# Chapter 1

Systems of Linear Equations and Matrices
Section 1.1
Exercise Set 1.1
1. (a), (c), and (f) are linear equations in x1 , x2 , and x3 .
(b) is not linear because of the term x1 x3 .
(d) is not linear because of the term x −2 .
(e) is not linear because of the term x13/ 5 .
3. (a) and (d) are linear systems.
(b) is not a linear system because the first and second equations are not linear.
(c) is not a linear system because the second equation is not linear.
5. By inspection, (a) and (d) are both consistent; x1 = 3, x2 = 2, x3 = −2, x4 = 1 is a solution of (a) and

x1 = 1, x2 = 3, x3 = 2, x4 = 2 is a solution of (d). Note that both systems have infinitely many solutions.
7. (a), (d), and (e) are solutions.
(b) and (c) do not satisfy any of the equations.
9. (a) 7 x − 5 y = 3
5
3
x= y+
7
7
Let y = t. The solution is
5
3
x = y+
7
7
y=t
(b) −8 x1 + 2 x2 − 5 x3 + 6 x4 = 1
1
5
3
1
x1 = x2 − x3 + x4 −
4
8
4
8
Let x2 = r , x3 = s, and x4 = t . The solution is

1
5
3 1
r− s+ t−
4
8
4 8
=r
=s
=t

x1 =

x2
x3
x4

⎡2 0 0⎤
11. (a) ⎢ 3 −4 0 ⎥ corresponds to

1 1⎥⎦
⎢⎣ 0
2 x1
=0
3x1 − 4 x2 = 0.
x2 = 1

1

Chapter 1: Systems of Linear Equations and Matrices

SSM: Elementary Linear Algebra

⎡ 3 0 −2 5⎤
(b) ⎢7
1 4 −3⎥⎥ corresponds to

1 7 ⎥⎦
⎢⎣ 0 −2
3x1
− 2 x3 = 5
7 x1 + x2 + 4 x3 = −3.
−2 x2 + x3 = 7

15. If (a, b, c) is a solution of the system, then

ax12 + bx1 + c = y1 , ax22 + bx2 + c = y2 , and
ax32 + bx3 + c = y3 which simply means that the
points are on the curve.
17. The solutions of x1 + kx2 = c are x1 = c − kt ,

x2 = t where t is any real number. If these
satisfy x1 + lx2 = d , then c − kt + lt = d or

(c) ⎢7 2 1 −3 5⎥ corresponds to
⎣ 1 2 4 0 1⎦
7 x1 + 2 x2 + x3 − 3x4 = 5
.
x1 + 2 x2 + 4 x3
=1
⎡1

(d) ⎢0
⎢0
⎢⎣0
x1

0
1
0
0

0
0
1
0

c − d = (k − l) t for all real numbers t. In
particular, if t = 0, then c = d, and if t = 1, then
k = l.
True/False 1.1

0 7⎤
0 −2 ⎥ corresponds to

0
3⎥
1 4 ⎥⎦

x2
x3
x4

(a) True; x1 = x2 =

(b) False; only multiplication by nonzero constants
is acceptable.

= 7
= −2 .
3
=
= 4

(c) True; if k = 6 the system has infinitely many
solutions, while if k ≠ 6, the system has no
solution.
(d) True; the equation can be solved for one variable
in terms of the other(s), yielding parametric
equations that give infinitely many solutions.

13. (a) The augmented matrix for −2 x1 = 6
3x1 = 8
9 x1 = −3

(e) False; the system 3x − 5 y = −7 has the
2 x + 9 y = 20
6 x − 10 y = −14
solution x = 1, y = 2.

⎡ −2 6 ⎤
is ⎢ 3 8⎥ .

⎢⎣ 9 −3⎥⎦
(b) The augmented matrix for

6 x1 − x2 + 3x3 = 4 is ⎢6 −1 3 4 ⎥ .
⎣0 5 −1 1⎦
5x − x = 1
2

(f) False; multiplying an equation by a nonzero
constant c does not change the solutions of the
system.

3

(g) True; subtracting one equation from another is
the same as multiplying an equation by −1 and

(c) The augmented matrix for
2 x2
− 3x4 + x5 = 0
−3x1 − x2 + x3
= −1
6 x1 + 2 x2 − x3 + 2 x4 − 3x5 = 6

(h) False; the second row corresponds to the
equation 0 x1 + 0 x2 = −1 or 0 = −1 which is false.

⎡ 0 2 0 −3
1 0⎤
is ⎢ −3 −1 1 0 0 −1⎥ .

⎢⎣ 6 2 −1 2 −3 6 ⎥⎦

Section 1.2
Exercise Set 1.2

(d) The augmented matrix for x1 − x5 = 7 is

[1

0

= xn = 0 will be a solution.

0

0

−1

1. (a) The matrix is in both row echelon and
reduced row echelon form.

7].

(b) The matrix is in both row echelon and
reduced row echelon form.

2

SSM: Elementary Linear Algebra

Section 1.2

(c) The matrix is in both row echelon and
reduced row echelon form.

(d) The last line of the matrix corresponds to the
equation 0 x1 + 0 x2 + 0 x3 = 1, which is not

(d) The matrix is in both row echelon and
reduced row echelon form.

satisfied by any values of x1 , x2 , and x3 ,
so the system is inconsistent.
⎡ 1
1 2 8⎤
5. The augmented matrix is ⎢⎢ −1 −2 3 1⎥⎥ .
⎢⎣ 3 −7 4 10 ⎥⎦
times the first row to the third row.
⎡1
8⎤
1 2
⎢0
9⎥
−1 5

⎢⎣0 −10 −2 −14⎥⎦

(e) The matrix is in both row echelon and
reduced row echelon form.
(f) The matrix is in both row echelon and
reduced row echelon form.
(g) The matrix is in row echelon form.
3. (a) The matrix corresponds to the system
x1 − 3x2 + 4 x3 = 7
x1 = 7 + 3x2 − 4 x3
x2 + 2 x3 = 2 or x2 = 2 − 2 x3
.
x3 = 5
x3 = 5

Multiply the second row by −1.
⎡1
8⎤
1 2
⎢0
1 −5 −9⎥⎥

⎢⎣0 −10 −2 −14⎥⎦

Thus x3 = 5, x2 = 2 − 2(5) = −8, and
x1 = 7 + 3(−8) − 4(5) = −37. The solution is
x1 = −37, x2 = −8, x3 = 5.

Add 10 times the second row to the third row.
⎡1 1
8⎤
2
⎢ 0 1 −5
−9 ⎥

⎣⎢0 0 −52 −104 ⎦⎥

(b) The matrix corresponds to the system
x1
+ 8 x3 − 5 x4 = 6
x2 + 4 x3 − 9 x4 = 3 or
x3 + x4 = 2

Multiply the third row by −

x1 = 6 − 8 x3 + 5 x4
x2 = 3 − 4 x3 + 9 x4 .
x3 = 2 − x4

1
.
52

⎡1 1 2
8⎤
⎢0 1 −5 −9 ⎥

1 2 ⎥⎦
⎢⎣0 0
Add 5 times the third row to the second row and
−2 times the third row to the first row.
⎡ 1 1 0 4⎤
⎢0 1 0 1⎥

⎢⎣0 0 1 2⎥⎦

Let x4 = t , then x3 = 2 − t ,
x2 = 3 − 4(2 − t ) + 9t = 13t − 5, and
x1 = 6 − 8(2 − t ) + 5t = 13t − 10. The solution
is x1 = 13t − 10, x2 = 13t − 5, x3 = −t + 2,
x4 = t .

Add −1 times the second row to the first row.
⎡ 1 0 0 3⎤
⎢0 1 0 1⎥

⎢⎣0 0 1 2⎥⎦

(c) The matrix corresponds to the system
x1 + 7 x2 − 2 x3
− 8 x5 = −3
x3 + x4 + 6 x5 = 5 or
x4 + 3x5 = 9

x1 = −3 − 7 x2 + 2 x3 + 8 x5
x3 = 5 − x4 − 6 x5
x4 = 9 − 3 x5

The solution is x1 = 3, x2 = 1, x3 = 2.
7. The augmented matrix is
⎡ 1 −1 2 −1 −1⎤
⎢ 2 1 −2 −2 −2 ⎥

⎥.
1 1⎥
⎢ −1 2 −4
⎣⎢ 3 0 0 −3 −3⎦⎥

Let x2 = s and x5 = t , then x4 = 9 − 3t ,
x3 = 5 − (9 − 3t ) − 6t = −3t − 4, and
x1 = −3 − 7 s + 2(−3t − 4) + 8t = −7 s + 2t − 11.
The solution is x1 = −7 s + 2t − 11, x2 = s,

Add −2 times the first row to the second row, 1
times the first row to the third row, and −3 times
the first row to the fourth row.

x3 = −3t − 4, x4 = −3t + 9, x5 = t .

3

Chapter 1: Systems of Linear Equations and Matrices

SSM: Elementary Linear Algebra

⎡ 1 −1 2 −1 −1⎤
⎢0 3 −6 0 0 ⎥

⎢0 1 −2 0 0 ⎥
⎢⎣0 3 −6 0 0 ⎥⎦

13. Since the system has more unknowns (4) than
equations (3), it has nontrivial solutions.
15. Since the system has more unknowns (3) than
equations (2), it has nontrivial solutions.

1
3
times the new second row to the third row and
add −3 times the new second row to the fourth
row.
⎡ 1 −1 2 −1 −1⎤
⎢0 1 −2 0 0 ⎥

⎢0 0 0 0 0 ⎥
⎣⎢0 0 0 0 0 ⎦⎥
Add the second row to the first row.
⎡ 1 0 0 −1 −1⎤
⎢ 0 1 −2 0 0 ⎥

⎢0 0 0 0 0⎥
⎢⎣0 0 0 0 0⎥⎦
The corresponding system of equations is
x = w −1
− w = −1
x
or
.
=0
y = 2z
y − 2z
Let z = s and w = t. The solution is x = t − 1,
y = 2s, z = s, w = t.
Multiply the second row by

⎡2 1 3 0⎤
17. The augmented matrix is ⎢⎢ 1 2 0 0 ⎥⎥ .
⎢⎣ 0 1 1 0 ⎥⎦
Interchange the first and second rows, then add
−2 times the new first row to the second row.
⎡ 1 2 0 0⎤
⎢0 −3 3 0 ⎥

1 1 0 ⎥⎦
⎢⎣0
1
Multiply the second row by − , then add −1
3
times the new second row to the third row.
⎡ 1 2 0 0⎤
⎢0 1 −1 0 ⎥

⎣⎢0 0 2 0 ⎦⎥
1
2
third row to the second row.
⎡ 1 2 0 0⎤
⎢0 1 0 0⎥

⎢⎣0 0 1 0⎥⎦
Multiply the third row by

9. In Exercise 5, the following row echelon matrix
occurred.
⎡1 1 2
8⎤
⎢0 1 −5 −9 ⎥

1 2 ⎥⎦
⎢⎣0 0

Add −2 times the second row to the first row.
⎡ 1 0 0 0⎤
⎢0 1 0 0⎥

⎣⎢0 0 1 0⎦⎥

The corresponding system of equations is
x1 + x2 + 2 x3 = 8
x1 = − x2 − 2 x3 + 8
x2 − 5 x3 = −9 or x2 = 5 x3 − 9
.
x3 = 2
x3 = 2

The solution, which can be read from the matrix,
is x1 = 0, x2 = 0, x3 = 0.

Since x3 = 2, x2 = 5(2) − 9 = 1, and
x1 = −1 − 2(2) + 8 = 3. The solution is x1 = 3,

19. The augmented matrix is ⎢3 1 1 1 0⎥ .
⎣ 5 −1 1 −1 0 ⎦
1
Multiply the first row by , then add −5 times
3
the new first row to the second row.
1
1
1 0⎤
⎡1
3
3
3

⎢⎣0 − 83 − 23 − 83 0 ⎥⎦
3
Multiply the second row by − .
8

x2 = 1, x3 = 2.
11. From Exercise 7, one row echelon form of the
⎡ 1 −1 2 −1 −1⎤

augmented matrix is ⎢0 1 −2 0 0 ⎥ .
⎢0 0 0 0 0 ⎥
⎢⎣0 0 0 0 0 ⎥⎦
The corresponding system is
x = y − 2z + w −1
x − y + 2 z − w = −1
or
.
y = 2z
y − 2z
=0
Let z = s and w = t. Then y = 2s and
x = 2s − 2s + t − 1 = t − 1. The solution is
x = t − 1, y = 2s, z = s, w = t.

⎡1 1
3

⎢⎣0 1
4

1
3
1
4

0⎤

1 0 ⎥⎦

1
3

SSM: Elementary Linear Algebra

Section 1.2
w= y
=0
= 0 or x = − y .
z=0
z=0
Let y = t. The solution is w = t, x = −t, y = t,
z = 0.
w

1
times the second row to the first row.
3

⎡ 1 0 1 0 0⎤
4

⎢⎣0 1 14 1 0 ⎥⎦
This corresponds to the system
1
1
+ x3
=0
x1
x1 = − x3
4
4
or
.
1
1
x2 + x3 + x4 = 0
x2 = − x3 − x4
4
4
Let x3 = 4s and x4 = t . Then x1 = − s and

−y
x+ y

⎡ 2 −1 3 4 9 ⎤
⎢ 1 0 −2 7 11⎥
23. The augmented matrix is ⎢
⎥.
1 5 8⎥
⎢ 3 −3
1 4 4 10⎦⎥
⎣⎢ 2
Interchange the first and second rows.
⎡ 1 0 −2 7 11⎤
⎢ 2 −1 3 4 9 ⎥

1 5 8⎥
⎢ 3 −3
1 4 4 10⎥⎦
⎢⎣ 2
Add −2 times the first row to the second and
fourth rows, and add −3 times the first row to the
third row.
7
11⎤
⎡ 1 0 −2
⎢0 −1 7 −10 −13⎥

⎢0 −3 7 −16 −25⎥
1 8 −10 −12⎦⎥
⎣⎢0

x2 = − s − t . The solution is x1 = − s, x2 = − s − t ,
x3 = 4s, x4 = t .

⎡ 0 2 2 4 0⎤
⎢ 1 0 −1 −3 0 ⎥
21. The augmented matrix is ⎢
⎥.
⎢ 2 3 1 1 0⎥
⎢⎣ −2 1 3 −2 0 ⎥⎦
Interchange the first and second rows.
⎡ 1 0 −1 −3 0 ⎤
⎢ 0 2 2 4 0⎥

⎢ 2 3 1 1 0⎥
⎣⎢ −2 1 3 −2 0 ⎦⎥

Multiply the second row by −1, then add 3 times
the new second row to the third row and −1
times the new second row to the fourth row.
7
11⎤
⎡ 1 0 −2
⎢0 1 −7
10
13⎥

⎢0 0 −14 14 14 ⎥
15 −20 −25⎦⎥
⎣⎢0 0

Add −2 times the first row to the third row and 2
times the first row to the fourth row.
⎡ 1 0 −1 −3 0 ⎤
⎢0 2 2 4 0 ⎥

⎢0 3 3 7 0 ⎥
⎢⎣0 1 1 −8 0 ⎥⎦
1
Multiply the second row by , then add −3
2
times the new second row to the third row and
−1 times the new second row to the fourth row.
⎡ 1 0 −1 −3 0 ⎤
⎢0 1 1
2 0⎥

1 0⎥
⎢0 0 0
⎣⎢0 0 0 −10 0 ⎦⎥

1
14
times the new third row to the fourth row.
11⎤
⎡ 1 0 −2 7
⎢0 1 −7 10
13⎥

1 −1 −1⎥
⎢0 0
⎣⎢0 0 0 −5 −10 ⎦⎥
Multiply the third row by −

1
Multiply the fourth row by − , then add the
5
new fourth row to the third row, add −10 times
the new fourth row to the second row, and add
−7 times the new fourth row to the first row.
⎡ 1 0 −2 0 −3⎤
⎢ 0 1 −7 0 −7 ⎥

1 0
1⎥
⎢0 0
0 1 2 ⎦⎥
⎣⎢0 0
Add 7 times the third row to the second row and
2 times the third row to the first row.

Add 10 times the third row to the fourth row, −2
times the third row to the second row, and 3
times the third row to the first row.
⎡ 1 0 −1 0 0 ⎤
⎢0 1 1 0 0 ⎥

⎢0 0 0 1 0 ⎥
⎢⎣0 0 0 0 0 ⎥⎦
The corresponding system is

5

2 X = 1 ⇒ x = ±1 Y=3⇒ y=± 3 ⎡ 1 0 0 0⎤ ⎢0 1 0 0⎥ ⎢ ⎥ ⎣0 0 1 0⎦ The system has only the trivial solution.SSM: Elementary Linear Algebra Section 1. For 0 ≤ β ≤ 2π . 2 2 For 0 ≤ γ ≤ 2π . π . 9 and −1 times the fourth row to the first row. then add the 2 new second row to the third row. cos β = 0 ⇒ β = Z=2⇒ z=± 2 The solutions are x = ±1. 7): a + b + c + d = 7 (3. then the system is X +Y + Z = 6 X − Y + 2Z = 2 2X + Y − Z = 3 ⎡ 1 1 1 6⎤ The augmented matrix is ⎢ 1 −1 2 2⎥ . − =1 =3. ⎡ 1 1 1 6⎤ ⎢ 0 −2 1 −4 ⎥ ⎢ ⎥ ⎣ 0 −1 −3 −9 ⎦ Add −27 times the first row to the second row and −64 times the first row to the third row. ⎢ ⎥ ⎣0 0 1 2⎦ This corresponds to the system 1 1 1 43 1 13 9 19 12 1 1 . 2π . 4 7 ⎤ 100 ⎥ 9 ⎥ 107 ⎥ 6 ⎥ 10 ⎦⎥ 19 times the fourth row to the third row. π . −11): 27a + 9b + 3c + d = −11 (4. which corresponds to sin α = 0. 1 1 1 7⎤ ⎡1 ⎢0 −18 −24 −26 −200 ⎥ ⎢ ⎥ ⎢0 −48 −60 −63 −462 ⎥ 0 0 1 10 ⎦⎥ ⎣⎢0 1 . ⎡ 1 1 0 4⎤ ⎢ 0 1 0 3⎥ ⎢ ⎥ ⎣0 0 1 2⎦ Multiply the third row by ⎡1 ⎢0 ⎢ ⎢0 ⎢ ⎣⎢0 X Y 0 1 0 0 Add − Add −1 times the second row to the first row. ⎢ 64 16 4 1 −14⎥ ⎣⎢ 0 0 0 1 10⎦⎥ π 3π . . 12 13 times the fourth row to the second row. ⎡ 1 0 0 1⎤ ⎢ 0 1 0 3⎥ . then add 2 7 times the new third row to the second row and −1 times the new third row to the first row. Let X = x 2 . 37. Thus. then add 48 18 times the new second row to the third row. tan γ = 0 ⇒ γ = 0. 10): d = 10 (1. sin α = 0 ⇒ α = 0. and Z = z 2 . ⎡1 1 1 6⎤ ⎢ ⎥ 2⎥ ⎢0 1 − 12 ⎢ ⎥ 7 ⎢⎣0 0 − 2 −7 ⎥⎦ 2 1 Multiply the third row by − . the original system has 3 ⋅ 2 ⋅ 3 = 18 solutions. 35. cos β = 0. Y = y 2 . (0. For 0 ≤ α ≤ 2π . 7 ⎤ ⎡1 1 1 1 ⎢0 1 4 13 100 ⎥ ⎢ 3 9 9 ⎥ ⎢0 0 4 19 214 ⎥ 3 3 ⎥ ⎢ ⎣⎢0 0 0 1 10 ⎦⎥ Multiply the second row by − 1 Multiply the second row by − . 2π . z = ± 2. y = ± 3. Z =2 7 . −14): 64a + 16b + 4c + d = −14 The system is a +b +c+d = 7 27a + 9b + 3c + d = −11 and the augmented 64a + 16b + 4c + d = −14 d = 10 1 1 1 1 7⎤ ⎡ ⎢ 27 9 3 1 −11⎥ matrix is ⎢ ⎥. tan γ = 0. ⎢ ⎥ ⎣ 2 1 −1 3⎦ Add −1 times the first row to the second row and −2 times the first row to the third row.

⎢ ⎥ ⎣0 0 0 ⎦ ⎡ 1 p q ⎤ ⎡0 1 p ⎤ ⎢0 0 0 ⎥ . the sequence of matrices is b ⎤ ⎡1 ⎡ a b ⎤ ⎡1 ba ⎤ a ⎥ ⎢ → → ⎢⎣ c d ⎥⎦ ⎢ c d ⎥ −bc ⎥ ad ⎢0 ⎣⎢ ⎦⎥ a ⎦ ⎣ ⎡1 b ⎤ 1 0⎤ a⎥ → ⎡ →⎢ ⎢⎣0 1 ⎥⎦ . ⎢0 1 q ⎥ . ⎢0 0 0 1⎥ ⎢0 0 0 1⎥ ⎢⎣0 0 0 0 ⎥⎦ ⎢⎣0 0 0 0⎥⎦ ⎡0 1 0 0⎤ ⎡ 1 0 p q ⎤ ⎢0 0 1 0⎥ ⎢ 0 1 r s ⎥ ⎢ ⎥. ⎢0 0 0 0 ⎥ ⎢ 0 0 0 0 ⎥ ⎢⎣0 0 0 0 ⎥⎦ ⎢⎣ 0 0 0 0 ⎥⎦ If a = 0. Since the homogeneous system has only the trivial solution. ⎢ ⎥. d = 10. using the same steps of GaussJordan elimination will reduce the augmented matrix of the nonhomogeneous system to the ⎡ 1 0 0 d1 ⎤ form ⎢0 1 0 d 2 ⎥ . 0⎥ 0 ⎦⎥ 0 1 p⎤ 0 0 0⎥ ⎥. ⎢0 0 0 ⎥ . the sequence of matrices is ⎡1 d ⎤ ⎡0 b ⎤ ⎡ c d ⎤ ⎡1 dc ⎤ c ⎢⎣ c d ⎥⎦ → ⎢⎣0 b ⎥⎦ → ⎢ 0 b ⎥ → ⎢ 0 1 ⎥ ⎣⎢ ⎦⎥ ⎣⎢ ⎦⎥ ⎡1 0 ⎤ . p and q are any real numbers. so the augmented ⎡a b k ⎤ matrix ⎢ will reduce to ⎣ c d l ⎥⎦ ⎡ 1 0 k1 ⎤ ⎢0 1 l ⎥ and the system has exactly one 1⎦ ⎣ solution. 0 0 0⎥ 0 0 0 ⎦⎥ 0 1⎤ ⎡0 0 ⎢0 0 0 0⎥ ⎥ . 0⎥ 0 ⎦⎥ ⎡0 ⎢0 ⎢ ⎢0 ⎣⎢0 q 0 0 0 r⎤ 0⎥ ⎥. ⎢ ⎥. ⎢0 0 1⎥ . →⎢ ⎣ 0 1 ⎥⎦ ⎡0 ⎢0 ⎢ ⎢0 ⎣⎢0 0 1 0⎤ ⎡ 1 p 0 0 1⎥ ⎢ 0 0 ⎥. b = −6. ⎢⎣ 0 1 ⎥⎦ ⎡1 p ⎢0 0 ⎢ ⎢0 0 ⎢⎣0 0 q 0⎤ 0 1⎥ ⎥. ⎢0 0 1 0⎥ ⎢ 0 0 1 r ⎥ ⎢⎣0 0 0 1⎥⎦ ⎢⎣ 0 0 0 0 ⎥⎦ ⎡ 1 0 p 0 ⎤ ⎡ 1 p 0 0⎤ ⎢0 1 q 0 ⎥ ⎢0 0 1 0⎥ ⎢ ⎥. ⎡ 1 p 0 q⎤ ⎢0 0 1 r ⎥ ⎢ ⎥. ⎢ ⎥ ⎣ 0 0 1 d3 ⎦ Thus the system has exactly one solution. ⎡ 1 0 0 0⎤ ⎡ 1 0 0 p ⎤ ⎢0 1 0 0⎥ ⎢ 0 1 0 q ⎥ ⎢ ⎥. and ⎢ 0 0⎥ ⎢0 0 0 0 ⎥⎦ ⎢⎣0 0 0 0 0 0 0⎤ 0⎥ ⎥ 0⎥ 0⎥⎦ . and s are any real numbers. (a) There are eight possibilities. q. ⎡ 1 1 0 0 −5⎤ ⎢0 1 0 0 −6 ⎥ ⎢ ⎥ ⎢0 0 1 0 2 ⎥ ⎢⎣0 0 0 1 10 ⎥⎦ Add −1 times the second row to the first row. 0 0⎥ 0 0 ⎥⎦ ⎡0 1 0 p ⎤ ⎡ 0 1 p 0 ⎤ ⎢0 0 1 q ⎥ ⎢ 0 0 0 1⎥ ⎢ ⎥. ⎢ ⎥. ⎡0 ⎢0 ⎢ ⎢0 ⎢⎣0 8 0 0 0 0 q⎤ 0⎥ ⎥. c = 2. 39. 1⎤ ⎡1 0 0 0 ⎢0 1 0 0 −6 ⎥ ⎢ ⎥ ⎢0 0 1 0 2 ⎥ ⎣⎢0 0 0 1 10 ⎦⎥ ⎡1 ⎢0 ⎢ ⎢0 ⎢0 ⎣ 43. ⎢ 0 0 0⎥ ⎢ 0 0 0 0 0⎦⎥ ⎢⎣ 0 0 ⎡0 1 p ⎢0 0 0 ⎢ ⎢0 0 0 ⎣⎢0 0 0 (b) Since ad − bc ≠ 0. (a) If a ≠ 0. ⎢0 0 0 0 ⎥ ⎢⎣0 0 0 0 ⎥⎦ 41. ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣0 0 1⎦ ⎣0 0 0 ⎦ ⎣0 0 0⎦ ⎡0 1 0 ⎤ ⎢0 0 1⎥ . ⎢ ⎥ ⎢ ⎥ ⎣0 0 0 ⎦ ⎣0 0 0 ⎦ ⎡0 0 0 ⎤ ⎡0 0 1⎤ ⎢0 0 0 ⎥ . ⎢ ⎥. the lines are neither identical nor parallel. ⎡ 1 0 0 ⎤ ⎡ 1 0 p ⎤ ⎡ 1 p 0⎤ ⎢0 1 0 ⎥ .Chapter 1: Systems of Linear Equations and Matrices SSM: Elementary Linear Algebra 1 1 0 −3 ⎤ ⎥ 1 43 0 − 10 3⎥ 0 1 0 2 ⎥ 0 0 1 10 ⎥⎦ 4 Add − times the third row to the second row 3 and −1 times the third row to the first row. r. and ⎢0 0 0 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣0 0 0 ⎦ ⎣0 0 0 ⎦ (b) There are 16 possibilities. ⎢0 0 0 1⎥ ⎢ 0 0 0 0 ⎥ ⎣⎢0 0 0 0⎦⎥ ⎣⎢ 0 0 0 0 ⎦⎥ The coefficients are a = 1. p.

(b) AC + D is 4 × 2.SSM: Elementary Linear Algebra Section 1. (d) True. (d) AB + B is undefined since AB is undefined. (a) D + E = ⎢ −1 + (−1) 0 + 1 1 + 2 ⎥ ⎢ ⎥ 2 + 1 4 + 3⎦ ⎣ 3+ 4 ⎡ 7 6 5⎤ = ⎢ −2 1 3⎥ ⎢ ⎥ ⎣ 7 3 7⎦ True/False 1. ⎡0 0⎤ (h) A − A = ⎢ 0 0 ⎥ ⎢ ⎥ ⎣0 0⎦ (i) tr(D) = 1 + 0 + 4 = 5 9 . adding a nonzero multiple of the first row to any other row will result in a matrix that is not in row echelon form. (e) True. a column cannot contain more than one leading 1.3 1. a row of zeros does not indicate infinitely many solutions. since there are n leading 1’s. (f) E(AC) is 5 × 2. the system could be inconsistent. (a) BA is undefined. (i) False. ⎡ −7 −28 −14 ⎤ (d) −7C = ⎢ ⎣ −21 −7 −35⎥⎦ (h) False.3 5 + 1 2 + 3⎤ ⎡ 1+ 6 3. 4 12 ⎤ ⎡ 2 10 4 ⎤ 4 8 ⎥ − ⎢ −2 0 2 ⎥ ⎥ ⎢ ⎥ 4 12 ⎦ ⎣ 6 4 8⎦ −6 8 ⎤ 4 6⎥ ⎥ 0 4⎦ (g) −3( D + 2 E ) ⎛ ⎡ 1 5 2⎤ ⎡ 12 2 6 ⎤ ⎞ = −3 ⎜ ⎢ −1 0 1⎥ + ⎢ −2 2 4 ⎥ ⎟ ⎜⎜ ⎢ ⎥ ⎢ ⎥ ⎟⎟ ⎝ ⎣ 3 2 4⎦ ⎣ 8 2 6 ⎦ ⎠ ⎡ 13 7 8⎤ = −3 ⎢ −3 2 5 ⎥ ⎢ ⎥ ⎣ 11 4 10 ⎦ ⎡ −39 −21 −24 ⎤ = ⎢ 9 −6 −15⎥ ⎢ ⎥ ⎣ −33 −12 −30 ⎦ (c) AE + B is undefined. reduced row echelon form is a row echelon form. if the system has more equations than unknowns. (b) False. ⎡ 1− 6 (b) D − E = ⎢ −1 − (−1) ⎢ ⎣ 3− 4 ⎡ −5 4 = ⎢ 0 −1 ⎢ ⎣ −1 1 (c) False. see Exercise 31. (g) E T A is undefined since ET is 4 × 5. the sum of the number of leading 1’s and the number of free variables is n. (g) True. so only the trivial solution is possible. for instance.2 (a) True. (e) E(A + B) is 5 × 5. 5 − 1 2 − 3⎤ 0 −1 1 − 2 ⎥ ⎥ 2 − 1 4 − 3⎦ −1⎤ −1⎥ ⎥ 1⎦ ⎡ 15 0 ⎤ (c) 5 A = ⎢ −5 10 ⎥ ⎢ ⎥ ⎣ 5 5⎦ (f) False. there are no free variables. (e) 2B − C is not defined since 2B is a 2 × 2 matrix and C is a 2 × 3 matrix.3 Exercise Set 1. there can be nonzero entries above the leading 1 in a column. in row echelon form. ⎡ 24 (f) 4 E − 2 D = ⎢ −4 ⎢ ⎣ 16 ⎡ 22 = ⎢ −2 ⎢ ⎣ 10 Section 1. (h) ( AT + E ) D is 5 × 2. since AE is 4 × 4 not 4 × 5.

(a) AB = ⎢ −1 2 ⎥ ⎢ ⎢ ⎥ ⎣ 0 2 ⎥⎦ ⎣ 1 1⎦ ⎡ (3 ⋅ 4) + (0 ⋅ 0) −(3 ⋅1) + (0 ⋅ 2) ⎤ = ⎢ −(1 ⋅ 4) + (2 ⋅ 0) (1⋅1) + (2 ⋅ 2) ⎥ ⎢ ⎥ ⎣ (1 ⋅ 4) + (1 ⋅ 0) −(1 ⋅1) + (1 ⋅ 2) ⎦ ⎡ 12 −3⎤ = ⎢ −4 5 ⎥ ⎢ ⎥ 1⎦ ⎣ 4 (b) BA is not defined since B is a 2 × 2 matrix and A is 3 × 2.Chapter 1: Systems of Linear Equations and Matrices (j) SSM: Elementary Linear Algebra ⎛ ⎡ 1 5 2 ⎤ ⎡ 18 3 9 ⎤ ⎞ tr( D − 3E ) = tr ⎜ ⎢ −1 0 1⎥ − ⎢ −3 3 6 ⎥ ⎟ ⎜⎜ ⎢ ⎥ ⎢ ⎥ ⎟⎟ ⎝ ⎣ 3 2 4 ⎦ ⎣ 12 3 9 ⎦ ⎠ ⎛ ⎡ −17 2 −7 ⎤ ⎞ = tr ⎜ ⎢ 2 −3 −5⎥ ⎟ ⎜⎜ ⎢ ⎥ ⎟⎟ ⎝ ⎣ −9 −1 −5⎦ ⎠ = −17 − 3 − 5 = −25 ⎛ ⎡ 28 −7 ⎤ ⎞ (k) 4 tr(7 B) = 4 tr ⎜ ⎢ ⎥⎟ ⎝ ⎣ 0 14 ⎦ ⎠ = 4(28 + 14) = 4(42) = 168 (l) tr(A) is not defined because A is not a square matrix. ⎡ 18 3 9 ⎤ ⎡ 1 5 2 ⎤ (c) (3E ) D = ⎢ −3 3 6 ⎥ ⎢ −1 0 1⎥ ⎢ ⎥⎢ ⎥ ⎣ 12 3 9 ⎦ ⎣ 3 2 4 ⎦ ⎡ (18 ⋅1) − (3 ⋅1) + (9 ⋅ 3) (18 ⋅ 5) + (3 ⋅ 0) + (9 ⋅ 2) (18 ⋅ 2) + (3 ⋅1) + (9 ⋅ 4) ⎤ = ⎢ −(3 ⋅1) − (3 ⋅1) + (6 ⋅ 3) −(3 ⋅ 5) + (3 ⋅ 0) + (6 ⋅ 2) −(3 ⋅ 2) + (3 ⋅1) + (6 ⋅ 4) ⎥ ⎢ ⎥ ⎣ (12 ⋅1) − (3 ⋅1) + (9 ⋅ 3) (12 ⋅ 5) + (3 ⋅ 0) + (9 ⋅ 2) (12 ⋅ 2) + (3 ⋅1) + (9 ⋅ 4) ⎦ ⎡ 42 108 75⎤ = ⎢ 12 −3 21⎥ ⎢ ⎥ ⎣ 36 78 63⎦ ⎡ 12 −3⎤ ⎡1 4 (d) ( AB)C = ⎢ −4 5⎥ ⎢ ⎢ ⎥ ⎣3 1 1⎦ ⎣ 4 ⎡ (12 ⋅1) − (3 ⋅ 3) = ⎢ −(4 ⋅1) + (5 ⋅ 3) ⎢ ⎣ (4 ⋅1) + (1 ⋅ 3) ⎡ 3 45 9 ⎤ = ⎢11 −11 17 ⎥ ⎢ ⎥ ⎣ 7 17 13⎦ 2⎤ 5⎥⎦ (12 ⋅ 4) − (3 ⋅1) (12 ⋅ 2) − (3 ⋅ 5) ⎤ −(4 ⋅ 4) + (5 ⋅1) −(4 ⋅ 2) + (5 ⋅ 5) ⎥ ⎥ (4 ⋅ 4) + (1 ⋅1) (4 ⋅ 2) + (1 ⋅ 5) ⎦ 10 . ⎡ 3 0⎤ ⎡ 4 −1⎤ 5.

3 ⎡ 3 0⎤ ⎛ ⎡ 4 −1⎤ ⎡ 1 4 2 ⎤ ⎞ (e) A( BC ) = ⎢ −1 2 ⎥ ⎜ ⎢ ⎢ ⎥ ⎝ ⎣ 0 2 ⎥⎦ ⎢⎣3 1 5⎥⎦ ⎟⎠ ⎣ 1 1⎦ ⎡ 3 0⎤ ⎡ (4 ⋅1) − (1 ⋅ 3) (4 ⋅ 4) − (1 ⋅1) (4 ⋅ 2) − (1 ⋅ 5) ⎤ = ⎢ −1 2 ⎥ ⎢ ⎢ ⎥ ⎣(0 ⋅1) + (2 ⋅ 3) (0 ⋅ 4) + (2 ⋅1) (0 ⋅ 2) + (2 ⋅ 5) ⎥⎦ ⎣ 1 1⎦ ⎡ 3 0⎤ ⎡ 1 15 3⎤ = ⎢ −1 2 ⎥ ⎢ ⎢ ⎥ ⎣6 2 10⎥⎦ ⎣ 1 1⎦ ⎡ (3 ⋅1) + (0 ⋅ 6) (3 ⋅15) + (0 ⋅ 2) (3 ⋅ 3) + (0 ⋅10) ⎤ = ⎢ −(1 ⋅1) + (2 ⋅ 6) −(1⋅15) + (2 ⋅ 2) −(1 ⋅ 3) + (2 ⋅10) ⎥ ⎢ ⎥ (1⋅15) + (1 ⋅ 2) (1 ⋅ 3) + (1 ⋅10) ⎦ ⎣ (1 ⋅1) + (1 ⋅ 6) ⎡ 3 45 9⎤ = ⎢11 −11 17 ⎥ ⎢ ⎥ ⎣ 7 17 13⎦ ⎡ 1 3⎤ ⎡ 1 4 2⎤ ⎢ 4 1⎥ (f) CCT = ⎢ ⎣3 1 5⎥⎦ ⎢ 2 5⎥ ⎣ ⎦ ⎡(1 ⋅1) + (4 ⋅ 4) + (2 ⋅ 2) (1 ⋅ 3) + (4 ⋅1) + (2 ⋅ 5) ⎤ =⎢ ⎣ (3 ⋅1) + (1 ⋅ 4) + (5 ⋅ 2) (3 ⋅ 3) + (1⋅1) + (5 ⋅ 5) ⎥⎦ ⎡ 21 17 ⎤ =⎢ ⎣17 35⎥⎦ T ⎛ ⎡ 1 5 2⎤ ⎡ 3 0⎤ ⎞ (g) ( DA) = ⎜ ⎢ −1 0 1⎥ ⎢ −1 2 ⎥ ⎟ ⎜⎜ ⎢ ⎥⎢ ⎥ ⎟⎟ ⎝ ⎣ 3 2 4⎦ ⎣ 1 1⎦ ⎠ T T ⎛ ⎡ (1⋅ 3) − (5 ⋅1) + (2 ⋅1) (1 ⋅ 0) + (5 ⋅ 2) + (2 ⋅1) ⎤ ⎞ = ⎜ ⎢ −(1 ⋅ 3) − (0 ⋅1) + (1 ⋅1) −(1 ⋅ 0) + (0 ⋅ 2) + (1 ⋅1) ⎥ ⎟ ⎜⎜ ⎢ ⎥ ⎟⎟ ⎝ ⎣ (3 ⋅ 3) − (2 ⋅1) + (4 ⋅1) (3 ⋅ 0) + (2 ⋅ 2) + (4 ⋅1) ⎦ ⎠ T ⎛ ⎡ 0 12⎤ ⎞ = ⎜ ⎢ −2 1⎥ ⎟ ⎜⎜ ⎢ ⎥ ⎟⎟ ⎝ ⎣ 11 8⎦ ⎠ ⎡ 0 −2 11⎤ =⎢ 1 8⎥⎦ ⎣12 11 .SSM: Elementary Linear Algebra Section 1.

Chapter 1: Systems of Linear Equations and Matrices SSM: Elementary Linear Algebra ⎛ ⎡ 1 3⎤ ⎞ ⎡ 4 −1⎤ ⎟ ⎡ 3 −1 1⎤ (h) (C T B) AT = ⎜ ⎢ 4 1⎥ ⎢ ⎜⎜ ⎢ ⎥ ⎣ 0 2⎥⎦ ⎟⎟ ⎢⎣ 0 2 1⎥⎦ ⎝ ⎣ 2 5⎦ ⎠ ⎡ (1 ⋅ 4) + (3 ⋅ 0) −(1 ⋅1) + (3 ⋅ 2) ⎤ ⎡ 3 −1 1⎤ = ⎢ (4 ⋅ 4) + (1 ⋅ 0) −(4 ⋅1) + (1 ⋅ 2) ⎥ ⎢ ⎢ ⎥ ⎣ 0 2 1⎥⎦ ⎣(2 ⋅ 4) + (5 ⋅ 0) −(2 ⋅1) + (5 ⋅ 2) ⎦ 5⎤ ⎡ 4 ⎡ 3 −1 1⎤ = ⎢16 −2⎥ ⎢ ⎢ ⎥ ⎣ 0 2 1⎥⎦ ⎣ 8 8⎦ ⎡ (4 ⋅ 3) + (5 ⋅ 0) −(4 ⋅1) + (5 ⋅ 2) (4 ⋅1) + (5 ⋅1) ⎤ = ⎢(16 ⋅ 3) − (2 ⋅ 0) −(16 ⋅1) − (2 ⋅ 2) (16 ⋅1) − (2 ⋅1) ⎥ ⎢ ⎥ ⎣ (8 ⋅ 3) + (8 ⋅ 0) −(8 ⋅1) + (8 ⋅ 2) (8 ⋅1) + (8 ⋅1) ⎦ 6 9⎤ ⎡ 12 = ⎢ 48 −20 14 ⎥ ⎢ ⎥ 8 16 ⎦ ⎣ 24 (i) ⎛ ⎡ 1 5 2 ⎤ ⎡ 1 − 1 3⎤ ⎞ tr( DDT ) = tr ⎜ ⎢ −1 0 1⎥ ⎢ 5 0 2 ⎥ ⎟ ⎜⎜ ⎢ ⎥⎢ ⎥ ⎟⎟ ⎝ ⎣ 3 2 4⎦ ⎣2 1 4⎦ ⎠ ⎛ ⎡ (1 ⋅1) + (5 ⋅ 5) + (2 ⋅ 2) −(1 ⋅1) + (5 ⋅ 0) + (2 ⋅1) (1⋅ 3) + (5 ⋅ 2) + (2 ⋅ 4) ⎤ ⎞ = tr ⎜ ⎢ −(1 ⋅1) + (0 ⋅ 5) + (1 ⋅ 2) (1⋅1) + (0 ⋅ 0) + (1 ⋅1) −(1 ⋅ 3) + (0 ⋅ 2) + (1⋅ 4) ⎥ ⎟ ⎜⎜ ⎢ ⎥ ⎟⎟ ⎝ ⎣ (3 ⋅1) + (2 ⋅ 5) + (4 ⋅ 2) −(3 ⋅1) + (2 ⋅ 0) + (4 ⋅1) (3 ⋅ 3) + (2 ⋅ 2) + (4 ⋅ 4) ⎦ ⎠ ⎛ ⎡30 1 21⎤ ⎞ 1⎥ ⎟ = tr ⎜ ⎢ 1 2 ⎜⎜ ⎢ ⎥ ⎟⎟ ⎝ ⎣ 21 1 29 ⎦ ⎠ = 30 + 2 + 29 = 61 (j) ⎛ ⎡ 24 −4 16 ⎤ ⎡ 1 5 2 ⎤ ⎞ tr(4 ET − D) = tr ⎜ ⎢ 4 4 4 ⎥ − ⎢ −1 0 1⎥ ⎟ ⎜⎜ ⎢ ⎥ ⎢ ⎥⎟ 8 12 ⎦ ⎣ 3 2 4 ⎦ ⎟⎠ ⎝ ⎣ 12 ⎛ ⎡ 23 −9 14 ⎤ ⎞ = tr ⎜ ⎢ 5 4 3⎥ ⎟ ⎜⎜ ⎢ ⎥ ⎟⎟ ⎝ ⎣ 9 6 8⎦ ⎠ = 23 + 4 + 8 = 35 12 .

SSM: Elementary Linear Algebra Section 1. ⎡ 6 −2 4 ⎤ 1 3⎥ [a1 B] = [3 −2 7] ⎢ 0 ⎢ ⎥ ⎣ 7 7 5⎦ = [18 + 0 + 49 −6 − 2 + 49 12 − 6 + 35] = [67 41 41] 13 . (a) The first row of AB is the first row of A times B.3 ⎛ ⎡ 1 3⎤ ⎡ 3 −1 (k) tr(C T AT + 2 ET ) = tr ⎜ ⎢ 4 1⎥ ⎢ ⎜⎜ ⎢ ⎥ ⎣0 2 ⎝ ⎣ 2 5⎦ ⎛ ⎡ (1⋅ 3) + (3 ⋅ 0) = tr ⎜ ⎢ (4 ⋅ 3) + (1 ⋅ 0) ⎜⎜ ⎢(2 ⋅ 3) + (5 ⋅ 0) ⎝⎣ ⎛⎡ 3 5 = tr ⎜ ⎢12 −2 ⎜⎜ ⎢ 8 ⎝⎣ 6 ⎛ ⎡15 3 = tr ⎜ ⎢14 0 ⎜⎜ ⎢ ⎝ ⎣12 12 = 15 + 0 + 13 = 28 (l) ⎡12 −2 1⎤ ⎢ + 2 2 1⎥⎦ ⎢ 6 4 ⎣ −(1⋅1) + (3 ⋅ 2) −(4 ⋅1) + (1 ⋅ 2) −(2 ⋅1) + (5 ⋅ 2) 8⎤ ⎞ 2⎥ ⎟ ⎥⎟ 6 ⎦ ⎟⎠ (1 ⋅1) + (3 ⋅1) ⎤ ⎡12 −2 8⎤ ⎞ (4 ⋅1) + (1 ⋅1) ⎥ + ⎢ 2 2 2 ⎥ ⎟ ⎥ ⎢ ⎥⎟ (2 ⋅1) + (5 ⋅1) ⎦ ⎣ 6 4 6 ⎦ ⎟⎠ 4 ⎤ ⎡12 −2 8⎤ ⎞ 5⎥ + ⎢ 2 2 2 ⎥ ⎟ ⎥ ⎢ ⎥⎟ 7 ⎦ ⎣ 6 4 6 ⎦ ⎟⎠ 12⎤ ⎞ 7⎥ ⎟ ⎥⎟ 13⎦ ⎟⎠ ⎛ ⎛ ⎡ 6 1 3⎤ ⎡ 1 3⎤ ⎞T ⎡ 3 0 ⎤ ⎞ ⎜ ⎟ tr(( ECT )T A) = tr ⎜ ⎜ ⎢ −1 1 2 ⎥ ⎢ 4 1⎥ ⎟ ⎢ −1 2 ⎥ ⎟ ⎜⎜ ⎢ ⎟ ⎜ ⎝ ⎣ 4 1 3⎥⎦ ⎢⎣ 2 5⎥⎦ ⎟⎠ ⎢⎣ 1 1⎥⎦ ⎟ ⎝ ⎠ ⎛ ⎛ ⎡ (6 ⋅1) + (1 ⋅ 4) + (3 ⋅ 2) (6 ⋅ 3) + (1 ⋅1) + (3 ⋅ 5) ⎤ ⎞T ⎜ = tr ⎜ ⎜ ⎢ −(1⋅1) + (1 ⋅ 4) + (2 ⋅ 2) −(1 ⋅ 3) + (1⋅1) + (2 ⋅ 5) ⎥ ⎟ ⎜ ⎟ ⎜ ⎝⎜ ⎢⎣ (4 ⋅1) + (1⋅ 4) + (3 ⋅ 2) (4 ⋅ 3) + (1 ⋅1) + (3 ⋅ 5) ⎥⎦ ⎟⎠ ⎝ ⎛ ⎛ ⎡16 34 ⎤T ⎞ ⎡ 3 0 ⎤ ⎞ ⎜⎜ ⎟ ⎟ = tr ⎜ ⎜ ⎢ 7 8⎥ ⎟ ⎢ −1 2 ⎥ ⎟ ⎢ ⎥ ⎢ ⎥ ⎜ ⎜ ⎣14 28⎦ ⎟ ⎣ 1 1⎦ ⎟ ⎠ ⎝⎝ ⎠ ⎛ ⎡ 3 0⎤ ⎞ ⎡16 7 14 ⎤ ⎢ = tr ⎜ ⎢ −1 2 ⎥ ⎟ ⎜⎜ ⎣34 8 28⎥⎦ ⎢ ⎥ ⎟⎟ ⎣ 1 1⎦ ⎠ ⎝ ⎡ 3 0 ⎤ ⎞⎟ ⎢ −1 2 ⎥ ⎟ ⎢ ⎥ ⎣ 1 1⎦ ⎟⎠ ⎛ ⎡ (16 ⋅ 3) − (7 ⋅1) + (14 ⋅1) (16 ⋅ 0) + (7 ⋅ 2) + (14 ⋅1)⎤ ⎞ = tr ⎜ ⎢ ⎥⎟ ⎝ ⎣(34 ⋅ 3) − (8 ⋅1) + (28 ⋅1) (34 ⋅ 0) + (8 ⋅ 2) + (28 ⋅1) ⎦ ⎠ ⎛ ⎡ 55 28⎤ ⎞ = tr ⎜ ⎢ ⎥⎟ ⎝ ⎣122 44 ⎦ ⎠ = 55 + 44 = 99 7.

Chapter 1: Systems of Linear Equations and Matrices SSM: Elementary Linear Algebra (f) The third column of AA is A times the third column of A. b = ⎢ −1⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 0⎦ ⎣ 1 5 4⎦ ⎣ x3 ⎦ ⎡ 2 −3 5⎤ ⎡ x1 ⎤ ⎡ 7 ⎤ The equation is ⎢ 9 −1 1⎥ ⎢ x2 ⎥ = ⎢ −1⎥ . (a) A = ⎢ 9 −1 1⎥ . ⎡ 3 −2 7 ⎤ ⎡ −2 ⎤ [ A b 2 ] = ⎢6 5 4 ⎥ ⎢ 1⎥ ⎢ ⎥⎢ ⎥ ⎣0 4 9 ⎦ ⎣ 7 ⎦ ⎡ −6 − 2 + 49 ⎤ = ⎢ −12 + 5 + 28⎥ ⎢ ⎥ ⎣ 0 + 4 + 63 ⎦ ⎡ 41⎤ = ⎢ 21⎥ ⎢ ⎥ ⎣67 ⎦ ⎡ −3 12 76 ⎤ 9. ⎡ 6 −2 4 ⎤ ⎡ 3 ⎤ [ B a1 ] = ⎢ 0 1 3⎥ ⎢ 6 ⎥ ⎢ ⎥⎢ ⎥ 7 7 5⎦ ⎣ 0 ⎦ ⎣ ⎡18 − 12 + 0 ⎤ = ⎢ 0+6+0 ⎥ ⎢ ⎥ ⎣ 21 + 42 + 0⎦ ⎡ 6⎤ = ⎢ 6⎥ ⎢ ⎥ ⎣63⎦ ⎡76 ⎤ ⎡ 3⎤ ⎡ −2 ⎤ ⎡7 ⎤ ⎢ 98⎥ = 7 ⎢6 ⎥ + 4 ⎢ 5⎥ + 9 ⎢ 4 ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣97 ⎦ ⎣0 ⎦ ⎣ 4⎦ ⎣9⎦ ⎡ 64 14 38⎤ (b) BB = ⎢ 21 22 18⎥ ⎢ ⎥ ⎣77 28 74⎦ ⎡ 64 ⎤ ⎡6⎤ ⎡ 4⎤ ⎢ 21⎥ = 6 ⎢ 0 ⎥ + 7 ⎢ 3⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣77 ⎦ ⎣7 ⎦ ⎣ 5⎦ (e) The third row of AA is the third row of A times A. [a3 A] ⎡ 3 −2 7 ⎤ = [ 0 4 9] ⎢ 6 5 4 ⎥ ⎢ ⎥ ⎣0 4 9⎦ = [0 + 24 + 0 0 + 20 + 36 0 + 16 + 81] = [ 24 56 97] ⎡ 14 ⎤ ⎡ 6 ⎤ ⎡ −2 ⎤ ⎡4⎤ ⎢ 22 ⎥ = −2 ⎢ 0 ⎥ + 1 ⎢ 1⎥ + 7 ⎢ 3⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 28⎦ ⎣7 ⎦ ⎣ 7⎦ ⎣ 5⎦ ⎡ 38⎤ ⎡6⎤ ⎡ −2 ⎤ ⎡ 4⎤ ⎢ 18⎥ = 4 ⎢ 0 ⎥ + 3 ⎢ 1⎥ + 5 ⎢ 3⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣74 ⎦ ⎣7 ⎦ ⎣ 7⎦ ⎣ 5⎦ ⎡ x1 ⎤ ⎡ 2 −3 5⎤ ⎡ 7⎤ 11. ⎡ 3 −2 7 ⎤ ⎡7 ⎤ [ A a3 ] = ⎢ 6 5 4 ⎥ ⎢ 4 ⎥ ⎢ ⎥⎢ ⎥ ⎣ 0 4 9 ⎦ ⎣9 ⎦ ⎡ 21 − 8 + 63 ⎤ = ⎢ 42 + 20 + 36 ⎥ ⎢ ⎥ ⎣ 0 + 16 + 81 ⎦ ⎡ 76 ⎤ = ⎢ 98 ⎥ ⎢ ⎥ ⎣97 ⎦ (b) The third row of AB is the third row of A times B. [ a3 B ] ⎡ 6 −2 4 ⎤ 1 3⎥ = [0 4 9] ⎢ 0 ⎢ ⎥ ⎣ 7 7 5⎦ = [0 + 0 + 63 0 + 4 + 63 0 + 12 + 45] = [63 67 57] (c) The second column of AB is A times the second column of B. x = ⎢ x2 ⎥ . (a) AA = ⎢ 48 29 98⎥ ⎢ ⎥ ⎣ 24 56 97 ⎦ ⎡ −3 ⎤ ⎡ 3⎤ ⎡ −2 ⎤ ⎢ 48⎥ = 3 ⎢6 ⎥ + 6 ⎢ 5⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 24 ⎦ ⎣0 ⎦ ⎣ 4⎦ ⎡ 12 ⎤ ⎡ 3⎤ ⎡ −2 ⎤ ⎡7 ⎤ ⎢ 29 ⎥ = −2 ⎢ 6⎥ + 5 ⎢ 5⎥ + 4 ⎢ 4 ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 56 ⎦ ⎣ 0⎦ ⎣ 4⎦ ⎣9⎦ (d) The first column of BA is B times the first column of A. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 1 5 4⎦ ⎣ x3 ⎦ ⎣ 0 ⎦ 14 .

3 21. b = −6. so tr( A + B) = (a11 + b11 ) + (a22 + b22 ) + + (ann + bnn ) = (a11 + a22 + + ann ) + (b11 + b22 + + bnn ) = tr(A) + tr( B) ⎡ x1 ⎤ 1⎤ ⎡ 1⎤ ⎡ 4 0 −3 ⎢x ⎥ ⎢5 ⎢ 3⎥ 1 0 −8 ⎥ (b) A = ⎢ ⎥. d = 1. j. [ k 1 1] ⎢ 1 0 2 ⎥ ⎢ 1⎥ ⎢ ⎥⎢ ⎥ ⎣ 0 2 −3⎦ ⎣ 1⎦ ⎡k ⎤ = [ k + 1 k + 2 −1] ⎢ 1⎥ ⎢ ⎥ ⎣ 1⎦ ⎡ a11 0 ⎢a a ⎢ 21 22 a a (c) ⎢ 31 32 ⎢ a41 a42 ⎢a a ⎢ 51 52 ⎢⎣ a61 a62 0 0 a33 a43 a53 a63 0 0 0 a44 a54 a64 0 0 0 0 a55 a65 0 ⎤ 0 ⎥ ⎥ 0 ⎥ 0 ⎥ 0 ⎥ ⎥ a66 ⎥⎦ ⎡ a11 a12 ⎢a a ⎢ 21 22 0 a32 (d) ⎢ ⎢ 0 0 ⎢ 0 0 ⎢ 0 0 ⎢⎣ 0 a23 a33 a43 0 0 0 0 a34 a44 a54 0 0 0 0 a45 a55 a65 0 ⎤ 0 ⎥ ⎥ 0 ⎥ 0 ⎥ a56 ⎥ ⎥ a66 ⎦⎥ 13. d + 2c = −1 ⇒ 2d = 2. d = 1 d = 1 ⇒ 2c = −2. The ij-entry of kA is kaij .SSM: Elementary Linear Algebra Section 1. ⎛ x ⎞ ⎡1 1⎤ ⎡ x1 ⎤ ⎛ x1 + x2 ⎞ f ⎜ 1⎟=⎢ ⎟ ⎥⎢ ⎥ =⎜ ⎝ x2 ⎠ ⎣ 0 1⎦ ⎣ x2 ⎦ ⎝ x2 ⎠ ⎛1⎞ ⎛1 + 1⎞ ⎛ 2 ⎞ (a) f ⎜ ⎟ = ⎜ ⎟=⎜ ⎟ ⎝1⎠ ⎝ 1 ⎠ ⎝ 1 ⎠ y 1 f(x) x 19. (a) The system is 5 x1 + 6 x2 − 7 x3 = 2 − x1 − 2 x2 + 3 x3 = 0 4 x2 − x3 = 3 = [ k 2 + k + k + 2 − 1] = [ k 2 + 2k + 1] The only solution of k 2 + 2k + 1 = 0 is k = −1. The diagonal entries of A + B are aii + bii . j x then either k = 0 or aij = 0 for all i. 3 ⎤ ⎡ 4 d − 2c ⎤ ⎡a 17. 25. b = ⎢ ⎥ 2 − 5 9 − 1 x ⎢ 3⎥ ⎢ ⎥ ⎢0⎥ ⎣⎢ 0 3 −1 7 ⎦⎥ ⎣⎢ 2 ⎦⎥ ⎣⎢ x4 ⎦⎥ The equation is 1⎤ ⎡ x1 ⎤ ⎡ 1⎤ ⎡ 4 0 −3 ⎢5 1 0 −8⎥ ⎢ x2 ⎥ ⎢ 3⎥ ⎢ ⎥ ⎢ ⎥ = ⎢ ⎥. c = −1. c = −1 The solution is a = 4. ⎢ 2 −5 9 −1⎥ ⎢ x3 ⎥ ⎢ 0 ⎥ 3 −1 7 ⎦⎥ ⎣⎢ x4 ⎦⎥ ⎣⎢ 2 ⎦⎥ ⎣⎢ 0 ⎡ a11 0 ⎢ 0 a 22 ⎢ 0 0 23. (a) ⎢ ⎢ 0 0 ⎢ 0 0 ⎢ 0 ⎣⎢ 0 0 0 a33 0 0 0 0 0 0 a44 0 0 0 0 0 0 a55 0 0 ⎤ 0 ⎥ ⎥ 0 ⎥ 0 ⎥ 0 ⎥ ⎥ a66 ⎦⎥ (b) The system is x1 + x2 + x3 = 2 =2 2 x1 + 3x2 5 x1 − 3x2 − 6 x3 = −9 ⎡ a11 a12 ⎢ 0 a 22 ⎢ 0 0 (b) ⎢ ⎢ 0 0 ⎢ 0 0 ⎢ 0 ⎢⎣ 0 a13 a23 a33 0 0 0 a14 a24 a34 a44 0 0 a15 a25 a35 a45 a55 0 a16 ⎤ a26 ⎥ ⎥ a36 ⎥ a46 ⎥ a56 ⎥⎥ a66 ⎥⎦ ⎡ 1 1 0⎤ ⎡k ⎤ 15. 15 2 . If kaij = 0 for all i. in which 1 case A = 0. ⎢ = gives the −2 ⎥⎦ ⎣ −1 a + b ⎥⎦ ⎢⎣ d + 2c equations a=4 d − 2c = 3 d + 2c = −1 a + b = −2 a = 4 ⇒ b = −6 d − 2c = 3. x = ⎢ 2⎥.

⎡ a11 a12 ⎢a a ⎢ 21 22 a ⎣⎢ 31 a32 ⎡ 1 3⎤ (g) False. a13 ⎤ ⎡ x ⎤ ⎡ a11 x + a12 y + a13 z ⎤ a23 ⎥ ⎢ y ⎥ = ⎢ a21 x + a22 y + a23 z ⎥ ⎥⎢ ⎥ ⎢ ⎥ a33 ⎦⎥ ⎣ z ⎦ ⎣⎢ a31 x + a32 y + a33 z ⎦⎥ ⎡x + y⎤ = ⎢x − y⎥ ⎢ ⎥ ⎣ 0 ⎦ The matrix equation yields the equations a11 x + a12 y + a13 z = x + y a21 x + a22 y + a23 z = x − y . Let A = [ aij ]. then AT is a 4 × 6 matrix and BT is an a22 = −1. ⎢ 0 −3⎥⎦ ⎣ y 1 f(x) = x 1 (c) x 2 ⎛ 4 ⎞ ⎛ 4 + 3⎞ ⎛ 7 ⎞ f ⎜ ⎟=⎜ ⎟=⎜ ⎟ ⎝ 3⎠ ⎝ 3 ⎠ ⎝ 3⎠ 5 SSM: Elementary Linear Algebra True/False 1. if A − C = B − C. 16 . matrix multiplication is not commutative. There is n × m matrix. only square matrices have main diagonals. then aij − cij = bij − cij . so it follows that aij = bij . a31 x + a32 y + a33 z = 0 For these equations to be true for all values of x. for example. the ith row vector of AB is found by multiplying the ith row vector of A by B. an m × n matrix has m row vectors and n column vectors. while B=⎢ . ⎢ ⎥ ⎣0 0 0⎦ (j) True. for example. (a) Both ⎢ and ⎢ are square roots ⎥ 1 1 ⎣ ⎦ ⎣ −1 −1⎥⎦ ⎡2 2⎤ of ⎢ . ⎣3 −2 ⎥⎦ (h) True. y. . and z it must be that a11 = 1. ⎛ 2⎞ ⎛ 2 − 2⎞ ⎛ 0 ⎞ (d) f ⎜ ⎟ = ⎜ ⎟=⎜ ⎟ ⎝ −2 ⎠ ⎝ −2 ⎠ ⎝ −2 ⎠ (d) False. for a square matrix A the main diagonals of A and AT are the same. tr(cA) = ca11 + ca22 + = c(a11 + a22 + = c tr( A) ⎡1 1⎤ ⎡ −1 −1⎤ 29. So BT AT is only defined if m = 4. y (e) True x 1 ⎡ 1 3⎤ (f) False. a21 = 1. and ⎢ 0 3⎥ . ⎡ 1 1 0⎤ one such matrix. (i) True. f(x) x (b) False.Chapter 1: Systems of Linear Equations and Matrices ⎛ 2⎞ ⎛ 2 + 0⎞ ⎛ 2⎞ (b) f ⎜ ⎟ = ⎜ ⎟=⎜ ⎟ ⎝ 0⎠ ⎝ 0 ⎠ ⎝ 0⎠ ⎡5 0 ⎤ (b) The four square roots of ⎢ are ⎣0 9 ⎥⎦ ⎡ 5 0⎤ ⎡− 5 0⎤ ⎡ 5 0 ⎤ . if A = ⎢ and ⎣ 4 −1⎥⎦ ⎡ 1 4⎤ ⎡ 1 0⎤ . 2 –1 –2 f(x) x 27. and aij = 0 for all other i. A = ⎢ 1 −1 0 ⎥ . then ( AB)T = ⎢ ⎥ 0 2 ⎣6 −2 ⎥⎦ ⎣ ⎦ ⎡ 1 8⎤ AT BT = ⎢ . j. x 8 4 (c) False.3 y (a) True. then AB = ⎢ ⎥ 0 2 ⎣ ⎦ ⎣ 4 −2⎥⎦ tr(AB) = −1. ⎢ 0 3⎥⎦ ⎢⎣ 0 −3⎥⎦ ⎣ ⎦ ⎣ ⎡− 5 0 ⎤ . while tr(A) = 0 and tr(B) = 3. if A = ⎢ and ⎣ 4 −1⎥⎦ ⎡ 1 6⎤ ⎡1 0 ⎤ and B=⎢ . if A is a 6 × 4 matrix and B is an m × n matrix. a12 = 1. ⎣ 2 2 ⎥⎦ + cann + ann ) (k) True. and it can only be a 2 × 6 matrix if n = 2.

