AEMx CH08

Chapter 8
Matrices
8.1 Matrix Algebra
1. 2 × 4 2. 3 × 2 3. 3 × 3 4. 1 × 3 5. 3 × 4
6. 8 × 1 7. Not equal 8. Not equal 9. Not equal 10. Not equal
11. Solving x = y − 2, y = 3x − 2 we obtain x = 2, y = 4.
12. Solving x2 = 9, y = 4x we obtain x = 3, y = 12 and x = −3, t = −12.
13. c23 = 2(0) − 3(−3) = 9; c12 = 2(3) − 3(−2) = 12
14. c23 = 2(1) − 3(0) = 2; c12 = 2(−1) − 3(0) = −2

4−2 5+6 2 11
15. (a) A + B = =
−6 + 8 9 − 10 2 −1

−2 − 4 6−5 −6 1
(b) B − A = =
8 + 6 −10 − 9 14 −19

8 10 −6 18 2 28
(c) 2A + 3B = + =
−12 18 24 −30 12 −12
   
−2 − 3 0 + 1 −5 1
16. (a) A − B =  4 − 0 1 − 2 =  4 −1
7+4 3+2 11 5
   
3 + 2 −1 − 0 −5 −1
(b) B − A =  0 − 4 2 − 1 =  −4 1
−4 − 7 −2 − 3 −11 −5

  
1 −1 2 −2
(c) 2(A + B) = 2 4 3 = 8 6
3 1 6 2
550
8.1 Matrix Algebra 551

−2 − 9 12 − 6 −11 6
17. (a) AB = =
5 + 12 −30 + 8 17 −22

−2 − 30 3 + 24 −32 27
(b) BA = =
6 − 10 −9 + 8 −4 −1

2 4 + 12 −6 − 12 19 −18
(c) A = =
−10 − 20 15 + 16 −30 31

2 1 + 18 −6 + 12 19 6
(d) B = =
−3 + 6 18 + 4 3 22
   
−4 + 4 6 − 12 −3 + 8 0 −6 5
18. (a) AB = −20 + 10 30 − 30 −15 + 20 = −10 0 5
−32 + 12 49 − 36 −24 + 24 −20 12 0

−4 + 30 − 24 −16 + 60 − 36 2 8
(b) BA = =
1 − 15 + 16 4 − 30 + 24 2 −2

9 24
19. (a) BC =
3 8

1 −2 9 24 3 8
(b) A(BC) = =
−2 4 3 8 −6 −16

0 2 0 0 0 0
(c) C(BA) = =
3 4 0 0 0 0

1 −2 6 5 −4 −5
(d) A(B + C) = =
−2 4 5 5 8 10
 
3
20. (a) AB = 5 −6 7  4 = −16
−1
  
3 15 −18 21
(b) BA =  4 5 −6 7 =  20 −24 28
−1 −5 6 −7
    
15 −18 21 1 2 4 78 54 99
(c) (BA)C =  20 −24 28 0 1 −1 =  104 72 132
−5 6 −7 3 2 1 −26 −18 −33
(d) Since AB is 1 × 1 and C is 3 × 3 the product (AB)C is not defined.

552 CHAPTER 8 MATRICES
 
4
21. (a) AT A = 4 8 −10  8 = 180

−10
   
2 4 8 10
(b) BT B) = 4 2 4 5 =  8 16 20

5 10 20 25
     
4 2 6
(c) A + BT =  8 + 4 =  12
−10 5 −5

T 1 2 −2 5 −1 7
22. (a) A + B = + =
2 4 3 7 5 11

2 4
T −2 5 4 −1
T
(b) 2A − B = − =
4 8 3 7 1 1

T 1 2 3 −1 −3 −7
(c) A (A − B) = =
2 4 −3 −3 −6 −14
T
T 7 10 7 38
23. (a) (AB) = =
38 75 10 75

T T 5 −2 3 8 7 38
(b) B A = =
10 −5 4 1 10 75

T 5 −4 −3 11 2 7
24. (a) A + B = + =
9 6 −7 2 2 8

T 10 18 −3 −7 7 11
(b) 2A + B = + =
−8 12 11 2 3 14
       
6 −5 −6 −5
−4 4 −6 −14
25. − + = 26.  3 + −5 + −8 = −10
8 16 9 1
−3 15 10 22
       
−7 −1 2 −10
−19 19 −38
27. − = 28.  17 +  1 −  8 =  10
18 20 −2
−6 4 −6 4
29. 4 × 5 30. 3 × 2

T2 −3 T 6 −6
31. A = ; 32. (A + B) = = AT + BT
4 2 14 10

T T 2 4
(A ) = =A
−3 2
8.1 Matrix Algebra 553
T
16 40 16 −8 4 2 2 −3 16 −8
33. (AB)T = = ; BT AT = =
−8 −20 40 −20 10 5 4 2 40 −20

T 12 −18
34. (6A) = = 6AT
24 12
   
2 1 5 15 9
2 6 2
35. B = AAT = 6 3 = 15 39 27 = BT
1 3 5
2 5 9 27 29
36. Using Problem 33 we have (AAT )T = (AT )T AT = AAT , so that AAT is symmetric.

1 0 0 0
37. Let A = and B = . Then AB = 0.
0 0 0 1
 
2 3 4
38. We see that A 6= B, but AC = 4 6 8 = BC.
6 9 12
39. Since (A + B)2 = (A + B)(A + B) 6= A2 + AB + BA + B2 , and AB 6= BA in general,

(A + B)2 6= A2 + 2AB + B2 .
40. Since (A + B)(A − B) = A2 − AB + BA − B2 , and AB 6= BA in general, (A + B)(A − B) 6=

A2 − B2 .
41. a11 x1 + a12 x2 = b1 ; a21 x1 + a22 x2 = b2

    
2 6 1 x1 7
42.  1 2 −1   x2 = −1
 
5 7 −4 x3 9

a b/2 x x
= ax2 + bxy/2 + bxy/2 + cy 2

43. x y = ax + by/2 bx/2 + cy
b/2 c y y
= ax2 + bxy + cy 2

    
0 −∂/∂z ∂/∂y P −∂Q/∂z + ∂R/∂y
44.  ∂/∂z 0 −∂/∂x Q =  ∂P/∂z − ∂R/∂x = curl F
−∂/∂y ∂/∂x 0 R −∂P/∂y + ∂Q/∂x
 π π
cos − sin
x1 2 2 1 0 −1 1 −1
45. = = =

y1 π π 
1 1 0 1 1
sin cos
2 2
Therefore b = h−1, 1i .
√
π π 3 1
 
cos − sin −  √ !
− 3 − 2

x1 6 6  −2  2 2 −2
46. = = = √

y1 π π 4 √  4
sin cos

1 3
 −1 + 2 3
6 6 2 2
√ √
Therefore b = h− 3 − 2, −1 + 2 3i .
3π 3π 1 1
   
√ − √ − √ √ !
cos 4 − sin 4 
!
2 2

x1  √
 2 2 
 √ −2
47. = = =
y1  3π 3π  2  1 1  2 0
sin cos √ −√
4 4 2 2
Therefore b = h−2, 0i .
√  √ 
2π 2π 1 3 1 3
   
cos − sin   − − − +
x1  3 3  1 2 2   1  2 2 
48. = = √ = √ 
y1  2π 2π  −1 
3 1
 −1 
3 1

sin cos − − +
3 3 2 2 2 2
1
√ √
Therefore b = 2 h−1 + 3, 1 + 3i .

−1 0
49. M = since
0 1

x1 −1 0 x −x
= =
y1 0 1 y y

1 0
50. M = since
0 −1

x1 1 0 x x
= =
y1 0 −1 y −y
        
x cos γ sin γ 0 x x cos γ + y sin γ xY
51. (a) MY  y  = − sin γ cos γ 0  y  = −x sin γ + y cos γ  =  yY 
z 0 0 1 z z zY
   
cos β 0 − sin β 1 0 0
(b) MR =  0 1 0 ; MP 0 cos α sin α
sin β 0 cos β 0 − sin α cos α
 
1 0 0
 
      √   1
1 1 0 0 1 3 1 1 1 √

 
(c) MP 1 = 0 cos 30◦ ◦
sin 30  1 = 
 0  1 = 
  3 + 1 
2 2 2

1 0 − sin 30◦ cos 30◦ 1
 √ 
1

√

 1 3   1 
0 − 3 − 1
2 2 2
 
    1
1 cos 45◦ 0 − sin 45◦  1 √ 
MR MP 1 =  0 1 0 
 3 + 1 
2


1 sin 45◦ 0 cos 45◦  1 √ 
3−1
2
√ √   1 √ √ 
2 2 1 3 2 − 6
0 − 4
 2 2  1 √
  
  1 √
  
= 0 1
 0  2
 3 + 1 = 3+1 

  2√
√ √  √
    
 2 2 1 1 √ 
0 3−1 2 + 6

2 2 2 4
8.2 Systems of Linear Algebraic Equations 555
√ √  1
    4 3 2 − 6
1 ◦ ◦
cos 60 sin 60 0 

1 √ 
MY MR MP 1 = − sin 60◦ cos 45◦ 0  3+1 
 
1 0 0 1  2√
 
1 √ 
2+ 6
4
√  1 √ √ 
1 3

3 2− 6
0  4
 √2 2

√

  1 
= 3 + 1

− 3 1   
0 
  2 
2 2 1 √ √ 

0 0 1 2+ 6
4
 1 √ √ √ 
3 2− 6+6+2 3
 8 
1
 √ √ √ 
=  −3 6 + 3 2 + 2 3 + 2 

8 

1 √ √ 
2+ 6
4
 
17 43
A11 A12 B1 A11 B1 + A12 B2
52. (a) AB = = =  3 75
A21 A22 B2 A21 B1 + A22 B2
−14 51
since

13 25 4 18 17 43
A11 B1 + A12 B2 = + =
−9 49 12 26 3 75
and

A21 B1 + A22 B2 = −24 34 + 10 17 = −14 51 .
(b) It is easier to enter smaller strings of numbers and the chance of error is decreased. Also,
if the large matrix has submatrices consisting of all zeros or diagonal matrices, these are
easily entered without listing all of the entries.
8.2 Systems of Linear Algebraic Equations
1

1 −1 11 −4R1 +R2 1 −1 11 R
7 2 1 −1 11 R3 +R1 1 0 4
1. −−−−−−→ −−−−−−→ −−−−−−→
4 3 −5 0 7 −49 0 1 −7 0 1 −7
The solution is x1 = 4, x2 = −7.

3 −2 4 R 1 −1 −2 −3R1 +R2 1 −1 −2 R +R1 1 0 8
2. −−−−12
−−→ −−−−−−→ −−−2−−−→
1 −1 −2 3 −2 4 0 1 10 0 1 10
The solution is x1 = 8, x2 = 10.

1
− 59 1
− 95 1
− 59
! ! ! !
9 3 −5 1
R
9 1
1 3 −2R1 +R2 1 3 3R2 1 3
3. −−−−−−→ −−−−−−→ 1 1
−−−−−−→ 1
2 −1 −1 2 1
−1 0 3 9 0 1 3
1 0 − 32
!
− 1 R2 +R1
−−3−−−−→ 1
0 1 3
The solution is x1 = − 32 , x2 = 13 .
3 1 3 1 3 1
! ! ! !
10 15 1 1
R
10 1
1 2 − 10 −3R1 +R2 1 2 − 10 − 52 R2 1 2 10
4. −−−−−−→ −−−−−−→ −−−−−−→
3 2 −1 3 2−1 0 − 52 − 13
10 0 1 13
25
17
!
− 3 R2 +R1 1 0 − 25
−−2−−−−→ 13
0 1 25
17 13
The solution is x1 = − 25 , x2 = 25 .
     
1 −1 −1 −3 1 −1 −1 −3 1 −1 −1 −3
1
  −2R1 +R2   5
R2  7 13 

5. 
2 3 5 7 −−R
−−−−−→  0 5 7 13  −−−−−−→ 0
 1 5 5 
1 +R3
1 −2 3 −11 0 −1 4 −8 0 −1 4 −8
0 25 − 25 0 52 − 52
   
1 1
R2 +R1  7
 275 R3 
13 
1 75 13 

−−−−−−→  0 1 5 5  −−−−−−→ 0

5 
R2 +R3 
0 0 27
5 − 27
5 0 0 1 −1
 
1 0 0 0
− 25 R3 +R1  
−−−−−−→  0 1 0 4
− 57 R3 +R2

0 0 1 −1
The solution is x1 = 0, x2 = 4, x3 = −1.

     
1 2 −1 0 1 2 −1 0 1 2 −1 0
 −2R1 +R2   − 31 R2 
1 − 43 −3
 
6. 
2 1 2 9 −−R
−−−−−→  0 −3 4 9 −−−−−−→ 0

1 +R3

1 −1 1 3 0 −3 2 3 0 −3 2 3
5 5
   
1 0 3 6 1 0 3 6
−2R2 +R1  4
 − 21 R3 
1 − 43 −3

−−−−−−→  0 1 − 3 −3  −−−−−−→ 0

3R2 +R3  
0 0 −2 −6 0 0 1 3
 
1 0 0 1
− 35 R3 +R1  
−−−−−−→ 
4
R +R2
0 1 0 1 
3 3
0 0 1 3
The solution is x1 = 1, x2 = 1, x3 = 3.

1 1 1 0 −R1 +R2 1 1 1 0
7. −−−−−−→
1 1 3 0 0 0 2 0
Since x3 = 0, setting x2 = t we obtain x1 = −t, x2 = t, x3 = 0.
1

1 2 −4 9 −5R1 +R2 1 2 −4 9 − 11 R2 1 2 −4 9
8. −−−−−−→ −−−− −−→
5 −1 2 1 0 −11 22 −44 0 1 −2 4

−2R2 +R1 1 0 0 1
−−−−−−→
0 1 −2 4
If x3 = t, the solution is x1 = 1, x2 = 4 + 2t, x3 = t.
   
1 −1 −1 8 1 −1 −1 8
  row  
9. 
 1 −1 1 3 −−−−−−→ 
 operations 0 0 2 −5 
−1 1 1 4 0 0 0 12
Since the bottom row implies 0 = 12, the system is inconsistent.
1 31 4
   
3 1 4 3
row
10. 4 3 −3 −−−−−−→ 0 1 −5
   
operations
2 −1 11 0 0 0
The solution is x1 = 3, x2 = −5.
   
2 2 0 0 1 1 0 0
row 1
11. −2 1 1 0 −−−−−−→ 0 1 3 0
   
operations
3 0 1 0 0 0 1 0
The solution is x1 = x2 = x3 = 0.
   
1 −1 −2 0 1 −1 −2 0
  row  3

12. 
2 4 5 0 −−−−−−→ 
 operations 0 1 2 0

6 0 −3 0 0 0 0 0
The solution is x1 = 12 t, x2 = − 23 t, x3 = t.
   
1 2 2 2 1 2 2 2
row
13.  1 1 1 0 −−−−−−→ 0 1
  1 2
1 −3 −1 0 operations 0 0 1 4
The solution is x1 = −2, x2 − 2, x3 = 4.
   
1 −2 1 2 1 −2 1 2
row
14. 3 −1 2 5 −−−−−−→ 0 1 − 15 − 15 
   
operations
2 1 1 1 0 0 0 −2
Since the bottom row implies 0 = −2, the system is inconsistent.
   
1 1 1 3 1 1 1 3
row
15.  1 −1 −1 −1 −−−−−−→ 0 1 1
  2
3 1 1 5 operations 0 0 0 0
If x3 = t the solution is x1 = 1, x2 = 2 − t, x3 = t.
   
1 −1 −2 −1 1 −1 −2 −1
row
16. −3 −2 1 −7 −−−−−−→ 0 1 1 2
2 3 1 8 operations 0 0 0 0
If x3 = t the solution is x1 = 1 + t, x2 = 2 − t, x3 = t.
1 0 1 −1 1 1 0 1 −1 1
   
0 1 1 3
2 1 1 3
row 0
 1 2 2 2
17.   −−−−−−→ 

1 −1 0 1 −1 operations 0 0 1 −5 1

1 1 1 1 2 0 0 0 1 0
The solution is x1 = 0, x2 = 1, x3 = 1, x4 = 0.
1 1 3
2 1 1 0 3 1 0
  
2 2 2
3 1 1 1 4 row 0 1 1 −2 1
 
18.   −−−−−−→ 
 
1 2 2 3 3 operations 0 0 1 −1 −1

4 5 −2 1 16 0 0 0 1 0
The solution is x1 = 1, x2 = 2, x3 = −1, x4 = 0.

   
1 3 5 −1 1 1 3 5 −1 1
0 1 1 −1 4
 row  0 1 1 −1 4
19. 
1 −−−−−−→  
2 5 −4 −2 operations 0 0 1 −4 1
1 4 6 −2 6 0 0 0 0 1
Since the bottom row implies 0 = 1, the system is inconsistent.

   
1 2 0 1 0 1 2 0 1 0
4 9 1 12 0 row 0 1 1 8 0
20.   −−−
3 9 6 21 0 operations−−−→  0 0 1 −2 0

1 3 1 9 0 0 0 0 0 0
If x4 = t the solution is x1 = 19t, x2 = −10t, x3 = 2t, x4 = t.

   
1 1 1 4.280 1 1 1 4.28
row
21.  0.2 −0.1 −0.5 −1.978 −−−−−−→ 0
  1 2.333 9.447
4.1 0.3 0.12 1.686 operations 0 0 1 4.1
The solution is x1 = 0.3, x2 = −0.12, x3 = 4.1.

   
2.5 1.4 4.5 2.6170 1 0.56 1.8 1.0468
row
22. 1.35 0.95 1.2 0.7545 −−−−−−→ 0 1 −6.3402 −3.3953
2.7 3.05 −1.44 −1.4292 operations 0 0 1 0.28
The solution is x1 = 1.45, x2 = −1.62, x3 = 0.28.
23. From x1 Na + x2 H2 O → x3 NaOH + x4 H2 we obtain the system x1 = x3 , 2x2 = x3 + 2x4 ,

x2 = x3 . We see that x1 = x2 = x3 , so the second equation becomes 2x1 = x1 + 2x4 or
x1 = 2x4 . A solution of the system is x1 = x2 = x3 = 2t, x4 = t. Letting t = 1 we obtain the
balanced equation 2Na + 2H2 O → 2NaOH + H2 .
24. From x1 KClO3 → x2 KCl + x3 O2 we obtain the system x1 = x2 , x1 = x2 , 3x1 = 2x3 . Letting
x3 = t we see that a solution of the system is x1 = x2 = 32 t, x3 = t. Taking t = 3 we obtain
the balanced equation
2KClO3 → 2KCl + 3O2 .
25. From x1 Fe3 O4 + x2 C → x3 Fe + x4 CO we obtain the system 3x1 = x3 , 4x1 = x4 , x2 = x4 .

Letting x1 = t we see that x3 = 3t and x2 = x4 = 4t. Taking t = 1 we obtain the balanced
equation
Fe3 O4 + 4C → 3Fe + 4CO.
26. From x1 C5 H8 + x2 O2 → x3 CO2 + x4 H2 O we obtain the system 5x1 = x3 , 8x1 = 2x4 ,

2x2 = 2x3 + x4 . Letting x1 = t we see that x3 = 5t, x4 = 4t, and x2 = 7t. Taking t = 1 we
obtain the balanced equation
C5 H8 + 7O2 → 5CO2 + 4H2 O.
27. From x1 Cu + x2 HNO3 → x3 Cu(NO3 )2 + x4 H2 O + x5 NO we obtain the system
x1 = 3, x2 = 2x4 , x2 = 2x3 + x5 , 3x2 = 6x3 + x4 + x5 .
Letting x4 = t we see that x2 = 2t and
2t = 2x3 + x5 2x3 + x5 = 2t
or
6t = 6x3 + t + x5 6x3 + x5 = 5t.
3 1 3
Then x3 = 4t and x5 = 2 t. Finally, x1 = x3 = 4 t. Taking t = 4 we obtain the balanced
equation
3Cu + 8HNO3 → 3Cu(NO3 )2 + 4H2 O + 2NO.
28. From x1 Ca3 (PO4 )2 + x2 H3 PO4 → x3 Ca(H2 PO4 )2 we obtain the system
3x1 = x3 , 2x1 + x2 = 2x3 , 8x1 + 4x2 = 8x3 , 3x2 = 4x3 .
Letting x1 = t we see from the first equation that x3 = 3t and from the fourth equation that
x2 = 4t. These choices also satisfy the second and third equations. Taking t = 1 we obtain
the balanced equation
Ca3 (PO4 )2 + 4H3 PO4 → 3Ca(H2 PO4 )2 .

29. The system of equations is
−i1 + i2 − i3 = 0 −i1 + i2 − i3 = 0
10 − 3i1 + 5i3 = 0 or 3i1 − 5i3 = 10
27 − 6i2 − 5i3 = 0 6i2 + 5i3 = 27
Gaussian elimination gives

   
−1 1 −1 0 1 −1 1 0
 3 0 −5 10 −−− row
−−−→ 0 1 −8/3 10/3
0 6 5 27 operations 0 0 1 1/3
35 38 1
The solution is i1 = , i2 = , i3 = .
9 9 3
30. The system of equations is
i1 − i2 − i3 = 0 i1 − i2 − i3 = 0
52 − i1 − 5i2 = 0 or i1 + 5i2 = 52
−10i3 + 5i2 = 0 5i2 − 10i3 = 0
Gaussian elimination gives

   
1 −1 −1 0 1 −1 −1 0
1 row
5 0 52 −−−−−−→ 0 1 1/6 26/3
0 5 −10 0 operations 0 0 1 4
The solution is i1 = 12, i2 = 8, i3 = 4.
31. AX = 0 is
    
1 1 1 x1 0
5 −2 2 x2  = 0
8 1 5 x3 0
Now
      
1 1 1 4c1 4c1 + 3c1 − 7c1 0
5 −2 2  3c1  = 20c1 − 6c1 − 14c1  = 0 for all values of c1 .
8 1 5 −7c1 32c1 + 3c1 − 35c1 0
32. AX = 0 is
    
1 1 1 1 x1 0
−1 1 −1 1 x2  0
   =  
 1 1 −1 −1 x3  0
3 1 1 −1 x4 0
Now
      
1 1 1 1 c1 c1 − c1 − c1 + c1 0
−1 1 −1 1 −c1  −c1 − c1 + c1 + c1  0
  =  =   for all values of c1 .
 1 1 −1 −1 −c1   c1 + c1 − c1 − c1  0
3 1 1 −1 c1 3c1 − c1 − c1 − c1 0
33. AX = 0 is
    
2 −3 1 x1 −12
1 1 −1 x2  =  1
4 −1 −1 x3 −10
Now
         
2 −3 1 −1 2 −3 1 2c1 −2 − 12 + 2 4c1 − 9c1 + 5c1
1 1 −1  4 + 1 1 −1 3c1  =  −1 + 4 − 2  + 2c1 + 3c1 − 5c1 
4 −1 −1 2 4 −1 −1 5c1 −4 − 4 − 2 8c1 − 3c1 − 5c1
   
−12 0
=  1 + 0
 
−10 0
 
−12
=  1 for all values of c1 .
−10
34. AX = 0 is
    
1 2 −3 x1 2
2 4 −6 x2  = 4
3 6 −9 x3 6
Now
            
1 2 −3 2 1 2 −3 c1 1 2 −3 −2c2 2 0
2 4 −6 0 + 2 4 −6 c1  + 2 4 −6  c2  = 4 + 0
3 6 −9 0 3 6 −9 c1 3 6 −9 0 6 0
 
2
= 4 for all values of c1 and c2 .
6
35. Interchange row 1 and row in I3 . 36. Multiply row 3 by c in I3 .
37. Add c times row 2 to row 3 in I3 . 38. Add row 4 to row 1 in I4 .

   
a21 a22 a23 a11 a12 a13
39. EA = a11 a12 a13  40. EA = a21 a22 a23 
a31 a32 a33 a31 a32 a33
 
a11 a12 a13
41. EA =  a21 a22 a23 
ca21 + a31 ca22 + a32 ca23 + a33
42. E1 E2 A
   
a11 a12 a13 a21 a22 a23
= E1  a21 a22 a23 = a11 a12 a13 
ca21 + a31 ca22 + a32 ca23 + a33 ca21 + a31 ca22 + a32 ca23 + a33
43. Using the Solve function in Mathematica we find x1 = −0.0717393 − 1.43084c,

x2 = −0.332591 + 0.855709c, x3 = c, where c is any real number.
44. Using the Solve function in Mathematica we find x1 = c/3, x2 = 5c/6, x3 = c, where c is any
real number.
45. Using the Solve function in Mathematica we find x1 = −3.76993, x2 = −1.09071,

x3 = −4.50461, x4 = −3.12221.
46. Using the Solve function in Mathematica we find x1 = 83 − 73 b + 23 c, x2 = 2

3 − 13 b − 31 c,
x3 = −3, x4 = b, x5 = c, where b and c are any real numbers.
8.3 Rank of a Matrix

3 −1 row 1 3
1. −−−−−−→ ; The rank is 2.
1 3 operations 0 1

2 −2 row 1 −1
2. −−−−−−→ ; The rank is 1.
0 0 operations 0 0
1 3
   
2 1 3 1 2 2
row
3.  6 3 9 −−−−−−→ 0 0 0 ; The rank is 1.
   
