¼
eigenvector corresponding to
1
ª3 / 10 º
= 15 / 2, so u = ± «
».
«¬ 1/ 10 »¼
(c) By Theorem 7, the maximum value of xT Ax subject to the constraints xT x = 1 and x T u = 0 is
the second greatest eigenvalue 2 of A, which is 5/2.
7. The eigenvalues of the matrix of the quadratic form are
1
T
= 2,
2
= −1, and
3
= −4. By Theorem
T
6, the maximum value of x Ax subject to the constraint x x = 1 occurs at a unit eigenvector u
ª1/ 2 º
corresponding to the greatest eigenvalue 1 of A. One may compute that « 1» is an eigenvector
« »
«¬ 1»¼
corresponding to
1
ª 1/ 3º
= 2, so u = ± « 2 / 3» .
«
»
«¬ 2 / 3»¼
8. The eigenvalues of the matrix of the quadratic form are
T
1
= 9, and
2
= −3. By Theorem 6, the
T
maximum value of x Ax subject to the constraint x x = 1 occurs at a unit eigenvector u
ª −1º
ª −2 º
«
»
corresponding to the greatest eigenvalue 1 of A. One may compute that 0 and « 1» are linearly
« »
« »
«¬ 1»¼
«¬ 0 »¼
independent eigenvectors corresponding to 1 = 9, so u can be any unit vector which is a linear
ª −1º
ª −2 º
«
»
combination of 0 and « 1» . Alternatively, u can be any unit vector which is orthogonal to the
« »
« »
«¬ 1»¼
«¬ 0 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
430
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª1 º
eigenspace corresponding to the eigenvalue 2 = −3. Since multiples of « 2 » are eigenvectors
« »
«¬ 1 »¼
ª1 º
corresponding to 2 = −3, u can be any unit vector orthogonal to « 2 » .
« »
«¬ 1 »¼
9. This is equivalent to finding the maximum value of xT Ax subject to the constraint x T x = 1. By
Theorem 6, this value is the greatest eigenvalue 1 of the matrix of the quadratic form. The matrix of
ª 7
the quadratic form is A = «
¬ −1
−1º
, and the eigenvalues of A are
3»¼
the desired constrained maximum value is
1
1
= 5 + 5,
2
= 5 − 5. Thus
= 5 + 5.
10. This is equivalent to finding the maximum value of xT Ax subject to the constraint xT x = 1 . By
Theorem 6, this value is the greatest eigenvalue 1 of the matrix of the quadratic form. The matrix of
ª −3
the quadratic form is A = «
¬ −1
−1º
, and the eigenvalues of A are
5»¼
the desired constrained maximum value is
1
1
= 1 + 17,
2
= 1 − 17. Thus
= 1 + 17.
11. Since x is an eigenvector of A corresponding to the eigenvalue 3, Ax = 3x, and xT Ax = xT (3x) =
3(xT x) = 3|| x ||2 = 3 since x is a unit vector.
12. Let x be a unit eigenvector for the eigenvalue λ. Then xT Ax = xT ( x) = (xT x) =
So λ must satisfy m ≤ λ ≤ M.
since xT x = 1 .
13. If m = M, then let t = (1 – 0)m + 0M = m and x = u n . Theorem 6 shows that uTn Au n = m. Now
suppose that m < M, and let t be between m and M. Then 0 ≤ t – m ≤ M – m and 0 ≤ (t – m)/(M – m)
≤ 1. Let
α = (t – m)/(M – m), and let x = 1 − α u n + α u1. The vectors 1 − α un and α u1 are orthogonal
because they are eigenvectors for different eigenvectors (or one of them is 0). By the Pythagorean
Theorem
xT x =|| x ||2 = || 1 − α un ||2 + || α u1 ||2 = |1 − α ||| un ||2 + | α ||| u1 ||2 = (1 − α ) + α = 1
since u n and u1 are unit vectors and 0 ≤ α ≤ 1. Also, since u n and u1 are orthogonal,
xT Ax = ( 1 − α un + α u1 )T A( 1 − α un + α u1 )
= ( 1 − α un + α u1 )T (m 1 − α un + M α u1 )
= |1 − α | muTn un + | α | M u1T u1 = (1 − α )m + α M = t
Thus the quadratic form xT Ax assumes every value between m and M for a suitable unit vector x.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.3
ª 0
« 1/ 2
14. [M] The matrix of the quadratic form is A = «
«3/ 2
«
¬« 15
1 = 17,
2 = 13,
3 = − 14, and
4 = − 16.
1/ 2
0
3/ 2
15
15
3/ 2
0
1/ 2
• Solutions
431
15º
3/ 2 »»
. The eigenvalues of A are
1/ 2 »
»
0 ¼»
(a) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 is the greatest
eigenvalue 1 of A, which is 17.
(b) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 occurs at a unit
ª1º
«1»
eigenvector u corresponding to the greatest eigenvalue 1 of A. One may compute that « » is an
«1»
«»
¬«1¼»
ª1/ 2 º
«1/ 2 »
eigenvector corresponding to 1 = 17, so u = ± « » .
«1/ 2 »
« »
¬«1/ 2 ¼»
(c) By Theorem 7, the maximum value of xT Ax subject to the constraints xT x = 1 and x T u = 0 is
the second greatest eigenvalue 2 of A, which is 13.
ª 0 3/ 2
« 3/ 2
0
15. [M] The matrix of the quadratic form is A = «
«5 / 2 7 / 2
«
«¬7 / 2 5 / 2
1 = 15 / 2,
2 = − 1/ 2,
3 = − 5 / 2, and
4 = − 9 / 2.
5/ 2
7/2
0
3/ 2
7 / 2º
5 / 2 »»
. The eigenvalues of A are
3/ 2 »
»
0 »¼
(a) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 is the greatest
eigenvalue 1 of A, which is 15/2.
(b) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 occurs at a unit
ª1º
«1»
eigenvector u corresponding to the greatest eigenvalue 1 of A. One may compute that « » is an
«1»
«»
¬«1¼»
ª1/ 2 º
«1/ 2 »
eigenvector corresponding to 1 = 15 / 2, so u = ± « » .
«1/ 2 »
« »
¬«1/ 2 ¼»
(c) By Theorem 7, the maximum value of xT Ax subject to the constraints xT x = 1 and x T u = 0 is
the second greatest eigenvalue 2 of A, which is –1/2.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
432
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª 4
« −3
16. [M] The matrix of the quadratic form is A = «
« −5
«
¬« −5
2 = 3,
3 = 1, and
4 = − 9.
−3
0
−5
−3
−3
−3
0
−1
−5º
−3»»
. The eigenvalues of A are
−1»
»
0¼»
1
= 9,
(a) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 is the greatest
eigenvalue 1 of A, which is 9.
(b) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 occurs at a unit
ª−2º
« 0»
eigenvector u corresponding to the greatest eigenvalue 1 of A. One may compute that « » is
« 1»
« »
¬« 1¼»
an eigenvector corresponding to
1
ª −2 /
«
«
= 9, so u = ± «
« 1/
« 1/
¬
6º
»
0»
.
6 »»
6 »¼
(c) By Theorem 7, the maximum value of xT Ax subject to the constraints xT x = 1 and x T u = 0 is the
second greatest eigenvalue 2 of A, which is 3.
ª −6
« −2
17. [M] The matrix of the quadratic form is A = «
« −2
«
¬« −2
1 = − 4,
2 = − 10,
3 = − 12, and
4 = − 16.
−2
−10
−2
0
0
0
−13
3
−2 º
0 »»
. The eigenvalues of A are
3»
»
−13¼»
(a) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 is the greatest
eigenvalue 1 of A, which is –4.
(b) By Theorem 6, the maximum value of xT Ax subject to the constraint xT x = 1 occurs at a unit
ª−3º
« 1»
eigenvector u corresponding to the greatest eigenvalue 1 of A. One may compute that « » is
« 1»
« »
¬« 1¼»
ª −3/
«
« 1/
an eigenvector corresponding to 1 = −4, so u = ± «
« 1/
«
¬ 1/
12 º
»
12 »
».
12 »
»
12 ¼
(c) By Theorem 7, the maximum value of xT Ax subject to the constraints xT x = 1 and x T u = 0 is
the second greatest eigenvalue 2 of A, which is –10.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.4
7.4
• Solutions
433
SOLUTIONS
Notes: The section presents a modern topic of great importance in applications, particularly in computer
calculations. An understanding of the singular value decomposition is essential for advanced work in
science and engineering that requires matrix computations. Moreover, the singular value decomposition
explains much about the structure of matrix transformations. The SVD does for an arbitrary matrix almost
what an orthogonal decomposition does for a symmetric matrix.
0º
ª1
. Then AT A = «
»
− 3¼
¬0
ª1
1. Let A = «
¬0
order)
1
= 9 and
ª −5
2. Let A = «
¬ 0
order)
1
2
2
2
2
0º
, and the eigenvalues of AT A are seen to be (in decreasing
0 »¼
= 0. Thus the singular values of A are σ 1 = 25 = 5 and σ 2 = 0 = 0.
ª 6
1º
T
» . Then A A = «
6 ¼»
¬« 6
6º
T
» , and the characteristic polynomial of A A is
7 ¼»
− 13 + 36 = ( − 9)( − 4), and the eigenvalues of AT A are (in decreasing order)
1
= 9 and
= 4. Thus the singular values of A are σ 1 = 9 = 3 and σ 2 = 4 = 2.
ª 3
4. Let A = «
«¬ 0
2
= 1. Thus the singular values of A are σ 1 = 9 = 3 and σ 2 = 1 = 1.
0º
ª 25
. Then AT A = «
»
0¼
¬ 0
= 25 and
ª 6
3. Let A = «
¬« 0
2
0º
, and the eigenvalues of AT A are seen to be (in decreasing
9»¼
ª 3
2º
T
» . Then A A = «
3 »¼
«¬ 2 3
2 3º
T
» , and the characteristic polynomial of A A is
7 »¼
− 10 + 9 = ( − 9)( − 1), and the eigenvalues of AT A are (in decreasing order)
1
= 9 and
= 1. Thus the singular values of A are σ 1 = 9 = 3 and σ 2 = 1 = 1.
ª9 0º
ª −3 0 º
T
5. Let A = «
. Then AT A = «
» , and the eigenvalues of A A are seen to be (in decreasing
»
0
0
0
0
¬
¼
¬
¼
order) 1 = 9 and 2 = 0. Associated unit eigenvectors may be computed:
ª 1º
ª0 º
= 9: « », = 0:« »
¬0¼
¬ 1¼
ª 1 0º
Thus one choice for V is V = «
» . The singular values of A are σ 1 = 9 = 3 and σ 2 = 0 = 0.
¬ 0 1¼
ª 3 0º
Thus the matrix Σ is Σ = «
» . Next compute
¬0 0 ¼
u1 =
ª −1º
Av1 = « »
σ1
¬ 0¼
1
Because Av2 = 0, the only column found for U so far is u1. Find the other column of U is found by
ª0º
extending {u1} to an orthonormal basis for 2. An easy choice is u2 = « » .
¬ 1¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
434
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª −1
Let U = «
¬ 0
0º
. Thus
1»¼
ª −1
A =U ΣVT = «
¬ 0
0º ª 3
1»¼ «¬ 0
0º ª 1
0 »¼ «¬ 0
0º
1»¼
0º
ª −2
ª 4 0º
6. Let A = «
. Then AT A = «
, and the eigenvalues of AT A are seen to be (in decreasing
»
»
¬ 0 −1¼
¬ 0 1¼
order) 1 = 4 and 2 = 1. Associated unit eigenvectors may be computed:
ª 1º
ª0º
= 4 : « » , = 1: « »
¬0 ¼
¬ 1¼
ª 1 0º
Thus one choice for V is V = «
» . The singular values of A are σ 1 = 4 = 2 and σ 2 = 1 = 1.