4 ⎡ 1 2⎤ ⎡ 1 5⎤ (l) False. then AC = BC = ⎢ . for example. if A is an m × n matrix. Then AB will be an m × m matrix and BA will be an n × n matrix. a column of zeros in B will result in a column of zeros in AB. I + 2 A = (( I + 2 A)−1 )−1 1 ⎡ 5 −2 ⎤ = −1⋅ 5 − 2 ⋅ 4 ⎢⎣ −4 −1⎥⎦ 1 ⎡ 5 −2 ⎤ =− ⎢ 13 ⎣ −4 −1⎥⎦ ⎡− 5 2 ⎤ = ⎢ 13 13 ⎥ 4 1⎥ ⎢⎣ 13 13 ⎦ −1 − e− x ) ⎤ ⎥ + e− x ) ⎥⎦ ⎡ 1 (e x + e − x ) − 1 (e x − e − x ) ⎤ 2 ⎥. B=⎢ . = ⎢ 2 x −x 1 (e x + e − x ) ⎥ ⎢ − 1 (e − e ) ⎣ 2 ⎦ 2 1 (e x 2 1 (e x 2 − 15 ⎤ ⎥ = ( B −1 )T 1⎥ 10 ⎦ 17 .SSM: Elementary Linear Algebra Section 1. 7 A = ((7 A)−1 )−1 1 ⎡ −2 −7 ⎤ = (−3)(−2) − 1⋅ 7 ⎢⎣ −1 −3⎥⎦ ⎡ −2 −7 ⎤ = −1 ⎢ ⎣ −1 −3⎥⎦ ⎡2 7⎤ =⎢ ⎣ 1 3⎥⎦ ⎡ 2 1⎤ ⎥. BT = ⎢ . so m = n in order for the sum to be defined. Thus A = ⎢ 7 ⎢⎣ 17 73 ⎥⎦ 1 1 ⎡ 3 0 ⎤ 1 ⎡ 3 0 ⎤ ⎡ 2 0⎤ ⎢ ⎥ = = 2 ⋅ 3 − 0 ⎢⎣0 2 ⎥⎦ 6 ⎢⎣0 2 ⎥⎦ ⎢ 0 1 ⎥ 3⎦ ⎣ 1 9. B −1 = 7. 1 1 ⎡ 4 −4 ⎤ ⎡ 5 ⎢ = 8 + 12 ⎢⎣ 3 2 ⎥⎦ ⎢ 3 ⎣ 20 ⎡ 70 45⎤ 13. ⎣3 4 ⎥⎦ ⎣3 7 ⎥⎦ ⎡ 1 0⎤ ⎡ 1 0⎤ and C = ⎢ . so 2 1 1 ad − bc = (e x + e − x )2 − (e x − e− x )2 4 4 1 2x 1 −2 x = (e + 2 + e ) − (e2 x − 2 + e −2 x ) 4 4 = 1. ABC = ⎢ ⎣122 79 ⎥⎦ 1 ⎡ 79 −45⎤ 70 ⋅ 79 − 122 ⋅ 45 ⎣⎢ −122 70 ⎦⎥ 1 ⎡ 79 −45⎤ = 40 ⎢⎣ −122 70⎥⎦ 1 ⎡ −1 −4 ⎤ 1 ⎡ −1 −4 ⎤ C −1 = = 6(−1) − 4(−2) ⎢⎣ 2 6 ⎥⎦ 2 ⎢⎣ 2 6 ⎥⎦ 1 ⎡ 4 3⎤ 1 ⎡ 4 3 ⎤ B −1 = = 2 ⋅ 4 − (−3)4 ⎢⎣ −4 2 ⎥⎦ 20 ⎢⎣ −4 2 ⎥⎦ ( ABC )−1 = (n) True. ⎡ 1 (e x + e − x ) Thus.4 A−1 = Exercise Set 1. then for both AB and BA to be defined B must be an n × m matrix. a = d = (e x + e − x ) and 2 1 x −x b = c = (e − e ). ⎣ −3 4⎥⎦ ( BT ) − 1 = (m) True. ⎢ = ⎣0 0⎥⎦ ⎢⎣3 7 ⎥⎦ ⎢⎣ 0 0 ⎥⎦ 1 ⎡ 2 −1⎤ ⎡ 2 −1⎤ = 3 ⋅ 2 − 1 ⋅ 5 ⎢⎣ −5 3⎥⎦ ⎢⎣ −5 3⎥⎦ 1 ⎛ ⎡ −1 −4 ⎤ ⎡ 4 3⎤ ⎡ 2 −1⎤ ⎞ C −1 B −1 A−1 = ⎜ ⎟ 40 ⎝ ⎢⎣ 2 6 ⎥⎦ ⎢⎣ −4 2 ⎥⎦ ⎢⎣ −5 3⎥⎦ ⎠ 1 ⎡ 79 −45⎤ = 40 ⎢⎣ −122 70⎥⎦ Section 1. D −1 = 15. ⎥ 0 0 ⎣ ⎦ ⎣3 0 ⎥⎦ ⎡ 2 4⎤ 11. Here. since the jth column vector of AB is A[jth column vector of B]. ⎢ 2 x − x ⎢ 1 (e − e ) ⎣2 17. if A = ⎢ . ⎡ 1 0 ⎤ ⎡ 1 5⎤ ⎡ 1 5 ⎤ (o) False.4 1 ⎡ 4 3⎤ 2 ⋅ 4 − (−3)4 ⎢⎣ −4 2 ⎥⎦ 1 ⎡ 4 3⎤ = 20 ⎢⎣ −4 2 ⎥⎦ 3 ⎤ = ⎡ 15 20 ⎢ ⎥ 1⎥ ⎢⎣ − 15 10 ⎦ 5.

Chapter 1: Systems of Linear Equations and Matrices 1⎤ 2⎤ ⎡− 9 ⎡ − 5 2 ⎤ ⎡ 1 0 ⎤ ⎡ − 18 13 ⎥ . 13 13 ⎥ and A = ⎢ 13 ⎢ Thus 2 A = ⎢ 13 13 ⎥ − ⎢ = 4 1 ⎥ ⎣ 0 1⎦⎥ ⎢ 4 − 12 ⎥ 2 − 6⎥ ⎢ 13 ⎢⎣ 13 13 ⎦ 13 ⎦ 13 ⎦ ⎣ 13 ⎣ ⎛ ⎡3 19. (a) A3 = ⎜ ⎢0 −1 3⎥ ⎢ 0 ⎜⎜ ⎢ ⎥⎢ ⎝ ⎣0 −3 −1⎦ ⎣ 0 ⎡9 0 0⎤ ⎡ 3 = ⎢ 0 −8 −6 ⎥ ⎢ 0 ⎢ ⎥⎢ ⎣ 0 6 −8 ⎦ ⎣ 0 0⎤ ⎡ 27 0 = ⎢ 0 26 −18⎥ ⎢ ⎥ ⎣ 0 18 26⎦ 0 0⎤ ⎞ ⎡ 3 0 0⎤ −1 3⎥ ⎟ ⎢0 −1 3⎥ ⎥⎟⎢ ⎥ −3 −1⎦ ⎟⎠ ⎣0 −3 −1⎦ 0 0⎤ − 1 3⎥ ⎥ −3 −1⎦ 18 SSM: Elementary Linear Algebra . (a) A3 = ⎜ ⎢ ⎝ ⎣2 ⎡11 =⎢ ⎣8 ⎡ 41 =⎢ ⎣30 1⎤ ⎡ 3 1⎥⎦ ⎢⎣ 2 4⎤ ⎡ 3 3⎥⎦ ⎢⎣ 2 15⎤ 11⎥⎦ 1⎤ ⎞ ⎡ 3 1⎤ 1⎥⎦ ⎟⎠ ⎢⎣ 2 1⎥⎦ 1⎤ 1⎥⎦ (b) A−3 = ( A3 ) −1 1 ⎡ 11 −15⎤ = 41⎥⎦ 41⋅11 − 15 ⋅ 30 ⎢⎣ −30 ⎡ 11 −15⎤ =⎢ 41⎥⎦ ⎣ −30 ⎡11 (c) A2 − 2 A + I = ⎢ ⎣8 ⎡6 =⎢ ⎣4 4 ⎤ ⎡ 6 2 ⎤ ⎡ 1 0⎤ − + 3⎥⎦ ⎢⎣ 4 2 ⎥⎦ ⎢⎣0 1⎥⎦ 2⎤ 2 ⎥⎦ ⎡ 3 1⎤ ⎡ 2 0 ⎤ ⎡ 1 1⎤ (d) p( A) = A − 2 I = ⎢ − = ⎣ 2 1⎥⎦ ⎢⎣ 0 2 ⎥⎦ ⎢⎣ 2 −1⎥⎦ (e) p( A) = 2 A2 − A + I ⎡ 22 8⎤ ⎡ 3 1⎤ ⎡ 1 0 ⎤ =⎢ − + ⎣ 16 6 ⎥⎦ ⎢⎣ 2 1⎥⎦ ⎢⎣0 1⎥⎦ ⎡ 20 7 ⎤ =⎢ ⎣ 14 6⎥⎦ (f) p( A) = A3 − 2 A + 4 I ⎡ 41 15⎤ ⎡ 6 2 ⎤ ⎡ 4 0⎤ =⎢ − + ⎣30 11⎥⎦ ⎢⎣ 4 2 ⎥⎦ ⎢⎣ 0 4⎥⎦ ⎡ 39 13⎤ =⎢ ⎣ 26 13⎥⎦ ⎛ ⎡ 3 0 0⎤ ⎡ 3 21.

Thus A−1 = ⎢ 0 − 10 10 ⎢ ⎥ 3 − 1 ⎢⎣ 0 10 10 ⎥⎦ 0⎤ ⎡ 13 0 0 ⎤ ⎞ ⎡ 13 0 0⎤ ⎥⎢ ⎥⎟⎢ ⎥ 1 − 3 ⎥ ⎢0 − 1 − 3 ⎥ ⎟ ⎢0 − 1 − 3 ⎥ − 10 10 10 10 ⎟ 10 10 ⎥⎢ ⎥ ⎢ ⎥ 3 − 1 3 − 1 ⎟ 0 3 − 1 0 10 10 ⎥⎦ ⎢⎣ 10 10 ⎥⎦ ⎠ ⎢⎣ 10 10 ⎥⎦ 0 0 ⎤ ⎡ 13 0 0⎤ ⎥⎢ ⎥ 3 ⎥ ⎢0 − 1 − 3 ⎥ 2 − 25 50 10 10 ⎥⎢ ⎥ 3 3 − 1 2 − 50 − 25 ⎥⎦ ⎢⎣ 0 10 10 ⎥⎦ 0 0 ⎤ ⎥ 0.018 0.026 0. 1⎥ − 10 ⎦ 0 0⎤ ⎡ 2 0 0 ⎤ −1 3⎥ − ⎢ 0 2 0 ⎥ ⎥ ⎢ ⎥ −3 −1⎦ ⎣ 0 0 2 ⎦ 0 0⎤ −3 3⎥ ⎥ −3 −3⎦ (e) p( A) = 2 A2 − A + I 0 0⎤ ⎡ 3 0 0 ⎤ ⎡ 1 0 0⎤ ⎡18 = ⎢ 0 −16 −12⎥ − ⎢0 −1 3⎥ + ⎢ 0 1 0⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 0 12 −16⎦ ⎣0 −3 −1⎦ ⎣ 0 0 1⎦ 0 0⎤ ⎡16 = ⎢ 0 −14 −15⎥ ⎢ ⎥ 15 −14 ⎦ ⎣ 0 (f) p( A) = A3 − 2 A + 4 I 0 ⎤ ⎡6 0 0⎤ ⎡ 4 0 0 ⎤ ⎡ 27 0 ⎢ = 0 26 −18⎥ − ⎢0 −2 6⎥ + ⎢ 0 4 0 ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 0 18 26 ⎦ ⎣0 −6 −2⎦ ⎣ 0 0 4 ⎦ 0⎤ ⎡ 25 0 ⎢ = 0 32 −24 ⎥ ⎢ ⎥ ⎣ 0 24 32 ⎦ 19 .SSM: Elementary Linear Algebra Section 1.018 ⎥ −0.026 ⎥ ⎦ 0 ⎡9 0 0⎤ ⎡6 0 0 ⎤ ⎡ 1 0 0⎤ (c) A2 − 2 A + I = ⎢0 −8 −6⎥ − ⎢0 −2 6 ⎥ + ⎢ 0 1 0⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣0 6 −8⎦ ⎣0 −6 −2 ⎦ ⎣ 0 0 1⎦ 0⎤ ⎡4 0 = ⎢ 0 −5 −12⎥ ⎢ ⎥ ⎣ 0 12 −5⎦ ⎡3 (d) p( A) = A − 2 I = ⎢0 ⎢ ⎣0 ⎡1 = ⎢0 ⎢ ⎣0 3⎤ − 10 ⎥.4 1 1 ⎡ −1 −3⎤ ⎡ − 10 ⎡ −1 3⎤ ⎢ is (b) Note that the inverse of ⎢ = 10 ⎢⎣ 3 −1⎥⎦ ⎢ 3 ⎣ −3 −1⎥⎦ ⎣ 10 A−3 ⎛ ⎡1 ⎜ ⎢3 = ( A− 1 ) 3 = ⎜ ⎢ 0 ⎜⎢ ⎜ ⎢0 ⎝⎣ ⎡1 ⎢9 = ⎢0 ⎢ ⎢⎣ 0 ⎡1 ⎢ 27 =⎢0 ⎢0 ⎣ ⎡1 0 0⎤ ⎢3 ⎥ 1 − 3 ⎥.

Let X = ⎢b21 b22 ⎢ ⎣⎢b31 b32 b13 ⎤ b23 ⎥ . AB will have a row of zeros. then by formula (8) of Section 1. 2 1⎥ 2 ⎥⎦ . 29. b11 + b31 = 1 b11 + b21 = 0 b21 + b31 = 0 1 1 1 Thus b11 = b31 = −b21 . (a) If A has a row of zeros. and B cannot be invertible. ⎥ b33 ⎦⎥ AX = I ⎡ b11 + b31 b12 + b32 b13 + b33 ⎤ ⎡1 0 0⎤ ⎢b + b b +b b + b ⎥ = ⎢ 0 1 0⎥ ⎢ 11 21 12 22 13 23 ⎥ ⎢ ⎥ ⎣⎢b21 + b31 b22 + b32 b23 + b33 ⎦⎥ ⎣ 0 0 1 ⎦ Consider the equations from the first columns of the matrices. ⎥ 1 ⎥ ann ⎥⎦ AB = BA = I so A is invertible and B = A−1 . ⎡1 0 ⎢ a11 1 ⎢0 a22 Let B = ⎢ ⎢ ⎢ 0 ⎢⎣ 0 SSM: Elementary Linear Algebra none of the diagonal entries are zero. ( AB)−1 ( AC −1 )( D −1C −1 )−1 D −1 = ( B −1 A−1 )( AC −1 )(CD) D −1 = B −1 ( A−1 A)(C −1C )( DD −1 ) = B −1 ⎡ b11 b12 35. so b11 = . (b) If B has a column of zeros. b21 = − . 0 ⎤ ⎥ 0 ⎥ ⎥. Since a11a22 ann ≠ 0. 2 2 2 ⎡ 1 ⎢ 2 Working similarly with the other columns of both matrices yields A−1 = ⎢ − 12 ⎢ 1 ⎢⎣ 2 20 1 2 1 2 − 12 − 12 ⎤ ⎥ 1⎥. AB will have a column of zeros.Chapter 1: Systems of Linear Equations and Matrices 27. CT B −1 A2 BAC −1DA−2 BT C −2 = C T 31.3.3. CA−1 B −1 A−2 B(CT )−1 ⋅ CT B −1 A2 BAC −1DA−2 BT C −2 = CA−1B −1 A−2 B(C T ) −1 C T DA−2 BT C −2 = CA−1B −1 A−2 B DA−2 BT C −2 ⋅ C 2 ( BT )−1 A2 = CA−1B −1 A−2 B ⋅ C 2 ( BT )−1 A2 D = CA−1B −1 A−2 BC 2 ( BT )−1 A2 33. so A cannot be invertible. then by formula (9) of Section 1. and b31 = .

Let X = ⎢b21 b22 ⎢ ⎣⎢b31 b32 Section 1. x2 = . 23 23 1⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ ⎡6 . The matrix form of the system is ⎢ ⎢ ⎥= 4 3 − ⎣ ⎦⎥ ⎣ x2 ⎦ ⎣⎢ −2 ⎦⎥ 1⎤ ⎡6 ⎢⎣ 4 −3⎥⎦ −1 1 ⎡ −3 −1⎤ 6(−3) − 1 ⋅ 4 ⎢⎣ −4 6 ⎥⎦ 1 ⎡ −3 −1⎤ =− ⎢ 22 ⎣ −4 6 ⎥⎦ = 1 ⎡ −3 −1⎤ ⎡ 0⎤ ⎡ x1 ⎤ ⎢ x ⎥ = − 22 ⎢ −4 6⎥ ⎢ −2⎥ ⎣ ⎦⎣ ⎦ ⎣ 2⎦ 1 ⎡ 0+2 ⎤ =− ⎢ 22 ⎣0 − 12 ⎥⎦ 1 ⎡ 2⎤ =− ⎢ 22 ⎣ −12⎥⎦ 1 6 The solution is x1 = − . b21 = −b11 and −2b11 + 1 = 0. so b11 = . so the solution is 39.SSM: Elementary Linear Algebra ⎡ b11 b12 37. 1 ⎡ 5 2⎤ 3 ⋅ 5 − (−2)4 ⎢⎣ −4 3⎥⎦ 1 ⎡ 5 2⎤ = 23 ⎢⎣ −4 2 ⎥⎦ = ⎡ x1 ⎤ 1 ⎡ 5 2 ⎤ ⎡ −1⎤ 1 ⎡ −5 + 6⎤ 1 ⎡ 1 ⎤ ⎢ x ⎥ = 23 ⎢ −4 3⎥ ⎢ 3⎥ = 23 ⎢ 4 + 9⎥ = 23 ⎢13⎥ ⎣ ⎦⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ 2⎦ 1 13 The solution is x1 = . x2 = .4 b13 ⎤ b23 ⎥ . The matrix form of the system is ⎢ = 5⎥⎦ ⎢⎣ x2 ⎥⎦ ⎢⎣ 3⎥⎦ ⎣4 ⎡ 3 −2 ⎤ ⎢⎣ 4 5⎥⎦ −1 ⎡ x1 ⎤ ⎡ 3 −2 ⎤ ⎢ x ⎥ = ⎢4 5⎥⎦ ⎣ 2⎦ ⎣ −1 ⎡ −1⎤ ⎢⎣ 3⎥⎦ . 11 11 21 1⎤ ⎡ x1 ⎤ ⎡ 6 ⎢ x ⎥ = ⎢ 4 −3⎥ ⎦ ⎣ 2⎦ ⎣ . b21 = − . −1 ⎡ 0⎤ ⎢⎣ −2 ⎥⎦ . and b31 = 1. b31 = 1 b11 + b21 = 0 −b11 + b21 + b31 = 0 1 1 Thus. ⎥ b33 ⎦⎥ AX = I b b b ⎡ ⎤ ⎡1 0 0 ⎤ 31 32 33 ⎢ b +b ⎥ = ⎢0 1 0 ⎥ b12 + b22 b13 + b23 11 21 ⎢ ⎥ ⎢ ⎥ ⎢⎣ −b11 + b21 + b31 −b12 + b22 + b32 −b13 + b23 + b33 ⎥⎦ ⎣0 0 1 ⎦ Consider the equations from the first columns of the matrices. so the solution is 41. 2 2 Working similarly with the other columns of both matrices yields ⎡ 1 1 − 1⎤ 2⎥ ⎢ 2 2 −1 1 1 1 A = ⎢− 2 2 ⎥. 2 ⎢ ⎥ 1 0 0 ⎢⎣ ⎥⎦ ⎡ 3 −2 ⎤ ⎡ x1 ⎤ ⎡ −1⎤ .

by Theorem 1. ( I − A)( I + A + A2 + + Ak −1 ) = ( I − A) I + ( I − A) A + ( I − A) A2 + = I − A + A − A2 + A2 − A3 + =I−A = I −0 =I + ( I − A) Ak −1 + Ak −1 − Ak k True/False 1. (kA + B)T = (kA)T + BT = kAT + BT . if A is an m × n matrix and B is and n × p matrix. (b) False. 22 . Since AB ≠ BA. (i) False. not a scalar. (e) False. (d) False. the two terms cannot be combined.Chapter 1: Systems of Linear Equations and Matrices SSM: Elementary Linear Algebra 51.4. (f) True. with m ≠ p. ( A + B)2 = ( A + B)( A + B) = A2 + AB + BA + B 2 Since AB ≠ BA the two terms cannot be combined. by Theorem 1. A and B are inverses if and only if AB = BA = I. then A−1 exists. (g) True. (a) If A is invertible.4 (a) False. then AB and ( AB)T are defined but AT BT is not defined. AB is invertible. 53. but ( AB)−1 = B −1 A−1 ≠ A−1 B −1 . AB = AC A−1 AB = A−1 AC IB = IC B=C (b) The matrix A in Example 3 is not invertible. ( A − B)( A + B) = A2 + AB − BA − B 2 . (c) False.9. (a) A( A−1 + B −1 ) B( A + B)−1 = ( AA−1 + AB −1 ) B( A + B)−1 = ( I + AB −1 ) B( A + B)−1 = ( IB + AB −1 B)( A + B) −1 = ( B + A)( A + B)−1 = ( A + B)( A + B)−1 =I (b) A−1 + B −1 ≠ ( A + B)−1 55.4.5. (h) True. p( I ) = (a0 + a1 + a2 + + am ) I which is a matrix.

⎡ 1 0 0⎤ E = ⎢ 0 1 0⎥ ⎢ ⎥ ⎣ −2 0 1⎦ (c) The operation is to add 5 times the first row ⎡ 1 0 0⎤ to the third row. (a) The matrix results from adding −5 times the first row of I 2 to the second row. ⎡ 1 0 4 ⎤ ⎡ 1 4 ⎤ ⎡13 28⎤ EA = ⎢0 1 0 ⎥ ⎢ 2 5⎥ = ⎢ 2 5⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 0 1⎦ ⎣ 3 6 ⎦ ⎣ 3 6⎦ (b) The matrix is not elementary. ⎡0 0 1⎤ E = ⎢0 1 0 ⎥ ⎢ ⎥ ⎣ 1 0 0⎦ 3. (a) The operation is to add 3 times the second ⎡ 1 3⎤ row to the first row. ⎡0 0 1⎤ E = ⎢0 1 0 ⎥ ⎢ ⎥ ⎣ 1 0 0⎦ (c) The matrix is not elementary. ⎢1 0 0 0 ⎥ ⎢⎣0 0 0 1 ⎥⎦ ⎡ 1 4 1 0⎤ 9. it is an elementary matrix.5 Exercise Set 1. the first and third rows must be interchanged. ⎣0 1⎥⎦ (b) The operation is to multiply the first row by ⎡− 1 0 0⎤ ⎢ 7 ⎥ 1 − . 23 . but their sum ⎢ is not ⎣0 0 ⎥⎦ invertible. The matrix is ⎢ 0 1 0 ⎥ . for example. 7 ⎢ 0 0 1⎥ ⎣ ⎦ (c) To obtain C from A. The matrix is ⎢ ⎥. more than one elementary row operation on I3 is required. if the square matrix A has a row of zeros. so AB = I is impossible. 5.5 1. (a) To obtain B from A. The matrix is ⎢ . more than one elementary row operation on I 2 is required. A similar statement can be made if A has a column of zeros. (c) E adds 4 times the third row to the first row.5 (j) True. add 2 times the first row to the third row.SSM: Elementary Linear Algebra Section 1. ⎢ ⎥ ⎣ 5 0 1⎦ (d) To obtain A from C. (b) To obtain A from B. (a) E interchanges the first and second rows of A. more than one elementary row operation on I 4 is required. ⎡ 1 0 0⎤ E = ⎢0 1 0⎥ ⎢ ⎥ ⎣ 2 0 1⎦ (d) The operation is to interchange the first and ⎡0 0 1 0⎤ ⎢0 1 0 0⎥ third rows. (b) E adds −3 times the second row to the third row. the product of A and any matrix B will also have a row or column of zeros (depending on whether the product is AB or BA). the first and third rows must be interchanged. add −2 times the first row to the third row. 5 −1⎤ ⎡ 0 1 ⎤ ⎡ −1 −2 EA = ⎢ ⎥ ⎢ − − −6 ⎥⎦ 1 0 3 6 6 ⎣ ⎦⎣ ⎡ 3 −6 −6 −6 ⎤ =⎢ 5 −1⎥⎦ ⎣ −1 −2 ⎡ 1 0⎤ ⎡ −1 0 ⎤ (k) False. (d) The matrix is not elementary. 7. ⎢ ⎣ 2 7 0 1⎥⎦ Add −2 times the first row to the second. The matrix is ⎢0 1 0 ⎥ . ⎡ 1 0 0⎤ ⎡ 2 −1 0 −4 −4 ⎤ EA = ⎢0 1 0⎥ ⎢ 1 −3 −1 5 3⎥ ⎢ ⎥⎢ ⎥ ⎣0 −3 1⎦ ⎣ 2 0 1 3 −1⎦ ⎡ 2 −1 0 −4 −4 ⎤ 5 3⎥ = ⎢ 1 −3 −1 ⎢ ⎥ ⎣ −1 9 4 −12 −10 ⎦ Section 1. ⎢ and ⎢ are ⎥ 0 1 ⎣ ⎦ ⎣ 0 −1⎥⎦ ⎡0 0 ⎤ both invertible.

⎢ 2 4 ⎢ ⎥ ⎣ −4 2 −9 0 0 1⎦ Multiply the first row by −1. ⎡ 1 0 −7 4 ⎤ ⎢⎣0 1 2 −1⎥⎦ ⎡1 4 ⎤ ⎢⎣ 2 7 ⎥⎦ −1 SSM: Elementary Linear Algebra 3 0 1 0⎤ ⎡1 0 ⎢0 5 −10 0 −2 1⎥ ⎢ ⎥ ⎣0 4 −10 1 −3 0 ⎦ Multiply the second row by 1 . ⎡ 1 −3 4 −1 0 0 ⎤ ⎢ 2 4 1 0 1 0⎥ ⎢ ⎥ 4 2 9 0 0 1⎦ − − ⎣ ⎡ 3 4 −1 1 0 0 ⎤ 3 0 1 0⎥ 13. 10 . ⎡ 1 −3 −1 0 ⎤ ⎢⎣0 7 3 1⎥⎦ 1 Multiply the second row by . ⎢ 1 0 ⎢ ⎥ ⎣ 2 5 −4 0 0 1⎦ Interchange the first and second rows. 5 3 0 1 0⎤ ⎡1 0 ⎢ 0 1 −2 0 − 2 1 ⎥ 5 5⎥ ⎢ ⎢⎣0 4 −10 1 −3 0⎥⎦ Add −4 times the second row to the third. ⎡ 1 4 1 0⎤ ⎢⎣0 1 2 −1⎥⎦ Add −4 times the second row to the first. ⎡ 1 −3 4 −1 0 0 ⎤ ⎢0 10 −7 2 1 0⎥ ⎢ ⎥ ⎣0 −10 7 −4 0 1⎦ Add −3 times the first row to the second and −2 times the first row to the third. 7 ⎡ 1 −3 −1 0 ⎤ ⎢0 1 73 17 ⎥⎥ ⎣⎢ ⎦ Add 3 times the second row to the first. 3 0 1 0⎤ ⎡1 0 ⎢ 3 4 −1 1 0 0 ⎥ ⎢ ⎥ ⎣ 2 5 −4 0 0 1⎦ Add −2 times the first row to the second and 4 times the first row to the third. 2 ⎡1 0 3 0 1 0⎤ ⎢ ⎥ 2 0 − 5 15 ⎥ ⎢ 0 1 −2 ⎢ ⎥ 7 2 1 − 12 10 ⎢⎣0 0 5 ⎥⎦ Add −3 times the third row to the first and 2 times the third row to the second.Chapter 1: Systems of Linear Equations and Matrices 1 0⎤ ⎡1 4 ⎢⎣0 −1 −2 1⎥⎦ Multiply the second row by −1. 3 0 1 0⎤ ⎡1 0 ⎢0 4 −10 1 −3 0 ⎥ ⎢ ⎥ ⎣0 5 −10 0 −2 1⎦ Multiply the second row by Interchange the second and third rows. ⎡1 0 2 3⎤ 7 7⎥ ⎢ ⎢0 1 73 17 ⎥ ⎣ ⎦ −1 ⎡ 2 3 ⎤ ⎡ −1 3 ⎤ 7 7 ⎢⎣ 3 −2 ⎥⎦ = ⎢⎢ 3 1 ⎥⎥ ⎣7 7⎦ ⎡ 3 4 −1⎤ ⎢1 0 3⎥ ⎢ ⎥ 2 5 4 − ⎣ ⎦ −1 ⎡ 3 − 11 − 6 ⎤ 10 5⎥ ⎢ 2 1 1⎥ = ⎢ −1 7 2⎥ ⎢− 1 10 5⎦ ⎣ 2 ⎡ −1 3 −4 1 0 0 ⎤ 1 0 1 0⎥ 15. ⎡ 1 −3 −1 0 ⎤ ⎢⎣3 −2 0 1⎥⎦ Add −3 times the first row to the second. ⎡1 0 3 0 1 0⎤ ⎢ ⎥ 2 1⎥ ⎢ 0 1 −2 0 − 5 5 ⎢ ⎥ 7 4 ⎢⎣0 0 −2 1 − 5 − 5 ⎥⎦ 1 Multiply the third row by − . ⎢ ⎣ 3 −2 0 1⎥⎦ Multiply the first row by −1. 3 − 11 − 6 ⎤ ⎡1 0 0 2 10 5⎥ ⎢ 1 1⎥ ⎢0 1 0 −1 7 2⎥ ⎢0 0 1 − 1 2 10 5⎦ ⎣ ⎡ −7 4 ⎤ =⎢ ⎣ 2 −1⎥⎦ ⎡ −1 3 1 0 ⎤ 11. 24 1 .

⎡ 1 2 12 0 0 1 0 0 ⎤ ⎢ 2 −4 0 0 1 0 0 0 ⎥ ⎢ ⎥ ⎢0 0 2 0 0 0 1 0⎥ ⎣⎢ 0 −1 −4 −5 0 0 0 1⎦⎥ Add −1 times the third row to both the first and second rows. ⎡ −1 3 −4 ⎤ ⎢ 2 4 1⎥ is not invertible. ⎡ 1 0 1 1 0 0⎤ ⎢0 1 1 0 1 0 ⎥ ⎢ ⎥ ⎣0 1 −1 −1 0 1⎦ Add −1 times the second row to the third.SSM: Elementary Linear Algebra Section 1. ⎡ 1 3 3 1 0 0⎤ 2 ⎢ ⎥ ⎢ 0 1 0 −1 1 0 ⎥ ⎢0 1 1 −1 0 1⎥ ⎣ ⎦ Add −1 times the second row to the third.5 ⎡ 1 3 3 1 0 0⎤ 2 ⎢ ⎥ 2 7 6 0 1 0⎥ ⎢ ⎢ 2 7 7 0 0 1⎥ ⎣ ⎦ Add −2 times the first row to both the second and third rows. ⎡1 3 0 1 3 −3⎤ 2 ⎢ ⎥ ⎢0 1 0 −1 1 0 ⎥ ⎢0 0 1 0 −1 1⎥ ⎣ ⎦ Add −3 times the second row to the first. 1 0 0⎤ ⎡ 1 2 12 0 0 ⎢0 −1 −4 −5 0 0 0 1⎥ ⎢ ⎥ 2 0 0 0 1 0⎥ ⎢0 0 ⎢⎣0 −8 −24 0 1 −2 0 0 ⎥⎦ ⎡2 6 6 1 0 0⎤ 19. 2 25 . 1 1 0 0⎤ ⎡1 0 ⎢0 1 1 0 1 0⎥ ⎢ ⎥ ⎣0 0 −2 −1 −1 1⎦ 1 Multiply the third row by − . ⎢0 1 1 0 1 0 ⎥ ⎢ ⎥ ⎣ 1 1 0 0 0 1⎦ Add −1 times the first row to the third. 2 ⎡1 0 1 1 0 ⎤ 0 ⎢ ⎥ 0 1 1 0 1 0 ⎢ ⎥ ⎢0 0 1 1 1 − 1 ⎥ 2 2 2⎦ ⎣ ⎡ 2 −4 0 0 1 0 0 0 ⎤ ⎢ 1 2 12 0 0 1 0 0 ⎥ 21. ⎢ ⎥ ⎢0 0 2 0 0 0 1 0⎥ ⎢⎣ 0 −1 −4 −5 0 0 0 1⎥⎦ Interchange the first and second rows. 1 0 0⎤ ⎡ 1 2 12 0 0 ⎢0 −8 −24 0 1 −2 0 0 ⎥ ⎢ ⎥ 2 0 0 0 1 0⎥ ⎢0 0 ⎣⎢0 −1 −4 −5 0 0 0 1⎦⎥ Interchange the second and fourth rows. 1 −1 1⎤ ⎡1 0 0 2 2 2⎥ ⎢ 1 1⎥ ⎢0 1 0 − 12 2 2 ⎢ 1 1 − 1⎥ ⎢⎣0 0 1 2 2 2 ⎥⎦ 1⎤ −1 ⎡ 1 − 1 2 2⎥ ⎡ 1 0 1⎤ ⎢ 2 1 1⎥ ⎢0 1 1⎥ = ⎢ − 1 2 2 2 ⎢ ⎥ ⎢ 1 1 − 1⎥ ⎣ 1 1 0⎦ ⎢⎣ 2 2 2 ⎥⎦ Add −2 times the first row to the second. 4 −1 0 0 ⎤ ⎡ 1 −3 ⎢0 7 1 1 0⎥ 1 − 10 ⎢ ⎥ 5 10 0 −2 1 1⎦⎥ ⎣⎢0 0 Since there is a row of zeros on the left side. ⎢ 2 7 6 0 1 0 ⎥ ⎢ ⎥ ⎣ 2 7 7 0 0 1⎦ Multiply the first row by 1 . ⎡1 3 3 1 0 0⎤ 2 ⎢ ⎥ ⎢0 1 0 −1 1 0 ⎥ ⎢0 0 1 0 −1 1⎥ ⎣ ⎦ Add −3 times the third row to the first. ⎡ 1 0 0 7 0 −3⎤ 2 ⎢ ⎥ ⎢0 1 0 −1 1 0 ⎥ ⎢0 0 1 0 −1 1⎥ ⎣ ⎦ −1 ⎡ 7 0 −3⎤ ⎡2 6 6⎤ 2 ⎢ ⎢ 2 7 6 ⎥ = ⎢ −1 1 0 ⎥⎥ ⎢ ⎥ ⎢ 0 −1 1⎥ ⎣2 7 7 ⎦ ⎣ ⎦ 4 −1 0 0 ⎤ ⎡ 1 −3 ⎢0 7 1 1 0⎥ 1 − 10 ⎢ ⎥ 5 10 ⎢⎣0 −10 7 −4 0 1⎥⎦ Add 10 times the second row to the third. ⎢ ⎥ ⎣ −4 2 −9 ⎦ ⎡ 1 0 1 1 0 0⎤ 17.

0 1 −6 0⎤ ⎡1 2 0 0 ⎢0 1 0 0 − 1 3 1 0 ⎥⎥ −2 ⎢ 8 4 1 ⎢0 0 1 0 0 0 0⎥ 2 ⎢ ⎥ ⎢0 0 0 1 1 − 1 − 1 − 1 ⎥ 40 20 10 5⎦ ⎣ Multiply the third row by −1 −3 − 32 1 2 1 − 10 0⎤ ⎥ 0⎥ ⎥ 0⎥ ⎥ − 15 ⎥ ⎦ 1 ⎡ 1 0⎤ −3 2 ⎢ 4 ⎥ 3 1 − 0 ⎢ − 18 ⎥ 4 2 =⎢ ⎥ 1 0 0 0⎥ ⎢ 2 ⎢ 1 1 1 1⎥ ⎢⎣ 40 − 20 − 10 − 5 ⎥⎦ 1 0 1 0 0 0⎤ ⎡ −1 0 ⎢ 2 3 −2 6 0 1 0 0 ⎥ 23. 1 0 0⎤ ⎡ 1 2 12 0 0 ⎢0 1 4 5 0 0 0 −1⎥ ⎢ ⎥ 1 0⎥ ⎢0 0 1 0 0 0 2 ⎢0 0 0 40 1 −2 −4 −8⎥ ⎣ ⎦ 1 . ⎢ ⎥ ⎢ 0 −1 2 0 0 0 1 0 ⎥ 1 5 0 0 0 1⎥⎦ ⎢⎣ 0 0 Multiply the first row by −1. 0 1 0 0⎤ ⎡ 1 2 12 0 ⎢0 1 4 0 − 1 1 1 0⎥ 8 4 2 ⎢ ⎥ 1 ⎢0 0 1 0 ⎥ 0 0 0 2 ⎢ ⎥ ⎢0 0 0 1 1 − 1 − 1 − 1 ⎥ 40 20 10 5⎦ ⎣ Add −12 times the third row to the first and −4 times the third row to the second. 1 0 0⎤ ⎡ 1 2 12 0 0 ⎢0 1 4 5 0 0 0 −1⎥ ⎢ ⎥ ⎢0 0 2 0 0 0 1 0 ⎥ ⎣⎢0 0 8 40 1 −2 0 −8⎦⎥ SSM: Elementary Linear Algebra ⎡1 ⎢ ⎢0 ⎢ ⎢0 ⎢ 0 ⎣⎢ 0 0 0 1 0 0 0 1 0 0 0 1 1 4 − 18 1 2 1 4 0 0 1 40 1 − 20 ⎡ 2 −4 0 0 ⎤ ⎢ 1 2 12 0 ⎥ ⎢ ⎥ ⎢0 0 2 0⎥ ⎣⎢ 0 −1 −4 −5⎦⎥ 1 . 1 0 0⎤ ⎡ 1 2 12 0 0 ⎢0 1 4 5 0 0 0 −1⎥ ⎢ ⎥ 2 0 0 0 1 0⎥ ⎢0 0 ⎢⎣0 −8 −24 0 1 −2 0 0 ⎥⎦ Add 8 times the second row to the fourth. ⎡ 1 0 −1 0 −1 0 0 0 ⎤ ⎢ 0 1 −2 0 0 0 −1 0 ⎥ ⎢ ⎥ 1 5 0 0 0 1⎥ ⎢0 0 ⎢⎣0 0 6 6 2 1 3 0 ⎥⎦ Add −6 times the third row to the fourth. ⎡ 1 0 −1 0 −1 0 0 0 ⎤ ⎢ 2 3 −2 6 0 1 0 0 ⎥ ⎢ ⎥ ⎢ 0 −1 2 0 0 0 1 0 ⎥ 1 5 0 0 0 1⎥⎦ ⎢⎣ 0 0 Add −2 times the first row to the second. ⎡ 1 0 −1 0 −1 0 0 0⎤ ⎢0 −1 2 0 0 0 1 0⎥ ⎢ ⎥ ⎢0 3 0 6 2 1 0 0⎥ ⎢⎣0 0 1 5 0 0 0 1⎥⎦ Multiply the second row by −1.Chapter 1: Systems of Linear Equations and Matrices Multiply the second row by −1. ⎡ 1 0 −1 0 −1 0 0 0⎤ ⎢0 3 0 6 2 1 0 0⎥ ⎢ ⎥ ⎢0 −1 2 0 0 0 1 0⎥ ⎣⎢0 0 1 5 0 0 0 1⎦⎥ Interchange the second and third rows. Multiply the fourth row by 40 1 0 0⎤ ⎡ 1 2 12 0 0 ⎢0 1 4 5 0 0 0 −1⎥ ⎢ ⎥ 1 0 0⎥ ⎢0 0 1 0 0 2 ⎢ 1 1 1 1⎥ ⎢⎣0 0 0 1 40 − 20 − 10 − 5 ⎥⎦ Add −5 times the fourth row to the second. 2 1 0 0⎤ ⎡ 1 2 12 0 0 ⎢0 1 4 5 0 0 0 −1⎥ ⎢ ⎥ 1 0⎥ ⎢0 0 1 0 0 0 2 ⎢0 0 8 40 1 −2 0 −8⎥ ⎣ ⎦ Add −8 times the third row to the fourth. ⎡ 1 0 −1 0 −1 0 0 0 ⎤ ⎢ 0 1 −2 0 0 0 −1 0 ⎥ ⎢ ⎥ ⎢0 3 0 6 2 1 0 0 ⎥ 1 5 0 0 0 1⎥⎦ ⎢⎣0 0 Add −3 times the second row to the third. Add −2 times the second row to the first. ⎡ 1 0 −1 0 −1 0 0 0 ⎤ ⎢ 0 1 −2 0 0 0 −1 0 ⎥ ⎢ ⎥ ⎢0 0 6 6 2 1 3 0 ⎥ 1 5 0 0 0 1⎦⎥ ⎣⎢0 0 Interchange the third and fourth rows. 26 .

k4 row by 1 k −1 0 ⎡1 − 1 k ⎢k 0 1 ⎢ = ⎢0 0 ⎢ 0 ⎣⎢ 0 0⎤ ⎥ 0 0⎥ 1 − 1⎥ k k⎥ 0 1⎦⎥ 0 ⎡c c c 1 0 0 ⎤ 27. 5 5 − 1⎤ ⎡1 0 0 0 − 7 12 24 8 4⎥ ⎢ 5 5 1 − 1⎥ ⎢0 1 0 0 6 12 4 2 ⎢ ⎥ 5 5 5 1 − 4⎥ ⎢0 0 1 0 12 24 8 ⎢ 1 − 1 1⎥ 0 0 0 1 − 12 − 81 24 4 ⎦⎥ ⎣⎢ 1 0⎤ ⎡ −1 0 ⎢ 2 3 −2 6 ⎥ ⎢ ⎥ ⎢ 0 −1 2 0 ⎥ 1 5⎥⎦ ⎢⎣ 0 0 −1 ⎡− 7 ⎢ 12 ⎢ 56 =⎢ 5 ⎢ 12 ⎢ 1 − ⎣⎢ 12 5 24 5 12 5 24 1 − 24 ⎡ k1 0 0 0 1 0 0 ⎢0 k 0 0 0 1 0 2 25.5 ⎡1 ⎢ ⎢0 ⎢ ⎢0 ⎢ ⎢0 ⎢⎣ 0 −1 0 −1 1 −2 0 0 0 1 5 0 0 0 −24 2 0 0 0⎤ 0 −1 0 ⎥ ⎥ 0 0 1⎥ 1 3 −6 ⎥⎦ 1 Multiply the fourth row by − . (a) ⎢ k 0 0 0 0 0 1 3 ⎢ ⎢⎣ 0 0 0 k4 0 0 0 1 Multiply the first row by . 24 −1 0 0 0⎤ ⎡ 1 0 −1 0 ⎢ 0 1 −2 0 0 0 −1 0 ⎥ ⎢ ⎥ 1 5 0 0 0 1⎥ ⎢0 0 ⎢0 0 0 1 − 1 − 1 − 1 1 ⎥ 12 24 8 4⎦ ⎣ Add −5 times the fourth row to the third.SSM: Elementary Linear Algebra ⎡1 ⎢0 ⎢ ⎢0 ⎢⎣0 Section 1. and the k2 k3 1 fourth row by . −1 0 0 0⎤ ⎡ 1 0 −1 0 ⎢ 0 1 −2 0 0 0 −1 0⎥ ⎢ 5 5 5 − 1⎥ 1 0 ⎢0 0 12 24 8 4⎥ ⎢ ⎥ 1 1 1 1 0 0 0 1 − 12 − 24 − 8 4 ⎦⎥ ⎣⎢ Add the third row to the first and 2 times the third row to the second. k1 5 8 1 4 5 8 − 81 0 0 0 1 k1 0 0 1 0 0 0 1 k2 0 0 0 0 1 k3 0 0 0 1 0 0 0 1 ⎡ k1 0 0 0⎤ ⎢0 k 0 0⎥ 2 ⎢ ⎥ 0⎥ ⎢ 0 0 k3 ⎢⎣ 0 0 0 k4 ⎥⎦ −1 ⎡1 ⎢ k1 ⎢0 =⎢ ⎢0 ⎢ ⎢0 ⎢⎣ ⎡k 1 0 0 1 0 0 ⎢0 1 0 0 0 1 0 (b) ⎢ ⎢0 0 k 1 0 0 1 ⎢⎣ 0 0 0 1 0 0 0 0⎤ ⎥ 0⎥ ⎥ 0⎥ ⎥ 1⎥ k4 ⎥⎦ 0 0 1 k2 0 0 1 k3 0 0 0⎤ 0⎥ ⎥ 0⎥ 1⎥⎦ Multiply the first and third rows by ⎡1 ⎢ ⎢0 ⎢0 ⎢ ⎢⎣0 0⎤ ⎥ 0⎥ ⎥ 0⎥ ⎥ 1⎥ k4 ⎥⎦ 1 . ⎢ 1 c c 0 1 0 ⎥ ⎢ ⎥ ⎣ 1 1 c 0 0 1⎦ Multiply the first row by requirement). 27 1 (so c ≠ 0 is a c . the third row by . k 0 0⎤ ⎥ 1 0 0 0 1 0 0⎥ 0 1 1k 0 0 k1 0 ⎥ ⎥ 0 0 1 0 0 0 1⎥⎦ 1 Add − times the fourth row to the third k 1 and − times the second row to the first. k ⎡1 0 0 0 1 − 1 0 0⎤ k k ⎢ ⎥ 1 0 0⎥ ⎢0 1 0 0 0 ⎢0 0 1 0 0 0 k1 − k1 ⎥ ⎢ ⎥ 0 0 1⎦⎥ ⎣⎢0 0 0 1 0 − 14 ⎤ ⎥ − 12 ⎥ ⎥ − 14 ⎥ 1⎥ 4 ⎦⎥ 0⎤ 0⎥ ⎥ 0⎥ 1⎥⎦ 1 k 0 0 ⎡k 1 0 0⎤ ⎢ 0 1 0 0⎥ ⎢ ⎥ ⎢ 0 0 k 1⎥ ⎣⎢ 0 0 0 1⎥⎦ the second 1 1 .

Use elementary matrices to reduce ⎢ to ⎣ 2 2 ⎥⎦ I2 . ⎡ 1 0⎤ ⎡ −3 1⎤ ⎡ −3 ⎢0 1 ⎥ ⎢ 2 2 ⎥ = ⎢ 1 ⎦ ⎣ ⎣ 2⎦ ⎣ ⎡ 1 −1⎤ ⎡ −3 1⎤ ⎡ −4 ⎢⎣0 1⎥⎦ ⎢⎣ 1 1⎥⎦ = ⎢⎣ 1 ⎡ − 1 0 ⎤ ⎡ −4 0 ⎤ ⎡1 ⎢ 4 ⎥⎢ ⎥=⎢ ⎣ 0 1⎦ ⎣ 1 1⎦ ⎣1 SSM: Elementary Linear Algebra 1⎤ 1⎥⎦ 0⎤ 1⎥⎦ 0⎤ 1⎥⎦ 0 0⎤ ⎡ 1 0 0 ⎤ 4 0⎥ = ⎢ 0 1 0 ⎥ ⎥ ⎢ ⎥ 0 1⎦ ⎣ 0 0 1⎦ 0 0 ⎤ ⎡ 1 0 2 ⎤ ⎡ 1 0 −2 ⎤ 1 −3 ⎥ ⎢ 0 1 0 ⎥ ⎢ 0 4 3⎥ ⎥⎢ ⎥⎢ ⎥ 0 1⎦ ⎣0 0 1⎦ ⎣ 0 0 1⎦ −1 −1 ⎡ 1 0 2 ⎤ ⎡ 1 0 0 ⎤ ⎡ 1 0 0⎤ = ⎢ 0 1 0 ⎥ ⎢0 1 −3⎥ ⎢ 0 14 0⎥ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ 1⎦ ⎢⎣ 0 0 1⎥⎦ ⎣ 0 0 1⎦ ⎣0 0 ⎡ 1 0 −2 ⎤ ⎡ 1 0 0 ⎤ ⎡ 1 0 0⎤ = ⎢ 0 1 0 ⎥ ⎢ 0 1 3⎥ ⎢ 0 4 0 ⎥ ⎢ ⎥⎢ ⎥⎢ ⎥ 1⎦ ⎣ 0 0 1⎦ ⎣0 0 1⎦ ⎣0 0 Note that other answers are possible. ⎡ 1 0 2 ⎤ ⎡ 1 0 −2 ⎤ ⎡ 1 0 0 ⎤ ⎢0 1 0 ⎥ ⎢ 0 4 3 ⎥ = ⎢ 0 4 3⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ 0 0 1 0 0 1⎦ ⎣0 0 1⎦ ⎣ ⎦⎣ ⎡ 1 0 0⎤ ⎡ 1 0 0⎤ ⎡ 1 0 0⎤ ⎢ 0 1 − 3 ⎥ ⎢ 0 4 3⎥ = ⎢ 0 4 0 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ 1⎦ ⎣0 0 1⎦ ⎣0 0 1⎦ ⎣0 0 ⎡ 1 0 0⎤ ⎡ 1 ⎢0 1 0 ⎥ ⎢0 4 ⎢ ⎥⎢ ⎢⎣0 0 1⎥⎦ ⎣0 Thus ⎡ 1 0 0⎤ ⎡ 1 ⎢0 1 0 ⎥ ⎢0 4 ⎢ ⎥⎢ ⎢⎣0 0 1⎥⎦ ⎣0 ⎡ 1 0 0⎤ = ⎢ 0 1 0⎥ ⎢ ⎥ ⎣ 0 0 1⎦ so ⎡ 1 0 −2 ⎤ ⎢0 4 3⎥ ⎢ ⎥ 0 0 1⎦ ⎣ ⎡ −3 1⎤ 29. ⎡ 1 0 −2 ⎤ 31. Use elementary matrices to reduce ⎢0 4 3⎥ ⎢ ⎥ 1⎦ ⎣0 0 to I3 . ⎡ 1 0 ⎤ ⎡1 0 ⎤ ⎡ 1 0 ⎤ ⎢⎣ −1 1⎥⎦ ⎢⎣1 1⎥⎦ = ⎢⎣0 1⎥⎦ Thus. ⎡ −3 1⎤ ⎢⎣ 2 2 ⎥⎦ −1 −1 ⎡ 1 0 ⎤ ⎡ 1 −1⎤ −1 ⎡ − 1 0⎤ ⎡ 1 0 ⎤ −1 4 =⎢ ⎥ ⎢ 1⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 0 2 ⎦ ⎣ 0 1⎦ ⎣ 0 1⎦ ⎣ −1 1⎦ ⎡ 1 0⎤ ⎡ 1 1⎤ ⎡ −4 0 ⎤ ⎡1 0 ⎤ =⎢ ⎣ 0 2⎥⎦ ⎢⎣ 0 1⎥⎦ ⎢⎣ 0 1⎥⎦ ⎢⎣1 1⎥⎦ Note that other answers are possible. ⎡1 1 1 1 0 0 ⎤ c ⎢ ⎥ 1 ⎢0 1 1 − 1 ⎥ 0 c (c −1) c −1 ⎢ ⎥ 1 ⎥ ⎢0 0 1 − 1 0 c (c −1) c −1 ⎦ ⎣ Add −1 times the third row to the first and second rows. ⎡ 1 0 ⎤ ⎡ − 14 0 ⎤ ⎡ 1 −1⎤ ⎡ 1 0⎤ ⎡ −3 1⎤ ⎢⎣ −1 1⎥⎦ ⎢ 0 1⎥ ⎢⎣0 1⎥⎦ ⎢0 1 ⎥ ⎢⎣ 2 2⎥⎦ ⎣ ⎦ ⎣ 2⎦ ⎡ 1 0⎤ =⎢ ⎣ 0 1⎥⎦ so 28 −1 .Chapter 1: Systems of Linear Equations and Matrices ⎡1 1 1 1 0 0 ⎤ c ⎢ ⎥ c c 1 0 1 0⎥ ⎢ ⎢1 1 c 0 0 0 ⎥ ⎣ ⎦ Add −1 times and the first row to the second and third rows. 1 ⎡1 1 0 0 − c1−1 ⎤ c −1 ⎢ ⎥ 1 1 ⎥ ⎢0 1 0 0 − c −1 c −1 ⎥ ⎢ 1 ⎥ ⎢0 0 1 − c(c1−1) 0 c −1 ⎦ ⎣ The matrix is invertible for c ≠ 0. 1 0 0⎤ ⎡1 1 1 c ⎢ ⎥ ⎢0 c − 1 c − 1 − 1c 1 0⎥ ⎢ ⎥ 0 c − 1 − 1c 0 1⎥ ⎢⎣0 ⎦ 1 Multiply the second and third rows by (so c −1 c ≠ 1 is a requirement). 1.

x2 = −1. ⎡ −3 1⎤ ⎢⎣ 2 2 ⎥⎦ −1 (c) True.6 Exercise Set 1. hence it does not affect whether a matrix is invertible. Find the inverse of the coefficient matrix. the product of two elementary matrices is not necessarily elementary. Ax = 0 cannot have exactly one solution. 5⎤ ⎡1 0 ⎢0 2 −2 ⎥ ⎢ ⎥ ⎣ 1 −1 6 ⎦ 1.5. E2′ . hence it does not affect whether a matrix is invertible.6 33.) ⎡ 1 2 3⎤ 37. From Exercise 31.2. The matrix form of the system is ⎡ 1 3 1⎤ ⎡ x1 ⎤ ⎡ 4 ⎤ ⎢ 2 2 1⎥ ⎢ x ⎥ = ⎢ −1⎥ ⎢ ⎥⎢ 2⎥ ⎢ ⎥ ⎣ 2 3 1⎦ ⎣ x3 ⎦ ⎣ 3⎦ (b) True. 3. True/False 1. For instance.5 (a) False. since the sequence of row operations that convert an invertible matrix A into I n is not unique. a homogeneous system has either exactly one solution or infinitely many solutions. … . 5⎤ ⎡1 0 ⎢0 2 −2 ⎥ = B ⎢ ⎥ ⎣ 1 1 4⎦ 1 ⎡ 6 −1⎤ ⎡ 6 −1⎤ ⎡ 1 1⎤ ⎢⎣5 6 ⎥⎦ = 6 − 5 ⎢⎣ −5 1⎥⎦ = ⎢⎣ −5 1⎥⎦ ⎡ 6 −1⎤ ⎡ 2 ⎤ ⎡ 12 − 9 ⎤ ⎡ 3⎤ ⎢⎣ −5 1⎥⎦ ⎢⎣ 9 ⎥⎦ = ⎢⎣ −10 + 9 ⎥⎦ = ⎢⎣ −1⎥⎦ The solution is x1 = 3. ⎡ −3 1⎤ ⎡ 1 0 ⎤ ⎡ 1 1⎤ ⎡ −4 0 ⎤ ⎡1 0⎤ ⎢⎣ 2 2 ⎥⎦ = ⎢⎣0 2 ⎥⎦ ⎢⎣0 1⎥⎦ ⎢⎣ 0 1⎥⎦ ⎢⎣1 1⎥⎦ ⎡ 1 −2 ⎤ ⎡ 1 0 ⎤ ⎡ 1 0 ⎤ ⎡ 1 5⎤ =⎢ 1⎥⎦ ⎢⎣ 2 1⎥⎦ ⎢⎣0 −8⎥⎦ ⎢⎣0 1⎥⎦ ⎣0 (This is the matrix from Exercise 29. A = ⎢ 1 4 1⎥ ⎢ ⎥ ⎣2 1 9⎦ Add −1 times the first row to the second row. Ek such 0⎤ ⎡ − 14 ⎢ 0 ⎤ ⎡ 1 −1⎤ ⎡ 1 0 ⎤ ⎡ 1 =⎢ ⎥ ⎢ 1⎥ ⎥ ⎣ −1 1⎦ ⎣ 0 1⎦ ⎢⎣0 1⎥⎦ ⎣0 2 ⎦ ⎡− 1 1 ⎤ = ⎢ 4 83 ⎥ ⎢⎣ 14 8 ⎥⎦ that B = Ek E2 E1 A. 3⎤ ⎡1 2 ⎢ 0 2 −2 ⎥ ⎢ ⎥ ⎣2 1 9⎦ Add −1 times the first row to the third row. there exist elementary matrices E1 . since A and B are row equivalent. −1 −2 ⎤ 3⎥ ⎥ 1⎦ 0 0⎤ ⎡ 1 0 0 ⎤ ⎡ 1 0 2 ⎤ 1 0 ⎥ ⎢ 0 1 −3⎥ ⎢ 0 1 0 ⎥ 4 ⎥⎢ ⎥⎢ ⎥ 1⎦ ⎣ 0 0 1⎦ 0 1⎥⎦ ⎣ 0 0 0 2⎤ 1 − 3⎥ 4 4⎥ 0 1⎥⎦ (e) True. (f) True. From Exercise 29. ⎣ 2⎦ −1 Add the second row to the third row.6 Add −1 times the second row to the first row. 35. The matrix form of the system is ⎡ 1 1⎤ ⎡ x1 ⎤ ⎡ 2 ⎤ ⎢⎣5 6 ⎥⎦ ⎢ x ⎥ = ⎢⎣ 9 ⎥⎦ . ⎡1 0 ⎢0 4 ⎢ ⎣0 0 ⎡1 = ⎢0 ⎢ ⎢⎣ 0 ⎡1 = ⎢0 ⎢ ⎢⎣ 0 E2 E1 A so A (d) True. Similarly there exist elementary matrices E1′. E2 . Since A is not invertible. El′ such that C = El′ E2′ E1′B = El′ E2′ E1′Ek and C are row equivalent. (g) False. the expression of A as a product of elementary matrices is not unique. 29 . by Theorem 1. adding a multiple of the first row to the second row is an elementary row operation. interchanging two rows is an elementary row operation. …. 3⎤ ⎡1 2 ⎢0 2 −2 ⎥ ⎢ ⎥ ⎣ 1 −1 6 ⎦ Section 1.SSM: Elementary Linear Algebra Section 1.

Chapter 1: Systems of Linear Equations and Matrices SSM: Elementary Linear Algebra ⎡ 1 3 1 1 0 0⎤ ⎢2 2 1 0 1 0⎥ ⎢ ⎥ ⎣ 2 3 1 0 0 1⎦ ⎡ 1 1 1 1 0 0⎤ ⎢0 5 5 4 0 1⎥ ⎢ ⎥ ⎣ 0 0 −5 −1 1 0 ⎦ ⎡ 1 3 1 1 0 0⎤ ⎢0 −4 −1 −2 1 0⎥ ⎢ ⎥ ⎣0 −3 −1 −2 0 1⎦ ⎡ 1 1 1 1 0 0⎤ ⎢0 1 1 4 0 1 ⎥ 5 5⎥ ⎢ ⎣⎢0 0 −5 −1 1 0 ⎥⎦ 0 0⎤ ⎡1 3 1 1 ⎢0 1 1 1 4 − 14 0 ⎥ 2 ⎢ ⎥ ⎢⎣0 −3 −1 −2 0 1⎥⎦ ⎡1 1 1 1 0 ⎢ 4 0 ⎢0 1 1 5 ⎢ 1 0 0 1 5 − 15 ⎣⎢ 0⎤ ⎥ 1⎥ 5 ⎥ 0⎥ ⎦ ⎡1 1 0 ⎢ ⎢0 1 0 ⎢ 0 0 1 ⎣⎢ 0⎤ ⎥ 1⎥ 5 ⎥ 0⎥ ⎦ ⎡1 3 1 1 0 0⎤ ⎢ ⎥ 1 1 1 − 0 ⎢0 1 ⎥ 4 2 4 ⎢ ⎥ 3 1 1 ⎢⎣0 0 − 4 − 2 − 4 1⎥⎦ 0 0⎤ ⎡1 3 1 1 ⎢0 1 1 1 − 1 0⎥ 4 2 4 ⎢ ⎥ ⎢⎣0 0 1 2 3 −4 ⎥⎦ 4 5 3 5 1 5 ⎡1 0 0 1 5 ⎢ ⎢0 1 0 53 ⎢ 1 ⎢⎣0 0 1 5 ⎡ 1 3 0 −1 −3 4 ⎤ ⎢0 1 0 0 −1 1⎥ ⎢ ⎥ ⎣ 0 0 1 2 3 −4 ⎦ 1 5 1 5 − 15 0 − 15 ⎤ ⎥ 1 1⎥ 5 5 ⎥ 0⎥ − 15 ⎦ ⎡1 0 − 15 ⎤ 5 ⎢5 ⎥ ⎡ ⎤ ⎡ 1 ⎤ ⎡ 1⎤ 1 1 ⎥ ⎢10 ⎥ = ⎢3 + 2 ⎥ = ⎢ 5 ⎥ ⎢ 53 5 5 ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢1 ⎥ 0 1 ⎣ ⎦ ⎣1 − 2 ⎦ ⎣ −1⎦ 0 − ⎢⎣ 5 ⎥ 5 ⎦ The solution is x = 1. −1 1 ⎡ 2 −5⎤ ⎡ 2 −5⎤ = 6 − 5 ⎢⎣ −1 3⎥⎦ ⎢⎣ −1 3⎥⎦ ⎡ 2 −5⎤ ⎡ b1 ⎤ ⎡ 2b1 − 5b2 ⎤ ⎢⎣ −1 3⎥⎦ ⎢b ⎥ = ⎢ −b + 3b ⎥ ⎣ 2⎦ ⎣ 1 2⎦ The solution is x1 = 2b1 − 5b2 . ⎡ 1 1 1 1 0 0⎤ ⎢ 1 1 −4 0 1 0 ⎥ ⎢ ⎥ ⎣ −4 1 1 0 0 1⎦ = 9. The matrix form of the system is ⎡3 5⎤ ⎡ x1 ⎤ ⎡ b1 ⎤ ⎢⎣ 1 2 ⎥⎦ ⎢ x ⎥ = ⎢b ⎥ . ⎣ 2⎦ ⎣ 2⎦ The solution is x1 = −1. x2 = 4. y = 5. z = −1. x2 = −b1 + 3b2 . x3 = −7. 1⎤ ⎡ 1 0 0 −1 0 ⎢0 1 0 0 −1 1⎥ ⎢ ⎥ ⎣ 0 0 1 2 3 −4 ⎦ 1⎤ ⎡ 4 ⎤ ⎡ −4 + 3 ⎤ ⎡ −1⎤ ⎡ −1 0 ⎢ 0 −1 1⎥ ⎢ −1⎥ = ⎢ 1 + 3 ⎥ = ⎢ 4 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 2 3 −4⎦ ⎣ 3⎦ ⎣8 − 3 − 12⎦ ⎣ −7 ⎦ 7. ⎡ 1 −5 1 −2 ⎤ ⎢⎣3 2 4 5⎥⎦ ⎡ 1 1 1 1 0 0⎤ ⎢ 0 0 −5 −1 1 0 ⎥ ⎢ ⎥ ⎣0 5 5 4 0 1⎦ ⎡ 1 −5 1 −2 ⎤ ⎢⎣0 17 1 11⎥⎦ 30 . ⎡3 5 ⎤ ⎢⎣ 1 2 ⎥⎦ 5. Write and reduce the augmented matrix for both systems. The matrix form of the equation is ⎡ 1 1 1⎤ ⎡ x ⎤ ⎡ 5 ⎤ ⎢ 1 1 −4 ⎥ ⎢ y ⎥ = ⎢10 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ −4 1 1⎦ ⎣ z ⎦ ⎣ 0 ⎦ Find the inverse of the coefficient matrix.