1 3 operations
−1 − 2 − 2 0 0 0
   
1 1 2 1 1 1
row
4. −1 2 4 −−−−−−→ 0 1 −3 ; The rank is 3.
−1 0 3 operations 0 0 1
   
1 1 1 1 1 1
row
5. 1 0 4 −−−−−−→ 0 1 −3 ; The rank is 3.
1 4 1 operations 0 0 1
! !
3 −1 2 0 row 1 − 31 2
3 0
6. −−−−−−→ 5 ; The rank is 2.
6 2 4 5 operations 0 1 0 4
   
1 −2 1 −2
3 −6 row 0 1
7.   −−−−−−→ 
7 −1 operations 0
; The rank is 2.
0
4 5 0 0
1 −2 3 4 1 −2 3 4
   
1 4 6 8 row 0 1 0 0
 
8.   −−−−−−→  ; The rank is 3.
 
0 4
1 0 0 operations 0 0 1 3

2 5 6 8 0 0 0 0
8.3 Rank of a Matrix 563
1 1 3 1
   
0 2 4 2 2 1 2 3 2 6
4 1
   
4 1 0 5 1 0 1 1 −3
row 3
9.   −−−−−−→  ; The rank is 3.
  
2 1 2 1  operations 
 3 3 3 0 0 1 0 2

6 6 6 12 0 0 0 0 0 0
   
1 −2 1 8 −1 1 1 6 1 −2 1 8 −1 1 1 6
0 0 1 3 −1 1 1 5 0 0 1 3 −1 1 1 5
  row  
10. 0
 0 1 3 −1 2 10 8 −−−−−−→ 0
  0 0 0 0 1 9 3;
0 0 0 0 0 1 1 3 operations 0 0 0 0 0 0 1 0
1 −2 1 8 −1 1 2 6 0 0 0 0 0 0 0 0
The rank is 4.
   
1 2 3 1 2 3
row
11. 1 0 1 −−−−−−→ 0 1 1 ;
1 −1 5 operations 0 0 1
Since the rank of the matrix is 3 and there are 3 vectors, the vectors are linearly independent.
2 6 3 1 −1 4
   
1 −1 4
row
0 1 − 85 
12.   −−−−−−→  ;
   
3 2 1 operations 0 0 1 
2 5 4 0 0 0 1
Since the rank of the matrix is 3 and there are 4 vectors, the vectors are linearly dependent.
   
1 −1 3 −1 1 −1 3 −1
row
13.  1 −1 4 2 −−−−−−→ 0 0 1 3 ;
1 −1 5 7 operations 0 0 0 1
   
2 1 1 5 1 1 1 −1
2 2 1 1 row 0 1 1 −7
14. 
3 −1 6
 −−− −−−→  ;
1 operations 0 0 1 −3

1 1 1 −1 0 0 0 1
15. Since the number of unknowns is n = 8 and the rank of the coefficient matrix is r = 3, the
solution of the system has n − r = 5 parameters.
16. (a) The maximum possible rank of A is the number of rows in A, which is 4.
(b) The system is inconsistent if rank(A) < rank(A/B) = 2 and consistent if

rank(A) = rank(A/B) = 2.
(c) The system has n = 6 unknowns and the rank of A is r = 3, so the solution of the system
has n − r = 3 parameters.
17. Since 2v1 + 3v2 − v3 = 0 we conclude that v1 , v2 , and v3 are linearly dependent. Thus, the
rank of A is at most 2.
18. Since the rank of A is r = 3 and the number of equations is n = 6, the solution of the system
has n − r = 3 parameters. Thus, the solution of the system is not unique.
19. The system consists of 4 equations, so the rank of the coefficient matrix is at most 4, and
the maximum number of linearly independent rows is 4. However, the maximum number of
linearly independent columns is the same as the maximum number of linearly independent
rows. Thus, the coefficient matrix has at most 4 linearly independent columns. Since there
are 5 column vectors, they must be linearly dependent.
20. Using the RowReduce in Mathematica we find that the reduced row-echelon form of the
augmented matrix is
834 261
1 0 0 0 0 − 443
 
2215
1818 282 
0 1 0 0 0

 2215 443 

13 6 
0 0 1 0 0 − 443 

 443 
0 0 0 1 0 4214 130 
 2215 − 443 
6079 677
0 0 0 0 1 − 2215 443
226 834
We conclude that the system is consistent and the solution is x1 = − 443 − 2215 c,
282 1818 6 13
x2 = 443 − 2215 c, x3 = − 443 − 443 c, x4 = − 130
443 −
4214
2215 c, x5 = 677
433 + 6079
2215 c, x6 = c.
8.4 Determinants

1 2 2 4
1. M12 = =9 2. M32 = =0
−2 5 1 2

1+3 1 −1 2+2 2 4

3. C13 = (−1) −2 =1 4. C22 = (−1) −2 5 = 18
3

0 2 0 2 4 0

5. M33 = 1 2 3 = 2 6. M41 = 2 −2
3 = 24
1 1 2 1 0 −1

0 2 4 0 2 0
3+4 2+3

7. C34 = (−1) 1 2 −2 = 10
8. C23 = (−1) 5 1 −1 = 22

1 1 1 1 1 2
9. −7 10. 2 11. 17 12. −1/2
13. (1 − λ)(2 − λ) − 6 = λ2 − 3λ − 4 14. (−3 − λ)(5 − λ) − 8 = λ2 − 2λ − 23

8.4 Determinants 565

0 2 0 5 0 0
2 0 −3 0
15. 3 0 1 = −3
= −48 16. 0 −3 0 = 5
= 5(−3)(2) = −30
0 5 8 5 8 0 0 2
0 2

3 0 2
7 1 2 7
17. 2 7 1 = 3
+ 2
= 3(22) + 2(−2) = 62
2 6 4 6 4 2 6

1 −1 −1
2 −2 −1 −1 −1 −1
18. 2
2 −2 =
− 2
+ = 20 − 2(−8) + 4 = 40
1 1 9 1 9 2 −2
1 9

4 5 3
2 3 1 3 1 2
19. 1 2 3 = 4

− 5 1 3 + 1
=0
1 2 3 2 3 2

1 6 0
4
1
20. 3 8 0 = 0, expanding along the
third column.
1

2 9 0

−2 −1 4
6 1 −1 4 −1 4
21. −3 6 1 = −2 + 3
4 8 − 3 6 1 = −2(44) + 3(−24) − 3(−25) = −85

−3 4 8
4 8

3 5 1
2 5 −1 5 −1 2
22. −1
2 5 = 3
− 5
+ = 3(40) − 5(−45) + (−10) = 335
7 −4 10 −4 10 7 10 7 −4

1 1 1
y z x z x y
23. x y z =
− + = (4y − 3z) − (4x − 2z) + (3x − 2y) = −x + 2y − z
2 3 4 3 4 2 4 2 3

1 1 1
y z y z x y
24. x
y z =
− +
2 + x 3 + y 4 + z 3 + y 4 + z 3 + y 4 + z 2 + x 3 + y
= (4y + yz − 3z − yz) − (4x + xz − 2z − xz) + (3x + xy − 2y − xy)

= −x + 2y − z

1 1 −3 0
1 1 −3
1 5 3 2 1 −3 1 −3
25.
= 2 1 −2
1 = 2(4)
− 2(8)
= 8(−5) − 16)4 = −104
1 −2 1 0 4 8 0 −2 1 1 1
4 8 0 0

2 1 −1 1
2 −2 1 2 1 −1
0 5 0 4
26. = 5 1 1 0 + 4 1 6 1 = 5(0) + 4(80) = 320
1 6 1 0
5

5 −1
1 1 5 −1 1
1 1
27. Expanding along the first column in the original matrix and each succeeding minor, we obtain
3(1)(2)(4)(2) = 48.
28. Expanding along the bottom row we obtain

2 0 0 −2 2 2 0 0

1 6 0 5 1 1 6 0
−1 + = −1(−48) + 0 = 48
1 2 −1 1 1 0 2 −1
2 1 −2 3 2 0 1 −2
29. Solving λ2 − 2λ − 15 − 20 = λ2 − 2λ − 35 = (λ − 7)(λ + 5) = 0 we obtain λ = 7 and −5.
30. Solving −λ3 + 3λ2 − 2λ = −λ(λ − 2)(λ − 1) = 0 we obtain λ = 0, 1, and 2.
8.5 Properties of Determinants
1. Theorem 8.5.4 2. Theorem 8.5.7
3. Theorem 8.5.7 4. Theorem 8.5.4 and 8.5.5
5. Theorem 8.5.5 (twice) 6. Theorem 8.5.4 (twice)
7. Theorem 8.5.3 8. Theorem 8.5.2 and 8.5.5
9. Theorem 8.5.1 10. Theorem 8.5.4 (twice)
11. det A = −5 12. det B = 2(3)(5) = 30
13. det C = −5 14. det D = 5
15. By Theorem 8.5.7

a1 a2 a3 a1 − 2b1 a2 − 2b2 a3 − 2b3 a1 − 2b1 + 3c1 a2 − 2b2 + 3c2 a3 − 2b3 + 3c3

b1 b2 b3 = b1 b 2 b 3
= b1 b 2 b 3

c1 c2 c3 c1 c2 c3 c1 c2 c3
=5
16. By Theorem 8.5.5,

4a1 a2 a3

4b1 b2 b3 = 4(5) = 20.

4c1 c2 c3
Then by Theorem 8.5.7 and Theorem 8.5.5,

4a1 − 2a3 a2 a3

4b1 − 2b3 b2 b3 = 20

4c1 − 2c3 c2 c3
8.5 Properties of Determinants 567
and

4a1 − 2a3 a2 a3 4a1 − 2a3 a2 a3

4b1 − 2b3 1
b2 b3 = 4b1 − 2b3 b2 b3 = (20) = 10
1
(4c1 − 2c3 ) 1 1 1 1 2
2 c2 2 c3
2c1 − c3
2 2 c2 2 c3
17. det A = 6( 23 )(−4)(−5) = 80 18. det B = −a13 a22 a31
19. det C = (−5)(7)(3) = −105 20. det D = 4(7)(−2) = −56
21. det A = 14 = det AT 22. det A = 96 = det T

0 −2 2

23. det AB = 10 7 23 = −80 = 20(−4) = det AB
8 4 16
24. From Theorem 8.5.6, (det A)2 = det A2 = det I = 1, so det A = ±1.
25. If A is an n × n matrix such that A2 = A, then
det A2 = det A
det A2 − det A = 0
det A · det A − det A = 0
det A (det A − 1) = 0
So det A = 0 or det A = 1.
26. If A and B are n × n matrices, then det AB = det B det A is true. We know by Theorem
8.5.6 and the commutative law of ordinary multiplication that
det AB = det A · det B = det B · A = det BA.
27. Using Theorems 8.5.7, 8.5.5, and 8.5.2,

a 1 2 a 1 1

det A = b 1 2 = 2 b 1 1 = 0.
c 1 2 c 1 1
28. Using Theorems 8.5.5 and 8.5.2,

1 1 1

1 1 1

det A = x y z = (x + y + z) x y z = 0.

x + y + z x + y + z x + y + z 1 1 1

1 1 5 1 1 5 1 1 5

29. 4
3 6 = 0 −1 −14 = 0 −1 −14 = 1(−1)(15) = −15

0 −1 1 0 −1 1 0 0 15

2 4 5 2 4 5 2 4 5 2 4 5

30. 4 2 0 = 0 −6 −10 = −2 0 3 5 = −2 0 3 5 = −2(2)(3)(−7) = 84
8 7 −2 0 −9 −22 0 −9 −22 0 0 −7

−1 2 3 −1 2 3 −1 2 3

31. 4 −5 −2 = 0 3 10 = 0 3 10 = −1(3)(3) = −9
9 −9 6 0 9 33 0 0 3

−2 2 −6 1 −2 2 1 −2 2 1 −2 2
19
32. 5 0 1 = − 5 0 1 = − 0 10 −9 = − 0 10 −9 = −1(10) −

5
0 − 19

1 −2 2 −2 2 −6 0 −2 −2 0
5
= 38

1 −2 2 1 1 −2 2 1 1 −2 2 1

2 1 −2 3 0 5 −6 1 0 5 −6 1
33. = = = 1(5)(−2)(0) = 0
3 4 −8 1 0 10 −14 −2 0 0 −2 −4
3 −11 12 2 0 −5 6 −1 0 0 0 0

0 1 4 5 1 2 2 0 1 2 2 0 1 2 2 0 1 2 2 0

2 5 0 1 2 5 0 1 0 1 −4 1 0 1 −4 1 0 1 −4 1

34. = − = − = − = −

1 2 2 0 0 1 4 5 0 1 4 5 0 0 8 4 0 0 8 4

3 1 3 2 3 1 3 2 0 −5 −3 2 0 0 −23 7 0 0 0 7

37
= −(1)(1)(8) = −148
2

1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4

1 3 5 7 0 1 2 3 0 1 2 3 0 1 2 3
35. = = = = 1(1)(2)(8) = 16
2 3 6 7 0 −1 0 −1 0 0 2 2 0 0 2 2

1 5 8 20 0 3 5 16 0 0 −1 7 0 0 0 8

2 9 1 8 1 3 7 4 1 3 7 4 1 3 7 4

1 3 7 4 2 9 1 8
= − 0 3 −13 0 0 1 6 5

36. = − =
0 1 6 5 0 1 6 5 0 1 6 5 0 3 −13 0

3 1 4 2 3 1 4 2 0 −8 −17 −10 0 −8 −17 −10

1 3 7 4 1 3 7 4

0 1 6 5 0 1 6 5
= = = 1(1)(−31)(15) = −465
0 3 −31 −15 0 0 −31 −15

0 0 31 30 0 0 0 15
37. We first use the second row to reduce the third row. Then we use the first row to reduce the
second row.

1 1 1 1 1 1 1 1 1

a b c = 0 b−a c − a = (b − a)(c − a) 0 1 1 .

0 b2 − ab c2 − ac 0 b(b − a) c(c − a) 0 b c
Expanding along the first row gives (b − a)(c − a)(c − b).

8.6 Inverse of a Matrix 569
38. In order, we use the third row to reduce the fourth row, the second row to reduce the third
row, and the first row to reduce the second row. We then pull out a common factor from
each column.

1 1 1 1 1 1 1 1 1 1 1 1

a b c d 0 b−a c−a d − a 0 1 1 1
a2 b2 c2 d2 0 b2 − ab c2 − ac d2 − ac = (b − a)(c − a)(d − a) 0 b c d .
=

a3 b3 c3 d3 0 b3 − ab2 c3 − ac2 d3 − ad2 0 b2 c2 d2
Expanding along the first column and using Problem 37 we obtain (b − a)(c − a)(d − a)(c −
b)(d − b)(d − c).
39. Since C11 = 4, C12 = 5, and C13 = −6, we have
a21 C11 + a22 C12 + a23 C13 = (−1)(4) + 2(5) + 1(−6) = 0.
Since C12 = 5, C22 = −7, and C23 = −3, we have
a13 C12 + a23 C22 + a33 C32 = 2(5) + 1(−7) + 1(−3) = 0.
40. Since C11 + −7, C12 = −8, and C13 = −10 we have
a21 C11 + a22 C12 + a23 C13 = −2(−7) + 3(−8) − 1(−10) = 0.
Since C12 = −8, C22 = −19, and C32 = −7 we have
a13 C12 + a23 C22 + a33 C32 = 5(−8) − 1(−19) − 3(−7) = 0.

10 0
41. det(A + B) = = −30; det A + det B = 10 − 31 = −21
0 −3
42. det(2A) = 25 det A = 32(−7) = −224
43. Factoring −1 out of each row we see that det(−A) = (−1)5 det A = − det A. Then − det A =
det(−A) = det AT = det A and det A = 0.
44. Cofactors: 25! ≈ 1.55(1025 ); Row reduction: 253 /3 ≈ 5.2(103 )
8.6 Inverse of a Matrix

3 − 2 −1 + 1 1 0
1. AB = =
6 − 6 −2 + 3 0 1
   
2−1 −1 + 1 −2 + 2 1 0 0
2. AB =  6 − 6 −3 + 4 6 − 6  = 0 1 0
2 + 1 − 3 −1 − 1 + 2 2 + 2 − 3 0 0 1
1 1
! !
−1 1 1 1 9 9
3. det A = 9. A is nonsigular. A = =
9 −4 5 − 94 5
9
3 1
! !
−1 1 3 1 5 5
4. det A = 5. A is nonsigular. A = 1
=
5 −4 3 − 54 1
15
1
! !
−1 1 2 0 6 0
5. det A = 12. A is nonsigular. A = = 1 1
12 3 6 4 2
1 1
! !
1 π π − 3π − 3π
6. det A = −3π 2 . A is nonsigular. A −1
=− 2 1
= 2
3π − 3π
π −2π 3π
   1 1 1

8 −8 −8 −2 2 2
1    1 1 3

7. det A = −16. A is nonsigular. A−1 =−  2 −4 6
 = − 8

4 −8
16 
3
−6 4 −2 8 − 41 1
8
8. det A = 0. A is singular.
  7 13 8

−14 13 16 15 − 30 − 15
−1 1   1 2 1 

9. det A = −30. A is nonsigular. A = −  −2  4 −2 =  15 − 15
 
15 
30
2 7 2
−4 −7 −4 15 30 15
   4 10 1

8 20 2 39 39 39
1 
−2 −5 19 = − 1 − 5 19 
  
10. det A = 78. A is nonsigular. A−1 = 39 78 78
78    
2 3 1
12 −9 3 13 − 26 26
  1 
−12 0 0 3 0 0
1 
11. det A = −36. A is nonsigular. A−1 = −  0 −6 0 =  0 16 0
  
36
0 0 18 0 0 − 21
0 0 18
   
0 0 2
1 
12. det A = 16. A is nonsigular. A−1 = 8 0 0 =  21 0 0
  
16
0 16 0 0 1 0
  2 7 1
− 34
 
6 21 −9 −36 9 9 −3
   1 1 2 1

1 −1 1 6 −3  − −
  27 27 9 9
13. det A = 27. A is nonsigular. A−1 = =

27  10 17 −6 −51  10 17 2 17 
 
   27 27 − 9 − 9 
4 4 1 4
4 −4 3 12 27 − 27 9 9
   
0 1 −3 3 0 − 16 1
2 −2
1
 0 − 1 − 1 3
   
1  0 1 3 −9 6 2 2
14. det A = −6. A is nonsigular. A−1 = −  =
   
6  0 −2 1

 0 0 0

3 0 0
1 1 5
−6 −1 −3 15 1 6 2 −2
1 − 31 1 1 1 1 1
! ! ! !
6 −2 1 0 1
R
6 1 6 0 1
R +R1
3 2
1 − 6 12 −1 6 12
15. −−−−−−→ −−−−−−→ ; A =
0 4 0 1 1
R
4 2 0 1 0 41 0 1 0 1
4 0 1
4
! ! !
1 1
8 0 1 0 1
R1 1 0 8 0 8 0
16. 1 −−−8−−−→ −1
; A =
0 2 0 1
2R2 0 1 0 2 0 2
! ! !
1 3 1 0 −5R1 +R2 1 3 1 0 1
− 12 R2 1 3 1 0
17. −−−−−−→ −−−−−−→ 5 1
5 3 0 1 0 −12 −5 1 0 1 12 − 12
1 0 − 14 1
− 14 1
! !
−3R2 +R1 4 4
−−−−−−→ 5 1
; A−1 = 5 1
0 1 12 − 12 12 − 12
! ! !
2 −3 1 0 1
R1 1 − 23 1
2 0 2R1 +R2 1 − 23 1
2 0
18. −−−2−−−→ −−−− −−→
−2 4 0 1 −2 4 0 1 0 1 1 1
! !
3 3
2
3
R2 +R1 1 0 2 2 −1 2 2
−− −−−−→ ; A =
0 1 1 1 1 1
   
1 2 3 1 0 0 1 2 3 1 0 0
row
19. 4 5 6 0 1 0 −−−−−−→ 0 1 2 43 − 31 0; A is singular.
   
operations
7 8 9 0 0 1 0 0 0 1 −2 1
5 1 2
     5 1 2

1 0 −1 1 0 0 1 0 0 9 −9 9 9 −9 9
  row  2 5 1
 −1
 2 5 1

0
20.  −2 1 0 1 0 −−−−−−→ 
 operations 0 1 0 − 9 − 9 9  ; A = − 9 − 9 9 

2 −1 3 0 0 1 0 0 1 − 49 − 91 92 − 94 − 91 29
   
4 2 3 1 0 0 −1 −2 0 0 0 1
  R13  
 2
21.  1 0 0 1 0 −−−−−−→  2
 1 0 0 1 0 
−1 −2 0 0 0 1 4 2 3 1 0 0
2 1 2 1
   
1 0 0 0 3 3 0 3 3
row  1 2
 −1
 1 2

−−−−−−→ 0 1 0 0 − 3 − 3  ; A =  0 − 3 − 3 
 
operations
0 0 1 13 − 32 0 1
3 −3
2
0
1
   
2 4 −2 1 0 0 1 2 −1 2 0 0
row
22. 4 2 −2 0 1 0 −−−−−−→ 0 1 − 31 1 1
3 − 6 0 ; A is singular.
   
operations
8 10 −6 0 0 1 0 0 0 −2 −1 1
   
−1 3 0 1 0 0 1 −3 0 −1 0 0
row
23.  3 −2 1 0 1 0 −−−−−−→ 0 1 1 1 1 0
0 1 2 0 0 1 operations 0 0 1 −1 −1 1
   
1 0 0 5 6 −3 5 6 −3
row
−−−−−−→ 0 1 0 2 2 −1 ; A−1 = 2 2 −1
operations 0 0 1 −1 −1 1 −1 −1 1
5 5
     
1 2 3 1 0 0 1 0 0 1 −2 8 1 −2 8
row
1 − 12  1 − 21 
    −1
 
24. 0 1 4 0 1 0 −−−
 −−−→ 
 operations 0 1 0 0  ; A = 0


1 1
0 0 8 0 0 1 0 0 1 0 0 8 0 0 8
   
1 2 3 1 1 0 0 0 1 2 3 1 1 0 0 0
5 1 1
   
−1 0 2 1 0 1 0 0 0 1 1 0 0
row 2 2 2
25.  −−−−−−→ 
   
 2 1 −3 0 0 0 1 0 operations 0 0 1 − 2 1 2
 
   3 3 −1 − 3 0
1 1 1
1 1 2 1 0 0 0 1 0 0 0 1 −2 1 2 2
1 0 0 0 − 12 − 32 − 61 7
 
6
1 1 4
 
0 1 0 0 1 −
row 3 3 3
−−−−−−→  ;

operations  1 1 1
0 0 1 0 0 − 3 − 3 3
0 0 0 1 − 12 1 1
2
1
2
 1
− 2 − 32 − 61 7

6
1 1 4
 
 1 −
3 3 3
A−1 = 
 
 0 −1 −1 1

 3 3 3
− 21 1 1
2
1
2
   
1 0 0 0 1 0 0 0 1 0 0 0 1 0 0 0
0 0 1 0 0 1 0 0 row 0 1 0 0 0 0 0 1
26. 0 0 0 1 0 0 1 0 operations 0 0 1 0 0 1 0 0 ;
  −−− −−−→  
0 1 0 0 0 0 0 1 0 0 0 1 0 0 1 0
 
1 0 0 0
 0 0 0 1
A−1 =  0 1 0 0

0 0 1 0
− 13 1
!
−1 −1 −1 3
27. (AB) = B A =
−1 10 3
 
−1 −4 20
28. (AB)−1 = B−1 A−1 =  2 6 −30
3 6 −32

−2 3
29. A = (A−1 )−1 =
3 −4
! ! ! !
1 2 5 −1 5 −2 5 −1
30. AT = ; (AT )−1 = 1 ; A−1 = 1 ; (A−1 )T = 1
4 10 −2 2 −1 2 −2 2

4 −3 4 −3 16 − 3x 0
31. Multiplying = . We see that the last matrix is
x −4 x −4 0 16 − 3x

1 0
I= when x = 5.
0 1

−1 sin θ − cos θ
32. A =
cos θ sin θ
√1 √1 √1
 
3 3 3
sin θ − cos θ √1
33. (a) AT = (b) AT =  − √12 
 
= A−1 0 2
= A−1
cos θ sin θ  
− √26 √1
6
√1
6
34. Since det A · det A−1 = det AA−1 = det I = 1, we see that det A−1 = 1/ det A. If A is
orthogonal, det A = det AT = det A−1 = 1/ det A and (det A)2 = 1, so det A = ±1.
35. Since A and B are nonsingular, det AB = det A · B 6= 0, and AB is nonsingular.
36. Suppose A is singular. Then det A = 0, det AB = det A · det B = 0, and AB is singular.
37. Since det A · det A−1 = det AA−1 = det I = 1, det A−1 = 1/ det A.
38. Suppose A2 = A and A is nonsingular. Then A2 A−1 = AA−1 , and A = I. Thus, if A2 = A,

either A is singular or A = I.
39. If A is nonsingular, then A−1 exists, and AB = 0 implies A−1 AB = A−1 0, so B = 0.
40. If A is nonsingular, A−1 exists, and AB = AC implies A−1 AB = A−1 AC, so B = C.