¬ 0 1¼
ª 2 0º
Thus the matrix Σ is Σ = «
» . Next compute
¬ 0 1¼
u1 =
ª −1º
ª 0º
1
Av1 = « » , u 2 =
Av 2 = « »
σ1
σ2
¬ 0¼
¬ −1¼
1
Since {u1 , u 2 } is a basis for
ª −1
A =U ΣVT = «
¬ 0
ª −1
, let U = «
¬ 0
0º
. Thus
−1»¼
2
0º ª 2
−1»¼ «¬ 0
0º ª 1
1»¼ «¬ 0
0º
1»¼
ª 8 2º
ª 2 −1º
7. Let A = «
. Then AT A = «
, and the characteristic polynomial of AT A is
»
»
2¼
¬ 2 5¼
¬2
2
− 13 + 36 = ( − 9)( − 4), and the eigenvalues of AT A are (in decreasing order) 1 = 9 and
2 = 4. Associated unit eigenvectors may be computed:
ª2 / 5 º
ª −1/ 5 º
= 9: «
», = 4: «
»
¬« 1/ 5 »¼
¬« 2 / 5 ¼»
ª2 / 5
Thus one choice for V is V = «
¬« 1/ 5
−1/ 5 º
» . The singular values of A are σ 1 = 9 = 3 and
2 / 5 ¼»
ª3
σ 2 = 4 = 2. Thus the matrix Σ is Σ = «
¬0
u1 =
0º
. Next compute
2 »¼
ª 1/ 5 º
ª −2 / 5 º
1
Av1 = «
Av 2 = «
» , u2 =
»
σ1
σ2
«¬ 1/ 5 ¼»
¬« 2 / 5 »¼
1
Since {u1 , u 2 } is a basis for
ª 1/ 5
, let U = «
«¬ 2 / 5
2
−2 / 5 º
» . Thus
1/ 5 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.4
ª 1/ 5
A =U ΣVT = «
¬« 2 / 5
−2 / 5 º ª 3
»«
1/ 5 ¼» ¬ 0
3º
ª4
. Then AT A = «
»
2¼
¬6
ª2
8. Let A = «
¬0
0º ª 2 / 5
«
2 »¼ ¬« −1/ 5
• Solutions
435
1/ 5 º
»
2 / 5 ¼»
6º
, and the characteristic polynomial of AT A is
»
13¼
2
− 17 + 16 = ( − 16)( − 1), and the eigenvalues of AT A are (in decreasing order)
2 = 1. Associated unit eigenvectors may be computed:
1
= 16 and
ª 1/ 5 º
ª −2 / 5 º
= 16 : «
» , = 1: «
»
«¬ 2 / 5 »¼
«¬ 1/ 5 »¼
ª 1/ 5
Thus one choice for V is V = «
¬« 2 / 5
−2 / 5 º
» . The singular values of A are σ 1 = 16 = 4 and
1/ 5 ¼»
ª4
σ 2 = 1 = 1. Thus the matrix Σ is Σ = «
¬0
u1 =
0º
. Next compute
1»¼
ª2 / 5 º
ª −1/ 5 º
1
Av1 = «
Av 2 = «
» , u2 =
»
σ1
σ2
«¬ 1/ 5 »¼
«¬ 2 / 5 »¼
1
Since {u1 , u 2 } is a basis for
ª2 / 5
A =U ΣVT = «
«¬ 1/ 5
ª7
9. Let A = « 0
«
«¬ 5
ª2 / 5
, let U = «
¬« 1/ 5
2
−1/ 5 º ª 4
»«
2 / 5 »¼ ¬ 0
1º
ª74
0 »» . Then AT A = «
¬32
5»¼
−1/ 5 º
» . Thus
2 / 5 ¼»
0 º ª 1/ 5
«
1»¼ «¬ −2 / 5
2/ 5º
»
1/ 5 »¼
32 º
, and the characteristic polynomial of AT A is
26 »¼
2
− 100 + 900 = ( − 90)( − 10), and the eigenvalues of AT A are (in decreasing order)
and 2 = 10. Associated unit eigenvectors may be computed:
1
= 90
ª2 / 5 º
ª −1/ 5 º
= 90 : «
» , = 10 : «
»
¬« 1/ 5 »¼
¬« 2 / 5 ¼»
ª2 / 5
Thus one choice for V is V = «
«¬ 1/ 5
−1/ 5 º
» . The singular values of A are σ 1 = 90 = 3 10 and
2 / 5 »¼
ª3 10
«
σ 2 = 10. Thus the matrix Σ is Σ = « 0
«
0
«¬
0º
»
10 » . Next compute
0»»
¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
436
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª1/ 2 º
ª −1/ 2 º
«
»
«
»
1
0» , u2 =
Av 2 = «
0»
u1 = Av1 = «
σ1
σ2
«1/ 2 »
« 1/ 2 »
¬
¼
¬
¼
3
Since {u1 , u 2 } is not a basis for , we need a unit vector u3 that is orthogonal to both u1 and u 2 .
1
The vector u3 must satisfy the set of equations u1T x = 0 and uT2 x = 0. These are equivalent to the
linear equations
x1 + 0 x2 + x3
− x1 + 0 x2 + x3
=
=
ª1/ 2
«
0
Therefore let U = «
«1/ 2
¬
ª1/ 2
«
T
A =U ΣV = «
0
«1/ 2
¬
ª4
10. Let A = « 2
«
¬« 0
ª0 º
ª0 º
0
«
»
, so x = « 1» , and u 3 = «« 1»»
0
«¬0 »¼
«¬0 »¼
−1/ 2
0
1/ 2
0º
»
1» . Thus
0»¼
−1/ 2
0
1/ 2
−2 º
ª 20
−1»» . Then AT A = «
¬ −10
0 »¼
0 º ª3 10
»«
1» «
0
«
»
0
0 ¼ «¬
0º
» ª 2/ 5
10 » «
−1/ 5
0 »» «¬
¼
1/ 5 º
»
2 / 5 »¼
−10 º
, and the characteristic polynomial of AT A is
»
5¼
2
− 25 = ( − 25) , and the eigenvalues of AT A are (in decreasing order)
Associated unit eigenvectors may be computed:
1
= 25 and
2
= 0.
ª 2/ 5º
ª 1/ 5 º
= 25 : «
», = 0: «
»
«¬ −1/ 5 »¼
«¬ 2 / 5 »¼
ª 2/ 5
Thus one choice for V is V = «
¬« −1/ 5
1/ 5 º
» . The singular values of A are σ 1 = 25 = 5 and
2 / 5 ¼»
ª5
σ 2 = 0 = 0. Thus the matrix Σ is Σ = «« 0
«¬ 0
0º
0 »» . Next compute
0 »¼
ª2 / 5 º
«
»
1
u1 =
Av1 = « 1/ 5 »
σ1
«
0 »»
«¬
¼
Because Av2 = 0, the only column found for U so far is u1. Find the other columns of U found by
extending {u1} to an orthonormal basis for 3. In this case, we need two orthogonal unit vectors u2
and u3 that are orthogonal to u1. Each vector must satisfy the equation u1T x = 0, which is equivalent
to the equation 2x1 + x2 = 0. An orthonormal basis for the solution set of this equation is
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.4
• Solutions
437
ª 1/ 5 º
ª0 º
«
»
u 2 = « −2 / 5 » , u3 = ««0 »» .
«
«¬ 1»¼
0 »»
«¬
¼
ª2 / 5
«
Therefore, let U = «1/ 5
« 0
«¬
ª2 / 5
«
A = U Σ V T = « 1/ 5
«
0
¬«
ª −3
11. Let A = « 6
«
«¬ 6
1/ 5
−2 / 5
0
0º
»
0 » . Thus
1 »»
¼
1/ 5
−2 / 5
0
0º ª5
»
0 » «« 0
1»» «¬ 0
¼
1º
ª 81
−2 »» . Then AT A = «
¬ −27
−2 »¼
0º
ª2 / 5
0 »» «
1/ 5
0 »¼ «¬
−1/ 5 º
»
2 / 5 »¼
−27 º
, and the characteristic polynomial of AT A is
9 »¼
2
− 90 = ( − 90), and the eigenvalues of AT A are (in decreasing order)
Associated unit eigenvectors may be computed:
1
= 90 and
2
= 0.
ª 3 / 10 º
ª 1/ 10 º
= 90 : «
», = 0:«
».
¬« −1/ 10 »¼
¬«3/ 10 ¼»
ª 3 / 10
Thus one choice for V is V = «
«¬ −1/ 10
1/ 10 º
» . The singular values of A are σ 1 = 90 = 3 10 and
3/ 10 »¼
ª3 10
«
σ 2 = 0 = 0. Thus the matrix Σ is Σ = « 0
«
0
¬
0º
»
0» . Next compute
0»¼
ª −1/ 3º
Av1 = «« 2 / 3»»
u1 =
σ1
«¬ 2 / 3»¼
1
Because Av2 = 0, the only column found for U so far is u1. The other columns of U can be found by
extending {u1} to an orthonormal basis for 3. In this case, we need two orthogonal unit vectors u2
and u3 that are orthogonal to u1. Each vector must satisfy the equation u1T x = 0, which is equivalent
to the equation − x1 + 2 x2 + 2 x3 = 0. An orthonormal basis for the solution set of this equation is
ª 2 / 3º
ª 2 / 3º
«
»
u 2 = « −1/ 3» , u 3 = «« 2 / 3»» .
«¬ 2 / 3»¼
«¬ −1/ 3»¼
ª −1/ 3
Therefore, let U = « 2 / 3
«
«¬ 2 / 3
2/3
−1/ 3
2/3
2 / 3º
2 / 3»» . Thus
−1/ 3»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
438
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª −1/ 3
«
A =U ΣV = « 2/3
«¬ 2 / 3
T
ª 1
12. Let A = « 0
«
«¬ −1
order) 1 = 3
2 / 3º ª3 10
«
2 / 3»» «
0
−1/ 3»¼ «¬
0
2/3
−1/ 3
2/3
1º
ª2
1»» . Then AT A = «
¬0
1»¼
and
2
0º
» ª3/ 10
0» «
1/ 10
0 »¼ «¬
−1/ 10 º
»
3/ 10 »¼
0º
, and the eigenvalues of AT A are seen to be (in decreasing
3»¼
= 2. Associated unit eigenvectors may be computed:
ª0 º
ª 1º
= 3: « », = 2: « »
¬ 1¼
¬0¼
ª 0 1º
Thus one choice for V is V = «
» . The singular values of A are σ 1 = 3 and σ 2 = 2. Thus the
¬ 1 0¼
ª 3
0º
«
»
matrix Σ is Σ = « 0
2 » . Next compute
« 0
0 »»
¬«
¼
ª1/ 3 º
ª 1/ 2 º
«
»
«
»
1
u1 =
Av1 = «1/ 3 » , u 2 =
Av 2 = «
0»
σ1
σ2
«
»
« −1/ 2 »
«¬1/ 3 »¼
¬
¼
Since {u1 , u 2 } is not a basis for 3, we need a unit vector u3 that is orthogonal to both u1 and u 2 .
1
The vector u3 must satisfy the set of equations u1T x = 0 and uT2 x = 0. These are equivalent to the
linear equations
x1 + x2 + x3
x1 + 0 x2 − x3
=
=
ª 1/ 6 º
ª1/ 3
ª 1º
«
»
«
0
,so x = «« −2 »» ,and u3 = « −2 / 6 » . Therefore let U = «1/ 3
0
«
»
«
«¬ 1»¼
¬« 1/ 6 »¼
¬«1/ 3
1/ 2
0
−1/ 2
1/ 6 º
»
−2 / 6 » .
»
1/ 6 ¼»
Thus
ª1/ 3
«
A = U Σ V T = «1/ 3
«
¬«1/ 3
ª3
13. Let A = «
¬2
2
3
1/ 2
0
−1/ 2
ª3
2º
T
. Then A = « 2
«
−2 »¼
¬« 2
1/ 6 º ª 3
»«
−2 / 6 » « 0
»«
1/ 6 ¼» ¬« 0
0º
» ª0
2» «
1
0 »» ¬
¼
1º
0 »¼
2º
ª17
3»» , ATT AT = AAT = «
¬ 8
−2 ¼»
ATT AT are seen to be (in decreasing order)
be computed:
1
= 25 and
2
8º
, and the eigenvalues of
17 »¼
= 9. Associated unit eigenvectors may
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.4
• Solutions
ª1/ 2 º
ª −1/ 2 º
= 25 : «
», = 9: «
»
¬«1/ 2 ¼»
¬« 1/ 2 »¼
ª1/ 2
Thus one choice for V is V = «
¬«1/ 2
−1/ 2 º
T
» . The singular values of A are σ 1 = 25 = 5 and
1/ 2 ¼»
ª5
σ 2 = 9 = 3. Thus the matrix Σ is Σ = «« 0
«¬ 0
0º
3»» . Next compute
0 »¼
ª1/ 2 º
ª −1/ 18 º
«
»
«
»
1
u1 =
AT v1 = «1/ 2 » , u 2 =
AT v 2 = « 1/ 18 »
σ1
σ2
«
«
»
0 »»
«¬
«¬ −4 / 18 »¼
¼
Since {u1 , u 2 } is not a basis for 3, we need a unit vector u3 that is orthogonal to both u1 and u 2 .