1 3 (iv) The solution is x1 = − . 17 17 (ii) The solution is x1 = ⎡ 1 0 1 b1 − 3 b2 ⎤ 7 7 ⎥ ⎢ ⎢⎣0 1 72 b1 + 17 b2 ⎥⎦ There are no restrictions on b1 and b2 . 15 15 (ii) The solution is x1 = 34 28 . x2 = . x2 = . 15 15 (iii) The solution is x1 = 19 13 . x2 = . x2 = . 5 5 31 . ⎢ 4 −5 8 b2 ⎥ ⎢ ⎥ ⎣ −3 3 −3 b3 ⎦ 1⎤ ⎡1 2 1 6 3 ⎢⎣ 4 −7 0 −4 −1 −5⎥⎦ 5 b1 ⎤ ⎡ 1 −2 ⎢0 3 −12 −4b1 + b2 ⎥ ⎢ ⎥ 3b1 + b3 ⎦ ⎣0 −3 12 2 1 6 3 1⎤ ⎡1 ⎢⎣0 −15 −4 −28 −13 −9 ⎥⎦ 5 b1 ⎡ 1 −2 ⎤ ⎢0 3 −12 −4b1 + b2 ⎥ ⎢ ⎥ 0 −b1 + b2 + b3 ⎦ ⎣0 0 ⎡ 1 2 1 6 3 1⎤ ⎢0 1 4 28 13 3 ⎥ ⎢⎣ 15 15 15 5 ⎥⎦ ⎡1 0 7 15 ⎢ 4 ⎢0 1 15 ⎣ 34 15 28 15 19 15 13 15 5 b1 ⎡ 1 −2 ⎤ ⎢ ⎥ 4 1 1 −4 − 3 b1 + 3 b2 ⎥ ⎢0 ⎢⎣0 0 0 −b1 + b2 + b3 ⎥⎦ − 15 ⎤ ⎥ 3⎥ 5⎦ (i) The solution is x1 = 7 4 . x2 = .SSM: Elementary Linear Algebra Section 1. ⎡ 4 −7 0 −4 −1 −5⎤ ⎢⎣ 1 2 1 6 3 1⎥⎦ ⎡ 1 −2 5 b1 ⎤ 15. Write and reduce the augmented matrix for the four systems.6 ⎡ 1 −5 1 −2 ⎤ ⎢0 1 11 ⎥ 1 17 17 ⎦ ⎣ ⎡ 1 0 22 17 ⎢ 1 ⎢⎣0 1 17 ⎡ 1 3 b1 ⎤ 13. 21 11 . ⎢ ⎥ ⎣ −2 1 b2 ⎦ b1 ⎤ ⎡1 3 ⎢0 7 2b + b ⎥ ⎣ 1 2⎦ 21 ⎤ 17 ⎥ 11 ⎥ 17 ⎦ b1 ⎡1 3 ⎤ ⎢ ⎥ 2 1 ⎢⎣0 1 7 b1 + 7 b2 ⎥⎦ 22 1 (i) The solution is x1 = . 15 15 ⎡ 1 0 −3 − 5 b1 + 2 b2 ⎤ 3 3 ⎢ ⎥ 4 1b ⎥ 0 1 4 b − − + ⎢ 1 2 3 3 ⎢ ⎥ ⎢⎣0 0 0 −b1 + b2 + b3 ⎥⎦ The only restriction is from the third row: −b1 + b2 + b3 = 0 or b3 = b1 − b2 . 17 17 11. x2 = .

⎡ 1 −1 1 1 0 0 ⎤ ⎢ 2 3 0 0 1 0⎥ ⎢ ⎥ ⎣ 0 2 −1 0 0 1⎦ True/False 1. so every matrix of the form x1 + x0 is a solution of Ax = b. b1 − b2 + b3 = 0 and −1 1 5 3 −1 3 ⎤ ⎡1 0 0 ⎢0 1 0 −2 1 −2⎥ ⎢ ⎥ ⎣0 0 1 −4 2 −5⎦ ⎡ 3 −1 3⎤ ⎡ 2 −1 X = ⎢ −2 1 −2 ⎥ ⎢ 4 0 ⎢ ⎥⎢ ⎣ −4 2 −5 ⎦ ⎣ 3 5 ⎡ 11 12 −3 27 = ⎢ −6 −8 1 −18 ⎢ ⎣ −15 −21 9 −38 b1 3 2 ⎡ 1 −1 ⎤ ⎢0 1 −11 −5 −2b1 − b2 ⎥ ⎢ ⎥ b1 − b2 + b3 ⎥ 0 0 ⎢0 0 0 0 −2b1 + b2 + b4 ⎦⎥ ⎣⎢0 0 ⎡ 1 −1 1⎤ 19. Let x1 be a fixed solution of Ax = b. −2b1 + b2 + b4 = 0. Ak is also invertible and Ak x = 0 has only the trivial solution. Thus b3 = −b1 + b2 and b4 = 2b1 − b2 . Thus. X = ⎢ 2 3 0 ⎥ ⎢ ⎥ ⎣ 0 2 −1⎦ 0⎤ 0 ⎥⎥ 2 −5⎥⎦ 0 5 −2 5⎤ ⎡ 1 −1 0 ⎢ 0 1 0 −2 1 −2 ⎥ ⎢ ⎥ ⎣ 0 0 1 −4 2 −5 ⎦ b1 ⎡ 1 −1 3 2 ⎤ ⎢0 −1 11 5 2b1 + b2 ⎥ ⎢ ⎥ b1 − b2 + b3 ⎥ ⎢0 0 0 0 ⎣⎢0 0 0 0 −2b1 + b2 + b4 ⎦⎥ ⎡1 ⎢0 ⎢ ⎢0 ⎢⎣0 SSM: Elementary Linear Algebra (a) True. and let x be any other solution. Expressing b1 and b2 in terms of b3 and b4 gives b1 = b3 + b4 and b2 = 2b3 + b4 . Since Ax = 0 has only the trivial solution. if a system of linear equations has more than one solution. Also A(x1 + x0 ) = Ax1 + Ax0 = b + 0 = b. then A is invertible and Ax = c has the unique solution A−1c. Note that ( Ak )−1 = ( A−1 ) k . so x can be expressed as x = x1 + x0 .Chapter 1: Systems of Linear Equations and Matrices ⎡ 1 −1 ⎢ −2 1 17. 0 0⎤ 1 0⎥ ⎥ 5 0 1⎥⎦ (b) True.6 ⎡ 1 −1 1 1 0 0⎤ ⎢0 5 −2 −2 1 0⎥ ⎢ ⎥ ⎣0 2 −1 0 0 1⎦ 1 1 ⎡ 1 −1 ⎢0 1 − 2 − 2 ⎢ 5 5 ⎢⎣0 2 −1 0 5 7 8⎤ −3 0 1⎥ ⎥ −7 2 1⎦ 26 ⎤ −17 ⎥ ⎥ −35⎦ 21. Then A(x − x1 ) = Ax − Ax1 = b − b = 0. Thus ⎡ 2 −1 5 7 8⎤ ⎢ 4 0 −3 0 1⎥ ⎢ ⎥ ⎣ 3 5 −7 2 1⎦ x0 = x − x1 is a solution to Ax = 0. 32 . 23. ⎢ − 3 2 ⎢ ⎣⎢ 4 −3 b1 ⎤ b2 ⎥ ⎥ b3 ⎥ b4 ⎦⎥ 3 2 5 1 2 −1 1 3 ⎡ 1 −1 1 1 0 0⎤ ⎢ ⎥ 1 0⎥ ⎢0 1 − 52 − 52 5 ⎢ ⎥ 1 4 −2 1 0 0 − ⎥⎦ 5 5 5 ⎣⎢ 1 1 ⎡ 1 −1 ⎢0 1 − 2 − 2 ⎢ 5 5 ⎢⎣0 0 1 −4 b1 ⎤ 3 2 ⎡ 1 −1 ⎢0 −1 11 5 2b + b ⎥ 1 2⎥ ⎢ 3b1 + b3 ⎥ ⎢0 −1 11 5 ⎣⎢0 1 −11 −5 −4b1 + b4 ⎦⎥ −b1 − b2 ⎤ 0 −8 −3 1 −11 −5 −2b1 − b2 ⎥ ⎥ 0 0 0 b1 − b2 + b3 ⎥ 0 0 0 −2b1 + b2 + b4 ⎥⎦ invertible. A is From the bottom two rows. if Ax = b has a unique solution. it has infinitely many solutions.

23. 4. ⎡12 9. The inverse is ⎢ 2 ⎢⎣ 0 − 15 ⎥⎦ 3. ⎡1 0⎤ ⎥. The matrix is not symmetric since a12 = −8 ≠ 0 = a21 . 33 . A2 = ⎢ ⎢ ⎢ ⎢⎣ ⎤ ⎡1 0 ⎤ =⎢ 1⎥ (−2) ⎥⎦ ⎣0 4 ⎦ 0 ⎤ 0 ⎤ ⎡1 = ⎢ 1 ⎥ ⎥ (−2)− k ⎥⎦ ⎢⎣0 ( −2)k ⎥⎦ 0 −2 ⎥ ⎤ 0 ⎥ ⎡1 0 ⎥ ⎢4 2 1 0 0 ⎥ = ⎢ 0 91 3 ⎥ ⎢ 2 1 ⎥ ⎢⎣ 0 0 0 0 4 ⎥ ⎦ ⎡ 1 −2 ⎤ 0 0 ⎥ ⎢ 2 ⎢ ⎥ ⎡4 −2 1 A−2 = ⎢ 0 0 ⎥ = ⎢0 3 ⎢ ⎥ ⎢0 −2 1 ⎢ 0 ⎥ ⎣ 0 4 ⎢⎣ ⎦⎥ − k ⎡ 1 ⎤ 0 0 ⎥ ⎢ 2 ⎢ ⎥ −k 1 0 A− k = ⎢ 0 ⎥ 3 ⎢ ⎥ −k 1 ⎢ 0 ⎥ 0 4 ⎣⎢ ⎦⎥ ⎡ 2k 0 ⎤ 0 ⎢ ⎥ k =⎢0 3 0⎥ ⎢0 0 4k ⎥⎥ ⎣⎢ ⎦ (f) True. the system Ax = 4x is equivalent to the system (A − 4I)x = 0. () Section 1. then the system (A − 4I)x = 0 has a unique solution and so does the equivalent system Ax = 4x. For A to be symmetric. ⎡ 3 0 0 ⎤ ⎡ 2 1⎤ ⎡ 6 3⎤ 5. then S ( S −1 AS )x = Sb or A(Sx) = Sb. 15. ⎡ −1 0 0 ⎤ The inverse is ⎢ 0 12 0 ⎥ . since a23 = −6 ≠ 6 = a32 . a12 = a21 or −3 = a + 5. and row equivalent matrices can be obtained from one another by elementary row operations. ( 12 ) 2 () (g) True.7 1. Thus x ≠ 1. hence A − 4I is invertible. then so will (A − 4I)x = 0. The matrix is a diagonal matrix with nonzero entries on the diagonal. ⎡1− k A− k = ⎢ ⎢⎣ 0 (e) True. ⎢0 2 0⎥ ⎢ 1 −5 3 ⎢ ⎥⎢ ⎣0 0 −3⎦ ⎣ −6 2 2 ⎡ −15 10 0 20 = ⎢ 2 −10 6 0 ⎢ ⎣ 18 −6 −6 −6 0 ⎤ ⎡1 0 ⎤ ⎥=⎢ ⎥ (−2) 2 ⎥⎦ ⎣0 4 ⎦ 19. since ( S −1 AS )x = b. ⎢ ⎥ ⎢⎣ 0 0 3⎥⎦ 0 () () () () () 0⎤ ⎥ 0⎥ ⎥ 1 16 ⎥⎦ 0 0⎤ 9 0⎥ ⎥ 0 16 ⎦ () 13. The matrix is a diagonal matrix with nonzero entries on the diagonal. A2 = ⎢ ⎢⎣ 0 ⎡1−2 A−2 = ⎢ ⎢⎣ 0 (d) True. so it is invertible. The matrix is not invertible because it is upper triangular and has a zero on the main diagonal. The matrix is symmetric. 25. ⎡ ⎢ ⎢ 11. the entries on the main diagonal must be nonzero. For A to be invertible. so it is invertible.7 Exercise Set 1. 17. then both A and B would have to be invertible. if AB = I n . since a13 = 1 ≠ 3 = a31 . 4 −4 ⎤ 0 3⎥ ⎥ 2 2⎦ −20 ⎤ 6⎥ ⎥ −6 ⎦ 21.SSM: Elementary Linear Algebra Section 1.7 (c) True. If A − 4I is invertible. then B = A−1 and BA = I n also. If the system Ax = 4x has a unique solution. −2. The matrix is not symmetric. so a = −8. elementary row operations do not change the solution set of a linear system. Sx is a solution of Ay = Sb. ⎢0 −1 0 ⎥ ⎢ −4 1⎥ = ⎢ 4 −1⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 0 2 ⎦ ⎣ 2 5⎦ ⎣ 4 10 ⎦ ⎡ 5 0 0⎤ ⎡ −3 2 0 7. if AB were invertible. The matrix is not symmetric.

since a diagonal matrix must be square and have zeros off the main diagonal. z 3 = −8 or x = 1. 2 33.7 (c) From the hint. + = ⎣0 1⎥⎦ ⎢⎣ 2 1⎥⎦ ⎢⎣ 2 2 ⎥⎦ (d) True. then (d) Consider a12 and a21.Chapter 1: Systems of Linear Equations and Matrices ⎡ 1 0 0⎤ 27. (c) A is symmetric since aij = a ji for all i. 39. ⎡ 0 0 −8 ⎤ A = ⎢ 0 0 −4 ⎥ ⎢ ⎥ ⎣ 8 4 0⎦ (b) i 2 − j 2 = j 2 − i 2 ⇒ 2i 2 = 2 j 2 or i 2 = j 2 ⇒ i = j. A = ⎢0 −1 0 ⎥ . 3 a12 = 2(1) + 2(2) = 18 while a21 = 2(2)2 + 2(1)3 = 10. its transpose is also diagonal. A is not symmetric unless n = 1. so A is symmetric. z = −2. in an upper triangular matrix. 2 (a) True. the entries on the main diagonal must be zero and the reflections of entries across the main diagonal must be opposites. (e) True. Then ⎣ 0 z ⎥⎦ ⎡ x3 y( x 2 + xz + z 2 ) ⎤ 3 A3 = ⎢ ⎥ . ⎡ 1 2 ⎤ ⎡ 3 0⎤ ⎡ 4 2 ⎤ (c) False. if A and B are commuting skew-symmetric matrices. 1 1 1 ( A + AT )T = ( AT + ( AT )T ) = ( A + AT ) 2 2 2 1 T Thus ( A + A ) is symmetric. the transpose of an upper triangular matrix is lower triangular. By inspection. it’s sufficient to prove that 1 1 ( A + AT ) is symmetric and ( A − AT ) is 2 2 skew-symmetric. No. ⎡x y⎤ 43. 34 . the entries below the main diagonal are all zero. so x = 1 and 3 z ⎣⎢ 0 ⎦⎥ 37. If AT A = A. Let A = ⎢ . 35. j. A is not symmetric unless n = 1. (b) Let A and B be skew-symmetric matrices. (b) False. Then 3y = 30 and y = 10. j. 2 SSM: Elementary Linear Algebra ( AB)T = ( BA)T = AT BT = (− A)(− B) = AB so the product of commuting skew-symmetric matrices is symmetric rather than skewsymmetric. then AT = ( AT A)T = AT ( AT )T = AT A = A. then A2 = AT A = A. Since A = AT . the entries above the main diagonal determine the entries below the main diagonal in a symmetric matrix. since i. From AT = − A. ⎢ ⎥ ⎣0 0 −1⎦ 1 1 ( A − AT )T = ( AT − ( AT )T ) 2 2 1 T = ( A − A) 2 1 = − ( A − AT ) 2 1 Thus ( A − AT ) is skew-symmetric. 41. (a) A is symmetric since aij = a ji for all i. ⎢ . (a) If A is invertible and skew-symmetric then ( A−1 )T = ( AT )−1 = (− A)−1 = − A−1 so A−1 is also skew-symmetric. for example. ⎡ 1 10⎤ A=⎢ ⎣0 −2⎥⎦ ( AT )T = (− A)T = − AT ( A + B)T = AT + BT = − A − B = −( A + B) ( A − B)T = AT − BT = − A + B = −( A − B) (kA)T = k ( AT ) = k (− A) = −kA True/False 1. j > 0 Thus.

(g) False.SSM: Elementary Linear Algebra Section 1. x1 = 50 = 30 = −10 = 10 60 50 ⎡ 1 −1 0 0 300 ⎤ ⎢0 0 1 −1 −500 ⎥ ⎢ ⎥ ⎢0 1 −1 0 100 ⎥ ⎣⎢0 −1 0 1 400 ⎦⎥ 40 Node A: x1 + x2 = 50 Node B: x1 + x3 = 30 Node C: x3 + 50 = 40 Node D: x2 + 50 = 60 The linear system is ⎡ 1 −1 0 0 300 ⎤ ⎢0 1 −1 0 100 ⎥ ⎢ ⎥ ⎢0 0 1 −1 −500 ⎥ ⎣⎢0 −1 0 1 400 ⎥⎦ 35 . and flow directions as shown. if (kA)T = kA.8 (b) Reduce the augmented matrix. ⎢ + = ⎣0 1⎥⎦ ⎢⎣ 2 6⎥⎦ ⎢⎣ 2 7 ⎥⎦ which is symmetric. (l) False. 300 + x2 = 400 + x3 750 + x3 = 250 + x4 100 + x1 = 400 + x2 200 + x4 = 300 + x1 The system is x2 − x3 = 100 x3 − x4 = −500 x1 − x2 = 300 − x1 + x4 = 100 2 (m) True. for example ⎢ =⎢ ⎣1 −1⎥⎦ ⎣ 0 1⎥⎦ 30 60 10 ⎡ 1 2 ⎤ ⎡ 1 3⎤ ⎡ 2 5 ⎤ (k) False. (a) Consider the nodes from left to right on the top. 50 (i) True. then the bottom. x3 = −10 and the completed network is shown. flow rates. the diagonal entries may be negative. as long as they are nonzero. A 10 30 ⎡ 1 2 ⎤ ⎡ 5 0⎤ ⎡ 6 2 ⎤ (j) False. x2 = 10. for example.8 1. By inspection. so also must be the entries above the main diagonal.8 x1 + x2 x1 + x3 x3 x2 (f) False. then A = AT . ⎡ 0 1 −1 0 100 ⎤ ⎢ 0 0 1 −1 −500 ⎥ ⎢ ⎥ ⎢ 1 −1 0 0 300 ⎥ ⎣⎢ −1 0 0 1 100 ⎦⎥ Exercise Set 1. 40 3. Label the network with nodes. the inverse of an invertible lower triangular matrix is lower triangular. 50 B x3 C ⎡ 1 −1 0 0 300 ⎤ ⎢ 0 0 1 −1 −500 ⎥ ⎢ ⎥ ⎢ 0 1 −1 0 100 ⎥ ⎢⎣ −1 0 0 1 100 ⎥⎦ x2 D 50 40 ⎡1 0 ⎤ ⎡ 1 0⎤ . ⎢ + = ⎣5 1⎥⎦ ⎢⎣ −5 1⎥⎦ ⎢⎣ 0 2 ⎥⎦ which is upper triangular. for example. (h) True. then 0 = (kA)T − kA = kAT − kA = k ( AT − A) since k ≠ 0. adding a diagonal matrix to a lower triangular matrix will not create nonzero entries above the main diagonal. Section 1. the solution is x1 = 40. since the entries below the main diagonal must be zero.

from x3 = −500 + t . ⎡ 1 1 −1 0 ⎤ ⎢ 2 −2 0 6 ⎥ ⎢ ⎥ ⎣ 0 2 4 8⎦ =0 I1 − I 2 − I 4 I 2 − I 4 + I5 = 0 =1 2 I1 + 2 I 2 −2 I 2 + 2 I5 = 1 Reduce the augmented matrix. x3 = 0. But. we have 2 I1 = 2 I 2 + 6 and 2 I 2 + 4 I3 = 8. and 5 5 11 2 13 I1 = I3 − I 2 = + = = 2.2 A. and from back substitution it 5 then follows that 22 2 I 2 = 4 − 2 I3 = 4 − = − = −0. Thus the currents I1. ⎡ 1 −1 −1 0 0⎤ ⎢0 1 −1 1 0 ⎥ ⎢ ⎥ ⎢ 2 2 0 0 1⎥ ⎣⎢ 0 −2 0 2 1⎦⎥ ⎡ 1 1 −1 0 ⎤ ⎢ 1 −1 0 3⎥ ⎢ ⎥ ⎣0 1 2 4 ⎦ ⎡ 1 1 −1 0 ⎤ ⎢ 0 −2 1 3⎥ ⎢ ⎥ 1 2 4⎦ ⎣0 36 .4 A. Thus x1 = 400. 20 I 2 = 20 I3 . From Kirchoff’s current law. I 6 . These four equations can be reduced to the following: I1 = I 2 + I 4 . let x4 = t . x4 = t . 2 I5 − 2 I3 = 1 This gives us six equations in the unknown currents I1 . I 4 = I3 + I5 . I 2 . To keep the traffic flowing on all roads.Chapter 1: Systems of Linear Equations and Matrices SSM: Elementary Linear Algebra ⎡ 1 1 −1 0 ⎤ ⎢0 1 2 4⎥ ⎢ ⎥ ⎣ 0 −2 1 3⎦ ⎡ 1 −1 0 0 300 ⎤ ⎢0 1 −1 0 100 ⎥ ⎢ ⎥ ⎢0 0 1 −1 −500 ⎥ ⎢⎣0 0 −1 1 500 ⎥⎦ ⎡ 1 1 −1 0 ⎤ ⎢0 1 2 4 ⎥ ⎢ ⎥ ⎣0 0 5 11⎦ From the last row we conclude that 11 I3 = = 2. we have I1 + I 2 = I3 . this can be simplified to a system of only four equations in the variables I1. and I3 satisfy the following system of equations: I1 + I 2 − I 3 = 0 =6 2 I1 − 2 I 2 2 I 2 + 4I3 = 8 Reduce the augmented matrix. applied to the left and right loops. These equations can be simplified to: 2 I1 + 2 I 2 = 1. x2 = 100. From Kirchoff’s voltage law. The solution to the system is x1 = −100 + t . Thus. I3 = I 2 . since I3 = I 2 and I 6 = I 4 . x2 = −400 + t . and I5 : (c) The flow from A to B is x4 . and I1 = I 2 + I 6 .6 A. 7. x4 = 500. the minimum flow from A to B is 500 vehicles per hour. x3 = −500 + t . I 6 = I3 + I5 . I 4 = I 6 From application of the voltage law to the three inner loops (left to right) we have 10 = 20 I1 + 20 I 2 . From application of the current law at each of the four nodes (clockwise from upper left) we have I1 = I 2 + I 4 . applied at either of the nodes. I 2 . 5. I 4 . Since I 2 is 5 5 5 negative. its direction is opposite to that indicated in the figure. … . each flow rate must be nonnegative. ⎡ 1 −1 0 0 300 ⎤ ⎢0 1 −1 0 100 ⎥ ⎢ ⎥ ⎢0 0 1 −1 −500 ⎥ 0 ⎦⎥ ⎣⎢0 0 0 0 ⎡ 1 0 0 −1 −100 ⎤ ⎢0 1 0 −1 −400 ⎥ ⎢ ⎥ ⎢0 0 1 −1 −500 ⎥ 0 ⎥⎦ ⎢⎣0 0 0 0 Since x4 is a free variable. I 4 = I3 + I5 . and 10 + 20 I3 = 20 I 5 .

SSM: Elementary Linear Algebra ⎡ 1 −1 −1 0 ⎢0 1 −1 1 ⎢ 0 4 2 0 ⎢ ⎢⎣0 −2 0 2 Section 1. and 2 2 1 1 I1 = I 2 + I 4 = 0 + = . x3 = x4 = t . the number of atoms on the left must be equal to the number of atoms on the right. 2 1 1 1 I 6 = I 4 = − + 2 I5 = − + 1 = .8 homogeneous linear system. 4 4 1 3 1 5 x2 = x3 + x4 = t + t = t . These equations form the 37 . This leads to the equations 3x1 = x3 . and x4 that will balance the chemical equation x1 (C3 H8 ) + x2 (O2 ) → x3 (CO 2 ) + x4 (H 2 O) For each of the atoms in the equation (C. in which case we obtain x1 = 1. H. 1 From this we conclude that I5 = . and that the general solution of the system is 3 3 given by x4 = t . x2 = 5. and 2 4 2 4 1 1 x1 = x3 = t . and x4 = 4. x2 . 2 2 2 1 1 I3 = I 2 = I 4 − I5 = − = 0. x3 . x3 = 3. Thus the balanced equation is C3 H8 + 5O 2 → 3CO 2 + 4H 2 O. ⎡ 3 0 −1 0 0⎤ ⎢ 8 0 0 −2 0⎥ ⎢ ⎥ ⎣ 0 2 −2 −1 0 ⎦ 0⎤ 0⎥ ⎥ 1⎥ 1⎥⎦ ⎡ 1 −1 −1 0 ⎢ 0 1 −1 1 ⎢ ⎢0 0 6 −4 ⎣⎢0 0 −2 4 0⎤ 0⎥ ⎥ 1⎥ 1⎦⎥ ⎡ 1 −1 −1 0 ⎢ 0 1 −1 1 ⎢ ⎢0 0 −2 4 ⎢⎣0 0 6 −4 0⎤ 0⎥ ⎥ 1⎥ 1⎥⎦ ⎡1 0 − 1 0 0⎤ 3 ⎢ ⎥ 0 −2 0 ⎥ ⎢8 0 ⎢0 2 −2 −1 0 ⎥ ⎣ ⎦ ⎡1 0 − 1 0 0⎤ 3 ⎢ ⎥ 8 −2 0 ⎥ ⎢0 0 3 ⎢ ⎥ ⎢⎣0 2 −2 −1 0 ⎥⎦ 0⎤ ⎡ 1 −1 −1 0 ⎢0 1 −1 1 0⎥ ⎢ 1⎥ ⎢0 0 1 −2 − 2 ⎥ ⎢0 0 6 −4 1⎥⎦ ⎣ ⎡1 0 − 1 0 0⎤ 3 ⎢ ⎥ ⎢0 2 −2 −1 0 ⎥ 8 −2 0 ⎥ ⎢0 0 3 ⎣ ⎦ 0⎤ ⎡ 1 −1 −1 0 ⎢0 1 −1 1 0⎥ ⎢ 1⎥ ⎢0 0 1 −2 − 2 ⎥ ⎢0 0 0 8 4 ⎥⎦ ⎣ ⎡1 0 − 1 0 0⎤ 3 ⎢ ⎥ ⎢0 1 −1 − 12 0⎥ ⎢ ⎥ 8 −2 0 ⎥ ⎢⎣0 0 3 ⎦ 0⎤ ⎡ 1 −1 −1 0 ⎢0 1 −1 1 0⎥ ⎢ 1⎥ ⎢0 0 1 −2 − 2 ⎥ ⎢ 1⎥ 1 ⎢⎣0 0 0 2 ⎥⎦ ⎡1 0 − 1 0 0⎤ 3 ⎢ ⎥ ⎢0 1 −1 − 12 0⎥ ⎢ ⎥ 1 − 34 0⎥⎦ ⎢⎣0 0 From this we conclude that x4 is a free variable. 3x1 − x3 =0 8 x1 − 2 x4 = 0 2 x2 − 2 x3 − x4 = 0 Reduce the augmented matrix. The smallest positive integer 3 4 solutions are obtained by taking t = 4. 8 x1 = 2 x4 . and O). and 2 x2 = 2 x3 + x4 . We seek positive integers x1 . 2 2 9.

and F). and from the last two rows we conclude that a2 = 0 and We conclude that a2 = 1. ⎡ 1 1 1 1⎤ ⎢0 1 3 1⎥ ⎢ ⎥ ⎣0 0 2 2 ⎦ ⎡1 ⎢0 ⎢ ⎢0 ⎣⎢0 0 0 0 1⎤ 1 −1 1 2 ⎥ ⎥ 1 1 1 2⎥ 4 16 64 −2 ⎦⎥ ⎡1 ⎢0 ⎢ ⎢0 ⎢⎣0 0 0 0 1⎤ 1 −1 1 2⎥ ⎥ 0 2 0 0⎥ 0 20 60 −10⎥⎦ ⎡1 ⎢0 ⎢ ⎢0 ⎣⎢0 0 0 0 1⎤ 1 −1 1 2⎥ ⎥ 0 1 0 0⎥ 0 20 60 −10⎦⎥ 1⎤ ⎡1 0 0 0 ⎢ 0 1 −1 1 2⎥ ⎢ ⎥ 0⎥ ⎢0 0 1 0 ⎢⎣0 0 0 60 −10 ⎥⎦ From the first row we see that a0 = 1. and x1 = x4 . The graph of the polynomial p( x) = a0 + a1 x + a2 x 2 passes through the points (1. and from back substitution it follows that a1 = 1 − 3a2 = −2. and x4 that balance the chemical equation x1 (CH3COF) + x2 (H 2 O) → x3 (CH3COOH) + x4 (HF) For each of the elements in the equation (C. 1). 1 a3 = − . a1 . (2. This is a very simple system of equations having (by inspection) the general solution x1 = x2 = x3 = x4 = t . 1). 3x1 + 2 x2 = 4 x3 + x4 . 5) if and only if the coefficients a0 . H. Thus. taking t = 1. 3). ⎡ 1 1 1 1⎤ ⎢0 1 3 1⎥ ⎢ ⎥ ⎣0 2 8 4 ⎦ Add −2 times row 2 to row 3. Thus the 6 1 3 13 polynomial is p( x) = − x + x + 1. 6 6 15. 2). if and only if the coefficients a0 . the number of atoms on the left must equal the number of atoms on the right. (1. Thus the quadratic polynomial is p( x) = x 2 − 2 x + 2. −1). The graph of the polynomial 2 SSM: Elementary Linear Algebra 3 p( x) = a0 + a1 x + a2 x + a3 x passes through the points (−1. a1 . x2 . from back substitution. a2 . Finally. it 6 13 follows that a1 = 2 + a2 − a3 = . and a3 satisfy the equations 38 . (0.Chapter 1: Systems of Linear Equations and Matrices a0 − a1 + a2 − a3 = −1 a0 =1 a0 + a1 + a2 + a3 = 3 a0 + 4a1 + 16a2 + 64a3 = −1 The augmented matrix of this system is ⎡1 −1 1 −1 −1⎤ ⎢1 0 0 0 1⎥ ⎢ ⎥ ⎢1 1 1 1 3⎥ ⎣⎢1 4 16 64 −1⎦⎥ 11. O. x3 . and (3. ⎡1 0 0 0 1⎤ ⎢1 −1 1 −1 −1⎥ ⎢ ⎥ ⎢1 1 1 1 3⎥ ⎣⎢1 4 16 64 −1⎦⎥ 1⎤ ⎡1 0 0 0 ⎢0 −1 1 −1 −2 ⎥ ⎢ ⎥ ⎢0 1 1 1 2 ⎥ ⎣⎢0 4 16 64 −2 ⎦⎥ 13. −1). and a0 = 1 − a1 − a2 = 2. We must find positive integers x1 . x1 + x2 = 2 x3 . and (4. This leads to the equations x1 = x3 . the balanced equation is CH3COF + H 2 O → CH3COOH + HF. ⎡1 1 1 1⎤ ⎢1 2 4 2 ⎥ ⎢ ⎥ ⎣1 3 9 5⎦ Add −1 times row 1 to rows 2 and 3. and a2 satisfy the equations a0 + a1 + a2 = 1 a0 + 2a1 + 4a2 = 2 a0 + 3a1 + 9a2 = 5 Reduce the augmented matrix.

5.25 0.3 0 .25 0.6 0 .6 0.25 0. Section 1.1 0 .08 ⎤ ≈⎢ .1 0 . 2) if and only if the coefficients a0 . 500⎥ . (b) False. this is only true if the n points have distinct x-coordinates.5 0.3 ⎤ ⎡ x1 ⎤ ⎡50 ⎤ .2 ⎦ (b) Below are graphs of four curves in the 2 family: y = 1 + x (k = 0). ⎢ ⎥ ⎣ 0 . 4 y k=0 (b) The Leontief equation (I − C)x = d is represented by the augmented matrix ⎡ 0. ⎣0. x = ( I − C )−1 d 1 ⎡0. 3. a1 = k . The Leontief equation (I − C)x = d for this ⎡ 0.90 0.3 ⎥ .3 3860 ⎥ .5 0.9 0.8 5790 ⎦ k=1 3 2 –2 –1 1 k=3 2 The reduced row echelon form of this matrix is ⎡1 0 0 31. 581⎥⎦ 2 p( x) = a0 + a1 x + a2 x passes through the points (0. a2 = 1 − k .2 0 . y = 1 + x (k = 1). and a2 satisfy the equations a0 =1 a0 + a1 + a2 = 2 The general solution of this system.1 0.4 1930 ⎤ ⎢ −0.3875 ⎢⎣0.000 worth of body work is given by x = ( I − C )−1 d 1 ⎡0. 000 ⎥⎦ ⎡\$25. 500 ⎤ ⎢0 1 0 26. 1) and (1. 300⎦ (a) True. ⎣ 202.10 ⎥⎦ 39 . only such networks were considered.9 1.25 0. and y = 1 + 3x − 2 x 2 (k = 3).8 −0.4 −0. using a1 as a parameter.4 ⎤ C = ⎢ 0 .6 ⎥⎦ ⎣ x2 ⎦ ⎢⎣60 ⎥⎦ required production vector is (c) True (d) False. (a) A consumption matrix for this economy is ⎡0.9 −0. 500 ⎥ .25⎤ .3⎤ ⎡50 ⎤ = 0.4 0 . = 0.9 Exercise Set 1. the number of atoms of each element must be the same on each side of the equation.SSM: Elementary Linear Algebra Section 1.25⎤ ⎡ 7000 ⎤ = 0. is a0 = 1.9⎥⎦ ⎢⎣ 60 ⎥⎦ 1 ⎡ 48⎤ = 0.39 ⎢⎣0.3 0. when a current passes through a resistor. there is a drop in electrical potential in the circuit. ⎢ ⎥ ⎣0 0 1 26. ⎢ ⎥ ⎣ −0. Thus this family of polynomials is represented by the equation p( x) = 1 + kx + (1 − k ) x 2 where −∞ < k < ∞. 300 ⎦ x 3 k=2 True/False 1.25 ⎤ C=⎢ .56 ⎥⎦ (e) False. a1 . (a) The quadratic polynomial ⎡ 0. 500 ⎤ x = ⎢\$26.50 −0.90 ⎥⎦ 1 ⎡ 0.3875 ⎢⎣ 0. thus the economy is ⎢ ⎢ ⎥= ⎣ −0.39 ⎢⎣79 ⎥⎦ ⎡123.9 −0. y = 1 + 2 x − x 2 (k = 2).6 −0.50 0.25 ⎤ (b) We have ( I − C ) −1 = ⎢ ⎣ −0. 290 ⎤ ≈⎢ ⎣ \$22.5 ⎥⎦ ⎢⎣14.50⎥⎦ and so the production needed to provide customers \$7000 worth of mechanical work and \$14.8 Thus the required production vector is ⎡ \$31. ⎢ ⎥ ⎣\$26.9 −1 17. (a) A consumption matrix for this economy is ⎡ 0 .

open sectors are those that do not produce output. If 1.Chapter 1: Systems of Linear Equations and Matrices SSM: Elementary Linear Algebra Chapter 1 Supplementary Exercises 7.9. The corresponding system is 2 x1 − 4 x2 + x3 = 6 −4 x1 + 3x3 = −1 x2 − x3 = 3 ( I − C ) −1 = (( I − C T )T )−1 = (( I − CT )−1 )T has nonnegative entries. If d1 = 2 and d 2 = 0. Thus the Leontief equation (I − C)x = d has a unique solution for every demand vector d. from Theorem 1. ⎡ 1 0⎤ (b) The consumption matrix C = ⎢ 2 ⎥ ⎣ 0 1⎦ indicates that the entire output of the second sector is consumed in producing that output. that the matrix I − C is invertible. by Theorem 1. x4 = t . it follows that (1 − C11 ) − C21C12 > 0. Since I − C = ( I T − C T )T = ( I − C T )T .9 (a) False. 2 2 2 9 1 5 x2 = − s − t − . x3 and x4 are free variables. Since c21c12 < 1 − c11. (e) True 40 . This shows that the economy is productive. x3 = s. The corresponding system is 3 x1 − x2 + 4 x4 = 1 2 x1 + 3x3 + 3x4 = −1 ⎡ 3 −1 0 4 1⎤ ⎢⎣ 2 0 3 3 −1⎥⎦ d1 = 2 and d 2 = 1 the system is inconsistent. 9. 9 1 5 9 1 5 x2 = − x3 − x4 − = − s − t − 2 2 2 2 2 2 1 4 1 x1 = x2 − x4 + 3 3 3 1⎛ 9 1 5⎞ 4 1 = ⎜− s − t − ⎟ − t + 3⎝ 2 2 2⎠ 3 3 3 3 1 =− s− t− 2 2 2 3 3 1 The solution is x1 = − s − t − . we can conclude that I − C is invertible and that 3. 2 2 2 11.1. the ⎡ 1 0⎤ Leontief matrix I − C = ⎢ 2 ⎥ is not ⎣ 0 0⎦ invertible. (a) The Leontief equation for this economy is ⎡ 1 0 ⎤ ⎡ x1 ⎤ ⎡ d1 ⎤ ⎢2 ⎥ ⎢ ⎥ = ⎢ ⎥ . then CT has column sums less than 1. Thus. ⎡1 − 1 0 3 ⎢ 0 3 ⎣2 4 3 1⎤ 3⎥ ⎡1 − 1 0 3 ⎢ 2 3 ⎢⎣0 3 4 3 1 3 1⎤ 3⎥ − 35 ⎥⎦ 3 −1⎦ 1⎤ ⎡1 − 1 0 4 3 3 3⎥ ⎢ 1 92 12 − 25 ⎥⎦ ⎢⎣0 Thus. and c12 ⎤ 1 ⎡ 1 ( I − C ) −1 = has ⎢ 1 − c c11 ⎥⎦ (1 − C11 ) − C21C12 ⎣ 21 nonnegative entries. Mathematically. ⎣ 0 0 ⎦ ⎣ x2 ⎦ ⎣ d 2 ⎦ the system is consistent with general solution x1 = 4. it follows that I − C T is invertible and that ( I − C T )−1 has nonnegative entries. ⎡ 2 −4 1 6 ⎤ ⎢ −4 0 3 −1⎥ ⎢ ⎥ 1 −1 3⎦ ⎣ 0 True/False 1.1. If C has row sums less than 1. ⎡ 1 −2 1 3⎤ 2 ⎢ ⎥ ⎢ −4 0 3 −1⎥ ⎢ 0 1 −1 3 ⎥ ⎣ ⎦ (b) True (c) True (d) True. From this we conclude Let x3 = s and x4 = t .9. x2 = t (0 ≤ t < ∞). thus there is nothing left to satisfy any outside demand.

x = 4. or 3. then x3 is a free variable and the system has a one-parameter solution. 2. b 2 ⎤ ⎡a 0 ⎢0 a 4 − b 2 ⎥ ⎢ ⎥ ⎣0 0 b − 2 b − 2⎦ ⎤ ⎥ y⎥ ⎦ (a) The system has a unique solution if the corresponding matrix can be put in reduced row echelon form without division by zero. For x. 3 5 3 ⎡1 − 4 3 ⎢ ⎢0 53 ⎣ 5x 3 − 43 x + ⎡1 − 4 3 ⎢ ⎢0 1 ⎣ ⎤ ⎥ − 54 x + 35 y ⎥ ⎦ 9⎤ 35 ⎥ 4⎦ ⎡a 0 b 2⎤ 9. a ≠ 0. Let z = s. 5 5 5 5 Add −1 times the second row to the third. The only possibility is s = 3. (c) If a = 0 and b = 2. and z to 4 4 4 4 be positive integers. s must be a positive integer such that 5s + 1 and −9s + 35 are positive and divisible by 4. b ≠ 2 is required. y ′ = − x + y.SSM: Elementary Linear Algebra Chapter 1 Supplementary Exercises ⎡ 1 −2 1 3⎤ 2 ⎢ ⎥ 0 8 5 11⎥ − ⎢ ⎢0 1 −1 3⎥ ⎣ ⎦ ⎡1 0 3 x + 4 y ⎤ 5 5 ⎥ ⎢ ⎢0 1 − 45 x + 53 y ⎥ ⎣ ⎦ The solution is x ′ = ⎡ 1 −2 1 3⎤ 2 ⎢ ⎥ 1 −1 3⎥ ⎢0 ⎢0 −8 5 11⎥ ⎣ ⎦ 7. Since −9s + 35 > 0. Reduce the augmented matrix. 2 3 6 2 17 26 The solution is x1 = − . y = 2. x2 = − . Reduce the augmented matrix. Then 9 35 5 1 y =− s+ and x = s + . Thus. y. and 3 3 1 52 35 17 x1 = 2 x2 − x3 + 3 = − + + 3 = − . x (b) If a ≠ 0 and b = 2. 41 . the system has a twoparameter solution. b 2⎤ ⎡a 0 ⎢0 a 4 − b 2⎥ ⎢ ⎥ b⎦ 2 ⎣0 a x⎤ ⎥ y⎥ ⎦ 5 3 1 9 4 ⎡1 0 − 5 1 ⎤ 4 4 ⎥ ⎢ 9 35 ⎢⎣0 1 ⎥ 4 4⎦ Thus z is a free variable. 5. so s = 1. ⎢ a a 4 4 ⎥ ⎢ ⎥ ⎣0 a 2 b⎦ Add −1 times the first row to the second. 2 3 35 x3 = − . ⎡3 − 4 x⎤ 5 ⎢5 ⎥ ⎢ 45 35 y ⎥ ⎣ ⎦ ⎡1 − 4 3 ⎢ ⎢ 45 35 ⎣ 3 4 4 3 x + y. s < 4. z = 3. x2 = x3 + 3 = − . ⎡1 1 1 9 ⎤ ⎢⎣1 5 10 44 ⎥⎦ ⎡ 1 1 1 9⎤ ⎢⎣0 4 9 35⎥⎦ ⎡ 1 −2 1 3⎤ 2 ⎢ ⎥ 1 −1 3⎥ ⎢0 ⎢0 0 −3 35⎥ ⎣ ⎦ ⎡1 1 ⎢0 1 ⎣ ⎡ 1 −2 1 3⎤ 2 ⎢ ⎥ 1 −1 3⎥ ⎢0 ⎢0 0 1 − 35 ⎥ 3⎦ ⎣ 35 26 Thus x3 = − . Thus.

Chapter 1: Systems of Linear Equations and Matrices SSM: Elementary Linear Algebra 13. ⎣1 1 ⎥⎦ ⎡ 1 0 −1 0 5⎤ ⎢ 0 1 1 2 1⎥ ⎢ ⎥ − − 1 1 3 1 3 ⎣ ⎦ ⎡ 1 0 −1 0 5 ⎤ ⎢0 1 1 2 1⎥ ⎢ ⎥ ⎣0 1 2 1 2 ⎦ ⎡ 1 0 −1 0 5 ⎤ ⎢0 1 1 2 1⎥ ⎢ ⎥ ⎣0 0 1 −1 1⎦ ⎡ 1 0 0 −1 6 ⎤ ⎢0 1 0 3 0 ⎥ ⎢ ⎥ ⎣0 0 1 −1 1⎦ Thus. ⎡ −1 1 3 1 −3⎤ ⎢ 0 1 1 2 1⎥ ⎢ ⎥ ⎣ 1 0 −1 0 5⎦ ⎡ 1 4⎤ ⎡ 8 6 −6 ⎤ ⎡ 2 a b −b ⎤ ⎢ or ⎢ −2 3⎥ ⎢ = 6 −1 1⎥ or ⎢ ⎥ ⎣ 2c d −d ⎥⎦ ⎢ ⎥ 1 − 2 − ⎣ ⎦ ⎣ 4 0 0⎦ −b − 4 d ⎤ b + 4d ⎡ 2a + 8c ⎢ −4a + 6c −2b + 3d 2b − 3d ⎥ ⎢ ⎥ −b + 2d ⎦ b − 2d ⎣ 2 a − 4c ⎡ 8 6 −6 ⎤ = ⎢ 6 −1 1⎥ . ⎢0 0 0 1 1⎥ ⎢0 0 0 0 0 ⎥ ⎢0 0 0 0 0 ⎥ ⎣ ⎦ Thus a = 0. ⎢ ⎥ ⎣ −4 0 0 ⎦ Thus 2a + 8c =8 b + 4d = 6 −4 a + 6c =6 −2b + 3d = −1 2a − 4c = −4 − 2d = 0 b Note that we have omitted the 3 equations obtained by equating elements of the last columns of these matrices because the information so obtained would be just a repeat of that gained by equating elements of the second columns. a = −1. Let ⎡a b ⎤ K =⎢ . the system has no solution. 9. Note that K must be a 2 × 2 matrix. d = 6. Let ⎡ a b c⎤ X =⎢ . 3 −1⎥ ⎢ 0 −2 0 ⎢ 2 0 −4 0 −4 ⎥ ⎢ 0 1 0 −2 0 ⎥⎦ ⎣ The reduced row-echelon form of this matrix is ⎡ 1 0 0 0 0⎤ ⎢0 1 0 0 2 ⎥ ⎢ ⎥ ⎢0 0 1 0 1⎥ . f = 1 and X = ⎢ ⎣ 6 0 1⎥⎦ 42 . c = 1. ⎣ d e f ⎥⎦ Then ⎡ −1 0 1⎤ ⎡ a b c⎤ ⎢ ⎥ ⎡ 1 2 0⎤ ⎢⎣ d e f ⎥⎦ ⎢ 1 1 0⎥ = ⎢⎣ −3 1 5⎥⎦ or ⎣ 3 1 −1⎦ (d) If b ≠ 2 but a = 0. b = 3. (a) Here X must be a 2 × 3 matrix. c = −1. The augmented matrix of the above system is 8 0 8⎤ ⎡ 2 0 ⎢ 0 1 0 4 6⎥ ⎢ ⎥ ⎢ −4 0 6 0 6 ⎥ . ⎣ c d ⎥⎦ Then ⎡ 1 4⎤ ⎡ 8 6 −6 ⎤ ⎡a b ⎤ ⎡ 2 0 0⎤ ⎢ ⎢ −2 3⎥ ⎢ = 6 −1 1⎥ ⎢ ⎥ ⎣ c d ⎦⎥ ⎣⎢ 0 1 −1⎦⎥ ⎢ ⎥ ⎣ 1 −2 ⎦ ⎣ −4 0 0 ⎦ ⎡ −a + b + 3c b + c a − c ⎤ ⎢⎣ − d + e + 3 f e + f d − f ⎥⎦ ⎡ 1 2 0⎤ . and d = 1. e = 0. ⎡ −1 3 −1⎤ . =⎢ ⎣ −3 1 5⎥⎦ Equating entries in the first row leads to the system − a + b + 3c = 1 b +c = 2 −c = 0 a Equating entries in the second row yields the system − d + e + 3 f = −3 e + f =1 d − f =5 These systems can be solved simultaneously by reducing the following augmented matrix. b = 2. ⎡0 2 ⎤ The matrix is K = ⎢ .

z=− . Consider the ith row of AB. 37 37 ⎡ − 113 − 160 ⎤ 46 37 ⎥ . ⎣3 0 1⎥⎦ ⎢⎣ z + 3w − z 2 z + w⎥⎦ If we equate matrix entries. X must be a 2 × 2 matrix. and c = 3. we have a +b+c = 2 p(1) = 2 ⇒ a −b +c = 6 p(−1) = 6 ⇒ p(2) = 3 ⇒ 4a + 2b + c = 3 This system has augmented matrix ⎡ 1 1 1 2⎤ ⎢ 1 −1 1 6 ⎥ which reduces to ⎢ ⎥ ⎣ 4 2 1 3⎦ 1⎤ ⎡1 0 0 ⎢0 1 0 −2 ⎥ . A similar argument shows that if AB = BA then AB −1 = B −1 A. x = − 37 ⎤ 0 0 − 113 37 ⎥ 0 0 − 160 ⎥ 37 . First suppose that AB −1 = B −1 A. Hence. b = −2. ⎡x y⎤ Let X = ⎢ . so that the matrix equation ⎣ z w⎥⎦ becomes 3y + w ⎤ ⎡ x + 2 y 4x⎤ ⎡ 3x + z ⎢⎣ − x + 2 z − y + 2w⎥⎦ − ⎢⎣ z + 2 w 4 z ⎥⎦ ⎡ 2 −2 ⎤ . so that the top two equations give y = −2 and w = 1. n 1 0 2⎤ ⎡ 2 −2 ⎢ −4 3 0 1 −2 ⎥ with matrix ⎢ ⎥ which 1 −2 5 ⎥ ⎢ −1 0 ⎣⎢ 0 −1 −4 2 4 ⎦⎥ ⎡1 0 ⎢ ⎢0 1 reduces to ⎢ ⎢0 0 ⎢ ⎢⎣0 0 113 . we have that ⎡1 − 2 ⎤ X =⎢ . By the definition of matrix multiplication. this is ai1b11 + ai 2b21 + + ain bn1 a + a + + ain = i1 i 2 n = ri 1 since all entries bi1 = . ⎥ 20 1 0 − 37 ⎥ 46 ⎥ 0 1 − 37 ⎥⎦ 160 20 y=− . ⎢ ⎥ ⎣ 0 0 1 3⎦ Thus. ⎣3 1 ⎥⎦ 15. Let ⎡x y⎤ X =⎢ . 21. 19. this gives us the equations z + 3w = 6 x + 3 y = −5 − z = −3 − x = −1 2z + w = 7 2x + y = 0 Thus x = 1 and z = 3. w = − .SSM: Elementary Linear Algebra Chapter 1 Supplementary Exercises (b) X must be a 2 × 2 matrix. Therefore ( AB −1 ) B = ( B −1 A) B or A = B −1 AB (c) Again. Since these values are consistent with the bottom two equations. and X = ⎢ 37 20 46 ⎥ 37 ⎢ − 37 − 37 ⎣ ⎦ 43 . =⎢ ⎣ 5 4 ⎥⎦ This yields the system of equations 2x − 2 y + z =2 −4 x + 3 y + w = −2 −x + z − 2w = 5 − y − 4 z + 2w = 4 so that BA = B( B −1 AB) = ( BB −1 )( AB) = AB. a = 1. Note that all matrices must be square and of the same size. Then ⎣ z w⎥⎦ ⎡ 1 −1 2 ⎤ ⎡ x + 3 y − x 2 x + y ⎤ = X⎢ . Since the coordinates of the given points must satisfy the polynomial.

upon first4 2 0 1 34 5 3 1 row cofactor expansion. (a) From Equation (11). y = 3x − 4. upon cofactor expansion along the first row. 2 dividing by 18. Using Equation (10) we obtain Section 10. x + 2y + z = 0.1 x2 0 0 4 4 16 Exercise Set 10. that is. In standard form this is ( x + 1)2 + ( y − 2)2 = 25. x y 2 −2 3 5 −4 6 1 1 1 = 0 which is the 1 1 1 160 x 2 + 320 xy + 160( y 2 + y) − 320 x = 0. 1 1 = 0 yields 1 1 50( x 2 + y 2 ) + 100 x − 200 y − 1000 = 0. 211 . 34 52 x y 0 0 0 −1 2 0 2 −5 4 −1 x 2 xy y 2 + y x 4 0 0 2 = 0. Completing the squares in x and y yields the standard form ( x − 2)2 + ( y − 3)2 = 9. x2 + y 2 8 (b) As in (a). x + y − 4 x − 6 y + 4 = 0. that x y z 2 3 1 (b) As in (a). 18( x 2 + y 2 ) − 72 x − 108 y + 72 = 0 or. 2x + y − 1 = 0 or y = −2x + 1. which is the equation of a parabola. 1 −1 1 1 0 −1 2 1 Expansion along the first row yields −2x − 4y − 2z = 0. x 2 + y 2 + 2 x − 4 y − 20 = 0. that is −x + y − 2z + 1 = 0 for the equation of the plane. yields x 2 + 2 xy + y 2 − 2 x + y = 0. 4 −10 20 2 16 −4 0 4 Now expand along the first row and get 2 2 y2 0 1 0 25 1 x 2 xy y 2 x y 0 0 1 0 −1 same as 4 0 0 2 0 = 0 by 4 −10 25 2 −5 16 −4 1 4 −1 expansion along the second row (taking advantage of the zeros there). 2 −1 −1 1 2 1 is. that is.Chapter 10 Applications of Linear Algebra 3. the equation of the x y z 1 1 1 −3 1 plane is = 0. (a) Substituting the coordinates of the points x y 1 into Equation (4) yields 1 −1 1 = 0 2 2 1 which. 2. Add column five to column three and take advantage of another row of all but one zero to get x y 1 (b) As in (a). 1 1 = 0 yields 1 1 −2x + 2y − 4z + 2 = 0. yields −3x + y + 4 = 0. (a) Equation (9) yields x +y x y 1 40 2 6 1 = 0 which.1 1. that is. 0 1 1 = 0 yields 1 −1 1 2 xy 0 0 0 −10 −4 4.

that is. Completing the squares in each variable yields the standard form 9. which is exactly Equation (11). Thus the determinant of this system is zero. y1 ). yi .. y4 ). 212 . As in the previous problem. c6 . Thus the determinant of the coefficient matrix is zero. y2 ). ( x2 ... 16( x 2 + y 2 + z 2 ) − 32 x − 64 y − 32 z + 32 = 0. which together with the above sphere equation form a homogeneous linear system for c1. x 2 + y 2 + z 2 − 2 x − 2 y − 3 = 0 or in standard form. (b) The parallel planes passing through the origin are x + 2y + z = 0 and −x + y − 2z = 0. For the plane passing through the origin parallel to the plane passing through the three points. zi ) of the four points into the equation c1 ( x 2 + y 2 + z 2 ) + c2 x + c3 y + c4 z + c5 = 0 of the sphere yields four equations. i = 1. ⎢ x2 y2 z2 1⎥ ⎢ c3 ⎥ ⎢ 0⎥ ⎣⎢ x3 y3 z3 1⎦⎥ ⎢⎣ c4 ⎥⎦ ⎢⎣ 0⎥⎦ Rows 2 through 4 show that the plane passes through the three points ( xi .. 11 5 11 yields x y z 0 1 −2 1 3 1 2 −1 0 3 1 −1 1 1 1 =0 1 1 −24( x 2 + y 2 + z 2 ) + 48 x + 48 y + 72 = 0. y5 ) into the equation c1 x 2 + c2 xy + c3 y 2 + c4 x + c5 y + c6 = 0 yields c1 x12 + c2 x1 y1 + c3 y12 + c4 x1 + c5 y1 + c6 = 0 # # # c1 x52 + c2 x5 y5 + c3 y52 + c4 x5 + c5 y5 + c6 = 0. it is useful to take advantage of the column of ones and elementary row operations. 3. c4 . 6. the constant term in the final equation will be 0.Chapter 10: Applications of Linear Algebra SSM: Elementary Linear Algebra 5. Substituting each of the points ( x1 . yi . Thus the determinant of the coefficient matrix is zero. 2 1 0 1 1 6 1 2 −1 1 Expanding by cofactors along the first row yields 8. which is exactly Equation (10). ( x4 . Substituting the coordinates ( xi . (a) Using Equation (12). yi .. while row 1 gives the equation c1 x + c2 y + c3 z + c4 = 0. which is Equation (12). the cofactor of x 2 + y 2 + z 2 above can be evaluated as follows: 1 2 3 1 1 2 3 1 −1 2 1 1 −2 0 −2 0 = = 16 by 1 0 1 1 0 −2 −2 0 1 2 −1 1 0 0 −4 0 cofactor expansion of the latter determinant along the last column. . zi ). which is accomplished by using x y z 0 x1 y1 z1 1 =0 x2 y2 z2 1 x3 y3 z3 1 x2 + y 2 + z 2 5 (b) As in (a). . ( x − 1) 2 + ( y − 1)2 + z 2 = 5... zi ) of each of the three points into the equation c1 x + c2 y + c3 z + c4 = 0 to obtain a homogeneous system with nontrivial solution for c1. . substitute the coordinates ( xi .. Note: When evaluating the cofactors. respectively. y3 ). ( x3 . These together with the original equation form a homogeneous linear system with a non-trivial solution for c1. 7.. 2. for example. and ( x5 . ( x − 1) 2 + ( y − 2)2 + ( z − 1)2 = 4. c2 . that is. ( x 2 + y 2 + z 2 ) − 2 x − 4 y − 2 x + 2 = 0. c5 with a nontrivial solution. the equation of the x2 + y2 + z 2 x y z 1 14 1 2 3 1 sphere is 6 −1 2 1 1 = 0. (a) Equation (11) involves the determinant of the coefficient matrix of the system y z 1⎤ ⎡ c1 ⎤ ⎡ 0⎤ ⎡x ⎢x y1 z1 1⎥ ⎢ c2 ⎥ ⎢ 0⎥ 1 ⎢ ⎥ ⎢ ⎥ = ⎢ ⎥.

The values of the objective function are shown in the following table: = 0. it is clear that the determinant is 0 and Equation (11) becomes 0 = 0. i. 2 x1 The intersection of the five half-planes defined by the constraints is empty. then the determinant in Equation (11) is x −ax − b −cx − d 1 x 0 0 1 . is attained when 3 2 x1 = 2 and x2 = . y x2 x 1 y1 x12 x22 x32 x1 1 y2 y3 x2 1 1 .2 10. Expanding the determinant in Equation (9) by cofactors of the first row makes it apparent that the coefficient of x 2 + y 2 in the final equation is x1 x2 x3 x 2 2 x1 − x2 = 0 x1 = 2 1. cx + d). If the three distinct points are collinear then two of the coordinates can be expressed in terms of the third. If the points are collinear. equivalent to 1 x2 0 0 1 x3 0 0 1 Expanding along the first row. x3 1 11.. Thus this problem has no feasible solutions. y1 ). 0) 6 13 2 Thus the maximum. we can say that y and z can be expressed in terms of x. ax + b. 1) ( 2. (0. c2 . y1 1 y2 1 .1 2 2.SSM: Elementary Linear Algebra Section 10. 22 . Without loss of generality. 12. c3 . ( x2 . so the coefficient of x 2 + y 2 = 0 and the resulting equation is that of the line through the three points. 3 2. 0) 0 ( 12 . y2 ) and ( x3 . Upon substitution of the coordinates of the three points ( x1 .2 Exercise Set 10. 213 . x2 x2 = 3 3 2 x1 − x2 = −2 4 x1 − x2 = 0 13.1 2 3 . 0) x2 = 1 2 3 2 x1 + 3 x2 = 6 5 x1 In the figure the feasible region is shown and the extreme points are labeled. y3 ). so the determinant of the coefficient matrix is zero. 23 ) 22 3 (2. c4 .2 c1 y + c2 x 2 + c3 x + c4 = 0 c1 y1 + c2 x12 + c3 x1 + c4 c1 y2 + c2 x22 + c3 x2 + c4 c1 y3 + c2 x32 + c3 x3 + c4 =0 =0 = 0. Extreme point ( x1 . then the y3 1 columns are linearly dependent ( yi = mxi + b). As in Exercise 11. the coefficient of x 2 + y 2 + z 2 will be 0. x is the parameter. so Equation (12) gives the equation of the plane in which the four points lie. we obtain the equations: Section 10. x2 ) Value of z = 3x1 + 2 x2 (0. This is a homogeneous system with a nontrivial solution for c1.e. If the line is (x. 0) (2. that is. 1) 7 2 ( 32 .

if we use z there is 18 7 25 cup of milk and ounces of corn flakes. then the feasible region is empty. 2000) x2 = 2000 (2000.5 x1 + 5.1x1 + . 0) x1 − 2 x2 = 0 3 3 . but now the maximum occurs for x1 = 925 and x2 = 0. where z = 2312. A truck should carry 550 containers from Company A and 300 containers from Company B for maximum shipping charges of \$2110. The maximum value of the objective function z = . x1 + x2 = 10. shown in the figure. 5000) 7. x1 1 1 1 x8 + x2 = 8 1 10 3 The feasible region and its extreme points are shown in the figure.50. 9 18 x1 = 0. is unbounded. x1 x1 − x2 = 0 The feasible region and vertices for this problem are shown in the figure. Letting x1 be the number of Company A’s containers shipped and x2 the number of Company B’s. so 8. 5. We must now maximize z = 2. which is attained at the 14 25 and x2 = .07 x2 is attained x2 when x1 = 6000 and x2 = 4000. Of the remaining constraints. (5000. 2000) 40 x1 + 50 x2 ≤ 37. (a) Both x1 ≥ 0 and x2 ≥ 0 are nonbinding. x1 = 6000 (c) x2 ≤ v will be nonbinding for v ≥ 8 and the feasible region will be empty for v < 0. If v < −6.6¢ is realized. 0) x1 The feasible region is shown in the figure. In the problem’s terms. 2 2 2 3 (925. it has a minimum..e. Thus. where z = 2110.Chapter 10: Applications of Linear Algebra SSM: Elementary Linear Algebra x2 3. Though the region is unbounded.0 x2 is also. 000 x2 (b) x1 − x2 ≤ v will be binding if the line x1 − x2 = v intersects the line x2 = 6 with 0 < x1 < 3. only 2 x1 + 3x2 ≤ 24 is binding. 4. In other words. 666 3 3x1 − 2 x1 = 0 2 3 9 . the problem is: Maximize z = 2. The feasible region is as in Exercise 7. 214 . The value of point where x1 = 9 18 335 . thus if −6 < v < −3. x1 and x2 are always positive.20 x1 + 3. she should invest \$6000 in bond A and \$4000 in bond B for annual yield of \$880. The feasible region for this problem. x2 5 14 25 .00 x2 . i. 21 21 4 x1 + 2 x2 = 9 40 x1 + 50 x2 = 37. a 9 18 minimum cost of 18. The value of z = −3x1 + 2 x2 cannot be minimized in this region since it becomes arbitrarily negative as we travel outward along the line − x1 + x2 = 1. so that z = 880. 300) 2 x1 + 3x2 = 2000 (0. the value of z is −3x1 + 2 x2 = −3x1 + 2( x1 + 1) = − x1 + 2 and x1 can be arbitrarily large. 5 − x1 + x2 = 1 3x1 − x2 = −5 2 x1 + 4 x2 = 12 x1 6. The vertex at which the maximum is attained is x1 = 550 and x2 = 300. 000 (550.50 x1 + 3. 2 4 40 20 .00 x2 subject to (6000. the objective function z = 7. 4000) (6000. 000 2 x1 + 3x2 ≤ 2000 x1 ≥ 0 x2 ≥ 0.