1 0 0 0
41. No, consider A = and B = .
0 0 0 1
42. From Theorem 8.5.1, det A = det AT . If A is a nonsingular matrix, then det A 6= 0 implies
det AT 6= 0 and therefore AT is nonsingular.
43. If A and B are n × n nonzero matrices and AB = 0, then from Theorem 8.5.6,
det AB = det A · det B = det 0 = 0.
Now suppose det A 6= 0 and det B = 0, then A is nonsingular. Then by Problem 39, B = 0
which is contrary to the assumption that B is nonzero. Similarly, if det A = 0 and det B 6= 0
then B is nonsingular and so by Problem 39, A = 0 which is contrary to the assumption that
A is nonzero. Hence we must have det A = 0 and det B = 0 implies A and B are singular.
44. A is nonsingular if a11 a22 a33 6= 0 or a11 , a22 , and a33 are all nonzero.
 
1/a11 0 0
A−1 =  0 1/a22 0 
0 0 1/a33
For any diagonal matrix, the inverse matrix is obtained by taking the reciprocals of the
diagonal entries and leaving all other entries 0.
1 1
! ! !
−1 3 3 −1
4 6
45. A = 2
; A = ; x1 = 6, x2 = −2
3 − 31 14 −2
2 1 1
! ! !
−1 3 6 −1
2 2 1 3
46. A = 1 1
; A = ; x1 = , x2 = −
−3 6 −5 − 32 2 2
1 3 3
! ! !
16 8 6 4 3 1
47. A−1 = ; A−1 = ; x1 = , x2 = −
− 18 1
4 1 − 12 4 2
! ! !
−1 −2 1 −1
4 −11 15
48. A = 3 ; A = ; x1 = −11, x2 =
2 − 21 −3
15
2
2
− 51 1 1
     
5 −4 5 2
49. A−1 =  −1 1 −1 
0 ; A  0 =  4 ; x1 = 2, x2 = 4, x3 = −6
    
6 1 1
5 −5 −5 6 −6
 5 1 1
− 12
    
12 − 12 4 1
 2 1
     1 3
50. A−1 = −3 3 0 −1 
 ; A  2 =  0 ; x1 = − 2 , x2 = 0, x3 = 2
  
1 5 1 3
− 12 12 − 4 −3 2
    
−2 −3 2 1

21
 1 1
   
51. A−1 =  4 −4 −1 
0 ; A −3 =  1 ; x1 = 21, x2 = 1, x3 = −11
   
5 7
−1 7 −11
4 4
  
  
2 −1 1 1 2 1
−1 2 −1 −1  1  2
 
52. A−1 =
 1 −1
; A−1 
−5 = −1 ;

1 1
1 −1 1 0 3 −4
x1 = 1, x2 = 2, x3 = −1, x4 = −4
1 1 9
! ! !
5

7 −2 x1 b 10 10 10
53. = 1 ; A −1
= ; X=A −1
=
3 2 x2 b2 3
− 20 7
4 13
20 20

10 6
−1 −1 0 −2
X=A = ; X=A =
50 16 −20 −7
          
1 2 5 x1 b1 2 −1 −1 −1 −12
54.  2 3 8 x2  = b2  ; A−1 =  12 −7 −2 ; X = A−1  4  = −52 ;
−1 1 2 x3 b3 −5 3 1 6 23
       
3 0 0 1
X = A−1 3 =  9 ; A−1 −5 =  27
3 −3 4 −11
55. det A = 18 6= 0, so the system has only the trivial solution.
56. det A = 0, so the system has a nontrivial solution.

8.7 Cramer’s Rule 575
57. det A = 0, so the system has a nontrivial solution.
58. det A = 12 6= 0, so the system has only the trivial solution.

    
1 1 1 i1 0
59. (a) −R1 R2 0  i2  = E2 − E1 
0 −R2 R3 i3 E3 − E2
(b) det A = R1 R2 + R1 R3 + R2 R3 > 0, so A is nonsingular.

 
R2 R3 −R2 − R3 −R2
1
(c) A−1 = R2 R3 R3 −R1  ;
R1 R2 + R1 R3 + R2 R3
R1 R2 R2 R1 + R2
   
0 R2 E1 − R2 E3 + R3 E1 − R3 E2
1
A−1 E2 − E1  =  R1 E2 − R1 E3 − R3 E1 + R3 E2 
R1 R2 + R1 R3 + R2 R3
E3 − E2 −R1 E2 + R1 E3 − R2 E1 + R2 E3
60. (a) We write the equations in the form
−4u1 + u2 + u4 = −200
u1 − 4u2 + u3 = −300
u2 − 4u3 + u4 = −300
u1 + u3 − 4u4 = −200
    
−4 1 0 1 u1 −200
 1 −4 1 0 u2  = −300.
   
In matrix form this becomes 
 0 1 −4 1 u3  −300
1 0 1 −4 u4 −200
     
7 1 1 1 225
− 24 − 12 − 24 − 12 −200 2
 1 7 1 1 
    275 
− − 24 − 12 − 24 −300  
 12   2 
(b) A−1 = 1 ; A−1   =  ;
 
− 1 7 1  −300  275 
 24 − 12 − 24 − 12     2 
1 1 1 7 225
− 12 − 24 − 12 − 24 −200 2
225 275
u1 = u4 = , u2 = u3 =
2 2
8.7 Cramer’s Rule
1. det A = 10, det A1 = −6, det A2 = 12; x1 = −6

10 = − 35 , x2 = 12
10 = 6
5
−6 −6
2. det A = −3, det A1 = −6, det A2 = −6; x1 = −3 = 2, x2 = −3 =2
0.03 −0.09
3. det A = 0.3, det A1 = 0.03, det A2 = −0.09; x1 = 0.3 = 0.1 , x2 = 0.3 = −0.3
−0.00315
4. det A = −0.015, det A1 = −0.00315, det A2 = −0.00855; x1 = −0.015 = 0.21,
x2 = −0.00855
−0.015 = 0.57
5. det A = 1, det A1 = 4, det A2 = −7; x = 4, y = −7
1 35
6. det A = −70, det A1 = −14, det A2 = 35; r = −14
−70 = 5 , s= −70 = − 12
−44 44
7. det A = 11, det A1 = −44, det A2 = 44, det A3 = −55; x1 = 11 = −4, x2 = 11 = 4,
x3 = −55
11 = −5
8. det A = −63, det A1 = 173, det A2 = −136, det A3 = − 61 173

2 ; x1 = − 63 , x2 =
136
63 ,
61
x3 = 126
48 18 3
9. det A = −12, det A1 = −48, det A2 = −18, det A3 = −12; u = 12 = 4, v = 12 = 2 ,
w=1
10. det A = 1, det A1 = −2, det A2 = 2, det A3 = 5; x = −2, y = 2, z = 5
12 − 7k 6 − 7k
11. det A = 6 − 5k, det A1 = 12 − 7k, det A2 = 6 − 7k; x1 = , x2 = . The
6 − 5k 6 − 5k
system is inconsistent for k = 6/5.
ǫ−2 ǫ−1−1 1
12. (a) det A = ǫ − 1, det A1 = ǫ − 2, det A2 = 1; x1 = = = 1− ,
ǫ−1 ǫ−1 ǫ−1
1
x2 =
ǫ−1
(b) When ǫ = 1.01, x1 = −99 and x2 = 100. When ǫ = 0.99, x1 = 101 and x2 = −100.
289.8 271.9
13. det A ≈ 0.6428, det A1 ≈ 289.8, det A2 ≈ 271.9; x1 ≈ 0.6428 ≈ 450.8, x2 ≈ 0.6428 ≈ 423
14. We have (sin 30◦ )F + (sin 30◦ )(0.5N ) + N sin 60◦ = 400 and
(cos 30◦ )F + (cos 30◦ )(0.5N ) − N cos 60◦ = 0. The system is
(sin 30◦ )F + (0.5 sin 30◦ + sin 60◦ )N = 400

(cos 30◦ )F + (0.5 cos 30◦ − cos 60◦ )N = 0.
det A ≈ −1, det A1 ≈ −26.795, det A2 ≈ −346.41; F ≈ 26.795, N ≈ 346.41
15. The system is
i1 + i2 − i3 = 0
r1 i1 − r2 i2 = E1 − E2
r2 i2 + Ri3 = E2
r1 E2 + r2 E1
det A = −r1 R − r2 R − r1 r2 , det A3 = −r1 E2 , −r2 E1 ; i3 =
r1 R + r2 R + r1 r2
8.8 The Eigenvalue Problem 577
8.8 The Eigenvalue Problem

4 2 −2 2 −2
1. K3 since = = (−1) ; λ = −1
5 1 5 −5 5
√
√ √

2 −1 1√ 2√ 1√
2. K1 and K2 since = = 2 ; λ= 2
2 −2 2− 2 −2 + 2 2 2− 2
√ √ √
√ 2+ 2 √

2 −1 2+ 2 2 +√2 2
= = 2 ; λ= 2
2 −2 2 2 2 2

6 3 −5 0 −5
3. K3 since = =0 ; λ=0
2 1 10 0 10

2 8 2 + 2i −4 + 4i 2 + 2i
4. K1 and K2 since = = 2i ; λ = 2i
−1 −2 −1 −2i −1
      
1 −2 2 4 12 4
5. K2 and K3 since −2
 1 −2  −4 = −12 = 3 −4 ;
    λ=3
2 2 1 0 0 0
    
1 −2 2 −1 −1
−2 1 −2  1 =  1 ; λ = 1
2 2 1 1 1
      
−1 1 0 1 3 1
6. K2 since  1 2 1 4 = 12 = 3 4 ; λ=3
0 3 −1 3 9 3

−1 − λ 2
7. We solve det(A − λI) = = (λ − 6)(λ − 1) = 0.
−7 8 − λ

−7 2 0 1 −2/7 0
For λ1 = 6 we have or
−7 2 0 0 0 0

2 2
so that k1 = 7 k2 . If k2 = 7 then K1 = . For λ2 = 1 we have
7

−2 2 0 1 −1 0
or
−7 7 0 0 0 0

1
so that k1 = k2 . If k2 = 1 then K2 = .
1

2 − λ 1
8. We solve det(A − λI) = = λ(λ − 3) = 0.
2 1 − λ

2 1 0 1 1/2 0
2 1 0 0 0 0

−1
so that k1 = − 21 k2 . If k2 = 2 then K1 = . For λ2 = 3 we have
2

−1 1 0 1 −1 0
or
2 −2 0 0 0 0

1
1

8 − λ −1
9. We solve det(A − λI) = = (λ + 4)2 = 0.
16 −λ

−4 −1 0 1 1/4 0
For λ1 = λ2 = −4 we have or
16 4 0 0 0 0

−1
so that k1 = − 41 k2 . If k2 = 4 then K1 = .
4

1 − λ 1
10. We solve det(A − λI) = = (λ − 3/2)(λ − 1/2) = 0.
1/4 1 − λ

−1/2 1 0 1 −2 0
For λ1 = 3/2 we have or
1/4 −1/2 0 0 0 0

2
so that k1 = 2k2 . If k2 = 1 then K1 = . For λ2 = 1/2 we have
1

1/2 1 0 1 2 0
or
1/4 1/2 0 0 0 0

−2
so that k1 = −2k2 . If k2 = 1 then K2 = .
1

−1 − λ 2
11. We solve det(A − λI) = = λ2 + 9 = (λ − 3i)(λ + 3i) = 0.
−5 1 − λ

−1 − 3i 2 0 1 −(1/5) + (3/5)i 0
For λ1 = 3i we have or
−5 1 − 3i 0 0 0 0

1 − 3i
so that k1 = 15 − 35 i k2 . If k2 = 5 then K1 =

. For λ2 = −3i we have
5

−1 + 3i 2 0 1 −(1/5) − (3/5)i 0
or
−5 1 + 3i 0 0 0 0

1 1 + 3i
+ 35 i k2 . If k2 = 5 then K2 =

so that k1 = 5 .
5

1 − λ −1
12. We solve det(A − λI) = = λ2 − 2λ + 2 = 0.
1 1 − λ

i −1 0 i −1 0
For λ1 = 1 − i we have or
1 i 0 0 0 0

−i i
so that k1 = −ik2 . If k2 = 1 then K1 = and K2 = K1 = .
1 1

4 − λ 8
13. We solve det(A − λI) = = (λ − 4)(λ + 5) = 0.
0 −5 − λ

0 8 0 0 1 0
0 −9 0 0 0 0

1
so that k2 = 0. If k1 = 1 then K1 = . For λ2 = −5 we have
0

9 8 0 1 8/9 0
or
0 0 0 0 0 0

−8
so that k1 = − 98 k2 . If k2 = 9 then K2 = .
9

7 − λ 0
14. We solve det(A − λI) = = (λ − 7)(λ − 13) = 0.
0 13 − λ

0 0 0 0 1 0
0 6 0 0 0 0

1
so that k2 = 0. If k1 = 1 then K1 = . For λ2 = 13 we have
0

−6 0 0 1 0 0
or
0 0 0 0 0 0

0
so that k1 = 0. If k2 = 1 then K2 = .
1
 
5 − λ −1 0 4−λ −1 0

15. We solve det(A − λI) = 0 −5 − λ 9 = 4 − λ −5 − λ 9 
5 −1 −λ 4−λ −1 −λ
= λ(4 − λ)(λ + 4) = 0.
   
5 −1 0 0 1 0 −9/25 0
For λ1 = 0 we have 0 −5 9 0 or 0 1 −9/5 0
5 −1 0 0 0 0 0 0
 
9
9
so that k1 = 25 k3 and k2 = 95 k3 . If k3 = 25 then K1 = 45. For λ2 = 4 we have
25
   
1 −1 0 0 1 0 −1 0
0 −9 9 0 or 0 1 −1 0
5 −1 −4 0 0 0 0 0
 
1
so that k1 = k3 and k2 = k3 . If k3 = 1 then K2 = 1. For λ3 = −4 we have

1
   
9 −1 0 0 1 0 −1 0
0 −1 9 0 or 0 1 −9 0
5 −1 4 0 0 0 0 0
 
1
so that k1 = k3 and k2 = 9k3 . If k3 = 1 then K3 = 9.
1

3 − λ 0 0

16. We solve det(A − λI) = 0 2−λ 0 = (3 − λ)(2 − λ)(1 − λ) = 0.
4 0 1 − λ
   
2 0 0 0 1 0 0 0
For λ1 = 1 we have 0 1 0 0 or 0 1 0 0
4 0 0 0 0 0 0 0
 
0
so that k1 = 0 and k2 = 0. If k3 = 1 then K1 = 0. For λ2 = 2 we have

1
   
1 0 0 0 1 0 0 0
0 0 0 0 or 0 0 1 0
4 0 −1 0 0 0 0 0
 
0

0
   
0 0 0 0 1 0 −1/2 0
0 −1 0 0 or 0 1 0 0
4 0 −2 0 0 0 0 0
 
1
so that k1 = 12 k3 and k2 = 0. If k3 = 2 then K3 = 0.
2

−λ 4 0
0 = −(λ + 2)3 = 0.

17. We solve det(A − λI) = −1 −4 − λ

0 0 −2 − λ
   
2 4 0 0 1 2 0 0
For λ1 = λ2 = λ3 = −2 we have −1 −2 0 0 or 0 0 0 0
0 0 0 0 0 0 0 0
so that k1 = −2k2 . If k2 = 1 and k3 = 1 then
   
−2 0
K1 =  1 and K2 = 0 .

0 1

1 − λ 6 0 1 − λ 6 0

18. We solve det(A − λI) = 0 2−λ 1 = 0 3 − λ 3 − λ
0 1 2−λ 0 1 2 − λ
= (3 − λ)(1 − λ)2 = 0.
   
−2 6 0 0 1 0 −3 0
For λ1 = 3 we have  0 0 0 0 or 0 1 −1 0
0 1 −1 0 0 0 0 0
 
3
so that k1 = 3k3 and k2 = k3 . If k3 = 1 then K1 = 1. For λ2 = λ3 = 1 we have
1
   
0 6 0 0 0 1 0 0
0 1 1 0 or 0 0 1 0
0 1 1 0 0 0 0 0
 
1
so that k2 = 0 and k3 = 0. If k1 = 1 then K2 = 0.

0

−λ 0 −1
0 = −(λ + 1)(λ2 + 1) = 0.

19. We solve det(A − λI) = 1 −λ
1 1 −1 − λ
   
1 0 −1 0 1 0 −1 0
For λ1 = −1 we have 1 1 0 0 or 0 1 1 0
1 1 0 0 0 0 0 0
 
1
so that k1 = k3 and k2 = −k3 . If k3 = 1 then K1 = −1. For λ2 = i we have
1
   
−i 0 −1 0 1 0 −i 0
 1 −i 0 0 or 0 1 −1 0
1 1 −1 − i 0 0 0 0 0
   
i −i
so that k1 = ik3 and k2 = k3 . If k3 = 1 then K2 = 1 and K3 = K2 =
   1.
1 1

2 − λ −1 0
4 = −λ3 + 6λ2 − 13λ + 10

20. We solve det(A − λI) = 5
2−λ
0 1 2 − λ
= (λ − 2)(−λ2 + 4λ − 5) = (λ − 2)(λ − (2 + i))(λ − (2 − i)) = 0.

   
0 −1 0 0 1 0 4/5 0
For λ1 = 2 we have 5 0 4 0 or 0 1 0 0
0 1 0 0 0 0 0 0
 
−4
so that k1 = − 45 k3 and k2 = 0. If k3 = 5 then K1 =  0. For λ2 = 2 + i we have
5
   
−i −1 0 0 1 −i 0 0
 5 −i 4 0 or 0 1 −i 0
0 1 −i 0 0 0 0 0

−i
so that k1 = ik2 and k2 = ik3 . If k3 = i then K3 = −1. For λ3 = 2 − i we have
i
   
i −1 0 0 1 i 0 0
5 i 4 0 or 0 1 i 0
0 1 i 0 0 0 0 0
 
−1
so that k1 = −ik2 and k2 = −ik3 . If k3 = i then K3 =  1.
i

1 − λ 2 3

21. We solve det(A − λI) = 0 5−λ 6 = −(λ − 1)(λ − 5)(λ + 7) = 0.
0 0 −7 − λ
   
0 2 3 0 0 1 0 0
For λ1 = 1 we have 0 4 6 0 or 0 0 1 0
0 0 −6 0 0 0 0 0
 
1

0
   
−4 2 3 0 1 −1/2 0 0
 0 0 6 0 or 0 0 1 0
0 0 −12 0 0 0 0 0
 
1
so that k3 = 0 and k2 = 2k1 . If k1 = 1 then K2 = 2. For λ3 = −7 we have
0
   
8 2 3 0 1 0 1/4 0
0 12 6 0 or 0 1 1/2 0
0 0 0 0 0 0 0 0
 
−1
so that k1 = − 14 k3 and k2 = − 21 k3 . If k3 = 4 then K3 = −2.
4

−λ 0 0
0 = λ2 (λ − 1) = 0.