1
The vector u3 must satisfy the set of equations u1T x = 0 and uT2 x = 0. These are equivalent to the
linear equations
ª −2 º
ª − 2 / 3º
x1 + x2 + 0 x3 = 0
«
»
, so x = « 2 » , and u 3 = «« 2 / 3»»
− x1 + x2 − 4 x3 = 0
«¬ 1»¼
«¬ 1/ 3»¼
ª1/ 2
«
Therefore let U = «1/ 2
«
0
«¬
−2 / 3º
»
2 / 3» . Thus
»
1/ 3»¼
−1/ 18
1/ 18
−4 / 18
ª1/ 2
«
AT = U Σ V T = «1/ 2
«
¬« 0
−1/ 18
1/ 18
−4 / 18
−2 / 3º ª 5
»
2 / 3 » ««0
»
1/ 3 ¼» «¬ 0
0º
ª 1/ 2
3 »» «
−1/ 2
0 »¼ «¬
1/ 2 º
»
1/ 2 »¼
An SVD for A is computed by taking transposes:
ª1/ 2
A=«
«¬1/ 2
−1/ 2 º ª5
»«
1/ 2 »¼ ¬ 0
0
3
ª 1/ 2 1/ 2
0º «
« −1/ 18 1/ 18
0 »¼ «
2/3
−2 / 3
¬«
ª2 / 5
14. From Exercise 7, A = U ΣV T with V = «
¬« 1/ 5
0º
»
−4 / 18 »
1/ 3»»
¼
−1/ 5 º
» . Since the first column of V is unit
2 / 5 ¼»
eigenvector associated with the greatest eigenvalue
vector at which || Ax || is maximized.
1
of AT A, so the first column of V is a unit
15. a. Since A has 2 nonzero singular values, rank A = 2.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
439
440
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
°
b. By Example 6, {u1 , u 2 } = ®
°
¯
for Nul A.
ª .40 º ª −.78º
« .37 » , « −.33»
«
» «
»
«¬ −.84 »¼ «¬ −.52 »¼
½
ª .58º
°
°«
»
¾ is a basis for Col A and {v3 } = ® « −.58»
°
° « .58»
¼
¿
¯¬
½
°
¾ is a basis
°
¿
16. a. Since A has 2 nonzero singular values, rank A = 2.
ª −.86 º ª −.11º ½
°
°
b. By Example 6, {u1 , u 2 } = ® «« .31»» , «« .68»» ¾ is a basis for Col A and
° « .41» « −.73» °
¼ ¬
¼¿
¯¬
ª .65 º ª −.34 º ½
°«
» «
»°
° .08 » « .42 » °
{v 3 , v 4 } = ® «
,
¾ is a basis for Nul A.
° « −.16 » « −.84 » °
° «¬« −.73»¼» «¬« −.08»¼» °
¯
¿
17. Let A = U ΣV T = U ΣV −1. Since A is square and invertible, rank A = n, and all of the entries on the
diagonal of Σ must be nonzero. So A−1 = (U ΣV −1 )−1 = V Σ−1 U −1 = V Σ−1 U T .
18. First note that the determinant of an orthogonal matrix is ±1, because 1 = det I = det U T U =
(det U T )(det U ) = (det U )2 . Suppose that A is square and A = U ΣV T . Then Σ is square, and
det A = (det U )(det Σ)(det V T ) = ±det Σ = ±σ1 …σ n .
19. Since U and V are orthogonal matrices,
AT A = (U ΣV T )T U ΣV T = V ΣTU TU ΣV T = V (ΣT Σ)V T = V (ΣT Σ)V −1
If σ 1,… , σ r are the diagonal entries in Σ, then ΣT Σ is a diagonal matrix with diagonal entries
2
2
σ 1 ,…,σ r and possibly some zeros. Thus V diagonalizes AT A and the columns of V are
eigenvectors of AT A by the Diagonalization Theorem in Section 5.3. Likewise
AAT = U ΣV T (U ΣV T )T = U ΣV T V ΣT U T = U (ΣΣT )U T = U (ΣΣT )U −1
so U diagonalizes AAT and the columns of U must be eigenvectors of AAT . Moreover, the
2
2
Diagonalization Theorem states that σ 1 ,…,σ r are the nonzero eigenvalues of AT A . Hence
σ 1,… , σ r are the nonzero singular values of A.
20. If A is positive definite, then A = PDPT , where P is an orthogonal matrix and D is a diagonal matrix.
The diagonal entries of D are positive because they are the eigenvalues of a positive definite matrix.
Since P is an orthogonal matrix, PPT = I and the square matrix PT is invertible. Moreover,
( PT )−1 = ( P−1 )−1 = P = ( PT )T , so PT is an orthogonal matrix. Thus the factorization A = PDPT has
the properties that make it a singular value decomposition.
21. Let A = U Σ V T . The matrix PU is orthogonal, because P and U are both orthogonal. (See Exercise
29 in Section 6.2). So the equation PA = ( PU ) ΣV T has the form required for a singular value
decomposition. By Exercise 19, the diagonal entries in Σ are the singular values of PA.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.4
22. The right singular vector v1 is an eigenvector for the largest eigenvector
in Section 7.3, the second largest eigenvalue
2
• Solutions
441
of AT A. By Theorem 7
1
is the maximum of xT ( AT A)x over all unit vectors
orthogonal to v1 . Since xT ( AT A)x = || Ax ||2 , the square root of 2 , which is the second largest
singular value of A, is the maximum of || Ax || over all unit vectors orthogonal to v1 .
23. From the proof of Theorem 10, U Σ = [σ1u1
of the product (U Σ)V
T
… σ r ur
0 … 0]. The column-row expansion
shows that
ª v1T º
«
»
T
T
A = (U Σ)V T = (U Σ) « # » = σ 1u1 v1 + …+ σ r u r v r
« T»
«¬ v n »¼
where r is the rank of A.
for i ≠ j
,
for i = j
0
24. From Exercise 23, AT = σ 1v1u1T +…+ σ r v r uTr . Then since uTi u j = ®
¯1
AT u j = (σ 1v1u1T + …+ σ r v r uTr )u j = (σ j v j uTj )u j = σ j v j (uTj u j ) = σ j v j
25. Consider the SVD for the standard matrix A of T, say A = U Σ V T . Let B = { v1 ,… , v n } and
n
and m constructed respectively from the columns of V and U. Since
C = {u1 ,… , u m } be bases for
the columns of V are orthogonal, V T v j = e j , where e j is the jth column of the n × n identity matrix.
To find the matrix of T relative to B and C, compute
T ( v j ) = Av j = U ΣV T v j = U Σe j = U σj e j = σ jUe j = σ j u j
so [T ( v j )]C = σj e j . Formula (4) in the discussion at the beginning of Section 5.4 shows that the
“diagonal” matrix Σ is the matrix of T relative to B and C.
ª −18
« 2
26. [M] Let A = «
« −14
«
«¬ −2
13
19
−4
−4
11
21
−12
4
4º
ª 528
« −392
»
12 »
T
. Then A A = «
« 224
8»
«
»
8»¼
«¬ −176
eigenvalues of AT A are found to be (in decreasing order)
4 = 0. Associated unit eigenvectors may be computed:
ª −.4 º
« .8»
« »,
1:
« −.2 »
« »
¬« .4 ¼»
ª.8º
«.4»
« »,
2:
«.4»
« »
¬«.2¼»
ª .4 º
« −.2 »
« »,
3:
« −.8»
« »
¬« .4 ¼»
−392
1092
224
−176
−176
536
192
−128
1
= 1600,
2
−176 º
536 »»
, and the
−128»
»
288»¼
= 400,
3
ª −.2º
« −.4»
« »
4:
« .4»
« »
¬« .8¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
= 100, and
442
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª −.4
« .8
Thus one choice for V is V = «
« −.2
«
¬« .4
.8
.4
.4
−.2
.4
.2
−.8
.4
−.2º
−.4»»
. The singular values of A are σ 1 = 40, σ 1 = 20,
.4»
»
.8¼»
ª 40
« 0
«
Σ
=
and
Thus
the
matrix
Σ
is
σ 3 = 10,
σ 4 = 0.
« 0
«
¬« 0
0
20
0
0
0
0
10
0
0º
0 »»
. Next compute
0»
»
0¼»
ª.5º
ª −.5º
«.5»
« .5»
1
1
u1 =
Av1 = « » , u 2 =
Av 2 = « » ,
«.5»
« −.5»
σ1
σ2
« »
« »
«¬.5»¼
«¬ .5»¼
ª −.5 º
« .5 »
1
u3 =
Av 3 = « »
« .5 »
σ3
« »
«¬ −.5 »¼
Because Av4 = 0, only three columns of U have been found so far. The last column of U can be found
by extending {u1, u2, u3} to an orthonormal basis for 4. The vector u4 must satisfy the set of
equations u1T x = 0, uT2 x = 0, and uT3 x = 0. These are equivalent to the linear equations
ª −1º
ª −.5º
x1 + x2 + x3 + x4 = 0
« −1»
« −.5»
− x1 + x2 − x3 + x4 = 0, so x = « » , and u 4 = « » .
« 1»
« .5»
− x1 + x2 + x3 − x4 = 0
« »
« »
¬« 1¼»
¬« .5»¼
ª.5
«.5
Therefore, let U = «
«.5
«
«¬.5
−.5
.5
−.5
.5
−.5
.5
.5
−.5
ª.5
«.5
A = U ΣV T = «
«.5
«
¬«.5
−.5
.5
−.5
.5
−.5
.5
.5
−.5
−.5º
−.5»»
. Thus
.5»
»
.5»¼
−.5º ª 40
−.5»» «« 0
.5» « 0
»«
.5¼» ¬« 0
0
20
0
0
0
0
10
0
0 º ª −.4
0 »» «« .8
0» « .4
»«
0 ¼» ¬« −.2
.8
.4
−.2
.4
−.2
−.4
−.8
.4
.4º
.2»»
.4»
»
.8¼»
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.4
ª 41 −32 −38
−4 º
« −32 118
−3
»
«
4»
T
. Then A A = « −38
−3 121
−1 −8
2
2»
«
10
»
« 14 −92
−2
4
4 −8»¼
«¬ −8
74 −52
eigenvalues of AT A are found to be (in decreasing order) 1 = 270.87,
4 = 18.55, and
5 = 0. Associated unit eigenvectors may be computed:
ª 6
« 2
27. [M] Let A = «
« 0
«
«¬ −1
ª −.10 º
« .61»
«
»
1 : « −.21» ,
«
»
« −.52 »
«¬ .55»¼
−8
7
−4
−5
ª −.39 º
« .29 »
«
»
2 : « .84 » ,
«
»
« −.14 »
«¬ −.19 »¼
14
−92
5
−6
ª −.74 º
« −.27 »
«
»
3 : « −.07 » ,
«
»
« .38»
«¬ .49 »¼
ª −.10
« .61
«
Thus one choice for V is V = « −.21
«
« −.52
«¬ .55
ª .41º
« −.50»
«
»
4 : « .45» ,
«
»
« −.23»
«¬ .58»¼
−.39
.29
.84
−.14
−.19
−.74
−.27
10
81
−72
2
• Solutions
443
−8º
74 »»
−52 » , and the
»
−72 »
100 »¼
= 147.85,
3
= 23.73,
ª −.36 º
« −.48»
«
»
5 : « −.19 »
«
»
« −.72 »
«¬ −.29»¼
−.36 º
−.48»»
.45 −.19 » . The nonzero singular values of A
−.07
»
.38 −.23 −.72 »
.49
.58 −.29 »¼
σ 4 = 4.31. Thus the matrix Σ is
.41
−.50
are σ 1 = 16.46, σ 1 = 12.16, σ 3 = 4.87, and
0
0
0 0º
ª16.46
«
0 12.16
0
0 0»»
. Next compute
Σ=«
«
0
0 4.87
0 0»
«
»
0
0
0 4.31 0¼»
¬«
ª −.57 º
ª −.65º
« .63»
« −.24»
1
1
«
»
»,
Av1 =
,u =
Av 2 = «
u1 =
« .07 » 2 σ 2
« −.63»
σ1
«
»
«
»
«¬ −.51»¼
«¬ .34»¼
ª −.42 º
ª .27 º
« −.68»
«
»
1
» , u 4 = 1 Av 4 = « −.29»
u3 =
Av 3 = «
« .53»
« −.56»
σ3
σ4
«
»
«
»
¬« −.29 ¼»
¬« −.73¼»
ª −.57
« .63
Since {u1 , u 2 , u 3 , u 4 } is a basis for 4, let U = «
« .07
«
«¬ −.51
−.65
−.24
−.42
−.68
−.63
.34
.53
−.29
.27 º
−.29»»
. Thus
−.56 »
»
−.73»¼
A = U ΣV T
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
444
CHAPTER 7
ª −.57
« .63
«
=
« .07
«
¬« −.51
−.65
−.24
−.63
.34
• Symmetric Matrices and Quadratic Forms
−.42
−.68
.53
−.29
.27 º ª16.46
0
−.29 »» ««
0
−.56 » «
»«
0
−.73¼» ¬«
ª 4
« −6
28. [M] Let A = «
« 7
«
¬ −1
0
9
−3
9
−5
2
10
4
0
12.16
0
0
0
0
4.87
0
ª −.10
0º «
−.39
0 »» «
« −.74
0» «
» .41
0 ¼» «
«¬ −.36
0
0
0
4.31
−7 º
ª102
« −91
»
9»
T
. Then A A = «
« 0
19 »
«
»
−1¼
¬108
−91
110
0
39
39
−16
206
246
.61
.29
−.21
.84
−.52
−.14
−.27
−.50
−.48
−.07
.45
−.19
.38
−.23
−.72
.55º
−.19 »»
.49 »
»
.58»
−.29 »¼
108º
−16 »»
, and the eigenvalues of
246 »
»
492 ¼
AT A are found to be (in decreasing order) 1 = 649.9059, 2 = 218.0033, 3 = 39.6345, and
4 = 2.4564. The singular values of A are thus σ 1 = 25.4933, σ 2 = 14.7649, σ 3 = 6.2956, and
σ 4 = 1.5673. The condition number σ 1 / σ 4 ≈ 6.266.