Then the problem is: Minimize z = 8 x1 + 9 x2 subject to 2 x1 + 5 x2 ≥ 10 2 x1 + 3x2 ≥ 8 6 x1 + 4 x2 ≥ 12 x1 ≥ 0 x2 ≥ 0 2. x2′′ ) and a(tx1′ + (1 − t ) x1′′) + b(tx2′ + (1 − t ) x2′′ ) = t (ax1′ + bx2′ ) + (1 − t )(ax1′′ + bx2′′ ) = tc + (1 − t )c = c. 0) 5 x1 Though the feasible region shown in the figure is unbounded. The corresponding array is (5. The minimum value of z is 5 5 124 or 24. so there are 350 oxen. tx2′ + (1 − t ) x2′′ ) is an arbitrary point on the line connecting ( x1′ . effectively.3 1. and there are 7 herds. D = 4(A + B + C). Note that this is. C = 3B. 5 Each sack of feed should contain 0. Let x1 be the number of pounds of ingredient A used and x2 the number of pounds of ingredient B.8¢. x2′ ) to ( x1′′. 5 5 3. and 21 34 units for an ox. Solving this linear system gives A = 4 (and B = 8.3 Exercise Set 10. (a) The equations are B = 2A. 21 (b) Let x. It is sufficient to show that if a linear function has the same value at two points. D = 240).4 lb of ingredient B for a minimum cost of 24. and z represent the number of bundles of each class. x 2 (0. Section 10.3 925 containers from Company A and no containers from Company B should be shipped for maximum shipping charges of \$2312. This minimum occurs at the vertex where 2 12 x1 = and x2 = . Then the equations are 2x + y =1 3y + z = 1 x 4z = 1 and the corresponding array is 215 . D = 4C. C = 45. and ax1′ + bx2′ = ax1′′ + bx2′′ = c. 2 5 5 2 8 10 and the elimination step subtracts twice column 2 from five times column 1. 132 = A + B + C + D.50. C = 24.8¢. then the equations are 5x + 2y = 10 and 2x + 5y = 8.1 2 2 x1 + 5 x2 = 10 6 x1 + 4 x2 = 12 4. x= 20 unit for a sheep. giving 5 10. ⎣ 4 ⎦ 2 x1 + 3x2 = 8 5 . D = 96). The number of oxen is 50 per herd.4 lb of ingredient A and 2.SSM: Elementary Linear Algebra Section 10. Let ax1 + bx2 be the function. then it has that value along the line connecting the two points. the objective function is always positive there and hence must have a minimum. y. Solving this linear system gives A = 5 (and B = 10. 300 = A + B + C + D. Then 21 2 20 10 and so y = (tx1′ + (1 − t ) x1′′. Gaussian elimination applied to the augmented matrix ⎡1 1 10 ⎤ ⎢1 1 7 ⎥ . 3) 2 12 . (a) Let x represent oxen and y represent sheep. (b) The equations are B = 2A. 9. Hence the total number of oxen and sheep is 350 + 350 = 700. C = 3(A + B).

y. . (a) We can write this as 5x + y + z − K = 0 x + 7y + z − K = 0 x + y + 8 − K = 0 (a 3 × 4 system). and third class class contains 25 4 measure. 2 4 1 1 Subtract two times the numbers in the third column from the second column to get 1 3 1 1 −8 4 1 −1 Now subtract three times the numbers in the second column from the first column to get 1 1 1 25 −8 4 4 −1 1 This is equivalent to the linear system x + 4z = 1 y − 8 z = −1. n). contains 25 6. 25 z = 4 From this. n) to find the other xi . 131 131 Since 131 is prime. the solution is that a bundle of the 9 measure. y= x j = a j − x1 (j = 2. 131 14 K 12 K . 3 3 a3 = (60) = 45. x j = a j − x1 (j = 2. we must choose K to be an integer multiple of 131 to get integer solutions. x3 represents tin. (a) From equations 2 through n. 5.. so n = 4 and a1 = 60. there is a unique solution x. K is an arbitrary parameter. 9 minae of brass. z for every K. . Then we can use 21K . hence. y = 14. Since the coefficient matrix of equations (5) is invertible (its determinant is 262).Chapter 10: Applications of Linear Algebra 2 3 (b) Exercise 7(b) may be solved using this technique.. for any choice of K. Using these equations in equation 1 gives x1 + (a2 − x1 ) + (a3 − x1 ) + " + (an − x1 ) = a1 (b) Gaussian elimination gives x = a + a + " + an − a1 x1 = 2 3 . and x4 represents much-wrought iron.. 216 . The obvious choice is K = 131... 2 a2 = (60) = 40. giving x = 21. x2 represents 1 1 1 1 SSM: Elementary Linear Algebra brass. second first class contains 25 7 measure. 14 minae of 2 2 1 tin.. z = 12. x1 represents gold. 5 (a2 + a3 + a4 ) − a1 x1 = n−2 40 + 45 + 36 − 60 = 4−2 61 = 2 61 19 x2 = a2 − x1 = 40 − = 2 2 61 29 x3 = a3 − x1 = 45 − = 2 2 61 11 x4 = a4 − x1 = 36 − = 2 2 1 The crown was made with 30 minae of 2 1 1 gold. 4 3 a4 = (60) = 36. z= . and K = 131. n−2 First find x1 in terms of the known quantities n and the ai . and 5 minae of iron.

b3 . the cubic runout spline will agree exactly with the single cubic curve.92158( x − .2915 ⎦⎥ ⎛2⎞ (b) The system is G + B = ⎜ ⎟ 60. 2 Solving this system yields M 2 = −. and d3 gives (M 4 − M 3 ) = −. y6 = .38942 = .3750 = −4.5) = −. M 5 = −.1880 ⎤ ⎢1 4 1 0 ⎥ ⎢ M 3 ⎥ ⎢ −2.4 (b) S (. 9 h2 6( y3 − 2 y4 + y5 ) h2 6( y4 − 2 y5 + y6 ) = −1.6 is the third interval. with solution G = 30.1) 2 + . the illegitimate son 9 2 receives 422 . B = 9.0.2. ⎟ ⎝3 ⎠ ⎛1⎞ ⎛1⎞ B = C + ⎜ ⎟ A. y2 = . ⎝4⎠ ⎝5⎠ G + B + T + I = 60. ⎝3⎠ ⎛3⎞ ⎛3⎞ G + T = ⎜ ⎟ 60. y1 = . c3 .4 6( y1 − 2 y2 + y3 ) (c) This solution corresponds to K = 262 = 2 ⋅ 131. y = 422 and . the second 1 1 has 37 .71736 x6 = 1.74712.6 is thus a3 = ⎛1 ⎞ (c) The system is A = B + ⎜ C . Using (14) to solve for a3 . 14 minae of tin.3295 ⎥ ⎥=⎢ ⎢ ⎥⎢ ⎥.5.40421. Since sin(.8.SSM: Elementary Linear Algebra Section 10. 2 2 S ( x) = −. y3 = .12643(.4.20211( x − .38942 x4 = . and 2 2 1 5 minae of iron.1880 = −2.5. with 3 ⎝ ⎠ ⎝3⎠ solution A = 45. and C = 22.4) 2 + . B = 37. T = 14. Exercise Set 10. From (19) and (20) we have M1 = M 2 = −.3295 = −3.74712. (a) The system is x + y = 1000.38942. (a) Given that the points lie on a single cubic curve.38942. and the third has 22 .5) = S(.2 and x1 = 0.6. the percentage error is zero. Section 10.47943 to five decimal places. The specific interval .5. y5 = . 217 .4)3 − .12643( x − .4 ≤ x ≤ . M 4 = −.00000 x2 = .5. y4 = .20211(.84147 Then 3.47943. The first person has 45 minae. G + I = ⎜ ⎟ 60.3750 ⎥ ⎣⎢0 0 1 5 ⎦⎥ ⎢⎣ M 5 ⎥⎦ ⎣⎢ −4.92158(. 2 1 1 9 minae of brass. and I = 5.4 ≤ x ≤ . The interpolating parabolic runout spline for . ⎛1⎞ ⎛1⎞ ⎜ ⎟ x − ⎜ ⎟ y = 10. ⎢0 1 4 1 ⎥ ⎢ M 4 ⎥ ⎢ −3.1)3 − .5) = . M 3 = −.19867 x3 = . M 6 = M 5 = −.2915 h2 and the linear system (21) for the parabolic runout spline becomes ⎡5 1 0 0 ⎤ ⎡ M 2 ⎤ ⎡ −1.15676.5. with solution x = 577 ⎝4⎠ ⎝5⎠ 7 2 and .92158 c3 = h 6 d3 = y3 = .5.12643 6h M b3 = 3 = −. (a) Set h = .4) + .20211 2 ( y4 − y3 ) ( M 4 + 2M 3 )h − = . The crown 1 was made with 30 minae of gold.55592. The legitimate son 9 9 7 receives 577 staters. C = ⎜ ⎟ B + 10.56464 x5 = .1) + . h2 6( y2 − 2 y3 + y4 ) 7.4 2.15676.

00000024.0000126. bi’s. 6h (M 5 − M 4 ) a4 = = −. Solving for the ai’s.00000022. y3 = 27 x4 = 3. 2 2 M3 M4 b3 = = 16.000214. b4 = = 25. 2 2 M3 M = = −. y4 = 79 x5 = 4. 2 2 b1 = ( y2 − y1 ) ( M 2 + 2M1 )h − = 5. M 3 = 32. 218 ( y2 − y1 ) ( M 2 + 2 M1 )h − = .00000042. ci’s. h 6 .0000132. For 1 ≤ x ≤ 2 we have S ( x) = S2 ( x) = 3( x − 1)3 + 7( x − 1)2 + 10( x − 1) + 7 = 3x3 − 2 x 2 + 5 x + 1. ⎥ ⎢ ⎢ ⎥⎢ ⎥ ⎣ 0 1 4 ⎦ ⎣ M 4 ⎦ ⎣ −. a3 = = 3. 6h (M 3 − M 2 ) a2 = = . and di’s from Equations (14) we have ( M 2 − M1 ) a1 = = −. ⎥ ⎢ ⎢ ⎥⎢ ⎥ ⎣0 0 6 ⎦ ⎣ M 4 ⎦ ⎣300 ⎦ Solving this system yields M 2 = 14. 6h (M 3 − M 2 ) (M 4 − M 3 ) a2 = = 3. h 6 ( y − y4 ) ( M 5 + 2 M 4 )h = 5 − = 74. For 2 ≤ x ≤ 3 we have S ( x) = S3 ( x) = 84 = 192 = 3( x − 2)3 + 16( x − 2)2 + 33( x − 2) + 27 = 3x3 − 2 x 2 + 5 x + 1. b2 = 2 = −. d 2 = y2 = 7.0000816 ⎥ . y1 = 1 x2 = 1. Thus S1 ( x) = S2 ( x) = S3 ( x) = S4 ( x). M 5 = 0. M 4 = 50.0000108. d3 = y3 = 27. h 6 c1 = c2 c3 c4 M1 M = 0.0000252. bi’s. 2 b1 = b3 b5 c1 = d1 = y1 = 1. h 6 ( y − y2 ) ( M 3 + 2M 2 )h = 3 − = 10.0000066. 6h 6h (M 5 − M 4 ) a4 = = 3. From (22) and (23) we have M 1 = 2 M 2 − M 3 = −4 M 5 = 2 M 4 − M 3 = 68. y5 = 181. ci’s. For 0 ≤ x ≤ 1 we have S ( x) = S1 ( x) = 3x3 − 2 x 2 + 5 x + 1. M 3 = −.0001116 ⎤ becomes ⎢1 4 1 ⎥ ⎢ M 3 ⎥ = ⎢ −. 6h (M 4 − M 3 ) a3 = = −. 4. Then 6( y1 − 2 y2 + y3 ) h2 6( y2 − 2 y3 + y4 ) h2 6( y3 − 2 y4 + y5 ) SSM: Elementary Linear Algebra d 4 = y4 = 79.0000054. y2 = 7 x3 = 2. and ( M 2 − M1 ) di’s we have a1 = = 3. For 3 ≤ x ≤ 4 we have S ( x) = S 4 ( x) = 300 h2 and the linear system (24) for the cubic runout spline becomes ⎡6 0 0 ⎤ ⎡ M 2 ⎤ ⎡ 84 ⎤ ⎢1 4 1 ⎥ ⎢ M 3 ⎥ = ⎢192 ⎥ . 6h M1 M = −2. 2 2 M5 = = 0.Chapter 10: Applications of Linear Algebra (b) Set h = 1 and x1 = 0.00000004. b4 = 4 = −. h 6 ( y − y3 ) ( M 4 + 2M 3 )h = 4 − = 33. or S ( x) = 3x3 − 2 x 2 + 5 x + 1 for 0 ≤ x ≤ 4.0000636 ⎦ Solving this system yields M 2 = −. Using (14) to solve for the ai’s. 6h = 3( x − 3)3 + 25( x − 3)2 + 74( x − 3) + 79 = 3x3 − 2 x 2 + 5 x + 1. From (17) and (18) we have M1 = 0. The linear system (16) for the natural spline ⎡ 4 1 0 ⎤ ⎡ M 2 ⎤ ⎡ −. b2 = 2 = 7. M 4 = −.

SSM: Elementary Linear Algebra

Section 10.4

( y3 − y2 ) ( M 3 + 2M 2 )h
( y − y3 ) ( M 4 + 2M 3 )h

= .000088, c3 = 4

= −.000092,
h
h
6
6
( y − y4 ) ( M 5 + 2M 4 )h
c4 = 5

= −.000212,
h
6
c2 =

d1 = y1 = .99815, d 2 = y2 = .99987, d3 = y3 = .99973, d 4 = y4 = .99823.
The resulting natural spline is
⎧−.00000042( x + 10)3 + .000214( x + 10) + .99815, − 10 ≤ x ≤ 0

⎪.00000024( x)3 − .0000126( x)2 + .000088( x) + .99987, 0 ≤ x ≤ 10
S ( x) = ⎨
3
2
⎪−.00000004( x − 10) − .0000054( x − 10) − .000092( x − 10) + .99973, 10 ≤ x ≤ 20
⎪−.00000022( x − 20)3 − .0000066( x − 20)2 − .000212( x − 20) + .99823, 20 ≤ x ≤ 30.

Assuming the maximum is attained in the interval [0, 10] we set S ′( x) equal to zero in this interval:
S ′( x) = .00000072 x 2 − .0000252 x + .000088 = 0.
To three significant digits the root of this quadratic equation in the interval [0, 10] is x = 3.93, and
S(3.93) = 1.00004.
⎡6 0 0 ⎤ ⎡ M 2 ⎤ ⎡ −.0001116 ⎤
5. The linear system (24) for the cubic runout spline becomes ⎢1 4 1 ⎥ ⎢ M 3 ⎥ = ⎢ −.0000816 ⎥ .
⎥ ⎢

⎥⎢

⎣0 0 6 ⎦ ⎣ M 4 ⎦ ⎣ −.0000636 ⎦
Solving this system yields M 2 = −.0000186, M 3 = −.0000131, M 4 = −.0000106.
From (22) and (23) we have M1 = 2M 2 − M 3 = −.0000241, M 5 = 2 M 4 − M 3 = −.0000081.
( M 2 − M1 )
= .00000009,
6h
(M 3 − M 2 )
(M 4 − M 3 )
(M 5 − M 4 )
a2 =
= .00000009, a3 =
= .00000004, a4 =
= .00000004.
6h
6h
6h

Solving for the ai’s, bi’s, ci’s, and di’s from Equations (14) we have a1 =

b1 =

M
M1
M
M
= −.0000121, b2 = 2 = −.0000093, b3 = 3 = −.0000066, b4 = 4 = −.0000053,
2
2
2
2

( y − y2 ) ( M 3 + 2M 2 )h
( y2 − y1 ) ( M 2 + 2 M1 )h

= .000282, c2 = 3

= .000070,
h
6
h
6
( y − y3 ) ( M 4 + 2M 3 )h
( y − y4 ) ( M 5 + 2M 4 )h
c3 = 4

= −.000087, c4 = 5

= −.000207,
h
6
h
6

c1 =

d1 = y1 = .99815, d 2 = y2 = .99987, d3 = y3 = .99973, d 4 = y4 = .99823.
The resulting cubic runout spline is
⎧.00000009( x + 10)3 − .0000121( x + 10) 2 + .000282( x + 10) + .99815, − 10 ≤ x ≤ 0

⎪.00000009( x)3 − .0000093( x) 2 + .000070( x) + .99987, 0 ≤ x ≤ 10
S ( x) = ⎨
3
2
⎪.00000004( x − 10) − .0000066( x − 10) − .000087( x − 10) + .99973, 10 ≤ x ≤ 20
⎪.00000004( x − 20)3 − .0000053( x − 20)2 − .000207( x − 20) + .99823, 20 ≤ x ≤ 30.

Assuming the maximum is attained in the interval [0, 10], we set S ′( x) equal to zero in this interval:
S ′( x) = .00000027 x 2 − .0000186 x + .000070 = 0.
To three significant digits the root of this quadratic equation in the interval [0, 10] is 4.00 and S(4.00) = 1.00001.

219

Chapter 10: Applications of Linear Algebra

SSM: Elementary Linear Algebra
For 1 ≤ x ≤ 1.5, we have
S ( x) = S2 ( x) = −2( x − 1) + 0 = −2 x + 2.
The resulting natural spline is
⎧ − 2 x + 2 0 .5 ≤ x ≤ 1
S ( x) = ⎨
.
⎩−2 x + 2 1 ≤ x ≤ 1.5

6. (a) Set h = .5 and
x1 = 0, y1 = 0

x2 = .5, y2 = 1
x3 = 1, y3 = 0
For a natural spline with n = 3,
M1 = M 3 = 0 and
6( y1 − 2 y2 + y3 )
h2
M 2 = −12.

(c) The three data points are collinear, so the
spline is just the line the points lie on.

= −48 = 4M 2 . Thus

7. (b) Equations (15) together with the three
equations in part (a) of the exercise
statement give
6( yn −1 − 2 y1 + y2 )
4M1 + M 2 + M n −1 =
h2
6( y1 − 2 y2 + y3 )
M1 + 4 M 2 + M 3 =
h2
6( y2 − 2 y3 + y4 )
M 2 + 4M 3 + M 4 =
h2
#
6( yn −3 − 2 yn − 2 + yn −1 )
M n −3 + 4 M n −2 + M n −1 =
h2
6( yn − 2 − 2 yn −1 + y1 )
M1 + M n −2 + 4 M n −1 =
.
h2
The linear system for M1 , M 2 , ..., M n −1 in
matrix form is
⎡ 4 1 0 0 " 0 0 0 1 ⎤ ⎡ M1 ⎤
⎢1 4 1 0 " 0 0 0 0 ⎥ ⎢ M 2 ⎥

⎥⎢
⎢0 1 4 1 " 0 0 0 0 ⎥ ⎢ M 3 ⎥
# # # # ⎥⎢ # ⎥
⎢# # # #
⎢0 0 0 0 " 0 1 4 1 ⎥ ⎢ M n −2 ⎥

⎢1 0 0 0 " 0 0 1 4 ⎥ ⎢

⎦ ⎣⎢ M n −1 ⎦⎥
⎡ yn −1 − 2 y1 + y2 ⎤

y1 − 2 y2 + y3

6 ⎢ y2 − 2 y3 + y4 ⎥
=
.

#
h2 ⎢
⎢y
− 2 yn −2 + yn −1 ⎥
⎢ n −3

⎣⎢ yn − 2 − 2 yn −1 + y1 ⎦⎥

M − M2
M 2 − M1
= −4, a2 = 3
= 4,
6h
6h
M
M
b1 = 1 = 0, b2 = 2 = −6,
2
2
y2 − y1 ( M 2 + 2M1 )h
c1 =

= 3,
h
6
y − y2 ( M 3 + 2 M 2 )h
c2 = 3

= 0, d1 = 0,
h
6
d2 = 1
For 0 ≤ x ≤ .5, we have
a1 =

S ( x) = S1 ( x) = −4 x3 + 3x.
For .5 ≤ x ≤ 1, we have
S ( x) = S2 ( x) = 4( x − .5)3 − 6( x − .5)2 + 1
= 4 x3 − 12 x 2 + 9 x − 1.
The resulting natural spline is
⎧⎪−4 x3 + 3 x
0 ≤ x ≤ 0.5
S ( x) = ⎨
.
3
2
⎪⎩4 x − 12 x + 9 x − 1 0.5 ≤ x ≤ 1
(b) Again h = .5 and
x1 = .5, y1 = 1
x2 = 1, y2 = 0
x3 = 1.5, y3 = −1

4M 2 =

6( y1 − 2 y2 + y3 )

=0
h2
Thus M1 = M 2 = M 3 = 0, hence all ai and
bi are also 0.
y − y2
y −y
= −2
c1 = 2 1 = −2, c2 = 3
h
h
d1 = y1 = 1, d 2 = y2 = 0
For .5 ≤ x ≤ 1, we have
S ( x) = S1 ( x) = −2( x − .5) + 1 = −2 x + 2.

220

SSM: Elementary Linear Algebra

Section 10.5

8. (b) Equations (15) together with the two
equations in part (a) give
6( y2 − y1 − hy1′ )
2 M1 + M 2 =
h2
6( y1 − 2 y2 + y3 )
M1 + 4 M 2 + M 3 =
h2
6( y2 − 2 y3 + y4 )
M 2 + 4M 3 + M 4 =
h2
#
6( yn − 2 − 2 yn −1 + yn )
M n − 2 + 4 M n −1 + M n =
h2
6( yn −1 − yn + hyn′ )
M n −1 + 2M n =
.
h2
This linear system for M1 , M 2 , ..., M n in
matrix form is
⎡2 1 0 0 " 0 0 0 0 ⎤ ⎡ M ⎤
⎢1 4 1 0 " 0 0 0 0 ⎥ ⎢ 1 ⎥

⎥ ⎢ M2 ⎥
⎢0 1 4 1 " 0 0 0 0 ⎥ ⎢ M 3 ⎥
⎢0 0 1 4 " 0 0 0 0 ⎥ ⎢
# ⎥
⎢# # # #
# # # # ⎥⎢

⎢0 0 0 0 " 1 4 1 0 ⎥ ⎢ M n− 2 ⎥

⎥ ⎢ M n −1 ⎥
⎢0 0 0 0 " 0 1 4 1 ⎥ ⎢

⎢⎣0 0 0 0 " 0 0 1 2 ⎥⎦ ⎣ M n ⎦
⎡ −hy1′ − y1 + y2 ⎤
⎢ y − 2y + y ⎥
2
3 ⎥
⎢ 1
6 ⎢ y2 − 2 y3 + y4 ⎥
.
=

#
h2 ⎢
⎢y
− 2 yn −1 + yn ⎥

⎢ n−2
⎣⎢ yn −1 − yn + hyn′ ⎦⎥

.6q1 − .5q2 = 0, or q1 =

5
q2 . Solutions are
6

⎡5⎤
thus of the form q = s ⎢ 6 ⎥ . Set
⎢⎣ 1 ⎥⎦

s=

1
5
6

+1

=

⎡5⎤
6
to obtain q = ⎢ 11 ⎥ .
6⎥
11
⎢⎣ 11

⎡. 7 ⎤
⎡.23⎤
2. (a) x(1) = Px(0) = ⎢.2 ⎥ ; likewise x( 2) = ⎢.52 ⎥
⎢ ⎥
⎢ ⎥
⎣.25⎦
⎣ .1⎦
⎡.273⎤
and x(3) = ⎢.396⎥ .

⎣ .331⎦
(b) P is regular because all of its entries are
positive. To solve (I − P)q = 0, i.e.
⎡ .8 −.1 −.7 ⎤ ⎡ q1 ⎤ ⎡ 0 ⎤
⎢ −.6 .6 −.2 ⎥ ⎢ q2 ⎥ = ⎢ 0 ⎥ , reduce the

⎥⎢ ⎥ ⎢ ⎥
⎣ −.2 −.5 .9 ⎦ ⎣ q3 ⎦ ⎣ 0 ⎦
coefficient matrix to row-echelon form:
5 −9 ⎤
⎡ 8 −1 −7 ⎤
⎡2
⎢ −6 6 −2⎥ → ⎢ 0 −21 29 ⎥

0 0⎦
⎣ −2 −5 9⎦
⎣0
⎡ 1 0 − 22 ⎤
21 ⎥

→ ⎢ 0 1 − 29 ⎥ .
21

0⎦
⎣0 0
This yields solutions (setting q3 = s) of the
⎡ 22 ⎤
⎢ 21 ⎥
form ⎢ 29 ⎥ s.
⎢ 21 ⎥
⎣1⎦
To obtain a probability vector, take
⎡ 22 ⎤
⎢ 72 ⎥
1
21
s=
=
yielding q = ⎢ 29
⎥.
72
22 + 29 + 1 72,

21 21
21
⎢⎣ 72 ⎥⎦

Section 10.5
Exercise Set 10.5

⎡. 4 ⎤
⎡.46 ⎤
1. (a) x(1) = Px(0) = ⎢ ⎥ , x( 2) = Px(1) = ⎢ ⎥ .
⎣. 6 ⎦
⎣.54 ⎦
Continuing in this manner yields
⎡.454⎤
.
x(3) = ⎢
⎣.546⎥⎦
⎡.4546 ⎤
⎡.45454⎤
.
x ( 4) = ⎢
and x(5) = ⎢

.
5454

⎣.54546⎥⎦

3. (a) Solve (I − P)q = 0, i.e.,
⎡ 2 − 3⎤ ⎡q ⎤
4 ⎥ 1 = ⎡0⎤ .
⎢ 3
3 ⎥ ⎢ q ⎥ ⎢⎣ 0 ⎥⎦
2
⎢− 3
⎣ 2⎦
4⎦

The only independent equation is

(b) P is regular because all of the entries of P
are positive. Its steady-state vector q solves
⎡ .6 −.5⎤ ⎡ q1 ⎤ ⎡0 ⎤
.
(I − P)q = 0; that is, ⎢
⎢ ⎥=
⎣ −.6 .5⎥⎦ ⎣ q2 ⎦ ⎢⎣0 ⎥⎦
This yields one independent equation,

221

Chapter 10: Applications of Linear Algebra

SSM: Elementary Linear Algebra

⎡ 1 n

⎢ 2 x1

n
P x=⎢
⎥.
n
x1 + x2 ⎥
⎢ 1 − 12

Already true for n = 1, 2. If true for n − 1,

()

⎡9 ⎤
2
3
q1 = q2 , yielding q = ⎢ 8 ⎥ s. Setting
3
4
⎢⎣ 1 ⎥⎦

(

⎡9⎤
8
s=
yields q = ⎢ 17 ⎥ .
8⎥
17
⎢ 17
⎣ ⎦

()

(

() )
( )( )
( () ()
()
( ())

⎡ 26 ⎤
⎡ 26 ⎤
19
to get q = ⎢ 45 ⎥ .
q = ⎢ 19 ⎥ s. Set s =
45
⎢ 19

⎢⎣ 1 ⎥⎦
⎣ 45 ⎦
⎡ 2 −1
0⎤ ⎡ q ⎤
2
⎡0 ⎤
⎢ 3
⎥ 1
1 − 14 ⎥ ⎢ q2 ⎥ = ⎢0 ⎥ by
(c) Again, solve ⎢ − 13
⎢ ⎥ ⎢ ⎥
⎢ 1
1
1 ⎥ ⎣ q3 ⎦ ⎣ 0 ⎦
⎢⎣ − 3 − 2

4⎦
reducing the coefficient matrix to row⎡1 0 − 1⎤
4⎥

echelon form: ⎢0 1 − 1 ⎥ yielding
3

0⎦
⎣0 0
⎡1⎤
⎢4⎥
solutions of the form q = ⎢ 1 ⎥ s.
⎢3⎥
⎣1 ⎦

⎡1⎤
⎢k ⎥
⎢1⎥
5. Let q = ⎢ k ⎥ . Then
⎢# ⎥
⎢1⎥
⎣k ⎦
1
1 k
1
pij = ∑ pij = ,
k j =1
k
j =1
j =1 k
since the row sums of P are 1. Thus ( Pq)i = qi
for all i.
( Pq ) i =

( n)
4. (a) Prove by induction that p12

true for n = 1. If true for n − 1, we have
P n = P n −1 P, so

( n)
( n −1)
( n −1)
p12
= p11
p12 + p12
p22 . But

p12 =

= 0 so

)

(c) The Theorem says that the entries of the
steady state vector should be positive; they
⎡0 ⎤
are not for ⎢ ⎥ .
⎣1 ⎦

⎡3⎤
⎢ 19 ⎥
12
4 ⎥.
to get q = ⎢ 19
Set s =
19
⎢ 12 ⎥
⎢⎣ 19 ⎥⎦

( n)
p12

)

then P n = P( P n −1x)
⎡ 1 n −1

x1
⎢ 2

= P⎢

n −1
x1 + x2 ⎥
⎢ 1 − 12

⎡ 1 1 n −1

x1
⎢ 2 2

=⎢

n −1
n
x1 + x2 ⎥
+ 12
⎢ 1 − 12

n
⎡ 1

⎢ 2 x1

=⎢
⎥.
n
1
x
x
+
⎢ 1− 2
1
2⎥

⎡ 0 ⎤ ⎡0 ⎤
if x is a
Since lim P n x = ⎢
⎥=
n→∞
⎣ x1 + x2 ⎦ ⎢⎣1 ⎥⎦
state vector.

⎡ .19 −.26 ⎤ ⎡ q1 ⎤ ⎡0 ⎤
(b) As in (a), solve ⎢
⎢ ⎥=
⎣ −.19 .26 ⎥⎦ ⎣ q2 ⎦ ⎢⎣0 ⎥⎦
i.e., .19q1 = .26q2 . Solutions have the form

( n −1)
p12

()

k

k

∑ pij q j = ∑

6. Since P has zeros entries, consider
⎡1 1 1⎤
⎢2 4 4⎥
P 2 = ⎢ 14 12 14 ⎥ , so P is regular. Note that all
⎢1 1 1⎥
⎢⎣ 4 4 2 ⎥⎦
the rows of P sum to 1. Since P is 3 × 3,
⎡1 ⎤
⎢3⎥
Exercise 5 implies q = ⎢ 13 ⎥ .
⎢1 ⎥
⎢⎣ 3 ⎥⎦

= 0 + 0 = 0. Thus,

no power of P can have all positive entries,
so P is not regular.

⎡ 1 x1

⎡x ⎤
⎥,
(b) If x = ⎢ 1 ⎥ , Px = ⎢ 2
1
⎢⎣ 2 x1 + x2 ⎥⎦
⎣ x2 ⎦
⎡ 1 x1

⎥ etc. We use
P2x = ⎢ 4
⎢⎣ 14 x1 + 12 x1 + x2 ⎥⎦
induction to show
222

SSM: Elementary Linear Algebra

7. Let x = [ x1

Section 10.6

Section 10.6

x2 ]T be the state vector, with

x1 = probability that John is happy and

Exercise Set 10.6

x2 = probability that John is sad. The transition

1. Note that the matrix has the same number of
rows and columns as the graph has vertices, and
that ones in the matrix correspond to arrows in
the graph.

⎡4 2⎤
matrix P will be P = ⎢ 5 3 ⎥ since the columns
⎢⎣ 15 13 ⎥⎦
must sum to one. We find the steady state vector
⎡ 1 − 2 ⎤ ⎡ q ⎤ ⎡0⎤
3⎥ 1 =
for P by solving ⎢ 5
, i.e.,
2 ⎥ ⎢ q ⎥ ⎢⎣ 0 ⎥⎦
⎢⎣ − 15
⎣ 2⎦
3⎦
⎡ 10 ⎤
1
2
3
q1 = q2 , so q = ⎢ 3 ⎥ s. Let s =
and get
5
3
13
⎣⎢ 1 ⎦⎥
⎡ 10 ⎤
10
is the probability that John will
q = ⎢ 13 ⎥ , so
3
13
⎢ 13 ⎥
⎣ ⎦
be happy on a given day.
8. The state vector x = [ x1 x2 x3 ]T will
represent the proportion of the population living
in regions 1, 2, and 3, respectively. In the
transition matrix, pij will represent the
proportion of the people in region j who move to
⎡.90 .15 .10 ⎤
region i, yielding P = ⎢.05 .75 .05⎥ .

⎣.05 .10 .85⎦

⎡0
⎢1
(a) ⎢
⎢1
⎣⎢0

0
0
1
0

0
1
0
0

1⎤
1⎥

1⎥
0⎦⎥

⎡0
⎢0

(b) ⎢1
⎢0
⎢⎣0

1
0
0
0
0

1
0
0
1
1

0
0
1
0
0

0⎤
1⎥

0⎥
0⎥
0 ⎥⎦

⎡0
⎢1

0
(c) ⎢
⎢0
⎢0
⎢0

1
0
1
0
0
0

0
0
0
0
0
1

1
0
1
0
0
0

0
0
1
0
0
1

0⎤
0⎥

1⎥
1⎥
1⎥
0 ⎥⎦

2. See the remark in problem 1; we obtain

⎡ .10 −.15 −.10⎤ ⎡ q1 ⎤ ⎡0 ⎤
⎢ −.05 .25 −.05⎥ ⎢ q2 ⎥ = ⎢0 ⎥

⎥⎢ ⎥ ⎢ ⎥
⎣ −.05 −.10 .15⎦ ⎣ q3 ⎦ ⎣0 ⎦

(a)

P1

P2

⎡1 0 − 13 ⎤
7⎥

First reduce to row echelon form ⎢0 1 − 4 ⎥ ,
7

⎣0 0 0 ⎦
⎡ 13 ⎤
⎢7⎥
7
and get
yielding q = ⎢ 4 ⎥ s. Set s =
24
⎢7⎥
⎣1⎦
q=

⎡ 13 ⎤
⎢ 24 ⎥
4 ⎥,
⎢ 24
⎢7⎥
⎢⎣ 24 ⎥⎦

i.e., in the long run

(b)

P4

P3

P1

13 ⎛
1 ⎞
⎜ or 54 % ⎟
24 ⎝
6 ⎠
P5

of the people reside in region 1,
in region 2, and

P4

4
24

2 ⎞

⎜ or 16 % ⎟
3 ⎠

P3

7 ⎛
1 ⎞
⎜ or 29 % ⎟ in region 3.
24 ⎝
6 ⎠

223

P2

a vertex must have “two-way” connections with at least two other vertices.6. P6 . P4 from P1 to P2 . leaving ( 2) (3) = 2 and m12 = 3 meaning there So m12 {P3 . containing all the other possibilities except {P4 . there still satisfy the conditions. so {P1 .e. P6}. P5 . P5 . P8}. Inspection 2-step: P1 → P4 → P2 and P1 → P3 → P2 shows that each of the sets {P2 . P6 } satisfy conditions (i) and (ii) in the definition of a clique. These are: added. P5} as the only possible clique. P8} P1 → P3 → P4 → P2 .1. P1 i. P4 . This consideration eliminates P1 and P2 . P6 . so there is one 1-step connection 5. P4 P2 (c) The ij entry of M T M is the number of family members who influence both member i and member j. {P2 . P6 . the sum of the squares of the i =1 entries in column k of M. P8} is a clique. Thus. (b) Not only must a clique vertex have two-way connections to at least two other vertices. But note that P8 can be added to the first set and we P1 → P4 → P3 → P2 . P3 (b) m12 = 1. (a) M M = ⎢0 ⎢0 ⎢⎣0 P3 P6 P5 3. one 2-step and two 3-step connections from P1 to P4 . but the vertices to which it is connected must share a two-way connection. These entries are 1 if family member i influences member k and 0 otherwise. P3 and P7 from being in a clique. ( 2) (3) (c) Since m14 = 1. P4 . P6 }. m14 = 1 and m14 = 2. {P4 . P2 .. P8}. 2⎥ 1 ⎦⎥ 2 could not be in a clique. so {P2 . are two 2-step and three 3-step connections from P1 to P2 by Theorem 10. (a) Note that to be contained in a clique. 1-step: P1 → P4 2-step: P1 → P3 → P4 3-step: P1 → P2 → P1 → P4 and P1 → P4 → P3 → P4 . 224 . P3} is the only possible clique. (a) 0 1 0 0 0 0 0 1 1 0 0 0 1 2 1 0⎤ 0⎥ ⎥ 0⎥ 1⎥ 2⎥⎦ (b) The kth diagonal entry of M T M is P4 5 ∑ mik 2 . P5 may not be are one 1-step. and and {P4 . Inspection shows that it is indeed a clique. {P2 . These are: 1-step: P1 → P2 (c) The above considerations eliminate P1 . which is also a clique. ⎡1 ⎢0 M =⎢ ⎢1 ⎢⎣1 2 1 1 1 1 1 1 0 1⎤ 1⎥ ⎥ and 0⎥ 1 ⎥⎦ ⎡2 ⎢1 M3 = ⎢ ⎢1 ⎣⎢1 3 2 2 2 2 1 1 2 2⎤ 1⎥ ⎥. P4 . P4 . 3-step: P1 → P2 → P1 → P2 . Inspection shows that this is indeed a clique.Chapter 10: Applications of Linear Algebra (c) P1 SSM: Elementary Linear Algebra P2 ⎡1 ⎢0 ⎢ T 4.

. Since a clique contains at least three vertices. ⎢ 3 1 1 0 3⎥ ⎢⎣1 1 3 3 0 ⎥⎦ 8. and of E is 6. and D last. Associating vertex P1 with team A. and of P4 is 2. ⎢1 0 0 0 1 ⎥ ⎢⎣0 0 1 1 0 ⎥⎦ ⎡0 3 1 3 1 ⎤ ⎢3 0 3 1 1 ⎥ ⎥ 3 ⎢ S = ⎢1 3 0 1 3 ⎥ . P2 with B. and P6 are members of a clique. the expected payoff of the game is ⎡1⎤ ⎢4⎥ 1⎤ ⎢ 1 ⎥ ⎡ −4 6 −4 pAq = ⎡⎣ 12 0 12 ⎤⎦ ⎢ 5 −7 3 8⎥ ⎢ 14 ⎥ ⎢ ⎥ ⎣ −8 0 6 −2 ⎦ ⎢ 4 ⎥ ⎢1⎥ ⎣⎢ 4 ⎦⎥ 5 =− .7 6. the game results yield the following dominance-directed graph: P1 P2 P3 in the graph represented by M. (a) With the given M we get ⎡0 1 0 1 0 ⎤ ⎢1 0 1 0 0 ⎥ ⎢ ⎥ S = ⎢0 1 0 0 1 ⎥ . we get that the power of A is 8. we must have {P3 . of D is 3. Since sii(3) = 0 for all i. ⎢7 1 8 2 7 5 ⎥ ⎢0 6 1 7 0 2⎥ ⎢2 3 4 5 2 2⎥ ⎣ ⎦ The elements along the main diagonal tell us that only P3 . P5 with E. P6 } as the only clique. 225 . the power of P2 is 3.. ⎢0 1 1 0 1 ⎥ ⎢⎣1 2 1 2 0 ⎥⎦ Summing the rows. 1 0 1 0 1 0 0 1 0 1 0 1 1 0 1 0 1 1 0 1 0 1 0 0 0⎤ 0⎥ ⎥ 1⎥ . there are no cliques ⎡0 ⎢1 ⎢ 0 (b) Here S = ⎢ ⎢1 ⎢0 ⎢0 ⎣ P5 Section 10.7 0 1 0 0 1 1 0 0 1. Thus ranking in decreasing order we get A in first place. ⎢0 0 1 0 1 ⎥ ⎢⎣0 2 1 1 0 ⎥⎦ ⎡0 2 2 2 2 ⎤ ⎢1 0 1 2 2 ⎥ ⎥ 2 ⎢ M + M = ⎢1 1 0 2 1 ⎥ . of C is 5. P4 . ⎡0 ⎢1 7. 1⎥ 0⎥ 0 ⎥⎦ ⎡0 6 1 7 0 2⎤ ⎢6 0 7 1 6 3⎥ ⎢ ⎥ 1 7 2 8 1 4⎥ S3 = ⎢ . of P3 is 4. 1⎥ 0 ⎥⎦ By summing the rows of M + M 2 . of B is 6. 8 1⎤ 1⎥ ⎥ and 0⎥ 0 ⎥⎦ 2⎤ 1⎥ ⎥..SSM: Elementary Linear Algebra Section 10. (a) From Equation (2). P4 . M = ⎢ ⎢0 ⎣⎢0 0 0 1 1 1 0 0 0 ⎡0 2 ⎢0 Then M = ⎢ ⎢1 ⎢⎣1 ⎡0 ⎢1 M +M =⎢ ⎢1 ⎢⎣1 2 1⎤ 0⎥ ⎥ 1⎥ 0⎦⎥ 2 0 1 0 2 0 2 1 P4 which has vertex matrix ⎡0 1 1 1 0 ⎤ ⎢0 0 1 0 1 ⎥ ⎢ ⎥ M = ⎢0 0 0 1 1 ⎥ .7 Exercise Set 10. we get that the power of P1 is 2 + 1 + 2 = 5. B and E tie for second place. . ⎢0 1 0 0 0 ⎥ ⎢⎣1 0 0 1 0 ⎥⎦ ⎡0 1 1 1 2⎤ ⎢1 0 0 2 1 ⎥ ⎥ 2 ⎢ Then M = ⎢1 1 0 1 0 ⎥ . C in fourth place.

Chapter 10: Applications of Linear Algebra

(b) If player R uses strategy [ p1

p2

SSM: Elementary Linear Algebra

p3 ]

(c) Here, a32 is a saddle point, so optimal
⎡0⎤
strategies are p* = [0 0 1], q* = ⎢1 ⎥ and
⎢ ⎥
⎣0⎦

⎡1⎤
⎢4⎥
⎢1⎥
against player C’s strategy ⎢ 14 ⎥ his payoff
⎢4⎥
⎢1⎥
⎣⎢ 4 ⎦⎥

v = a32 = 2.
(d) Here, a21 is a saddle point, so

⎛ 1⎞
⎛9⎞
will be pAq = ⎜ − ⎟ p1 + ⎜ ⎟ p 2 − p3 . Since
⎝ 4⎠
⎝4⎠
p1, p2 , and p3 are nonnegative and add
up to 1, this is a weighted average of the
1 9
numbers − , , and −1. Clearly this is the
4 4
largest if p1 = p3 = 0 and p2 = 1; that is,

⎡1 ⎤
p* = [0 1 0 0], q* = ⎢ 0 ⎥ and
⎢ ⎥
⎣0⎦

v = a21 = −2.
4. (a) Calling the matrix A, the formulas of
Theorem 10.7.2 yield p* = ⎡ 85 83 ⎤ ,

1
⎡ ⎤
27
q* = ⎢ 8 ⎥ , v =
8
⎢⎣ 78 ⎥⎦

p = [0 1 0].
(c) As in (b), if player C uses
[ q1 q2

q3

q4 ]T against ⎡⎣ 12

0

1⎤,
2⎦

40 20 ⎤ = ⎡ 2 1 ⎤ ,
(b) As in (a), p* = ⎡ 60
60 ⎦ ⎣ 3 3 ⎦

10
⎡ ⎤ ⎡1⎤
1400 70
q* = ⎢ 60 ⎥ = ⎢ 6 ⎥ , v =
=
(Again, A
50
5
60
3
⎢ ⎥ ⎢ ⎥
⎣ 60 ⎦ ⎣ 6 ⎦

1
we get pAq = −6q1 + 3q2 + q3 − q4 .
2
Clearly this is minimized over all strategies
by setting q1 = 1 and q2 = q3 = q4 = 0.
That is q = [1 0 0 0]T .
2. As per the hint, we will construct a 3 × 3 matrix
with two saddle points, say a11 = a33 = 1. Such a

(c) For this matrix, a11 is a saddle point, so
⎡1 ⎤
p* = [1 0], q* = ⎢ ⎥ , and v = a11 = 3.
⎣0⎦

⎡1 2 1 ⎤
matrix is A = ⎢ 0 7 0 ⎥ . Note that

⎣1 2 1 ⎦
a13 = a31 = 1 are also saddle points.

(d) This matrix has no saddle points, so, as in
(a), p* = ⎡ −−53 −−25 ⎤ = ⎡ 35 25 ⎤ ,

⎦ ⎣

3
3
⎡ ⎤ ⎡ ⎤
−19 19
q* = ⎢ −5 ⎥ = ⎢ 5 ⎥ , and v =
= .
−5
5
⎢⎣ −−25 ⎥⎦ ⎢⎣ 25 ⎥⎦

3. (a) Calling the matrix A, we see a22 is a saddle
point, so the optimal strategies are pure,
⎡0⎤
namely: p* = [0 1], q* = ⎢ ⎥ , the value of
⎣1 ⎦
the game is v = a22 = 3.

(e) Again, A has no saddle points, so as in (a),
⎡1⎤
3 10 ⎤ , q* = ⎢ 13 ⎥ , and v = −29 .
p* = ⎡ 13
13 ⎦

13
⎢⎣ 12

13 ⎦

(b) As in (a), a21 is a saddle point, so optimal
⎡1 ⎤
strategies are p* = [0 1 0], q* = ⎢ ⎥ , the
⎣0⎦
value of the game is v = a21 = 2.

5. Let
a11 = payoff to R if the black ace and black two
are played = 3.
a12 = payoff to R if the black ace and red three
are played = −4.
a21 = payoff to R if the red four and black two
are played = −6.
a22 = payoff to R if the red four and red three
226

SSM: Elementary Linear Algebra

Section 10.8

⎡ 1 0 − 78 ⎤ ⎡ p ⎤
⎡0⎤
1
79 ⎥

which reduces to ⎢0 1 − 54 ⎥ ⎢ p2 ⎥ = ⎢ 0 ⎥ .
⎥ ⎢ ⎥
79 ⎢

0
0 ⎦ ⎣ p3 ⎦ ⎣ ⎦
⎣0 0
⎡ 78 ⎤
⎢ 79 ⎥
Solutions are of the form p = ⎢ 54 ⎥ . Let
⎢ 79 ⎥
⎣1⎦
⎡ 78 ⎤
s = 79 to obtain p = ⎢54 ⎥ .
⎢ ⎥
⎣79 ⎦

are played = 7.
So, the payoff matrix for the game is
⎡ 3 −4 ⎤
A=⎢
.
⎣ −6 7 ⎥⎦
A has no saddle points, so from Theorem 10.7.2,
⎡ 11 ⎤
7 ⎤ , q* = ⎢ 20 ⎥ ; that is, player R
p* = ⎡ 13
9 ⎥
⎣ 20 20 ⎦
⎢⎣ 20

should play the black ace 65 percent of the time,
and player C should play the black two
55 percent of the time. The value of the game is
3
− , that is, player C can expect to collect on
20
the average 15 cents per game.

2. (a) By Corollary 10.8.4, this matrix is
productive, since each of its row sums is .9.

Section 10.8

(b) By Corollary 10.8.5, this matrix is
productive, since each of its column sums is
less than one.

Exercise Set 10.8
1. (a) Calling the given matrix E, we need to solve
⎡ 1 − 1 ⎤ ⎡ p ⎤ ⎡0 ⎤
3⎥
1 =
( I − E )p = ⎢ 2
.
1 ⎥ ⎢ p ⎥ ⎣⎢ 0 ⎦⎥
⎢⎣ − 12

2⎦
3⎦
⎡1⎤
1
1
This yields p1 = p2 , that is, p = s ⎢ 3 ⎥ .
2
3
⎣2⎦
2
⎡ ⎤
Set s = 2 and get p = ⎢ ⎥ .
⎣3⎦

⎡2⎤
⎡1.9 ⎤
(c) Try x = ⎢ 1 ⎥ . Then Cx = ⎢ .9 ⎥ , i.e., x > Cx,
⎢ ⎥
⎢ ⎥
⎣1 ⎦
⎣ .9 ⎦
so this matrix is productive by Theorem
10.8.3.
3. Theorem 10.8.2 says there will be one linearly
independent price vector for the matrix E if some
positive power of E is positive. Since E is not

positive, try E 2 .
⎡.2 .34 .1⎤
E 2 = ⎢.2 .54 .6⎥ > 0

⎣.6 .12 .3⎦

(b) As in (a), solve
⎡ 1
0 − 12 ⎤ ⎡ p ⎤
⎡0⎤
⎢ 2
⎥ 1
1 − 12 ⎥ ⎢ p2 ⎥ = ⎢ 0 ⎥ .
( I − E )p = ⎢ − 13
⎢ 1
⎥⎢ ⎥ ⎢ ⎥
1⎥ ⎣ p3 ⎦ ⎣ 0 ⎦
⎢⎣ − 6 −1

In row-echelon form, this reduces to
⎡ 1 0 −1⎤ ⎡ p1 ⎤ ⎡ 0 ⎤
⎢0 1 − 5 ⎥ ⎢ p ⎥ = ⎢ 0 ⎥ .

6⎥ ⎢ 2⎥ ⎢ ⎥
⎢⎣0 0
0 ⎥⎦ ⎣ p3 ⎦ ⎣ 0 ⎦

4. The exchange matrix for this arrangement (using
⎡1 1 1⎤
⎢2 3 4⎥
A, B, and C in that order) is ⎢ 13 13 14 ⎥ .
⎢1 1 1⎥
⎢⎣ 6 3 2 ⎥⎦
For equilibrium, we must solve (I − E)p = 0.
⎡ 1 − 1 − 1⎤
3
4 ⎥ ⎡ p1 ⎤ ⎡ 0 ⎤
⎢ 2
1
2
1
That is ⎢ − 3
− 4 ⎥ ⎢ p2 ⎥ = ⎢0 ⎥ .
3
⎢ ⎥ ⎢ ⎥
⎢ 1
1
1 ⎥ ⎣ p3 ⎦ ⎣ 0 ⎦
⎢⎣ − 6 − 3

2⎦
Row reduction yields solutions of the form
⎡ 18 ⎤
⎢ 16 ⎥
1600
p = ⎢ 15 ⎥ s. Set s =
and obtain
15
⎢ 16 ⎥
⎣1⎦

Solutions of this system have the form
⎡1⎤
⎡6 ⎤
p = s ⎢⎢ 56 ⎥⎥ . Set s = 6 and get p = ⎢5 ⎥ .
⎢ ⎥
⎢⎣ 1 ⎥⎦
⎣6 ⎦

(c) As in (a), solve
⎡ .65 −.50 −.30 ⎤ ⎡ p1 ⎤ ⎡ 0⎤
( I − E )p = ⎢ −.25 .80 −.30 ⎥ ⎢ p2 ⎥ = ⎢ 0⎥ ,

⎥⎢ ⎥ ⎢ ⎥
⎣ −.40 −.30 .60 ⎦ ⎣ p3 ⎦ ⎣ 0⎦

227

Chapter 10: Applications of Linear Algebra

SSM: Elementary Linear Algebra

⎡120 ⎤
p = ⎢100 ⎥ ; i.e., the price of tomatoes was

⎣106.67 ⎦
\$120, corn was \$100, and lettuce was \$106.67.

(b) The increase in value is the second column
⎡542 ⎤
1 ⎢
of ( I − C )−1,
690⎥ . Thus the value of
503 ⎢170 ⎥

the coal-mining operation must increase by
542
.
503

5. Taking the CE, EE, and ME in that order, we
form the consumption matrix C, where cij = the

amount (per consulting dollar) of the i-th
engineer’s services purchased by the j-th
⎡ 0 .2 .3 ⎤
engineer. Thus, C = ⎢.1 0 .4 ⎥ .

⎣.3 .4 0 ⎦

n

7. The i-th column sum of E is

∑ e ji ,

and the

j =1

elements of the i-th column of I − E are the
negatives of the elements of E, except for the
ii-th, which is 1 − eii . So, the i-th column sum of

We want to solve (I − C)x = d, where d is the
demand vector, i.e.
⎡ 1 −.2 −.3⎤ ⎡ x1 ⎤ ⎡500 ⎤
⎢ −.1 1 −.4 ⎥ ⎢ x2 ⎥ = ⎢700 ⎥ .

⎥⎢ ⎥ ⎢

⎣ −.3 −.4 1 ⎦ ⎣ x3 ⎦ ⎣600 ⎦

n

I − E is 1 − ∑ e ji = 1 − 1 = 0. Now, ( I − E )T has
j =1

zero row sums, so the vector x = [1 1 " 1]T

In row-echelon form this reduces to
⎡1 −.2 −.3
⎤ ⎡ x1 ⎤ ⎡ 500 ⎤
⎢0 1
−.43877 ⎥ ⎢ x2 ⎥ = ⎢ 765.31⎥ .

⎥⎢ ⎥ ⎢

1
⎣0 0
⎦ ⎣ x3 ⎦ ⎣1556.19⎦
Back-substitution yields the solution
⎡1256.48 ⎤
x = ⎢1448.12 ⎥ .

⎣1556.19 ⎦

solves ( I − E )T x = 0. This implies
det( I − E )T = 0. But det( I − E )T = det( I − E ),
so (I − E)p = 0 must have nontrivial (i.e.,
nonzero) solutions.

8. Let C be a consumption matrix whose column
sums are less than one; then the row sums of C T
are less than one. By Corollary 10.8.4, C T is

6. (a) The solution of the system (I − C)x = d is

productive so ( I − C T )−1 ≥ 0. But

−1

x = ( I − C ) d. The effect of increasing the
demand di for the ith industry by one unit

( I − C )−1 = ((( I − C )T ))−1 )T

= (( I − C T )−1 )T
≥ 0.
Thus, C is productive.

⎡0 ⎤
⎢# ⎥
⎢ ⎥
⎢0 ⎥
is the same as adding ⎢1 ⎥ to d where the 1
⎢0 ⎥
⎢# ⎥
⎢ ⎥
⎢⎣0 ⎥⎦
is in the ith row. The new solution is

⎡0 ⎤ ⎞
⎡ 0⎤

⎢# ⎥ ⎟
⎢# ⎥

⎢ ⎥⎟
⎢ ⎥
0

⎢0 ⎥ ⎟
−1
−1 ⎢ ⎥
( I − C ) ⎜ d + ⎢1 ⎥ ⎟ = x + ( I − C ) ⎢ 1 ⎥

⎢0 ⎥ ⎟
⎢ 0⎥

⎢# ⎥ ⎟
⎢# ⎥

⎢ ⎥ ⎟⎟
⎢ ⎥

⎢⎣0 ⎥⎦ ⎠
⎢⎣ 0⎥⎦

which has the effect of adding the ith

Section 10.9
Exercise Set 10.9
1. Using Equation (18), we calculate
30s
= 15s
Yld 2 =
2
50s 100s
.
=
Yld3 =
7
2+ 3
2

So all the trees in the second class should be
harvested for an optimal yield (since s = 1000)
of \$15,000.

column of ( I − C ) −1 to the original solution.
228

SSM: Elementary Linear Algebra

Section 10.10

2. From the solution to Example 1, we see that for the fifth class to be harvested in the optimal case we must have
p5 s
> 14.7 s, yielding p5 > \$222.63.
( .28−1 + .31−1 + .25−1 + .23−1 )
3. Assume p2 = 1, then Yld 2 =
p3 s

(.28

−1

+ .31−1 )
p4 s

s

(.28)−1

= .28s. Thus, for all the yields to be the same we must have

= .28s

(.28−1 + .31−1 + .25−1 )
p5 s

= .28s

(.28−1 + .31−1 + .25−1 + .23−1 )
p6 s

= .28s

= .28s
(.28−1 + .31−1 + .25−1 + .23−1 + .37−1 )
Solving these successively yields p3 = 1.90, p4 = 3.02, p5 = 4.24 and p6 = 5.00. Thus the ratio
p2 : p3 : p4 : p5 : p6 = 1 : 1.90 : 3.02 : 4.24 : 5.00.
n

5. Since y is the harvest vector, N = ∑ yi is the number of trees removed from the forest. Then Equation (7) and
i =1

the first of Equations (8) yield N = g1 x1 , and from Equation (17) we obtain
N=

g1s

1+

g1
g2

+" +

g1
g k −1

=

s
1
g1

+" +

1
g k −1

.

6. Set g1 = " = g n −1 = g , and p2 = 1. Then from Equation (18), Yld 2 =
to be the same, we need to solve Yld k =

pk s

(k − 1) g1

p2 s
1
g1

= gs. Since we want all of the Yld k ’s

= gs for pk for 3 ≤ k ≤ n. Thus pk = k − 1. So the ratio

p2 : p3 : p4 : " : pn = 1 : 2 : 3 : " : (n − 1).

Section 10.10
Exercise Set 10.10
⎡0 1 1 0⎤
1. (a) Using the coordinates of the points as the columns of a matrix we obtain ⎢0 0 1 1 ⎥ .

⎣0 0 0 0⎦

229

Chapter 10: Applications of Linear Algebra

SSM: Elementary Linear Algebra

⎡ 3 0 0⎤
⎢2

(b) The scaling is accomplished by multiplication of the coordinate matrix on the left by ⎢ 0 1 0 ⎥ , resulting in
2

⎣ 0 0 1⎦
⎡0 3 3 0 ⎤
2 2

⎛3
⎞ ⎛3 1 ⎞
⎛ 1 ⎞
the matrix ⎢0 0 1 1 ⎥ , which represents the vertices (0, 0, 0), ⎜ , 0, 0 ⎟ , ⎜ , , 0 ⎟ and ⎜ 0, , 0 ⎟ as
2 2
⎝2
⎠ ⎝2 2 ⎠
⎝ 2 ⎠

⎣0 0 0 0 ⎦
shown below.

⎡ −2 −2 −2 −2 ⎤
⎡ −2 −1 −1 −2 ⎤
(c) Adding the matrix ⎢ −1 −1 −1 −1⎥ to the original matrix yields ⎢ −1 −1 0 0 ⎥ , which represents

⎣ 3 3 3 3⎦
⎣ 3 3 3 3⎦
the vertices (−2, −1, 3), (−1, −1, 3), (−1, 0, 3), and (−2, 0, 3) as shown below.

⎡cos(−30°) − sin(−30°) 0 ⎤
(d) Multiplying by the matrix ⎢ sin(−30°) cos(−30°) 0 ⎥ , we obtain

0
0
1⎦

.866 1.366 .500 ⎤
⎡0 cos(−30°) cos(−30°) − sin(−30°) − sin(−30°) ⎤ ⎡0
⎢0 sin(−30°) cos(−30°) + sin(−30°) cos(−30°) ⎥ = ⎢0 −.500 .366 .866 ⎥ .

⎥ ⎢

0
0
0
0
0
⎣0
⎦ ⎣0 0

The vertices are then (0, 0, 0), (.866, −.500, 0), (1.366, .366, 0), and (.500, .866, 0) as shown:

⎡1 1 0 ⎤ ⎡ xi ⎤ ⎡ xi + 1 yi ⎤
2 ⎥
2

2. (a) Simply perform the matrix multiplication ⎢0 1 0 ⎥ ⎢ yi ⎥ = ⎢ yi ⎥ .

⎥⎢ ⎥
⎢⎣0 0 1 ⎥⎦ ⎢⎣ zi ⎥⎦ ⎢⎣ zi ⎥⎦

230

⎢ ⎥ ⎣0 sin 20° cos 20° ⎦ ⎡ cos(−45°) 0 sin(−45)° ⎤ ⎥ . 0). P ′ = M 7 ( M 6 + M 5 M 4 ( M 3 + M 2 M1P)).800. (1. ⎢ ⎥ ⎣ 0 0 1⎦ Negating the x-coordinates of the 12 points in view 1 yields the 12 points (−1. 0 ⎟ . 0. . 0. 0 " 0⎦⎥ 1 2 1 0 ⎤ ⎡1 M 3 = ⎢0 cos 20° − sin 20° ⎥ . 3. 0). and ⎜ .. 1. translation. ⎢ ⎥ ⎣0 0 0 " 0⎦ ⎡ cos 35° 0 sin 35° ⎤ M4 = ⎢ 0 1 0 ⎥. 0 ⎟ . (a) The formulas for scaling.866). etc. ⎢ ⎥ ⎣0 0 1⎦ ⎡. ⎢ ⎥ 0 1⎦ ⎣ 0 (b) Clearly P ′ = M 5 M 4 M 3 ( M1 P + M 2 ). (−. ⎢ ⎥ sin 35 0 cos 35°⎦ − ° ⎣ ⎡cos(−45°) − sin(−45°) 0 ⎤ M 5 = ⎢ sin(−45°) cos(−45°) 0 ⎥ .5 0⎥ . ⎢ ⎥ ⎣ 0 0 1⎦ 0 0 ⎤ ⎡1 M 2 = ⎢0 cos 45° − sin 45° ⎥ . yi .SSM: Elementary Linear Algebra Section 10.6. as shown: (c) Here we want to negate the z-coordinates. yi . 231 . (1. −. ⎢ ⎥ ⎣0 0 1 ⎦ (b) For this reflection we want to transform ( xi . (a) As in 4(a).800. .500. zi ) to (− xi . as shown: 1 2 0 0 " 12 ⎤ ⎥ 0 " 0⎥ . −1.10 (b) We multiply ⎡1 1 0 ⎤ ⎡ 0 1 1 0 ⎤ ⎡ 0 1 3 1 ⎤ 2 2 2⎥ ⎢ ⎥ ⎢0 0 1 1 ⎥ = ⎢ 0 1 0 0 0 1 1⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎢⎣0 0 1 ⎥⎦ ⎣0 0 0 0 ⎦ ⎣⎢0 0 0 0 ⎦⎥ yielding the vertices (0.000).000. ⎢ ⎥ 0 0 1⎦ ⎣ ⎡0 0 0 " 0 ⎤ M 6 = ⎢0 0 0 " 0 ⎥ . respectively and indeed its ⎡ 1 0 0⎤ matrix is ⎢0 −1 0 ⎥ . ⎢ ⎥ ⎣0 0 −1⎦ This does not change View 1. ⎛3 ⎞ ⎛1 ⎞ ⎜ . and ⎢ ⎥ ⎣1 1 1 " 1 ⎦ ⎡2 0 0⎤ M 7 = ⎢0 1 0⎥ . M1 = ⎢ 0 . 0). −. 0).6. −. and (0. 1. 0.6 1 0 ⎥ ⎢0 0 1 1 ⎥ = ⎢0 . 0). (1. ⎢ ⎥ ⎣0 sin 45° cos 45° ⎦ ⎡1 1 1 " 1 ⎤ M 3 = ⎢0 0 0 " 0⎥ . 3 0 0 ⎤ 5.6 1. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 0 0 1 ⎦ ⎣0 0 0 0 ⎦ ⎣0 0 0 0⎦ yielding (0. 1. ⎢0 0 1⎥ 3⎦ ⎣ ⎡1 2 M2 = ⎢0 ⎢ ⎣⎢ 0 (c) Obtain the vertices via ⎡ 1 0 0 ⎤ ⎡0 1 1 0 ⎤ ⎡0 1 1 0⎤ ⎢. 1. 1. and M4 = ⎢ 0 1 0 ⎢ ⎥ sin( 45 ) 0 cos( 45 ) − − ° − ° ⎣ ⎦ cos 90 ° − sin 90 ° 0 ⎡ ⎤ M 5 = ⎢ sin 90° cos 90° 0 ⎥ . zi ) with the matrix ⎡ −1 0 0 ⎤ ⎢ 0 1 0⎥ . ⎡ 1 0 0⎤ with the matrix ⎢0 1 0 ⎥ . ⎝2 ⎠ ⎝2 ⎠ 4. 0). and rotation yield the matrices ⎡ 1 0 0⎤ ⎢2 ⎥ M1 = ⎢ 0 2 0 ⎥ .6 1 ⎥ . (b) As in 4(b). (a) This transformation looks like scaling by the factors 1.

and ⎢ ⎥ 0 1⎦ ⎣ 0 Section 10. (a) The discrete mean value property yields the four equations 1 t1 = (t2 + t3 ) 4 1 t2 = (t1 + t4 + 1 + 1) 4 1 t3 = (t1 + t4 ) 4 1 t4 = (t2 + t3 + 1 + 1). for the rotation matrix about the xaxis we obtain 0 0 ⎤ ⎡1 0 0 ⎤ ⎡1 RRT = ⎢ 0 cos θ − sin θ ⎥ ⎢0 cos θ sin θ ⎥ ⎢ ⎥⎢ ⎥ ⎣ 0 sin θ cos θ ⎦ ⎣0 − sin θ cos θ ⎦ ⎡1 0 0 ⎤ = ⎢0 1 0⎥ . this is ⎡1 − 1 − 1 0 ⎤ ⎡ t1 ⎤ ⎡ 0 ⎤ 4 4 ⎢ ⎥ ⎢ ⎥ 1 −15 4 ⎥ ⎢t2 ⎥ ⎢ 0 ⎥ ⎢0 ⎢ ⎥= 1 . ⎢0 0 1 − 12 ⎥ ⎢ t3 ⎥ ⎢ − 8 ⎥ ⎢ ⎥ ⎢t ⎥ ⎢ 3 ⎥ 0 0 1⎦⎥ ⎣ 4 ⎦ ⎣⎢ 4 ⎦⎥ ⎣⎢0 x0 ⎤ ⎡ xi ⎤ y0 ⎥ ⎢ yi ⎥ ⎥ ⎢ ⎥. ⎢ ⎥ ⎣ − sin β 0 cos β ⎦ ⎡cos α − sin α 0 ⎤ R2 = ⎢ sin α cos α 0 ⎥ . ⎢ ⎥ ⎣ − sin(− β ) 0 cos(− β ) ⎦ 7.11 1. This can be done most easily performing the multiplication RRT and showing that this is I. so x0 = −5. this ⎡0 1 1 0⎤ 0 4 4 ⎡ t1 ⎤ ⎢ ⎥ ⎡ t1 ⎤ ⎡⎢ ⎤⎥ 1 1 ⎢t ⎥ ⎢ 0 0 4 ⎥ ⎢t2 ⎥ 12 becomes ⎢ 2 ⎥ = ⎢ 14 ⎢ ⎥ + ⎢ ⎥. zi by −3. ⎢ ⎥ ⎣0 0 1 ⎦ 232 . 1 −3⎥ 0 1⎥⎦ ⎡1⎤ ⎢4⎥ ⎢3⎥ Back substitution yields the result t = ⎢ 14 ⎥ . 1⎥ t ⎢0⎥ t 0 0 3 ⎢ 3⎥ ⎢4 4⎥ ⎢ ⎥ ⎢ ⎥ ⎢⎣t4 ⎥⎦ ⎢ ⎢ ⎥ t ⎥ 1 1 ⎣ 4 ⎦ ⎢⎣ 12 ⎥⎦ ⎢⎣ 0 4 4 0 ⎥⎦ ⎡ cos(− β ) 0 sin(− β ) ⎤ R5 = ⎢ 0 1 0 ⎥. ⎢ ⎥ sin 0 cos − θ θ⎦ ⎣ ⎡ cos(−α ) − sin(−α ) 0 ⎤ R4 = ⎢ sin(−α ) cos(−α ) 0 ⎥ . (a) We rewrite the formula for vi′ as ⎡1 ⋅ xi + x0 ⋅1⎤ ⎢1 ⋅ y + y ⋅1⎥ 0 ⎥ . yi by +9. For example. So vi′ = ⎢ i z z ⋅ + 1 ⎢ i 0 ⋅1 ⎥ ⎢⎣ ⎥⎦ 1 ⋅1 ⎡1 ⎢0 vi′ = ⎢ ⎢0 ⎢⎣0 0 1 0 0 0 0 1 0 (b) To solve the system in part (a). ⎢ ⎥ 0 1⎦ ⎣ 0 ⎡ cos θ 0 sin θ ⎤ R3 = ⎢ 0 1 0 ⎥. 4 Translated into matrix notation. we solve (I − M)t = b for t: ⎡ 1 −1 −1 0⎤ 0 4 4 ⎢ ⎥ ⎡ t1 ⎤ ⎡⎢ ⎤⎥ 1 1 1 ⎢ ⎥ 1 0 − 4 ⎥ t2 ⎢− 4 ⎢2⎥. Using the hing given. ⎡1 ⎢0 The matrix is ⎢ ⎢0 ⎢⎣0 0 1 0 0 0 −5⎤ 0 9⎥ ⎥.Chapter 10: Applications of Linear Algebra SSM: Elementary Linear Algebra 6. we have ⎡ cos β 0 sin β ⎤ R1 = ⎢ 0 1 0 ⎥. z0 ⎥ ⎢ zi ⎥ 1 ⎥⎦ ⎢⎣ 1 ⎥⎦ (b) We want to translate xi by −5. ⎢4⎥ ⎢3⎥ ⎣⎢ 4 ⎦⎥ 8. ⎢ ⎥ = ⎢ 1 ⎥ 0 1 − 14 ⎥ ⎢ t3 ⎥ ⎢ 0 ⎥ ⎢− 4 ⎢ ⎥ ⎢t ⎥ ⎢ 1 ⎥ 1 1 1⎥⎦ ⎣ 4 ⎦ ⎣⎢ 2 ⎦⎥ ⎢⎣ 0 − 4 − 4 In row-echelon form. z0 = −3. y0 = 9.11 Exercise Set 10.

we have 233 . but using M and b as in the problem statement. and for . x02 using part (b) of Exercise 1. . x32 22 (d) Using percentage error computed value − actual value = × 100% actual value we have that the percentage error for t1 and 2.0371 × 100% = −12. −π π <θ < and is zero Clearly f(θ) = 1 for 2 2 otherwise.0371 t2 and t4 was × 100% = 5.SSM: Elementary Linear Algebra (c) t (1) = Mt (0) + b ⎡0 1 1 4 4 ⎢ ⎢ 14 0 0 =⎢ 1 ⎢4 0 0 ⎢ 1 1 ⎣⎢ 0 4 4 ⎡ 0⎤ ⎢1⎥ = ⎢2⎥ ⎢ 0⎥ ⎢1⎥ ⎢⎣ 2 ⎥⎦ Section 10. which has the solution 31 * x31 = 22 27 * = .12 ⎡ 15 ⎤ ⎢ 64 ⎥ ⎢ 47 ⎥ t (5) = Mt( 4) + b = ⎢ 64 ⎥ 15 ⎢ 64 ⎥ ⎢ 47 ⎥ ⎣⎢ 64 ⎦⎥ 1. As in 1(c). ⎢8⎥ ⎢5⎥ ⎢⎣ 8 ⎥⎦ ⎡3⎤ ⎢ 16 ⎥ ⎢ 11 ⎥ t (3) = Mt ( 2) + b = ⎢ 16 ⎥ 3 ⎢ 16 ⎥ ⎢ 11 ⎥ ⎣⎢ 16 ⎦⎥ 3. (a) Setting ( ) ( ) (1) (1) = (0) (0) = (0. 2 ⎡1⎤ ⎢8⎥ ⎢5⎥ t ( 2) = Mt(1) + b = ⎢ 81 ⎥ . where r is the radius circle is 2π r ∫−π of the circle and f(θ) is the temperature at the point of the circumference where the radius to that point makes the angle θ with the horizontal. the value of the integral above (which equals the temperature at 1 the center of the circle) is . Consequently.2871 . and x(01) = x01 x31 .2%. (c) The linear system 1 * * * x31 = [ 28 + x31 − x32 ] 20 1 * * * x32 = [ 24 + 3x31 − 3x32 ] 20 can be rewritten as ⎡ 15 ⎤ ⎡ 1 ⎤ ⎡ − 1 ⎤ ⎢ 64 ⎥ ⎢ 4 ⎥ ⎢ 64 ⎥ ⎢ 47 ⎥ ⎢ 3 ⎥ ⎢ − 1 ⎥ t (5) − t = ⎢ 64 ⎥ − ⎢ 4 ⎥ = ⎢ 64 ⎥ 15 1 1 ⎢ 64 ⎥ ⎢ 4 ⎥ ⎢ − 64 ⎥ ⎢ 47 ⎥ ⎢ 3 ⎥ ⎢ 1 ⎥ − ⎣⎢ 64 ⎦⎥ ⎢⎣ 4 ⎥⎦ ⎣⎢ 64 ⎦⎥ * * 19 x31 + x32 = 28 * * −3x31 + 23x32 = 24.9%. x32 .12 −. 0). The average value of the temperature on the 1 π f (θ )r dθ .7129 t3 was 0⎤ 0 ⎥ ⎡0 ⎤ ⎡ ⎤ 1⎥ ⎢ ⎥ ⎢1⎥ 4 0 + ⎢2⎥ 1 ⎥ ⎢0 ⎥ ⎢ 0 ⎥ 4⎥ ⎢ ⎥ ⎢1⎥ ⎥ ⎢0 ⎥ 0 ⎦⎥ ⎣ ⎦ ⎣⎢ 2 ⎦⎥ 2. we obtain t (1) = Mt (0) + b = ⎡⎣ 34 5 4 1 2 5 4 1 2 5 4 13 16 7 16 1 3⎤ 4⎦ 1 T t ( 2) = Mt(1) + b ⎡7⎤ ⎢ 32 ⎥ ⎢ 23 ⎥ t ( 4) = Mt(3) + b = ⎢ 32 ⎥ . Section 10.12 Exercise Set 10. 7 ⎢ 32 ⎥ ⎢ 23 ⎥ ⎢⎣ 32 ⎥⎦ 13 = ⎡ 16 ⎣ 9 8 9 16 11 8 21 16 1 15 ⎤ 8⎦ T .