22. We solve det(A − λI) = 0 −λ
0 0 1 − λ
   
0 0 0 0 0 0 1 0
For λ1 λ2 = 0 we have 0 0 0 0 or 0 0 0 0
0 0 1 0 0 0 0 0
 
1
so that k3 = 0. If k1 = 1 and k2 = 0 then K1 = 0 and if k1 = 0 and k2 = 1 then

0
 
0
K2 = 1. For λ3 = 1 we have
0
   
−1 0 0 0 1 0 0 0
 0 −1 0 0 or 0 1 0 0
0 0 0 0 0 0 0 0
 
0
so that k1 = 0 and k2 = 0. If k3 = 1 then K3 = 0.
1

5 − λ 1
23. We solve det(A − λI) = = λ2 − 10λ + 24 = 0.
1 5 − λ

1 1 0 1 1 0
1 1 0 0 0 0

1
so that k1 = −k2 . If k2 = −1 then K1 = . For λ2 = 6 we have
−1

−1 1 0 1 −1 0
or
1 −1 0 0 0 0

1
1
By Theorem 8.8.3, A−1 has eigenvalues λ1 = 41 and λ2 = 16 with correspoding eigenvectors

1 1
K1 = and K2 = respectively.
−1 1

4 − λ 2
24. We solve det(A − λI) = = (λ − 6)(λ + 3) = 0.
7 −1 − λ

−2 2 0 1 −1 0
7 −7 0 0 0 0

1
so that k1 = k2 . If k2 = 1 then K1 = . For λ2 = −3 we have
1

7 2 0 1 2/7 0
or
7 2 0 0 0 0

−2
so that k1 = − 27 k2 . If k2 = −7 then K2 = .
7
1
By Theorem 8.8.3, A−1 has eigenvalues λ1 = 6 and λ2 = − 31 with correspoding eigenvectors

1 −2
K1 = and K2 = respectively.
1 7
25. By Theorem 8.8.4, the eigenvalues of the

 given
 upper triangular matrixare λ1 = 4, λ2 = 3,
1 −2
and λ3 = 5. For λ1 = 4 we get K1 = 0. For λ2 = 3 we get K2 =  1. For λ3 = 5,
0 0
 
−3
K3 = −1. By Theorem 8.8.3, A−1 has eigenvalues λ1 = 14 , λ2 = 13 , and λ3 = 51 with
1
     
1 −2 −3
corresponding eigenvectors K1 = 0, K2 =  1, and K3 = −1 respectively.
0 0 1
26. The eigenvalues

  of the given matrix are λ1 =  1, λ2 = 2, and λ3 = 3. Forλ1 = 1 we get
−1 −2 −1
K1 =  1. For λ2 = 2 we get K2 =  1. For λ3 = 3, K3 =  1. By Theorem
2 4 4
8.8.3, A−1 has eigenvalues λ1 = 1, λ2 = 12 , and λ3 = 31 with corresponding eigenvectors
     
−1 −2 −1
K1 =  1, K2 =  1, and K3 =  1 respectively.
2 4 4
27. Expand det(A − λI) along the row or column with the most zeros in it so for example
 
a11 a12 a13 a11 − λ
a12 a13
A =  0 a22 a23  gives det(A − λI) = 0 a22 − λ a23
0 0 a33 0 0 a33 − λ
Expanding along the first column gives us

a11 − λ a12 a 13
a22 − λ a23
det(A − λI) = 0
a22 − λ a23 = (a11 − λ)

0 0 a33 − λ
0 a33 − λ
= (a11 − λ)(a22 − λ)(a33 − λ)
The corresponding eigenvalues are λ1 = a11 , λ2 = a22 , and λ3 = a33 and these are the entries
on the main diagonal.
28. This is true. Since λ is an eigenvalue, then det(A − λI) = 0 which tells us the matrix A − λI
is singular.
29. (a) The eigenvalue λ has corresponding eigenvector K which, by definition 8.8.1, means
AK = λK, so
AK = λK
A (AK) = λAK
A2 K = λ(λK)
A2 K = λ 2 K
According to the last equation, λ2 is an eigenvalue of A2 and has the same corresponding
eigenvector K.

3
(b) The eigenvalues λ1 = 7 and λ2 = −1 with corresponding eigenvectors K1 = and
5

−1
K2 = .
1
By part (a), the eigenvalues and eigenvectors of A2 are

2 3
λ1 = 49; K1 =
5

−1
λ22 = 1; K2 =
1
So to check notice that

2 19 18
3 147 3
A K1 = = = 49 = λ21 K1
30 31
5 245 5

2 19 18 −1 −1 −1
A K2 = = =1 = λ22 K2
30 31 1 1 1
(c) Repeating the same procedure as in part (a), we can continue to multiply both sides of
the equation by A and simplifying,
A2 K = λ2 K
A A2 K = λ2 AK

A3 K = λ2 (λK)
A3 K = λ3 K
..
.
An K = λ n K
30. If an n × n matrix B is similar to the matrix A, then there exists a nonsingular matrix S
such that B = S−1 AS. But since S and S−1 are nonsingular we can write
SB = SS−1 AASB = AS
A = SBS−1 .
−1
The last expression can be written A = S−1 BS−1 and so A is similar to B.
31. If A and B are similar, then the matrices have the same characteristic polynomial. To see
this we write

det(B − λI) = det S−1 AS − λI = det S−1 AS − λS−1 S = det S−1 (A − λI)S
= det S−1 · det(A − λI) · det S ←− det S−1 = 1/ det S
= det(A − λI).
1 2
!
3 3
32. (a) For A1 = 1 3
, all entries are nonnegative numbers, and each row sums to 1. There-
4 4
fore, A is stochastic.
1 1 1
2 4 4
1 1 1

For A = 
3 3 3 , all entries are nonnegative numbers, and each row sums to 1.
1 1 1
6 3 2
Therefore, A is stochastic.
1 1 1 1
+ √ , and λ3 = − √ .
(b) For A from part (a), λ1 = 1, λ2 =
6 6 2 6 6 2
 

9 15 √  
9 15 15 √ 
1  17 − 34 2 + 2   17 − 34 − 34 2 − 2 
     
  
30
K1 = 1 , K2 = − + 8 √ 
, and K =

30 8 √ 
3
     
 17 17 2 + 2   − 17 + 17 2 − 2 
  
 
   
1 1 1
From this and additional stochastic matrices, we find that λ1 = 1 is a dominant

eigenvalue of stochastic matrix A with the remaining eigenvalues being
1 = λ1 ≥ λ2 ≥ · · · ≥ λn . Another characteristic of stochastic matrices is that the
eigenvector for λ1 = 1 has either only positive or only negative values.
(c) Using CAS to calculate An for the previous matrices, we see that the matrix elements of
stochastic matrices raised to the power n, tend to converge, and the rows become almost
identical.
8.9 Powers of Matrices
1. The characteristic equation is λ2 − 6λ + 13 = 0. Then

2 −7 −12 6 −12 13 0 0 0
A − 6A + 13I = − + = .
24 17 24 30 0 13 0 0
2. The characteristic equation is −λ3 + λ2 + 4λ − 1. Then

         
2 6 13 1 2 5 0 1 2 1 0 0 0 0 0
3 2
−A + A + A − I = 4 5  17 + 0 4 5 + 4 1 0 3 + 0 1 0 = 0 0 0 .
1 5 9 1 1 4 0 1 1 0 0 1 0 0 0
8.9 Powers of Matrices 587
3. The characteristic equation is λ2 − 3λ − 10 = 0, with eigenvalues −2 and 5. Substituting the

eigenvalues into λm = c0 + c1 λ generates
(−2)m = c0 − 2c1
5m = c0 + 5c1 .
Solving the system gives

1 1
c0 = [5(−2)m + 2(5)m ], c1 = [−(−2)m + 5m ] .
7 7
Thus
1 3
!
3(−1)m 2m+1 + 5m [−(−2)m + 5m ]

m 7 7
A = c0 I + c1 A = 2 1
7 [−(−2)m + 5m ] 7 [(−2)m + 6(5)m ]
and

3 11 57
A = .
38 106
4. The characteristic equation is λ2 − 10λ + 16 = 0, with eigenvalues 2 and 8. Substituting the

2m = c0 + 2c1
8m = c0 + 8c1 .

1 m+2 1
c0 = (2 − 8m ), c1 = (−2m + 8m ).
3 6
Thus
1 1
!
m 2 (2m + 8m ) 2 (2m − 8m )
A = c0 I + c1 A = 1 1
2 (2m − 8m ) 2 (2m + 8m )
and

4 2056 −2040
A = .
−2040 2056
5. The characteristic equation is λ2 − 8λ − 20 = 0, with eigenvalues −2 and 10. Substituting

the eigenvalues into λm = c0 + c1 λ generates
(−2)m = c0 − 2c1
10m = c0 + 10c1 .
1 1
c0 = [5(−2)m + 10m ], c1 = [−(−2)m + 10m ].
6 12
Thus
1 5
!
(−2)m + 2m 5m+1 m + 10m ]

m 6 12 [−(−2)
A = c0 I + c1 A = 1 1
3 [−(−2)m + 10m ] 6 [5(−2)
m + 10m ]
and

5 83328 41680
A =
33344 16640
6. The characteristic equation is λ2 + 4λ + 3 = 0, with eigenvalues −3 and −1. Substituting the

(−3)m = c0 − 3c1
(−1)m = c0 − c1 .
1 1
c0 = [−(−3)m + 3(−1)m ], c1 = [−(−3)m + (−1)m ].
2 2
Thus
(−1)m −(−3)m + (−1)m

m
A = c0 I + c1 A =
0 (−3)m
and

6 1 −728
A = .
0 729
7. The characteristic equation is −λ3 +2λ2 +λ−2 = 0, with eigenvalues −1, 1, and 2. Substituting
the eigenvalues into λm = c0 + c1 λ + c2 λ2 generates
(−1)m = c0 − c1 + c2
1 = c0 + c1 + c2
2m = c0 + 2c1 + 4c2 .
1 1 1
c0 = [3 + (−1)m − 2m ], c1 = [1 − (−1)m ], c2 = [−3 + (−1)m + 2m+1 ].
3 2 6
Thus
 
1 −1 + 2m −1 + 2m
Am = c0 I + c1 A + c2 A2 =  1
− 23 [(−1)m − 2m ]
 
(−1)m + 2m+1

0 3 
1 1
0 3 [−(−1)m + 2m ] 3 [2(−1)m + 2m ]
and
 
1 1023 1023
A10 = 0 683 682 .
0 341 342
√ √
8. The characteristic equation is −λ3 − λ2 + 2λ + 2 = 0, with eigenvalues −1, − 2 , and 2 .
Substituting the eigenvalues into λm = c0 + c1 λ + c2 λ2 generates
(−1)m = c0 − c1 + c2
√ √
(− 2 )m = c0 − 2 c1 + 2c2
√ √
( 2 )m = c0 + 2 c1 + 2c2 .

√ √ √ √
c0 = [2 − ( 2 )m−1 − ( 2 )m−2 ](−1)m + ( 2 − 1)( 2 )m−2 ,
1 √
c1 = [1 − (−1)m ]( 2 )m−1 ,
2
1 √ √ 1 √ √
c2 = (−1)m+1 + (1 + 2 )(−1)m ( 2 )m−1 + ( 2 − 1)( 2 )m−1 .
2 2
Thus Am = c0 I + c1 A + c2 A2 and
 
1 0 7
A6 = 7 8 −7 .
0 0 8
9. The characteristic equation is −λ3 + 3λ2 + 6λ − 8 = 0, with eigenvalues −2, 1, and 4.

(−2)m = c0 − 2c1 + 4c2

1 = c0 + c1 + c2
4m = c0 + 4c1 + 16c2 .

1
c0 = [8 + (−1)m 2m+1 − 4m ],
9
1
c1 = [4 − 5(−2)m + 4m ],
18
1
c2 = [−2 + (−2)m + 4m ].
18
Thus
1 1

(−2)m + (−1)m 2m+1 + 3 · 22m+1 [−(−2)m + 4m ]

9 3 0
Am = c0 I + c1 A + c2 A2 =  − 23 [(−2)m − 4m ] 1
 m m+1 + 4m

 3 (−1) 2 0

1 m + 22m+1 1 m m

3 −3 + (−2) 3 [−(−2) + 4 ] 1
and
 
699392 349184 0
A10 = 698368 350208 0
699391 349184 1
10. The characteristic equation is −λ3 − 23 λ2 + 32 λ + 1 = 0, with eigenvalues −2, − 12 , and 1.

(−2)m = c0 − 2c1 + 4c2

1 m

1 1
− = c0 − c1 + c2
2 2 4
1 = c0 + c1 + c2 .
1
c0 = [2−m [(−4)m + 8(−1)m + 2m+1 − (−1)m 22m+1 ],
9
1
c1 = − 2−m [(−4)m + 4(−1)m − 5 · 2m ],
9
2
c2 = [1 + (−2)m − (−1)m 2m−1 ].
9
Thus Am = c0 I + c1 A + c2 A2
 1 −m 1
m 
−1 + − 12

32 [2(−1)m + 2m ] 3 0
 
1 m 1 m
2
1

= −1 + − 2 3 2 + −2 0
 
3

 
1 m
− 91 2−m 7(−4)m − 6(−1)m − 3 · 2m + (−1)m 22m+1 1 1

3 −1 + − 2 3 (−2)m + (−1)m 2m+1
and
43 85
 
128 − 256 0
A8 =  − 85 171
 
 128 256 0
32725 85
− 128 − 256 256
11. The characteristic equation is λ2 − 8λ + 16 = 0, with eigenvalues 4 and 4. Substituting the

4m = c0 + 4c1
4m−1 m = c1 .
c0 = −4m (m − 1), c1 = 4m−1 m.
Thus
4m−1 (3m + 4) 3 · 4m−1 m

m
A = c0 I + c1 A = m−1 m−1
−3 · 4 m 4 (−3m + 4)
and

6 22528 18432
A = .
−18432 −14336
12. The characteristic equation is −λ3 − λ2 + 21λ + 45 = 0, with eigenvalues −3, −3, and 5.
(−3)m = c0 − 3c1 + 9c2

(−3)m−1 m = c1 − 6c2
5m = c0 + 5c1 + 25c2 .

1
c0 = [73(−3)m − 2(−1)m 3m+2 + 9 · 5m − 40(−3)m m],
64
1
c1 = [−(−1)m 3m+2 + 9 · 5m − 8(−3)m m],
96
1
c2 = [−(−3)m + 5m − 8(−3)m−1 m].
64
Thus Am = c0 I + c1 A + c2 A2
 
1 1 1

31(−3)m − (−1)m 3m+1 + 4 · 5m −(−3)m − (−1)m 3m+1 + 4 · 5m (−3)m + (−1)m 3m+1 − 4 · 5m
 32 16 32 
 
=  1 −(−3)m − (−1)m 3m+1 + 4 · 5m 1 1

7(−3)m − (−1)m 3m+1 + 4 · 5m (−3)m + (−1)m 3m+1 − 4 · 5m 

 16 8 16 
 
3 3 1

32
(−3)m + (−1)m 3m+1 − 4 · 5m 16
(−3)m + (−1)m 3m+1 − 4 · 5m 32
29(−3)m − (−1)m 3m+2 + 12 · 5m
and
 
178 842 −421
A5 =  842 1441 −842
−1263 −2526 1020
13. (a) The characteristic equation is λ2 − 4λ = λ(λ − 4) = 0, so 0 is an eigenvalue. Since the

matrix satisfies the characteristic equation, A2 = 4A, A3 = 4A2 = 42 A, A4 = 42 A2 =
43 A, and, in general,
m
4m

m m 4
A =4 A=
3(4)m 3(4)m
(b) The characteristic equation is λ2 = 0, so 0 is an eigenvalue. Since the matrix satisfies

the characteristic equation, A2 = 0, A3 = AA2 = 0, and, in general, Am = 0.
(c) The characteristic equation is −λ3 +5λ2 −6λ = 0, with eigenvalues 0, 2, and 3. Substitut-
ing λ = 0 into λm = c0 + c1 λ + c2 λ2 we find that c0 = 0. Using the nonzero eigenvalues,
we find
2m = 2c1 + 4c2
3m = 3c1 + 9c2 .

1 1
c1 = [9(2)m − 4(3)m ], c2 = [−3(2)m + 2(3)m ].
6 6
Thus Am = c1 A + c2 A2 and
 
2(3)m−1 3m−1 3m−1
 1 1 1
Am =  m m m m m

m 
 6 [9(2) − 4(3) ] 6 [3(2) − 2(3) ] 6 [−3(2) − 2(3) ]
1 m m 1 m m 1 m m
6 [−9(2) + 8(3) ] 6 [−3(2) + 4(3) ] 6 [3(2) + 4(3) ]
14. (a) Let

xn−1 1 1
Xn−1 = and A=
yn−1 0 1
Then

1 1 xn−1 xn−1 + yn+1
Xn = AXn−1 = = .
1 0 yn−1 xn−1
√
(b) The characteristic equation of A is λ2 − λ − 1 = 0, with eigenvalues λ1 = 12 (1 − 5 )
√
and λ2 = 21 (1 + 5 ). From λm = c0 + c1 λ we get λm m
1 = c0 + c1 λ1 and λ2 = c0 + c1 λ2 .
Solving this system gives
c0 = (λ2 λm m
1 − λ1 λ2 )/(λ2 − λ1 ) and c1 = (λm m
2 − λ1 )/(λ2 − λ1 ).
Thus
Am = c0 I + c1 A
√ m+1 √ √ √ !
1 (1 + 5) − (1 − 5 )m+1 2(1 + 5 )m − 2(1 − 5 )m
= √ √ √ √ √ √ √
2m+1 5 2(1 + 5 )m − 2(1 − 5 )m (1 + 5 )(1 − 5 )m − (1 − 5 )(1 + 5 )m
(c) From part (a), X2 = AX1 , X3 = AX2 = A2 X1 , X4 = AX3 = A3 X1 , and, in general,

Xn = An−1 X1 . With

1 11 144 89 1 233
X1 = we have X12 = A X1 = = ,
1 89 55 1 144
so the number of adult pairs is 233. With

1 11 144 89 1 144
X1 = we have A X1 = = ,
0 89 55 0 89
so the number of baby pairs is 144. With

2 11 144 89 2 377
X1 = we have X12 = A X1 = = ,
1 89 55 1 233
so the total number of pairs is 377.
1
15. The characteristic equation of A is λ2 −5λ+10 = 0, so A2 −5A+10I = 0 and I = − 10 A2 + 21 A.
Multiplying by A−1 we find
3 2
! ! !
1 1 1 2 −4 1 1 0 10 5
A−1 = − A + I = − = = .
10 2 10 1 3 2 0 1 −1 1 10 5
16. The characteristic equation of A is −λ3 + 2λ2 + λ − 2 = 0, so −A3 + 2A2 + A − 2I = 0 and

I = − 21 A3 + A2 + 21 A. Multiplying by A−1 we find
3 1 5

2 2 −2
1 1
A−1 = − A2 + A + I = 
1 1
− 21 

2 2  2 2 .
1 1 3
2 2 −2
17. (a) Since

2 1 0 m 1 0
A = we see that A =
−1 0 −1 0
for all integers m ≥ 2. Thus A is not nilpotent.
(b) Since A2 = 0, the matrix is nilpotent with index 2.
(c) Since A3 = 0, the matrix is nilpotent with index 3.
(d) Since A2 = 0, the matrix is nilpotent with index 2.
(e) Since A4 = 0, the matrix is nilpotent with index 4.
(f ) Since A4 = 0, the matrix is nilpotent with index 4.
18. (a) If Am = 0 for some m, then (det A)m = det Am = det 0 = 0, and A is a singular matrix.
(b) By (1) of Section 8.8 we have AK = λK, A2 K = λAK = λ2 K, A3 K = λ2 AK = λ3 K,

and, in general, Am K = λm K. If A is nilpotent with index m, then Am = 0 and
λm = 0.
8.10 Orthogonal Matrices
      
0 0 −4 0 0 0
1. (a)-(b)  0 −4 0 1 = −4 = −4 1 ; λ1 = −4
−4 0 15 0 0 0
      
0 0 −4 4 −4 4
 0 −4 0 0 =  0 = (−1) 0 ; λ2 = −1
−4 0 15 1 1 1
      
0 0 −4 1 16 1
 0 −4 0  0 =
  0 = 16
  0 ; λ3 = 16
−4 0 15 −4 −64 −4
   
4 1
(c) KT1 K2 = 0 1 0 0 = 0; KT1 K3 = 0 1 0  0  = 0;

1 −4
 
1
KT2 K3 = 4 0 1  0  = 0
−4
      
1 −1 −1 −2 −4 −2
2. (a)-(b)  −1 1 −1  1 =
  2 =2
  1 ; λ1 = 2
−1 −1 1 1 2 1
      
1 −1 −1 0 0 0
−1 1 −1  1 =  2 = 2  1 ; λ2 = 2
−1 −1 1 −1 −2 −1
      
1 −1 −1 1 −1 1
−1 1 −1  1 = −1 = (−1) 1 ;
    λ3 = −1
−1 −1 1 1 −1 1
 
0
(c) KT1 K2 = −2 1 1  1  = 1 − 1 = 0;

−1
 
1
KT1 K3 = −2 1 1 1 = −2 + 1 + 1 = 0;

1
 
T
1
K2 K3 = 0 1 −1 1 = 1 − 1 = 0
1
8.10 Orthogonal Matrices 595
   √2   √   √2 
5 13 0 2 9 2 2
  √   √  √ 
3. (a)-(b) 
13 5 0   = 9 2 = 18  2 
  2     ; λ1 = 18
2 2 
0 0 −8 0 0 0
 √   √   √ 
  3 8 3 3
5 13 0 3  − 3  3 
  √   √ 
 
 √ 

13 5
 0
 − 33  =  8 3 3  = (−8) − 33  ; λ2 = −8
 √   √   √ 
0 0 −8 3
− 38 3 3
3 3
   √6   8√6   √6 
5 13 0 6 − 6 6
 √   √   √ 
 − 6   8 6  6
13 5 0  =  6  = (−8) − 6  ; λ3 = −8

  √6   √   √ 
0 0 −8 − 36 8 6 − 36
3
 √ 
3
√ √  3
√  √6 √6
(c) KT1 K2 = 2 2 3
0 − 3  = − = 0;
 
2 2  √  6 6
3
3
 √ 
6
√ √  6
√  √12 √12
KT1 K3 = 2 2 6
0 − 6  = − = 0;
 
2 2  √  12 12
− 36
 √ 
6
6  √ √ √
√ √ √  √  18 18 18
KT2 K3 = 3
− 33 33 − 66  = + − =0

3  √  18 18 9
− 36
      
3 2 2 −2 0 −2
4. (a)-(b) 2 2 0  2 = 0 = 0  2 ; λ1 = 0
2 0 4 1 0 1
      
3 2 2 1 3 1
2 2 0  2 =  6 = 3  2 ; λ2 = 3
2 0 4 −2 −6 −2
      
3 2 2 2 12 2
2 2 0  1 = 6 = 6 1  ; λ3 = 6
   
2 0 4 2 12 2
 
1
T

(c) K1 K2 = −2 2 1  2  = −2 + 4 − 2 = 0;
−2
 
2
KT1 K3 = −2 2 1 1 = −4 + 2 + 2 = 0;

2
 
T
2
K2 K3 = 1 2 −2 1 = 2 + 2 − 4 = 0
2
5. Orthogonal. Columns form an orthonormal set.
6. Not orthogonal. Columns one and three are not unit vectors.
7. Orthogonal. Columns form an orthonormal set.
8. Not orthogonal. The matrix is singular.
9. Not orthogonal. Columns are not unit vectors.
10. Orthogonal. Columns form an orthogonal set.
√1 √1
! ! !
1 1 2 2
11. λ1 = −8, λ2 = 10, K1 = , K2 = , P= 1
−1 1 − √2 √1
2

1 0 1 0
12. λ1 = 7, λ2 = 4, K1 = , K2 = , P=
0 1 0 1
√3 √1
! ! !
3 1 10 10
13. λ1 = 0, λ2 = 10, K1 = , K2 = , P= 1
−1 3 − √10 √3
10
√ √ √ √
1 5 1 5 1+ 5 1− 5
14. λ1 = + , λ2 = − , K1 = , K2 = ,
2 2 2 2 2 2
 √ √ 
√1+ 5√ √1− 5√
 10+2 5 10−2 5 
P=
2 √ 2 √