ª5
«6
«
29. [M] Let A = «7
«
«9
«¬ 8
3
4
5
6
5
eigenvalues of AT A
9º
ª 255 168 90 160
« 168 111 60 104
»
−8»
«
3 10
9 » . Then AT A = « 90
60 34
39
«
»
4 −9 −5»
« 160 104 39 415
«¬ 47
»
2 11
4¼
30
8 178
are found to be (in decreasing order) 1 = 672.589,
1
2
47 º
30 »»
8» , and the
»
178»
267 »¼
7
8
2
= 280.745,
−7
= 127.503, 4 = 1.163, and 5 = 1.428 ×10 . The singular values of A are thus σ 1 = 25.9343,
σ 2 = 16.7554, σ 3 = 11.2917, σ 4 = 1.07853, and σ 5 = .000377928. The condition number
σ 1 / σ 5 = 68, 622.
3
7.5
SOLUTIONS
Notes: The application presented here has turned out to be of interest to a wide variety of students,
including engineers. I cover this in Course Syllabus 3 described in the front mater of the text, but I only
have time to mention the idea briefly to my other classes.
ª19
1. The matrix of observations is X = «
¬12
22
6
6
9
3
15
2
13
20 º
and the sample mean is
5»¼
1 ª 72 º ª12º
M = « » = « » . The mean-deviation form B is obtained by subtracting M from each column of X,
6 ¬ 60 ¼ ¬10¼
8º
ª 7 10 −6 −9 −10
so B = «
. The sample covariance matrix is
5
3 −5»¼
¬ 2 −4 −1
S=
1
1 ª 430
BBT = «
6 −1
5 ¬ −135
−135º ª 86
=
80 »¼ «¬ −27
−27 º
16 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
7.5
• Solutions
445
7
3º
ª1 5 2 6
2. The matrix of observations is X = «
» and the sample mean is
¬3 11 6 8 15 11¼
1 ª 24 º ª 4 º
M = « » = « » . The mean-deviation form B is obtained by subtracting M from each column of X,
6 ¬ 54 ¼ ¬ 9 ¼
ª −3
so B = «
¬ −6
S=
1
2
−2
−3
2
−1
1
1 ª 28
BBT = «
6 −1
5 ¬ 40
3
6
−1º
. The sample covariance matrix is
2 »¼
40 º ª5.6
=
90 »¼ «¬ 8
8º
18»¼
3. The principal components of the data are the unit eigenvectors of the sample covariance matrix S.
ª 86 −27 º
are 1 = 95.2041 and
One computes that (in descending order) the eigenvalues of S = «
16 »¼
¬ −27
2
ª −2.93348º
» and
1
¬
¼
= 6.79593. One further computes that corresponding eigenvectors are v1 = «
ª.340892 º
v2 = «
» . These vectors may be normalized to find the principal components, which are
¬ 1 ¼
ª .946515º
u1 = «
» for
¬ −.322659 ¼
1
ª.322659 º
» for
¬.946515¼
= 95.2041 and u 2 = «
2
= 6.79593.
4. The principal components of the data are the unit eigenvectors of the sample covariance matrix S.
8º
ª5.6
One computes that (in descending order) the eigenvalues of S = «
» are 1 = 21.9213 and
¬ 8 18 ¼
2
ª.490158º
» and
¬ 1 ¼
= 1.67874. One further computes that corresponding eigenvectors are v1 = «
ª −2.04016 º
v2 = «
» . These vectors may be normalized to find the principal components, which are
1
¬
¼
ª −.897934 º
ª .44013º
for 1 = 21.9213 and u 2 = «
u1 = «
» for 2 = 1.67874.
»
¬ .44013¼
¬.897934 ¼
ª164.12
5. [M] The largest eigenvalue of S = « 32.73
«
«¬ 81.04
32.73
539.44
249.13
81.04 º
249.13»» is
189.11»¼
1
= 677.497, and the first
principal component of the data is the unit eigenvector corresponding to 1 , which is
ª.129554 º
u1 = «« .874423»» . The fraction of the total variance that is contained in this component is
«¬.467547 »¼
1
/ tr( S ) = 677.497 / (164.12 + 539.44 + 189.11) = .758956 so 75.8956% of the variance of the data is
contained in the first principal component.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
446
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª 29.64
6. [M] The largest eigenvalue of S = « 18.38
«
«¬ 5.00
18.38
20.82
14.06
5.00 º
14.06 »» is
29.21»¼
1
= 51.6957, and the first principal
ª.615525º
component of the data is the unit eigenvector corresponding to 1 , which is u1 = «.599424 » . Thus
«
»
«¬.511683»¼
one choice for the new variable is y1 = .615525 x1 + .599424 x2 + .511683 x3 . The fraction of the total
variance that is contained in this component is
1 / tr( S ) = 51.6957 /(29.64 + 20.82 + 29.21) = .648872, so 64.8872% of the variance of the data is
explained by y1 .
ª .946515º
» , one choice for the
¬ −.322659 ¼
new variable is y1 = .946515 x1 − .322659 x2 . The fraction of the total variance that is contained in this
component is 1 / tr( S ) = 95.2041/(86 + 16) = .933374, so 93.3374% of the variance of the data is
explained by y1 .
7. Since the unit eigenvector corresponding to
1
= 95.2041 is u1 = «
ª .44013º
» , one choice for the new
¬.897934 ¼
variable is y1 = .44013 x1 + .897934 x2 . The fraction of the total variance that is contained in this
component is 1 / tr( S ) = 21.9213 /(5.6 + 18) = .928869, so 92.8869% of the variance of the data is
explained by y1 .
8. Since the unit eigenvector corresponding to
ª5
9. The largest eigenvalue of S = « 2
«
«¬ 0
2
6
2
0º
2 »» is
7 »¼
1
= 21.9213 is u1 = «
1
= 9, and the first principal component of the data is
ª 1/ 3º
the unit eigenvector corresponding to 1 , which is u1 = « 2 / 3» . Thus one choice for y is
«
»
«¬ 2 / 3»¼
y = (1/ 3) x1 + (2 / 3) x2 + (2 / 3) x3 , and the variance of y is 1 = 9.
ª5
10. [M] The largest eigenvalue of S = « 4
«
«¬ 2
4
11
4
2º
4 »» is
5 »¼
1
= 15, and the first principal component of the
ª 1/ 6 º
«
»
data is the unit eigenvector corresponding to 1 , which is u1 = « 2 / 6 » . Thus one choice for y is
«
»
«¬ 1/ 6 »¼
y = (1/ 6) x1 + (2/ 6) x2 + (1/ 6) x3 , and the variance of y is 1 = 15.
11. a. If w is the vector in N with a 1 in each position, then [ X1
the X k are in mean-deviation form. Then
… X N ] w = X1 +…+ X N = 0 since
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 7
[ Y1
YN ] w = ª¬ PT X1
…
…
PT X N º¼ w = PT [ X1
…
• Supplementary Exercises
447
X N ] w = PT 0 = 0
Thus Y1 + … + YN = 0, and the Yk are in mean-deviation form.
b. By part a., the covariance matrix S Y of Y1 ,… , YN is
1
[ Y1 … YN ][ Y1 … YN ]T
N −1
1
=
PT [ X1 … X N ] ( PT [ X1 …
N −1
SY =
§ 1
= PT ¨
[ X1 … X N ][ X1
© N −1
since the X k are in mean-deviation form.
…
X N ] )T
T ·
X N ] ¸ P = PT SP
¹
12. By Exercise 11, the change of variables X = PY changes the covariance matrix S of X into the
covariance matrix P T SP of Y. The total variance of the data as described by Y is tr( PT SP).
However, since P T SP is similar to S, they have the same trace (by Exercise 25 in Section 5.4). Thus
the total variance of the data is unchanged by the change of variables X = PY.
ˆ = X − M. Let B = ª X
ˆ
13. Let M be the sample mean for the data, and let X
k
k
¬ 1
…
ˆ º be the
X
N¼
matrix of observations in mean-deviation form. By the row-column expansion of BBT , the sample
covariance matrix is
S=
=
1
BBT
N −1
1
ˆ
ªX
1
N −1 ¬
=
…
ˆT º
ªX
1
« »
ˆ
º
XN ¼ « # »
«X
ˆT »
¬ N¼
1 N ˆ ˆT
1 N
Xk Xk =
( Xk − M )( Xk − M )T
¦
¦
N − 1 k =1
N −1 k =1
Chapter 7 SUPPLEMENTARY EXERCISES
1. a. True. This is just part of Theorem 2 in Section 7.1. The proof appears just before the statement
of the theorem.
ª 0 −1º
b. False. A counterexample is A = «
.
0 »¼
¬1
c. True. This is proved in the first part of the proof of Theorem 6 in Section 7.3. It is also a
consequence of Theorem 7 in Section 6.2.
d. False. The principal axes of xT Ax are the columns of any orthogonal matrix P that
diagonalizes A. Note: When A has an eigenvalue whose eigenspace has dimension greater than
1, the principal axes are not uniquely determined.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
448
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª1 −1º
e. False. A counterexample is P = «
. The columns here are orthogonal but not
1»¼
¬1
orthonormal.
f. False. See Example 6 in Section 7.2.
0º
ª1 º
ª2
g. False. A counterexample is A = «
and x = « » . Then x T A x = 2 > 0 , but xT Ax is an
»
¬0 ¼
¬ 0 − 3¼
indefinite quadratic form.
h. True. This is basically the Principal Axes Theorem from Section 7.2. Any quadratic form can be
written as xT Ax for some symmetric matrix A.
i. False. See Example 3 in Section 7.3.
j. False. The maximum value must be computed over the set of unit vectors. Without a restriction
on the norm of x, the values of xT Ax can be made as large as desired.
k. False. Any orthogonal change of variable x = Py changes a positive definite quadratic form into
another positive definite quadratic form. Proof: By Theorem 5 of Section 7.2., the classification
of a quadratic form is determined by the eigenvalues of the matrix of the form. Given a form
xT Ax, the matrix of the new quadratic form is P −1 AP , which is similar to A and thus has the
same eigenvalues as A.
l. False. The term “definite eigenvalue” is undefined and therefore meaningless.
m. True. If x = Py, then xT Ax = ( Py )T A( Py) = yT PT APy = yT P −1 APy .
ª1 −1º
n. False. A counterexample is U = «
» . The columns of U must be orthonormal to make
¬1 −1¼
UU T x the orthogonal projection of x onto Col U.
o. True. This follows from the discussion in Example 2 of Section 7.4., which refers to a proof
given in Example 1.
p. True. Theorem 10 in Section 7.4 writes the decomposition in the form U ΣV T , where U and V
are orthogonal matrices. In this case, V T is also an orthogonal matrix. Proof: Since V is
orthogonal, V is invertible and V −1 = V T . Then (V T )−1 = (V −1 )T = (V T )T , and since V is square
and invertible, V T is an orthogonal matrix.
ª2
q. False. A counterexample is A = «
¬0
values of AT A are 4 and 1.
0º
. The singular values of A are 2 and 1, but the singular
1»¼
2. a. Each term in the expansion of A is symmetric by Exercise 35 in Section 7.1. The fact that
( B + C )T = BT + CT implies that any sum of symmetric matrices is symmetric, so A is
symmetric.
b. Since u1T u1 = 1 and uTj u1 = 0 for j ≠ 1,
Au1 = ( 1u1u1T )u1 +…+ ( nunuTn )u1 = 1u1 (u1T u1 ) +…+
Since u1 ≠ 0 ,
1
T
nu n (u n u1 )
= 1u1
is an eigenvalue of A. A similar argument shows that
j
is an eigenvalue of A
for j = 2, …, n.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 7
• Supplementary Exercises
449
3. If rank A = r, then dim Nul A = n – r by the Rank Theorem. So 0 is an eigenvalue of A with
multiplicity n – r, and of the n terms in the spectral decomposition of A exactly n – r are zero. The
remaining r terms (which correspond to nonzero eigenvalues) are all rank 1 matrices, as mentioned
in the discussion of the spectral decomposition.
4. a. By Theorem 3 in Section 6.1, (Col A)⊥ = Nul AT = Nul A since AT = A.
b. Let y be in
n
. By the Orthogonal Decomposition Theorem in Section 6.3, y = ŷ + z, where ŷ is
in Col A and z is in (Col A)⊥ . By part a., z is in Nul A.