22727] 20 = 1.40909) − 3(1.59500 20 (1) x31 = 234 .65000 20 1 ( 2) x31 = [ 28 + 9.22727)] 20 = 1..4091 − 1.227)] 20 = 1.Chapter 10: Applications of Linear Algebra SSM: Elementary Linear Algebra 1 [ 28] = 1. 1) = x31 ..23] = 1.2273] = 1.41 − 1.2] = 1.22727 1 ( 6) x31 = [ 28 + 1.40909 − 1.2 20 Since x(31) in this part is the same as x(31) in part (a).4091) − 3(1.. − 15) = x31 .41000 20 1 ( 2) x32 = [ 24 + 3(1. . x32 1 (1) x31 = [ 28 + 1 − 1] = 1.4 20 1 ( 2) x32 = [ 24 + 3(1) − 3(1)] = 1.23)] = 1.40910 20 1 ( 4) x32 = [ 24 + 3(1.55 − 25.20000 20 1 ( 2) x31 = [ 28 + 1. we will get x(32) as in part (a) and therefore x(33) .41) − 3(1.40909 20 1 (5) x32 = [ 24 + 3(1. x32 ) 1 [ 28 + 148 − (−15)] = 9.227] = 1.22700 20 1 ( 4) x31 = [ 28 + 1.40900 20 1 (3) x32 = [ 24 + 3(1.55000 20 1 (1) x32 = [ 24 + 3(148) − 3(−15)] = 25.22727.23000 20 1 (3) x31 = [ 28 + 1.4) − 3(1.22730 1 (5) x31 = [ 28 + 1.409 − 1.40000 20 1 (1) x32 = [ 24] = 1. ( ( 0) (0) (c) x(01) = (148.409) − 3(1. x(36) will also be the same as in part (a).65] = 0.2)] = 1. (1) x31 = ( ) ( 0) (0) (b) x(01) = (1.40909 1 ( 6) x32 = [ 24 + 3(1.2273)] 20 = 1.4 − 1.

40095 20 1 = [ 24 + 3(1.40095 − 1.40991) − 3(1.55) − 3(25. it follows that x21 − x22 = 2. x1(1) is projected to x(21) on L2 .40095) − 3(1.65)] 20 = −1. Referring to the figure below and starting with x(01) = (0. x2 L1 L3 (2) (1) x1 x1 L2 A x2 = 1 x(1) 3 x1 (1) x0 B (1) x2 x1 – x2 = 0 x1 – x2 = 2 As seen from the graph the points of the limit cycle are x1∗ = A.22972] = 1.215] 20 = 1.20285 1 = [ 28 + 1.49050 1 = [ 24 + 3(0.595 + 1.40991 1 = [ 24 + 3(1.47150 1 = [ 28 + 1.20285] 20 = 1.12 1 [ 24 + 3(9.21500 1 = [ 28 + 0. 0) : x(01) is projected to x1(1) on L1 .4905 − 1.22972)] 20 = 1.40901 20 1 = [ 24 + 3(1. Since x1∗ is the point of intersection of L1 and L3 it follows on solving the system ∗ x12 =1 ∗ * x11 − x12 =0 ∗ ∗ ∗ ∗ .215)] 20 = 1. Since x∗2 = ( x21 235 .4715)] 20 = 1.22703 ( 2) x32 = (3) x31 (3) x32 ( 4) x31 ( 4) x32 (5) x31 (5) x32 ( 6) x31 ( 6) x32 4. Now x1∗ x∗2 is perpendicular to L2 . x∗3 = A. that x1* = (1. 1). x(21) is projected to x(31) on L3 .22972 1 = [ 28 + 1.4715] = 1. x22 ) is on L2 .4905) − 3(1.SSM: Elementary Linear Algebra Section 10. x∗2 = B.20285)] 20 = 1.595) + 3(1. and so on.40991 − 1.

the hypotenuse of triangle A is the portion of the centerline of the 10th beam that lies in the 2nd pixel. which in turn is 2 − 1. 9 = 2 = 1. 6 = a12. respectively. Also from the diagram. 2 = a10. Since these diagonals have length 2. Thus. 3 = 2 − 2 = . 3.11 in the text. x∗3 = (1. a10. and 9 cross the pixels squarely. Let us choose units so that each pixel is one unit wide. In the following diagram. 8 = a62 = a64 = a46 = a48 = . 1). 8. the centerlines of beams 5 and 11 lie along the diagonals of pixels 3. we have a53 = a55 = a57 = 2 = 1. If the i-th beam crosses the j-th pixel squarely.Chapter 10: Applications of Linear Algebra SSM: Elementary Linear Algebra ⎛ x∗ − 1 ⎞ ∗ ∗ ∗ ∗ therefore ⎜ 22 ⎟ (1) = −1 so we have x22 − 1 = 1 − x21 or x21 + x22 = 2. 5 = a11.82843. 1 = a11.82843. 7.12. 2 = 2 ( ) 2 − 1 = .41421. x∗2 = (2. 9. 10. 236 . 2. 0). Thus. we see that the hypotenuse of triangle B is the portion of the centerline of the 10th beam that lies in the 3rd pixel. 5. it is then clear that a17 = a18 = a19 = 1 a24 = a25 = a26 = 1 a31 = a32 = a33 = 1 a73 = a76 = a79 = 1 a82 = a85 = a88 = 1 a91 = a94 = a97 = 1 since beams 1. Next.58579. The length of this hypotenuse is twice the height of triangle A. Then aij = length of the center line of the i-th beam that lies in the j-th pixel. 4 = a12. From Fig. it follows that aij = 1. Thus the points on the limit cycle are x1∗ = (1.41421 a11. gives x21 7. a10. By symmetry we also have a10. 7 and 1. 5. 1). ⎜ x∗ − 1 ⎟ ⎝ 21 ⎠ Solving the system ∗ ∗ x21 − x22 =2 ∗ ∗ x21 + x22 =2 ∗ ∗ = 2 and x22 = 0.

3.12. and 9 cross the pixels squarely. 7 = a61 = a49 = .11 in the text. we then have 237 .79 1. 2. 10.58579 x7 = 7.58579 x9 = 14. and so the 12 beam equations (4) are x7 + x8 + x9 = 13.00 x4 + x5 + x6 = 15.41421( x3 + x5 + x7 ) = 14. Let us choose units so that each pixel is one unit wide. The remaining aij ’s are all zero.12 2−1 2 1 A 2( 2 − 1) B 1 2 2− 2 1 − 2 2 1− 1 − 2 2 1 By symmetry we have a10.00 = 12.00 x1 + x2 + x3 = 8. For the remaining aij ’s.00 . Since the width of each beam is also one unit it follows that aij = 1 if the i-th beam crosses the j-th pixel squarely. From the diagram of the nine pixels.51 1.58579. first observe from the figure that an isosceles right triangle of height h.82843( x6 + x8 ) + .SSM: Elementary Linear Algebra Section 10. Then aij = area of the i-th beam that lies in the j-th pixel.58579 x1 = 3. as indicated. 7. 3 = a12. From Fig.81 x3 + x6 + x9 x2 + x5 + x8 x1 + x4 + x7 . it is then clear that a17 = a18 = a19 = 1 a24 = a25 = a26 = 1 a31 = a32 = a33 = 1 a73 = a76 = a79 = 1 a82 = a85 = a88 = 1 a91 = a94 = a97 = 1 since beams 1.04 8. has area h 2 .41421( x1 + x5 + x9 ) = 16.82843( x4 + x8 ) + .13 .82843( x2 + x6 ) + . 8.00 = 6.31 .00 = 10.58579 x3 = 18.82843( x2 + x4 ) + .

25000 4 ⎣2⎦ 2 ⎡3 ⎤ Area of triangle F = ⎢ ( 2 − 1) ⎥ ⎣2 ⎦ 9 = (3 − 2 2 ) 4 = .Chapter 10: Applications of Linear Algebra SSM: Elementary Linear Algebra h Area = h 2 1 − 2 ( 2 − 1) 2 B 1 F D A E C 1 − 2 B 3 − 2 3 −( 2 3 − 2 2 1 ⎡1 ⎤ Area of triangle B = ⎢ ( 2 − 1) ⎥ ⎣2 ⎦ 1 = (3 − 2 2 ) 4 = 0.38604 We also have Area of polygon A = 1 − 2 × (Area of triangle B) 1 = 2− 2 = .91421 238 2 − 1) .4289 2 2 1 ⎡1⎤ Area of triangle C = ⎢ ⎥ = = .

25000 a10.61396 x9 = 14.12.91421( x1 + x5 + x9 ) + 0. and so the 12 beam equations (4) are x7 + x8 + x9 = 13.04289( x3 + x5 + x7 ) + 0. 9 = a53 = a55 = a57 = . 6 = a11.00 x1 + x4 + x7 = 6.04 239 .04289 a11. 4 = a12.79 0. 1 = Area of triangle B = . 1 = Area of polygon A = .75000 a10.61396 By symmetry we then have a11. we see that a11. 1 = a11. 2 = Area of triangle C = .04289( x1 + x5 + x9 ) + 0.61396 Referring back to Fig.61396 x7 = 7. 2 = a11.75000 a10. 1 = a12. 9 = a63 = a65 = a67 = a43 = a45 = a47 = . 4 = a11.75( x6 + x8 ) + 0.04289 a11.91421( x3 + x5 + x7 ) + 0. 5 = a12.00 0. 8 = a62 = a64 = a46 = a48 = .12 Area of polygon D = 1 − (Area of triangle C ) 1 = 1− 4 3 = 4 = . 5 = a11. 7 = a61 = a49 = . 5 = a10.00 x4 + x5 + x6 = 15. 10.31 0. 2 = Area of polygon D = .SSM: Elementary Linear Algebra Section 10.25( x2 + x4 + x6 + x8 ) = 16.81 x3 + x6 + x9 = 18. The remaining aij ’s are all zero. 6 = a12.00 0.13 0.04289( x3 + x5 + x7 ) + 0. 3 = a12.61396 x1 = 3.25000 a10. 9 = a12. 8 = a52 = a54 = a56 = a58 = .75( x2 + x6 ) + 0.61396 x3 = 10.51 0.00 x1 + x2 + x3 = 8.91421 a10.61396.7500 Area of polygon E = 1 − (Area of triangle F ) 1 = (18 2 − 23) 4 = . 2 = a10.25( x2 + x4 + x6 + x8 ) = 14.75( x4 + x8 ) + 0.00 x2 + x5 + x8 = 12.04289( x1 + x5 + x9 ) + 0.75( x2 + x4 ) + 0. 3 = Area of polygon E = . 1 = a10.11.91421 a10.

S3 . S4 . S2 . Each of the subsets S1 . respectively.. 2. . S2 .Chapter 10: Applications of Linear Algebra SSM: Elementary Linear Algebra Section 10. and Ti ⎜ ⎢ ⎥ ⎟ = + i = 1. . are ⎢f ⎥ ⎢⎣0 ⎥⎦ ⎢ 0 ⎥ ⎢ 13 ⎥ ⎢⎣ 0 1 ⎥⎦ ⎢⎣ y ⎥⎦ ⎢ f ⎥ y 25 ⎣ ⎦ ⎢ 13 ⎥ ⎣ i⎦ ⎣ i⎦ ⎝ ⎠ ⎣⎢ ⎦⎥ ⎣⎢ 25 ⎦⎥ ⎣ 25 ⎦ 12 Because s = and k = 4 in the definition of a self-similar set. Also. 4. 1 25 ln s ln 12 () 2. and S4 .. . S4 in the figure is congruent to the entire set scaled by a factor of y 1 12 25 1 25 S3 S4 S1 S2 x 1 ⎡ 13 ⎤ ⎤ ⎡0⎤ ⎛ ⎡ x ⎤ ⎞ 12 ⎡1 0 ⎤ ⎡ x ⎤ ⎡ ei ⎤ ⎡ ei ⎤ ⎡0 ⎤ ⎡ 13 25 ⎢ 25 ⎥ .8 to two significant digits.888. S3 . Examination of the figure reveals rotation angles of 180°. The displacement distances can be determined from the figure to find the four similitudes that map the entire set onto the four subsets S1 . the Hausdorff dimension of the set is 25 ln(k ) ln(4) d H (S ) = = = 1. 3. S3 . Since k = 4. 1. respectively. The set is a fractal because its Hausdorff dimension is not an integer. 0°. These are. the Hausdorff dimension of the set is approximately d H ( S ) = 2 ln(k ) ln () 1 s = ln(4) ln ( .471 ) = 1 . S2 . where the four values of .. 240 . 180°.13 Exercise Set 10. ( ) 15 ( 16 ) The rough measurements indicated in the figure give an approximate scale factor of s ≈ = . and −90° for the sets S1 .47 to two decimal places. the 25 rotation angles for the four subsets are all 0°.13 12 .

771.. 2) the upper right and lower right iterations are both rotated 180° (2. By inspection. ln(3) ln 1 (s) 241 . 0) the upper right iteration is rotated 270° (0.. and the lower right is rotated 270° (2. 3. the 3 rotations angles are 0° for all seven subsets. 0) none are rotated (1. 0) the upper right iteration is rotated 180° (3. 0) the upper right iteration is rotated 90° (2. 3) the upper right iteration is rotated 180°. reading left to right and top to bottom. each of which is congruent to the original set scaled by a factor s = . Because its Hausdorff dimension is not an integer. 2) the lower right iteration is rotated 180° (1. (a) The figure shows the original self-similar set and a decomposition of the set into seven nonoverlapping 1 congruent subsets. the lower left is rotated 90°. 0. 0. the triplets are: (0. 0. 3) the lower left and lower right iterations are both rotated 270° 4. By inspection. 0. The Hausdorff dimension of the set is ln(k ) ln(7) d H (S ) = = = 1. 1. 2. 0) the upper right iteration is rotated 90° and the lower left is rotated 180° (2.13 15 " 16 s3 s4 s1 s2 2" 3.SSM: Elementary Linear Algebra Section 10. 2. 1) the lower right iteration is rotated 90° (0.. 0) the upper right and lower left iterations are both rotated 180° (0. 0. the set is a fractal. 1) the upper right iteration is rotated 180° and the lower right is rotated 90° (2. 0. 0. 0.

180°. By inspection. and S3 .. each of which is congruent to the original set scaled by a factor s = . By inspection. the 2 rotation angles are 180°. each of which is congruent to the original set scaled by a factor s = . ln(2) ln 1 (s) (c) The figure shows the original self-similar set and a decomposition of the set into three nonoverlapping 1 congruent subsets.. The Hausdorff dimension of the set is d H ( S ) = ln(k ) () ln 1s = ln(3) = 1.. the set is a fractal. respectively.. Because its Hausdorff dimension is not an integer.Chapter 10: Applications of Linear Algebra SSM: Elementary Linear Algebra (b) The figure shows the original self-similar set and a decomposition of the set into three nonoverlapping 1 congruent subsets. the 2 rotation angles are 180° for all three subsets.584. Because its Hausdorff dimension is not an integer.584. ln(2) S3 S1 242 S2 . S2 .. the set is a fractal. The Hausdorff dimension of the set is ln(k ) ln(3) d H (S ) = = = 1.. and −90° for S1 .

and −90° for S1 .SSM: Elementary Linear Algebra Section 10. S2 . text is d H ( S ) = ln(4) ln 1 (s) 243 .85 .075⎤ ⎡. 1 each of side length . ⎡ x ⎤ ⎡.180⎥⎦ = ⎢⎣.... Solving this matrix equation gives rounded to three decimal places. The Hausdorff dimension of the set is d H ( S ) = ln(k ) ln () 1 s = ln(3) = 1.996 ⎥⎦ 7.584. we must have s cos θ = .04 ⎞ ⎟ = −2. the Hausdorff dimension of the unit square as given by Equation (2) of the 4 ln(k ) ln(16) = = 2.04 ⎤ ⎢⎣ y ⎥⎦ = ⎢⎣.04.180 ⎥⎦ .04⎤ 5.°.04 .766 ⎤ ⎢⎣.15 −. Letting ⎢ ⎥ be the vector to the tip of the fern and using the hint. The matrix portion of a similitude is of the ⎣ −.85 .04)2 = .04 . the 2 rotation angles are 180°. 180°. The matrix of the affine transformation in question is ⎢ .04 ⎤ ⎡ x ⎤ ⎡.85⎥⎦ ⎡cos θ − sin θ ⎤ form s ⎢ . Consequently.15⎥⎦ −1 ⎡.13 (d) The figure shows the original self-similar set and a decomposition of the set into three nonoverlapping 1 congruent subsets.85 and cos θ ⎥⎦ ⎣ sin θ s sin θ = −. we have ⎢ ⎥ = T2 ⎜ ⎢ ⎥ ⎟ or ⎣ y⎦ ⎣ y⎦ ⎝ ⎣ y⎦ ⎠ ⎡ x ⎤ ⎡ . ln(2) S3 S2 S1 ⎡ .075⎤ ⎢⎣ y ⎥⎦ = ⎢⎣ −. By inspection. the unit square can be expressed as the union of 16 nonoverlapping congruent squares. Consequently. the set is a fractal. and S3 ..85)2 + (−. As the figure indicates. Because its Hausdorff dimension is not an integer. each of which is congruent to the original set scaled by a factor s = .69. ⎝ ..8509..85 ⎠ θ = tan −1 ⎛⎜ ⎛ ⎡ x⎤ ⎞ ⎡ x⎤ ⎡x⎤ 6.. and −. Solving this pair of equations gives s = (. respectively.04 .85⎥⎦ ⎢⎣ y ⎥⎦ + ⎢⎣.

Chapter 10: Applications of Linear Algebra SSM: Elementary Linear Algebra 1 4 1 8. Equation (2) of the text gives d H ( S ) = 2 ln(2) ln 1s () of a unit cube. ⎟ . Each of these four smaller squares has side length of . and ⎜ ... and ⎜ . (1. 9.726. and (0. Finally the similitude T4 maps the unit square onto the square ⎝ 4⎠ ⎝4 4⎠ ⎝4 ⎠ 3 ⎛1 1⎞ ⎛ 1⎞ ⎛1 ⎞ whose vertices are ⎜ . the Menger sponge is a fractal. Because k = 2 and s = . 0 ⎟ . Because the Hausdorff dimension of the cube is the same as its topological dimension. 1⎟ .. The similitude T2 maps the unit square onto the square whose ⎝4 ⎠ ⎝4 4⎠ ⎝ 4⎠ ⎛1 ⎞ ⎛ 3⎞ ⎛1 3⎞ vertices are ⎜ .. 1⎟ . 244 . (1. and ⎜ 0.6309. ⎜1. A careful examination of Figure Ex-10 in the text shows that the Menger sponge can be expressed as the union of 1 1 20 smaller nonoverlapping congruent Menger sponges each of side length . (1.. 0). k = 20 and s = . ⎜1.. 3 3 ln(k ) ln(20) = = 2. The similitude T3 maps the unit square onto the square whose ⎝4 ⎠ ⎝ 4⎠ ⎝4 4⎠ ⎛ 1⎞ ⎛3 1⎞ ⎛3 ⎞ vertices are ⎜ 0.818. ⎟ . 1). ⎜ . ⎜ . Because s = ln(k ) ln(8) 1 = = 3 for the Hausdorff dimension and k = 8. This is not the correct Hausdorff dimension of the square (which is 2) because the four ln 1s ln 43 () () smaller squares overlap. Because its and so the Hausdorff dimension of the Menger sponge is d H ( S ) = ln(3) ln 1 (s) Hausdorff dimension is not an integer. ⎟ . ⎜ . ⎟ .. 11. 0 ⎟ . ⎜ . 1)) onto the square whose ⎛3 ⎞ ⎛3 3⎞ ⎛ 3⎞ vertices are (0. ⎟ . and (0. The figure shows the first four iterates as determined by Algorithm 1 and starting with the unit square as the 1 initial set. ⎟ . ⎟ . 0). The similitude T1 maps the unit square (whose vertices are (0. 0). (1. 10. 4 ⎝4 4⎠ ⎝ 4⎠ ⎝4 ⎠ 3 so that the common scale factor of the similitudes is s = .. the Hausdorff dimension of the Cantor set is 3 ln(k ) ln(2) d H (S ) = = = 0. The right-hand side of Equation (2) of the text gives 4 ln(k ) ln(4) = = 4. 1). the cube is not a fractal.. 0). Notice that the Cantor set is a subset of the unit interval along the x-axis and ln(3) ln 1 (s) that its topological dimension must be 0 (since the topological dimension of any set is a nonnegative integer less than or equal to its Hausdorff dimension). ⎟ . 1). Consequently.

245 . Because 250 = 2 ⋅ 53 it follows from (i) that Π (250) = 3 ⋅ 250 = 750. T2 . (Notice that this 9 ⎝9⎠ ⎝9⎠ ⎝9⎠ n ⎛8⎞ implies that the area of the Sierpinski carpet is 0.. 1..) ⎝9⎠ Section 10.14 Initial set First Iterate Second Iterate Third Iterate Fourth Iterate 12.. since the limit of ⎜ ⎟ as n tends to infinity is 0. Because 25 = 52 it follows from (ii) that Π (25) = 2 ⋅ 25 = 50. their total area is . which is then the area of the set S1 ... 8 ⎛ 8 ⎞ 2 ⎛ 8 ⎞3 ⎛ 8 ⎞ 4 .14 Exercise Set 10. we find that the 9 areas of S0 . Because 30 = 6 ⋅ 5 it follows from (ii) that Π (30) = 2 ⋅ 30 = 60. Continuing the argument further. ⎜ ⎟ . . . Because 125 = 53 it follows from (ii) that Π(125) = 2 ⋅125 = 250.. Each of the eight similitudes T1 . Because 10 = 2 ⋅ 5 it follows from (i) that Π (10) = 3 ⋅10 = 30. ⎜ ⎟ . S3 .14 1.SSM: Elementary Linear Algebra Section 10. The area of the unit square S0 is.. T8 given in Equation (8) of 1 1 the text has scale factor s = . S2 . ⎜ ⎟ . S4 . form the geometric sequence 1.. and so each maps the unit square onto a smaller square of area . of course. . the area of the set S2 is -th the area of the set S1 . By a similar 9 8 argument.. Because these 3 9 8 eight smaller squares are nonoverlapping. S1.

and so Π (6) = 12. We continue in this way. 246 . ⎟ .Chapter 10: Elementary Linear Algebra SSM: Elementary Linear Algebra Because 50 = 2 ⋅ 52 it follows from (i) that Π (50) = 3 ⋅ 50 = 150. we arrive at ⎝ 6⎠ ⎡ 2 ⎤ ⎡0⎤ ⎡ 4 ⎤ ⎡0⎤ the 4-cycle ⎢ 2 ⎥ → ⎢ 6 ⎥ → ⎢ 4 ⎥ → ⎢ 6 ⎥ . We now choose another of the 36 points of the form ⎜ . 0) is obviously a 1-cycle. Taking one of the remaining points. Because 3750 = 6 ⋅ 54 it follows from (ii) that Π (3750) = 2 ⋅ 3750 = 7500. ⎟ . each time starting with some point of the form ⎢⎣ 64 ⎥⎦ ⎣ 6 ⎦ ⎢⎣ 62 ⎥⎦ ⎣6⎦ ⎛m n⎞ ⎜ . and another 12-cycle: 6 ⎢ → ⎢ ⎥ ⎢3⎥ ⎢ ⎥ ⎢ ⎥ ⎢ 63 ⎥ ⎢⎣ 64 ⎥⎦ ⎣ 0 ⎦ ⎢⎣ 62 ⎥⎦ ⎣⎢ 0 ⎦⎥ ⎣⎢ 6 ⎦⎥ ⎣0⎦ ⎣ ⎦ ⎡1⎤ ⎡2⎤ ⎡5⎤ ⎡1⎤ ⎡4⎤ ⎡1⎤ 6 6 6 6 6 ⎢6⎥ → ⎢1⎥ → ⎢3⎥ → ⎢2⎥ → ⎢3⎥ → ⎢1⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ 0 ⎣ ⎦ ⎣6⎦ ⎣6⎦ ⎣6⎦ ⎣6⎦ ⎣ 6 ⎥⎦ ⎡5⎤ ⎡5⎤ ⎡4⎤ ⎡1⎤ ⎡5⎤ → ⎢6⎥ → ⎢6⎥ → ⎢6⎥ → ⎢6⎥ → ⎢6⎥ ⎢ 56 ⎥ ⎢⎣ 63 ⎥⎦ ⎢ 36 ⎥ ⎢⎣ 64 ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎣ ⎦ ⎣ ⎦ 2 ⎡ ⎤ → ⎢ 65 ⎥ . ⎛m n⎞ 2. Its iterates produce the 12-cycle ⎝ 6⎠ ⎡3⎤ ⎡1⎤ ⎡2⎤ ⎡3⎤ ⎡1⎤ ⎡0⎤ 6⎥ → ⎢6⎥ → ⎢6⎥ → ⎢6⎥ → ⎢6⎥ ⎢ → ⎢1⎥ 5 4 ⎢ 65 ⎥ ⎢⎣ 62 ⎥⎦ ⎢⎣ 16 ⎥⎦ ⎣6⎦ ⎣⎢ 6 ⎦⎥ ⎣⎢ 6 ⎦⎥ ⎣ ⎦ ⎡5⎤ ⎡3⎤ ⎡4⎤ ⎡3⎤ ⎡5⎤ ⎡0⎤ → ⎢5⎥ → ⎢6⎥ → ⎢6⎥ → ⎢6⎥ → ⎢6⎥ → ⎢6⎥. Because 6 = 6 ⋅ 50 it follows from (ii) that Π(6) = 2 ⋅ 6 = 12 Because 5 = 51 it follows from (ii) that Π (5) = 2 ⋅ 5 = 10. ⎢6⎥ ⎣ ⎦ ⎛m n⎞ The possible periods of points for the form ⎜ . The least common multiple of these ⎝ 6 6⎠ four numbers is 12. 3. This yields a 3-cycle: ⎝ 6 6⎠ ⎡3⎤ ⎡4⎤ ⎡2⎤ 4 2 0 ⎡3⎤ 6 ⎥ → ⎡ ⎤ . The point (0. say ⎝ 6 6⎠ 1 ⎛ ⎞ ⎜ 0. 4 and 12. another 4-cycle: ⎡ 6 ⎤ → ⎢ 6 ⎥ → ⎡ 6 ⎤ → ⎢ 6 ⎥ . ⎢⎣ 64 ⎥⎦ ⎢⎣ 16 ⎥⎦ ⎢⎣ 65 ⎥⎦ ⎢⎣ 62 ⎥⎦ ⎢⎣ 16 ⎥⎦ ⎣⎢ 6 ⎦⎥ ⎛ 2⎞ So far we have accounted for 13 of the 36 points. until we exhaust all such points. ⎟ that has not yet appeared in a cycle. ⎟ are thus 1. say ⎜ 0. ⎟ .

mod 15 = 19 mod 15 = 4. mod 15 = 5 mod 15 = 5. mod 15 = 7 mod 15 = 7.SSM: Elementary Linear Algebra Section 10. x11 = x10 + x9 mod 15 = 7 + 1 mod 15 = 8 mod 15 = 8 x12 = x11 + x10 mod 15 = 8 + 7 mod 15 = 15 mod 15 = 0. mod 15 = 5 mod 15 = 5. With p = 15 we have x2 = x1 + x0 mod 15 = 7 + 3 mod 15 = 10 mod 15 = 10. mod 15 = 14 mod 15 = 14. mod 15 = 17 mod 15 = 2. x35 = x34 + x33 mod 15 = 13 + 13 mod 15 = 26 mod 15 = 11. x39 = x38 + x37 mod 15 = 14 + 5 mod 15 = 19 mod 15 = 4. mod 15 = 18 mod 15 = 3. x9 = x8 + x7 mod 15 = 6 + 10 mod 15 = 16 mod 15 = 1. x33 = x32 + x31 mod 15 = 0 + 13 mod 15 = 13 mod 15 = 13. x3 = x2 + x1 mod 15 = 10 + 7 mod 15 = 17 mod 15 = 2. x34 = x33 + x32 mod 15 = 13 + 0 mod 15 = 13 mod 15 = 13. x41 = x40 + x39 mod 15 = 3 + 4 mod 15 = 7 mod 15 = 7. x4 = x3 + x2 mod 15 = 2 + 10 mod 15 = 12 mod 15 = 12 x5 = x4 + x3 mod 15 = 12 + 2 mod 15 = 14 mod 15 = 14. x6 = x5 + x4 mod 15 = 14 + 12 mod 15 = 26 mod 15 = 11. x15 = x14 + x13 mod 15 = 8 + 8 mod 15 = 16 mod 15 = 1. 247 . mod 15 = 19 mod 15 = 4. x10 = x9 + x8 mod 15 = 1 + 6 mod 15 = 7 mod 15 = 7. mod 15 = 6 mod 15 = 6. x16 = x15 + x14 mod 15 = 1 + 8 mod 15 = 9 mod 15 = 9.14 3. mod 15 = 12 mod 15 = 12. x13 = x12 + x11 mod 15 = 0 + 8 mod 15 = 8 mod 15 = 8. x31 = x30 + x29 mod 15 = 2 + 11 mod 15 = 13 mod 15 = 13. x37 = x36 + x35 mod 15 = 9 + 11 mod 15 = 20 mod 15 = 5. x36 = x35 + x34 mod 15 = 11 + 13 mod 15 = 24 mod 15 = 9. x14 = x13 + x12 mod 15 = 8 + 0 mod 15 = 8 mod 15 = 8. Thus this sequence is periodic with period 40. mod 15 = 11 mod 15 = 11. x7 = x6 + x5 mod 15 = 11 + 14 mod 15 = 25 mod 15 = 10 x8 = x7 + x6 mod 15 = 10 + 11 mod 15 = 21 mod 15 = 6. mod 15 = 17 mod 15 = 2. (a) We are given that x0 = 3 and x1 = 7. x38 = x37 + x36 mod 15 = 5 + 9 mod 15 = 14 mod 15 = 14. and finally: x40 = x0 and x41 = x1 . mod 15 = 16 mod 15 = 1. x40 = x39 + x38 mod 15 = 4 + 14 mod 15 = 18 mod 15 = 3. x17 = x16 + x15 mod 15 = 9 + 1 x18 = x17 + x16 mod 15 = 10 + 9 x19 = x18 + x17 mod 15 = 4 + 10 x20 = x19 + x18 mod 15 = 14 + 4 x21 = x20 + x19 mod 15 = 3 + 14 x22 = x21 + x20 mod 15 = 2 + 3 x23 = x22 + x21 mod 15 = 5 + 2 x24 = x23 + x22 mod 15 = 7 + 5 x25 = x24 + x23 mod 15 = 12 + 7 x26 = x25 + x24 mod 15 = 4 + 12 x27 = x26 + x25 mod 15 = 1 + 4 x28 = x27 + x26 mod 15 = 5 + 1 x29 = x28 + x27 mod 15 = 6 + 5 x30 = x29 + x28 mod 15 = 11 + 6 mod 15 = 10 mod 15 = 10. x32 = x31 + x30 mod 15 = 13 + 2 mod 15 = 15 mod 15 = 0.

the ⎛ 1 ⎞ period of the periodic point ⎜ . and so 12 ⎠ ⎝ 10 ⎞ ⎛ T 2 ( x. C 3 ⎜ ⎢ 101 ⎥ ⎟ = ⎢ ⎜ 0 ⎟ ⎣1 2⎥⎦ ⎢ 3 ⎥ ⎢ 8 ⎥ ⎝⎣ ⎦⎠ ⎣ 101 ⎦ ⎣ 101 ⎦ 5 ⎤ ⎡ 13 ⎤ ⎛ ⎡ 1 ⎤ ⎞ ⎡1 1 ⎤ ⎡ 101 ⎢ ⎥ mod 1 = ⎢ 101 ⎥ . (c) We have that ⎡ 1 ⎤ ⎛ ⎡ 1 ⎤ ⎞ ⎡1 1 ⎤ ⎡ 1 ⎤ 101 ⎥ mod 1 = ⎢ 101 ⎥ . 0 ⎟ must ⎝ 101 ⎠ be greater than 5. y) = ⎜ x + . Replacing n in this formula by n + 1 gives xn + 2 = xn +1 + xn mod p = ( xn + xn −1 ) + xn mod p = 2 xn + xn −1 mod p.Chapter 10: Elementary Linear Algebra SSM: Elementary Linear Algebra ⎡ x12 ⎤ ⎡1 1 ⎤ ⎡ 4 ⎤ ⎢ x ⎥ = ⎢1 2⎥ ⎢ 6 ⎥ mod 21 ⎦⎣ ⎦ ⎣ 13 ⎦ ⎣ ⎡10 ⎤ = ⎢ ⎥ mod 21 ⎣16 ⎦ ⎡10 ⎤ =⎢ ⎥ ⎣16 ⎦ (b) Step (ii) of the algorithm is xn +1 = xn + xn −1 mod p. ⎝ 12 ⎠ 248 . ⎡ x10 ⎤ ⎡1 1 ⎤ ⎡ 2 ⎤ ⎢ x ⎥ = ⎢1 2⎥ ⎢ 2 ⎥ mod 21 ⎦⎣ ⎦ ⎣ 11 ⎦ ⎣ ⎡4⎤ = ⎢ ⎥ mod 21 ⎣6⎦ ⎡4⎤ =⎢ ⎥ ⎣6⎦ 5. If 0 ≤ x < 1 and 0 ≤ y < 1. y ⎟ mod 1. then 5 ⎛ ⎞ T ( x. y ⎟ mod 1. C ⎜ ⎢ 101 ⎥ ⎟ = ⎢ ⎢ 1 ⎥ ⎜ 0 ⎟ ⎣1 2 ⎥⎦ 0 ⎢⎣ 101 ⎣ ⎦ ⎝⎣ ⎦⎠ ⎦ 1 2 ⎡ ⎤ ⎡ ⎤ 1 ⎛ ⎡ ⎤ ⎞ ⎡1 1 ⎤ 101 ⎢ 101 ⎥ . ⎢ ⎥ 1 mod C 2 ⎜ ⎢ 101 ⎥ ⎟ = ⎢ = 3 ⎥ ⎜ ⎟ 1 2 ⎥⎦ ⎢ 1 ⎥ ⎢⎣ 101 ⎝⎣ 0 ⎦⎠ ⎣ ⎣ 101 ⎦ ⎦ 2 5 ⎡ ⎤ ⎡ ⎤ 1 ⎛ ⎡ ⎤ ⎞ ⎡1 1 ⎤ 101 ⎢ ⎥ mod 1 = ⎢ 101 ⎥ . ⎦⎣ n ⎦ ⎣ n+2 ⎦ ⎣ ⎡ x14 ⎤ ⎡1 1 ⎤ ⎡10 ⎤ ⎢ x ⎥ = ⎢1 2 ⎥ ⎢16 ⎥ mod 21 ⎦⎣ ⎦ ⎣ 15 ⎦ ⎣ ⎡ 26 ⎤ = ⎢ ⎥ mod 21 ⎣ 42 ⎦ ⎡5 ⎤ =⎢ ⎥ ⎣0 ⎦ ⎡ x16 ⎤ ⎡1 1 ⎤ ⎡ 5 ⎤ ⎢ x ⎥ = ⎢1 2⎥ ⎢ 0 ⎥ mod 21 ⎦⎣ ⎦ ⎣ 17 ⎦ ⎣ ⎡5⎤ = ⎢ ⎥ mod 21 ⎣5⎦ ⎡5⎤ =⎢ ⎥ ⎣5⎦ ⎡ x ⎤ ⎡5 ⎤ (c) Beginning with ⎢ 0 ⎥ = ⎢ ⎥ . y) = ⎜ x + . C 4 ⎜ ⎢ 101 ⎥ ⎟ = ⎢ 21 ⎥ ⎜ ⎟ 1 2 ⎥⎦ ⎢ 8 ⎥ ⎢⎣ 101 ⎝⎣ 0 ⎦⎠ ⎣ ⎣ 101 ⎦ ⎦ 34 13 ⎡ ⎤ ⎛ ⎡ 1 ⎤ ⎞ ⎡1 1 ⎤ ⎡ 101 ⎤ ⎢ ⎥ mod 1 = ⎢ 101 ⎥ . These equations can be written as xn +1 = xn −1 + xn mod p xn + 2 = xn −1 + 2 xn mod p which in matrix form are ⎡ xn +1 ⎤ ⎡1 1 ⎤ ⎡ xn −1 ⎤ ⎢ x ⎥ = ⎢1 2⎥ ⎢ x ⎥ mod p. ⎣ x17 ⎦ ⎣ x1 ⎦ ⎡ x4 ⎤ ⎡1 1 ⎤ ⎡10 ⎤ ⎢ x ⎥ = ⎢1 2 ⎥ ⎢15⎥ mod 21 ⎦⎣ ⎦ ⎣ 5⎦ ⎣ ⎡ 25⎤ = ⎢ ⎥ mod 21 ⎣ 40 ⎦ ⎡4⎤ =⎢ ⎥ ⎣19 ⎦ ⎡ x6 ⎤ ⎡1 1 ⎤ ⎡ 4 ⎤ ⎢ x ⎥ = ⎢1 2 ⎥ ⎢19 ⎥ mod 21 ⎦⎣ ⎦ ⎣ 7⎦ ⎣ ⎡ 23⎤ = ⎢ ⎥ mod 21 ⎣ 42 ⎦ ⎡2⎤ =⎢ ⎥ ⎣0⎦ ⎡ x8 ⎤ ⎡1 1 ⎤ ⎡ 2 ⎤ ⎢ x ⎥ = ⎢1 2⎥ ⎢ 0 ⎥ mod 21 ⎦⎣ ⎦ ⎣ 9⎦ ⎣ ⎡2⎤ = ⎢ ⎥ mod 21 ⎣2⎦ ⎡2⎤ =⎢ ⎥ ⎣2⎦ 4. C 5 ⎜ ⎢ 101 ⎥ ⎟ = ⎢ ⎥ 55 ⎥ 21 ⎜ ⎟ 1 2⎦ ⎢ ⎥ ⎢ 101 ⎝⎣ 0 ⎦⎠ ⎣ ⎣ 101 ⎦ ⎣ ⎦ Because all five iterates are different. we obtain ⎣ x1 ⎦ ⎣5⎦ ⎡ x2 ⎤ ⎡1 1 ⎤ ⎡5⎤ ⎢ x ⎥ = ⎢1 2⎥ ⎢5⎥ mod 21 ⎦⎣ ⎦ ⎣ 3⎦ ⎣ ⎡10 ⎤ = ⎢ ⎥ mod 21 ⎣15⎦ ⎡10 ⎤ =⎢ ⎥ ⎣15⎦ ⎡x ⎤ ⎡x ⎤ and we see that ⎢ 16 ⎥ = ⎢ 0 ⎥ .

. its determinant is 1. we first ⎡0 ⎤ observe that the origin ⎢ ⎥ is a fixed point. ⎡3 2 ⎤ are The entries of the matrix ⎢ ⎣1 1 ⎥⎦ integers. has magnitude 1. and so both have magnitude 1. this cannot be the matrix of an Anosov automorphism. ⎝ 12 ⎠ 60 ⎞ ⎛ T 12 ( x. both of which have magnitude 1.14 15 ⎞ ⎛ T 3 ( x. this is the matrix of an Anosov automorphism. this is not the matrix of an Anosov automorphism. 2 3− 5 = 0. with 0 < x < 1 and 0 < y < 1) we obtain ⎡ 0 1⎤ ⎡ x ⎤ ⎡y⎤ ⎡ y ⎤ ⎢⎣ −1 0 ⎥⎦ ⎢⎣ y ⎥⎦ mod 1 = ⎢⎣ − x ⎥⎦ mod 1 = ⎢⎣1 − x ⎥⎦ . y) = ⎜ x + . The geometric effect of this transformation.3819. y) = ⎜ x + . is to rotate each point in the interior of S clockwise by 90° about the ⎡1⎤ center point ⎢ 2 ⎥ of S. Consequently. y ⎟ mod 1. has magnitude 1. ⎡ 0 1⎤ (c) The eigenvalues of the matrix ⎢ are ⎣ −1 0⎥⎦ ±i. ⎣0 ⎦ which can easily be verified. For points not in the interior of S. =⎢ ⎣1 − y ⎥⎦ ⎡ 0 1⎤ ⎡1 − x ⎤ ⎡ 1− y ⎤ ⎢⎣ −1 0 ⎥⎦ ⎢⎣1 − y ⎥⎦ mod 1 = ⎢⎣ −1 + x ⎥⎦ mod 1 ⎡1 − y ⎤ .. each ⎢⎣ 12 ⎥⎦ point in the interior of S has period 4 with ⎡1⎤ the exception of the center point ⎢ 2 ⎥ which ⎢⎣ 12 ⎥⎦ is a fixed point. and neither of its eigenvalues. Consequently.. By part (iii) of the definition of an Anosov automorphism. ⎡1 0 ⎤ The eigenvalues of the matrix ⎢ are ⎣0 1 ⎥⎦ both equal to 1. both of which have magnitude 1.. y ⎟ mod 1 ⎝ 12 ⎠ = ( x + 5. ⎡ 0 1⎤ ⎡ y ⎤ ⎡1 − x ⎤ ⎢⎣ −1 0 ⎥⎦ ⎢⎣1 − x ⎥⎦ mod 1 = ⎢⎣ − y ⎥⎦ mod 1 ⎡1 − x ⎤ . and so the mapping cannot be chaotic. By part (iii) of the definition.. ⎡0 1 ⎤ (b) The eigenvalues of the matrix ⎢ are ⎣1 0 ⎥⎦ ±1. ⎣5 2 ⎥⎦ 249 .. =⎢ ⎣ x ⎥⎦ ⎡ 0 1⎤ ⎡1 − y ⎤ ⎡ x ⎤ ⎢⎣ −1 0 ⎥⎦ ⎢⎣ x ⎥⎦ mod 1 = ⎢⎣ −1 + y ⎥⎦ mod 1 ⎡ x⎤ = ⎢ ⎥. ⎣ y⎦ Thus every point in the interior of S is a periodic point with period at most 4.6180. and neither of its eigenvalues.SSM: Elementary Linear Algebra Section 10. and so by part (ii) of the definition. we ⎣0 ⎦ obtain a 4-cycle ⎡ x⎤ ⎡ 0 ⎤ ⎡1 − x ⎤ ⎡0⎤ ⎡ x⎤ ⎢⎣ 0 ⎥⎦ → ⎢⎣1 − x ⎥⎦ → ⎢⎣ 0 ⎥⎦ → ⎢⎣ x ⎥⎦ → ⎢⎣ 0 ⎥⎦ if ⎡1 1 ⎤ 6. no point can have a dense set of iterates. 4 ± 15. (ii) the determinant is 1. y).. . Consequently.. this matrix is not the matrix of an Anosov automorphism. Starting with a ⎡ x⎤ point of the form ⎢ ⎥ with 0 < x < 1. By part (iii) of the definition. ⎢ . Because every point is a periodic point. its determinant is 1. ⎡5 7 ⎤ The entries of the matrix ⎢ are ⎣ 2 3⎥⎦ integers. ⎡ x⎤ Starting with an arbitrary point ⎢ ⎥ in the ⎣ y⎦ interior of S. this is the matrix of an Anosov automorphism. Thus every point in S returns to its original position after 12 iterations and so every point in S is a periodic point with period at most 12. do not have magnitude 2 1. y) mod 1 = ( x. ⎡6 2 ⎤ The determinant of the matrix ⎢ is 2. is ⎣1 2 ⎥⎦ one in which (i) the entries are all integers. (a) The matrix of Arnold’s cat map. and eigenvalues. this is not the matrix of an Anosov automorphism. The three conditions of an Anosov automorphism are thus satisfied. 2 ± 3. as seen by the iterates. and (iii) the 3+ 5 = 2. (that is.

1) I' IV' ( 12 . + = 2 ⎥⎦ ⎢⎣ y3 ⎥⎦ ⎢⎣ b ⎥⎦ ⎢⎣1⎥⎦ ⎡1 1 ⎤ ⎡ 2 −1⎤ The inverse of the matrix ⎢ is ⎢ . we wish to find the regions in S that map onto the four indicated regions in the figure below. 0) For region II ′. 2 ) II 1 III (1. 1) 9. Finally because no point in S can have a dense set of iterates. respectively. ⎜ . and (1. 0). and ( x3 . otherwise we obtain the 2-cycle 2 ⎡1⎤ ⎡0⎤ ⎡1⎤ ⎢2⎥ → ⎢1⎥ → ⎢2⎥. and (1. and (1. We seek points ( x1 . This then gives a = b = 0 and the ⎡ x ⎤ ⎡1 1 ⎤ ⎡ x ⎤ mapping ⎢ ⎥ → ⎢ maps the three ⎣ y ⎦ ⎣1 2 ⎥⎦ ⎢⎣ y ⎥⎦ ⎛ 1⎞ points (0. 0). and (1. 1) 1 ⎤ ⎡ x1 ⎤ ⎡ a ⎤ ⎡0 ⎤ + = . Notice ⎣ d ⎦ ⎣ −1 1⎥⎦ ⎢⎣ b ⎥⎦ that c and d must be integers. with entries that lie in [0.Chapter 10: Elementary Linear Algebra SSM: Elementary Linear Algebra ⎡1 ⎢⎣1 ⎡1 ⎢⎣1 ⎡1 ⎢⎣1 1 . y2 ). otherwise we obtain the 2-cycle 2 ⎡0⎤ ⎡1⎤ ⎡0⎤ ⎢ 1 ⎥ → ⎢ 2 ⎥ → ⎢ 1 ⎥ . ⎦⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ 1⎦ ⎣ 1 0⎤ ⎡ c ⎤ ⎡ ⎤ ⎡ x ⎡ 2 ⎤ ⎡ 2 −1⎤ 2 ⎡ c ⎤ ⎢ y ⎥ = ⎢ −1 1⎥ ⎢ ⎥ − ⎢ d ⎥ = ⎢ 1 ⎥ − ⎢ d ⎥ . 0) (1. 2. we obtain a 4-cycle ⎡0⎤ ⎡ y⎤ ⎡ 0 ⎤ ⎡1 − y ⎤ ⎡ 0 ⎤ ⎢⎣ y ⎥⎦ → ⎢⎣ 0 ⎥⎦ → ⎢⎣1 − y ⎥⎦ → ⎢⎣ 0 ⎥⎦ → ⎢⎣ y ⎥⎦ if 1 y ≠ . that map onto these three points under the ⎡ x ⎤ ⎡1 1 ⎤ ⎡ x ⎤ ⎡ a ⎤ for certain mapping ⎢ ⎥ → ⎢ + ⎣ y ⎦ ⎣1 2 ⎥⎦ ⎢⎣ y ⎥⎦ ⎢⎣ b ⎥⎦ integer values of a and b to be determined. which then maps onto the region I ′. 2 ⎥⎦ ⎢⎣ y2 ⎥⎦ ⎢⎣ b ⎥⎦ ⎢⎣ 1 ⎥⎦ 1 ⎤ ⎡ x3 ⎤ ⎡ a ⎤ ⎡1⎤ . 1). 1⎟ . This leads to ⎡ x1 ⎤ ⎡ 2 −1⎤ ⎡0 ⎤ ⎡ c ⎤ ⎡c⎤ ⎢ y ⎥ = ⎢ −1 1⎥ ⎢0 ⎥ − ⎢ d ⎥ = − ⎢ d ⎥ . y1 ). ⎟ . the calculations are as follows: 250 . 0). As per the hint. ⎝2 ⎠ ( x2 . 0) 1 ⎤ ⎡ x2 ⎤ ⎡ a ⎤ ⎡ 12 ⎤ + = . 1) III' (0. 0) define ⎝ 2⎠ the triangular region labeled I in the diagram below. ⎜ 0. 0) to the three ⎝ 2⎠ ⎛1 ⎞ points (0. Thus every point not in ⎣2⎦ ⎣0⎦ ⎣2⎦ the interior of S is a periodic point with 1. 1]. or 4. y3 ). 2 ⎥⎦ ⎢⎣ y2 ⎥⎦ ⎢⎣ b ⎥⎦ ⎢⎣0 ⎥⎦ II' (1. 2 ) I (0. 0) (1. This leads to the three equations IV 1 (0. ⎟ . ⎛1 ⎞ ⎜ . 0). x≠ 1 ( 2 . ⎣0⎦ ⎣2⎦ ⎣0⎦ Similarly. ⎜ 0. 0) We first consider region I ′ with vertices (0. 1) (1. (0. 1). ⎥ 1 2 ⎣ ⎦ ⎣ −1 1⎥⎦ We multiply the above three matrix equations by ⎡ c ⎤ ⎡ 2 −1⎤ ⎡ a ⎤ this inverse and set ⎢ ⎥ = ⎢ . ⎦ ⎣1⎦ ⎣ ⎦ ⎣ 2 ⎦ ⎣ ⎦ ⎣ 2⎦ ⎣ ⎡ x3 ⎤ ⎡ 2 −1⎤ ⎡1⎤ ⎡ c ⎤ ⎡1 ⎤ ⎡ c ⎤ ⎢ ⎥=⎢ ⎥ ⎢ ⎥ − ⎢ ⎥ = ⎢ ⎥ − ⎢ ⎥. ⎣ y3 ⎦ ⎣ −1 1⎦ ⎣1⎦ ⎣ d ⎦ ⎣0 ⎦ ⎣ d ⎦ The only integer values of c and d that will give values of xi and yi in the interval [0. 1] are c = d = 0. starting with a point of the form ⎡0⎤ ⎢⎣ y ⎥⎦ with 0 < y < 1. (0. it follows that the mapping cannot be chaotic. ⎝2 ⎠ ⎛ 1⎞ The three points (0. 1⎟ .

(0. 5 5 ⎛2 1⎞ ⎛3 4⎞ We can then check that ⎜ . ⎦⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ 1⎦ ⎣ ⎡1 1 ⎤ ⎡ x2 ⎤ ⎡ a ⎤ ⎡1⎤ ⎢⎣1 2 ⎥⎦ ⎢ y ⎥ + ⎢⎣ b ⎥⎦ = ⎢⎣1⎥⎦ . ⎦⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ 3⎦ ⎣ Only c = −1 and d = 0 will work. ⎣ 3⎦ ⎣ ⎦ ⎡ x1 ⎤ ⎡ 2 −1⎤ ⎡0 ⎤ ⎡ c ⎤ ⎡c⎤ ⎢ y ⎥ = ⎢ −1 1⎥ ⎢0 ⎥ − ⎢ d ⎥ = − ⎢ d ⎥ . 1). 5 5 3 4 r = 3 and s = 5. ⎦⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ 3⎦ ⎣ Only c = 1 and d = −1 will work. ⎟ ⎝5 5⎠ ⎝5 5⎠ ⎛1 3⎞ ⎛4 2⎞ form one 2-cycle and ⎜ . which give x0 = and y0 = . ⎣ 1⎦ 12. ⎦ ⎣1⎦ ⎣ ⎦ ⎣ 2 ⎦ ⎣ ⎦ ⎣ 2⎦ ⎣ ⎡ x3 ⎤ ⎡ 2 −1⎤ ⎡0 ⎤ ⎡ c ⎤ ⎡ −1⎤ ⎡ c ⎤ ⎢ y ⎥ = ⎢ −1 1⎥ ⎢1 ⎥ − ⎢ d ⎥ = ⎢ 1⎥ − ⎢ d ⎥ . b = −1 and the mapping ⎡ x ⎤ ⎡1 1⎤ ⎡ x ⎤ ⎡ 0 ⎤ ⎢⎣ y ⎥⎦ → ⎢⎣1 2 ⎥⎦ ⎢⎣ y ⎥⎦ + ⎢⎣ −1⎥⎦ maps region II with ⎛ 1⎞ vertices ⎜ 0. and ⎜1. This leads to a = −1. ⎣ y3 ⎦ ⎣ −1 1⎦ ⎣ 0 ⎦ ⎣ d ⎦ ⎣ − 2 ⎦ ⎣ d ⎦ Only c = 0 and d = −1 will work. 0).. ⎣ 2⎦ ⎡1 1 ⎤ ⎡ x3 ⎤ ⎡ a ⎤ ⎡1 ⎤ ⎢⎣1 2⎥⎦ ⎢ y ⎥ + ⎢⎣ b ⎥⎦ = ⎢⎣0 ⎥⎦ . This leads to a = 0. which give x0 = and y0 = . 1. ⎡ x3 ⎤ ⎡ 2 −1⎤ ⎡1 ⎤ ⎡ c ⎤ ⎡ 2⎤ ⎡ c ⎤ ⎢ y ⎥ = ⎢ −1 1⎥ ⎢0 ⎥ − ⎢ d ⎥ = ⎢ −1⎥ − ⎢ d ⎥ . ⎣ 1⎦ 251 . ⎦⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ 2⎦ ⎣ ⎡ x2 ⎤ ⎡ 2 −1⎤ ⎡1⎤ ⎡ c ⎤ ⎡1 ⎤ ⎡ c ⎤ ⎢ y ⎥ = ⎢ −1 1⎥ ⎢1⎥ − ⎢ d ⎥ = ⎢ 0⎥ − ⎢ d ⎥ . then r = 1 and s = 0. ⎟ form ⎝5 5⎠ ⎝5 5⎠ another 2-cycle.. ⎟ and ⎜ . ⎣ 1⎦ ⎣ ⎦ ⎡1 1 ⎤ ⎡ x2 ⎤ ⎡ a ⎤ ⎡1⎤ ⎢⎣1 2 ⎥⎦ ⎢ y ⎥ + ⎢⎣ b ⎥⎦ = ⎢⎣1⎥⎦ . ⎣ 3⎦ ⎡ x1 ⎤ ⎡ 2 −1⎤ ⎡ 12 ⎤ ⎡ c ⎤ ⎡ 1 ⎤ ⎡ c ⎤ ⎢ y ⎥ = ⎢ −1 1⎥ ⎢ ⎥ − ⎢ d ⎥ = ⎢ − 1 ⎥ − ⎢ d ⎥ . For region IV ′. 0) onto region ⎝ 2⎠ ′ II . ⎣ 2⎦ ⎡1 1 ⎤ ⎡ x3 ⎤ ⎡ a ⎤ ⎡ 12 ⎤ ⎢⎣1 2 ⎥⎦ ⎢ y ⎥ + ⎢⎣ b ⎥⎦ = ⎢ 0 ⎥ . 5 5 1 3 r = 2 and s = 3. etc.SSM: Elementary Linear Algebra Section 10. As per the hint.14 ⎡1 1 ⎤ ⎡ x1 ⎤ ⎡ a ⎤ ⎡ 12 ⎤ ⎢⎣1 2⎥⎦ ⎢ y ⎥ + ⎢⎣ b ⎥⎦ = ⎢ 0 ⎥ . b = −2 and the mapping ⎡ x ⎤ ⎡1 1 ⎤ ⎡ x ⎤ ⎡ −1⎤ ⎢⎣ y ⎥⎦ → ⎢⎣1 2 ⎥⎦ ⎢⎣ y ⎥⎦ + ⎢⎣ −2 ⎥⎦ maps region IV with ⎛ 1⎞ vertices (0. This leads to a = −1. This equation can be ⎡1 3 ⎤ ⎡ x0 ⎤ ⎡ r ⎤ rewritten as ⎢ . . 5 5 4 2 r = 2 and s = 4. 1) are: 2 1 r = 1 and s = 2. ⎣ 2⎦ ⎣ ⎦ ⎡1 1 ⎤ ⎡ x3 ⎤ ⎡ a ⎤ ⎡0 ⎤ ⎢⎣1 2⎥⎦ ⎢ y ⎥ + ⎢⎣ b ⎥⎦ = ⎢⎣1 ⎥⎦ . 1. which has the ⎢ ⎥= ⎣3 4 ⎥⎦ ⎣ y0 ⎦ ⎢⎣ s ⎥⎦ −4r + 3s 3r − s and y0 = . ⎟ ... and (1. 1). For region III ′. 1) onto region ⎝ 2⎠ III ′. ⎟ onto region ⎝ 2⎠ IV ′. ⎟ . and r and s are nonnegative integers. we find that the only values of r and s that yield values of x0 and y0 lying in [0. ⎟ and ⎜ . 1). 2. 2. ⎡1 1 ⎤ ⎡ x2 ⎤ ⎡ a ⎤ ⎡ 12 ⎤ ⎢⎣1 2 ⎥⎦ ⎢ y ⎥ + ⎢⎣ b ⎥⎦ = ⎢ 1 ⎥ . ⎦⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ 1⎦ ⎣ 1 ⎡ x2 ⎤ ⎡ 2 −1⎤ ⎡ 2 ⎤ ⎡ c ⎤ ⎡ 0 ⎤ ⎡ c ⎤ ⎢ y ⎥ = ⎢ −1 1⎥ ⎢ ⎥ − ⎢ d ⎥ = ⎢ 1 ⎥ − ⎢ d ⎥ . (1. which give x0 = and y0 = . . the calculations are as follows: ⎡1 1 ⎤ ⎡ x1 ⎤ ⎡ a ⎤ ⎡0 ⎤ ⎢⎣1 2⎥⎦ ⎢ y ⎥ + ⎢⎣ b ⎥⎦ = ⎢⎣0 ⎥⎦ . the calculations are as follows: ⎡1 1 ⎤ ⎡ x1 ⎤ ⎡ a ⎤ ⎡0 ⎤ ⎢⎣1 2⎥⎦ ⎢ y ⎥ + ⎢⎣ b ⎥⎦ = ⎢⎣0 ⎥⎦ . First solution x0 = 5 5 trying r = 0 and s = 0. which give x0 = and y0 = . ⎦ ⎣0⎦ ⎣ ⎦ ⎣ 2 ⎦ ⎣ ⎦ ⎣ 1⎦ ⎣ ⎡ x2 ⎤ ⎡ 2 −1⎤ ⎡1⎤ ⎡ c ⎤ ⎡1 ⎤ ⎡ c ⎤ ⎢ y ⎥ = ⎢ −1 1⎥ ⎢1⎥ − ⎢ d ⎥ = ⎢ 0⎥ − ⎢ d ⎥ . and (0. ⎣ 3⎦ ⎡ x1 ⎤ ⎡ 2 −1⎤ ⎡0 ⎤ ⎡ c ⎤ ⎡c⎤ ⎢ y ⎥ = ⎢ −1 1⎥ ⎢0 ⎥ − ⎢ d ⎥ = − ⎢ d ⎥ . we want to find all solutions of ⎡ x ⎤ ⎡ 2 3⎤ ⎡ x0 ⎤ ⎡ r ⎤ the matrix equation ⎢ 0 ⎥ = ⎢ ⎥⎢ ⎥−⎢ ⎥ ⎣ y0 ⎦ ⎣ 3 5⎦ ⎣ y0 ⎦ ⎣ s ⎦ where 0 ≤ x0 < 1.. ⎜1.. ⎦⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ 2⎦ ⎣ ⎡ x3 ⎤ ⎡ 2 −1⎤ ⎡ 12 ⎤ ⎡ c ⎤ ⎡ 1 ⎤ ⎡ c ⎤ ⎢ ⎥=⎢ ⎥ ⎢ ⎥ − ⎢ ⎥ = ⎢ 1⎥ − ⎢ ⎥.. 0 ≤ y0 < 1. b = −1 and the mapping ⎡ x ⎤ ⎡1 1 ⎤ ⎡ x ⎤ ⎡ −1⎤ ⎢⎣ y ⎥⎦ → ⎢⎣1 2 ⎥⎦ ⎢⎣ y ⎥⎦ + ⎢⎣ −1⎥⎦ maps region III with ⎛ 1⎞ vertices (1.