√ √
10+2 5 10−2 5
 1
√1

− √2 0
     
−1 1 0 2
 
15. λ1 = 0, λ2 = 2, λ3 = 1, K1 =  0 , K2 = 0 , K3 = 1 , P =  0 0 1
     
 
1 √1
1 1 0 √ 0
2 2
     
−1 1 1
√ √  √  √ 
16. λ1 = −1, λ2 = 1 − 2, λ3 = 1 + 2, K1 =  0 , K2 = − 2 , K3 =  2 ,
 
1 1 1
 1 1 1

− √2 2 2
 √ √ 
P=  0 − 22 22  
√1 1 1
2 2 2
     
−3 1 1
17. λ1 = −11, λ2 = 0, λ3 = 6, K1 =  1 , K2 = −4 , K3 = 2 ,
     
1 7 1
 3
√1 √1

− √11 66 6
 
 √1 − 4 2 
P= 11
√
66
√
6
 
√1 √7 √1
11 66 6
     
1 −2 2
     
18. λ1 = −18, λ2 = 0, λ3 = 9, K1 = 
−2 , K2 =  1 , K3 = 2 ,
    
2 2 1
 1 2 2

3 −3 3
 2 1 2

P= − 3 3 3
2 2 1
3 3 3
3 3 4
! ! !
5 a 5 5 1 0 9 16
19. 4
= implies 25 + a2 = 1 and 25 + b2 = 1. These equations give
5 b a b 0 1
a = ± 54 , b = ± 53 . But 12
25 + ab = 0 indicates a and b must have opposite signs. Therefore
choose a = − 54 , b = 35 .
3 4
!
5 − 5
The matrix 4 3
is orthogonal.
5 5
√1 √1
! ! !
5
b 5
a 1 0 1 1
20. = implies 5 + b2 = 1 and a2 + 5 = 1. These equations give
a √1 b √1 0 1
5 5
a = ± √25 , b = ± √25 . But √a

5
+ √b
5
= 0indicates a and b must have opposite signs. Therefore
choose a = − √25 , b = √2 .
5
 1
√2

√
5 5
The matrix   is orthogonal.
− √25 √1
5
21. (a)-(b) We compute

      
0 2 2 1 −2 1
AK1 = 2 0 2 −1 =  2 = −2 −1 = −2K1
2 2 0 0 0 0
      
0 2 2 1 −2 1
AK2 = 2
 0 2   0 =
  0 = −2
  0 = −2K2
2 2 0 −1 2 −1
      
0 2 2 1 4 1
AK1 = 2 0 2 1 = 4 = 4 1 = 4K3
2 2 0 1 4 1
and observe that K1 is an eigenvector with corresponding eigenvalue −2, K2 is an eigen-

vector with corresponding eigenvalue −2, and K3 is an eigenvector with corresponding
eigenvalue 4.
(c) Since K1 · K2 = 1 6= 0, K1 and K2 are not orthogonal, while K1 · K3 = 0 and K2 · K3 = 0

so K3 is orthogonal to both K1 and K2 , To transform {K1 , K2 } into an orthogonal set
we let V1 = K1 and compute K2 · V1 = 1 and V1 · V1 = 2. Then
     1
1 1 2
K 2 · V1   1    1
V2 = K2 − V1 =  0 − −1 =  2 
    
V1 · V1 2 
−1 0 −1
Now, {V1 , V2 , K3 } is an orthogonal set of eigenvectors with

√ 3 √
||V1 || = 2, ||V2 || = √ , and ||K3 || = 3.
6
An orthonormal set of vectors is

√1 √1
 1     
√
2 6 3
     
 √1   1   1 
− 2  ,  √ ,
 6
and √ 
 3
     
0 − √26 √1
3
and so the matrix

√1 √1 √1
 
2 6 3
 
 1
√1 √1 

− 2
P= √
6 3
 
2 √1
0 − √6 3
is orthogonal.
22. (a)-(b) We compute

      
1 1 1 1 −1 0 −1
1 1 1 1  0 0
      0
AK1 =
1 =   = 0  = 0K1
1 1 1  0  0  0
1 1 1 1 1 0 1
      
1 1 1 1 −1 0 −1
1 1 1 1  0 0
      0
AK2 =
1 =   = 0  = 0K2
1 1 1  1  0  1
1 1 1 1 0 0 0
      
1 1 1 1 −1 0 −1
1 1 1 1  1 0  1 = 0K3
      
AK3 =
1 = = 0
1 1 1  0 0  0
1 1 1 1 0 0 0
      
1 1 1 1 1 4 1
1 1 1 1 1 4 1 = 4K4
      
AK4 =
1 = = 4
1 1 1 1 4 1
1 1 1 1 1 4 1
and observe that K1 is an eigenvector with corresponding eigenvalue 0, K2 is an eigenvec-

tor with corresponding eigenvalue 0, K3 is an eigenvector with corresponding eigenvalue
0, and K4 is an eigenvector with corresponding eigenvalue 4.
(c) Since K1 · K2 = 1 6= 0, K1 and K2 are not orthogonal. Similarly, K1 · K3 = 1 6= 0

and K2 · K3 = 1 6= 0 so K1 and K3 and K2 and K3 are not orthogonal. However,
K1 · K4 = 0, K2 · K4 = 0, and K3 · K4 = 0, so each of K1 , K2 , and K3 is orthogonal
to K4 . To transform {K1 , K2 , K3 } into an orthogonal set we let V1 = K1 and compute
K2 · V1 = 1 and V1 · V1 = 2. Then
   1
−1 −1 −2
 
K 2 · V1  0 1  0  0 
     
V2 = K2 − V1 =   −   =   .
V1 · V1  1 2  0  1 
0 1 − 12
Next, using K3 · V1 = 1, K3 · V2 = 12 , and V2 · V2 = 23 , we obtain

     1  1
−1 −1 −2 −3
       
K3 · V1 K 3 · V2  1 1  0 1/2  0  1
V3 = K 3 = V1 − V2 =   −   −   =  .
       
V1 · V1 V2 · V2  0 2  0 3/2  1 − 1 
       3
0 1 − 12 − 13
Now, {V1 , V2 , V3 , K4 } is an orthogonal set of eigenvectors with

√ 3 2
||V1 || = 2, ||V2 || = √ , ||K3 || = √ and ||K4 k = 2.
6 3
An orthonormal set of vectors is

 1 
− √16
   1  1
− √2 − 2√3 2
       
3  1

 0


 0

 √
 2 3 2
 ,  ,  , and  ,
√2  1  1
0 − 2 3 
     √
  6
2
      
√1 − √16 1
− 2√3 1
2 2
and so the matrix

1
− √16 1 1

− sqrt2 − 2√ 3 2
 
3 1

 0 0 √
2 3 2
P= 
0 √2 1
− 2√ 1


 6 3 2

√1 − √16 1
− 2√ 1
2 3 2
is orthogonal.
   
0 a
23. If we take K1 = 1 as in Example 4 in the text then we look for a vector K2 =  b  such
1 c
1 1
that 1(a) + 4 b− 4 c = 0and K1 · K2 = 0 or b + c = 0. The last equation implies c = −b so
a + 14 b − 41 (−b) = a + 21 b = 0. If we let b = −2, then a = 1 and c = 2, so a second eigenvector
 
1
with eigenvalue −9 and orthogonal to K1 is K2 = −2.
2
24. The eigenvalues and corresponding eigenvectors of A are
     
−1 0 0
1 0 0
λ1 = λ2 = −1, λ3 = λ4 = 3, and K1 =   0 ,
 K2 = 
−1 ,
 K3 = 
1 ,

0 1 1
 
1
1
K4 = 0 .

Since K1 · K2 = K1 · K3 = K1 · K4 = K2 · K3 = K2 · K4 = K3 · K4 = 0, the vectors are

√
orthogonal. Using kK1 k = kK2 k = kK3 k = kK4 k = 2 , we construct the orthogonal matrix
 1
0 √12

− √2 0
 
 √1 0 0 √12 
 2 
P= .
 
 0 − √12 √12 0 
 
 1 1 
0 √
2
√
2
0
8.11 Approximation of Eigenvalues 601
25. Suppose A and B are orthogonal matrices. Then A−1 = AT and B−1 = BT and
(AB)−1 = B−1 A−1 = BT AT = (AB)T .
Thus AB is an orthogonal matrix.
26. To show that A2 is orthogonal we must 27. To show that A−1 is orthogonal we must
−1 T −1 T
demonstrate that A2 = A2 . demostrate that A−1 = A−1 .
A2
−1 T
= AA A−1 = A = AT
−1
A2

= (AA)−1 = A−1 A−1
−1 T
A−1 = A−1 ← Since AT = A−1
−1 2
A2 = AT = AT AT
2 T
A−1 = A2 .
28. Since AT = A−1 , 29. Since A2 = I,
AAT = I AA = I
det AAT = 1

A = A−1
det A · det AT = 1 A = AT .
(det A)2 = 1
det A = ±1.
cos2 θ + sin2 θ

T cos θ − sin θ cos θ sin θ 0 1 0
30. AA = = =
sin θ cos θ − sin θ cos θ 0 sin2 θ + cos2 θ 0 1
= I.
Therefore AT = A−1 .
8.11 Approximation of Eigenvalues

1
1. Taking X0 = and computing Xi = AXi−1 for i = 1, 2, 3, 4 we obtain
1

2 4 8 16
X1 = , X2 = , X3 = , X4 = .
2 4 8 16

1
We conclude that a dominant eigenvector is K = with corresponding eigenvalue
1
AK · K 4
λ= = = 2.
K·K 2

1
2. Taking X0 = and computing Xi = AXi−1 for i = 1, 2, 3, 4, 5 we obtain
1

−5 49 −437 3937
X1 = ,
X2 = , X3 = , X4 = ,
7 −47 439 −3935

−35429
X5 = .
35431

1 −35429 −0.99994
We conclude that a dominant eigenvector is K = ≈ with
35439 35431 1
AK · K
corresponding eigenvalue λ = = −8.9998.
K·K

1 6 1 6 0.375
3. Taking X0 = and computing AX0 = , we define X1 = = .
1 16 16 16 1
Continuing in this manner we obtain

0.3363 0.3335 0.3333
X2 = , X3 = , X4 = .
1 1 1

0.3333
1
λ = 14.

1 1 1 1 0.2
4. Taking X0 = and computing AX0 = , we define X1 = = . Continuing
1 5 5 5 1
in this manner we obtain

0.2727 0.2676 0.2680 0.2679
X2 = , X3 = , X4 = , X5 = .
1 1 1 1

0.2679
1
λ = 6.4641.
       
1 11 11 1
1   
5. Taking X0 = 1 and computing AX0 = 11, we define X1 = 11 = 1 .
11
1 6 6 0.5455
Continuing in this manner we obtain
    

1 1 1
X2 =  1  , X3 =  1  , X4 =  1  .
0.5045 0.5005 0.5
 
1
We conclude that a dominant eigenvector is K =  1  with corresponding eigenvalue
0.5
λ = 10.
       
1 5 5 1
1   
6. Taking X0 = 1 and computing AX0 = 2 , we define X1 =
    2 = 0.4 . Contin-
5
1 2 2 0.4
uing in this manner we obtain
       
1 1 1 1
X2 = 0.2105 , X3 = 0.1231 , X4 = 0.0758 , X5 = 0.0481 .
0.2105 0.1231 0.0758 0.0481
   
1 1
At this point if we restart with X0 = 0 we see that K = 0 is a dominant eigenvector
  
0 0
with corresponding eigenvalue λ = 3.

1
7. Taking X0 = and using scaling we obtain
1

0.625 0.5345 0.5098 0.5028
X1 = , X2 = , X3 = , X4 = ,
1 1 1 1

0.5008
X5 = .
1

0.5
Taking K = as the dominant eigenvector we find λ1 = 7. Now the normalized
1

0.4472 1.6 −0.8 1
eigenvector is K1 = and B = . Taking X0 = and using
0.8944 −0.8 0.4 1

1 1 1
scaling again we obtain X1 = , X2 = . Taking K = we find λ2 = 2.
−0.5 −0.5 −0.5
The eigenvalues are 7 and 2.

1 0.3333 0.3333
8. Taking X0 = and using scaling we obtain X1 = , X2 = . Taking
1 1 1

1/3
K= as the dominant eigenvector we find λ1 = 10. Now the normalized eigenvector
1

0.3162 0 0
is K1 = and B = . An eigenvector for the zero matrix is λ2 = 0. The
0.9486 0 0
eigenvalues are 10 and 0.
 
1
9. Taking X0 = 1 and using scaling we obtain
1
       
1 1 1 1
X1 = 0 , X2 = −0.6667 , X3 = −0.9091 , X4 = −0.9767 ,
1 1 1 1
 
1
X5 = −0.9942 .
1
 
1
Taking K = −1 as the dominant eigenvector we find λ1 = 4. Now the normalized
1
     
0.5774 1.6667 0.3333 −1.3333 1
eigenvector is K1 = −0.5774 and B =  0.3333 0.6667 0.3333. If X0 = 1
0.5774 −1.3333 0.3333 1.6667 1
 
1
is now chosen only one more eigenvalue is found. Thus, try X0 = 1. Using scaling we

0
obtain
       
1 1 1 1
X1 =  0.5  , X2 =  0.2  , X3 =  0.0714  , X4 =  0.0244  ,
−0.5 −0.8 −0.9286 −0.9756
 
1
X5 =  0.0082  .
−0.9918
 
1
Taking K =  0 as the eigenvector we find λ2 = 3. The normalized eigenvector in this
−1
     
0.7071 0.1667 0.3333 0.1667 1
case is K2 =  0 and C = 0.3333 0.6667 0.3333. If X0 = 1 is chosen, and
−0.7071 0.1667 0.3333 0.1667 1
     
0.5 0.5 0.5
scaling is used we obtain X1 =  1 , X2 =  1 . Taking K =  1 we find λ3 = 1.
0.5 0.5 0.5
 
1
The eigenvalues are 4, 3, and 1. The difficulty in choosing X0 = 1to find the second

1
eigenvector results from the fact that this vector is a linear combination of the eigenvectors
corresponding to the other two eigenvalues, with 0 contribution from the second eigenvector.
When this occurs the development of the power method, shown in the text, breaks down.
 
1
10. Taking X0 = 1 and using scaling we obtain

1
       
−0.3636 −0.2431 −0.2504 −0.2499
X1 = −0.3636 , X2 =  0.0884  , X3 = −0.0221 , X4 = −0.0055 .
1 1 1 1
 
−0.25
Taking K =  0  as the dominant eigenvector we find λ1 = 16. The normalized
1
     
−0.2425 −0.9412 0 −0.2353 1
eigenvector is K1 =  0  and B =  0 −4 0  . Taking X0 = 1
0.9701 −0.2353 0 −0.0588 1
and using scaling we obtain

       
−0.2941 0.0735 −0.0184 0.0046
X1 =  −1  , X2 =  1  , X3 =  −1  , X4 =  1  .
−0.0735 0.0184 −0.0046 0.0011
 
0
Taking K = 1 as the eigenvector we find λ2 = −4. The normalized eigenvector in this
0
     
0 −0.9412 0 −0.2353 1
case is K2 = K = 1 and C =  0 0 0 . Taking X0 = 1 and using
0 −0.2353 0 −0.0588 1
     
−1 1 1
scaling we obtain X1 =  0 , X2 =  0 . Using K =  0 we find λ3 = −1. The
−0.25 0.25 0.25
eigenvalues are 16, −4, and −1.

4 −1 1
11. The inverse matrix is . Taking X0 = and using scaling we obtain
−3 1 1

1 1 1 1
X1 = , X2 = , X3 = , X4 = .
−0.6667 −0.7857 −0.7910 −0.7913

1 1 1
Using K = we find λ = 4.7913. The minimum eigenvalue of is
−0.7913 3 4
1/4.7913 ≈ 0.2087.

1 3 1
12. The inverse matrix is . Taking X0 = and using scaling we obtain
4 2 1

0.6667 0.7857 0.75
X1 = , X2 = , . . . , X10 = .
1 1 1

0.75 −0.2 0.3
Using K = we find λ = 5. The minimum eigenvalue of is 1/5 = 0.2
1 0.4 −0.1
13. (a) Replacing the second derivative with the difference expression we obtain
yi+1 − 2yi + yi−1
EI + P yi = 0 or EI(yi+1 − 2yi + yi−1 ) + P h2 yi = 0.
h2
(b) Expanding the difference equation for i = 1, 2, 3 and using h = L/4, y0 = 0, and y4 = 0
we obtain
P L2
EI(y2 − 2y1 + y0 ) + y1 = 0 P L2
16 2y1 − y2 = y1
16EI
P L2
EI(y3 − 2y2 + y1 ) + y2 = 0
or P L2
16 −y1 + 2y2 − y3 = y2
16EI
P L2
EI(y4 − 2y3 + y2 ) + y3 = 0 P L2
16 −y2 + 2y3 = y3 .
16EI
In matrix form this becomes

    
2 −1 0 y1 2 y1
−1 P L
2 −1 y2  = y2  .
16EI
0 −1 2 y3 y3
 
0.75 0.5 0.25
(c) A−1 =  0.5 1 0.5 
0.25 0.5 0.75
 
1
(d) Taking X0 = 1 and using scaling we obtain

1
       
0.75 0.7143 0.7083 0.7073
X1 =  1  , X2 =  1  , X3 =  1  , X4 =  1  ,
0.75 0.7143 0.7083 0.7073
 
0.7071
X5 =  1  .
0.7071
 
0.7071
Using K =  1  we find λ = 1.7071. Then 1/λ = 0.5859 is the minimum
0.7071
eigenvalue of A.
P L2 EI
(e) Solving = 0.5859 for P we obtain P = 9.3726 2 . In Example 3 of Section 3.9
16EI L
we saw
EI EI
P = π 2 2 ≈ 9.8696 2 .
L L
14. (a) The difference equation is
EIi (yi+1 − 2yi + yi−1 ) + P h2 yi = 0, i = 1, 2, 3,
where I0 = 0.00200, I1 = 0.00175, I2 = 0.00150, I3 = 0.00125, and I4 = 0.00100. The

system of equations is
P L2
0.00175E(y2 − 2y1 + y0 ) + y1 = 0
16
P L2
0.00150E(y3 − 2y2 + y1 ) + y2 = 0
16
P L2
0.00125E(y4 − 2y3 + y2 ) + y2 = 0
16
or
P L2
0.0035y1 − 0.00175y2 = y1
16E
P L2
−0.0015y1 + 0.003y2 − 0.0015y3 = y2
16E
P L2
−0.00125y2 + 0.0025y3 = y3 .
16E
In matrix form this becomes
    
0.0035 −0.00175 0 y1 2 y1
−0.0015 P L
0.003 −0.0015 y2  = y2  .
16E
0 −0.00125 0.0025 y3 y3
(b) The inverse of A is  

428.571 333.333 200
A−1 = 285.714 666.667 400 .
142.857 333.333 600
 
1
Taking X0 = 1 and using scaling we obtain

1
       
0.7113 0.6710 0.6645 0.6634
X1 =  1  , X2 =  1  , X3 =  1  , X4 =  1  ,
0.7958 0.7679 0.7635 0.7628
 
0.6632
X5 =  1  .
0.7627
This yields the eigenvalue λ = 1161.23. The smallest eigenvalue of A is then 1/λ =
0.0008612. The lowest critical load is
16E E
P = (0.0008612) − 0.01378 2 .
L2 L
 
67,745,349 −43,691,832 8,258,598
15. (a) A10 = −43,691,832 28,182,816 −5,328,720 
8,258,598 −5,328,720 1,008,180
     
1 67,745,349 1
(b) X10 = A10 0 = −43,691,832  ≈ 67,745,349 −0.644942
0 8,258,598 0.121906
     
1 2,680,201,629 1
X12 = A12 0 = −1,728,645,624  ≈ 2,680,201,629 −0.644968 .
0 326,775,222 0.121922
The vectors appear to be approaching scalar multiples of K = (1, −0.644968, 0.121922),
which approximates the dominant eigenvector.
(c) The dominant eigenvalue is λ1 = (AK · K)/(K · K) = 6.28995.

8.12 Diagonalization
1. Distinct eigenvalues λ1 = 1, λ2 = 5 imply A is diagonalizable.

−3 1 1 0
P= , D=
1 1 0 5
2. Distinct eigenvalues λ1 = 0, λ2 = 6 imply A is diagonalizable.

−5 −1 0 0
P= , D=
4 2 0 6

1
3. For λ1 = λ2 = 1 we obtain the single eigenvector K1 = . Hence A is not diagonalizable.
1
√ √
4. Distinct eigenvalues λ1 = 5, λ2 = − 5 imply A is diagonalizable.
√ √ √
5 − 5 5 √ 0
P= , D=
1 1 0 − 5
5. Distinct eigenvalues λ1 = −7, λ2 = 4 imply A is diagonalizable.

13 1 −7 0
P= , D=
2 1 0 4
6. Distinct eigenvalues λ1 = −4, λ2 = 10 imply A is diagonalizable.

−3 1 −4 0
P= , D=
1 −5 0 10
1 2
7. Distinct eigenvalues λ1 = , λ2 = imply A is diagonalizable.
3 3
1
!
0

1 1 3
P= , D=
−1 1 0 2 3

1
8. For λ1 = λ2 = −3 we obtain the single eigenvector K1 = . Hence A is not diagonalizable.
1
9. Distinct eigenvalues λ1 = −i, λ2 = i imply A is diagonalizable.

1 1 −i 0
P= , D=
−i i 0 i
10. Distinct eigenvalues λ1 = 1 + i, λ2 = 1 − i imply A is diagonalizable.

2 2 1+i 0
P= , D=
i −i 0 1−i
8.12 Diagonalization 609
11. Distinct eigenvalues λ1 = 1, λ2 = −1, λ3 = 2 imply A is diagonalizable.

   
1 0 1 1 0 0
P = 0 1 1 , D = 0 −1 0
0 0 1 0 0 2
12. Distinct eigenvalues λ1 = 3, λ2 = 4, λ3 = 5 imply A is diagonalizable.

   
1 2 0 3 0 0
P= 0  2 1 , D = 0 4 0
1 1 −1 0 0 5

   
1 1 1 0 0 0
P =  0 1 0 , D = 0 1 0
−1 1 1 0 0 2
14. Distinct eigenvalues λ1 = 1, λ2 = −3i, λ3 = 3i imply A is diagonalizable.