5. If Av = λv for some nonzero λ, then v =
combination of the columns of A.
−1
Av = A(
−1
v), which shows that v is a linear
6. Because A is symmetric, there is an orthonormal eigenvector basis {u1 ,… , u n } for n. Let r = rank A.
If r = 0, then A = O and the decomposition of Exercise 4(b) is y = 0 + y for each y in n; if r = n then
the decomposition is y = y + 0 for each y in n.
Assume that 0 < r < n. Then dim Nul A = n – r by the Rank Theorem, and so 0 is an eigenvalue of A
with multiplicity n – r. Hence there are r nonzero eigenvalues, counted according to their
multiplicities. Renumber the eigenvector basis if necessary so that u1 ,… , u r are the eigenvectors
corresponding to the nonzero eigenvalues. By Exercise 5, u1 ,… , u r are in Col A. Also, u r +1 ,… , u n
are in Nul A because these vectors are eigenvectors corresponding to the eigenvalue 0. For y in n,
there are scalars c1 , … , cn such that
y = c1u1 + … + cr u r + cr +1u r +1 + … + cn u n
Z
yˆ
This provides the decomposition in Exercise 4(b).
7. If A = RT R and R is invertible, then A is positive definite by Exercise 25 in Section 7.2.
Conversely, suppose that A is positive definite. Then by Exercise 26 in Section 7.2, A = BT B for
some positive definite matrix B. Since the eigenvalues of B are positive, 0 is not an eigenvalue of B
and B is invertible. Thus the columns of B are linearly independent. By Theorem 12 in Section 6.4, B
= QR for some n × n matrix Q with orthonormal columns and some upper triangular matrix R with
positive entries on its diagonal. Since Q is a square matrix, QT Q = I , and
A = BT B = (QR)T (QR) = RT QT QR = RT R
and R has the required properties.
8. Suppose that A is positive definite, and consider a Cholesky factorization of A = RT R with R upper
triangular and having positive entries on its diagonal. Let D be the diagonal matrix whose diagonal
entries are the entries on the diagonal of R. Since right-multiplication by a diagonal matrix scales the
columns of the matrix on its left, the matrix L = RT D −1 is lower triangular with 1’s on its diagonal.
If U = DR, then A = R T D −1 DR = LU .
9. If A is an m × n matrix and x is in
, then xT AT Ax = ( Ax)T ( Ax) = || Ax ||2 ≥ 0. Thus AT A is positive
n
semidefinite. By Exercise 22 in Section 6.5, rank AT A = rank A.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
450
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
10. If rank G = r, then dim Nul G = n – r by the Rank Theorem. Hence 0 is an eigenvalue of G with
multiplicity n – r, and the spectral decomposition of G is
G = 1u1u1T +…+
Also
1 ,… ,
G=
(
r
T
r ur ur
are positive because G is positive semidefinite. Thus
1 u1
)(
T
1 u1
) +…+ (
r ur
)(
r ur
T
)
By the column-row expansion of a matrix product, G = BB T where B is the n × r matrix
T
T
º
B = ª¬ 1 u1 …
r u r ¼ . Finally, G = A A for A = B .
11. Let A = U ΣV T be a singular value decomposition of A. Since U is orthogonal, U T U = I and
A = U ΣU TUV T = PQ where P = U ΣU T = U ΣU −1 and Q = UV T . Since Σ is symmetric, P is
symmetric, and P has nonnegative eigenvalues because it is similar to Σ, which is diagonal with
nonnegative diagonal entries. Thus P is positive semidefinite. The matrix Q is orthogonal since it is
the product of orthogonal matrices.
12. a. Because the columns of Vr are orthonormal,
AA+ y = (U r DVrT )(Vr D−1U rT )y = (U r DD−1U rT )y = U rU rT y
Since U rU rT y is the orthogonal projection of y onto Col U r by Theorem 10 in Section 6.3, and
since Col U r = Col A by (5) in Example 6 of Section 7.4, AA+ y is the orthogonal projection of
y onto Col A.
b. Because the columns of U r are orthonormal,
A+ Ax = (Vr D−1U rT )(U r DVrT )x = (Vr D−1DVrT )x = VrVrT x
Since VrVrT x is the orthogonal projection of x onto Col Vr by Theorem 10 in Section 6.3, and
since Col Vr = Row A by (8) in Example 6 of Section 7.4, A+ Ax is the orthogonal projection of
x onto Row A.
c. Using the reduced singular value decomposition, the definition of A+ , and the associativity of
matrix multiplication gives:
AA+ A = (U r DVrT )(Vr D−1U rT )(U r DVrT ) = (U r DD−1U rT )(U r DVrT )
= U r DD−1DVrT = U r DVrT = A
A+ AA+ = (Vr D−1U rT )(U r DVrT )(Vr D−1U rT ) = (Vr D−1DVrT )(Vr D−1U rT )
= Vr D−1DD−1U rT = Vr D−1U rT = A+
13. a. If b = Ax, then x + = A + b = A + Ax. By Exercise 12(a), x+ is the orthogonal projection of x onto
Row A.
b. From part (a) and Exercise 12(c), Ax+ = A( A+ Ax) = ( AA+ A)x = Ax = b.
c. Let Au = b. Since x+ is the orthogonal projection of x onto Row A, the Pythagorean Theorem
shows that || u ||2 = || x+ ||2 + || u − x+ ||2 ≥ || x+ ||2 , with equality only if u = x + .
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Chapter 7
• Supplementary Exercises
451
14. The least-squares solutions of Ax = b are precisely the solutions of Ax = bˆ , where b̂ is the
orthogonal projection of b onto Col A. From Exercise 13, the minimum length solution of Ax = b̂ is
A+ bˆ , so A + bˆ is the minimum length least-squares solution of Ax = b. However, bˆ = AA + b by
Exercise 12(a) and hence A + bˆ = A + AA + b = Α + b by Exercise 12(c). Thus A + b is the minimum
length least-squares solution of Ax = b.
15. [M] The reduced SVD of A is A = U r DVrT , where
ª .966641
«.185205
Ur = «
«.125107
«
¬«.125107
.253758
−.786338
−.398296
−.398296
ª −.313388
« −.313388
«
and Vr = « −.633380
«
« .633380
«¬ .035148
−.034804 º
ª9.84443
−.589382 »»
, D = ««
0
»
.570709
«
0
»
¬
.570709 ¼»
.009549
.009549
.023005
−.023005
.999379
0
2.62466
0
0º
0 »» ,
1.09467 »¼
.633795º
.633795»»
−.313529 »
»
.313529 »
.002322 »¼
So the pseudoinverse A+ = Vr D −1U rT may be calculated, as well as the solution xˆ = A+ b for the
system Ax = b:
ª −.05
«
« −.05
A+ = « −.05
«
« .05
«¬ .10
−.35
−.35
.325
.325
.15
−.15
−.30
−.175
.175
−.150
.325º
ª .7 º
»
« »
.325»
« .7 »
−.175» , xˆ = « −.8»
»
« »
.175»
« .8»
»
«¬ .6 »¼
−.150 ¼
Row reducing the augmented matrix for the system AT z = xˆ shows that this system has a solution, so
ª 0 º ª −1º ½
°« » « »°
°° « 0 » « 1» °°
T
x̂ is in Col A = Row A. A basis for Nul A is {a1 , a 2 } = ® «1 » , « 0 » ¾ , and an arbitrary element of
° «1 » « 0 » °
°« » « »°
°¯ «¬ 0 »¼ «¬ 0 »¼ °¿
Nul A is u = ca1 + da 2 . One computes that || xˆ || = 131/50 , while || xˆ + u ||= (131/50) + 2c 2 + 2d 2 .
Thus if u ≠ 0, || x̂ || < || x̂ + u ||, which confirms that x̂ is the minimum length solution to Ax = b.
16. [M] The reduced SVD of A is A = U r DVrT , where
ª −.337977
« .591763
Ur = «
« −.231428
«
«¬ −.694283
.936307
.290230
−.062526
−.187578
.095396 º
ª12.9536
−.752053»»
, D = ««
0
−.206232 »
«¬
0
»
−.618696 »¼
0
1.44553
0
0º
0 »» ,
.337763»¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
452
CHAPTER 7
• Symmetric Matrices and Quadratic Forms
ª −.690099
«
0
«
and Vr = « .341800
«
« .637916
«¬
0
.721920
0
.387156
.573534
0
.050939 º
0 »»
−.856320 »
»
.513928»
0 »¼
So the pseudoinverse A+ = Vr D −1U rT may be calculated, as well as the solution xˆ = A+ b for the
system Ax = b:
ª.5
«0
«
A+ = « 0
«
«.5
¬« 0
0
0
−.05
0
2
−1
0
.5
−.35
0
−.15º
ª 2.3º
»
«
»
0»
« 0»
1.5» , xˆ = « 5.0 »
»
« »
−1.05»
« −.9 »
0 ¼»
¬« 0 »¼
Row reducing the augmented matrix for the system AT z = xˆ shows that this system has a solution, so
ª0º ª0º ½
°« » « »°
°° «1 » « 0 » °°
T
x̂ is in Col A = Row A . A basis for Nul A is {a1 , a 2 } = ® « 0 » , « 0 » ¾ , and an arbitrary element of
° «0» «0» °
°« » « »°
°¯ «¬ 0 »¼ «¬1 »¼ ¿°
Nul A is u = ca1 + da 2 . One computes that || xˆ || = 311/10 , while || xˆ + u || = (311/10) + c 2 + d 2 .
Thus if u ≠ 0, || x̂ || < || x̂ + u ||, which confirms that x̂ is the minimum length solution to Ax = b.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
!
!
!
"#$!%$&'$()*!&+!
,$-(&)!./0-$1!
!
8.1
SOLUTIONS
Notes. This section introduces a special kinds of linear combination used to describe the sets created
when a subspace is shifted away from the origin. An affine combination is a linear combination in which
the coefficients sum to one. Theorems 1, 3, and 4 connect affine combinations directly to linear
combinations. There are several approaches to solving many of the exercises in this section, and some of
the alternatives are presented here.
1.
ª 1º
ª −2 º
ª 0º
ª 3º
ª5º
v1 = « » , v 2 = « » , v3 = « » 4, v 4 = « » , y = « »
¬ 2¼
¬ 2¼
¬ 4¼
¬7 ¼
¬ 3¼
ª −3 º
ª −1º
ª 2º
ª4º
v 2 − v1 = « » , v 3 − v1 = « » , v 4 − v1 = « » , y − v1 = « »
¬ 0¼
¬ 2¼
¬ 5¼
¬ 1¼
Solve c2(v2 − v1) + c3(v3 − v1) + c4(v4 − v1) = y − v1 by row reducing the augmented matrix.
4 º ª −3 0 4.5 4.5º ª 1 0 −1.5 −1.5º
ª −3 −1 2 4 º ª −3 −1 2
« 0 2 5 1» « 0
1 2.5 .5»¼ «¬ 0 1 2.5 .5»¼ «¬0 1 2.5
.5»¼
¬
¼ ¬
The general solution is c2 = 1.5c4 −1.5, c3 = −2.5c4 + .5, with c4 free. When c4 = 0,
y − v1 = −1.5(v2 − v1) + .5(v3 − v1) and y = 2v1 − 1.5v2 + .5v3
If c4 = 1, then c2 = 0 and
y − v1 = −2(v3 − v1) + 1(v4 − v1) and y = 2v1 − 2v3 + v4
If c4 = 3, then
y − v1 = 3(v2 − v1) − 7(v3 − v1) + 3(v4 − v1) and y = 2v1 + 3v2 − 7v3 + 3v4
Of course, many other answers are possible. Note that in all cases, the weights in the linear
combination sum to one.
2.
ª −2 º
ª2º
ª 4º
ª1º
ª −1º
ª 3º
ª 5º
v1 = « » , v 2 = « » , v 3 = « » , y = « » , so v 2 − v1 = « » , v 3 − v1 = « » , and y − v1 = « »
¬1¼
¬ 2¼
¬ 2¼
¬7 ¼
¬ 1¼
¬ 1¼
¬ 6¼
Solve c2(v2 – v1) + c3(v3 – v1) = y – v1 by row reducing the augmented matrix:
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley"!
453!
454
CHAPTER 8
ª −2
« 1
¬
2
1
• The Geometry of Vector Spaces!
4º
ª 1 −1 −2 º
ª 1 0 2º
~ «
~ «
»
»
»
6¼
2
8¼
¬ 0
¬0 1 4¼
The general solution is c2 = 2 and c3 = 4, so y – v1 = 2(v2 – v1) + 4(v3 – v1) and
y = –5v1 + 2v2 + 4v3. The weights sum to one, so this is an affine sum.
3. Row reduce the augmented matrix [v2-v1 v3-v1 y-v1] to find a solution for writing y-v1 in terms of
v2-v1 and v3-v2. Then solve for y to get y = −3v1 + 2v2 + 2v3. The weights sum to one, so this is an
affine sum.