15 Exercise Set 10. First group the plaintext into pairs.3. which is not divisible by 2 or 13. ⎣ 1 2 ⎥⎦ reducing everything mod 26. =⎢ ⎣0 1⎥⎦ ⎡3 1⎤ (b) For A = ⎢ . 2. The next application of the mapping will return you to the letter ‘A’ with the pixels for the letters ‘B’ through ‘E’ scattered in the background. 252 .3. Therefore by Corollary 10. DA RK NI GH TT 4 1 18 11 14 9 7 8 20 20 ⎡ 1 3⎤ (a) For the enciphering matrix A = ⎢ . which ⎣5 3⎥⎦ is divisible by 2.15. ⎡12 7 ⎤ ⎡ 9 1⎤ A−1 A = ⎢ ⎣ 23 15⎥⎦ ⎢⎣ 7 2 ⎥⎦ ⎡157 26 ⎤ =⎢ ⎣312 53⎥⎦ ⎡ 1 0⎤ (mod 26). ⎡9 1 ⎤ . ⎣ 1 9 ⎥⎦ det(A) = 72 − 11 = 61 = 9 (mod 26). =⎢ ⎣ 23 15⎥⎦ Checking: ⎡ 9 1⎤ ⎡12 7 ⎤ AA−1 = ⎢ ⎣ 7 2⎥⎦ ⎢⎣ 23 15⎥⎦ ⎡131 78⎤ =⎢ ⎣130 79 ⎥⎦ ⎡ 1 0⎤ (mod 26) =⎢ ⎣ 0 1⎥⎦ Section 10. Apply the mapping T again and then superimpose the letter ‘C’ in black pixels onto the resulting image. Apply the mapping T to this image. we have ⎡ 1 3⎤ ⎡ 4 ⎤ ⎡ 7 ⎤ G ⎢⎣ 2 1⎥⎦ ⎢⎣ 1 ⎥⎦ = ⎢⎣9 ⎥⎦ I ⎡1 3⎤ ⎡18⎤ ⎡ 51⎤ ⎡ 25⎤ Y ⎢⎣ 2 1⎥⎦ ⎢⎣11⎥⎦ = ⎢⎣ 47 ⎥⎦ = ⎢⎣ 21⎥⎦ U O ⎡1 3⎤ ⎡14 ⎤ ⎡ 41⎤ ⎡15⎤ ⎢⎣ 2 1⎥⎦ ⎢⎣ 9 ⎥⎦ = ⎢⎣37 ⎥⎦ = ⎢⎣11⎥⎦ K ⎡ 1 3⎤ ⎡ 7 ⎤ ⎡ 31⎤ ⎡ 5⎤ E ⎢⎣ 2 1⎥⎦ ⎢⎣ 8 ⎥⎦ = ⎢⎣ 22 ⎥⎦ = ⎢⎣ 22 ⎥⎦ V ⎡1 3⎤ ⎡ 20 ⎤ ⎡80 ⎤ ⎡ 2 ⎤ B ⎢⎣ 2 1⎥⎦ ⎢⎣ 20 ⎥⎦ = ⎢⎣60 ⎥⎦ = ⎢⎣ 8 ⎥⎦ H The Hill cipher is GIYUOKEVBH. A is invertible. =⎢ ⎣ 23 24 ⎥⎦ Checking: ⎡4 3⎤ (b) For the enciphering matrix A = ⎢ .15.Chapter 10: Elementary Linear Algebra SSM: Elementary Linear Algebra 14. Therefore by Corollary 10. From (2): ⎡ 9 −11⎤ A−1 = (9) −1 ⎢ 8⎥⎦ ⎣ −1 ⎡ 9 −11⎤ = 3⎢ 8⎥⎦ ⎣ −1 ⎡ 27 −33⎤ =⎢ ⎣ −3 24 ⎥⎦ ⎡ 1 19 ⎤ (mod 26). Repeat this procedure with the letter ‘D’ and ‘E’.3. A is not invertible. A is invertible. Begin with a 101 × 101 array of white pixels and add the letter ‘A’ in black pixels to it. ⎣ 2 1⎥⎦ reducing everything mod 26. add the dummy letter T. (a) For A = ⎢ ⎣ 7 2 ⎥⎦ which is not divisible by 2 or 13. we have ⎡ 4 3 ⎤ ⎡ 4 ⎤ ⎡19 ⎤ S ⎢⎣ 1 2 ⎥⎦ ⎢⎣ 1⎥⎦ = ⎢⎣ 6 ⎥⎦ F ⎡ 4 3⎤ ⎡18⎤ ⎡105⎤ ⎡ 1⎤ A ⎢⎣1 2⎥⎦ ⎢⎣11⎥⎦ = ⎢⎣ 40 ⎥⎦ = ⎢⎣14 ⎥⎦ N E ⎡ 4 3 ⎤ ⎡14 ⎤ ⎡ 83⎤ ⎡ 5⎤ ⎢⎣1 2 ⎥⎦ ⎢⎣ 9 ⎥⎦ = ⎢⎣32⎥⎦ = ⎢⎣ 6 ⎥⎦ F Z ⎡ 4 3 ⎤ ⎡ 7 ⎤ ⎡52⎤ ⎡ 0 ⎤ ⎢⎣ 1 2 ⎥⎦ ⎢⎣ 8⎥⎦ = ⎢⎣ 23⎥⎦ = ⎢⎣ 23⎥⎦ W J ⎡ 4 3 ⎤ ⎡ 20 ⎤ ⎡140 ⎤ ⎡10 ⎤ ⎢⎣1 2 ⎥⎦ ⎢⎣ 20 ⎥⎦ = ⎢⎣ 60 ⎥⎦ = ⎢⎣ 8⎥⎦ H The Hill cipher is SFANEFZWJH. From Eqation (2): ⎡ 2 −1⎤ A−1 = (11)−1 ⎢ ⎣ −7 9 ⎥⎦ ⎡ 2 −1⎤ = 19 ⎢ ⎣ −7 9 ⎥⎦ ⎡ 38 −19⎤ =⎢ ⎣ −133 171⎥⎦ ⎡12 7 ⎤ (mod 26). det(A) = 18 − 7 = 11. det(A) = 9 − 5 = 4. Then superimpose the letter ‘B’ in black pixels onto this image.15.15 1. and get the numerial equivalents from Table 1. which will scatter the black pixels throughout the image. ⎡8 11⎤ (c) For A = ⎢ . Therefore by Corollary 10.

det(A) = 3 − 8 = −5 = 21 ⎣1 3⎥⎦ (mod 26). det(A) = 6 − 6 = 0.15. multiply each ciphertext ⎡ 3 1⎤ (e) For A = ⎢ . ⎡16 5⎤ ⎡19 ⎤ ⎡309 ⎤ ⎡ 23⎤ W ⎢⎣15 6 ⎥⎦ ⎢⎣ 1⎥⎦ = ⎢⎣ 291⎥⎦ = ⎢⎣ 5⎥⎦ E L ⎡16 5 ⎤ ⎡11⎤ ⎡ 246 ⎤ ⎡12⎤ ⎢⎣15 6 ⎥⎦ ⎢⎣14 ⎥⎦ = ⎢⎣ 249 ⎥⎦ = ⎢⎣15⎥⎦ O ⎡16 5 ⎤ ⎡ 15⎤ ⎡360⎤ ⎡ 22⎤ V ⎢⎣15 6 ⎥⎦ ⎢⎣ 24 ⎥⎦ = ⎢⎣369⎥⎦ = ⎢⎣ 5⎥⎦ E ⎡16 5 ⎤ ⎡ 1⎤ ⎡ 91⎤ ⎡13⎤ M ⎢⎣15 6 ⎥⎦ ⎢⎣15⎥⎦ = ⎢⎣105⎥⎦ = ⎢⎣ 1⎥⎦ A ⎡16 5 ⎤ ⎡ 10 ⎤ ⎡ 280 ⎤ ⎡ 20 ⎤ T ⎢⎣15 6 ⎥⎦ ⎢⎣ 24 ⎥⎦ = ⎢⎣ 294 ⎥⎦ = ⎢⎣ 8⎥⎦ H The plaintext is thus WE LOVE MATH.4. =⎢ ⎣ 21 5⎥⎦ Checking: vector by A−1 and reduce the results modulo 26. From Table 1 the numerical equivalent of this ciphertext is 19 1 11 14 15 24 1 15 10 24 ⎡ 4 1⎤ Now we have to find the inverse of A = ⎢ . ⎣ 3 2 ⎥⎦ Since det(A) = 8 − 3 = 5. ⎣ 1 7 ⎥⎦ which is divisible by 13. From (2): ⎡ 3 −8 ⎤ A−1 = (21)−1 ⎢ ⎣ −1 1⎥⎦ ⎡ 3 −8 ⎤ = 5⎢ ⎣ −1 1⎥⎦ ⎡ 15 −40 ⎤ =⎢ 5⎥⎦ ⎣ −5 ⎡15 12 ⎤ (mod 26). A is not invertible. ⎡1 8⎤ (f) For A = ⎢ .15. From Table 1 the numerical equivalent of the known plaintext is AR MY 1 18 13 25 and the numerical equivalent of the corresponding ciphertext is 253 . Therefore by Corollary 10. so that ⎣6 2 ⎥⎦ A is not invertible by Corollary 10. =⎢ ⎣15 6 ⎥⎦ To obtain the plaintext.4. det(A) = 14 − 1 = 13. =⎢ ⎣0 1⎥⎦ ⎡15 A−1 A = ⎢ ⎣ 21 ⎡ 27 =⎢ ⎣ 26 ⎡1 =⎢ ⎣0 ⎡ 2 1⎤ (d) For A = ⎢ . Therefore by Corollary 10. 4. 12 ⎤ ⎡1 8⎤ 5⎥⎦ ⎢⎣1 3⎥⎦ 156⎤ 183⎥⎦ 0⎤ (mod 26). we have by (2): ⎡ 2 −1⎤ A−1 = (5)−1 ⎢ ⎣ −3 4 ⎥⎦ ⎡ 2 −1⎤ = 21 ⎢ ⎣ −3 4 ⎥⎦ ⎡ 42 −21⎤ =⎢ ⎣ −63 84⎥⎦ ⎡16 5⎤ (mod 26).4. A is invertible. 1⎥⎦ 3.15.15 ⎡8 11⎤ ⎡ 1 19 ⎤ AA−1 = ⎢ ⎣ 1 9⎥⎦ ⎢⎣ 23 24 ⎥⎦ ⎡ 261 416 ⎤ =⎢ ⎣ 208 235⎥⎦ ⎡ 1 0⎤ (mod 26) =⎢ ⎣ 0 1⎥⎦ ⎡1 8⎤ ⎡15 12⎤ AA−1 = ⎢ ⎣1 3⎥⎦ ⎢⎣ 21 5⎥⎦ ⎡183 52 ⎤ =⎢ ⎣ 78 27 ⎥⎦ ⎡ 1 0⎤ (mod 26) =⎢ ⎣ 0 1⎥⎦ ⎡ 1 19 ⎤ ⎡8 11⎤ A−1 A = ⎢ ⎣ 23 24 ⎥⎦ ⎢⎣ 1 9 ⎥⎦ ⎡ 27 182 ⎤ =⎢ ⎣ 208 469 ⎥⎦ ⎡ 1 0⎤ (mod 26).SSM: Elementary Linear Algebra Section 10. which is not divisible by 2 or 13.

⎡ 1 132 11 198⎤ ⎢⎣8 11 13 25⎥⎦ 5. ⎣15 13⎥⎦ The calculations are as follows: ⎡10 25 1 20⎤ Form the matrix ⎢⎣17 15 15 13⎥⎦ [C P ] .Chapter 10: Elementary Linear Algebra SSM: Elementary Linear Algebra Since det( A−1 ) = 35 − 90 = −55 = 23(mod 26). 11 16 ⎤ ⎡1 2 ⎢⎣0 −5 −75 −103⎥⎦ −8 times the first row to the second. Add −2 times the second row to the first. ⎡cT ⎤ ⎡19 12 ⎤ We want to reduce C = ⎢ 1T ⎥ = ⎢ ⎥ to I by ⎢⎣c2 ⎥⎦ ⎣ 8 11⎦ elementary row operations and simultaneously ⎡pT ⎤ ⎡ 1 18⎤ apply these operations to P = ⎢ 1T ⎥ = ⎢ ⎥. ⎡ 1 2 11 16 ⎤ ⎢⎣8 11 13 25⎥⎦ Replace 132 and 198 by their residues modulo 26. ⎡ 1 0 −19 6 ⎤ ⎢⎣0 1 15 5⎥⎦ the first (since 10−1 does not exist mod 26). ⎡ 1 2 11 16 ⎤ Replace the entries in ⎢⎣0 21 3 1⎥⎦ the second row by their residues modulo 26. matrix is A−1 = ⎢ ⎣ 6 5⎥⎦ 254 . 1 2 11 16 ⎡ ⎤ Multiply the second row ⎢⎣0 1 15 5⎥⎦ ⎡ 27 40 16 33⎤ ⎢⎣ 17 15 15 13⎥⎦ by 21−1 = 5 (mod 26). residue modulo 26. ⎢⎣p 2 ⎥⎦ ⎣13 25⎦ The calculations are as follows: ⎡19 12 1 18⎤ Form the matrix ⎢⎣ 8 11 13 25⎥⎦ [C P ] . 1 14 16 7 ⎡ ⎤ ⎢⎣0 −223 −257 −106 ⎥⎦ Add −17 times the first row to the second. SL HK 19 12 8 11 so the corresponding plaintext and ciphertext vectors are ⎡ 1⎤ ⎡19 ⎤ p1 = ⎢ ⎥ ↔ c1 = ⎢ ⎥ 18 ⎣ ⎦ ⎣12 ⎦ ⎡ 13⎤ ⎡ 8⎤ p 2 = ⎢ ⎥ ↔ c2 = ⎢ ⎥ 25 ⎣ ⎦ ⎣11⎦ A = ( A−1 )−1 ⎡ 5 −15⎤ = 23−1 ⎢ 7 ⎥⎦ ⎣ −6 ⎡ 5 −15⎤ = 17 ⎢ 7 ⎥⎦ ⎣ −6 ⎡ 85 −255⎤ =⎢ ⎣ −102 119 ⎥⎦ ⎡7 5⎤ (mod 26) =⎢ ⎣ 2 15⎥⎦ is the enciphering matrix. ⎡ 7 6⎤ so the deciphering Thus ( A−1 )T = ⎢ ⎣15 5⎥⎦ ⎡7 15⎤ (mod 26). From Table 1 the numerical equivalent of the known plaintext is AT OM 1 20 15 13 and the numerical equivalent of the corresponding ciphertext is JY QO 10 25 17 15 The corresponding plaintext and ciphertext vectors are: ⎡ 1⎤ ⎡ 10 ⎤ p1 = ⎢ ⎥ ↔ c1 = ⎢ ⎥ ⎣ 20 ⎦ ⎣ 25⎦ ⎡15⎤ ⎡17 ⎤ p2 = ⎢ ⎥ ↔ c2 = ⎢ ⎥ ⎣13⎦ ⎣15⎦ ⎡10 25⎤ We want to reduce C = ⎢ to I by ⎣17 15 ⎥⎦ elementary row operations and simultaneously ⎡ 1 20 ⎤ apply these operations to P = ⎢ . ⎡ 1 0 7 6⎤ ⎢⎣0 1 15 5⎥⎦ Add the second row to ⎡ 1 14 16 7 ⎤ ⎢⎣17 15 15 13⎥⎦ Replace −19 by its Replace the entries in the first row by their residues modulo 26. Multiply the first row by 19−1 = 11 (mod 26).

⎡ 1 14 16 7 ⎤ ⎢⎣0 1 5 14 ⎥⎦ Replace the entries in the second row by their residues modulo 26.SSM: Elementary Linear Algebra ⎡ 1 14 16 7 ⎤ ⎢⎣0 11 3 24 ⎥⎦ Section 10. From Table 1 the numerical equivalents of the known plaintext are I H A V E C O M E 9 8 1 22 5 3 15 13 5 and the numerical equivalent of the corresponding ciphertext are H P A F Q G G D U 8 16 1 6 17 7 7 4 21 The corresponding plaintext and ciphertext vectors are ⎡9 ⎤ ⎡ 8⎤ p1 = ⎢ 8⎥ ↔ c1 = ⎢16 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 1⎦ ⎣ 1⎦ ⎡ 22 ⎤ ⎡ 6⎤ p 2 = ⎢ 5⎥ ↔ c2 = ⎢17 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 3⎦ ⎣ 7⎦ ⎡15⎤ ⎡ 7⎤ p3 = ⎢13⎥ ↔ c3 = ⎢ 4 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 5⎦ ⎣ 21⎦ ⎡ 8 16 1⎤ We want to reduce C = ⎢ 6 17 7 ⎥ to I by ⎢ ⎥ ⎣7 4 21⎦ elementary row operations and simultaneously ⎡ 9 8 1⎤ apply these operations to P = ⎢ 22 5 3⎥ . ⎢ ⎥ ⎣ 15 13 5⎦ The calculations are as follows: ⎡ 8 16 1 9 8 1⎤ ⎢ 6 17 7 22 5 3⎥ Form the matrix ⎢ ⎥ ⎣7 4 21 15 13 5⎦ Multiply the second row by 11−1 = 19 (mod 26). Replace the entries in the second row by their residues modulo 26. 255 .15 which yields the message THEY SPLIT THE ATOM. ⎡15 20 22 24 21 6 ⎤ ⎢ 6 17 7 22 5 3⎥ ⎢ ⎥ ⎣ 7 4 21 15 13 5⎦ Add the third row to the first since 8−1 does not exist modulo 26. ⎡ 24 19 ⎤ . Since we want a Hill 3-cipher. we multiply each ciphertext vector by A−1: ⎡ 24 5⎤ ⎡12 ⎤ ⎡ 358⎤ ⎡ 20 ⎤ ⎢⎣ 19 14 ⎥⎦ ⎢⎣14 ⎥⎦ = ⎢⎣ 424 ⎥⎦ = ⎢⎣ 8⎥⎦ ⎡ 24 5⎤ ⎡7 ⎤ ⎡ 213⎤ ⎡ 5⎤ ⎢⎣ 19 14 ⎥⎦ ⎢⎣ 9 ⎥⎦ = ⎢⎣ 259 ⎥⎦ = ⎢⎣ 25⎥⎦ ⎡ 24 5⎤ ⎡ 8⎤ ⎡ 227 ⎤ ⎡19 ⎤ ⎢⎣ 19 14 ⎥⎦ ⎢⎣7 ⎥⎦ = ⎢⎣ 250 ⎥⎦ = ⎢⎣16 ⎥⎦ ⎡ 24 5⎤ ⎡ 25⎤ ⎡610 ⎤ ⎡12 ⎤ ⎢⎣ 19 14 ⎥⎦ ⎢⎣ 2 ⎥⎦ = ⎢⎣ 503⎥⎦ = ⎢⎣ 9 ⎥⎦ ⎡ 24 5⎤ ⎡ 22 ⎤ ⎡ 618⎤ ⎡ 20 ⎤ ⎢⎣ 19 14 ⎥⎦ ⎢⎣ 18⎥⎦ = ⎢⎣ 670 ⎥⎦ = ⎢⎣ 20 ⎥⎦ ⎡ 24 5⎤ ⎡ 5⎤ ⎡190⎤ ⎡8⎤ ⎢⎣ 19 14 ⎥⎦ ⎢⎣14 ⎥⎦ = ⎢⎣ 291⎥⎦ = ⎢⎣5⎥⎦ ⎡ 24 5⎤ ⎡ 10 ⎤ ⎡365⎤ ⎡ 1⎤ ⎢⎣ 19 14 ⎥⎦ ⎢⎣ 25⎥⎦ = ⎢⎣540 ⎥⎦ = ⎢⎣ 20⎥⎦ ⎡ 24 5⎤ ⎡17 ⎤ ⎡ 483⎤ ⎡15⎤ ⎢⎣ 19 14 ⎥⎦ ⎢⎣15⎥⎦ = ⎢⎣ 533⎥⎦ = ⎢⎣13⎥⎦ T H E Y S P L (mod 26) I T T H E A T O M [C P]. ⎡ 1 0 −54 −189 ⎤ ⎢⎣0 1 5 14 ⎥⎦ Add −14 times the second row to the first. 7⎤ ⎡ 1 14 16 ⎢⎣0 1 57 456 ⎥⎦ 6. deciphering matrix is A−1 = ⎢ ⎣ 19 14 ⎥⎦ From Table 1 the numerical equivalent of the given ciphertext is LN GI HG YB VR EN JY 12 14 7 9 8 7 25 2 22 18 5 14 10 25 QO 17 15 To obtain the plaintext pairs. ⎡ 1 0 24 19 ⎤ ⎢⎣0 1 5 14 ⎥⎦ Replace −54 and −189 by their residues modulo 26. and so the Thus ( A−1 )T = ⎢ ⎣ 5 14 ⎥⎦ ⎡ 24 5⎤ . we will group the letters in triples. ⎡ 1 140 154 168 147 42⎤ ⎢ 6 17 7 22 5 3⎥ ⎢ ⎥ 4 21 15 13 5⎦ ⎣7 Multiply the first row by 15−1 = 7 (mod 26).

16 ⎤ 11⎥ ⎥ 23⎦ Replace the entries in the second and third rows by their residues modulo 26. ( A−1 )T = ⎢ 9 ⎢ ⎣15 Add −6 times the first row to the second and −7 times the first row to the third. the message is I HAVE COME TO BURY CAESAR. 4 15 24⎤ 6 21 7 ⎥ ⎥ 15 6 17 ⎦ Replace the entries in the first and second row by their residues modulo 26. 256 . −48 −193 −54 ⎤ 6 21 7⎥ ⎥ −63 −228 −61⎦ ⎡ 4 9 15⎤ ⎡ 6 ⎤ ⎡ 282 ⎤ ⎡ 22 ⎤ ⎢ 15 17 6 ⎥ ⎢17 ⎥ = ⎢ 421⎥ = ⎢ 5⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 24 0 17 ⎦ ⎣ 7 ⎦ ⎣ 263⎦ ⎣ 3⎦ ⎡ 4 9 15⎤ ⎡ 7 ⎤ ⎡ 379⎤ ⎡15⎤ ⎢ 15 17 6 ⎥ ⎢ 4 ⎥ = ⎢ 299⎥ = ⎢13⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 24 0 17 ⎦ ⎣ 21⎦ ⎣ 525⎦ ⎣ 5⎦ I H A V E C O M E ⎡ 4 9 15⎤ ⎡7 ⎤ ⎡ 124 ⎤ ⎡ 20 ⎤ T ⎢ 15 17 6 ⎥ ⎢ 4⎥ = ⎢ 197 ⎥ = ⎢ 15⎥ O ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 24 0 17 ⎦ ⎣ 4⎦ ⎣ 236 ⎦ ⎣ 2 ⎦ B Add −10 times the second row to the first and −12 times the second row to the third. 24 ⎤ 0⎥ ⎥ 17 ⎦ Replace the entries in the second row by their residues modulo 26. ⎢ ⎥ ⎣ 24 0 17 ⎦ From Table 1 the numerical equivalent of the given ciphertext is H P A F Q G G D U G G D 8 16 1 6 17 7 7 4 21 7 4 4 H P G O D Y N O R 8 16 7 15 4 25 14 15 18 To obtain the plaintext triples. 12 17 16 ⎤ 6 21 7 ⎥ ⎥ 9 24 23⎦ Replace the entries in the second row by their residues modulo 26. we multiply each ciphertext vector by A−1 : ⎡ 4 9 15⎤ ⎡ 8⎤ ⎡ 191⎤ ⎡9 ⎤ ⎢ 15 17 6 ⎥ ⎢16 ⎥ = ⎢ 398⎥ = ⎢8⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 24 0 17 ⎦ ⎣ 1⎦ ⎣ 209 ⎦ ⎣ 1⎦ −1 ⎡ 1 10 24 ⎢0 1 5 ⎢ ⎣0 12 9 ⎡ 1 0 −26 ⎢0 1 5 ⎢ 0 0 − 51 ⎣ ⎡1 0 0 ⎢0 1 5 ⎢ ⎣0 0 1 ⎡1 0 0 ⎢0 1 0 ⎢ ⎣0 0 1 Add −5 times the third row to the second. 24 12 17 16⎤ ⎡ 1 10 ⎢0 −43 −137 −50 −97 −93⎥ ⎢ ⎥ ⎣0 −66 −147 −69 −106 −107 ⎦ ⎡ 1 10 24 12 17 ⎢0 9 19 2 7 ⎢ ⎣0 12 9 9 24 ⎡ 1 10 24 12 17 ⎢0 1 57 6 21 ⎢ ⎣0 12 9 9 24 ⎡ 1 0 0 4 15 ⎢0 1 0 9 17 ⎢ ⎣0 0 1 15 6 ⎡ 4 Thus. 15 24⎤ 17 0⎥ and so the ⎥ 6 17 ⎦ by 9 = 3 (modulo 26). 16 ⎤ 33⎥ ⎥ 23⎦ Multiply the second row ⎡ 4 9 15⎤ deciphering matrix is A−1 = ⎢ 15 17 6 ⎥ . 4 15 24 ⎤ −69 −9 −78⎥ ⎥ 15 6 17 ⎦ ⎡ 4 9 15⎤ ⎡ 8⎤ ⎡ 281⎤ ⎡ 21⎤ U ⎢ 15 17 6 ⎥ ⎢16 ⎥ = ⎢ 434 ⎥ = ⎢ 18⎥ R ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 24 0 17 ⎦ ⎣ 7 ⎦ ⎣ 311⎦ ⎣ 25⎦ Y ⎡ 4 9 15⎤ ⎡ 15⎤ ⎡ 471⎤ ⎡3⎤ C ⎢ 15 17 6 ⎥ ⎢ 4 ⎥ = ⎢ 443⎥ = ⎢ 1⎥ A ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 24 0 17 ⎦ ⎣ 25⎦ ⎣785⎦ ⎣5⎦ E ⎡ 4 9 15⎤ ⎡14 ⎤ ⎡ 461⎤ ⎡19 ⎤ S ⎢ 15 17 6 ⎥ ⎢15⎥ = ⎢ 573⎥ = ⎢ 1⎥ A ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 24 0 17 ⎦ ⎣18⎦ ⎣642 ⎦ ⎣18⎦ R Finally.Chapter 10: Elementary Linear Algebra SSM: Elementary Linear Algebra ⎡ 1 10 24 12 17 16⎤ ⎢ 6 17 7 22 5 3⎥ ⎢ ⎥ ⎣7 4 21 15 13 5⎦ Replace the entries in the first row by their residues modulo 26.

257 .2 a matrix A with entries in Z 29 is invertible if and only if det(A) ≠ 0 (mod 29). Using Table 1 and notations of Example 1. residue modulo 2. 1. (b) Reduce [ A I ] to ⎡⎣ I ⎡ 1 1 0 1 0 0⎤ ⎢0 1 1 0 1 0 ⎥ ⎢ ⎥ ⎣ 1 1 1 0 0 1⎦ ⎡ 1 1 0 1 0 0⎤ ⎢0 1 1 0 1 0⎥ ⎢ ⎥ ⎣ 2 2 1 1 0 1⎦ ⎡0 1 1⎤ ⎡1 ⎤ ⎡1 ⎤ ⎡1 ⎤ ⎢1 1 1⎥ ⎢1 ⎥ = ⎢ 2 ⎥ = ⎢0 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣1 0 1⎦ ⎣ 0 ⎦ ⎣1 ⎦ ⎣1 ⎦ ⎡0 1 1⎤ ⎡ 0⎤ ⎡1⎤ ⎢ 1 1 1⎥ ⎢ 0⎥ = ⎢1⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 1 0 1⎦ ⎣1 ⎦ ⎣1⎦ The decoded message is 110101111. the case n = 1 being already given. ⎡ 1 1 0 1 0 0⎤ ⎢0 1 1 0 1 0 ⎥ ⎢ ⎥ ⎣0 0 1 1 0 1⎦ = PD n −1 ( P −1P) DP −1 Replace 2 by its = PD n −1DP −1 = PD n P −1 .16 ⎡0 1 1⎤ Thus A−1 = ⎢ 1 1 1⎥ .16 I ].16 [A Exercise Set 10. proving the result.SSM: Elementary Linear Algebra Section 10. Use induction on n. by Corollary 10. ⎡ 1 1 0 1 0 0⎤ ⎢0 1 2 1 1 1⎥ ⎢ ⎥ ⎣0 0 1 1 0 1⎦ Add the third row 2. ⎡ 1 0 0 0 1 1⎤ ⎢0 1 0 1 1 1⎥ ⎢ ⎥ ⎣0 0 1 1 0 1⎦ Replace 2 by its residue modulo 2. Form the matrix Section 10. Since 29 is a prime number. 8.15. ⎡ 1 1 0 1 0 0⎤ ⎢0 1 0 1 1 1⎥ ⎢ ⎥ ⎣0 0 1 1 0 1⎦ Replace 2 by its residue modulo 2. −1 ⎤ A ⎦ modulo 2. 4 2 ⎡ 1 1 0⎤ ⎢2 4 ⎥ The transition matrix is thus M = ⎢ 12 12 12 ⎥ . ⎡ 1 1 0 ⎤ ⎡ 1⎤ ⎡ 2 ⎤ ⎡0 ⎤ ⎢0 1 1⎥ ⎢ 1⎥ = ⎢ 1⎥ = ⎢ 1⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 1 1 1⎦ ⎣ 0 ⎦ ⎣ 2 ⎦ ⎣0 ⎦ ⎡ 1 1 0 ⎤ ⎡ 1⎤ ⎡ 1⎤ ⎡ 1⎤ ⎢0 1 1⎥ ⎢ 0 ⎥ = ⎢ 1⎥ = ⎢ 1⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 1 1 1⎦ ⎣ 1⎦ ⎣ 2 ⎦ ⎣0 ⎦ ⎡ 1 1 0 ⎤ ⎡1⎤ ⎡ 2 ⎤ ⎡ 0 ⎤ ⎢0 1 1⎥ ⎢1⎥ = ⎢ 2 ⎥ = ⎢ 0 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 1 1 1⎦ ⎣1⎦ ⎣ 3⎦ ⎣ 1⎦ The encoded message is 010110001. which is the original message. ⎡ 1 2 0 2 1 1⎤ ⎢0 1 0 1 1 1⎥ ⎢ ⎥ ⎣0 0 1 1 0 1⎦ Add the second row to the first row. If the result is true for n − 1. we derive the following equations: 1 1 an = an −1 + bn −1 2 4 1 1 1 bn = an −1 + bn −1 + cn −1 2 2 2 1 1 cn = bn −1 + cn −1. ⎢ ⎥ ⎣ 1 0 1⎦ ⎡0 1 1⎤ ⎡ 0 ⎤ ⎡1 ⎤ ⎢ 1 1 1⎥ ⎢1 ⎥ = ⎢1 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 1 0 1⎦ ⎣ 0 ⎦ ⎣0 ⎦ 7. ⎢ 1 1⎥ ⎢⎣ 0 4 2 ⎥⎦ The characteristic polynomial of M is to the second row. then Add the first row to M n = M n −1M = ( PD n −1 P −1 )( PDP −1 ) the third row. (a) Multiply each of the triples of the message ⎡ 1 1 0⎤ by A = ⎢ 0 1 1⎥ and reduce the results ⎢ ⎥ ⎣ 1 1 1⎦ modulo 2.

4. and 1 1⎤ 1 0 0 ⎤ ⎡ 14 4 4⎥ ⎡ 1 1 1⎤ ⎢⎡ ⎢ ⎥ 1 n 0 − 12 ⎥ . we obtain () () () 1 1 1 ⎛1⎞ a0 + b0 + c0 + ⎜ ⎟ 4 4 4 ⎝2⎠ n +1 ⎛1⎞ a0 − ⎜ ⎟ ⎝2⎠ b2 n n +1 c0 c2 n 1 ⎛1⎞ + ⎜ ⎟ (a0 − c0 ) 4 ⎝2⎠ 1 1 1 1 bn = a0 + b0 + c0 = 2 2 2 2 = cn = = and 2 1 + (2a0 − b0 − 4c0 ) 3 6 ⋅ 4n 1 1 b2 n +1 = − (2a0 − b0 − 4c0 ) 3 6 ⋅ 4n c2 n +1 = 0. Then 258 . and M 2 the matrix of Exercise 2. e2 = [ −1 0 1]T . so the eigenvalues are λ1 = 1. Thus T T Mn = PD n P −1 are e1 = [5 6 1]T . 0 ⎢ 2 2 2 2 2 1 ⎥ ⎢0 0 0 ⎥ ⎢ 0 1 1 ⎥ ⎦ ⎢⎣ 2 ⎦⎥ ⎣ 8 4 ⎦⎥ The characteristic polynomial of this matrix is 5 1 λ3 − λ 2 + λ. = ⎢ 2 0 −2 ⎥ ⎢0 1 ⎢2 0 ⎥ 2 ⎢ ⎥ ⎥⎢ 1⎥ ⎣ 1 −1 1⎦ ⎢⎣0 0 0 ⎦ ⎢⎣ 14 − 14 4 ⎥⎦ This yields ⎡ an ⎤ x( n) = ⎢ bn ⎥ ⎢ ⎥ ⎢⎣ cn ⎥⎦ ⎡ 1 1 n +1 1 1 1 n +1 ⎤ − 2 ⎢4 + 2 ⎥ ⎡ a0 ⎤ 4 4 ⎢ ⎥⎢ ⎥ 1 1 1 =⎢ 2 2 2 ⎥ ⎢ b0 ⎥ . easily found to be e1 = [1 0]T and e2 = [1 −1]T . the verification of Equation (7) is in the text. We have 0 ⎤ ⎡1 1 0 ⎤ ⎡ 12 83 14 ⎤ 2 ⎥ ⎥ ⎥ ⎢ 1⎥ ⎢ 1 1⎥ = ⎢ 1 1 1 ⎥ . ⎢ 1 1 n +1 1 1 1 n +1 ⎥ ⎣⎢ c0 ⎦⎥ + 2 ⎢⎣ 4 − 2 ⎥⎦ 4 4 Remembering that a0 + b0 + c0 = 1. The characteristic polynomial of M is 1⎞ ⎛ (λ − 1) ⎜ λ − ⎟ . Thus. T e2 = [1 0 −1] . ⎝ 2⎠ x( 2 n) = ( M 2 M1 )n x( 0) and x( 2 n +1) = M1 ( M 2 M1 )n x( 0) . = ⎢ 6 0 −2 ⎥ ⎢ 0 1 0⎥ ⎢− 3 6 3 4 ⎢ ⎥⎢ ⎢ ⎥ ⎥ ⎣ 1 1 1⎦ ⎣0 0 0 ⎦ ⎢⎣ 14 − 14 14 ⎥⎦ Using the notation of Example 1 (recall that a0 + b0 + c0 = 1). Corresponding eigenvectors are 2 approaches zero as n → ∞. bn → . we 1 1 1 obtain an → . Corresponding eigenvectors (found by solving (λI − M)x = 0) are e1 = [1 2 1] . Call M1 the matrix of Example 1.Chapter 10: Elementary Linear Algebra SSM: Elementary Linear Algebra ⎛3⎞ ⎛1⎞ det(λI − M ) = λ3 − ⎜ ⎟ λ 2 + ⎜ ⎟ λ ⎝2⎠ ⎝2⎠ 1⎞ ⎛ = λ(λ − 1) ⎜ λ − ⎟ . and e3 = [1 −2 1] . λ 2 = . and 2 λ3 = 0. a2n +1 = n +1 n +1 1 1 1 ⎛1⎞ ⎛1⎞ a0 + b0 + c0 − ⎜ ⎟ a0 + ⎜ ⎟ c0 4 4 2 ⎝2⎠ ⎝2⎠ 1 ⎛1⎞ −⎜ ⎟ 4 ⎝2⎠ ⎛1⎞ Since ⎜ ⎟ ⎝2⎠ n +1 n +1 5 1 + (2a0 − b0 − 4c0 ) 12 6 ⋅ 4n 1 = 2 1 1 = − (2a0 − b0 − 4c0 ) 12 6 ⋅ 4n a2 n = () n +1 1 4 1 2 1 4 (a0 − c0 ). 4 4 1 λ 2 = . From this point. () () an = ( M 2 M1 ) n = PD n P −1 1 1 1⎤ 1 0 0 ⎤ ⎡ 12 12 12 ⎥ ⎡ 5 −1 1⎤ ⎢⎡ ⎢ ⎥ n 1 1 2⎥. and cn → as 4 2 4 n → ∞. Corresponding eigenvectors 4 ⎡1 ⎢2 M 2 M1 = ⎢ 12 ⎢ ⎣⎢ 0 1 so the eigenvalues of M are λ = 1. we obtain e3 = [1 −2 1]T . 3. so the eigenvalues are λ1 = 1 2⎠ ⎝ 1 and λ 2 = . λ3 = 0.

10 in 12 generations. and. AA) is 2 1 2 1 a0 + b0 + c0 + d0 + e0 . 3 3 3 3 The proportion of A genes in the population at the outset is as follows: all the type 259 . 384 2 bn = 1 ⎡ ⎤ 2 ⎢ ⎥ 1 ⎢ ⎥ 2 ⎢ ⎥ 0 ⎥ 6.25 = . According to Equation (8). under the controlled program 8+ n 20 1 1 the percentage would be in 12 generations. 11 2 2 .16 1 4 9 8 2 = . then b2 = 9 2 9 10 9 1 = . From (13) we have that the probability that the limiting sibling-pairs will be type (A. 14 16 . b3 = 5 1 5 11 10 = 2 . P −1x(0) = ⎢ 0 ⎢ ⎥ ⎢1 ( ⎥ ) ⎢ 12 1 + 5 ⎥ ⎢ 1 (1 − 5 ) ⎥ ⎣ 12 ⎦ 1 ⎡ ⎤ 2 ⎢ ⎥ 1 ⎢ ⎥ 2 ⎢ ⎥ 0 ⎥ n −1 (0) ⎢ D P x =⎢ 0 ⎥ ⎢1 1 n +1 ⎥ ⎢ 3 ⎡⎣ 4 (1 + 5 ) ⎤⎦ ⎥ ⎢ n +1 ⎥ ⎢ 1 ⎡ 1 (1 − 5 ) ⎤ ⎥ ⎦ ⎦ ⎣3 ⎣4 n +1 n +1 ⎡1 + 1 ⋅ 1 ⎡ + ( −3 + 5 )(1 − 5 ) ⎤⎦ ⎤ ⎢ 2 3 4n + 2 ⎣( −3 − 5 )(1 + 5 ) ⎥ ⎢ ⎥ n +1 n +1 ⎤ 1⋅ 1 ⎡ + (1 − 5 ) ⎦ 3 4n +1 ⎣(1 + 5 ) ⎢ ⎥ ⎢ ⎥ n n⎤ 1⋅ 1 ⎡ ⎥ 3 4n +1 ⎣(1 + 5 ) + (1 − 5 ) ⎦ n −1 (0) ⎢ PD P x = ⎢ ⎥ n n 1⋅ 1 ⎡ ⎤ ⎢ ⎥ 3 4n +1 ⎣(1 + 5 ) + (1 − 5 ) ⎦ ⎢ ⎥ n +1 n +1 1⋅ 1 ⎡ ⎢ ⎥ + (1 − 5 ) ⎤⎦ 3 4n +1 ⎣(1 + 5 ) ⎢ ⎥ ⎢ 1 + 1 ⋅ n1+ 2 ⎡( −3 − 5 )(1 + 5 )n +1 + ( −3 + 5 )(1 − 5 )n +1 ⎤ ⎥ ⎦⎦ ⎣2 3 4 ⎣ ⎡1⎤ ⎢2⎥ ⎢0⎥ ⎢0⎥ 1 1 [As n tends to infinity. n+ 2 n +1 0 4 4 ⎢ ⎥ ⎢0⎥ ⎢1⎥ ⎣2⎦ 7. if b0 = . or = . We will reach = . we get b1 = 4 Section 10. so x( n) approaches ⎢ ⎥ .006%. in general.00006 = . and approach 0.SSM: Elementary Linear Algebra 1 5. From Equation (9).

(a) The characteristic polynomial of L is 3 λ 2 − λ − . . x = Lx ( 4 ) = ⎢ ⎣125⎥⎦ ⎣ 191⎥⎦ (b) From Equation (10). aa) genes. aa).49611. which is the net reproduction rate. From an (A. The eigenvector 3 ⎡1⎤ ⎢1⎥ 3 corresponding to λ = is ⎢ 3 ⎥ .17 1.9% of the youngest age class. AA) pairs and similarly for (a. Aa) genes. 3 3 3 3 (A. From either (A. ⎡100 ⎤ ⎡175⎤ (b) x(1) = Lx(0) = ⎢ ⎥ . 50 ⎣ ⎦ ⎣ 50 ⎦ ⎡ 250⎤ x(3) = Lx( 2) = ⎢ . say the third. who will not mature. aa) or (a. Equation (9) tells 11 11 .. AA) pairs we must get an Aa female. and follows the analysis in the text. x( 2) = Lx(1) = ⎢ ⎥ . (a) The characteristic polynomial of L is 3 ⎛ 1⎤ 3⎞⎡ ⎛ 3⎞ λ3 − 2λ − = ⎜ λ − ⎟ ⎢λ 2 + ⎜ ⎟ λ + ⎥ . etc. so we harvest 19 19 or 57. AA) genes.18 Exercise Set 10. so the eigenvalues of L are 4 3 1 3 λ = and λ = − . Since Lx1 = ⎡ 19 ⎣8 260 1 2 1⎤ 8⎦ T .Chapter 10: Elementary Linear Algebra SSM: Elementary Linear Algebra ⎡857 ⎤ (c) x(6) = Lx(5) = ⎢ ⎣ 285⎥⎦ ⎡ 855⎤ x(6)  λ1x(5) = ⎢ ⎣ 287 ⎥⎦ 2 1 of the type (A. ⎛ 1 ⎞ ⎛ 1 ⎞⎛ 1 ⎞ 19 7. the sum of these.e. ⎢0 0 0 0⎥ ⎣⎢0 0 0 1 ⎦⎥ in the first age period.. aa) must be of type a. a1 is the average number of offspring produced 8. The transition matrix is then ⎡1 0 0 0 ⎤ ⎢0 0 0 0⎥ ⎢ ⎥. thus λ1 = and 2 2 2 ⎡1⎤ ⎢ 1 ⎥ ⎡1 ⎤ x1 = ⎢ 2 ⎥ = ⎢ 1 ⎥ is the corresponding ⎢ 3 ⎥ ⎣3⎦ ⎣2⎦ eigenvector. Section 10.18 Section 10. Thus h. R = 0 + (. the fraction harvested of 2 2 1 each age group. is the expected number of offspring produced by a given female during her expected lifetime. and similarly for the last column. R = 0 + 4 ⎜ ⎟ + 3 ⎜ ⎟⎜ ⎟ = = 2. is 1 − = so the yield is 3 3 1 33 % of the population.. and so on for all the periods. For the middle two columns.00240)(.00024)(. because of the way the genes are inherited. Thus no offspring will come from such pairs. 8. this is the 2 ⎢1⎥ ⎣ 18 ⎦ age distribution vector after each harvest.987) = 1. the youngest class . note that male offspring from (A. and females are of type Aa.375 ⎝ 2 ⎠ ⎝ 2 ⎠⎝ 4 ⎠ 8 9. 3 3 the type (A. a2b1 is the number of offspring produced in the second period times the probability that the female will live into the second period. The fifth column is like the second column. For the first column of M we realize that parents of type (A. the age distribution vector x1 is ⎡1 ⎣ us that h1 = 1 − 1 2 1 19 8 = 1⎤ 8⎦ T . it is the expected number of offspring per female during the second period. Thus. ⎣ 88⎥⎦ ⎡382 ⎤ 5 ⎡570 ⎤ x( 4) = Lx(3) = ⎢ . so 8 ⎝ 4⎦ ⎝ 2⎠ 2⎠⎣ Exercise Set 10. AA) can produce offspring only of that type.17 1. 5. i. 3 λ1 = . AA) pair we get only (A.99651)" (.99651) + " + (.yielding 2 1 2 1 a0 + b0 + c0 + d0 + e0 .

So the least-squares trigonometric polynomial of order 3 is 4 + 4 cos t + cos 2t + cos 3t .755 .824 . and h1 = h2 = " = hI −1 = hI +1 = " = hn = 0.1. a + a b + " + aI −1b1b2 " bI − 2 − 1 So hI = 1 2 1 +1 aI b1b2 " bI −1 + " + an b1b2 " bn −1 a + a b + " + aI −1b1b2 " bI − 2 − 1 + aI b1b2 " bI −1 + " + an b1b2 " bn −1 = 1 21 aI b1b2 " bI −1 + " + an b1b2 " bn −1 R −1 = aI b1b2 " bI −1 + " + an b1b2 " bn −1 5.699 .19 1.626 . The Leslie matrix of Example 1 has b1 = . t dt = T = = and a t cos dt k 3 T ∫0 T ∫0 T k 2π 2 2 T 2 2 kπ t T2 = − t sin dt .418 0 0 0]T .19. Then Equation (4) becomes a1 + a2 b1 + " + aI −1b1b2 " bI − 2 + (1 − hI )[ aI b1b2 " bI −1 + " + aJ −1b1b2 " bJ − 2 ] = 1. or 19. etc. and all the other hk’s are zero. 2. From Theorem 10.8% of the population. ak = π 3 2π ∫0 (t − π )2 cos kt dt = 4 k2 . and (t − π )2 sin kt dt = 0.975. This.845 . a1 + a2b1 + " + aI −1b1b2 " bI − 2 − 1 +1 aI b1b2 " bI −1 + " + aJ −1b1b2 " bJ −2 a + a b + " + aJ −1b1b2 " bJ − 2 − 1 .584.755 .9% of the population. = 1 21 aI b1b2 " bI −1 + " + aJ −1b1b2 " bJ −2 We solve for hI to obtain hI = Section 10.626 .090 . Equation (4) then takes the form a1 + a2 b1 + a3b1b2 + " + aI b1b2 " bI −1 (1 − hI ) + aI +1b1b2 " bI (1 − hI ) + " + an b1b2 " bn −1 (1 − hI ) = 1.795 . Here hJ = 1.824 .795 .19 Exercise Set 10. 3 π2 ⎝ T T T 16 T ⎠ π ⎝ T T T T ⎠ 4 9 2 3 4 261 . From Theorem 10.5323 0 0 0 0]T Lx1 = [ 2.9% of 79.19. 9 2.845 .2% of the population.532 . we compute a0 = bk = π2 3 1 2π π ∫0 1 2π π ∫0 2 1 (t − π )2 dt = π 2 .SSM: Elementary Linear Algebra Section 10. b3 = . (1 − hI )[ aI b1b2 " bI −1 + aI +1b1b2 " bI + " + an b1b2 " bn −1 ] = 1 − a1 − a2 b1 − " − aI −1b1b2 " bI − 2 . or 45.2%. hI ≠ 0.845. together with the harvesting data from Equations (13) and the formula of Equation (5) yields x1 = [1 .19 contains 79.2.699 . we compute a0 = 2 T 2 2 2 2 T 2 2 kπ t T2 . b2 = . 4. In this situation we have hI ≠ 0. The proportion of sheep harvested is h1 ( Lx1 )2 + h9 ( Lx1 )9 = 1. Thus the yield is 57.965.51. T ∫0 T kπ So the least-squares trigonometric polynomial of order 4 is bk = 2 T2 T2 ⎛ 2π t 1 4π t 1 6π t 1 8π t ⎞ T ⎛ 2π t 1 4π t 1 6π t 1 8π t ⎞ + − cos cos cos cos + + + + sin + sin + sin ⎜ ⎟ ⎜ sin ⎟. The total of the entries of Lx1 is 7.

3⋅5 5⋅7 (2n − 1)(2n + 1) ⎟⎠ π ⎜⎝ 1 ⋅ 3 . =⎨ if k is odd ⎪ ( 2k ) 2 π 2 ⎩ So the least-squares trigonometric polynomial of order n is: 2π t 1 6π t 1 10π t 1 2π nt ⎞ T 8T ⎛ 1 + cos + +" + cos cos cos − ⎜ 2 2 2 2 T T 10 T T ⎟⎠ 4 π2 ⎝2 6 ( 2 n) 5. a0 = if n is even. 262 cos 3t cos nt ⎞ + + +" + .19. π 0 ⎪⎩0 if k > 1 So the least-squares trigonometric polynomial of order 4 is 4. the last term involves n − 1 if n is odd. + sin t − 2 3π 15π 1 4 sin t dt = . From Theorem 10. 1 π π ∫0 sin t cos kt dt π 1⎛ 1 ⎞ = ⎜ 2 [ k sin kt sin t + cos kt cos t ⎟ π ⎝ k −1 ⎠0 1⎛ 1 ⎞ = ⎜ 2 [0 + (−1)k −1 − 1] ⎟ π ⎝ k −1 ⎠ 0 if is odd k ⎧ ⎪ 2 =⎨ if k is even.19. 2 π 1 sin t cos kt dt 2 4 π (4k 2 − 1) 4 .Chapter 10: Elementary Linear Algebra 3. 0 2 π =− So the least-square trigonometric polynomial of order n is 2 π − 4 ⎛ cos t cos 2t 2 T 2 1T 2 T T f (t )dt = ∫ 2 t dt + ∫ 1 (T − t )dt = .2. π (2k − 1)(2k + 1) 1 2π 1 bk = ∫ sin t sin kt dt = 0. − ⎪ π (k 2 − 1) ⎩ ⎧1 1 π ⎪ if k = 1 bk = ∫ sin kt sin t dt = ⎨ 2 . ∫ 2 T 0 T 0 T 2T 1T 2 2kπ t 2 T 2 kπ t ak = ∫ 2 t cos dt + ∫ 1 (T − t ) cos dt T 0 T T 2T T 4T ((−1)k − 1) = 4k 2π 2 if k is even ⎧0 ⎪ 8T . (Note the upper limit on the second integral). From Theorem 10. a0 = ak = 1 2π π ∫0 SSM: Elementary Linear Algebra f (t )dt = 1 π π 2 ∫0 sin t dt = π . From Theorem 10. a0 = ak = 1 2π π ∫0 =− 1 2π π ∫0 1 π 1 2 2 cos 2t − cos 4t .2.2.19.

v 2 . or 7 = 7. (b) As in part (a) the system for c1 . c2 = . and = π Section 10. and c3 ⎡3 −2 3⎤ ⎡ c1 ⎤ ⎡1 ⎤ is ⎢3 −2 0 ⎥ ⎢c2 ⎥ = ⎢0 ⎥ which has the ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 1 1 1⎦ ⎣ c3 ⎦ ⎣1 ⎦ 4 6 unique solution c1 = . it follows that v is a convex combination of the vectors v1 .20 Exercise Set 10. it follows that v is a convex combination of the vectors v1 . m = 2n − 2 − k. and 15 15 5 c3 = . These equations can be written in combined matrix ⎡1 3 4 ⎤ ⎡ c1 ⎤ ⎡3⎤ form as ⎢1 5 2 ⎥ ⎢c2 ⎥ = ⎢3⎥ . and v2 v3 v4 v5 v6 v3 . ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣1 1 1 ⎦ ⎣ c3 ⎦ ⎣1⎦ v3 . and 2π ⎛ 1 + cos 2t ⎞ ∫0 ⎜ ⎝ 2 ⎟ dt ⎠ = π. 5 2 2 c2 = . Because one of these coefficients 5 is negative. and v3 . c2 = .20 (c) As in part (a) the system for c1 . (a) 2 4 unique solution c1 = .20 ⎡1⎤ ⎡3⎤ ⎡ 4 ⎤ ⎡3⎤ 1. the number of vertex points. c2 .SSM: Elementary Linear Algebra 6. 2. c2 . is equal to 7. (a) (b) 1 = 2π ∫0 cos kt = = Section 10. Because these coefficients are all 15 nonnegative. is equal to 5. and 5 5 1 c3 = − . v 2 . Combining everything that is given in the statement of the problem. c2 . and c3 ⎡1 3 4 ⎤ ⎡ c1 ⎤ ⎡ 2 ⎤ is ⎢1 5 2 ⎥ ⎢c2 ⎥ = ⎢ 4 ⎥ which has the ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣1 1 1 ⎦ ⎣ c3 ⎦ ⎣1 ⎦ v1 4. becomes 7 = 2(7) − 2 − 5. k. Equation (7). Because these 5 5 coefficients are all nonnegative. (c) 2π sin kt = ∫0 = ∫0 sin 2 kt dt v3 . v 2 . c2 = . and c3 = . and c3 ⎡3 −2 3⎤ ⎡ c1 ⎤ ⎡0 ⎤ is ⎢3 −2 0 ⎥ ⎢c2 ⎥ = ⎢0 ⎥ which has the ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 1 1 1⎦ ⎣ c3 ⎦ ⎣1 ⎦ dt = 2π 2π ∫0 cos2 kt dt 2 3 unique solution c1 = . we obtain: w = Mv + b = M (c1 v1 + c2 v 2 + c3 v3 ) + (c1 + c2 + c3 )b = c1 ( Mv1 + b) + c2 ( Mv 2 + b) + c3 ( Mv3 + b) = c1w1 + c2 w 2 + c3 w 3 . v 2 . For both triangulations the number of triangles. 1 This system has the unique solution c1 = . it follows that v is not a convex combination of the vectors v1 . is equal to 7. 2π ⎛ 1 − cos 2t ⎞ ⎜ ⎝ 2 ⎟ dt ⎠ (d) As in part (a) the system for c1 . 3. (a) Equation (2) is c1 ⎢ ⎥ + c2 ⎢ ⎥ + c3 ⎢ ⎥ = ⎢ ⎥ ⎣1⎦ ⎣5⎦ ⎣ 2 ⎦ ⎣3⎦ and Equation (3) is c1 + c2 + c3 = 1. and 5 5 c3 = 0. n. 263 v7 . m. it follows that v is a convex combination of the vectors v1 . and the number of boundary vertex points. Because these coefficients are all nonnegative.

⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎣ 2 1 1⎦ ⎣ b1 ⎦ ⎣ 5 ⎦ fourth. must equal zero. we are led to the following two linear systems: ⎡ −2 2 1⎤ ⎡ m11 ⎤ ⎡ −8⎤ ⎢ 0 0 1⎥ ⎢ m12 ⎥ = ⎢ 0⎥ and ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎣ 2 1 1⎦ ⎣ b1 ⎦ ⎣ 5⎦ v 2 then from Exercise 6(a) we must have that v = c1 v1 + c2 v 2 + (0) v3 where c1 + c2 = 1. 2.Chapter 10: Elementary Linear Algebra (b) v1 SSM: Elementary Linear Algebra v2 Solving these two linear systems leads to ⎡3 −1⎤ ⎡0 ⎤ and b = ⎢ ⎥ . Thus we obtain M = ⎢ ⎣2⎦ ⎣0 1 ⎥⎦ (b) If. b1 = 1 and the second system has the solution m21 = 0. and fifth equations can be written in matrix form as ⎡1 1 1⎤ ⎡ m11 ⎤ ⎡ 4 ⎤ ⎢ 2 3 1⎥ ⎢ m12 ⎥ = ⎢9 ⎥ and the second. and v3 = (0) v1 + (0) v 2 + (1) v3 . ⎡1 2 ⎤ ⎡1 ⎤ and b = ⎢ ⎥ . Solving these two linear systems leads to ⎡ 1 1⎤ ⎡ 1⎤ M = ⎢2 ⎥ and b = ⎢ 2 ⎥ . v 2 . M =⎢ ⎣1 ⎦ ⎣ 1 1⎥⎦ v3 v4 (c) As in part (a). ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎣ 2 1 1⎦ ⎣ b2 ⎦ ⎣ 4 ⎦ 264 . Thus at least one of the coefficients. precisely two of the coefficients are zero and one coefficient is one. In each of these cases. ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎣ −4 −2 1⎦ ⎣ b2 ⎦ ⎣ −9 ⎦ 2m11 + 3m12 + b1 = 9 2m21 + 3m22 + b2 = 5 2m11 + m12 + b1 = 5 2m21 + m22 + b2 = 3 The first. can be written as the six scalar equations m11 + m12 + b1 = 4 m21 + m22 + b2 = 3 (d) As in part (a). and v3 of a triangle can be written as the convex combinations v1 = (1) v1 + (0) v 2 + (0) v3 . m12 = 2. Then m m ⎣ 21 ⎣ 2⎦ 22 ⎦ the three matrix equations Mvi + b = w i . v lies on the side of the triangle determined by the vectors v1 and (b) As in part (a). and sixth equations as ⎡1 1 1⎤ ⎡ m21 ⎤ ⎡3⎤ ⎢ 2 3 1⎥ ⎢ m22 ⎥ = ⎢5⎥ . i = 1. v 2 = (0) v1 + (1) v 2 + (0) v3 . (a) The vertices v1 . 3. b2 = 2. third. m22 = 1. we are led to the following two linear systems: ⎡ 5⎤ ⎡ 0 2 1⎤ ⎡ m11 ⎤ ⎢ 2 ⎥ ⎢ 2 2 1⎥ ⎢ m12 ⎥ = ⎢ 7 ⎥ and ⎥ ⎢ 2⎥ ⎢ ⎥⎢ ⎣ −4 −2 1⎦ ⎣ b1 ⎦ ⎢ − 7 ⎥ ⎣ 2⎦ ⎡ 0 2 1⎤ ⎡ m21 ⎤ ⎡ −1⎤ ⎢ 2 2 1⎥ ⎢ m22 ⎥ = ⎢ 3⎥ . for example. ⎣ 2 0⎦ ⎣ −1⎦ 7. we are led to the following two linear systems: ⎡ −2 1 1⎤ ⎡ m11 ⎤ ⎡ 0⎤ ⎢ 3 5 1⎥ ⎢ m12 ⎥ = ⎢ 5⎥ and ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎣ 1 0 1⎦ ⎣ b1 ⎦ ⎣ 3⎦ v5 v6 v7 ⎡ −2 1 1⎤ ⎡ m21 ⎤ ⎡ −2 ⎤ ⎢ 3 5 1⎥ ⎢ m22 ⎥ = ⎢ 2 ⎥ . ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎣ 1 0 1⎦ ⎣ b2 ⎦ ⎣ −3⎦ Solving these two linear systems leads to ⎡1 0 ⎤ ⎡ 2⎤ M =⎢ and b = ⎢ ⎥ . ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎣ 2 1 1⎦ ⎣ b2 ⎦ ⎣3⎦ The first system has the solution m11 = 1. in this example c3 . ⎡ −2 2 1⎤ ⎡ m21 ⎤ ⎡1 ⎤ ⎢ 0 0 1⎥ ⎢ m22 ⎥ = ⎢1 ⎥ . (a) Let M = ⎢ 11 ⎥ and b = ⎢b ⎥ . ⎥ 0 1 ⎣ ⎦ ⎣ −3⎦ m12 ⎤ ⎡m ⎡ b1 ⎤ 5.

SSM: Elementary Linear Algebra Section 10. Consequently. (a) Consider the vertex v1 of the triangle and its opposite side determined by the vectors v 2 and v3 . if at least one of the coefficients in the convex combination is zero. 3 3 3 (b) 1 ⎡ 2 ⎤ 1 ⎡ 5 ⎤ 1 ⎡1⎤ 1 ⎡8 ⎤ ⎡ 83 ⎤ + + = =⎢ ⎥ 3 ⎢⎣ 3 ⎥⎦ 3 ⎢⎣ 2 ⎥⎦ 3 ⎢⎣1⎥⎦ 3 ⎢⎣6 ⎥⎦ ⎣⎢ 2 ⎦⎥ 265 .20 (c) From part (b). none of the coefficients can be zero if the vector lies in the interior of the triangle. then the vector must lie on one of the sides of the triangle. 8. The midpoint v m of this ( v 2 + v3 ) and the point on 2 the line segment from v1 to v m that is twothirds of the distance to v m is given by opposite side is 1 2 1 2 ⎛ v + v3 ⎞ v1 + v m = v1 + ⎜ 2 ⎟ 3 3 3 3⎝ 2 ⎠ 1 1 1 = v1 + v 2 + v3 .