   
0 −3i 3i 1 0 0
P = 0 1 1 , D = 0 −3i 0
1 0 0 0 0 3i
15. The eigenvalues are λ1 = λ2 = 1, λ3 = 2. For λ1 = λ2 = 1 we obtain the single eigenvector

 
1
K1 = 0. Hence A is not diagonalizable.

0
   
1 1 0 1 0 0
P = 0 1 0 , D = 0 2 0
0 0 1 0 0 3
√ √
17. Distinct eigenvalues λ1 = 1, λ2 = 5 , λ3 = − 5 imply A is diagonalizable.
 √ √   
0 1+ 5 1− 5 1 √0 0
P = 0 2 2  , D = 0 5 √0

1 0 0 0 0 − 5
 
1
18. For λ1 = λ2 = λ3 = 1 we obtain the single eigenvector K1 = −2. Hence A is not
1
diagonalizable.
19. For the eigenvalues λ1 = λ2 = 2, λ3 = 1, λ4 = −1 we obtain four linearly independent

eigenvectors. Hence A is diagonalizable and
   
−3 −1 −1 1 2 0 0 0
 0 1 0 0 , D = 0 2 0 0

P= −3
.
0 0 1   0 0 1 0
1 0 1 0 0 0 0 −1
20. The eigenvalues are λ1 = λ2 = 2, λ3 = λ4 = 3. For λ3 = λ4 = 3 we obtain the single

 
1
0
1. Hence A is not diagonalizable.
eigenvector K1 =  
 
√1 √1
1 1 2 2 0 0
21. λ1 = 0, λ2 = 2, K1 = , K2 = , P= , D =
−1 1 − √1 √1 0 2
2 2
 
√1 √2
1 2 5 5 −1 0
22. λ1 = −1, λ2 = 4, K1 = , K2 = , P=  , D =
−2 1 −√2 √1 0 4
5 5
 √ √ 
√ √ − √10 √10
− 10 10 14 35  3 0
23. λ1 = 3, λ2 = 10, K1 = , K2 = , P= , D=
2 5 √2 √15 0 10
14 35
 
√1 √1
1 1 2 2 −1 0
24. λ1 = −1, λ2 = 3, K1 = , K2 = , P=  , D =
1 −1 √1
− √12 0 3
2
 1
√1

     − √2 0
−1 1 0 
2

25. λ1 = −1, λ2 = λ3 = 1, K1 =  1, K2 = 1, K3 = 0, P =  1 √1 ,
 √2 2
0
0 0 1
0 0 1
 
−1 0 0
D=  0 1 0
0 0 1
 1
√1 √1

      − √2 2 3
−1 1 1  
26. λ1 = λ2 = −1, λ3 = 5, K1 =  0, K2 = 1, K3 = −1, P = 
 0 √1 − √13 
,
2
1 0 1
 
√1 0 √1
  2 3
−1 0 0
D =  0 −1 0
0 0 5
2 2 1

     
2 2 1 2
3 3 3
− 31 − 32 

27. λ1 = 3, λ2 = 6, λ3 = 9, K1 = 2 , K2 = −1 , K3 = −2 , P =  3
      
,
1 −2 2 1
− 32 2
  3 3
3 0 0
D = 0 6 0
0 0 9
   √   √ 
√ √ 0 1− 2 1+ 2
28. λ1 = 1, λ2 = 2 − 2 , λ3 = 2 + 2 , K1 = 1, K2 =  0 , K3 =  0 ,
0 1 1
 √ √ √ √ 
0 − 2− 2 2+ 2  
2 2 1 0 0
  √
P = 1 0 0 , D = 0 2 − 2 2 0 
   
 √ √ √ √  √
2+ 2 2− 2 0 0 2+ 2
0 2 2
√1 √1
 
 
0
 
1
 
1 0 2 2
 
29. λ1 = 1, λ2 = −6, λ3 = 8, K1 = 1, K2 =  0, K3 = 0, P = 
 1 0 ,
0
0 −1 1 1
0 − √2 √1
  2
1 0 0
D = 0 −6 0
0 0 8
       
−1 0 1 1
 0 −1 −1 1
30. λ1 = λ2 = 0, λ3 = −2, λ4 = 2, K1 = 
 1, K2 =
   , K3 =   , K4 =   ,
 0  1 1
0 1 −1 1
 1 1 1

− √2 0 2 2  
0 0 0 0
√1 1 1
 
 0 − 2
− 2 2 0 0 0 0
P= 1 , D=
 
1 1 0 0 −2 0
 √
2
0 2 2

0 0 0 2

1 1 1
0 √
2
−2 2
31. The given equation can be written as XT AX = 24:

y
5 −1 x 1 Y
x y = 24. Using λ1 = 6, λ2 = 4, K1 = ,
−1 5 y −1
 
√1 √1
1 2 2
 and X = PX′ we find x
K2 = , P=
1 −√ 1 √1
2 2
X
6 0 X 2 2
X Y = 24 or 6X + 4Y = 24.
0 4 Y
The conic section is an ellipse. Now from X′ = PT X we see that the XY -coordinates of
√ √
(1, −1) and (1, 1) are ( 2 , 0) and (0, 2 ), respectively. From this we conclude that the
X-axis and Y -axis are as shown in the accompanying figure.

y
13 −5 x 1 X
x y = 288. Using λ1 = 8, λ2 = 18, K1 = ,
−5 13 y 1 Y
 
√1 − √1 2 2
−1 2 2
 and X = PX′ we find
K2 = , P= x
1 √1 √1
2 2

8 0 X
X Y = 288 or 8X 2 + 18Y 2 = 288.
0 18 Y
The conic section is an ellipse. Now from X′ = PT X we see that the XY -coordinates of (1, 1)
√ √
and (1, −1) are ( 2 , 0) and (0, − 2 ), respectively. From this we conclude that the X-axis
and Y -axis are as shown in the accompanying figure.

y X
−3 4 x 1
x y = 20. Using λ1 = 5, λ2 = −5, K1 = ,
4 3 y 2 Y
  4
√1 − √2 x
−2 5 5
 and X = PX′ we find
K2 = , P= 2
1 √ √1
5 5

5 0 X
X Y = 20 or 5X 2 − 5Y 2 = 20.
0 −5 Y
The conic section is a hyperbola. Now from X′ = PT X we see that the XY -coordinates
√ √
of (1, 2) and (−2, 1) are ( 5 , 0) and (0, 5 ), respectively. From this we conclude that the
X-axis and Y -axis are as shown in the accompanying figure.

y X
16 12 x x
x y + −3 4 = 0. Using λ1 = 25, λ2 = 0,
12 9 y y Y
4
4 3
!
x

4 −3 5 5
K1 = , K2 = , P= 3 and X = PX′ we find
3 4 4
5 −5

25 0 X X
X Y + 0 5 =0 or 25X 2 + 5Y = 0.
0 0 Y Y
The conic section is a parabola. Now from X′ = PT X we see that the XY -coordinates of
(4, 3) and (3, −4) are (5, 0) and (0, −5), respectively. From this we conclude that the X-axis
and Y -axis are as shown in the accompanying figure.
35. Since D = P−1 AP we have A = PDP−1 . Hence

1 1 2 0 −1 1 4 −1
A= = .
2 1 0 3 2 −1 2 1
36. Since eigenvectors are mutually orthogonal we use an orthogonal matrix P and A = PDPT .
 1
√1 √1
 1
√1 √1
  
− 8 4 1
√ √

3 2 6
  3 3 3 −
  1 0 0    3 3 3
 1 0 √26   1 0 − √12  4 11 4

A = − √3  0 3 0  √2 = 3 3 3 
0 0 5
   
1 1 1 1 2 1 1 4 8
√ − √2 √6 √ √ √ −3 3 3
3 6 6 6
37. Since D = P−1 AP we have
A = PDP−1
A2 = PDP−1 PDP−1 = PDDP−1 = PD2 P−1
A3 = A2 A = PD2 P−1 PDP−1 = PD2 DP−1 = PD3 P−1
and so on.
 4   
2 0 0 0 16 0 0 0
 0 34 0 0   =  0 81 0 0 
 
38. 
 0 0 (−1)4 0   0 0 1 0 
0 0 0 (5)4 0 0 0 625
2 1
!
1 −1 1 −1 3 3
39. λ1 = 2, λ2 = −1, K1 = , K2 = , P= , P−1 =
1 2 1 2 −31 1
3
2 1
!
1 −1 32 0 3 3 21 11
A5 = =
1 2 0 −1 1
−3 3 1 22 10

5 2 5 2 −1 −1 2
40. λ1 = 0, λ2 = 1, K1 = , K2 = , P= , P =
3 1 3 1 3 −5

5 2 0 0 −1 2 6 −10
A10 = =
3 1 0 1 3 −5 3 −5
41. If A is a nonsingular diagonalizable matrix, then there exists a nonsingular matrix P such
that D = P−1 AP. Because A and P are nonsingular so is D. Hence
−1 −1
D−1 = P−1 AP = P−1 A−1 P−1 = P−1 A−1 P.
D−1 is a diagonal matrix, and so by definition A−1 is diagonalizable.
42. If A is a diagnalizable matrix, then D = P−1 AP. The matrix P is not unique because
interchanging columns of P or multiplying thecolumns
by a nonzero constant
still
diagonalizes
−3 1 1 −3
A. For example, in Problem 1 we used P = . But if we take ,
1 1 1 1

1 1 3
P−1
1 =
4−1 1

1 1 3 2 3 1 −3 5 0
P−1
1 AP = = .
4 −1 1 1 4 1 1 0 1
8.13 LU-Factorization
! ! ! !
2 −2 1 0 u11 u12 u11 u12
1. = = . Thus,
1 2 l21 1 0 u22 u11 l21 u12 l21 + u22
(
u11 = 2 u12 = −2
u11 l21 = 1 u12 l21 = 2
Now u11 l21 = (2)l21 = 2 and so l21 = 1/2. Also, u12 l21 +u22 = (−2)(1/2)+u22 = 2 so u22 = 3.
! ! !
1 0 2 −2 2 −2
LU = 1 = =A
2 1 0 3 1 2
! ! ! !
6 2 1 0 u11 u12 u11 u12
2. Proceeding as in Problem 1, = = .
4 1 l21 1 0 u22 u11 l21 u12 l21 + u22
Thus,
(
u11 = 6 u12 = 2
u11 l21 = 4 u12 l21 + u22 = 1
Now u11 l21 = (6)l21 = 4 and so l21 = 2/3. Also, u12 l21 + u22 = (2)(2/3) + u22 = 1 so
u22 = −1/3.
! !
1 0 6 2

6 2
LU = 2 = =A
3 1 0 − 13 4 1
! ! ! !
−1 4 1 0 u11 u12 u11 u12
3. Proceeding as in Example 2, = = .
2 2 l21 1 0 u22 u11 l21 u12 l21 + u22
Thus,
(
u11 = −1 u12 = 4
u11 l21 = 2 u12 l21 + u22 = 2
Now u11 l21 = (−1)l21 = 2 and so l21 = −2. Also, u12 l21 + u22 = (4)(−2) + u22 = 2 so u22 = 10

1 0 −1 4 −1 4
LU = = =A
−2 1 0 10 2 2
(
5 = u11 −4 = u12

1 0 5 −4
4. gives A = LU = .
15 = l12 u11 2 = l21 u12 + u22 3 1 0 14
8.13 LU-Factorization 615
5. Proceeding as in Example 2,
    
4 −2 1 1 0 0 u11 u12 u13
−5 1 2 = l21 1 0  0 u22 u23 
    
12 1 3 l31 l32 1 0 0 u33

 
u11 u12 u13
= l21 u11 l21 u12 + u22 l21 u13 + u23
 

l31 u11 l31 u12 + l32 u22 l31 u13 + l32 u23 + u33
Equating corresponding entries, we get
u11 = 4 u12 = −2 u13 = 1

l21 u11 = −4 l21 u12 + u22 = 1 l21 u13 + u23 = 2
l31 u11 = 12 l31 u12 + l32 u22 = 1 l31 u13 + l32 u23 + u33 = 3
Thus,
  
1 0 0 4 −2 1
LU = −1 1 0 0 −1 3 = A
3 −7 1 0 0 21
    
−3 2 1 1 0 0 u11 u12 u13
 9 3 2 = l21 1 0  0 u22 u23 
    
3 1 −1 l31 l32 1 0 0 u33

 
u11 u12 u13
= l21 u11 l21 u12 + u22 l21 u13 + u23
 

l31 u11 l31 u12 + l32 u22 l31 u13 + l32 u23 + u33
u11 = −3 u12 = 2 u13 = 1

l21 u11 = 9 l21 u12 + u22 = 3 l21 u13 + u23 = 2
l31 u11 = 3 l31 u12 + l32 u22 = 1 l31 u13 + l32 u23 + u33 = −1
Thus,
  
1 0 0 −3 2 1
LU = −3 1 0  0 9 5 = A
  
1
−1 3 1 0 0 − 53
    
1 2 7 1 0 0 u11 u12 u13
2 5 6 = l21 1 0  0 u22 u23 
    
7 6 4 l31 l32 1 0 0 u33

 
u11 u12 u13
= l21 u11 l21 u12 + u22 l21 u13 + u23 
l31 u11 l31 u12 + l32 u22 l31 u13 + l32 u23 + u33
u11 = 1 u12 = 2 u13 = 7
l21 u11 = 2 l21 u12 + u22 = 5 l21 u13 + u23 = 6
l31 u11 = 7 l31 u12 + l32 u22 = 6 l31 u13 + l32 u23 + u33 = 4
Thus,
  
1 0 0 1 2 7
LU = 2 1 0 0 1 −8 = A
7 −8 1 0 0 −109
    
−4 2 −10 1 0 0 u11 u12 u13
 4 1 4 = l21 1 0  0 u22 u23 
    
6 6 −8 l31 l32 1 0 0 u33

 
u11 u12 u13
= l21 u11 l21 u12 + u22 l21 u13 + u23
 

l31 u11 l31 u12 + l32 u22 l31 u13 + l32 u23 + u33
u11 = −4 u12 = 2 u13 = −10
l21 u11 = 4 l21 u12 + u22 = 1 l21 u13 + u23 = 4
l31 u11 = 6 l31 u12 + l32 u22 = 6 l31 u13 + l32 u23 + u33 = −8
Thus,
  
1 0 0 −4 2 −10
LU =  −1 1 0  0 3 −6 = A
  
− 32 3 1 0 0 −5
    
1 −2 1 1 0 0 u11 u12 u13
0 1 2 = l21 1 0  0 u22 u23 
    
2 6 1 l31 l32 1 0 0 u33

 
u11 u12 u13
= l21 u11 l21 u12 + u22 l21 u13 + u23
 

l31 u11 l31 u12 + l32 u22 l31 u13 + l32 u23 + u33
u11 = 1 u12 = −2 u13 = 1

l21 u11 = 0 l21 u12 + u22 = 1 l21 u13 + u23 = 2
l31 u11 = 2 l31 u12 + l32 u22 = 6 l31 u13 + l32 u23 + u33 = 1
Thus,
  
1 0 0 1 −2 1
LU = 0 1 0 0 1 2 = A
2 10 1 0 0 −21

    
1 0 1 1 0 0 u11 u12 u13
1 9 1 = l21 1 0  0 u22 u23 
    
1 0 −1 l31 l32 1 0 0 u33

 
u11 u12 u13
= l21 u11 l21 u12 + u22 l21 u13 + u23 
 
l31 u11 l31 u12 + l32 u22 l31 u13 + l32 u23 + u33
u11 = 1 u12 = 0 u13 = 1

l21 u11 = 1 l21 u12 + u22 = 9 l21 u13 + u23 = 1
l31 u11 = 1 l31 u12 + l32 u22 = 0 l31 u13 + l32 u23 + u33 = −1
Thus,
  
1 0 0 1 0 1
LU = 1 1 0 0 9 0 = A
1 0 1 0 0 −2
11. Proceeding as in Examples 3 and 4,

3 9 1 11 −3R1 +R2 1 11
A= = −−−−−−→
1 11 3 9 0 −24

1 0 1 0 Record 3 1 0
I= = −−−−−−→
0 1 0 1 3 1
Thus,

1 0 1 11
LU = =A
3 1 0 −24

−2 10 1 −4 2R1 +R2 1 −4
A= = −−−−−−→
1 −4 −2 10 0 2

1 0 1 0 Record −2 1 0
I= = −−−−−−→
0 1 0 1 −2 1
Thus,

1 0 1 −4
LU = =A
−2 1 0 2

−4 −2 −1 −3 −4R1 +R2 −1 −3
A= = −−−−−−→
−1 −3 −4 −2 0 10

1 0 1 0 Record 4 1 0
I= = −−−−−−→
0 1 0 1 4 1
Thus,

1 0 −1 −3
LU = =A
4 1 0 10

0.2 0.3 −4R1 +R2 0.2 0.3
A= −−−−−−→
0.8 3.5 0 2.3

1 0 1 0 Record 4 1 0
I= = −−−−−−→
0 1 0 1 4 1
Thus,

1 0 0.2 0.3
LU = =A
4 1 0 2.3

  −3R1 +R2    
1 1 1 −1R1 +R3 1 1 1 1 1 1
−1R +R3 
A = 3 1 2 −−−−−−→ 0 −2 −1 −−−−2−−→ 0 −2 −1
1 −1 1 0 −2 0 0 0 1
  Record 3    
1 0 0 Record 1
1 0 0 1 0 0
Record 1
I = 0 1 0 −−−−−−→ 3 1 0 −−−−−−→ 3 1 0
0 0 1 1 0 1 1 1 1
Thus,
  
1 0 0 1 1 1
LU = 3 1 0 0 −2 −1 = A
1 1 1 0 0 1
  2R1 +R2    
−1 2 −4 3R1 +R3
−1 2 −4 −1 2 −4
7R2 +R3
A =  2 −5 10 = −−−−−−→  0 −1 2 −−−− −−→  0 −1 2
3 1 6 0 7 −6 0 0 8
  Record −2    
1 0 0 Record −3 1 0 0 1 0 0
Record −7
I = 0 1 0 −−−−−−→ −2 1 0 −−−−−−→ −2 1 0
0 0 1 −3 0 1 −3 −7 1
Thus,
  
1 0 0 −1 2 −4
LU = −2 1 0  0 −1 2 = A
−3 −7 1 0 0 8
  − 21 R1 +R2    
4 2 12 −3R1 +R3 4 2 12 2
R +R
4 2 12
5 2 3
A= 2 26 −4 −−−−−−→ 0 25 −10 −− −−−−→ 0 25 −10
12 −14 56 0 −10 20 0 0 16
  Record 12    
1 0 0 Record 3
1 0 0 Record − 2 1 0 0
I = 0 1 0 −−−−−−→  21 1 0 −−−−−−→ 5 1
2 1 0
0 0 1 3 0 1 3 − 52 1
Thus,
  
10 0 4 2 12
LU =  21 1 0 0 25 −10 = A
3 − 52 1 0 0 16
  − 14 R1 +R2    
16 4 20 − 5 R1 +R3 16 4 20 1
R +R
16 4 20
4 2 2 3
A =  4 5 3 −−−−−−→  0 4 −2 −− −−−−→  0 4 −2
20 3 29 0 −2 4 0 0 3
  Record 14    
1 0 0 Record 54
1 0 0 Record − 1 1 0 0
I = 0 1 0 −−−−−−→  14 1 0 −−−−−−→ 2 1
4 1 0
5 5 1
0 0 1 4 0 1 4 −2 1
Thus,
  
1 0 0 16 4 20
LU =  41 1 0  0 4 −2 = A
5 1
4 2 1 0 0 3
2R1 +R2
     
1 −2 1 0 −1R1 +R3 1 −2 1 0 4R2 +R3 1 −2 1 0
−2 3 −2 1 −5R1 +R4 0 −1 0 1 11R2 +R4 0 −1 0 1
A=
 1
 −−−−−−→   −−−− −−→  
2 4 3 0 4 3 3 0 0 3 7
5 1 −1 1 0 11 −6 1 0 0 −6 12
     
1 −2 1 0 1 0 0 0 1 0 0 0
2R3 +R4  0 −1 0 1 0 1 0 0 −2 1 0 0
−−−− −−→ 
0
, I=  −→  
0 3 7 0 0 1 0  1 −4 1 0
0 0 0 26 0 0 0 1 5 −11 −2 1
Thus,
  
1 0 0 0 1 −2 1 0
−2 1 0 0
 0 −1 0 1 = A
 
LU = 
 1 −4 1 0 0 0 3 7
5 −11 −2 1 0 0 0 26
−3R1 +R2
     
1 −1 3 2 −2R1 +R3 1 −1 3 2 1 −1 3 2
 3 −5 19 3 +1R +R4 0 −2 10 −3 2R2 +R4 0 −2 10 −3
A=  −−−−1−−→   −−−−−−→  
 2 −2 3 10   0 0 −3 6  0 0 −3 6
−1 5 −5 −2 0 4 −2 0 0 0 18 −6
     
1 −1 3 2 1 0 0 0 1 0 0 0
6R3 +R4  0 −2 10 −3 0 1 0 0  3 1 0 0
−−−− −−→ 
0
, I=  −→  
0 −3 6 0 0 1 0   2 0 1 0
0 0 0 30 0 0 0 1 −1 −2 −6 1
Thus,
  
1 0 0 0 1 −1 3 2
 3 1 0 0 0 −2 10 −3 = A
 
LU = 
 2 0 1 0   0 0 −3 6
−1 −2 −6 1 0 0 0 30

2 −2 x1 1
21. Using the results from Problem 1, =
1 2 x2 −2

1 0 2 −2 x1 1
Thus, 1 =
2 1 0 3 x2 −2

y1
Now let UX = Y where Y = so we get
y2
! ! ! !
1 0 y1 1 1
1 = or Y =
2 1 y2 −2 − 25
Now solve UX = Y to get
− 13
! ! ! ! !
2 −2 x1 1 x1
= or X= =
0 3 x2 − 25 x2 − 56

6 2 x1 12
4 1 x2 2
Thus,
! ! ! !
1 0 6 2 x1 12
2 =
3 1 0 − 31 x2 2

y1
y2

1 0 y1 12 12
2 = or Y=
3 1 y2 2 −6

! ! !
6 2 x1 12

x1 −4
= or X= =
0 − 31 x2 −6 x2 18

−1 4 x1 15
2 2 x2 5
Thus,

1 0 −1 4 x1 15
=
−2 1 0 10 x2 5

y1
y2

1 0 y1 15 15
= or Y=
−2 1 y2 5 35

! !
x1 −1

−1 4 x1 15
= or X= = 7
0 10 x2 35 x2 2

5 −4 x1 1
15 2 x2 7
Thus,

1 0 5 −4 x1 1
=
3 1 0 14 x2 7

y1
y2

1 0 y1 1 1
= or Y=
3 1 y2 7 4
3
! !
x1

5 −4 x1 1 7
= or X= =
0 14 x2 4 x2 2
7
    
4 −2 1 x1 7
25. Using the results from Problem 5, −4
 1 2 x2 =
  7
12 1 3 x3 28
Thus,
     
1 0 0 4 −2 1 x1 7
−1 1 0 0 −1 3 x2  =  7
3 −7 1 0 0 21 x3 28
 
y1
Now let UX = Y where Y = y2  so we get
y3
      
1 0 0 y1 7 7
−1 1 0   y2 =
  7 or Y =  14
3 −7 1 y3 28 105

        
4 −2 1 x1 7 x1 1
0 −1 3 x2  =  14 or X = x2 = 1
  
0 0 21 x3 105 x3 5
    
−3 2 1 x1 15
26. Using the results of Problem 6,  9 3 2 x2  =  3
3 1 −1 x3 −24
Thus,
    
1 0 0 −3 2 1

x1 15
−3 1 0  0 9 5 x2  =  3
  
−1 1
0 0 − 53 x3 −24
3 1
 
y1
Now let UX = Y where Y = y2  so we get

y3
    
1 0 0 y1 15
 
15
−3 1 0 y2  =  3 or Y =  48
    
−1 1 −25
3 1 y3 −24

    
−3 2 1 x1 15
   
x1 −2
 0 9 5 x2  =  48 or X = x2  = −3
    
0 0 − 53 x3 −25 x3 15
    
1 2 7 x1 109
27. Using the results from Problem 7, 2 5 6 x2  = 109
7 6 4 x3 218
Thus,
     
1 0 0 1 2 7 x1 109
2 1 0 0 1 −8 x2  = 109
7 −8 1 0 0 −109 x3 218
 
y1
Now let UX = Y where Y = y2  so we get

y3
      
1 0 0 y1 109 109
2 1 0 y2  = 109 or Y =  −109
7 −8 1 y3 218 −1417

        
1 2 7 x1 109 x1 28
0 1 −8 x2  =  −109 or X = x2  = −5
0 0 −109 x3 −1417 x3 13
    
−4 2 −10 x1 −20
28. Using the results from Problem 8,  4 1 4 x2  =  14
6 6 −8 x3 −18
Thus,
     