4. Row reduce the augmented matrix [v2-v1 v3-v1 y-v1] to find a solution for writing y-v1 in terms of
v2-v1 and v3-v2. Then solve for y to get y = 2.6v1 − .4v2 − 1.2v3. The weights sum to one, so this is
an affine sum.
5. Since {b1, b2, b3} is an orthogonal basis, use Theorem 5 from Section 6.2 to write
pj=
p j 2; n< v 2 = −2 < 2; n< v 3 = 2. Hence v2 is on the same side of H
as 0, v1 is on the other side, and v3 is in H.
12. Let H = [ f : d ], where f (x1, x2, x3) = 3x1 + x2 − 2x3. f (a1) = − 5, f (a2) = 4. f (a3) =3, f (b1) =7,
f (b 2) = 4, and f (b 3) = 6. Choose d = 4 so that all the points in A are in or on one side of H and all
the points in B are in or on the other side of H. There is no hyperplane parallel to H that strictly
separates A and B because both sets have a point at which f takes on the value of 4. There may be
(and in fact is) a hyperplane that is not parallel to H that strictly separates A and B.
13. H1 = {x : n1 ⋅ x = d1} and H2 = {x : n 2 ⋅ x = d2}. Since p1 ∈ H1, d1 = n1 ⋅ p1 = 4. Similarly,
d2 = n 2 ⋅ p2 = 22. Solve the simultaneous system [1 2 4 2]x = 4 and [2 3 1 5]x = 22:
ª 1
« 2
¬
2
3
4
1
2 4º ª
~
5 22»¼ «¬
1
0
0 −10
1
7
4 32º
−1 −14»¼
The general solution provides one set of vectors, p, v1, and v2. Other choices are possible.
ª 32 º
ª 10 º
ª − 4º
ª 32º
ª 10 º
ª − 4º
« −14»
« −7 »
« »
« −14 »
« −7 »
« 1»
» + x « » + x « » = p + x v + x v , where p = «
» , v = « » , v = « 1»
x=«
3 1
4 2
« 0 » 3 « 1» 4 « 0 »
« 0» 1 « 1» 2 « 0 »
«
»
« »
« »
«
»
« »
« »
¬ 0¼
¬ 0¼
¬ 1¼
¬ 0¼
¬ 0¼
¬ 1¼
Then H1 ∩ H2 = {x : x = p + x3v1 + x4v2}.
14. Since each of F1 and F2 can be described as the solution sets of A1x=b1 and A2x=b2 respectively,
ªA º
ªb º
where A1 and A2 have rank 2, their intersection is described as the solution set to « 1 » x = « 1 » .
¬ A2 ¼
¬b 2 ¼
§ ª A1 º ·
» ¸ ≤ 4 , the solution set will have dimensions 6 − 2=4,6 − 3=3, or 6 − 4=2.
© ¬ A2 ¼ ¹
Since 2 ≤ rank ¨ «
15. f (x1, x2, x3) =Ax= x1 – 3x2 + 4x3 – 2x4 and d =b= 5
16. f (x1, x2, x3) = Ax= 2x1 + 5x2 – 3x3 + 6x5 and d = b= 0
17. Since by Theorem 3 in Section 6.1, Row B=(Nul B) ⊥ , choose a nonzero vector n ∈ Nul B . For
ª1º
« »
example take n= −2 . Then f (x1, x2, x3) = x1 – 2x2 + x3 and d = 0
« »
«¬ 1 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
470
CHAPTER 8
• The Geometry of Vector Spaces!
18. Since by Theorem 3, Section 6.1, Row B=(Nul B) ⊥ , choose a nonzero vector n ∈ Nul B . For
ª −11º
«
»
example take n= 4 . Then f (x1, x2, x3) = –11x1 + 4x2 + x3 and d = 0
«
»
«¬ 1 »¼
19. Theorem 3 in Section 6.1 says that (Col B)⊥ = Nul BT. Since the two columns of B are clearly linear
independent, the rank of B is 2, as is the rank of BT. So dim Nul BT = 1, by the Rank Theorem, since
there are three columns in BT. This means that Nul BT is one-dimensional and any nonzero vector n
in Nul BT will be orthogonal to H and can be used as its normal vector. Solve the linear system
BTx = 0 by row reduction to find a basis for Nul BT:
ª 1
« 0
¬
!
4 −7
2 −6
0º
ª 1
~ «
»
0¼
¬ 0
0 5
1 −3
ª −5º
0º
« »
n = « 3»
0»¼
«¬ 1»¼
Now, let f (x1, x2, x3) = –5x1 + 3x2 + x3. Since the hyperplane H is a subspace, it goes through the
origin and d must be 0.
The solution is easy to check by evaluating f at each of the columns of B.
20. Since by Theorem 3, Section 6.1, Col B=(Nul BT) ⊥ , choose a nonzero vector n ∈ Nul BT . For
ª −6 º
« »
example take n= 2 . Then f (x1, x2, x3) = – 6x1 + 2x2 + x3 and d = 0
« »
«¬ 1 »¼
21. a. False. A linear functional goes from n to . See the definition at the beginning of this section.
b. False. See the discussion of (1) and (4). There is a 1×n matrix A such that f (x) = Ax for all x in
n. Equivalently, there is a point n in n such that f (x) = n ⋅ x for all x in n.
c. True. See the comments after the definition of strictly separate.
d. False. See the sets in Figure 4.
22. a. True. See the statement after (3).
b. False. The vector n must be nonzero. If n = 0, then the given set is empty if d ≠ 0 and the set
is all of n if d = 0.
c. False. Theorem 12 requires that the sets A and B be convex. For example, A could be the
boundary of a circle and B could be the center of the circle.
d. False. Some other hyperplane might strictly separate them. See the caution at the end of
Example 8.
ª3º
».
¬ −2 ¼
23. Notice that the side of the triangle closest to p is v 2 v 3 . A vector orthogonal to v 2 v 3 .is n= «
Take f (x1, x2) = 3x1 – 2x2 . Then f (v2) = f (v3)=9 and f (p)=10 so any d satisfying 9 < d < 10 will
work. There are other possible answers.
!
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
8.4
• Solutions
471!
ª −2 º
».
¬3¼
24. Notice that the side of the triangle closest to p is v1 v 3 A vector orthogonal to v1 v 3 is n= «
Take f (x1, x2) = − 2x1 +3x2 . Then f (v1) = f (v3)=4 and f (p)=5 so any d satisfying 4 < d < 5 will
work. There are other possible answers.
25. Let L be the line segment from the center of B(0, 3) to the center of B(p, 1). This is on the line
through the origin in the direction of p. The length of L is (42 + 12)1/2 4.1231. This exceeds the
sum of the radii of the two disks, so the disks do not touch. If the disks did touch, there would be no
hyperplane (line) strictly separating them, but the line orthogonal to L through the point of tangency
would (weakly) separate them. Since the disks are separated slightly, the hyperplane need not be
exactly perpendicular to L, but the easiest one to find is a hyperplane H whose normal vector is p.
So define f by f (x) = p ⋅ x.
To find d, evaluate f at any point on L that is between the two disks. If the disks were tangent, that
point would be three-fourths of the distance between their centers, since the radii are 3 and 1. Since
the disks are slightly separated, the distance is about 4.1231. Three-fourths of this distance is greater
than 3, and one-fourth of this distance is greater than 1. A suitable value of d is f (q), where q =
(.25)0 + (.75)p = (3, .75). So d = p ⋅ q = 4(3) + 1(.75) = 12.75.
26. The normal to the separating hyperplane has the direction of the line segment between p and q. So,
ª 4º
let n = p − q = « » . The distance between p and q is 20 , which is more than the sum of the radii
¬−2¼
of the two balls. The large ball has center q. A point three-fourths of the distance from q to p will be
greater than 3 units from q and greater than 1 unit from p. This point is
ª6º
ª2º ª5.0º
x = .75p + .25q = .75 « » + .25 « » = « »
¬ 1¼
¬ 3¼ ¬1.5¼
°ª x º
½°
Compute n ⋅ x = 17. The desired hyperplane is ®« » : 4 x − 2 y = 17 ¾ .
°¯¬ y ¼
°¿
27. Exercise 2(a) in Section 8.3 gives one possibility. Or let S = {(x, y) : x2y 2 = 1 and y > 0}. Then
conv S is the upper (open) half-plane.
28. One possibility is B = {(x, y) : x2y 2 = 1 and y > 0} and A = {(x, y) : | x | ≤ 1 and y = 0}.
29. Let x, y ∈ B( p, δ ) and suppose z = (1 − t) x + t y, where 0 ≤ t ≤ 1. Then
||z − p|| = || [(1 − t) x + t y] − p|| = ||(1 − t)(x − p) + t (y − p)||
≤ (1 − t) ||x − p|| + t ||y − p|| < (1 − t) δ + tδ = δ
where the first inequality comes from the Triangle Inequality (Theorem 17 in Section 6.7) and the
second inequality follows from x, y ∈ B( p, δ ). It follows that z ∈ B( p, δ ) and B( p, δ ) is convex.
30. Let S be a bounded set. Then there exists a δ > 0 such that S B(0, δ ). But B(0, δ ) is
convex by Exercise 29, so Theorem 9 in Section 8.3 (or Exercise 17 in Section 8.3) implies that
conv S B( p, δ ) and conv S is bounded.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
472
CHAPTER 8
8.5
• The Geometry of Vector Spaces!
SOLUTIONS
Notes: A polytope is the convex hull of a finite number of points. Polytopes and simplices are important
in linear programming, which has numerous applications in engineering design and business
management. The behavior of functions on polytopes is studied in this section.
1. Evaluate each linear functional at each of the three extreme points of S. Then select the extreme
point(s) that give the maximum value of the functional.
a. f (p1) = 1, f (p2) = –1, and f (p3) = –3, so m = 1 at p1.
b. f (p1) = 1, f (p2) = 5, and f (p3) = 1, so m = 5 at p2.
c. f (p1) = –3, f (p2) = –3, and f (p3) = 5, so m = 5 at p3"!
!"# Evaluate each linear functional at each of the three extreme points of S. Then select the point(s) that
give the maximum value of the functional.
a. f (p1) = − 1, f (p2) = 3, and f (p3) = 3, so m = 3 on the set conv {p2, p3}.
b. f (p1) = 1, f (p2) = 1, and f (p3) = − 1, so m = 1 on the set conv {p1, p2}.
c. f (p1) = –1, f (p2) = –3, and f (p3) = 0, so m = 0 at p3"!
3. Evaluate each linear functional at each of the three extreme points of S. Then select the point(s) that
give the minimum value of the functional.
a. f (p1) = 1, f (p2) = –1, and f (p3) = –3, so m = –3 at the point p3
b. f (p1) = 1, f (p2) = 5, and f (p3) = 1, so m = 1 on the set conv {p1, p3}.
c. f (p1) = –3, f (p2) = –3, and f (p3) = 5, so m = –3 on the set conv {p1, p2}.
4. Evaluate each linear functional at each of the three extreme points of S. Then select the point(s) that
give the maximum value of the functional.
a. f (p1) = − 1, f (p2) = 3, and f (p3) = 3, so m = –1 at the point p1.
b. .f (p1) = 1, f (p2) = 1, and f (p3) = − 1, so m = –1 at the point p3.
c. f (p1) = –1, f (p2) = –3, and f (p3) = 0, so m = –3 at the point p2.
5. The two inequalities are (a) x1 + 2x2 ≤ 10 and (b) 3x1 + x2 ≤ 15. Line (a) goes from (0,5) to (10,0).
Line (b) goes from (0,15) to (5,0). One vertex is (0,0). The x1-intercepts (when x2 = 0) are 10 and 5,
so (5,0) is a vertex. The x2-intercepts (when x1 = 0) are 5 and 15, so (0,5) is a vertex. The two lines
°ª0 º ª 5º ª 4º ª0º ½°
intersect at (4,3) so (4,3) is a vertex. The minimal representation is ®« » , « » , « » , « » ¾
°¯¬0 ¼ ¬0 ¼ ¬ 3¼ ¬ 5¼ °¿
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
8.5
• Solutions
473!
6. The two inequalities are (a) 2x1 + 3x2 ≤ 18 and (b) 4x1 + x2 ≤ 16. Line (a) goes from (0,6) to (9,0).
Line (b) goes from (0,16) to (4,0). One vertex is (0,0). The x1-intercepts (when x2 = 0) are 9 and 4, so
(4,0) is a vertex. The x2-intercepts (when x1 = 0) are 6 and 16, so (0,6) is a vertex. The two lines
°ª0 º ª 4 º ª 3º ª 0º °½
intersect at (3,4) so (3,4) is a vertex. The minimal representation is ®« » , « » , « » , « » ¾
°¯¬0 ¼ ¬ 0 ¼ ¬ 4¼ ¬ 6¼ °¿
7. The three inequalities are (a) x1 + 3x2 ≤ 18, (b) x1 + x2 ≤ 10, and (c) 4x1 + x2 ≤ 28. Line (a) goes from
(0,6) to (18,0). Line (b) goes from (0,10) to (10,0). And line (c) goes from (0,28) to (7,0). One
vertex is (0,0). The x1-intercepts (when x2 = 0) are 18, 10, and 7, so (7,0) is a vertex. The x2intercepts (when x1 = 0) are 6, 10, and 28, so (0,6) is a vertex. All three lines go through (6,4), so
°ª0 º ª7 º ª 6º ª 0º ½°
(6,4) is a vertex. The minimal representation is ®« » , « » , « » , « » ¾ .