1 Exercise Set 2. C33 = (−1)3+3 M 33 = 19 3. 1 3 = = −1 − 18 = −19 6 −1 C32 = (−1) M 33 = 3+ 2 ⎡ 3 5⎤ ⎢⎣ −2 4⎥⎦ M 32 = 19 1 −2 = 7 − (−12) = 19 6 7 7. M11 = = 28 − (−1) = 29 1 4 C11 = (−1)1+1 M11 = 29 M12 = 6 −1 = 24 − 3 = 21 −3 4 C23 = (−1)2 +3 M 23 = 96 C12 = (−1)1+ 2 M12 = −21 M13 = 4 1 6 (c) M 22 = 4 0 14 4 3 2 1 6 4 6 4 1 = −4 +0 − 14 3 2 4 2 4 3 = −4(2 − 18) − 14(12 − 4) = −48 6 7 = 6 − (−21) = 27 −3 1 C13 = (−1)1+3 M13 = 27 M 21 = −2 3 = −8 − 3 = −11 1 4 C22 = (−1) 2+ 2 M 22 = −48 C21 = (−1)2 +1 M 21 = 11 M 22 = 1 3 = 4 − (−9) = 13 −3 4 −1 1 6 (d) M 21 = 1 0 14 1 3 2 1 6 −1 6 −1 1 = −1 +0 − 14 3 2 1 2 1 3 = −(2 − 18) − 14(−3 − 1) = 72 C22 = (−1) 2+ 2 M 22 = 13 M 23 = 1 −2 = 1 − 6 = −5 −3 1 C23 = (−1)2 +3 M 23 = 5 M 31 = −2 3 = 2 − 21 = −19 7 −1 C21 = (−1)2 +1 M 21 = −72 C31 = (−1)3+1 M 31 = −19 M 32 5.Chapter 2 Determinants 4 −1 6 (b) M 23 = 4 1 14 4 1 2 1 14 4 14 4 1 =4 − (−1) +6 1 2 4 2 4 1 = 4(2 − 14) + (8 − 56) + 6(4 − 4) = −96 Section 2.1 7 −1 1. (a) M13 3 5 = 12 − (−10) = 22 −2 4 −1 C13 = (−1)1+3 M13 = 0 44 5 2 1 ⎡ 4 −5⎤ ⎡ 11 − 22 ⎤ ⎢ ⎥ = 3 ⎥ 3⎥⎦ ⎢ 1 22 ⎢⎣ 2 22 ⎦ ⎣ 11 −5 7 = 10 − (−49) = 59 −7 −2 ⎡ −5 7 ⎤ ⎢⎣ −7 −2 ⎥⎦ 0 0 3 4 1 = 4 1 14 = 3 = 3(4 − 4) = 0 4 1 4 1 2 = −1 = 2 1 ⎡ −2 −7 ⎤ ⎡ − 59 ⎢ = 59 ⎢⎣ 7 −5⎥⎦ ⎢ 7 ⎣ 59 7 ⎤ − 59 ⎥ 5⎥ − 59 ⎦ .

det( A) = (λ − 2)(λ + 4) − (−5) = λ 2 + 2λ − 8 + 5 = λ 2 + 2λ − 3 = (λ − 1)(λ + 3) det(A) = 0 for λ = 1 or −3.1 a −3 5 = (a − 3)(a − 2) − (5)(−3) −3 a − 2 = a 2 − 5a + 6 + 15 = a 2 − 5a + 21 11. −2 1 4 −2 1 4 −2 1 3 5 −7 = 3 5 −7 3 5 1 6 2 1 6 2 1 6 = [(−2)(5)(2) + (1)(−7)(1) + (4)(3)(6)] − [(4)(5)(1) + (−2)(−7)(6) + (1)(3)(2)] = [ −20 − 7 + 72] − [ 20 + 84 + 6] = −65 3 0 0 3 0 03 0 2 −1 5 = 2 −1 5 2 −1 1 9 −4 1 9 −4 1 9 = [(3)(−1)(−4) + (0)(5)(1) + (0)(2)(9)] − [0(−1)(1) + (3)(5)(9) + (0)(2)(−4)] = 12 − 135 = −123 15. det(A) = (λ − 1)(λ + 1) − 0 = (λ − 1)(λ + 1) det(A) = 0 for λ = 1 or −1. Section 2. 13. 17.SSM: Elementary Linear Algebra 9. 3 0 0 −1 5 2 5 2 −1 19. (a) 2 −1 5 = 3 −0 +0 9 −4 1 −4 1 9 1 9 −4 = 3(4 − 45) = −123 3 0 0 −1 5 0 0 0 0 (b) 2 −1 5 = 3 −2 +1 9 −4 9 −4 −1 5 1 9 −4 = 3(4 − 45) − 2(0 − 0) + (0 − 0) = −123 (c) 3 0 0 0 0 3 0 3 0 + (−1) −5 2 −1 5 = −2 9 −4 1 −4 1 9 1 9 −4 = −2(0 − 0) − (−12 − 0) − 5(27 − 0) = 12 − 135 = −123 45 .

Expand along the first column. Expand along the third column. Expand along the second column. then along the first rows of the 3 × 3 matrices. −3 7 det( A) = 5 = 5[ −15 − (−7)] = −40 −1 5 23.Chapter 2: Determinants SSM: Elementary Linear Algebra 3 0 0 2 5 3 0 3 0 (d) 2 −1 5 = −0 + (−1) −9 1 −4 1 −4 2 5 1 9 −4 = −(−12 − 0) − 9(15 − 0) = 12 − 135 = −123 (e) (f) 3 0 0 0 0 3 0 3 0 −9 + (−4) 2 −1 5 = 1 −1 5 2 5 2 −1 1 9 −4 = (0 − 0) − 9(15 − 0) − 4(−3 − 0) = −135 + 12 = −123 3 0 0 2 −1 3 0 3 0 2 −1 5 = 0 −5 + (−4) 1 9 1 9 2 −1 1 9 −4 = −5(27 − 0) − 4(−3 − 0) = −135 + 12 = −123 21. 1 0 0 0 −1 0 = (1)(−1)(1) = −1 0 0 1 46 . 3 3 5 3 3 5 det( A) = −3 2 2 −2 − 3 2 2 −2 2 10 2 4 1 0 ⎛ 2 −2 2 −2 2 2 ⎞ ⎛ 2 −2 2 −2 2 2⎞ = 3⎜ 3 −3 +5 ⎟ − 3⎜ 3 1 0 − 3 4 0 + 5 4 1 ⎟ 10 2 2 2 2 10 ⎝ ⎠ ⎝ ⎠ = −3[3(24) − 3(8) + 5(16)] − 3[3(2) − 3(8) + 5(−6)] = −3(128) − 3(−48) = −240 27. det( A) = 1 k k2 k k2 −1 k k2 k k2 +1 k k2 k k2 = (k 3 − k 3 ) − (k 3 − k 3 ) + (k 3 − k 3 ) =0 25.

1 0 0 = (0)(2)(3)(8) = 0 0 8 33. (e) True. 1 True/False 2. Similar f arguments show that C23 = C32 and C13 = C31 . let A = ⎢ b d ⎢ ⎣c e c⎤ b e ⎥ . then (−1)i + j = 1 so i + j must be even. 47 . (f) True (g) False. 0 0 2 7 −3 1 −4 1 = (1)(1)(2)(3) = 6 0 2 7 0 0 3 Section 2. sin(θ ) cos(θ ) 0 sin(θ ) cos(θ ) sin(θ ) 0 = − cos(θ ) − cos(θ ) sin(θ ) sin(θ ) − cos(θ ) sin(θ ) + cos(θ ) 1 = sin 2 (θ ) − (− cos2 (θ )) = sin 2 (θ ) + cos2 (θ ) =1 1 35. then det(A) = 1. M11 = 0 f = 1 in both matrices. but det(cA) = ⎥ 0 c 0 1 ⎣ ⎦ det(cA) = c n A. (b) False. if Cij = (−1)i + j M ij = M ij . let A = ⎢ . ⎡a b (d) True. In general. c 0 ⎡1 0 ⎤ = c 2 . the determinant is ad − bc. not the sum. the value is det(A) regardless of the row or column chosen. then (h) False.SSM: Elementary Linear Algebra 0 1 29. if i + j is even then (−a)i + j = 1 and Cij = (−1)i + j M ij = M ij . if A is an n × n matrix. 1 0 0 1 0 = 0 1 0 =1 0 1 0 0 1 (c) True.1 (a) False. the determinant would be the product. Similarly. so expanding along the first row or column gives that d 2 = d1 + λ. 0 1 0 2 4 2 0 0 3 3 1 0 31. then C12 = (−1) c ⎥ f⎦ e b = −(bf − ce) and C21 = (−1) f e c = −(bf − ce). Expand along the third column.

then ⎣ 0 −1⎥⎦ ⎣0 1 ⎥⎦ det(A) = det(B) = 1. let A = ⎢ so that ⎢ ⎥ and ⎣ c d ⎥⎦ ⎢⎣ ac + cd bc + d 2 ⎦⎥ det( A2 ) = (a 2 + bc)(bc + d 2 ) − (ab + bd )(ac + cd ) = a 2 bc + a 2 d 2 + bcd 2 + b 2 c 2 − (a 2bc + 2abcd + bcd 2 ) = a 2 d 2 − 2abcd + b 2 c 2 = (ad − bc)2 = (det( A))2 Section 2. 1 0 0 0 0 1 0 −9 =1 0 0 1 0 0 0 0 1 48 .Chapter 2: Determinants SSM: Elementary Linear Algebra ⎡1 0 ⎤ ⎡ −1 0 ⎤ (i) False. let A = ⎢ and B = ⎢ . but det(A + B) = 0. det( A) = [(2)(2)(6) + (−1)(4)(5) + (3)(1)(−3)] − [(3)(2)(5) + (2)(4)(−3) + (−1)(1)(6)] = [ 24 − 20 − 9] − [30 − 24 − 6] = −5 det( AT ) = [(2)(2)(6) + (1)(−3)(3) + (5)(−1)(4)] − [(5)(2)(3) + (2)(−3)(4) + (1)(−1)(6)] = [ 24 − 9 − 20] − [30 − 24 − 6] = −5 5. 1 0 0 0 0 1 0 0 = −5 0 0 −5 0 0 0 0 1 7. The matrix is I 4 with −9 times the fourth row added to the second row. 2 ⎡a b⎤ 2 ⎡ a + bc ab + bd ⎤ = A (j) True. The matrix is I 4 with the third row multiplied by −5.2 1. det(A) = (−2)(4) − (3)(1) = −8 − 3 = −11 ⎡ −2 1⎤ AT = ⎢ ⎣ 3 4 ⎥⎦ det( AT ) = (−2)(4) − (1)(3) = −8 − 3 = −11 3. The matrix is I 4 with the second and third rows interchanged.2 Exercise Set 2. 1 0 0 0 0 0 1 0 = −1 0 1 0 0 0 0 0 1 9.

SSM: Elementary Linear Algebra 11. 3 −6 9 1 −2 3 2 7 2 3 2 7 2 − − = − − 0 1 5 0 1 5 1 −2 3 3 4 =30 0 1 5 1 −2 3 1 5 = −3 0 0 3 4 1 −2 3 1 5 = −3 0 0 0 −11 = −3(−11) = 33 17. 2 1 0 0 1 0 2 1 3 1 1 2 1 1 =− 0 3 1 0 1 1 2 1 3 1 0 2 1 0 0 1 2 3 1 0 1 1 0 1 1 −1 =− 0 2 1 0 0 1 2 3 1 0 1 1 0 1 1 −1 =− 0 0 −1 2 0 0 1 4 1 0 1 1 0 1 1 −1 = 0 0 1 −2 0 0 1 4 1 0 1 1 0 1 1 −1 = 0 0 1 −2 0 0 0 6 =6 1 3 1 5 3 1 3 −2 −7 0 −4 2 0 −1 0 0 1 0 1= 0 0 0 0 2 1 1 0 0 0 0 0 1 1 0 0 1 3 0 1 =−0 0 0 0 0 0 1 0 =−0 0 0 1 0 =−0 0 0 = −2 49 1 2 1 2 0 1 −2 1 2 0 5 6 0 1 1 3 8 1 1 1 5 3 −6 −8 0 1 1 1 1 1 3 1 5 1 −2 −6 0 1 0 0 0 1 0 0 1 3 1 5 1 −2 −6 0 1 0 0 0 1 0 0 0 3 −8 1 −1 1 3 −8 1 −1 2 . Section 2.2 0 3 1 1 1 2 1 1 2 =−0 3 1 3 2 4 3 2 4 1 1 2 1 =−0 3 0 −1 −2 1 1 2 = 0 −1 −2 0 3 1 1 1 2 =−0 1 2 0 3 1 1 1 2 =−0 1 2 0 0 −5 = −(−5) =5 15. 13.

0 1 2 2 3 − 13 1 1 1 1 2 1 3 2 3 1 3 1 0 1 2 1 2 2 3 − 13 0 0 0 = 1 0 1 1 0 1 2 1 3 1 0 3 2 0 0 3 1 0 1 2 2 = (−1) 23 13 13 − 13 32 0 1 0 1 2 2 1 = − 23 13 3 − 53 0 − 23 1 1 ⎛1⎞ 2 = − ⎜ ⎟ 25 ⎝ 3 ⎠ − 3 − 23 1 ⎡ 1 ⎛ 5 ⎞⎤ = − ⎢− − ⎜ − ⎟⎥ 3 ⎣ 3 ⎝ 6 ⎠⎦ 1 =− 6 Exercise 17: Add 2 times the first row to the second row. then expand by cofactors along the first column. then expand by cofactors along the first column. For the 3 × 3 determinant. Finally. then evaluate the 2 × 2 determinant. evaluate the 2 × 2 determinant. 1 3 1 5 3 1 3 1 5 3 −2 −7 0 −4 2 0 −1 2 6 8 0 0 1 0 1= 0 0 1 0 1 0 0 2 1 1 0 0 2 1 1 0 0 0 1 1 0 0 0 1 1 −1 2 6 8 0 1 0 1 = 0 2 1 1 0 0 1 1 1 0 1 = (−1) 2 1 1 0 1 1 1 0 1 = − 0 1 −1 0 1 1 1 −1 =− 1 1 = −[1 − (−1)] = −2 50 . Finally. Repeat these steps with the 3 × 3 determinant. then expand by cofactors of the first column. add −2 times the second row to the third row. For the 3 × 3 determinant. add −2 times the first row to the second row. Finally. then expand by cofactors along the second column. add multiples of the first row to the remaining row. For the 3 × 3 determinant. 2 1 3 1 0 1 1 −1 1 0 1 1 1 0 1 1 = 0 2 1 0 0 2 1 0 0 1 2 2 0 1 2 3 1 1 −1 = (−1) 2 1 0 1 2 3 1 1 −1 = − 0 −1 2 0 1 4 −1 2 =− 1 4 = −(−4 − 2) =6 1 Exercise 16: Add − times the first row to the 2 second row. Exercise 14: First add multiples of the first row to the remaining rows. 1 −2 3 1 1 −2 3 1 5 −9 6 3 0 1 −9 −2 = −1 2 −6 −2 0 0 −3 −1 2 8 6 1 0 12 0 −1 1 −9 −2 = 0 −3 −1 12 0 −1 1 −9 −2 = 0 −3 −1 0 108 23 −3 −1 = 108 23 = (−3)(23) − (−1)(108) = 39 Exercise 15: Add −2 times the second row to the first row.Chapter 2: Determinants SSM: Elementary Linear Algebra 19. evaluate the 2 × 2 determinant. then expand by cofactors along the first column two times. then expand by cofactors along the fourth column. evaluate the 2 × 2 determinant. then expand by cofactors along the first column.

and 4 from the third row to arrive at the determinant whose value is given.3 1. det(A) = −4 − 6 = −10 −2 4 det(2 A) = = −16 − 24 = −40 = 22 (−10) 6 8 51 . (d) False. Remove the common factor of −3 from the first row. −3a −3b −3c d e f g − 4 d h − 4e i − 4 f a b c 3 d e f =− g − 4d h − 4e i − 4 f a b c = −3 d e f g h i = −3(−6) = 18 (e) True. −1 times the second and fourth rows can be added to the sixth row without changing det(A). 27. (b) True. then the first and second rows to arrive at the determinant whose value is given. multiplying each row by its row number causes the determinant to be multiplied by the row number. 3a 3b 3c a b c − d −e − f = 3 − d −e − f 4 g 4h 4i 4 g 4h 4i a b c e f = −3 d 4 g 4h 4i a b c = −12 d e f g h i = −12(−6) = 72 = 1 b−a 1 c−a 0 b2 − a 2 c2 − a 2 b−a c−a b2 − a2 c2 − a2 = (b − a)(c 2 − a 2 ) − (c − a)(b 2 − a 2 ) = (b − a)(c + a)(c − a) − (c − a)(b + a)(b − a) = (b − a)(c − a)[c + a − (b + a)] = (b − a)(c − a)(c − b) True/False 2. d e f d e f a b c g h i = − a b c = d e f = −6 a b c g h i g h i 29. a+ g b+h c+i a b c d e f = d e f = −6 g h i g h i (c) False. so det(B) = n! det(A).5. −1 from the second row. det(B) = det(A).2 (a) True. Interchange the second and third rows. consider the transposes of the matrices so that the column operations performed on A to get B are row operations on AT .3 21. 25.2. then add 4 times the second row to the third to arrive at the matrix whose value is given. (f) True. since the columns are identical. Section 2. they are proportional. Remove the common factors of 3 from the first row. adding a multiple of one row to another does not change the determinant. Add −1 times the third row to the first row to arrive at the matrix whose value is given. Since this introduces a row of zeros. In this case. 1 a 1 b 1 c a2 b2 c2 1 = 0 23. so det(A) = 0 by Theorem 2. det(A) = 0.3 Exercise Set 2. so two such interchanges result in a multiplication by (−1)(−1) = 1. interchanging two rows of a matrix causes the determinant to be multiplied by −1.SSM: Elementary Linear Algebra Section 2.

C32 = −(−6) = 6. det( AB) = 31 1 17 10 0 2 = [18 − 170 − 0] − [80 + 0 − 62] = −170 −1 −3 6 det( BA) = 17 11 4 10 5 2 = [ −22 − 120 + 510] − [660 − 20 − 102] = −170 2 1 det( A) = 2 = 2(8 − 3) = 10 3 4 det( B) = [1 − 10 + 0] − [15 + 0 − 7] = −17 k 2 − 5k + 2 = 0. C13 = 0. so det(A) = 0 and A is not invertible. A is the same matrix as in Exercise 7. C23 = −(8 − 10) = 2. C31 = 9 − 5 = 4. C32 = −(0 + 5) = −5. det(A) = [20 − 1 + 36] − [6 + 8 − 15] = 56 −4 2 −6 det(−2 A) = −6 −4 −2 −2 −8 −10 = [ −160 + 8 − 288] − [ −48 − 64 + 120] = −448 13. 1 −3 det( A) = 2 = 2(2) = 4 0 2 21. 11. The cofactors of A are C11 = −3. which was shown to be invertible (det(A) = −1). Since det(A) ≠ 0. Expand by cofactors of the first row. 3 0 3 det( A + B) = 10 5 2 5 0 3 3 3 =5 5 3 = 5(9 − 15) = −30 ≠ det( A) + det( B) C13 = −4 + 2 = −2. 52 . C31 = 0 + 5 = 5. A is invertible. ⎡ −3 3 −2 ⎤ The matrix of cofactors is ⎢ 5 −4 2 ⎥ . 1 0 det( A) = 2 = 2(6) = 12 3 6 Since det(A) ≠ 0. 9. det( A) = (k − 3)(k − 2) − 4 = k 2 − 5k + 6 − 4 = (−2)3 (56) = k 2 − 5k + 2 Use the quadratic formula to solve 9 −1 8 5. k= −(−5) ± (−5)2 − 4(1)(2) 5 ± 17 = 2(1) 2 A is invertible for k ≠ 5 ± 17 . Add multiples of the second row to the first and third row to simplify evaluating the determinant. A is invertible. and C33 = 2. C22 = 4. and C33 = −2 + 5 = 3. Expand by cofactors of the first column. which was shown to be invertible (det(A) = 4). A is invertible. 19. C12 = 0. The third column of A is twice the first column. C21 = −(−6) = 6. C22 = 6 − 10 = −4. C12 = −(−3) = 3. 15. The cofactors of A are C11 = 2. C21 = −(15 − 20) = 5.Chapter 2: Determinants SSM: Elementary Linear Algebra 3. 0 3 5 3 5 det( A) = −1 −1 0 = −(−1) = 9 − 10 = −1 2 3 0 2 3 Since det(A) ≠ 0. C23 = 0. A is the same matrix as in Exercise 9. ⎢ ⎥ 3⎦ ⎣ 5 −5 ⎡ −3 5 5⎤ 1 1 ⎢ A−1 = adj( A) = 3 −4 −5⎥ ⎥ −1 ⎢ −2 2 det( A) 3⎦ ⎣ ⎡ 3 −5 −5 ⎤ = ⎢ −3 4 5⎥ ⎢ ⎥ ⎣ 2 −2 −3⎦ 7. det( A) = [ 2 + 12k + 36] − [ 4k + 18 + 12] = 8k + 8 A is invertible for k ≠ −1. 2 17.

1 3 1 1 0 −1 0 0 det( A) = 0 0 7 8 0 0 1 1 −1 0 0 = 0 7 8 0 1 1 7 8 = (−1) 1 1 = −(7 − 8) =1 Since det(A) ≠ 0. Use row operations and cofactor expansion to simplify the determinant. A is invertible.SSM: Elementary Linear Algebra Section 2.3 ⎡2 0 0⎤ The matrix of cofactors is ⎢ 6 4 0 ⎥ . ⎢ 0 0 −1 1⎥ 8 −7 ⎦⎥ ⎣⎢ −1 0 2 5 2 C14 = (−1) 1 3 8 = −(−6) = 6 1 3 2 A 3 1 1 C21 = (−1) 3 8 9 = −(−3) = 3 3 2 2 1 1 1 C22 = 1 8 9 = −1 1 2 2 53 −1 ⎡ −4 1 1⎢ 2 adj( A) = ⎢ = det( A) 1 ⎢ −7 ⎢⎣ 6 ⎡ −4 ⎢ 2 =⎢ ⎢ −7 ⎢⎣ 6 3 0 −1⎤ −1 0 0 ⎥ ⎥ 0 −1 8⎥ 0 1 −7 ⎥⎦ 3 0 −1⎤ 0⎥ −1 0 ⎥ 0 −1 8 ⎥ 0 1 −7 ⎥⎦ . The cofactors of A are 5 2 2 C11 = 3 8 9 = −4 3 2 2 1 3 1 C33 = 2 5 2 = −1 1 3 2 1 3 1 C34 = (−1) 2 5 2 = −(−1) = 1 1 3 2 3 1 1 C41 = (−1) 5 2 2 = −(1) = −1 3 8 9 1 1 1 C42 = 2 2 2 = 0 1 8 9 1 3 1 C43 = (−1) 2 5 2 = −(−8) = 8 1 3 9 2 2 2 C12 = (−1) 1 8 9 = −(−2) = 2 1 2 2 1 3 1 C44 = 2 5 2 = −7 1 3 8 2 5 2 C13 = 1 3 9 = −7 1 3 2 6⎤ ⎡ −4 2 −7 ⎢ 3 −1 0 0⎥ The matrix of cofactors is ⎢ ⎥. ⎢ ⎥ ⎣4 6 2⎦ ⎡2 6 4⎤ 1 1⎢ −1 adj( A) = 0 4 6 ⎥ A = 4 ⎢0 0 2⎥ det( A) ⎣ ⎦ ⎡1 ⎢2 = ⎢0 ⎢ ⎢⎣ 0 1 3 1 C23 = (−1) 1 3 9 = 0 1 3 2 1 3 1 C24 = 1 3 8 = 0 1 3 2 1⎤ ⎥ 1 32 ⎥ ⎥ 0 12 ⎥ ⎦ 3 2 3 1 1 C31 = 5 2 2 = 0 3 2 2 1 1 1 C32 = (−1) 2 2 2 = 0 1 2 2 23.

z = − . y = . 1⎤ ⎡ 1 −3 27. 11 x1 = 1 3 5 2 5 2 − 11 + 5 1 5 1 1 3 = 4(1 − 15) − 11(5 − 10) + (15 − 2) = 12 det( Ax ) −36 3 = = x= det( A) −132 11 det( Ay ) −24 2 y= = = det( A) −132 11 det( Az ) 12 1 z= = =− det( A) −132 11 3 2 1 The solution is x = .Chapter 2: Determinants SSM: Elementary Linear Algebra 1⎤ ⎡ 4 −3 A1 = ⎢ −2 −1 0 ⎥ ⎢ ⎥ ⎣ 0 0 −3⎦ 4 −3 det( A1 ) = (−3) = −3(−4 − 6) = 30 −2 −1 ⎡ 4 5 0⎤ 25. A = ⎢ 2 −1 0⎥ ⎢ ⎥ ⎣ 4 0 −3⎦ 2 −1 1 −3 + (−3) 4 0 2 −1 = (0 + 4) − 3(−1 + 6) = −11 det( A) = 54 . x2 = − . 11 11 11 det( Az ) = 4 ⎡ 3 −1 1⎤ 29. ⎡ 4 1 1 1⎤ ⎢ 3 7 −1 1⎥ 31. det(A) = 0 and Cramer’s rule does not apply. Thus. A = ⎢11 1 2⎥ ⎢ ⎥ ⎣ 1 5 2⎦ 4 5 4 5 det( A) = −2 +2 1 5 11 1 = −2(20 − 5) + 2(4 − 55) = −132 ⎡ 2 5 0⎤ Ax = ⎢ 3 1 2⎥ ⎢ ⎥ ⎣ 1 5 2⎦ 1⎤ ⎡1 4 A2 = ⎢ 2 −2 0 ⎥ ⎢ ⎥ ⎣ 4 0 −3⎦ 2 −2 1 4 det( A2 ) = + (−3) 4 0 2 −2 = (0 + 8) − 3(−2 − 8) = 38 ⎡ 1 −3 4 ⎤ A3 = ⎢ 2 −1 −2⎥ ⎢ ⎥ ⎣ 4 0 0⎦ 2 5 2 5 +2 1 5 3 1 = −2(10 − 5) + 2(2 − 15) = −36 ⎡ 4 2 0⎤ Ay = ⎢11 3 2 ⎥ ⎢ ⎥ ⎣ 1 1 2⎦ det( Ax ) = −2 det( A3 ) = 4 4 2 4 2 +2 1 1 11 3 = −2(4 − 2) + 2(12 − 22) = −24 ⎡ 4 5 2⎤ Az = ⎢11 1 3⎥ ⎢ ⎥ ⎣ 1 5 1⎦ det( Ay ) = −2 −3 4 = 4(6 + 4) = 40 −1 −2 det( A1 ) 30 30 = =− det( A) −11 11 det( A2 ) 38 38 x2 = = =− det( A) −11 11 det( A3 ) 40 40 x3 = = =− det( A) −11 11 30 38 The solution is x1 = − . A = ⎢ −1 7 −2 ⎥ ⎢ ⎥ ⎣ 2 6 −1⎦ Note that the third row of A is the sum of the first and second rows. A = ⎢ ⎥ ⎢ 7 3 −5 8⎥ ⎢⎣ 1 1 1 2 ⎥⎦ Use row operations and cofactor expansion to simplify the determinant. 11 11 40 x3 = − .

then all the entries A−1 = det( A) −1 1 det(2 A) 1 1 1 1 = =− det( A) −7 7 55 . ⎣0 1⎥⎦ ⎣ 0 −1⎥⎦ Then det(A) = det(B) = 1. ⎡ −1 0 ⎤ ⎡ 1 0⎤ (b) False.3 8 ⎛ 1⎞ (c) det(2 A−1 ) = 23 det( A−1 ) = 8 ⎜ − ⎟ = − 7 7 ⎝ ⎠ 0 −3 −3 −7 0 4 −4 −5 det( A) = 0 −4 −12 −6 1 1 1 2 −3 −3 −7 = −(1) 4 −4 −5 −4 −12 −6 −3 −3 −7 = − 4 −4 −5 0 −16 −11 (d) det((2 A)−1 ) = = 23 det( A) 1 = 8(−7) 1 =− 56 ⎛ −4 −5 −3 −7 ⎞ = − ⎜ −3 −4 − − − 16 11 16 −11 ⎟⎠ ⎝ = 3(44 − 80) + 4(33 − 112) = −424 1 1⎤ ⎡4 6 ⎢3 1 −1 1⎥ Ay = ⎢ ⎥ ⎢ 7 −3 −5 8 ⎥ 1 2 ⎦⎥ ⎣⎢ 1 3 ⎡a (e) det ⎢ b ⎢ ⎣c 0 −6 −3 7 0 −8 −4 −5 det( Ay ) = 0 −24 −12 −6 1 3 1 2 −6 −3 7 = −(1) −8 −4 −5 −24 −12 −6 −6 −3 7 = − −8 −4 −5 0 0 −34 −6 −3 = −(−34) −8 −4 = −(−34)(24 − 24) =0 det( Ay ) 0 = =0 y= det( A) −424 in A (b) det( A−1 ) = d⎤ ⎡a b e ⎥ = det ⎢ g h ⎥ ⎢ f⎦ ⎣d e ⎡a b = − det ⎢ d e ⎢ ⎣g h = − det( A) =7 c⎤ i⎥ ⎥ f⎦ c⎤ f⎥ ⎥ i⎦ 1 1 = det( A) 7 ⎛1⎞ 8 (c) det(2 A−1 ) = 23 det( A−1 ) = 8 ⎜ ⎟ = ⎝7⎠ 7 (d) det((2 A)−1 ) = = 1 det(2 A) 1 23 det( A) 1 = 8(7) 1 = 56 True/False 2. (a) det(3 A) = 33 det( A) = 27(−7) = −189 (b) det( A−1 ) = g h i 37. then all the cofactors of A are integers. det(2 A) = 23 det( A) if A is a 3 × 3 matrix. If all the entries in A are integers. (a) det(3 A) = 33 det( A) = 27(7) = 189 33.SSM: Elementary Linear Algebra Section 2. consider A = ⎢ and B = ⎢ .3 (a) False. 35. Since det(A) = 1 and 1 adj( A) = adj( A). so all the entries in adj(A) are integers. but det(A + B) = 0 ≠ det(2A) = 4. are integers.

5. a square matrix A is invertible if and only if det(A) ≠ 0. hence the −1 5 2 2 −1 5 2 0 2 −1 = −1 + (−3) 1 1 2 −1 −3 1 1 = −(2 + 1) − 3(−5 − 4) = 24 5 2 −1 5 2 −1 0 2 −1 = 0 2 −1 0 −14 −5 −3 1 1 2 −1 5 = 0 2 −1 0 0 −12 = (−1)(2)(−12) = 24 3 0 −1 1 1 0 −1 1 1 1 =3 − 4 2 4 2 0 4 2 = 3(2 − 4) − (0 + 4) = −10 3 0 −1 1 1 1 1 1=−3 0 0 4 2 0 4 1 1 = − 0 −3 0 4 1 −1 2 1 −4 2 1 1 1 = − 0 −3 −4 0 0 − 10 3 cofactor matrix has a row of zeros and the corresponding column (not row) of adj(A) is a column of zeros. (j) True. Thus. det( A) (b) (d) False. if the kth row of A is a row of zeros.Chapter 2: Determinants SSM: Elementary Linear Algebra (c) True. since E is invertible. and ( BT )T = B for any matrix B. (b) (k) True. if det(A) were not 0. A ⋅ adj(A) is a diagonal matrix with diagonal entries all equal to det(A) or (det( A)) I n . (h) True. then 1 det( A−1 ) = and det( A) 1 det( A−1BA) = ⋅ det( B) ⋅ det( A) = det( B). if A is invertible. the system Ax = 0 would have exactly one solution (the trivial solution). then so is A−1 and so both Ax = 0 and A−1x = 0 have only the trivial solution. (e) True. if A is invertible. 3. then there could be no such matrix b. (a) (f) True. if the reduced row echelon form of A were I n . since adj(A) is the transpose of the matrix of cofactors of A. (a) 3 3 −4 2 =− 3 3 −4 2 1 1 = −3 −4 2 1 1 = −3 0 6 = −3(1)(6) = −18 −4 2 = −4(3) − 2(3) = −18 3 3 56 . (l) False. ⎛ 10 ⎞ = −(1)(−3) ⎜ − ⎟ ⎝ 3⎠ = −10 Chapter 2 Supplementary Exercises 1. (a) (i) True. (b) (g) True. since A is invertible. then det(A) ≠ 0 and A−1 is invertible. the ij entry of A ⋅ adj(A) is ai1C j1 + ai 2 C j 2 + " + ain C jn . and when i = j then this sum is det(A). then every cofactor Cij = 0 for i ≠ k. If i ≠ j then this sum is 0. so det( A) ⋅ A−1 = adj( A) is invertible.

(a) 1 −9 6 0 3 1 0 −1 2 −2 1 −2 −2 3 3 1 4 1 4 1 4 = 3 0 −1 1 − 6 1 −1 1 − 1 1 0 −1 1 −9 −2 2 −9 2 −2 2 −2 2 2 ⎛ −1 1 −1 1 −2 3 ⎞ 1 4⎞ ⎛ 1 1 1 −1 ⎞ ⎛ 3 1 = 3⎜ 3 +2 ⎟ − 6 ⎜ −2 −2 2 − 1 −9 2 + 4 −9 −2 ⎟ − ⎜ −1 2 −2 − (−1) −9 2 ⎟ − − 2 2 1 1 ⎝ ⎠ ⎝ ⎠ ⎝ ⎠ = 3[3(−2 + 2) + 2(1 + 4)] − 6[ −2(−2 + 2) − (2 + 9) + 4(−2 − 9)] − [ −(−6 − 2) + (−4 + 27)] = 30 + 330 − 31 = 329 3 −2 (b) 1 −9 6 0 3 1 0 −1 2 −2 1 4 =− 1 2 1 0 −2 3 3 6 −9 2 1 0 0 3 =− 0 6 0 2 1 0 0 3 =−0 0 0 0 −1 1 0 −2 −1 −1 3 −11 −1 −1 5 31 −3 1 4 1 2 1 6 −2 11 1 6 −14 7 0 −1 1 3 −1 6 0 5 −14 0 0 − 329 15 ⎛ 329 ⎞ = −(1)(3)(5) ⎜ − ⎟ ⎝ 15 ⎠ = 329 1 0 =−0 0 9.SSM: Elementary Linear Algebra Chapter 2 Supplementary Exercises 3 −2 7. Exercise 3: −1 5 2 −1 5 2 −1 5 0 2 −1 = 0 2 −1 0 2 −3 1 1 −3 1 1 −3 1 = [(−1)(2)(1) + (5)(−1)(−3) + (2)(0)(1)] − [(2)(2)(−3) + (−1)(−1)(1) + (5)(0)(1)] = 13 − (−11) = 24 Exercise 4: −1 −2 −3 −1 −2 −3 −1 −2 −4 −5 −6 = −4 −5 −6 −4 −5 −7 −8 −9 −7 −8 −9 −7 −8 = [(−1)(−5)(−9) + (−2)(−6)(−7) + (−3)(−4)(−8)] − [(−3)(−5)(−7) + (−1)(−6)(−8) + (−2)(−4)(−9)] = −225 − (−225) =0 57 .

Let A = ⎢ 0 2 −1⎥ . C12 = −3. C13 = 0 + 6 = 6. and C33 = −2 − 0 = −2. Then det(A) = 24. so those matrices are invertible. Let A = ⎢ . 0 0 5 0 0 0 −3 0 0 −4 0 0 −1 0 0 = (5)(2)(−1)(−4)(−3) 2 0 0 0 0 0 0 0 = −120 ⎡ −4 2 ⎤ 17. The ⎣ 3 3⎥⎦ ⎡ 3 −3⎤ matrix of cofactors is ⎢ . and 4 were evaluated in Exercises 1. 3. C12 = −(0 − 3) = 3. Exercise 2: 7 −1 = 7(−6) − (−1)(−2) = −44 −2 −6 The determinants of the matrices in Exercises 1−3 are nonzero. C23 = −(−1 + 15) = −14. 13.Chapter 2: Determinants SSM: Elementary Linear Algebra Exercise 5: 3 0 −1 3 0 −1 3 0 1 1 1= 1 1 11 1 0 4 2 0 4 20 4 = [(3)(1)(2) + (0)(1)(0) + (−1)(1)(4)] − [(−1)(1)(0) + (3)(1)(4) + (0)(1)(2)] = 2 − 12 = −10 Exercise 6: −5 1 4 −5 1 4 −5 1 3 0 2 = 3 0 2 3 0 1 −2 2 1 −2 2 1 −2 = [(−5)(0)(2) + (1)(2)(1) + (4)(3)(−2)] − [(4)(0)(1) + (−5)(2)(−2) + (1)(3)(2)] = −22 − 26 = −48 11. 3. and C22 = −4. C22 = −1 + 6 = 5. Then det(A) = −18. ⎣ −2 −4 ⎥⎦ A−1 = 1 1 1 ⎡ 3 −2 ⎤ ⎡ − 6 ⎢ = adj( A) = −18 ⎢⎣ −3 −4 ⎥⎦ ⎢ 1 det( A) ⎣ 6 1⎤ 9⎥ 2⎥ 9⎦ ⎡ −1 5 2 ⎤ 19. C21 = −2. 58 . C31 = −5 − 4 = −9. The determinants for Exercises 1. C21 = −(5 − 2) = −3. The cofactors of A are C11 = 3. and 9. since its determinant is zero. 5 b−3 = 5(−3) − (b − 3)(b − 2) b − 2 −3 = −15 − b2 + 5b − 6 = −b 2 + 5b − 21 0 0 15. ⎢ ⎥ ⎣ −3 1 1⎦ The cofactors of A are C11 = 2 + 1 = 3. The matrix in Exercise 4 is not invertible. C32 = −(1 − 0) = −1.

SSM: Elementary Linear Algebra Chapter 2 Supplementary Exercises 6⎤ ⎡ 3 3 The matrix of cofactors is ⎢ −3 5 −14⎥ . The matrix form of the system is ⎡0 c b ⎤ ⎡ cos α ⎤ ⎡ a ⎤ ⎢ c 0 a ⎥ ⎢ cos β ⎥ = ⎢ b ⎥ . 1 det( A) = 0 C41 = −27. C14 = −31. α β −α 0 β −α 1−α 2 β −α 0 = β − α 1− α 2 C31 = 52. A similar construction gives the other two equations in the system. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣b a 0 ⎦ ⎣ cos γ ⎦ ⎣ c ⎦ Then cos α = 25. C13 = −21. Let A = ⎢ ⎥. C22 = −11. C42 = 16. ⎢ ⎥ ⎣ −9 −1 −2⎦ A−1 = 9 16 + =1 25 25 ⎡x − 4⎤ 5⎥ Ax′ = ⎢ 3⎥ ⎢⎣ y 5⎦ 3 4 det( Ax′ ) = x + y 5 5 3 ⎡ x⎤ ⎥ Ay′ = ⎢ 5 ⎢⎣ 54 y ⎥⎦ 3 4 4 3 det( Ay′ ) = y − x = − x + y 5 5 5 5 3 4 4 3 The solution is x ′ = x + y. C33 = −175. A = ⎢ 1 ⎢ ⎣α C21 = −2. 1 adj( A) det( A) 52 −27 ⎤ ⎡ 10 −2 1 ⎢ 55 −11 −43 16 ⎥ = ⎢ ⎥ 329 ⎢ −21 70 −175 −42 ⎥ ⎢⎣ −31 72 102 −15⎥⎦ 10 52 27 ⎤ 2 ⎡ − 329 − 329 329 ⎢ 329 ⎥ 55 43 16 ⎥ 11 − 329 − 329 ⎢ 329 329 =⎢ ⎥ 3 10 25 6 − − ⎢ − 47 47 47 47 ⎥ ⎢ 31 72 102 − 15 ⎥ 329 329 329 ⎦⎥ ⎣⎢ − 329 A−1 = 29. (a) Draw the perpendicular from the vertex of angle γ to side c as shown. ⎢ 52 −43 −175 102 ⎥ ⎣⎢ −27 16 −42 −15⎦⎥ 1 0 = 0 − (β − α )2 = −( β − α ) 2 det(A) = 0 if and only if α = β . 1 α⎤ 1 β⎥ ⎥ β 1⎦ ⎡1 27. so b a c = c1 + c2 = a cos β + b cos α . b α c1 γ c a c2 β c1 c and cos β = 2 . C23 = 70. C43 = −42. and the system has a nontrivial solution if and only if det(A) = 0. The matrix of cofactors is ⎡ 10 55 −21 −31⎤ ⎢ −2 −11 70 72 ⎥ ⎢ ⎥. C24 = 72. C34 = 102. the system is ⎡ 3 − 4 ⎤ x′ 5 ⎥ ⎡ ⎤ = ⎡ x⎤ . In matrix form. C12 = 55. ⎢5 3 ⎥ ⎣⎢ y ′⎦⎥ ⎣⎢ y ⎦⎥ ⎢ 45 5⎦ ⎣ ⎡3 − 4⎤ 5⎥ A = ⎢5 3⎥ ⎢ 45 5⎦ ⎣ 59 . C32 = −43. ⎢ 1 0 −1 1⎥ ⎣⎢ −9 2 −2 2⎦⎥ Then det(A) = 329. and C44 = −15. The cofactors of A are C11 = 10. 5 5 5 5 det( A) = ⎡ 3 −3 −9 ⎤ 1 1 ⎢ adj( A) = 3 5 −1⎥ det( A) 24 ⎢6 −14 −2 ⎥ ⎣ ⎦ ⎡ 1 − 1 − 3⎤ 8 8⎥ ⎢8 5 1 ⎥ = ⎢ 18 − 24 24 ⎢1 ⎥ 7 1 ⎢⎣ 4 − 12 − 12 ⎥⎦ ⎡ 3 6 0 1⎤ ⎢ −2 3 1 4⎥ 23. y ′ = − x + y.

223C45 + 79.223. 798C25 + 34. 1000 times the second column.Chapter 2: Determinants SSM: Elementary Linear Algebra 31. and 79. 34. 38. and 10 times the fourth column to the fifth column. That is. so det(A) = 0. c b c 0 +a b c b a = −c(c 2 − b 2 ) + a(ac − 0) = c( a 2 + b 2 − c 2 ) cos γ = c( a 2 + b 2 − c 2 ) det( A) 2abc a 2 + b2 − c 2 = 2ab det( Aγ ) 1 A. Let A be an n × n matrix for which the sum of the entries in each row is zero and let x1 be the = a (c 2 + b 2 − a 2 ) det( Aα ) cos α = det( A) a(c 2 + b 2 − a 2 ) = 2abc c 2 + b2 − a 2 = 2bc n × 1 column matrix whose entries are all one. so det( A) A−1 is invertible.154C55 .798. adj( A) = det( A) A−1 is invertible and (adj( A))−1 = (det( A) A−1 )−1 1 = ( A−1 )−1 det( A) 1 = A det( A) = adj(A −1 ) b a a b −a c 0 b a since adj( A−1 ) = det( A−1 )( A−1 )−1 = = −c(0 − ac) − a(a 2 − b 2 ) 33. 100 times the third column. Note that the column operations do not change the value of the determinant⎯consider them as row operations on the transpose. Evaluating the determinant by expanding by cofactors of the fifth column.162. Add 10. respectively.375. = −a(0 − ab) + b(c 2 − b2 ) cos β = = b( a 2 + c 2 − b 2 ) det( Aβ ) det( A) b( a 2 + c 2 − b 2 ) = 2abc 2 a + c 2 − b2 = 2ac ⎡0 c a ⎤ Aγ = ⎢ c 0 b ⎥ ⎢ ⎥ ⎣b a c ⎦ det( Aγ ) = −c Since each coefficient of Ci5 is divisible by 19. Thus. ⎡0 a b ⎤ (b) Aβ = ⎢ c b a ⎥ ⎢ ⎥ ⎣b c 0 ⎦ c a c b det( Aβ ) = −a +b b 0 b c 35. gives 21. The entries in the fifth column are now 21.154. If A is invertible then A−1 is also invertible and 0 c b det( A) = c 0 a b a 0 c a c 0 = −c +b b 0 b a = −c(0 − ab) + b(ac − 0) = 2abc ⎡a c b⎤ Aα = ⎢ b 0 a ⎥ ⎢ ⎥ ⎣ c a 0⎦ det( Aα ) = −c det(A) ≠ 0. 40. this sum must also be divisible by 19. det( A) = 60 .000 times the first column. Ax = 0 has a nontrivial solution. ⎡ 0⎤ ⎢ 0⎥ Then Ax1 = ⎢ ⎥ because each entry in the ⎢# ⎥ ⎣⎢ 0⎦⎥ product is the sum of the entries in a row of A. 375C15 + 38.162C35 + 40.

(a) (–3. 4. 4. 5) (c) y (d) z x y x z (d) y y x (3.1 (f) z y Exercise Set 3. 4. –5) x (e) (–3.Chapter 3 Euclidean Vector Spaces Section 3. –5) (3.1 x z 1. 5) z x y 61 . –4. 5) y (b) x y x z (c) (3. (a) y x x z (b) (–3. 4. 5) y 3. –4.

so y a1 = −2. − 4 − 2) = (−5. 3). 2) so b1 = 2 and b2 = 3. − 2. z) satisfies (3 − x. − 3) − (8. −2. 29) → (c) P1 P2 = (−2 − 3. − 1) y x 13. 2. (a) P1 P2 = (2 − 3. −2.Chapter 3: Euclidean Vector Spaces SSM: Elementary Linear Algebra z (e) 9. 1) (b) 2u + 7 v = (−6. One possibility is (−1. −3) + (45. − 7 − (−5)) = (−7. −4) (b) v − 3u = (0. so that x = 7. and z = −4. a2 = −2. b2 ). 2. 2. then the initial point (x. then the initial point (x. − 15)) = 2(19. −9)) = (−8. a2 . 14) = (22. 14) = (4. 1. − 2) + (0. x z (f) 11. − 6) z (e) −3(w − 2u + v) = −3((−3. and a3 = −1. 8 − 5) = (−1. (a) u + w = (4. 15) − 2((4. 0 − y. 1). − 3. 4) = (33. 2 − a3 ) = (1. 9. 5) + (−9. − 6)) = (0. x → 5. −4) = (−8. 0 − y. a3 ). Then (b1 − 1. The terminal point is B(2. −5 − z) = (4. 5) − (12. −1) + (−3. − 19. 3) 15. − 1 − a2 . (a) Let the terminal point be B(b1 . 53. b2 − 1) = (1. 7 − 8) = (−1. − 12) y x (f) (−2u − v) − 5(v + 3w) = ((−8. 1) → (b) P1 P2 = (2 − 5. y (b) One possibility is u = −v. and z = −6. y. −4). 2) − (0. (b) Let the initial point be A(a1. − 21. 6. − 11. 4 − (−2). 17) → 7. −3) = (−12. One possibility is (7. −5 − z) = (−4. Then (−1 − a1 . (a) One possibility is u = v. y = −2. −6). 20) = (37. −1). −3) = (1. −3) − 5(−9. 5 − (−7). The initial point is A(−2. (a) v − w = (4. 28) y x (d) 3v − 2(u + 2w ) = (0. − 1) − (−15. 8) → (b) P1 P2 = (−4 − 3. y. − 7) = (0. 5)) = −3(−11. 8. so that x = −1. 4. y = 2. − 2) (c) 2(u − 5w ) = 2((4. 12. 2) − (5. 5)) − 5((0. 2 − 1) = (−3. 7. (a) P1 P2 = (3 − 4. 14) 62 . 15) − 2(−2. 15) + (4. − 1) + (−6. −2. −1). 49. z) satisfies (3 − x. 1) = (−1. 3). 14) = (38. 0) + (28.

−35. 7. − 19. 2. 7. 5) (b) 2 v + 3u = (−2. −2. − 3. − 9. −2. 4. − 5. − 36) (e) − v − w = (−4. − 1. −14) = (125. −8. 6) + (6. − 3. −1. − 5. 15. 21. −36. 0) 2 9 ⎛ 11 5 ⎞ = ⎜ . 7. 2. −2. −3)) = (4. 2) − (5. 0. − 9) = (13. 5. − 2. − 1. −4. 0. −25. − 10. 12) − (5. 1. − 10. 0) 2 ⎝2 2 ⎠ 5 ⎛9 3 ⎞ = ⎜ . − 4. −2) + 3(−6. 4) = (−10. − 2 ⎟ + (−1. . −2.1 (c) −u + (v − 4w) = (3.SSM: Elementary Linear Algebra Section 3. 24) + (8. 7) (b) 6u + 2 v = (−18. 7. 24. 6) + (−1. 1. − 5) = (−2. 0) + (4. 14. (a) v − w = (4. − 2. 2)) = −2(−13. − 4) = (26. −7. 0. 2) − (6. 7. 27. 10. 3. −2. 3. − 2. −2. − 8. 3. −3. − 4. 6. − 4. 2)) = 5(25. 35. 6. 9) 17. 2. 21. 7. 7. − 5. 0) + ((4. − 1. 0) + (15. 3. . 8. 1. − 4) = (32. 12. − 3. − . 5. 0) + (−16. − 3) = (−9. 8. 15. −2. − 38. 1) = (−9. −1. 0) − (12. − 2) − (5. 2. −1. 3) = (4. 13. − 9. − 114. − 3) (f) (6v − w) − (4u + v) = ((24. 3. 1. −2) = (−13. 13. −1. 0)) + ((10. 0. 0. −2. 26. 6. − 2. −5. − 8. 0. 5. − 9) (c) −w + 3(v − u) = (4. − 16. −2) + (−18. − 1. 2) − (−5. −2) (d) 6(u − 3v) = 6((−3. 0. − 6) + (−4. 0) + (10. −26. 1. 29. 2. 0) − (5. 0. 7. −20. −18. −12) − (−4. 2. 4. 0. 4. 44. − 3. − 2. 14. 3. 2. − 4) + (−1. 0. 1. 12. 2) = (27. − 7. − 19. 6. 4)) = (3. 26. − 12) + (6. 2. 1)) − ((−12. −3. 0. −70) (e) −2(3w + v) + (2u + w) = −2((−12. − 5. − . −5. −27. 2. 0) + (20. −4. 24. − 2. − 1. 2. −1. 75. − 3. 9. 1. − 5. − 5. 2. 60. − 6)) + (−1. 5. − 6) = (−90. − 1. 10. 15. 8. 0. −3. 3. − 13. 5. 11) − (−8. − 16) (f) 1 1 (w − 5v + 2u) + v = ((−4. 32. −2) + 3((−1. (a) w − u = (−4. 0) 2 2 1 = (11. − 3. 6. 2)) = (19. 7) (d) 5(−v + 4u − w) = 5((1. 42. 6)) = 6(−15. − 2 ⎟ 2 ⎝2 2 ⎠ 19. −3. 2) − (20. 3. − 3. − 9. − 12. − 15. 28) 63 . − 5. 1. 2. 9) = (−14.

. b = −1. 12. 5a) + (2b. 1. − 64. − 62. − 3b) = (a + 2b. 0. au + bv = (a. . 7 + x4 . − 28. 3a. −2) The given vector is parallel to u. − a. 7 x5 − 5) Equate components. 7 x3 . 7 x4 . 12 + x3 . −2. 4)) = (−48. 1. 4. (c) The given vector is 0. 32. 4. b. 7 x2 . 6 + x5 ) 7x + w = (7 x1. 7 x4 + 3. 2. 0. 64 . x5 ) = (−10 + x1. 0. − 13. 5a − 3b) = (1. c3 = 5. Equating any other components gives b = −1. The scalars are a = 3. x2 . 2u − v + x = (−6. − 8. 4. − a + b. − 5) = (7 x1 + 6. Equating components gives the following system of equations. so the given vector is not parallel to u. 25. ⎢ ⎥ ⎢ ⎥ ⎣ 0 0 1 5⎦ ⎣ 0 1 4 19⎦ The scalars are c1 = 2. 8 −10 + x1 = 7 x1 + 6 ⇒ x1 = − 3 1 2 + x2 = 7 x2 − 1 ⇒ x2 = 2 8 12 + x3 = 7 x3 − 4 ⇒ x3 = 3 2 7 + x4 = 7 x4 + 3 ⇒ x4 = 3 11 6 + x5 = 7 x5 − 5 ⇒ x5 = 6 ⎛ 8 1 8 2 11 ⎞ x = ⎜− . 9. 3a. 8) − (42. 21. 2) + ( x1 . 8) − (4.Chapter 3: Euclidean Vector Spaces SSM: Elementary Linear Algebra (c) (2u − 7w ) − (8v + u) = ((−6. 2 + x2 . x5 ). which is parallel to all vectors. c1 + 3c2 = −1 −c1 + 2c2 + c3 = 1 c2 + 4c3 = 19 ⎡ 1 0 0 2⎤ ⎡ 1 3 0 −1⎤ The reduced row echelon form of ⎢ −1 2 1 1⎥ is ⎢0 1 0 −1⎥ . − 1. − 25. x3 . 16) + (−3. − 7. 18) Equating the third components give 3a = 9 or a = 3. 23) 21. 0. 7 x2 − 1. − 4. x4 . 8. 7 x5 ) + (6. 8. (b) −2u = (4. − 35)) − ((32. x4 . 43) − (29. 27. 2. (a) There is no scalar k such that ku is the given vector. 9. 94. . 1. Let x = ( x1. c2 = −1. ⎟ ⎝ 3 2 3 3 6⎠ 23. x2 . 8. x3 . 8. − 4. 20) = (−77. 3. 7 x3 − 4. −6. 2. −10.

3) = ⎜ . x = (5v + 4w ) 8 (k) False. − 7. − 2 + ⎟ 4 4⎠ ⎝ 4⎠ ⎝ ⎛ 23 9 1 ⎞ =⎜ . 3 + ⎜ − ⎟. equivalent vectors must have the same number of components. −c1 + 2c2 + 7c3 + 6c4 = 0 3c1 + c3 + 3c4 = 5 2c1 + 4c2 + c3 + c4 = 6 −c2 + 4c3 + 2c4 = −3 The reduced row echelon form of ⎡ 1 0 0 0 1⎤ ⎡ −1 2 7 6 0 ⎤ ⎢0 1 0 0 1⎥ ⎢ 3 0 1 3 5⎥ ⎥. not location. vector addition is defined component-wise. ⎛ 5 3⎞ ⎛9 1 1⎞ ⎛ 7⎞ ⎜ 2 + . if v 2 = − v1 then a1 v1 + a2 v 2 = b1 v1 + b2 v 2 as long as 1 → PQ with its initial point at P 2 gives the midpoint. c4 = 1. ⎟ 4 4⎠ ⎝ 4 65 . (a) (h) True. b. b) is a vector in 2-space. 0) is a vector in 3-space. ⎛ 15 9⎞ ⎛ 21 ⎞ ⎜ 2 + . ⎢ ⎥ ⎢ ⎥ 1 5 4⎦ ⎣0 0 0 1⎦ ⎣ 6 (e) True. 1 → 1 PQ = (7 − 2. 3) 2 7 3⎞ ⎛5 =⎜ . (b).2 29. (a. (g) False. − 7.2 Exercise Set 3. a and b must be nonzero scalars for the statement to be true.2 3 → 3 21 9 ⎞ ⎛ 15 PQ = (5. the vector −u can be added to both sides of the equation. Locating (b) Section 3. Equating components gives the following system of equations. while (a. for example. (f) False. the vectors v and −v have the same length and can be placed to be collinear. − 2 + ⎟ = ⎜ . From the bottom row. (a) v = 42 + (−3)2 = 25 = 5 v 1 ⎛ 4 3⎞ = (4. ⎟ 4 4 4 4 4⎠ ⎝ 1. v 1 ⎛ 4 3⎞ = − (4. 1 − (−2)) 2 2 1 = (5. The scalars are c1 = 1. u1 = 3 → Locating PQ with its initial point at P 4 gives the desired point. Equating components gives the following system of equations. c3 = −1. c2 = 1. vector equivalence is determined solely by length and direction. − 4 − 3. − . 1 (j) True. − ⎟ 2 2⎠ ⎝2 2 2⎠ ⎝ 2⎠ ⎝ a1 − a2 = b1 − b2 . ⎢ ⎥ is ⎢ 2 4 1 1 6 ⎢0 0 1 0 −1⎥ ⎢ ⎥ ⎣⎢0 0 0 1 1⎦⎥ ⎣⎢ 0 −1 4 2 −3⎦⎥ True/False 3. ⎟ has the u2 = − 5 v ⎝ 5 5⎠ direction opposite v. − . −2c1 − 3c2 + c3 = 0 9c1 + 2c2 + 7c3 = 5 6c1 + c2 + 5c3 = 4 The reduced row echelon form of ⎡ 1 0 1 0⎤ ⎡ −2 −3 1 0 ⎤ ⎢ 9 2 7 5⎥ is ⎢0 1 −1 0 ⎥ . so no such scalars exist. (d) True. Alternatively. (b) False. − ⎟ has the v 5 ⎝5 5⎠ same direction as v. and (a) again: v + (u + w) = v + (w + u) = ( v + w) + u = (w + v) + u. − 3) = ⎜ . 3 + ⎜ − ⎟. − .1.1 parts (a).SSM: Elementary Linear Algebra Section 3. v and kv are parallel for all values of k. (c) False. − 3) = ⎜ − . ⎟ 2 2⎠ ⎝2 (i) False. 33. 31. (k + m)(u + v) = (k + m)u + (k + m)v. but are not equal. − .1 (a) False. apply Theorem 3. the system has no solution.

Chapter 3: Euclidean Vector Spaces (b) SSM: Elementary Linear Algebra v = 22 + 22 + 22 = 12 = 2 3 ⎛ 1 1 1 ⎞ v 1 = (2. v 2 3 ⎝ 3 3 3⎠ v u2 = − v 1 =− (2. . 2) = (−2)2 + (−2) 2 + 22 = 12 =2 3 66 . − 3. 2. 2) = ⎜ . 2. 0. − 6) + (2. 4) = (3. 1.− ⎟ 15 15 15 ⎠ ⎝ 15 has the direction opposite v. ⎟ has the same direction as v. 0. 3) 15 ⎛ 1 2 1 3 ⎞ . 3) + (1. − 2. u1 = (c) v = 12 + 02 + 22 + 12 + 32 = 15 u1 = = v v 1 (1. . − 5. . 3. 7) = 32 + (−5)2 + 72 = 83 (b) u + v = (2. (a) u + v = (2. 0. 3) + (1. 3) 15 ⎛ 1 2 1 3 ⎞ = ⎜− . − 2. 1. − 6. − 3. 2) 2 3 ⎛ 1 1 1 ⎞ . 2.− . =⎜ ⎟ 15 15 15 ⎠ ⎝ 15 has the same direction as v. 2. 4. 8) = (−2.− . 4) = 22 + (−2)2 + 32 + 12 + (−3)2 + 42 = 17 + 26 (c) −2u + 2 v = (−4. v u2 = − v 1 =− (1.− = ⎜− ⎟ 3 3 3⎠ ⎝ has the direction opposite v. − 2. 0. − .

(a) u ⋅ v = (3)(2) + (1)(2) + (4)(−4) = −8 u ⋅u = u 2 = 32 + 12 + 42 = 26 v⋅v = v 2 = 22 + 22 + (−4)2 = 24 (b) u ⋅ v = (1)(2) + (1)(−2) + (4)(3) + (6)(−2) =0 u ⋅u = u 2 = 12 + 12 + 42 + 62 = 54 v⋅v = v 2 = 22 + (−2)2 + 32 + (−2)2 = 21 67 . − 6. − 5. 7) + (−6. so k = or k = − . kv = k (−2. 38. − 15. − 15) = 42 + 152 + (−15)2 = 466 5. 6) = k (−2)2 + 32 + 0 + 62 = k 49 =7 k 5 5 5 7 k = 5 if k = . 35) + (−6. − 5. 7) = 46 32 + 12 + (−5) 2 + 7 2 = 46 84 = 2 966 7. 20) + (3. 15) − 5 (3. 1. 9) − (5. 5. 1. − 4) = (4. 12. 7 7 7 9. 2. 2. 3. 15. 1. 7) = − 46 (3. − 19) = (−27)2 + (−6)2 + 382 + (−19)2 = 2570 (b) 3u − 5 v + w = (−6. 1.SSM: Elementary Linear Algebra (d) Section 3. (a) 3u − 5v + w = (−6. − 5. 1) = (−27. 6. − 6. − 25. 1) = (−6)2 + (−3)2 + 122 + 152 − 5 32 + 12 + (−5) 2 + 7 2 + (−6)2 + 22 + 12 + 12 = 414 − 5 84 + 42 = 3 46 − 10 21 + 42 (c) − u v = − (−2)2 + (−1)2 + 42 + 52 (3. 15) − (15. 1. 12. − 3.2 3u − 5v + w = (6. − 3. 0.

θ is obtuse. v) = u − v = (3 − (−4))2 + (−3 − 1)2 + (−2 − (−1)) 2 + (0 − 5) 2 + (−3 − 0)2 + (13 − (−11))2 + (5 − 4)2 = 72 + (−4)2 + (−1)2 + (−5)2 + (−3)2 + 242 + 12 = 677 13. v) = u − v = (3 − 1)2 + (3 − 0)2 + (3 − 4)2 = 22 + 32 + (−1)2 = 14 (b) d (u. θ is acute. 68 .Chapter 3: Euclidean Vector Spaces SSM: Elementary Linear Algebra 11. v) = u − v = (0 − (−3))2 + (−2 − 2)2 + (−1 − 4)2 + (1 − 4)2 = 32 + (−4)2 + (−5)2 + (−3)2 = 59 (c) d (u. (b) cos θ = u⋅v u v (0)(−3) + (−2)(2) + (−1)(4) + (1)(4) = 2 0 + (−2)2 + (−1) 2 + 12 (−3)2 + 22 + 42 + 42 −4 = =− 6 45 4 6 45 Since cos θ is negative. θ is obtuse. (c) cos θ = u⋅v u v = = =− (3)(−4) + (−3)(1) + (−2)(−1) + (0)(5) + (−3)(0) + (13)(−11) + (5)(4) 32 + (−3)2 + (−2) 2 + 02 + (−3)2 + 132 + 52 (−4)2 + 12 + (−1)2 + 52 + 02 + (−11) 2 + 42 −136 225 180 136 225 180 Since cos θ is negative. (a) cos θ = = u⋅v u v (3)(1) + (3)(0) + (3)(4) 32 + 32 + 32 12 + 02 + 42 15 = 27 17 Since cos θ is positive. (a) d (u.

(c) u⋅ v u v (2)(5) + (3)(−7) v . 2. − 3) = (−4.SSM: Elementary Linear Algebra Section 3. 19. 3 ) ( −3. 5) (1. 2. 2. ⎜ 4 2 4 ⎟⎟ ⎝ ⎠ 40 16 24 u⋅ v u v (1)(3) + (−5)(3) + (4)(3) = = 962 (−6)2 + (−2)2 42 + 02 −24 =− (c) cos θ = 13 74 11 u⋅v u v (−6)(4) + (−2)(0) = (1. (b) cos θ = (d) (u ⋅ v) − u makes sense since both u ⋅ v = and u are scalars. 5) = 2 1 + 22 + 32 + 42 + 52 (1. 5) (1. (a) u ⋅ (v ⋅ w) does not make sense because v ⋅ w is a scalar. v 69 . . 2. 7) (1. 3 ) 22 + 32 52 + (−7) 2 −11 = (b) u ⋅ (v + w) makes sense. 3 ) = (d) cos θ = (d) = =0 42 27 (−2)2 + 22 + 32 12 + 72 + (−4)2 0 17 66 25. The angle θ between the vectors is 30°. 4.866 21. − 3) = 25 3⎞ ⎛ 4 = ⎜− . ⎟ 2 2 (1. ⎟ 55 55 55 55 ⎠ ⎝ 55 8 10 3 12 + (−5)2 + 42 32 + 32 + 32 0 =0 16 ⎛ 3 1 3⎞ = ⎜− . − 3) (−4. =− u ⋅ v does not make sense because u ⋅ v is a scalar. 3.2 15. (a) cos θ = = 17. 2. 3. 3 ) 2 (−3)2 + 22 + ( 3 ) ( −3. (a) (b) (c) (−4. 4. 4. 2. . 5) = 55 ⎛ 1 2 3 4 5 ⎞ =⎜ . Divide v by v to get a unit vector that has the same direction as v. 2. then multiply by m: m 10 u⋅v u v (−2)(1) + (2)(7) + (3)(−4) = (1. 4. . . − ⎟ 5 5 ⎝ ⎠ =− = ( −3. 3. 7) 50 ⎝ 5 2 5 2 ⎠ 1 +7 ( −3. (a) u ⋅ v = (3)(4) + (2)(−1) = 10 u v = 32 + 22 42 + (−1) 2 = 13 17 ≈ 14. 7) ⎛ 1 7 ⎞ = = =⎜ . − 3) (−4)2 + (−3)2 (−4. 2. 7) (1. 3. 3 a ⋅ b = a b cosθ = 9 ⋅ 5 cos 30° = 45 2 23.