1 0 0 −4 2 −10 x1 20
 −1 1 0  0 3 −6 x2  =  14
     
− 23 3 1 0 0 −5 x3 −18
 
y1
y3
      
1 0 0 y1 −20 −20
 −1 1 0 y2  =  14 or Y =  −6
− 32 3 1 y3 −8 −30

       
−4 2 −10 x1 −20 x1 −5
 0 3 −6 x2  =  −6 or X = x2  =  10
0 0 −5 x3 −30 x3 6
    
1 −2 1 x1 14
29. Using the results from Problem 9, 0 1 2 x2  = −42
2 6 1 x3 7
Thus,
     
1 0 0 1 −2 1 x1 14
0 1 0 0 1 2 x2  = −42
2 10 1 0 0 −21 x3 7
 
y1
y3
      
1 0 0 y1 14 14
0 1 0 y2  = −42 or Y = −42
2 10 1 y3 7 399

        
1 2 1 x1 14 x1 25
0 1 2 x2  = −42 or X = x2  =  −4
0 0 −21 x3 399 x3 −19
    
1 0 1 x1 18
30. Using the results from Problem 10, 1 9 1 x2  =  27
1 0 −1 x3 −12
Thus,
     
1 0 0 1 0 1 x1 18
1 1 0 0 9 0 x2  =  27
1 0 1 0 0 −2 x3 −12
 
y1
y3
      
1 0 0 y1 18 18
1 1 0 y2  =  27 or Y =  9
1 0 1 y3 −12 −30

        
1 0 1 x1 18 x1 3
0 9 0 x2  =  9 or X = x2  =  1
0 0 −2 x3 −30 x3 15
    
1 1 1 x1 2
31. Proceeding as in Problems 20-30, 1 2 2
   x2 = 4 
 
1 2 3 x3 1
Thus,
     
1 0 0 1 1 1 x1 2
1 1 0 0 1 1 x2  = 4
1 1 1 0 0 1 x3 1
 
y1
y3
      
1 0 0 y1 2 2
1 1 0 y2  = 4 or Y =  2
1 1 1 y3 1 −3

        
1 1 1 x1 2 x1 0
0 1 1 x2  =  2 or X =  x2  =  5 
0 0 1 x3 −3 x3 −3
    
1 1 1 x1 −4
32. Proceeding as in Problem 31, 1 2 2 x2  =  7
1 2 3 x3 10
Thus,
     
1 0 0 1 1 1 x1 −4
1 1 0 0 1 1 x2  =  7
1 1 1 0 0 1 x3 10
 
y1
y3
      
1 0 0 y1 −4 −4
1 1 0 y2  =  7 or Y = 11

1 1 1 y3 10 3

        
1 1 1 x1 −4 x1 −15
0 1 1 x2  =  11 or X = x2  =  8
0 0 1 x3 3 x3 3
     1
1 1 1 x1 2
     3
33. Proceeding as in Problems 30 and 31, 1 2 2 x2  =  4 
    

1
1 2 3 x3 −2
Thus,
      1
1 0 0 1 1 1 x1 2
      3
1 1 0 0 1 1 x2  =  
      4
1 1 1 0 0 1 x3 − 12
 
y1
y3
     1  1

1 0 0 y1 2 2
     3  1
1 1 0 y2  =   or Y=  
     4  4
1 1 1 y3 − 12 − 54

     1    1

1 1 1 x1 2 x1 4
     1    3
0 1 1 x2  =   or X=
 x2  =
  
     4  2
0 0 1 x3 − 54 x3 − 54
    
1 1 1 x1 30
34. Proceeding as in Problems 30-33, 1 2 2 x2  = −42
1 2 3 x3 −18
Thus,
     
1 0 0 1 1 1 x1 30
1 1 0 0 1 1 x2  = −42
1 1 1 0 0 1 x3 −18
 
y1
y3
     

1 0 0 y1 30 30
1 1 0 y2  = −42 or Y = −72
1 1 1 y3 −18 24

        
1 1 1 x1 30 x1 102
0 1 1 x2  = −72 or X = x2  = −96
0 0 1 x3 24 x3 24
35. Using the results of Problem 15, A = LU

    
1 1 1 1 0 0 1 1 1
3 1 2 = 3 1 0 0 −2 −1
1 −1 1 1 1 1 0 0 1
Therefore det A = det L · det U = (1 · 1 · 1) · (1 · (−2) · 1) = −2
36. Using the results of Problem 16 and Theorem 8.5.8,

   
1 0 0 −1 2 −4
det A = det −2 1 0 · det  0 −1 2 = (1)(8) = 8
−3 −7 1 0 0 8

   
1 0 0 4 2 12
det A = det  21 1 0 · det 0 25 −10 = (1)(1600) = 1600
   
3 − 52 1 0 0 16

   
1 0 0 16 4 20
1   
det A = det 
4 1 0  · det  0 4 −2 = (1)(192) = 192
 
5
4 − 12 1 0 0 3
   
1 0 0 0 1 −2 1 0
−2 1 0 0
 · det 0 −1 0 1 = (1)(−78) = −78
 
det A = det 
 1 −4 1 0 0 0 3 7
5 −11 −2 1 0 0 0 26
   
1 0 0 0 1 −1 3 2
 3 1 0 0 · det 0 −2 10 −3 = (1)(180) = 180
 
det A = det 
 2 0 1 0   0 0 −3 6
−1 −2 −6 1 0 0 0 30
    
1 1 1 l11 0 0 1 u12 u13
41. 3 1 2 = l21 l22 0  0 1 u23 
    
1 −1 1 l31 l32 l33 0 0 1

 
l11 l11 u12 l11 u13
= l21 l21 u12 + l22 l21 u13 + l22 u23 
 
l31 l31 u12 + l32 l31 u13 + l32 u23 + l33
 
l11 = 1 l11 u12 = 1 l11 u13 = 1
l21 = 3 l21 u12 + l22 = 1 l21 u13 + l22 u23 = 2 
 
l31 = 1 l31 u12 + l32 = −1 l31 u13 + l32 u23 + l33 = 1
Thus,
  
1 0 0 1 1 1
1
LU = 3 −2 0 0 1 =A
 
2
1 −2 1 0 0 1
    
−1 2 −4 l11 0 0 1 u12 u13
42.  2 −5 10 = l21 l22 0  0 1 u23 
    
3 1 6 l31 l32 l33 0 0 1

 
l11 l11 u12 l11 u13
= l21 l21 u12 + l22 l21 u13 + l22 u23 
 
l31 l31 u12 + l32 l31 u13 + l32 u23 + l33
 
l11 = −1 l11 u12 = 2 l11 u13 = −4
 l21 = 2 l21 u12 + l22 = −5 l21 u13 + l22 u23 = 10 
 
l31 = 3 l31 u12 + l32 = 1 l31 u13 + l32 u23 + l33 = 6
Thus,
  
−1 0 0 1 −2 4
LU =  2 −1 0 0 1 −2 = A
3 7 8 0 0 1
    
4 2 12 l11 0 0 l11 l21 l31
43.  2 26 −4 = l21 l22 0   0 l22 l32 
    
12 −4 56 l31 l32 l33 0 0 l33

 2 
l11 l11 l21 l11 l31
= l21 l11
 2
l21 + l222 l21 l31 + l22 l32 

l31 l11 l31 l21 + l32 l22 2 + l2 + l2

l31 32 33
 2 
l11 = 4 l11 l21 = 2 l11 l31 = 12
 l21 l11 = 2 2 + l2 = 26
l21 l21 l31 + l22 l32 = −4
22
2 + l2 + l2 = 56
l31 l11 = 12 l31 l21 + l32 l22 = −4 l31 32 33
Thus,
  
2 0 0 2 1 6
LLT = 1 5 0 0 5 −2
6 −2 4 0 0 4
    
16 4 20 l11 0 0 l11 l21 l31
44.  4 5 3 = l21 l22 0   0 l22 l32 
    
20 3 29 l31 l32 l33 0 0 l33

 2 
l11 l11 l21 l11 l31
= l21 l11
 2
l21 + l222 l21 l31 + l22 l32 

l31 l11 l31 l21 + l32 l22 2 + l2 + l2

l31 32 33
 2 
l11 = 16 l11 l21 = 4 l11 l31 = 20
2 + l2 = 5
 l21 l11 = 4 l21 l21 l31 + l22 l32 = 3 
 
22
2 + l2 + l2 = 29
l31 l11 = 20 l31 l21 + l32 l22 = 3 l31 32 33
Thus,
  
4 0 0 4 1 5
T
LL = 1 2 0 0 2 −1
  
√ √
5 −1 3 0 0 3
8.14 Cryptography

19 5 14 4 0
1. (a) The message is M = . The encoded message is
8 5 12 16 0

1 2 19 5 14 4 0 35 15 38 36 0
B = AM = = .
1 1 8 5 12 16 0 27 10 26 20 0
8.14 Cryptography 631
(b) The decoded message is

−1 2 35 15 38 36 0 19 5 14 4 0
M = A−1 B = = .
1 −1 27 10 26 20 0 8 5 12 16 0

20 8 5 0 13 15 14 5 25
0 9 19 0 8 5 18 5 0

3 5 20 8 5 0 13 15 14 5 25
B = AM =
1 2 0 9 19 0 8 5 18 5 0

60 69 110 0 79 70 132 40 75
= .
20 26 43 0 29 25 50 15 25

−1 2 −5 60 69 110 0 79 70 132 40 75
M=A B=
−1 3 20 26 43 0 29 25 50 15 25

20 8 5 0 13 15 14 5 25
= .
0 9 19 0 8 5 18 5 0

16 8 15 14 5
0 8 15 13 5

3 5 16 8 15 14 5 48 64 120 107 40
B = AM = = .
2 3 0 8 15 13 5 32 40 75 67 25

−1 −3 5 48 64 120 107 40 16 8 15 14 5
M=A B= = = .
2 −3 32 40 75 67 25 0 8 15 13 5
 
7 15 0 14 15 18 20
4. (a) The message is M = 8 0 15 14 0 13 1. The encoded message is
9 14 0 19 20 0 0
  
1 2 3 7 15 0 14 15 18 20
B = AM = 1 1 2
   8 0 15 14 0 13 1
0 1 2 9 14 0 19 20 0 0
 
50 57 30 99 75 44 22
= 33 43 15 66 55 31 21 .
26 28 15 52 40 13 1

  
0 1 −1 50 57 30 99 75 44 22
M = A−1 B =  2 −2 −1 33 43 15 66 55 31 21
−1 1 1 26 28 15 52 40 13 1
 
7 15 0 14 15 18 20
= 8 0 15 14 0
 13 1 .
9 14 0 19 20 0 0
 
7 15 0 14 15 18 20
5. (a) The message is M = 8 0 15 14 0 13 1. The encoded message is
9 14 0 19 20 0 0
  
2 1 1 7 15 0 14 15 18 20
B = AM =  1 1 1 8 0 15 14 0 13 1
−1 1 0 9 14 0 19 20 0 0
 
31 44 15 61 50 49 41
= 24 29 15 47 35 31 21 .
1 −15 15 0 −15 −5 −19

  
1 −1 0 31 44 15 61 50 49 41
−1
M=A B=  1 −1 1   24 29 15 47 35 31 21
−2 3 −1 1 −15 15 0 −15 −5 −19
 
7 15 0 14 15 18 20
= 8 0 15 14 0
 13 1 .
9 14 0 19 20 0 0
 
4 18 0 10 15 8
6. (a) The message is M = 14 0 9 19 0 20. The encoded message is
8 5 0 19 16 25
  
5 3 0 4 18 0 10 15 8
B = AM = 4
 3 −1 14 0 9 19 0 20
5 2 2 8 5 0 19 16 25
 
62 90 27 107 75 100
= 50 67 27 78 44 67 .
64 100 18 126 107 130

  
8 −6 −3 62 90 27 107 75 100
−1
M = A B = −13 10
 5 50 67 27 78 44 67
−7 5 3 64 100 18 126 107 130
 
4 18 0 10 15 8
= 14 0 9 19 0 20 .
8 5 0 19 16 25
7. The decoded message is

−1 2 −3 152 184 171 86 212 19 20 21 4 25
M=A B= = .
−5 8 95 116 107 56 133 0 8 1 18 4
From correspondence (1) we obtain: STUDY HARD.

8.14 Cryptography 633

−1 1 −1 46 −7 −13 22 −18 1 10
M=A B=
1 −2 23 −15 −14 2 −18 −12 5

23 8 1 20 0 13 5
= .
0 23 15 18 18 25 0
From correspondence (1) we obtain: WHAT ME WORRY .

    
0 0 1 31 21 21 22 20 9 13 1 20 8 0 9
M = A−1 B = 0 1 0 19 0 9 13 16 15 = 19 0 9 13 16 15 .
1 0 −1 13 1 20 8 0 9 18 20 1 14 20 0
From correspondence (1) we obtain: MATH IS IMPORTANT.

  
1 0 −1 36 32 28 61 26 56 10 12
M = A−1 B = −1 1 2 −9 −2 −18 −1 −18 −25 0 0
0 −1 0 23 27 23 41 26 43 5 12
 
13 5 5 20 0 13 5 0
=  1 20 0 20 8 5 0 12 .
9 2 18 1 18 25 0 0
From correspondence (1) we obtain: MEET ME AT THE LIBRARY .

−1 u v
11. Let A = . Then
x y

−1 u v 17 16 18 5 34 0 34 20 9 5 25
A B= ,
x y −30 −31 −32 −10 −59 0 −54 −35 −13 −6 −50

2 1
so 17u − 30v = 4, 16u − 31v = 1 and 5x − 6y = 1, 25x − 50y = 25. Then A−1 =
−1 −1
and
−1 4 1 4 0 9 0 14 5 5 4 0
A B .
13 15 14 5 25 0 20 15 4 1 25
From correspondence (1) we obtain: DAD I NEED MONEY TODAY.

 
22 8 19 27 21 3 3 27 21 18 21
12. (a) MT = 13 3 21 22 3 25 27 6 7 14 23
2 27 21 7 27 5 21 17 2 25 7
  
1 1 0 37 38 61 56 51 33 51 50 30 57 51
(b) BT = M = 1 0 1 = 24 35 40 34 48 8 24 44 23 43 28
1 1 −1 11 −24 0 15 −24 20 6 −11 5 −11 16
 
−1 1 1
(c) BA−1 = B  2 −1 −1 = M
1 0 −1
 
15 22 20 8 23 6 21 22
13. (a) B′ = 10 22 18 23 25 2 23 25
3 26 26 14 23 16 26 12
(b) Using correspondence (1) the encoded message is:

OVTHWFUVJVRWYBWYCZZNWPZL.
 
1 4 −3
(c) M′ A−1 B′ =  2 3 −2 B′
−2 −4 3
 
46 32 14 58 54 −34 35 86
=  54 58 42 57 75 −14 59 95
−61 −54 −34 −66 −77 28 −56 −108
 
19 5 14 4 0 20 8 5
M = M′ mod 27 =  0 4 15 3 21 13 5 14.
20 0 20 15 4 1 25 0
Using correspondence (1) the encoded message is: SEND THE DOCUMENT TO-
DAY.
8.15 An Error-Correcting Code

1. 0 1 1 0 2. 1 1 1 1

3. 0 0 0 1 1 4. 1 0 1 0 0

5. 1 0 1 0 1 0 0 1 6. 0 1 1 0 1 0 1 0

7. 1 0 0 8. 0 0 1

9. Parity error 10. 1 0 1 0

11. 1 0 0 1 1 12. Parity error
 
1 1 0 1
In Problems 13–18, D = c1 c2 c3 and P = 1 0 1 1.
0 1 1 1
T T
13. DT = P 1 1 1 0

= 0 0 0
; C= 0 0 1 0 1 1 0
T T
14. DT = P 0 0 1 1 = 1 0 0 ; C = 1 0 0 0 0 1 1

T T
15. DT = P 0 1 0 1 = 0 1 0 ; C = 0 1 0 0 1 0 1

8.15 An Error-Correcting Code 635
T T
16. DT = P 0 0 0 1

= 1 1 1
; C= 1 1 0 1 0 0 1
T T
17. DT = P 0 1 1 0 = 1 1 0 ; C = 1 1 0 0 1 1 0

T T
18. DT = P 1 1 0 0 = 0 1 1 ; C = 0 1 1 1 1 0 0

In Problems 19–28, W represents the correctly decoded message.

T
19. S = HRT = H 0 0 0 0 0 0 0 =

0 0 0 ; a code word. W = 0 0 0 0
T
20. S = HRT = H 1 1 0 0 0 0 0 =

0 1 1 ; not a code word. The error is in the

third bit. W = 1 0 0 0
T
21. S = HRT = H 1 1 0 1 1 0 1 =


fifth bit. W = 0 0 0 1
T
22. S = HRT = H 0 1 0 1 0 1 0 =

0 0 0 ; a code word. W = 0 0 1 0
T
23. S = HRT = H 1 1 1 1 1 1 1 =

0 0 0 ; a code word. W = 1 1 1 1
T
24. S = HRT = H 1 1 0 0 1 1 0 =

0 0 0 ; a code word. W = 0 1 1 0
T
25. S = HRT = H 0 1 1 1 0 0 1 =


second bit. W = 1 0 0 1
T
26. S = HRT = H 1 0 0 1 0 0 1 =


T
27. S = HRT = H 1 0 1 1 0 1 1 =


seventh bit. W = 1 0 1 0
T
28. S = HRT = H 0 0 1 0 0 1 1 =


29. (a) 27 = 128 (b) 24 = 16

(c) 0 0 0 0 0 0 0 , 1 1 0 1 0 0 1 , 0 1 0 1 0 1 0 , 1 0 0 0 0 1 1 ,

1 0 0 1 1 0 0 , 0 1 0 0 1 0 1 , 1 1 0 0 1 1 0 , 0 0 0 1 1 1 1 ,

1 1 1 0 0 0 0 , 0 0 1 1 0 0 1 , 1 0 1 1 0 1 0 , 0 1 1 0 0 1 1 ,

0 1 1 1 1 0 0 , 1 0 1 0 1 0 1 , 0 0 1 0 1 1 0 , 1 1 1 1 1 1 1

30. (a) c4 = 0, c3 = 1, c2 = 1, c1 = 0; 0 1 1 0 0 1 1 0
 
0 0 0 0 1 1 1 1
0 0 1 1 0 0 1 1
(b) H = 
0

1 0 1 0 1 0 1
1 1 1 1 1 1 1 1
T T
(c) S = HRT = H 0 0 1 1 1 1 0 0 = 0 0 0 0
8.16 Method of Least Squares

2 3 4 5
YT AT

1. We have = 1 2 3 2 and = .
1 1 1 1

T 54 14 T −1 1 4 −14
Now A A = and (A A) =
14 4 20 −14 54
2
!
5
so X = (AT A)−1 AT Y = 3
and the least squares line is y = 0.4x + 0.6.
5

0 1 2 3
YT AT

2. We have = −1 3 5 7 and = .
1 1 1 1

14 6 1 4 −6
Now AT A = and (AT A)−1 =
6 4 20 −6 14
13
!
5
so X = (AT A)−1 AT Y = and the least squares line is y = 2.6x − 0.4.
− 25

1 2 3 4 5
YT AT

3. We have = 1 1.5 3 4.5 5 and = .
1 1 1 1 1

T 55 15 T −1 1 5 −15
15 5 50 −15 55

1.1
−0.3

T
T 0 2 3 4 5
4. We have Y = 0 1.5 3 4.5 5 and A = .
1 1 1 1 1

T 54 14 T −1 1 5 −14
14 5 74 −14 54

1.06757
−0.189189

T
T 0 1 2 3 4 5 6
5. We have Y = 2 3 5 5 9 8 10 and A = .
1 1 1 1 1 1 1

T 91 21 T −1 1 7 −21
21 7 196 −21 91
19
!
14
so X = (AT A)−1 AT Y = 27
and the least squares line is y = 1.35714x + 1.92857.
14

1 2 3 4 5 6 7
Y T = 2 2.5 1 1.5 2 3.2 5 and AT =

6. We have .
1 1 1 1 1 1 1

140 28 1 7 −28
Now AT A = and (AT A)−1 =
28 7 196 −28 140

0.407143
so X = (AT A)−1 AT Y = and the least squares line is y = 0.407143x + 0.828571.
0.828571
8.16 Method of Least Squares 637

20 40 60 80 100 120
YT AT

7. We have = 220 200 180 170 150 135 and = .
1 1 1 1 1 1

T 36400 420 T −1 1 6 −420
420 6 42000 −420 36400
117
!
− 140
so X = (AT A)−1 AT Y = 703
and the least squares line is v = −0.835714T + 234.333.
3
At T = 140, v ≈ 117.333 and at T = 160, v ≈ 100.619.

400 450 500 550 600 650
YT AT

8. We have = 0.47 0.90 2.0 3.7 7.5 15 and = .
1 1 1 1 1 1

T 1697500 3150 T −1 1 6 −3150
3150 6 262500 −3150 1697500

0.0538
so X = (AT A)−1 AT Y = and the least squares line is R = 0.0538T − 23.3167.
−23.3167
At T = 700, R ≈ 14.3433.
9. The data yields

   
1 1 1 1  
1  4 1 4 9 16
2 1
Y= 2 ,
 A=
 9
, AT = 1 2 3 4
3 1
1 1 1 1
5 16 4 1
and
 
  1 1 1  
1 4 9 16  354 100 30
4 2 1
AT A = 1 2 3 4   9 3 1 = 100 30 10 ,
  
1 1 1 1 30 10 4
16 4 1
 
5 −25 25
−1 1
AT A

= −25 129 −135
20
25 −135 155
Hence,
 
 1   
5 −25 25 1 4 9 16   0.75
−1 T 1  1 
X = AT A

A Y= −25 129 −135 1 2 3 4 
2 = −2.45

20
25 −135 155 1 1 1 1 2.75
5
Therefore, f (x) = 0.75x2 − 2.45x + 2.75.
10. The data yields

   
1 4 −2 1  
0 1 −1 1 4 1 1 4
Y= 2 ,
 A=
1
, AT = −2 −1 1 2
1 1
1 1 1 1
3 4 2 1
and
 
  4 −2 1  
4 1 1 4  34 0 10
1 −1 1 
AT A = −2 −1 1

2 
1 = 0 10 0 ,
1 1
1 1 1 1 10 0 4
4 2 1
1 5
 
9 0 − 18
T
−1 1  1
A A = 0 0

20
 10
5 17
− 18 0 18
Hence,
 
 1 5
  1 1
9 0 − 18 4 1 1 4   3
T
−1 T 1  1
  0 3
X= A A A Y= 0 0 −2 −1 1 2  = 
    
20  10 2  5 
5 17 2
− 18 0 18 1 1 1 1   3
3
Therefore, f (x) = 31 x2 + 35 x + 32 .
8.17 Discrete Compartmental Models
In Problems 1-5 we use the fact that the element τij in the transfer matrix T is the rate of transfer
from compartment j to compartment i, and the fact that the sum of each column in T is 1.
1. (a) The initial state and the transfer matrix are

90 0.8 0.4
X0 = and T= .
60 0.2 0.6
(b) We have

0.8 0.4 90 96
X1 = TX0 = =
0.2 0.6 60 54
and
0.8 0.4 96 98.4
X2 = TX1 = = .
0.2 0.6 54 51.6
(c) From TX̂ − X̂ = (T − I)X̂ = 0 and the fact that the system is closed we obtain
−0.2x1 + 0.4x2 = 0
x1 + x2 = 150.