°¯¬0 ¼ ¬ 0 ¼ ¬ 4¼ ¬ 6¼ °¿
8. The three inequalities are (a) 2x1 + x2 ≤ 8, (b) x1 + x2 ≤ 6, and (c) x1 + 2x2 ≤ 7. Line (a) goes from
(0,8) to (4,0). Line (b) goes from (0,6) to (6,0). And line (c) goes from (0,3.5) to (7,0). One vertex is
(0,0). The x1-intercepts (when x2 = 0) are 4, 6, and 7, so (4,0) is a vertex. The x2-intercepts (when x1
= 0) are 8, 6, and 3.5, so (0,3.5) is a vertex. All three lines go through (3,2), so (3,2) is a vertex. The
° ª0º ª 4º ª 3º ª 0º °½
minimal representation is ® « » , « » , « » , « » ¾
°¯ ¬0¼ ¬ 0¼ ¬ 2¼ ¬3.5¼ °¿
9. The origin is an extreme point, but it is not a vertex. It is an
extreme point since it is not in the interior of any line segment
that lies in S. It is not a vertex since the only supporting
hyperplane (line) containing the origin also contains the line
segment from (0,0) to (3,0).
10. One possibility is a ray. It has an extreme point at one end.
11. One possibility is to let S be a square that includes part of the boundary but not all of it. For example,
include just two adjacent edges. The convex hull of the profile P is a triangular region.
!"
"#$%!#!!&!
12. a. f 0(S 5) = 6, f 1(S 5) = 15, f 2(S 5) = 20, f 3(S 5) = 15, f 4(S 5) = 6, and 6 − 15 + 20 − 15 + 6 = 2.
b.
f0
f1
f2
f3
S
1
2
S
2
3
3
S
3
4
6
4
S4
5
10
10
5
5
6
15
20
15
S
f4
6
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
474
CHAPTER 8
• The Geometry of Vector Spaces!
§a·
§ n + 1·
a!
f k (S n ) = ¨
is the binomial coefficient.
¸ , where ¨ ¸ =
© b ¹ b!(a − b)!
© k + 1¹
13. a. To determine the number of k-faces of the 5-dimensional hypercube C 5, look at the pattern that is
followed in building C 4 from C 3. For example, the 2-faces in C 4 include the 2-faces of C 3 and
the 2-faces in the translated image of C 3. In addition, there are the 1-faces of C 3 that are
“stretched” into 2-faces. In general, the number of k-faces in C n equals twice the number of kfaces in C n – 1 plus the number of (k – 1)-faces in C n – 1. Here is the pattern: fk(C n) = 2 fk (C n – 1)
+ fk – 1(C n – 1). For k = 0, 1, …, 4, and n = 5, this gives f 0(C 5) = 32, f 1(C 5) = 80, f 2(C 5) = 80,
f 3(C 5) = 40, and f 4(C 5) = 10. These numbers satisfy Euler’s formula since, 32 − 80 + 80 − 40 +
10 = 2.
§ n·
§a·
a!
b. The general formula is f k (C n ) = 2n−k ¨ ¸ , where ¨ ¸ =
is the binomial coefficient.!
©k ¹
© b ¹ b!(a − b)!
14. a.! X 1 is a line segment
!"
#"!
X 2 is a parallelogram
#"!
##!
b. f 0 (X 3) = 6, f 1 (X 3) = 12, f 2 (X 3) = 8. X 3 is an octahedron.
c. f 0 (X 4 ) = 8, f 1 (X 4 ) = 24, f 2 (X 4 ) = 32, f 3 (X 4 ) = 16, 8 − 24 + 32 − 16 = 0
d.
§ n ·
f k ( X n ) = 2k +1 ¨
¸ , 0 ≤ k ≤ n −1, where
© k + 1¹
§a·
a!
is the binomial coefficient.
¨ ¸=
© b ¹ b!(a − b)!
15. a. f 0 (P n ) = f 0 (Q) + 1
b. f k (P n ) = f k (Q) + f k − 1 (Q)
c. f n − 1 (P n ) = f n − 2 (Q) + 1
16. a.
b.
c.
d.
True. See the definition at the beginning of this section.
True. See the definition after Example 1.
False. S must be compact. See Theorem 15.
True. See the comment after Fig. 7.
17. a. False. It has six facets (faces).
b. True. See Theorem 14.
c. False. The maximum is always attained at some extreme point, but there may be other points that
are not extreme points at which the maximum is attained. See Theorem 16.
d. True. Follows from Euler’s formula with n = 2.
18. Let v be an extreme point of the convex set S and let T = {y ∈ S : y ≠ v}. If y and z are in T, then
yz ⊆ S since S is convex. But since v is an extreme point of S, v ∉ yz , so yz T . Thus T is
convex.
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
8.6
• Solutions
475!
Conversely, suppose v ∈ S, but v is not an extreme point of S. Then there exist y and z in S such
that v ∈ yz , with v ≠ y and v ≠ z. It follows that y and z are in T, but yz T . Hence T is not
convex.
19. Let S be convex and let x ∈ cS + dS, where c > 0 and d > 0. Then there exist s1 and s2 in S such that
x = cs1 + ds2. But then
§
·
x = cs1 + ds 2 = (c + d ) ¨ c s1 + d s 2 ¸ .
c
+
d
c
+
d
©
¹
Now show that the expression on the right side is a member of (c + d)S.
For the converse, pick an typical point in (c + d)S and show it is in cS + dS.
20. For example, let S = {1, 2} in . Then 2S = {2, 4}, 3S = {3, 6} and (2 + 3)S = {5, 10}.
However, 2S + 3S = {2, 4} + {3, 6} = {2 + 3, 4 + 3, 2 + 6, 4 + 6} = {5, 7, 8, 10} ≠ (2 + 3)S.
1
21. Suppose A and B are convex. Let x, y ∈ A + B. Then there exist a, c ∈ A and b, d ∈ B such that
x = a + b and y = c + d. For any t such that 0 ≤ t ≤ 1, we have
w = (1 − t )x + ty = (1 − t )(a + b) + t (c + d)
= [ (1 − t )a + tc] + [ (1 − t )b + td]
But (1 − t)a + tc ∈ A since A is convex, and (1 − t)b + td ∈ B since B is convex. Thus w is in A + B,
which shows that A + B is convex.
22. a. Since each edge belongs to two facets, kr is twice the number of edges: k r = 2e. Since each edge
has two vertices, s v = 2e.
b. v − e + r = 2, so 2se − e + 2ke = 2 1s + 1k = 12 + 1e
c. A polygon must have at least three sides, so k ≥ 3. At least three edges meet at each vertex,
so s ≥ 3. But both k and s cannot both be greater than 3, for then the left side of the equation
in (b) could not exceed 1!2.
When k = 3, we get 1s − 16 = 1e , so s = 3, 4, or 5. For these values, we get e = 6, 12, or 30,
corresponding to the tetrahedron, the octahedron, and the icosahedron, respectively.
When s = 3, we get 1k − 16 = 1e , so k = 3, 4, or 5 and e = 6, 12, or 30, respectively.
These values correspond to the tetrahedron, the cube, and the dodecahedron.
8.6
SOLUTIONS
Notes: This section moves beyond lines and planes to the study of some of the curves that are used to
model surfaces in engineering and computer aided design. Notice that these curves have a matrix
representation.
1. The original curve is x(t) = (1 – t)3p0 + 3t(1 – t)2p1 + 3t 2(1 – t)p2 + t 3p3 (0 < t < 1). Since the
curve is determined by its control points, it seems reasonable that to translate the curve, one
should translate the control points. In this case, the new Bézier curve y(t) would have the
equation
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
476
• The Geometry of Vector Spaces!
CHAPTER 8
y(t) = (1 – t)3(p0 + b) + 3t(1 – t)2(p1 + b) + 3t 2(1 – t)(p2 + b) + t 3(p3 + b)
= (1 – t)3p0 + 3t(1 – t)2p1 + 3t 2(1 – t)p2 + t 3p3
+ (1 – t)3b + 3t(1 – t)2b + 3t 2(1 – t)b + t 3b
A routine algebraic calculation verifies that (1 – t)3 + 3t(1 – t)2 + 3t 2(1 – t) + t 3 = 1 for all t.
Thus y(t) = x(t) + b for all t, and translation by b maps a Bézier curve into a Bézier curve.
2. a. Equation (15) reveals that each polynomial weight is nonnegative for 0 < t < 1, since 4 − 3t >
0. For the sum of the coefficients, use (15) with the first term expanded: 1 – 3t + 6t 2 − t 3.
The 1 here plus the 4 and 1 in the coefficients of p1 and p2, respectively, sum to 6, while the
other terms sum to 0. This explains the 1/6 in the formula for x(t), which makes the
coefficients sum to 1. Thus, x(t) is a convex combination of the control points for 0 < t < 1.
b. Since the coefficients inside the brackets in equation (14) sum to 6, it follows that
b = 16 [ 6b] = 16 ª(1 − t )3 b + (3t 3 − 6t 2 + 4)b + (−3t 3 + 3t 2 + 3t + 1)b + t 3b º and hence x(t) + b
¬
¼
may be written in a similar form, with pi replaced by pi + b for each i. This shows that
x(t) + b is a cubic B-spline with control points pi + b for i = 0, …, 3.
3. a. x' (t) = (–3 + 6t – 3t 2)p0 + (3 –12t + 9t 2)p1 + (6t – 9t 2)p2 + 3t 2p3, so x' (0) = –3p0 + 3p1=3(p1 – p0),
and x' (1) = –3p2 + 3p3 = 3(p3 – p2). This shows that the tangent vector x' (0) points in the
direction from p0 to p1 and is three times the length of p1 – p0. Likewise, x' (1) points in the
direction from p2 to p3 and is three times the length of p3 – p2. In particular, x' (1) = 0 if and only
if p3 = p2.
b. x'' (t) = (6 – 6t)p0 + (–12 + 18t)p1 + (6 – 18t)p2 + 6tp3, so that
x'' (0) = 6p0 – 12p1 + 6p2 = 6(p0 – p1) + 6(p2 – p1)
and x'' (1) = 6p1 – 12p2 + 6p3 = 6(p1 – p2) + 6(p3 – p2)
For a picture of x'' (0), construct a coordinate system with the origin at p1, temporarily, label p0
as p0 − p1, and label p2 as p2 − p1. Finally, construct a line from this new origin through the sum
of p0 − p1 and p2 − p1, extended out a bit. That line points in the direction of x'' (0).
0 = p1
p2 – p1
p0 – p1
w = (p 0 − p1 ) + (p 2 − p1 ) =
w
(
)
(
)
(
1 x′′(0)
6
)
4. a. x' (t) = 16 ª −3t 2 + 6t − 3 p0 + 9t 2 − 12t p1 + −9t 2 + 6t + 3 p 2 + 3t 2p3 º
¬
¼
1
1
x' (0) = 2 ( p 2 − p 0 ) and x' (1) = 2 ( p 3 − p1 )
(Verify that, in the first part of Fig. 10, a line drawn through p0 and p2 is parallel to the
tangent line at the beginning of the B-spline.)
When x' (0) and x' (1) are both zero, the figure collapses and the convex hull of the set of
control points is the line segment between p0 and p3, in which case x(t) is a straight line.
Where does x(t) start? In this case,
x(t) = 16 ª(−4t 3 + 6t 2 + 2)p0 + (4t 3 − 6t 2 + 4)p3 º
¬
¼
x(0) = 13 p 0 + 23 p 3 and x(1) = 23 p 0 + 13 p 3
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
8.6
• Solutions
477!
The curve begins closer to p3 and finishes closer to p0. Could it turn around during its travel?
Since x' (t) = 2t(1 − t)(p0 − p3), the curve travels in the direction p0 − p3, so when x' (0) = x' (1)
= 0, the curve always moves away from p3 toward p0 for 0 < t < 1.
b. x'' (t) = (1 – t)p0 + (–2 + 3t)p1 + (1 – 3t)p2 + tp3
x'' (0) = p0 – 2p1 + p2 = (p0 − p1) + (p2 − p1)
and
x'' (1) = p1 – 2p2 + p3 = (p1 − p2) + (p3 − p2)
For a picture of x'' (0), construct a coordinate system with the origin at p1, temporarily, label p0 as
p0 − p1, and label p2 as p2 − p1. Finally, construct a line from this new origin to the sum of p0 − p1
and p2 − p1. That segment represents x'' (0).