100
The solution is x1 = 100, x2 = 50, so the equilibrium state is X̂ = .
50
8.17 Discrete Compartmental Models 639

   
100 0.7 0 0.5
X0 = 200 and T = 0.3 0.8 0 .
150 0 0.2 0.5
(b) We have    
145 159
X1 = TX0 = 190 and X2 = TX1 = 195.5 .
115 95.5
−0.8x1 + 0.5x2 =0
0.3x1 − 0.9x2 =0
x1 + x2 + x3 = 450.
The solution is x1 = 145.161, x2 = 217.742, x3 = 87.0968, so the equilibrium state is

 
145.161
X̂ = 217.742.
87.097

   
100 0.2 0.5 0
X0 =  0 and T = 0.3 0.1 0 .
0 0.5 0.4 1
(b) We have    
20 19
X1 = TX0 = 30
 and X2 = TX1 =  9 .
50 72
−0.8x1 + 0.5x2 =0
0.3x1 − 0.9x2 =0
x1 + x2 + x3 = 100.
 
0
The solution is x1 = x2 = 0, x3 = 100, so the equilibrium state is X̂ =  0.
100
4. (a) The transfer matrix is  

0.7 0.05 0.15
T = 0.3 0.75 0 .
0 0.2 0.85
(b)
5. From TX̂ = 1X̂ we see that the equilibrium state vector X̂ is the eigenvector of the transfer
matrix T corresponding to the eigenvalue 1. It has the properties that its components add
up to the sum of the components of the initial state vector.

   
0 0.88 0.02 0
X0 = 100 and T = 0.06 0.97 0.05 .
0 0.06 0.01 0.95
(b)
Chapter 8 in Review
 
2 3 4 2. 4 × 3
3 4 5
1. 
4

5 6
5 6 7
! !
3 4 1 4 −2 −2 1
3. AB = ; BA = 11 4. A −1
=− =
6 8 2 −3 1 3
− 21
2
Chapter 8 in Review 641

1 0
5. False; consider A = and 6. True
0 1

0 1
B=
1 0
3
7. det 21 A = 21 (5) = 5

8 ; 8. det AB−1 = det A/ det B = 6/2 = 3
det(−AT ) = (−1)3 (5) = −5
9. 0 10. det C = (−1)3 / det B = −1/103 (2)

= −1/2000
11. False; an eigenvalue can be 0. 12. True
13. True 14. True, since complex roots of real polynomials

occur in conjugate pairs.
15. False; if the characteristic equation of an n × n matrix has repeated roots, there may not be
n linearly independent eigenvectors.
16. True 17. True
18. True 19. False; A is singular and thus not orthogonal.
20. True
21. A = 12 (A+AT )+ 21 (A−AT ) where 12 (A+AT ) is symmetric and 21 (A−AT ) is skew-symmetric.

2 2 0 1 2 0 1
22. Since det A = (det A) ≥ 0 and det = −1, there is no A such that A = .
1 0 1 0
2
1 1 1 1 0 0
23. is nilpotent since = .
−1 −1 −1 −1 0 0

i 0 0 −1 0 i
24. (a) σx σy = = −σy σx ; σx σz = = −σz σx ; σy σz = = −σz σy
0 −i 1 0 i 0
(b) We first note that for anticommuting matrices AB = −BA, so C = 2AB. Then

2i 0 0 2i 0 2
Cxy = , Cyz = , and Czx = .
0 −2i 2i 0 −2 0
     
5 −1 1 −9 1 1 5 9 1 1 5 9
R row
25. 2 4 0 27 −−−−13 −−→ 2 4 0 27 −−−−−−→ 0 1 −5 92 
1 1 5 9 5 −1 1 −9 operations 0 0 1 12
1 0 0 − 12
 
row
−−−−−−→ 0 1 0 7 .
operations 0 0 1 1
2
1 T
The solution is X = − 12

7 2 .
     
1 1 1 6 1 1 1 6 1 0 0 3
row row
26. 1 −2 3 2 −−−−−−→ 0 1 − 32 43  −−− −−−→ 0 1 0 2.
2 0 −3 3 operations 0 0 1 1 operations 0 0 1 1
The solution is x1 = 3, x2 = 2, x3 = 1.
27. Multiplying the second row by abc we obtain the third row. Thus the determinant is 0.
28. Expanding along the first row we see that the result is an expression of the form
ay + bx2 + cx + d = 0, which is a parabola since, in this case a 6= 0 and b 6= 0. Letting
x = 1 and y = 2 we note that the first and second rows are the same. Similarly, when x = 2
and y = 3, the first and third rows are the same; and when x = 3 and y = 5, the first and
fourth rows are the same. In each case the determinant is 0 and the points lie on the parabola.
29. 4(−2)(3)(−1)(2)(5) = 240 30. (−3)(6)(9)(1) = −162

1 −1 1

31. Since 5 1 −1 = 18 6= 0, the system has only the trivial solution.
1 2 1

1 −1 −1

32. Since 5 1 −1 = 0, the system has infinitely many solutions.
1 2 1
33. From x1 I2 + x2 HNO3 → x3 HIO3 + x4 NO2 + x5 H2 O we obtain the system

2x1 = x3 , x2 = x3 +2x5 , x2 = x4 , 3x2 = 3x3 +2x4 +x5 . Letting x4 = x2 in the fourth equation
we obtain x2 = 3x3 + x5 . Taking x1 = t we see that x3 = 2t, x2 = 2t + 2x5 , and x2 = 6t + x5 .
From the latter two equations we get x5 = 4t. Taking t = 1 we have x1 = 1, x2 = 10, x3 = 2,
x4 = 10, and x5 = 4. The balanced equation is I2 + 10HNO3 → 2HIO3 + 10NO2 + 4H2 O.
34. From x1 Ca + x2 H3 PO4 → x3 Ca3 P2 O8 + x4 H2 we obtain the system x1 = 3x3 , 3x2 = 2x4 ,
x2 = 2x3 , 4x2 = 8x3 . Letting x3 = t we see that x1 = 3t, x2 = 2t, and x4 = 3t. Taking t = 1
we obtain the balanced equation 3Ca + 2H3 PO4 → Ca3 P2 O8 + 3H2 .
42 1 −21 1
35. det A = −84, det A1 = 42, det A2 = −21, det A3 = −56; x1 = = − , x2 = = ,
−84 2 −84 4
−56 2
x3 = = .
−84 3
16 −4 0
36. det A = 4, det A1 = 16, det A2 = −4, det A3 = 0; x1 = = 4, x2 = = −1, x3 = = 0
4 4 4
37. det A = cos2 θ + sin2 θ, det A1 = X cos θ − Y sin θ, det A2 = Y cos θ + X sin θ;
x1 = X cos θ − Y sin θ, y = Y cos θ + X sin θ
38. (a) i1 − i2 − i3 − i4 = 0, i2 R1 = E, i2 R1 − i3 R2 = 0, i3 R2 − i4 R3 = 0
 
1 −1 −1 −1
0 R1 0 0
(b) det A = 
0 R1 −R2
 = R1 R2 R3 ;
0
0 0 R2 −R3
 
0 −1 −1 −1
E R1 0 0
det A1 = 
 0 R1 −R2
 = −E (−R2 R3 − R1 (R3 + R2 ))
0
0 0 R2 −R3
= E (R2 R3 + R1 R3 + R1 R2 ) ;

det A1 E(R2 R3 + R1 R3 + R1 R2 ) 1 1 1
i1 = = =E + +
det A R1 R2 R3 R1 R2 R3
      
2 3 −1 x1 6 −2 −3 −2
1
39. AX = B is  1 −2 0 x2  = −3. Since A−1 = − −1 0 −1, we have
3
−2 0 1 x3 9 −4 −6 −7
 
7
−1
X=A B=  5.
23
 3 1 9
   
2 −4 −4 1 −1
1 3  
     
40. (a) A−1 B = 
−1 2 2  1 =  1
 
1 1 1
2 −4 −4 1 0
3
− 41 − 49
 
  
2 −2 −10
1 3 
    
(b) A−1 B = 
−1 2 2   1 = 
  7

1
2 − 41 1
−4 3 −2
41. From the characteristic equation λ2 − 4λ − 5 = 0 we see that the eigenvalues are λ1 = −1

−1
and λ2 = 5. For λ1 = −1 we have 2k1 + 2k2 = 0, 4k1 4k2 = 0 and K1 = . For λ2 = 5
1

1
we have −4k1 + 2k2 = 0, 4k1 − 2k2 = 0 and K2 = .
2
42. From the characteristic equation λ2 = 0 we see that the eigenvalues are λ1 = λ2 = 0. For

0
λ1 = λ2 = 0 we have 4k1 = 0 and K1 = is a single eigenvector.
1
43. From the characteristic equation −λ3 + 6λ2 + 15λ + 8 = −(λ + 1)2 (λ − 8) = 0 we see that the
eigenvalues are λ1 = λ2 = −1 and λ3 = 8. For λ1 = λ2 = −1 we have
1 21 1 0
   
4 2 4 0
row
2 1 2 0 −−−−−−→ 0 0 0 0 .
   
operations
4 2 4 0 0 0 0 0
T T
Thus K1 = 1 −2 0 and K2 = 1 0 −1 . For λ3 = 8 we have
1 − 25 − 54 0
   
−5 4 0 2
row
1 − 21 0
   
 2 −8 2 0 −−−−−−→ 
  operations 0 .
4 2 −5 0 0 0 0 0
T
Thus K3 = 2 1 2 .
44. From the characteristic equation −λ3 + 18λ2 − 99λ + 162 = −(λ − 9)(λ − 6)(λ − 3) = 0 we
see that the eigenvalues are λ1 = 9, λ2 = 6, and λ3 = 3. For λ1 = 9 we have
   
−2 −2 0 0 1 1 0 0
−2 −3 row
2 0 −−−−−−→ 0 1 −2 0 .
0 2 −4 0 operations 0 0 0 0
T
Thus K1 = −2 2 1 . For λ2 = 6 we have
   
1 −2 0 0 1 −2 0 0
row
−2 0 2 0 −−−−−−→ 0 1 − 21 0 .
   
operations
0 2 −1 0 0 0 0 0
T
Thus K2 = 2 1 2 . For λ3 = 3 we have
1 − 12 0 0
   
4 −2 0 0
row
−2 3 2 0 −−−−−−→ 0 1 1 0 .
   
operations
0 2 2 0 0 0 0 0
T
Thus K3 = 1 2 −2 .
45. From the characteristic equation −λ3 − λ2 + 21λ + 45 = −(λ + 3)2 (λ − 5) = 0 we see that the
eigenvalues are λ1 = λ2 = −3 and λ3 = 5. For λ1 = λ2 = −3 we have
   
1 2 −3 0 1 2 −3 0
 2 row
4 −6 0 −−−−−−→ 0 0 0 0 .
−1 −2 3 0 operations 0 0 0 0
T T
Thus K1 = −2 1 0 and K2 = 3 0 1 . For λ3 = 5 we have
1 − 72 3
   
−7 2 −3 0 7 0
row
 2 −4 −6 0 −−−−−−→ 0 1 2 0 .
   
operations
−1 −2 −5 0 0 0 0 0
T
Thus K3 = −1 −2 1 .
46. From the characteristic equation −λ3 + λ2 + 2λ = −λ(λ + 1)(λ − 2) = 0 we see that the
eigenvalues are λ1 = 0, λ2 = −1, and λ3 = 2. For λ1 = 0 we have k3 = 0, 2k1 + 2k2 + k3 = 0
T
and K1 = 1 −1 0 . For λ2 = −1 we have
   
1 0 0 0 1 0 0 0
0 1 1 0 −−− row
−−−→ 0 1 1 0 .
2 2 2 0 operations 0 0 0 0
T
Thus K2 = 0 1 −1 . For λ3 = 2 we have
   
−2 0 0 0 1 0 0 0
row
 0 −2 1 0 −−−−−−→ 0 1 − 21 0 .
   
operations
2 2 −1 0 0 0 0 0
T
Thus K3 = 0 1 2 .
T
47. Let X1 = a b c be the first column of the matrix. Then
T 1 T 1
XT1 − √12 0 √1
2
= √ (c − a) = 0 and XT1 √1
3
√1
3
√1
3
= √ (a + b + c) = 0.
2 3
Also XT1 X1 = a2 + b2 + c2 = 1. We see that c = a and b = −2a from the first two equations.
T
Then a2 + 4a2 + a2 = 6a2 = 1 and a = √16 . Thus X1 = √16 − √26 √16 .
48. (a) Eigenvalues are λ1 = λ2 = 0 and λ3 = 5 with corresponding eigenvectors K1 =

T T T √
0 1 0 , K2 = 2 0 1 , and K3 = −1 0 2 . Since kK1 k = 1, kK2 k = 5 ,
√
and kK3 k = 5 , we have
0 1 0
   
0 √2 − √15
5
 2
and P−1 = PT =  0 √1  .
  
P= √
1 0 0 
  5 5
0 √1 √2
5 5 − √15 0 √2
5
 
0 0 0
(b) P−1 AP = 0 0 0
0 0 5
3
! !
1 2 − 21 0
49. We identify A = . Eigenvalues are λ1 = − 21 and λ2 = 25 so D =
3 5 and the
2 1 2 0

X
= − 21 X 2 + 25 Y 2 = 1. The graph is a hyperbola.

equation becomes X Y D
Y
50. We measure years in units of 10, with 0 corresponding to 1890. Then

T
T 0 1 2 3 4 T 30 10
Y = 63 76 92 106 123 and A = , so A A = . Thus
1 1 1 1 1 10 5

T −1 T 1 5 −10 T 15
X = (A A) A Y= A Y= ,
50 −10 30 62
and the least squares line is y = 15t + 62. At t = 5 (corresponding to 1940) we have y = 137.
The error in the predicted population is 5 million or 3.7%.
51. The encoded message is

10 1 19 1 20 5 12 12 9 20 5 0 12 1 21
B = AM =
9 1 14 3 8 5 4 0 15 14 0 6 18 9 0

204 13 208 55 124 120 105 214 50 6 138 19 210
= .
185 12 188 50 112 108 96 194 45 6 126 18 189
52. The encoded message is

10 1 19 5 3 0 1 7 14 20 0 1 18
B = AM =
9 1 18 22 19 0 20 21 5 19 0 1 13

208 72 49 0 30 91 145 219 0 11 193
= .
189 67 46 0 29 84 131 199 0 10 175

    
−3 2 −1 19 0 15 14 0 20 8 5 12 16 0 9
M = A−1 B =  1 0 0 35 10 27 53 1 54 = 19 0 15 14 0 20 .
2 −1 1 5 15 −3 48 2 39 8 5 0 23 1 25
From correspondence (1) we obtain: HELP IS ON THE WAY.

    
−3 2 −1 5 2 21 18 15 19
M = A−1 B =  1 0 0 27 17 40 =  5 2 21 .
2 −1 1 21 13 −2 4 0 0
From correspondence (1) we obtain: ROSEBUD .

55. (a) The parity is even so the decoded message is 1 1 0 0 1
(b) The parity is odd; there is a parity error.

 
   1  
c1 1 1 0 1   0
0
56. From c2  1 0 1 1 
0
 = 0 we obtain the code word 0 0 1 1 0 0 1 .
c3 0 1 1 1 1
1
     
1 1 1 1 1 1 − 2 R +R 1 1 1
−R1 +R2 2 3
57. 1 −2 3 −−−−−−→ 0 −3 2 −−3−−−−→ 0 −3 2 = U
−2R1 +R3
2 0 −3 0 −2 −5 0 0 −6.333
   
1 0 0 1 0 0
+R2
0 1 0 −−R−1− −−→ 1 1 0 = L
0 0 1 2R 1 +R3 2 0.667 1
− 32 R2 +R3
  
1 0 0 1 1 1
A = LU = 1 1 0 0 −3 2
2 0.667 1 0 0 −6.333
    
1 0 0 y1 6
LY = B = 1
 1 0  y2 = 2 ,
  hence y1 = 6, y2 = −4, y3 = −6.333
2 0.667 1 y3 3
    
1 1 1 x1 6
UX = Y = 0 −3 2   x2  =  −4 , hence x1 = 3, x2 = 2, x3 = 1
0 0 −6.333 x3 −6.333
     
1 0 1 1 1 1 − 4 R +R 1 0 1
−2R +R2  2 3
58. 2 3 4 −−−−1−−→ 0 3 2 −−3−−−−→ 0 3 2 = U
−R1 +R3
1 4 5 0 4 4 0 0 1.333
   
1 0 0 1 0 0
2R1 +R2 
0 1 0 −−−−−−→ 2 1 0 = L
0 0 1 R 1 +R3 1 1.333 1
4
R +R3
3 2
  
1 0 0 1 0 1
A = LU = 2 1 0 0 3 2
1 1.333 1 0 0 1.333
    
1 0 0 y1 4
LY = B = 2 1 0 y2  = 5 , hence y1 = 4, y2 = −3, y3 = 0
1 1.333 1 y3 0
    
1 0 1 x1 4
UX = Y = 0 3
 2   x2 = −3 ,
  hence x1 = 4, x2 = −1, x3 = 0
0 0 1.333 x3 0
59. The data yields

   
−2 1 1
 0 2 1 T 1 2 3 4
 5 ,
Y=  A=
3
, A =
1 1 1 1 1
−1 4 1
and
 
1 1
1 2 3 4 2  = 30 10 ,
1
AT A =
1 1 1 1 3 1 10 4
4 1
1 1
!
−1 5 −2
AT A =
− 12 3
2
Hence,
 
−2
1
− 12 4
! !
−1 T 5 1 2 3 4 
 0 = 5
X = AT A

A Y=
− 21 3 1 1 1 1  5 − 32
2
−1
Therefore, f (x) = 54 x − 32 .
60. The data yields

   
−2 1 1 1  
 0  4 1 4 9 16
2 1
Y=
 5 ,
 A=
 9
, AT = 1 2 3 4
3 1
1 1 1 1
−1 16 4 1
and
 
  1 1 1  
1 4 9 16  354 100 30
4 2 1
AT A = 1 2 3 4   9 3 1 = 100 30 10 ,
  
1 1 1 1 30 10 4
16 4 1
 
5 −25 25
−1 1
AT A

= −25 129 −135
20
25 −135 155
Hence,
 
 −2    
5 −25 25 1 4 9 16   −2
−1 T 1  0 
X = AT A

A Y= −25  5 =
129 −135 1 2 3 4  10.8
20
25 −135 155 1 1 1 1 −11.5
−1
Therefore, f (x) = −2x2 + 10.8x − 11.5 .

AEMx CH08

Uploaded by

Copyright:

Available Formats

AEMx CH08

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AEMx CH08

Uploaded by

Copyright:

Available Formats

Chapter 8

8.1 Matrix Algebra

6. 8 × 1 7. Not equal 8. Not equal 9. Not equal 10. Not equal

11. Solving x = y − 2, y = 3x − 2 we obtain x = 2, y = 4.

12. Solving x2 = 9, y = 4x we obtain x = 3, y = 12 and x = −3, t = −12.

13. c23 = 2(0) − 3(−3) = 9; c12 = 2(3) − 3(−2) = 12

14. c23 = 2(1) − 3(0) = 2; c12 = 2(−1) − 3(0) = −2

(d) Since AB is 1 × 1 and C is 3 × 3 the product (AB)C is not defined.

39. Since (A + B)2 = (A + B)(A + B) 6= A2 + AB + BA + B2 , and AB 6= BA in general,

40. Since (A + B)(A − B) = A2 − AB + BA − B2 , and AB 6= BA in general, (A + B)(A − B) 6=

41. a11 x1 + a12 x2 = b1 ; a21 x1 + a22 x2 = b2

8.2 Systems of Linear Algebraic Equations

The solution is x1 = 4, x2 = −7.

The solution is x1 = 8, x2 = 10.

The solution is x1 = 0, x2 = 4, x3 = −1.

The solution is x1 = 1, x2 = 2, x3 = −1, x4 = 0.

Since the bottom row implies 0 = 1, the system is inconsistent.

If x4 = t the solution is x1 = 19t, x2 = −10t, x3 = 2t, x4 = t.

The solution is x1 = 0.3, x2 = −0.12, x3 = 4.1.

The solution is x1 = 1.45, x2 = −1.62, x3 = 0.28.

23. From x1 Na + x2 H2 O → x3 NaOH + x4 H2 we obtain the system x1 = x3 , 2x2 = x3 + 2x4 ,

25. From x1 Fe3 O4 + x2 C → x3 Fe + x4 CO we obtain the system 3x1 = x3 , 4x1 = x4 , x2 = x4 .

26. From x1 C5 H8 + x2 O2 → x3 CO2 + x4 H2 O we obtain the system 5x1 = x3 , 8x1 = 2x4 ,

C5 H8 + 7O2 → 5CO2 + 4H2 O.

27. From x1 Cu + x2 HNO3 → x3 Cu(NO3 )2 + x4 H2 O + x5 NO we obtain the system

x1 = 3, x2 = 2x4 , x2 = 2x3 + x5 , 3x2 = 6x3 + x4 + x5 .

Letting x4 = t we see that x2 = 2t and

3x1 = x3 , 2x1 + x2 = 2x3 , 8x1 + 4x2 = 8x3 , 3x2 = 4x3 .

Ca3 (PO4 )2 + 4H3 PO4 → 3Ca(H2 PO4 )2 .

29. The system of equations is

Gaussian elimination gives

Gaussian elimination gives

The solution is i1 = 12, i2 = 8, i3 = 4.

35. Interchange row 1 and row in I3 . 36. Multiply row 3 by c in I3 .

37. Add c times row 2 to row 3 in I3 . 38. Add row 4 to row 1 in I4 .

43. Using the Solve function in Mathematica we find x1 = −0.0717393 − 1.43084c,

45. Using the Solve function in Mathematica we find x1 = −3.76993, x2 = −1.09071,

46. Using the Solve function in Mathematica we find x1 = 83 − 73 b + 23 c, x2 = 2

8.3 Rank of a Matrix

(b) The system is inconsistent if rank(A) < rank(A/B) = 2 and consistent if

9. −7 10. 2 11. 17 12. −1/2

13. (1 − λ)(2 − λ) − 6 = λ2 − 3λ − 4 14. (−3 − λ)(5 − λ) − 8 = λ2 − 2λ − 23

= (4y + yz − 3z − yz) − (4x + xz − 2z − xz) + (3x + xy − 2y − xy)

28. Expanding along the bottom row we obtain

29. Solving λ2 − 2λ − 15 − 20 = λ2 − 2λ − 35 = (λ − 7)(λ + 5) = 0 we obtain λ = 7 and −5.

30. Solving −λ3 + 3λ2 − 2λ = −λ(λ − 2)(λ − 1) = 0 we obtain λ = 0, 1, and 2.

8.5 Properties of Determinants

1. Theorem 8.5.4 2. Theorem 8.5.7

3. Theorem 8.5.7 4. Theorem 8.5.4 and 8.5.5

5. Theorem 8.5.5 (twice) 6. Theorem 8.5.4 (twice)

7. Theorem 8.5.3 8. Theorem 8.5.2 and 8.5.5

9. Theorem 8.5.1 10. Theorem 8.5.4 (twice)

11. det A = −5 12. det B = 2(3)(5) = 30

13. det C = −5 14. det D = 5

15. By Theorem 8.5.7

16. By Theorem 8.5.5,

Then by Theorem 8.5.7 and Theorem 8.5.5,

17. det A = 6( 23 )(−4)(−5) = 80 18. det B = −a13 a22 a31