For a picture of x'' (1), construct a coordinate system with the origin at p2, temporarily, label p1 as
p1 − p2, and label p3 as p3 − p2. Finally, construct a line from this new origin to the sum of
p1 − p2 and p3 − p2. That segment represents x'' (1).
p1 – p2
w
p2 = 0
p3 – p2
w = (p1 − p2 ) + (p3 − p2 ) = x′′(1)
5. a. From Exercise 3(a) or equation (9) in the text,
x' (1) = 3(p3 – p2)
Use the formula for x'(0), with the control points from y(t), and obtain
y' (0) = –3p3 + 3p4 = 3(p4 – p3)
For C1 continuity, 3(p3 – p2) = 3(p4 – p3), so p3 = (p4 + p2)/2, and p3 is the midpoint of the
line segment from p2 to p4.
b. If x' (1) = y' (0) = 0, then p2 = p3 and p3 = p4. Thus, the “line segment” from p2 to p4 is just
the point p3. [Note: In this case, the combined curve is still C1 continuous, by definition.
However, some choices of the other control points, p0, p1, p5, and p6 can produce a curve
with a visible “corner” at p3, in which case the curve is not G1 continuous at p3.]
6. a.
With x(t) as in Exercise 2,
x(0) = (p0 + 4p1 + p2)/6 and x(1) = (p1 + 4p2 + p3)/6
Use the formula for x(0), but with the shifted control points for y(t), and obtain
y(0) = (p1 + 4p2 + p3)/6
This equals x(1), so the B-spline is G0 continuous at the join point.
b. From Exercise 4(a),
x' (1) = (p3 – p1)/2 and x' (0) = (p2 – p0)/2
Use the formula for x' (0) with the control points for y(t), and obtain
y' (0) = (p3 – p1)/2 = x' (1)
Thus the B-spline is C1 continuous at the join point.
7. From Exercise 3(b),
x!! (0) = 6(p0 – p1) + 6(p2 – p1) and x!! (1) = 6(p1 – p2) + 6(p3 – p2)
Use x!! (0) with the control points for y(t), to get
y!! (0) = 6(p3 – p4) + 6(p5 – p4)
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
478
CHAPTER 8
• The Geometry of Vector Spaces!
Set x'' (1) = y'' (0) and divide by 6, to get
(p1 – p2) + (p3 – p2) = (p3 – p4) + (p5 – p4)
(*)
Since the curve is C1 continuous at p3, the point p3 is the midpoint of the segment from p2 to p4, by
Exercise 5(a). Thus p 3 = 12 (p 2 + p 4 ) , which leads to p4 – p3 = p3 – p2. Substituting into (*) gives
(p1 – p2) + (p3 – p2) = –(p3 – p2) + p5 – p4
(p1 – p2) + 2(p3 – p2) + p4 = p5
Finally, again from C1 continuity, p4 = p3 + p3 – p2. Thus,
p5 = p3 + (p1 – p2) + 3(p3 – p2)
So p4 and p5 are uniquely!"#$#%&'(#"!)*!!+,!!-,!.("!!/0!!!1(2*!!3!4.(!)#!4567#(!.%)'$%.%'2*0!
8. From Exercise 4(b), x!! (0) = p0 – 2p1 + p2 and x!! (1) = p1 – 2p2 + p3. Use the formula for x!! (0), with
the shifted control points for y(t), to get
y!! (0) = p1 – 2p2 + 2p3 = x!! (1)
Thus the curve has C2 continuity at x(1).
9. Write a vector of the polynomial weights for x(t), expand the polynomial weights and factor the
vector as MBu(t):
ª1 − 4t + 6t 2 − 4t 3 + t 4 º
6 −4
1º ª 1 º
«
» ª 1 −4
« »
« 4t − 12t 2 + 12t 3 − 4t 4 » «0 4 −12 12 −4 » « t »
» 2
«
» «
2
3
4
6 −12 6 » «t » , MB =
« 6t − 12t + 6t
» = «0 0
»« »
«
» «0 0
3
4
0
4 −4 » « t 3 »
«
4
t
4
t
−
«
»
« »
0
0
1»¼ «t 4 »
«
» «¬0 0
4
t
¬ ¼
«¬
»¼
6 −4
1º
ª 1 −4
«0 4 −12 12 −4 »
«
»
«0 0
6 −12 6 »
«
»
0
4 −4 »
«0 0
«¬0 0
0
0
1»¼
10. Write a vector of the polynomial weights for x(t), expand the polynomial weights, taking care to
write the terms in ascending powers of t, and factor the vector as MSu(t):
ª 1 − 3t + 3t 2 − t 3 º
ª 1 −3 3 −1º ª 1 º
ª 1 −3 3 −1º
«
»
«t »
«
»
«
»
2
3
1 « 4 − 6t + 3t » 1 « 4 0 −6 3» « »
1 « 4 0 −6 3»
= MSu(t), MS =
«
»=
6 «1 + 3t + 3t 2 − 3t 3 » 6 « 1 3 3 −3» «t 2 »
6 « 1 3 3 −3»
« »
«
»
«
»
3
«
»
0 0 0
1»¼ «¬ t »¼
0 0 0
1»¼
3
«
«
¬
¬
t
¬«
¼»
11. a. True. See equation (2).
b. False. Example 1 shows that the tangent vector x′(t) at p0 is two times the directed line segment
from p0 to p1.
c. True. See Example 2.
12. a. False. The essential properties are preserved under translations as well as linear transformations.
See the comment after Figure 1.
b. True. This is the definition of G0 continuity at a point.
c. False. The Bézier basis matrix is a matrix of the polynomial coefficients of the control points.
See the definition before equation (4).
13. a. From (12), q1 − q 0 = 12 (p1 − p 0 ) = 12 p1 − 12 p 0 . Since q 0 = p 0 , q1 = 12 (p1 + p 0 ) .
b. From (13), 8(q3 – q2) = –p0 – p1 + p2 + p3. So 8q3 + p0 + p1 – p2 – p3 = 8q2.
c. Use (8) to substitute for 8q3, and obtain
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
8.6
• Solutions
479!
8q2 = (p0 + 3p1 + 3p2 + p3) + p0 + p1 – p2 – p3 = 2p0 + 4p1 + 2p2
Then divide by 8, regroup the terms, and use part (a) to obtain
q 2 = 14 p 0 + 12 p1 + 14 p 2 = ( 14 p 0 + 14 p1 ) + ( 14 p1 + 14 p 2 ) = 12 q1 + 14 (p1 + p 2 )
= 12 (q1 + 12 (p1 + p 2 ))
14. a. 3(r3 − r2) = z!(1), by (9) with z!(1) and rj in place of x!(1) and pj.
z!(1) = .5x!(1), by (11) with t = 1.
.5x!(1) = (.5)3(p3 − p2), by (9).
b. From part (a), 6(r3 − r2) = 3(p3 − p2), r3 − r2 = 12 p 3 − 12 p 2 , and r3 − 12 p 3 + 12 p 2 = r2 .
Since r3 = p3, this equation becomes r2 = 12 (p 3 + p 2 ) .
c. 3(r1 − r0) = z! (0), by (9) with z!(0) and rj in place of x!(0) and pj.
z! (0) = .5x!(.5), by (11) with t = 0.
d. Part (c) and (10) show that 3(r1 − r0) =
3
8
(−p0 − p1 + p2 + p3). Multiply by
8
3
and rearrange to
obtain 8r1 = −p0 − p1 + p2 + p3 + 8r0.
e. From (8), 8r0 = p0 + 3p1 + 3p2 + p3. Substitute into the equation from part (d), and obtain
8r1 = 2p1 + 4p2 + 2p3. Divide by 8 and use part (b) to obtain
r1 = 14 p1 + 12 p 2 + 14 p 3 = ( 14 p1 + 14 p 2 ) +
1
4
( p 2 + p 3 ) = 12 < 12 (p1 + p 2 ) + 12 r2
Interchange the terms on the right, and obtain r1 = 12 [r2 + 12 (p1 + p 2 )] .
15. a. From (11), y!(1) = .5x!(.5) = z!(0).
b. Observe that y!(1) = 3(q 3 – q 2). This follows from (9), with y(t) and its control points in
place of x(t) and its control points. Similarly, for z(t) and its control points, z!(0) = 3(r1 – r0).
By part (a) 3(q 3 – q 2) = 3(r1 – r0). Replace r0 by q 3, and obtain q 3 – q 2 = r1 – q 3, and hence
q 3 = (q 2 + r1)/2.
c. Set q 0 = p0 and r3 = p3. Compute q1 = (p0 + p1)/2 and r2 = (p2 + p3)/2. Compute m = (p1 + p2)/2.
Compute q 2 = (q1 + m)/2 and r1 = (m + r2)/2. Compute q 3 = (q 2 + r1)/2 and set r0 = q 3.
16. A Bézier curve is completely determined by its four control points. Two are given directly: p0 =
x(0) and p3 = x(1). From equation (9), x! (0) = 3(p1 – p0) and x! (1) = 3(p3 – p2). Solving gives
!
p1 = p0 +
1
3
x! (0) and p2 = p3 −
!
1
3
!
x! (1)
!
17. a. The quadratic curve is w(t) = (1 – t) 2 p0 + 2t(1 − t)p1 + t 2 p2. From Example 1, the tangent
vectors at the endpoints are w! (0) = 2p1 − 2p0 and w! (1) = 2p2 − 2p1. Denote the control
points of x(t) by r0, r1, r2, and r3. Then
r0 = x(0) = w(0) = p0 and r3 = x(1) = w(1) = p2
From equation (9) or Exercise 3(a) (using ri in place of pi) and Example 1,
–3r0 + 3r1 = x! (0) = w! (0) = 2p1 − 2p0
!
so −p0 + r1 =
2p1 − 2p0
3
and
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
480
CHAPTER 8
• The Geometry of Vector Spaces!
p0 + 2p1
(i)
3
Similarly, using the tangent data at t = 1, along with equation (9) and Example 1, yields
–3r2 + 3r3 = x! (1) = w! (1) = 2p2 − 2p1,
r1 =
!
2p 2 − 2p1
2p − 2p1
, r2 = p 2 − 2
, and
3
3
2p + p 2
r2 = 1
(ii)
3
b. Write the standard formula (7) in this section, with ri in place of pi for i = 1, …, 4, and then
replace r0 and r3 by p0 and p2, respectively:
x(t) = (1 – 3t + 3t2 – t 3)p0 + (3t – 6t2 + 3t 3)r1 + (3t2 – 3t 3)r2 + t 3p2
(iii)
Use the formulas (i) and (ii) for r1 and r2 to examine the second and third terms in (iii):
(3t − 6t 2 + 3t 3 )r1 = 13 (3t − 6t 2 + 3t 3 )p 0 + 23 (3t − 6t 2 + 3t 3 )p1
−r2 + p2 =
= (t − 2t 2 + t 3 )p 0 + (2t − 4t 2 + 2t 3 )p1
(3t 2 − 3t 3 )r2 = 23 (3t 2 − 3t 3 )p1 + 13 (3t 2 − 3t 3 )p 2
= (2t 2 − 2t 3 )p1 + (t 2 − t 3 )p 2
When these two results are substituted in (iii), the coefficient of p0 is
(1 − 3t + 3t 2 − t 3) + (t − 2t 2 + t 3) = 1 − 2t + t 2 = (1 − t)2
The coefficient of p1 is
(2t − 4t 2 + 2t 3) + (2t 2 − 2t 3) = 2t − 2t 2 = 2t(1 − t)
The coefficient of p2 is (t 2 − t 3) + t 3 = t 2. So x(t) = (1 − t)2p0 + 2t(1 − t)p1 + t 2p2, which
shows that x(t) is the quadratic Bézier curve w(t).
p0
ª
º
«
»
−3p 0 + 3p1
«
»
18.
« 3p 0 − 6p1 + 3p 2 »
«
»
«¬ −p 0 + 3p1 − 3p 2 + p3 »¼
Copyright © 2012 Pearson Education, Inc. Publishing as Addison-Wesley.
Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.7 Linearized : No Author : Creator : -|- this layout: pidus -|- Modify Date : 2015:04:15 19:51:31-04:00 Subject : Create Date : 2015:04:15 19:41:52-04:00 PXC Viewer Info : PDF-XChange Viewer;2.5.196.0;Jun 30 2011;20:40:17;D:20150415195131-04'00' Page Count : 485 Has XFA : No XMP Toolkit : XMP Core 4.1.1 Metadata Date : 2015:04:15 19:41:52-04:00 Creator Tool : -|- this layout: pidus -|- Format : application/pdf Description : Title : Document ID : uuid:02502e73-ade2-4426-a332-2802c897d6ef Instance ID : uuid:20e955ba-b28c-4f91-899e-ecb1d1abe540 Producer : -|- this layout: pidus -|- Page Mode : UseOutlines Page Layout : SinglePageEXIF Metadata provided by EXIF.tools