⎛ 1 1 1 ⎞ The possible vectors are ± ⎜ . y0 . 5. 27. the vectors are (e) True. and w = (1. if u = (1.3 = 10 14 ≈ 11. use the triangle inequality twice. if u = (u1 . u2 v2 < 0. u ⋅ v = 0 indicates that the angle between u and v is 90°. Let x = ( x1. x2 . v v and − . so u and v are not orthogonal. then u ⋅ v = u ⋅ w = 0. the vectors do not form an orthogonal set. (d) True. 3. 1). True/False 3. hence u1v1 + u2 v2 < 0.− ⎟. so u and v are not orthogonal. 3 3 3⎠ ⎝ (j) True. − 1) 3 ⎛ 1 1 1 ⎞ =⎜ . for example. One vector orthogonal to both u and v is x = (1. cos π 3 = v 2 ⋅ v3 = (1)(−2) + (0)(−5) + (2)(1) = 0 The vectors form an orthogonal set. v2 < 0 so u1v1 . v v (d) v1 ⋅ v 2 = (−3)(1) + (4)(2) + (−1)(5) = 0 v1 ⋅ v3 = (−3)(4) + (4)(−3) + (−1)(0) = −24 1 u⋅v .− ⎟ 3⎠ ⎝ 3 3 u1 . u2 ) and v = (v1 . It is a sphere of radius 1 centered at ( x0 . (u ⋅ v) + w does not make sense because u ⋅ v is a scalar. 1). x (1. u ⋅ x = (1)( x1 ) + (0)( x2 ) + (1)( x3 ) = x1 + x3 v ⋅ x = (0)( x1 ) + (1)( x2 ) + (1)( x3 ) = x2 + x3 (g) False. let u = (1.Chapter 3: Euclidean Vector Spaces SSM: Elementary Linear Algebra (b) u ⋅ v = (−3)(2) + (1)(−1) + (0)(3) = −7 = 7 Section 3. (b) True (c) v1 ⋅ v 2 = (−2)(1) + (1)(0) + (1)(2) = 0 v1 ⋅ v3 = (−2)(−2) + (1)(−5) + (1)(1) = 0 (c) False. so u and v are not orthogonal. u ⋅ (v + w) is a meaningful expression. 1. 1. 0 = 0. −1). (d) u ⋅ v = (2)(5) + (4)(3) + (−8)(7) = −34 ≠ 0. u v = 02 + 22 + 22 + 12 12 + 12 + 12 + 12 = 9 4 =6 (c) u ⋅ v = (−6)(3) + (0)(1) + (4)(6) = 6 ≠ 0. so u and v are orthogonal. v2 ) then (1. u + v + w = (u + v) + w ≤ u + v + w u+ v + w ≤ u + v + w 70 .2 (b) v1 ⋅ v 2 = (−1)(1) + (1)(1) = 0 The vectors form an orthogonal set. 1). v = (−1. 1) and v = (−1. 2 v = 2 v . u ⋅ v = (0)(1) + (2)(1) + (2)(1) + (1)(1) = 5 (b) u ⋅ v = (0)(1) + (0)(1) + (−1)(1) = −1 ≠ 0. x3 ). − 1) = x 12 + 12 + (−1)2 (h) False. . (f) False.3 u v = (−3)2 + 12 + 02 22 + (−1)2 + 32 Exercise Set 3. −1). .832 (c) 1. = 2 u v Since v1 ⋅ v3 ≠ 0. u2 > 0 and v1. (a) v1 ⋅ v 2 = (2)(3) + (3)(2) = 12 ≠ 0 The vectors do not form an orthogonal set. then u ⋅ v = (1)(−1) + (1)(1) = 0. (a) u ⋅ v = (6)(2) + (1)(0) + (4)(−3) = 0. z0 ). = (i) True. 1. Thus x1 = x2 = − x3 . (a) True.

a= 2 2 a a 1 1 1 1 z+ y ⇒ x− y− z =0 2 4 4 2 1 1⎞ ⎛ A normal to the plane is n 2 = ⎜1. (a) → BC = (−3 − (−2). 0. 1 − 1) = (4. u ⋅ a = (6)(3) + (2)(−9) = 0 The vector component of u along a is u⋅a 0 proja u = (3. 5) = 26 80 ⎞ ⎛ 16 = ⎜ − . 4). − 1.3 → 7. 13 ⎠ ⎝ 13 Since n1 and n 2 are not orthogonal. − 2. = u ⋅a u ⋅a 2 a a a u ⋅a a (1)(−4) + (−2)(−3) (−4) 2 + (−3)2 2 = 9. 2). the planes are not parallel. 0(x − 2) + 0(y − 0) + 2(z − 0) = 0 2z = 0 proja u = (b) 13. − . 2). = 15. AB and BC are orthogonal and the points form the vertices of a right triangle. − 4). 2) proja u = 19. 1 − 3) = (−1. 23. 2). − ⎟ . u ⋅ a = (3)(1) + (1)(0) + (−7)(5) = −32 17. − ⎟ . 4 2⎠ ⎝ Since n1 = 8n 2 . − 2) = → CA = (1 − (−3). − 1. 0. 71 . 2. 3 − 1) = (−3. 1) and a normal to x + 2z = −1 is n 2 = (1. AB = (−2 − 1. 0) = (6. 0. − 1. 1. − 1 − 0. − ⎟ . = Since n1 and n 2 are not parallel. −2(x − (−1)) + 1(y − 3) + (−1)(z − (−2)) = 0 −2(x + 1) + (y − 3) − (z + 2) = 0 2 a 25 2 = 5 11. A normal to 4x − y + 2z = 5 is n1 = (4. the planes are not perpendicular. 13 ⎠ ⎝ 13 The vector component of u orthogonal to a is 80 ⎞ ⎛ 16 u − proja u = (3. 0) → → = AB ⋅ BC = (−3)(−1) + (−1)(−1) + (2)(−2) = 0 → → → → Since AB ⋅ BC = 0. 1. − 1. − 7) − ⎜ − . A normal to 3x − y + z − 4 = 0 is n1 = (3. − 9) = (0. the planes are parallel. 2y = 8x − 4z + 5 ⇒ 8x − 2y − 4z = −5 A normal to the plane is n1 = (8. 0 − 1.SSM: Elementary Linear Algebra Section 3. 2) − (0. x= The vector component of u orthogonal to a is u − proja u = (6. 1 − (−1). 0). a n1 ⋅ n 2 = (3)(1) + (−1)(0) + (1)(2) = 5 2 = 12 + 02 + 52 = 26 The vector component of u along a is u⋅a proja u = a 2 a −32 (1. 0. u ⋅a a (3)(2) + (0)(3) + (4)(3) 22 + 32 + 32 18 22 21. − ⎟ 13 ⎠ ⎝ 13 11 ⎞ ⎛ 55 = ⎜ . A normal to 7x − 3y + 4z = 8 is n 2 = (7. − 3.

. ax + by0 + cz0 + d D= 0 a2 + b2 + c2 2(−1) + 3(2) − 4(1) − 1 = 22 + 32 + (−4)2 −1 = 29 1 = 29 37. The distance between the planes is 0. − ⎟ . − 2) 40 1⎞ ⎛1 1 1 = ⎜ .Chapter 3: Euclidean Vector Spaces SSM: Elementary Linear Algebra 25. − ⎟ ⎝ 5 5⎠ ⎛ 3 6⎞ = ⎜ 1. The plane is 2x + 3y − 4z − 1 = 0. = 6 = =1 −5 42 + 32 25 39. 2) − ⎜ . . u ⋅ a = (2)(4) + (1)(−4) + (1)(2) + (2)(−2) = 2 a 2 = 42 + (−4)2 + 22 + (−2)2 = 40 The vector component of u along a is u⋅a proja u = a 2 a 2 = (4. The ax0 + by0 + c distance between P0 and the plane a2 + b2 4(−3) + 3(1) + 4 −4x + 2y + 2z = 12 is −4(0) + 2(−5) + 2(0) − 12 D= (−4)2 + 22 + 22 −22 = 24 22 = 2 6 11 . − ⎟. . 5⎠ ⎝ 5 The vector component of u orthogonal to a is ⎛ 2 1⎞ u − proja u = (1. 0). − ⎟ ⎝ 5 5 10 10 ⎠ ⎛ 9 6 9 21 ⎞ = ⎜ . ⎝ 5 5 10 10 ⎠ 29. . ax + by0 + cz0 + d D= 0 a2 + b2 + c2 (1)(3) + 2(1) − 2(−2) − 4 = 12 + 22 + (−2)2 5 = 9 5 = 3 27. u ⋅ a = (1)(0) + (1)(2) + (1)(−1) = 1 a 2 2 2 31. ⎟ . 2. 5 5 10 10 ⎝ ⎠ The vector component of u orthogonal to a is 1⎞ ⎛1 1 1 u − proja u = (2. 1. − 4. . so the planes coincide. 1) − ⎜ 0. ⎝ 5 5⎠ 33. 72 . . ⎟. − . Note that the equation of the second plane is −2 times the equation of the first plane. − . 1. A point on 2x − y − z = 5 is P0 (0. The line is 4x + y − 2 = 0. D = = 35. The plane is x + 2y − 2z − 4 = 0. 1. 2. . − 1) 5 ⎛ 2 1⎞ = ⎜ 0. ax + by0 + c D= 0 a 2 + b2 4(2) + (1)(−5) − 2 = 42 + 12 1 = 17 2 = 0 + 2 + (−1) = 5 The vector component of u along a is u⋅a proja u = a 2 a 1 = (0. − 5.

0) and v 2 = (5. 0) + t2 (−1. For t = 0. 0. 3. For t = 0. The vector equation is x = x0 + tv. x = −3 − 5t2 . since projb u has the same direction as b. 6 − 6t) The coefficients of t give the parallel vector (−6. the vector equation is (x. the orthogonal projection of u on a has the same direction as a while the vector component of u orthogonal to a is orthogonal (perpendicular) to a. 1) Equating components gives the parametric equations. y. − 1. then proja v = v. y) = t(3. z ) = (−3. ( x. 4) and the parametric equations are x = 5t2 . z = 6t1 + 2t2 43. 1. 0) = (4 − 6t. y) = (−4. the vector equation is (e) True. cos2 α + cos2 β + cos2 γ = 1 or cos2 α + cos2 β + cos2 γ = 1. z = t c . so 5. 7). the point is (3. z ) = t1 (0. its magnitude is v 7. the zero vector is orthogonal to all vectors. u + v ≤ u + v . y. Two possible vectors that are orthogonal to v are v1 = (0. 2). cos γ ) v ⎜⎝ v v v ⎟⎠ (d) Since 1. (d) True. z = 4 + t2 (a) True. Using v1 and v 2 . 9. 4). 1. The coefficients of t give the parallel vector (−5. (b) True.4 41. v v ⎛ a b c ⎞ =⎜ . y = 2t. . (f) False. (x. x = −4. x = −3t. y. y = t1. 1. Since the given point is the origin. 6). 1. (c) True. the vector equation is x = tv. 1) Equating components gives the parametric equations. ( x. 4) + t1 (6. u = (1. z ) = (−1. Using w. True/False 3. y = 1 − t1 + 3t2 . proja u = proja v. 0). cos β = cos γ = (c) Section 3. 0) + t2 (5. so v ⋅i cos α = v i (a)(1) + (b)(0) + (c)(0) = v ⋅1 a = . for a = (1. z) = t(−3. −8) Equating components gives the parametric equations. 10) and v = (1. 6 − 6t) + (−2t. (x.3 ( x. x = (4 − 4t. The vector equation is x = x0 + t1v1 + t2 v 2 . 1) + t(0. 2) and the parametric equations are x = 3t. cos β . ⎟ = (cos α . the point is (4. 6) + t2 (−5. x = −1 + 6t1 − t2 . 0. 73 . z = 4t2 . 0) + t1 (0. One vector orthogonal to v is w = (3. 0. (g) False. − 3. (ku) ⋅ (mv) = km(u ⋅ v) = km(0) = 0. v is a unit vector.4 1. v ⋅ (k1w1 + k2 w 2 ) = v ⋅ (k1w1 ) + v ⋅ (k2 w 2 ) = k1 ( v ⋅ w1 ) + k2 ( v ⋅ w 2 ) = k1 (0) + k2 (0) =0 11. 15. y. −1). if v has the same direction as a. y = 0.4 Exercise Set 3. v (b) Similar to part (a). y = 1 − 8t b and v 3. −6). 13. −6). (a) α is the angle between v and i. 2) Equating components gives the parametric equations. 1.SSM: Elementary Linear Algebra Section 3. The vector equation is x = x0 + t1v1 + t2 v 2 . y = 1 − 3t1 + t2 .

(b) Geometrically. The system is −2 x + 3 y =0 ⎡1 0 ⎡ 1 1 1 0⎤ ⎢ (b) The reduced row echelon form of ⎢ is ⎢⎣0 1 ⎣ −2 3 0 0⎥⎦ 3 5 2 5 0⎤ ⎥ so the solution will require one 0 ⎥⎦ parameter. the solution space is a line through the origin in R3 . t ⎟ in vector form. 2). (a) If x = (x. (a) A particular solution of x + y + z = 1 is (1. ⎢ ⎥ ⎣0 0 0 0⎦ ⎡ 1 1 1⎤ Since the row vectors of ⎢ 2 2 2 ⎥ are all multiples of (1. 1. 0) and parallel to the vectors (−1. x5 = t . or x = (−r − s. − r + s + t . 5. x2 = − r + s + t . 3. 0) + t(−1. − 1. 1. 1). 7 7 7 7 7 7 ⎝ ⎠ For this system. 2. The equation can be represented by (1. 74 . x2 = r . ⎢ ⎥ ⎣ 3 3 3⎦ r1 ⋅ x = (1. z = t. which is s(−1. x + y+z =0. 0) + t(−1. passing through the point P(1. The augmented matrix of the system is ⎢ 2 2 2 0 ⎥ which has reduced row echelon form ⎢ ⎥ ⎣ 3 3 3 0⎦ The solution of the system is x1 = −r − s. 0. 1. 0. 0. 1). r . s. 1. then a ⋅ x = x + y + z = 0 and b ⋅ x = −2x + 3y = 0. y. 0) and (−1. 1) ⋅ (− r − s. or 7 7 7 7 7 7 19 8 2 1 3 ⎛3 ⎞ x = ⎜ r − s − t . only r1 needs to be considered. 1). 2 − 1 − 3 0⎥ ⎢0 1 7 7 7 ⎣ ⎦ 3 19 8 2 1 3 The solution of the system is x1 = r − s − t . x3 = r . 0) + s(−1. Since only one parameter is required. r . z) is orthogonal to a and b. r. 0. 1) in vector form. The general solution of x + y + z = 0 is x = −s − t. 1. the form of the equation indicates that it is a plane in R3 . x3 = s. 19 8 ⎞ ⎛ 2 1 3 ⎞ ⎛3 r1 ⋅ x = 1⎜ r − s − t ⎟ + 5 ⎜ − r + s + t ⎟ + 1(r ) + 2( s) − 1(t ) 7 7 7 7 7 7 ⎝ ⎠ ⎝ ⎠ 3 19 8 10 5 15 = r − s − t − r + s + t + r + 2s − t 7 7 7 7 7 7 =0 19 8 ⎞ ⎛ 2 1 3 ⎞ ⎛3 r2 ⋅ x = 1⎜ r − s − t ⎟ − 2 ⎜ − r + s + t ⎟ − 1(r ) + 3( s) + 2(t ) 7 7 ⎠ ⎝ 7 7 7 ⎠ ⎝7 3 19 8 4 2 6 = r − s − t + r − s − t − r + 3s + 2t 7 7 7 7 7 7 =0 21. s) = 1(−r − s) + 1(r ) + 1( s) =0 ⎡1 5 1 2 −1 0 ⎤ which has reduced row echelon form 19. y = s. The augmented matrix of the system is ⎢ ⎣1 −2 −1 3 2 0 ⎥⎦ 8 0⎤ ⎡ 1 0 − 3 19 7 7 7 ⎢ ⎥. r1 = (1. ⎡ 1 1 1 0⎤ ⎢0 0 0 0⎥ .Chapter 3: Euclidean Vector Spaces SSM: Elementary Linear Algebra ⎡ 1 1 1 0⎤ 17. 0). x4 = s. 1. − 2. − 1) and r2 = (1. 23. s) in vector form. 0. 0.

6 7 2 7 2 6 ⎟⎠ ⎝ = (14 + 18.SSM: Elementary Linear Algebra Section 3. 1 4 32 − − −4 32 −6 ⎟⎠ ⎝ = (−8 − 6. 1) to 1 ⎛ 2 ⎞ ⎜ − s + t . − 18 − 64) = (−14. 3 3 1. x2 = s. − 4) ⎛ 2 −6 3 −1 3 2 ⎞ =⎜ . and if x0 is a solution vector of the system. x3 = 0.− .5 x2 = s. x3 = t . If b ≠ 0. The reduced row echelon form of ⎡1 4 1 0 1⎤ ⎡ 3 4 1 2 3⎤ 3 3 3 ⎢6 8 2 5 7 ⎥ is ⎢⎢0 0 0 1 1⎥⎥ . a general solution of 4 1 the homogeneous system is x1 = − s − t . (d) True. then by matrix multiplication. 0. 3 3 3 From the general solution of the nonhomogeneous system. (b) False. 1. − 82) 75 . − 4) 27. x2 = s. 0). two non collinear vectors that are parallel to the plane are needed.4. x2 = 0. 3 3 (b) u × ( v × w ) = (3. 3 x2 = 0. a particular solution of Ax = b must be used to obtain the general solution of Ax = b from the general solution of Ax = 0.5 Exercise Set 3. x4 = 0 and a particular solution (c) From part (b). 6. or 5 5 2 ⎞ ⎛ 3 x = ⎜ − t . (a) v × w = (0. if b = 0. the vector equation is then x = x0 + tv where x0 is the given point and v is the given vector. 3 3 (c) True. − 6.− . 3. x2 = s. True/False 3. 0 − 4) = (32. x3 = 1 + t . x3 = 1 is a solution of 3x1 + 2 x2 − x3 = 2. y = − t . then the statement is true by Theorem 3. the first. the equation of the line is x = tv where v is any nonzero vector on the line.4 ⎛ 3 ⎞ ⎛ 2 ⎞ r1 ⋅ x = 1⎜ − t ⎟ + 1⎜ − t ⎟ + 1(t ) = 0 ⎝ 5 ⎠ ⎝ 5 ⎠ ⎛ 3 ⎞ ⎛ 2 ⎞ r2 ⋅ x = −2 ⎜ − t ⎟ + 3 ⎜ − t ⎟ + 0(t ) = 0 ⎝ 5 ⎠ ⎝ 5 ⎠ (a) True. s. A(x1 − x2 ) = Ax1 − Ax2 = b − b = 0. Here. (b) Since the second and third rows of ⎡ 3 2 −1 2 ⎤ ⎢ 6 4 −2 4 ⎥ are scalar multiples of ⎢ ⎥ 1 −2 ⎦ ⎣ −3 −2 (e) False. 2. − (−12 + 32). 1 of the nonhomogeneous system is x1 = . 2. − t . add (1. x4 = 1. − 20. x3 = t . 7) ⎛ 2 −3 0 −3 0 2 ⎞ =⎜ . (a) The reduced row echelon form of ⎡ 1 2 − 1 0⎤ ⎡ 3 2 −1 0 ⎤ 3 3 ⎥ ⎢ 6 4 −2 0 ⎥ is ⎢⎢0 0 0 0⎥ . ⎢ ⎥ ⎢0 0 0 0 0⎥ ⎣9 12 3 10 13⎦ ⎣ ⎦ A general solution of the system is 1 4 1 x1 = − s − t . Section 3. ⎢ ⎥ ⎢0 0 1 0⎦ 0 0⎥ ⎣ −3 −2 ⎣ ⎦ The solution of the system is 2 1 x1 = − s + t . t ⎟ to get 3 3 ⎝ ⎠ 2 1 x1 = 1 − s + t . − 3) × (2. x3 = t . so that there is a nonzero entry. 1) and r2 = (−2. x4 = 1. say bi of b. 25.3. z = t. (f) True. t ⎟ in vector form. − (0 + 6). it suffices to show that x1 = 1. − 6. ri ⋅ x0 = bi where ri is the ith row vector of A.5 (c) In vector form. 3(1) + 2(0) − 1 = 3 − 1 = 2. − 1) × (32. 5 ⎠ ⎝ 5 r1 = (1. the solution of the system is 3 2 x = − t .

⎛4 0 1 0 1 4⎞ . 0 1 0 3 ⎟⎠ ⎝ 3 1 = (−1 − 6. 3. − 24 − 18) = (27. 1) ⎛ −1 2 1 2 1 −1 ⎞ =⎜ . 2) 3. − 5. but the direction is opposite.− . − 3) ⎛ 2 −1 3 −1 3 2 ⎞ =⎜ . − 3) u × v = 02 + 02 + (−3) 2 = 3 The area is 3. 6. 9. 0) and v = (3. 0). → The triangle can be considered in R3 as being half of the parallelogram formed by u = (1. − 42) → P1 P4 = (3. 7) ⎛9 6 −4 6 −4 9 ⎞ =⎜ . 10) u × v = (−7)2 + (−1) 2 + 32 = 59 u × v = (−15)2 + 7 2 + 102 = 374 59. 6 − 0) = (−4. u × v = (−6. 0. 7) → 1 374 u× v = .− . 0) and v = (−3. ⎛2 0 3 0 3 2⎞ . 0). 3. u× v = ⎜ −3 0 −3 2 ⎟⎠ ⎝2 0 = (0. 3) The area is → 13.Chapter 3: Euclidean Vector Spaces SSM: Elementary Linear Algebra (c) u × v = (3. 7. 3). − (−9 − 0). 2. 9. u × v = (1. − 1. 2. 5) ⎛4 2 −6 2 −6 4 ⎞ =⎜ . 3 − 0) = (−7. 14) 5. 1. 5) × (3. 2) × (3. 2) × (0. u ⋅ ( v × w) = 0 4 −2 2 2 −4 4 −2 −6 2 =2 +2 2 −4 4 −2 = 2(−12) + 2(4) = −16 The volume of the parallelepiped is −16 = 16. Let u = P1 P2 = (−1. u× v = ⎜ 3 0 3 1 ⎟⎠ ⎝1 0 = (0. thus P1 P2 and P1 P4 are adjacent sides.− . ⎛ −5 2 −1 2 −1 −5 ⎞ .− . 40. 4. − (−4 − 0). AB = (1. 4) and AC = (−3. 1 5 3 5 3 1 ⎟⎠ ⎝ = (20 − 2. 0 3 3 −3 3 0 ⎟⎠ − ⎝ = (−3 − 0. 0. 2) and P3 P4 = (−3. 0) × (−1. 1) The parallelogram can be considered in R3 as being determined by u = (3. P1 P2 = (3.− . 2. − (−30 − 6). u× v = ⎜ 0 3 2 3 2 0 ⎟⎠ ⎝ = (−15. 2 2 1 −2 −1 2 ⎟⎠ − − ⎝ = (−6 − 0. 1. − (−28 − 12). 4. 9. − 3) u × v = 02 + 02 + 142 = 14 The area of the triangle is → → v = P1P3 = (2. → 11. − (1 − 0).− .− . 4 + 3) = (−6. 6 7 2 7 2 6 ⎟⎠ ⎝ = (63 − 36. 6) (u × v) × w = (−4. u × v = (−2. 0. 0 − 3) = (−3. u × v = (−6)2 + 42 + 72 = 101 The area is 1 u × v = 7. u × v is orthogonal to both u and v. − 1) × (0. − 1. 0. 2 3 0 − −3 0 2 ⎟⎠ ⎝ = (−6 + 2. 2 2 2 −6 2 17.− .− . − 2) ⎛3 0 2 0 2 3⎞ =⎜ . − 6 − 12) = (18. 76 . − 2) so the side determined by P1 and P2 is parallel to the side determined by P3 and P4 . − 18) 101. − 3) ⎛1 5 −2 5 −2 1 ⎞ =⎜ . 2. 2 15. 1. 36. 4. 2. 2) and 7. − (6 − 15). u × v is orthogonal to both u and v. u × v = (2. 6)× (2. The area of the triangle is 9.

⎛2 2 −1 2 −1 2 ⎞ . − 3) u × v = (−4) 2 + 12 + (−3) 2 = 26 The area of the triangle is 1 26 u× v = . 2) and → v = AC = (1.5 1 −1 −2 19. 1. 4) 1 2 −1 a ⋅ (b × c) = 3 4 0 −1 −3 4 3 4 1 = −1 +4 3 −1 −3 = −(−5) + 4(−2) = −3 1 The volume is a ⋅ (b × c) = 6 → 27. 0) → c = PS = (−1.) (b) (v × w) ⋅ u = u ⋅ (v × w) = 3 ⎡ w1 w2 w3 ⎤ (c) w ⋅ (u × v) = det ⎢ u1 u2 u3 ⎥ = 3 ⎢ ⎥ ⎣⎢ v1 v2 v3 ⎦⎥ (Rows 2 and 3 were interchanged. 6 2 . 6 6 ⎡ u1 u2 u3 ⎤ (a) u ⋅ (w × v) = det ⎢ w1 w2 w3 ⎥ = −3 ⎢ ⎥ ⎣⎢ v1 v2 v3 ⎦⎥ (Rows 2 and 3 were interchanged. − 4. ⎢ ⎥ ⎢⎣ w1 w2 w3 ⎥⎦ c = PS = (4. (a) Let u = AB = (−1. − 3.SSM: Elementary Linear Algebra Section 3. u ⋅ ( v × w) = 29. then 1 → 26 x = 2 AB 2 26 26 x= = → 3 AB 21. u ⋅ ( v × w) = 0 b 0 = abc 0 0 c 37. − 1.) → (b) a = PQ = (1. 4. − 1) → b = PR = (3. 1. Since u ⋅ (v × w) = 3. − 3) → → b = PR = (2. (a) a = PQ = (3. (b) −2 0 6 1 −3 1 −5 −1 1 1 −3 −3 1 = −2 +6 −1 1 −5 −1 = −2(−2) + 6(−16) = −92 → = (−1)2 + 22 + 22 = 3 AB Let x be the length of the altitude from vertex C to side AB. 2.− . − 1. (u + v) × (u − v) = (u + v ) × u − (u + v ) × v = (u × u) + ( v × u) − ((u × v) + ( v × v)) = 0 + ( v × u ) − (u × v ) − 0 = ( v × u) + ( v × u) = 2( v × u) a 0 0 23. then rows 1 and 2 were interchanged. 2. then ⎡ u1 u2 u3 ⎤ det ⎢ v1 v2 v3 ⎥ = 3. u× v = ⎜ 1 1 1 −1 1 1 ⎟⎠ − ⎝ = (−4. u ⋅ ( v × w) = 3 0 −2 5 −4 0 3 0 −1 −2 = +2 5 −4 5 −4 = −12 + 2(14) = 16 The vectors do not lie in the same plane. − 1). 1) → 25. 3) 3 −1 −3 a ⋅ (b × c) = 2 −1 1 4 −4 3 2 1 2 −1 −1 1 =3 +1 −3 4 3 4 −4 −4 3 = 3(1) + 1(2) − 3(−4) = 17 1 17 The volume is a ⋅ (b × c) = . 2 2 77 2 4 3 1 = .

but v ≠ w. 0) − (−4.e. − 5) = (36. − 2430. 5) 3. = 22 + (−5)2 + (−5)2 = 54 projw u = u⋅w w 2 w −24 (2. 0. 12. if v and w are distinct vectors that are both parallel to u. 1 + 0 − 6) = (4. 7) u+ v+w = (−2 − 3 + 9. 36 −90 36 −90 ⎟⎠ ⎝ −90 −90 = (−3150.5 w (a) True. − 12) − (3. (d) True (e) False. − 1. 10) (b) 2 u+v+w = (−2 + 3 + 2. 4 + 6 − 5) = (3. 18) − (−4. for nonzero vectors u and v. −2 0 4 (e) u ⋅ ( v × w) = 3 −1 6 2 −5 −5 3 −1 −1 6 = −2 +4 2 −5 −5 −5 = −2(35) + 4(−13) = −122 (c) False. − 30) + (2. − 6. − 5) = 18(2. − 5) 27 (b) True. − 35) (u ⋅ v)w = ((−2)(3) + (0)(−1) + (4)(6))(2. − 5. − 25) = (−7. 5. − 90) (−5 v + w ) × (u ⋅ v)w ⎛ 0 −35 −13 −35 −13 0⎞ =⎜ . 20. − 25. − 5) = 42 + 7 2 + 42 + (−5) 2 = 106 = (−7)2 + 262 + 7 2 = 774 78 . the scalar triple product is a scalar. − 5. −3u − v − 5w = (6. 2) = (−5. hence normal to the plane containing the vectors. 0. the cross product of two nonzero and non collinear vectors will be perpendicular to both vectors. − 3. 6 + 0 + 1. for example. 7. (a) 3v − 2u = (−9. 26. 8) = (13. − 2) = 32 + (−6)2 + 52 = 70 (b) (c) The distance between −3u and v + 5w is −3u − ( v + 5w ) = −3u − v − 5w . − 3. 4. − 5. − 5) = (−13. u × v = u v sin θ will only be 0 if θ = 0. 0. 1170) (f) False. 6) − (10. for example (i × i) × j = 0 × j = 0 i × (i × j) = i × k = −j (f) −5v + w = (−15. − 12. 0.. not a vector. Chapter 3 Supplementary Exercises 1. 24. − 5. − 90. i. then u × v = u × w = 0. 0 − 1 − 5. − 5) = 54 12 = − (2.Chapter 3: Euclidean Vector Spaces SSM: Elementary Linear Algebra (d) u ⋅ w = (−2)(2) + (0)(−5) + (4)(−5) = −24 True/False 3. if u and v are parallel. − . (a) 3v − 2u = (9. 2 + 8 − 6. − 5. 4.

27) = (−36) 2 + (−23)2 + 162 + 272 = 2810 (d) u ⋅ w = (−2)(9) + (6)(1) + (2)(−6) + (1)(−6) = −30 w 2 = 92 + 12 + (−6)2 + (−6)2 = 154 projw u = u⋅w w 2 w −30 (9.SSM: Elementary Linear Algebra Chapter 3 Supplementary Exercises (c) The distance between −3u and v + 5w is −3u − ( v + 5w ) = −3u − v − 5w . 0. 16. The point is S(−1. −1. if u + v 2 2 = u + u ⋅ v + v ⋅u + v 2 2 + v 2 . 7. 79 . − 23. w = (1. s3 ). 0) − (45. 9. 2) u ⋅ v = (−32)(3) + (−1)(−1) + (19)(5) = 0 u ⋅ w = (−32)(1) + (−1)(6) + (19)(2) = 0 v ⋅ w = (3)(1) + (−1)(6) + (5)(2) = 7 Since v ⋅ w ≠ 0. u + v 2 = (u + v) ⋅ (u + v) = u ⋅u + u ⋅ v + v ⋅u + v ⋅ v = u Thus. − 30) = (−36. ⎛ 1 3 3 1 ⎞ −2 −2 . 5). 3s2 + 3) If u × v = 0. − 3) − (−3. − 6. (b) The set of all such vectors is the plane through the origin which is perpendicular to the given vector. which also causes the first component to be 0. − 6. 19). 5). − (3( s3 − 1) − 12). 1. s2 . −1. u and v are parallel if u × v = 0. then from the second and third components. 1. − 6) = 154 15 = − (9. Let S be S (−1. (c) The only vector in R 2 that can be orthogonal to two non collinear vectors is 0. the origin. s3 − 1). so u and v are orthogonal.− . the set is {0}. 5. u = (−32. − 30. u ⋅ v = v ⋅ u = 0. − 3s3 + 15. → → 11. − 6. 6. u× v = ⎜ 1 1 6 1 6 s s s s − − − − − − 1 ⎟⎠ 2 3 3 2 ⎝ = ( s3 − 1 + 2( s2 − 1). (a) The set of all such vectors is the line through the origin which is perpendicular to the given vector. the vectors do not form an orthogonal set. 3( s2 − 1) + 6) = (2s2 + s3 − 3. − 2) and v = RS = (−6. s2 = −1 and s3 = 5. (d) The set of all such vectors is the line through the origin which is perpendicular to the plane containing the two given vectors. 8. True. −1. −3u − v − 5w = (6. v = (3. − 18. s2 − 1. − 6) 77 5. Then u = PQ = (3. 1.

− 1. 29. − 1) and and v = PR = (5. −3) + t(8. − 1. → 13. y = 1 − 2t1 − t2 . From the vector equation. Equating components gives the parametric equations x = 8t. 6) is a point on the plane and (0. 0) ⎛ −1 3 0 3 0 −1 ⎞ =⎜ . − 2) + t2 (5. 19. −9 −6 −9 6 ⎟⎠ ⎝ 6 −6 = (−18. − 51. 2 0 2 −1 ⎟⎠ ⎝ −1 0 = (3. −5). 3) is parallel 1 to the line. − 2). v = PR = (−9. 1. 1. −1. − 3) cos θ = = = u⋅v u v (3)(2) + (1)(2) + (−2)(−3) 32 + 12 + (−2)2 22 + 22 + (−3)2 14 23. v = PR = (2. y) = (0. a point-normal equation for the plane is −18(x − 9) − 51y − 24(z − 4) = 0. y. Since the slope 3 of the line is m = . 14 17 14 = 17 15. z = 3 − 2t1 − 5t2 . − 5). Equating components gives the parametric equations x = −2 + t1 + 5t2 . (−1. − 1. − 1. 3). z ) = (−2. 80 . One point on the line is (0. The plane is 5x − 3y + z + 4 = 0. Using this point and vector gives the vector equation (x. 5(−3) − 3(1) + 3 + 4 11 D= = 35 52 + (−3) 2 + 12 → 17. 6. −1. 5. − 24) Using point P(9. A normal to the plane is u × v. 4. 3) × (2. u = PQ = (3. The equation represents a plane perpendicular to the xy-plane which intersects the xy-plane along the line Ax + By = 0. → Using P(−2. 0) will be a normal to the plane. − 2. the vector (1. 2.− . 3). the vector equation is ( x. y) = (0.− . (0. ⎛ 4 −1 −10 −1 −10 4 ⎞ u× v = ⎜ . 1. 3) × (2. 6. 4). −5) + t(1. Equating components gives the parametric equations x = t. y = −3 − t. − 2) 25. −1). 2) A point-normal equation for the plane is 3(x + 1) + 6(y − 5) + 2(z − 6) = 0. The vector equation is (x.Chapter 3: Euclidean Vector Spaces → SSM: Elementary Linear Algebra 21. 0. Two vectors in the plane are → → u = PQ = (−10. − 6). − 5). 3) + t1 (1. − 2. The plane will contain u = PQ = (1. y = −5 + 3t.

6) (b) The sum of any two real numbers is a real number. 4) = (−1 + 3.1 (a) False.1 1. u2 ) = (0. vectors are not restricted to being n-tuples of real numbers. 3. u+w = v+w (u + w) + (− w) = ( v + w) + (− w) u + [ w + (−w )] = v + [ w + (−w)] u+0 = v+0 u=v Hypothesis Add − w to both sides. (c) Axioms 1−5 hold in V because they also hold in R 2 . (a) u + v = (−1. Axiom 6 fails to hold for k < 0. 23. The set is a vector space with the given operations. 5.Chapter 4 General Vector Spaces Section 4. 2) = (0. u2 ) ≠ u. (1) Axiom 7 (2) Axiom 4 (3) Axiom 5 (4) Axiom 1 (5) Axiom 3 (6) Axiom 5 (7) Axiom 4 True/False 4. (c) True 81 . 7. The set is not a vector space. vectors are not restricted to being directed line segments. The set is a vector space with the given operations. 6) 3u = 3(−1. The set is a vector space with the given operations.1 Exercise Set 4. Axiom 3 Axiom 5 Axiom 4 25. Axiom 5 fails to hold because of the restriction that x ≥ 0. Then 1u = 1(u1 . (b) False. The product of two real numbers is a real number and 0 is a real number. (e) Let u = (u1 . u2 ) with u1 ≠ 0. 11. 2 + 4) = (2. The set is not a vector space. 9. 2) + (3. Axiom 8 fails to hold because (k + m)2 ≠ k 2 + m 2 .

Call the other vector u. (d) (0. Since u 2 2 1 1 would be an element of V. Consider the scalar . (d) This is not a subspace of R3 . 3 (e) This is a subspace of R . Thin (0k1 + k2 . 0 + u = u. c). Then 0 + 0 = 0. u = 0 or u = u. Section 4. then 2 1 1 ⎛1 1⎞ u = 1u = ⎜ + ⎟ u = u + u = u + u = 0. 7. (e) False. ⎢ −2 2 1 −1 0 3 7 ⎥ ⎣⎢ −2 3 4 −8 0 8 1⎦⎥ 3. ⎢ ⎥ 2k1 − k2 = c ⎣ 2 −1⎦ ⎡a ⎤ ⎡k ⎤ x = ⎢ 1 ⎥ .2 1. 5) = 4u + 3v is a linear combination of u and v. since for k ≠ 1. since for k ≠ 1 k (a + c + 1) ≠ ka + kc + 1. then −u = u and 1 1 u + u = 0. (a) This is a subspace of P3 . (d) This is a subspace of P3 . and C can be accomplished by reducing the matrix ⎡ 4 1 0 6 0 6 −1⎤ ⎢ 0 −1 2 −8 0 0 5⎥ ⎢ ⎥. ⎢ ⎥ ⎣ 2 −1 2 5 5 0 ⎦ ⎡ 1 0 2 4 0 0⎤ which reduces to ⎢0 1 2 3 0 0 ⎥ . kv is not in the set. 1) = (a1 + a2 . (b) (3. Consider k1u + k2 v = (a. so k(a. the zero vector would be 0 = 0 + 0x which does not have degree exactly 1. one of them would necessarily be the zero vector 0. B. 2) = 2u + 2v is a linear combination of u and v. 2k1 − k2 ) = (a. Since −u must exist and −u ≠ 0. Equating components gives the system k2 = a ⎡ 0 1⎤ −2k1 + 3k2 = b or Ax = b where A = ⎢ −2 3⎥ . (b) This is not a subspace of R ∞ . 0. (a) (2. 5) is not a linear combination of u and v. (c) This is a subspace of R3 . (c) This is a subspace of R ∞ . (d) This is a subspace of R ∞ . and b = ⎢ b ⎥ . 1. determining whether the given matrices are linear combinations of A. 5. 4. (b) This is a subspace of P3 . b. since (a1. 2 2 ⎝2 2⎠ Either way we get u = 0 which contradicts our assumption that we had exactly two elements. 0) = 0u + 0v is a linear combination of u and v. 9. c). 2 2 ⎝2 2⎠ 1 If u = u. b. The ⎢ ⎥ ⎣0 0 0 0 1 0 ⎦ results can be read from the reduced matrix. 1) + (a2 . 2. ⎢ ⎥ ⎣ k2 ⎦ ⎣c ⎦ The system can be solved simultaneously by ⎡ 0 1 2 3 0 0⎤ reducing the matrix ⎢ −2 3 2 1 4 0 ⎥ . 1. (a) This is a subspace of R ∞ . (c) This is not a subspace of P3 since k (a0 + a1 x + a2 x 2 + a3 x3 ) is not in the set for all noninteger values of k. Similar to the process in Exercise 7. 82 . and u + 0 = u. − 2k1 + 3k2 . (b) This is not a subspace of R3 . b. If 2 2 1 u = 0. (c) (0. 1. then 2 1 1 ⎛1 1⎞ u = 1u = ⎜ + ⎟ u = u + u = 0 + 0 = 0. 2) which is not in the set.Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra (d) False. if a vector space V had exactly two elements. 2.2 Exercise Set 4. (a) This is a subspace of R3 . c) is not in the set.

the given vectors do not span R3 . 0 1 0⎥ 0 0 1⎥⎦ ⎡ 6 −8 ⎤ (a) ⎢ = 1A + 2 B − 3C is a linear combination of A. 4k1 + 5k2 + 9k3 − k4 ) = (b1. and b3 so the given vectors ⎢ ⎥ ⎣ 2 3 1⎦ span R3 . v3 . B. b3 ) or 3k1 + 2k2 + 5k3 + k4 = b1 k1 − 3k2 − 2k3 + 4k4 = b2 . b2 .2 0 1 0 2 1 −3 0 0 0 1 0⎤ 0 2 0⎥ ⎥. and v 4 is not all of R3 . (b) ⎢ ⎣0 0⎥⎦ ⎡6 0⎤ = 1A + 2 B + 1C is a linear combination of A. then (2k1 . b2 . and C. and b3 so the given vectors ⎢ ⎥ ⎣ 3 2 8⎦ do not span R3 . the system is consistent for all values of b1 . then (2k1 + 4k2 + 8k3 . 2k1 + 3k2 + k3 ) = (b1 . and C.. k1 − 3k2 − 2k3 + 4k4 . (b) If k1v1 + k2 v 2 + k3 v3 = b. This restriction means that the 11 11 span of v1 . b3 ) be an arbitrary vector in R3 . b2 . v 2 . b2 . and C. b3 ) or 2k1 + k3 = b2 . 3k1 + 2k2 + 8k3 ) = (b1 . 2k1 = b1 (a) If k1v1 + k2 v 2 + k3 v3 = b. 3k1 + 2k2 + 8k3 = b3 ⎡ 2 4 8⎤ Since det ⎢ −1 1 −1⎥ = 0. 2k1 + k3 . b2 . b3 ) or 2k1 + 4k2 + 8k3 = b1 − k1 + k2 − k3 = b2 . and C. (d) ⎢ ⎣ 7 1⎥⎦ 11. i. then (3k1 + 2k2 + 5k3 + k4 . the system is not consistent for all values of b1 . ⎢0 0 0 0 − 17 b + 7 b + b ⎥ 3⎦ 11 1 11 2 ⎣ 17 7 b1 + b2 + b3 = 0 or 17b1 = 7b2 + 11b3 . b2 .e. Let b = (b1 . 83 . ⎣ −1 −8⎥⎦ ⎡0 0⎤ = 0 A + 0 B + 0C is a linear combination of A. B. 4k1 + 5k2 + 9k3 − k4 = b3 5 1 b1 ⎤ ⎡3 2 Reducing the matrix ⎢ 1 −3 −2 4 b2 ⎥ leads to ⎢ ⎥ ⎣ 4 5 9 −1 b3 ⎦ The system has a solution only if − ⎡1 −3 −2 4 b2 ⎤ ⎢ ⎥ 3 1 −1 11 b1 − 11 b2 ⎢0 1 1 ⎥. (c) ⎢ ⎣ 3 8⎥⎦ ⎡ −1 5 ⎤ is not a linear combination of A. − k1 + k2 − k3 . 2k1 + 3k2 + k3 = b3 ⎡2 0 0⎤ Since det ⎢ 2 0 1⎥ = −6 ≠ 0. B. (c) If k1v1 + k2 v 2 + k3 v3 + k4 v 4 = b. B.SSM: Elementary Linear Algebra ⎡1 ⎢0 The matrix reduces to ⎢ ⎢0 ⎢⎣0 0 1 0 0 Section 4.

2k1 + 4k3 + 2k4 = a2 ⎡ 1 3 5 −2 a0 ⎤ Reducing the matrix ⎢ −1 1 −1 −2 a1 ⎥ leads to ⎢ ⎥ ⎣ 2 0 4 2 a2 ⎦ a0 ⎡ 1 3 5 −2 ⎤ ⎢ ⎥ 1 1a − + 0 1 1 1 a ⎢ ⎥. ⎡ 1 −1 1 0 ⎤ (e) Since det(A) = 0. 84 ⎡ 1 0 3 0⎤ ⎢0 1 2 0⎥ .e. z = 0. ⎡ −1 1 1 0 ⎤ 15. − 39 b1 + 39 0 2 39 3 ⎥ 22 b − 17 b + 2 b 1 39 1 39 2 39 3 ⎥⎦ k1 . The solution ⎢ ⎥ ⎣0 0 0 0 ⎦ 1 3 set is x = − t . which is a line through the origin. (a) Since det(A) = 0. for any values of b1 ..e. they do not span P3 . then (k1 + 3k2 + 4k3 + 3k4 . The matrix reduces to ⎢ ⎥ ⎣ 2 −4 −5 0 ⎦ ⎡ 1 0 1 0⎤ 2 ⎢ ⎥ ⎢0 1 32 0 ⎥ . then (k1 + 3k2 + 5k3 − 2k4 ) + (−k1 + k2 − k3 − 2k4 ) x + (2k1 + 4k3 + 2k4 ) x 2 = a0 + a1 x + a2 x 2 or k1 + 3k2 + 5k3 − 2k4 = a0 − k1 + k2 − k3 − 2k4 = a1 . the system has only the trivial solution. 2k1 + 4k2 + 3k3 + 3k4 . Let p = a0 + a1 x + a2 x 2 be an arbitrary polynomial in P2 . 13. and p 4 is not all of P3 .Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra (d) If k1v1 + k2 v 2 + k3 v3 + k4 v 4 = b. which is a line through the origin. The matrix reduces to ⎢ ⎥ ⎣ 3 1 11 0⎦ is x = −3t. reduce the matrix ⎢ 2 −1 4 0⎥ . i. The given vectors span 0 1 39 16 39 17 39 R3 . the system has only the trivial solution.. k2 . If k1p1 + k2p 2 + k3p3 + k4 p 4 = p. values of ⎡1 3 4 3 b1 ⎤ The matrix ⎢ 2 4 3 3 b2 ⎥ reduces to ⎢ ⎥ ⎣6 1 1 1 b3 ⎦ 1 b − 1 b + 7 b ⎤ − 39 1 39 2 39 3 ⎥ 16 23 b − 5 b ⎥ . b2 . k3 . p 2 . y = t. This restriction means that the span of 2 2 p1 . (d) Since det(A) = 8 ≠ 0. z = t. 6k1 + k2 + k3 + k4 ) = (b1. 6k1 + k2 + k3 + k4 = b3 ⎡1 0 ⎢ ⎢0 1 ⎢ ⎢⎣0 0 Thus. 4 0 4 1 ⎢0 0 0 0 − 1 a + 3 a + a ⎥ 2⎦ ⎣ 2 0 2 1 1 3 The system has a solution only if − a0 + a1 + a2 = 0 or a0 = 3a1 + 2a2 . p3 . The matrix reduces to ⎢ ⎥ ⎣ −2 4 −6 0 ⎦ set is x = 2t. i. y = − t .. The solution set ⎢ ⎥ ⎣0 0 0 0⎦ . y = −2t. b2 . ⎡ 1 −2 0 0 ⎤ ⎢0 0 1 0 ⎥ . and b3 . 2 2 3 0⎤ ⎡ 1 −2 (b) Since det(A) = 0. and k4 can be found. b3 ) or k1 + 3k2 + 4k3 + 3k4 = b1 2k1 + 4k2 + 3k3 + 3k4 = b2 . z = t which is a line through the origin.e. reduce the matrix ⎢ 3 −1 0 0 ⎥ . reduce the matrix ⎢ −3 6 9 0 ⎥ . the origin. The solution ⎢ ⎥ ⎣0 0 0 0 ⎦ (c) Since det(A) = −1 ≠ 0. the origin. i.

Let f = f(x) and g = g(x) be elements of the set.2 (a) True. Section 4. but is not a subspace of R 2 . u1 = (−1. (h) False. if b ≠ 0. the span of x − 1.3.e. y) y = ± x} . Then b f + g = ∫ ( f ( x) + g ( x))dx a b b a a = ∫ f ( x)dx + ∫ g ( x)dx =0 b b a a and f + g is in the set. by Theorem 4. u 2 is a scalar multiple of u1 . (j) True. 0). v 2 }.3 1. − 1).2. not all of P3 . the system is not homogeneous.3 ⎡ 1 −3 1 0 ⎤ (f) Since det(A) = 0. (d) False. Then {v1 . v 2 } and {u1. True/False 4. y) y = x} and W2 = {( x. b].. but v1 + v 2 = (2. i. which is a plane through the origin. R 2 is not a subset of R3 . 85 . for instance. the set W in Example 4 contains the origin (zero vector). − 1) is in W2 . Let k be any scalar. consider the subspaces W1 = {( x. ⎡ 1 −3 1 0 ⎤ ⎢0 0 0 0 ⎥ ..2. this is part of the definition of a subspace. p 2 is a scalar multiple of p1 .e. 0).SSM: Elementary Linear Algebra Section 4. let v1 = (1. and ( x − 1)3 will only contain polynomials for which p(1) = 0. reduce the matrix ⎢ 2 −6 2 0 ⎥ . (i) False. this is by the definition of the span of a set of vectors.3. (f) True. 1) is in W1 and v 2 = (1. i.e. and u 2 = (0. (e) False. i.. (a) u 2 = −5u1 . The matrix reduces to ⎢ ⎥ ⎣ 3 −9 3 0 ⎦ is x − 3y + z = 0. Then kf = ∫ kf ( x)dx = k ∫ f ( x)dx = k ⋅ 0 = 0 so kf is in the set. the set is linearly dependent by Theorem 4. Thus the set is a subspace of C[a. i. 1). since a vector space is a subset of itself. The solution set ⎢ ⎥ ⎣0 0 0 0 ⎦ 17.e. the sum of two upper triangular matrices is upper triangular and the scalar multiple of an upper triangular matrix is upper triangular. 0) is not in W1 ∪ W2 = {( x. (b) True. (c) p 2 = 2p1 . u 2 } both span R 2 but {u1. then the solution set does not contain the zero vector. v 2 = (0. y) y = − x} in R 2 . u 2 } ≠ {v1 . ( x − 1)2 . (k) False.3 Exercise Set 4. (g) True. (b) Since there are 3 vectors and 3 > 2. v1 = (1. (c) False..

1) + k3 (0. 1.e. 0) generates the homogeneous system. 6) + k2 (0. 3. (c) The equation k1 (0. the system has only the trivial solution and the vectors are linearly ⎢ −3 0 −2 4 ⎥ ⎣⎢ −6 −6 −2 −4 ⎦⎥ independent. 0) generates the homogeneous system 3k1 − 2 k4 = 0 2k2 − 2k3 + k4 = 0 . − 8. 0. − 6) + k3 (0. 3k1 + k2 + 2k3 + k4 = 0 8k1 + 5k2 − k3 + 4k4 = 0 7k1 + 3k2 + 2k3 =0 −3k1 − k2 + 6k3 + 3k4 = 0 ⎡ 3 1 2 ⎢ 8 5 −1 Since det ⎢ ⎢ 7 3 2 ⎢⎣ −3 −1 6 independent. 2. − 2.. 0. 0. 2. 0. (a) The equation k1 (3. − 4. 4. 2) + k2 (3. 0. 2. 0) generates the homogeneous system 3k2 + k3 = 0 3k2 + k3 = 0 2k1 =0 2k1 − k3 = 0 The third equation gives k1 = 0. 3. 0) + k3 (1. 3.Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra (d) B = −A. 0. − 4) = (0. − 2) + k4 (0. 3. 1. 3) = (0. B is a scalar multiple of A. − 3) + k2 (1. Since the system has only the trivial solution. − 1) + k3 (2. 0 0 −2 ⎤ 2 −2 1⎥ ⎥ = 36 ≠ 0. 4. k3 = 0 gives k2 = 0 in the first two equations. − 1) = (0. the vectors are linearly independent. 7. − 3. the system has only the trivial solution and the vectors are linearly 0⎥ 3⎥⎦ (b) The equation k1 (0. − 3. 0. 0. 0) + k4 (−2. 0. 0. 1) = (0. 6) + k4 (1. 0) generates the homogeneous system −2 k 2 =0 3k1 − 4 k3 − 8 k 4 = 0 . − 2. 0. − 2. i. the system has only the trivial solution and the vectors are linearly 3 −2 2⎥ 1 0 1⎥⎦ 86 . 2. 0. which gives k3 = 0 in the fourth equation. 0. 8. 0. − 1. 0. − 6) + k2 (−2. −3k1 − 2k3 + 4k4 = 0 −6k1 − 6k2 − 2k3 − 4k4 = 0 ⎡ 0 −2 0 0 ⎤ ⎢ 3 0 −4 −8 ⎥ Since det ⎢ ⎥ = 480 ≠ 0. 1⎤ 4⎥ ⎥ = 128 ≠ 0. 3. (d) The equation k1 (3. 5. −3k1 + 3k2 − 2k3 + 2k4 = 0 6k1 + k2 + k4 = 0 ⎡ 3 ⎢ 0 Since det ⎢ ⎢ −3 ⎢⎣ 6 independent.

SSM: Elementary Linear Algebra

Section 4.3

5. If the vectors lie in a plane, then they are linearly
dependent.

3
7
3
t = − : v 2 = v1 + v3
2
2
2
7
2
t = 1: v3 = − v1 + v 2
3
3

(a) The equation k1v1 + k2 v 2 + k3 v3 = 0
generates the homogeneous system
2k1 + 6k2 + 2k3 = 0
−2k1 + k2
= 0.
4 k 2 − 4 k3 = 0

9. The equation k1v1 + k2 v 2 + k3 v3 = 0 generates
the homogeneous system
1
1
λk1 − k2 − k3 = 0
2
2
1
1
− k1 + λk2 − k3 = 0 .
2
2
1
1
− k1 − k2 + λk3 = 0
2
2
⎡ λ − 1 − 1⎤
2
2⎥

1
det ⎢ − 12
λ − 12 ⎥ = (4λ3 − 3λ − 1)
⎢ 1
⎥ 4
1
λ ⎥⎦
⎢⎣ − 2 − 2
1
= (λ − 1)(2λ + 1)2
4
1
For λ = 1 and λ = − , the determinant is zero
2
and the vectors form a linearly dependent set.

⎡ 2 6 2⎤
Since det ⎢ −2 1 0⎥ = −72 ≠ 0, the

⎣ 0 4 −4 ⎦
vectors are linearly independent, so they do
not lie in a plane.
(b) The equation k1v1 + k2 v 2 + k3 v3 = 0
generates the homogeneous system
−6k1 + 3k2 + 4k3 = 0
7k1 + 2k2 − k3 = 0 .
2k1 + 4k2 + 2k3 = 0

⎡ −6 3 4 ⎤
Since det ⎢ 7 2 −1⎥ = 0, the vectors are

⎣ 2 4 2⎦
linearly dependent, so they do lie in a plane.

11. Let {v a , vb , … , v n } be a (nonempty) subset of
S. If this set were linearly dependent, then there
would be a nonzero solution (ka , kb , … , kn ) to

7. (a) The equation k1v1 + k2 v 2 + k3 v3 = 0
generates the homogeneous system
6 k 2 + 4 k3 = 0
3k1
− 7 k3 = 0
.
k1 + 5k2 + k3 = 0
− k1 + k2 + 3k3 = 0

⎡ 0
⎢ 3
The matrix ⎢
⎢ 1
⎣⎢ −1

ka v a + kb vb + + kn v n = 0. This can be
expanded to a nonzero solution of
k1v1 + k2 v 2 + + kr v r = 0 by taking all other
coefficients as 0. This contradicts the linear
independence of S, so the subset must be linearly
independent.

6
4 0⎤
0 −7 0 ⎥
⎥ reduces to
5
1 0⎥
1 3 0 ⎥⎦

13. If S is linearly dependent, then there is a nonzero
solution (k1 , k2 , …, kr ) to

⎡ 1 0 − 7 0⎤
3

2 0⎥
⎢0 1
. Since the system has a
3

0
0
0
0

0 0 ⎦⎥
⎣⎢0 0
nontrivial solution, the vectors are linearly
dependent.
(b) The solution of the system is k1 =

k1v1 + k2 v 2 +

+ kr v r = 0. Thus

(k1 , k2 , … , kr , 0, 0, … , 0) is a nonzero solution
to
k1v1 + k2 v 2 + + kr v r + kr +1 v r +1 + + kn v n
=0
so the set {v1 , v 2 , … , v r , v r +1 , … , v n } is
linearly dependent.

7
t,
3

15. If {v1, v 2 , v3} were linearly dependent, there
would be a nonzero solution to
k1v1 + k2 v 2 + k3 v3 = 0 with k3 ≠ 0 (since v1

2
k 2 = − t , k3 = t .
3
3
2
3
t = : v1 = v 2 − v3
7
7
7

and v 2 are linearly independent). Solving for

87

Chapter 4: General Vector Spaces

SSM: Elementary Linear Algebra

determinant was simplified by adding the first
row to the third row.

k
k
v3 gives v3 = − 1 v1 − 2 v 2 which contradicts
k3
k3
that v3 is not in span {v1 , v 2 }. Thus,

True/False 4.3

{v1, v 2 , v3} is linearly independent.

(a) False; the set {0} is linearly dependent.

19. (a) If v1 , v 2 , and v3 are placed with their
initial points at the origin, they will not lie in
the same plane, so they are linearly
independent.

(b) True
(c) False; the set {v1 , v 2 } in R 2 where v1 = (0, 1)
and v 2 = (0, 2) is linearly dependent.

(b) If v1 , v 2 , and v3 are placed with their
initial points at the origin, they will lie in the
same plane, so they are linearly dependent.
Alternatively, note that v1 + v 2 = v3 by

(d) True; suppose {kv1, kv 2 , kv3} were linearly
dependent, then there would be a nontrivial
solution to k1 (kv1 ) + k2 (kv 2 ) + k3 (kv3 ) = 0
which would give a nontrivial solution to
k1v1 + k2 v 2 + k3 v3 = 0.

placing v1 along the z-axis with its terminal
point at the origin.
21. The Wronskian is
x
cos x
W ( x) =
= − x sin x − cos x.
1 − sin x

(e) True; consider the sets {v1 , v 2 }, {v1 , v 2 , v3},

..., {v1 , v 2 , … , v n }. If {v1 , v 2 } is linearly
dependent, then v 2 = kv1 for some k and the

W(0) = −cos 0 = −1, so the functions are linearly
independent.

condition is met. If v 2 ≠ kv1 then there is some
2 < k ≤ n for which the set {v1 , v 2 , …, v k −1} is
linearly independent but {v1, v 2 , … , v k } is
linearly dependent. This means that there is a
nontrivial solution of
a1 v1 + a2 v 2 + + ak −1 v k −1 + ak v k = 0 with

1 x ex
23. (a) W ( x) = 0

x
1 ex = e

0 0 ex
Since W ( x) = e x is nonzero for all x, the
vectors are linearly independent.

ak ≠ 0 and the equation can be solved to give
v k as a linear combination of v1 , v 2 , ...,
v k −1 .

1 x x2
(b) W ( x) = 0 1 2 x = 2
0 0
2
Since W(x) = 2 is nonzero for all x, the
vectors are linearly independent.

(f) False; the set is
⎧ ⎡1 1 ⎤ ⎡1 0 ⎤ ⎡1 0⎤ ⎡ 0 1 ⎤ ⎡ 0 1⎤
⎨ ⎢ 0 0 ⎥ , ⎢1 0 ⎥ , ⎢0 1 ⎥ , ⎢1 0 ⎥ , ⎢ 0 1⎥ ,
⎦ ⎣
⎦ ⎣
⎦ ⎣
⎦ ⎣

⎩⎣

⎡0 0⎤ ⎫
⎢⎣1 1 ⎥⎦ ⎬ and

⎡1 1 ⎤ ⎡0 0 ⎤
⎡1 0⎤
⎡0 1⎤
1⎢
+ 1⎢
+ (−1) ⎢
+ (−1) ⎢

0
0
1
1
1
0

⎦ ⎣

⎣0 1⎥⎦
⎡0 0⎤
=⎢
⎣ 0 0 ⎥⎦
so the set is linearly dependent.

x cos x
sin x
cos x
25. W ( x) = cos x − sin x
cos x − x sin x
− sin x − cos x − x cos x − 2 sin x
x cos x
sin x cos x
= cos x − sin x cos x − x sin x
−2 sin x
0
0
sin x cos x
= −2 sin x
cos x − sin x
2

(g) True; the first polynomial has a nonzero constant
term, so it cannot be expressed as a linear
combination of the others, and since the two
others are not scalar multiples of one another, the
three polynomials are linearly independent.

2

= −2 sin x(− sin x − cos x)
= 2 sin x
Since 2sin x is not identically zero, the vectors
are linearly independent, hence span a threedimensional subspace of F(−∞, ∞). Note that the
88

SSM: Elementary Linear Algebra

Section 4.4

(h) False; the functions f1 and f 2 are linearly dependent if there are scalars k1 and k2 such that
k1 f1 ( x) + k2 f 2 ( x) = 0 for all real numbers x.
Section 4.4
Exercise Set 4.4
1. (a) The set S = {u1, u 2 , u3} is not linearly independent; u3 = 2u1 + u 2 .
(b) The set S = {u1, u 2 } spans a plane in R3 , not all of R3 .
(c) The set S = {p1, p 2 } does not span P2 ; x 2 is not a linear combination of p1 and p 2 .
(d) The set S = {A, B, C, D, E} is not linearly independent; E can be written as linear combination of A, B, C, and
D.

⎡ 1 2 3⎤
3. (a) As in Example 3, it is sufficient to consider det ⎢ 0 2 3⎥ = 6 ≠ 0.

⎣ 0 0 3⎦
Since this determinant is nonzero, the set of vectors is a basis for R3 .

⎡ 3 2 1⎤
(b) It is sufficient to consider det ⎢ 1 5 4 ⎥ = 26 ≠ 0. Since this determinant is nonzero, the set of vectors is a

⎣ −4 6 8⎦
basis for R3 .
0⎤
⎡ 2 4
(c) It is sufficient to consider det ⎢ −3 1 −7 ⎥ = 0. Since this determinant is zero, the set of vectors is not

1⎦
⎣ 1 1
linearly independent.

⎡ 1 2 −1⎤
(d) It is sufficient to consider det ⎢ 6 4 2⎥ = 0. Since this determinant is zero, the set of vectors is not

⎣ 4 −1 5⎦
linearly independent.
⎡3 6 ⎤
⎡ 0 −1⎤
⎡ 0 −8 ⎤
⎡ 1 0 ⎤ ⎡ 0 0⎤
5. The equations c1 ⎢
+ c2 ⎢
+ c3 ⎢
+ c4 ⎢
=
and

3
6
1
0
12
4

⎣ −1 2 ⎥⎦ ⎢⎣ 0 0⎥⎦
⎡3 6 ⎤
⎡ 0 −1⎤
⎡ 0 −8 ⎤
⎡ 1 0⎤ ⎡ a b ⎤
c1 ⎢
+ c2 ⎢
+ c3 ⎢
+ c4 ⎢
=
generate the systems

3
6
1
0
12
4

⎣ −1 2 ⎥⎦ ⎢⎣ c d ⎥⎦
3c1
+ c4 = 0
3c1
+ c4 = a
6c1 − c2 − 8c3
=0
6c1 − c2 − 8c3
=b
and
3c1 − c2 − 12c3 − c4 = 0
3c1 − c2 − 12c3 − c4 = c
−6c1
− 4c3 + 2c4 = 0
−6c1
− 4c3 + 2c4 = d

89

Chapter 4: General Vector Spaces

SSM: Elementary Linear Algebra
c1 − 4c2 + 7c3 = 5
2c1 + 5c2 − 8c3 = −12 . The matrix
3c1 + 6c2 + 9c3 = 3

which have the same coefficient matrix. Since
0 1⎤
⎡ 3 0
⎢ 6 −1 −8 0 ⎥
det ⎢
⎥ = 48 ≠ 0 the matrices
⎢ 3 −1 −12 −1⎥
⎢⎣ −6 0 −4 2 ⎥⎦
are linearly independent and also span M 22 , so
they are a basis.

5⎤
⎡ 1 −4 7
⎢2
5 −8 −12⎥ reduces to

3⎦
⎣3 6 9
⎡ 1 0 0 −2 ⎤
⎢0 1 0 0 ⎥ .

⎣0 0 1 1⎦

7. (a) Since S is the standard basis for R 2 ,
(w) S = (3, − 7).

The solution of the system is c1 = −2,
c2 = 0, c3 = 1 and ( v) S = (−2, 0, 1).

(b) Solve c1u1 + c2 u 2 = w, or in terms of

components, (2c1 + 3c2 , − 4c1 + 8c2 ) = (1, 1).
2c1 + 3c2 = 1
Equating components gives
.
−4c1 + 8c2 = 1

11. Solve c1 A1 + c2 A2 + c3 A3 + c4 A4 = A. By
inspection, c3 = −1 and c4 = 3. Thus it remains

−c1 + c2 = 2
which can be
c1 + c2 = 0

⎡ 2 3 1⎤
The matrix ⎢
reduces to
⎣ −4 8 1⎥⎦

to solve the system

⎡1 0 5 ⎤
3⎞
28 ⎥ . Thus, ( w) = ⎛ 5

⎜ , ⎟.
S
3
⎝ 28 14 ⎠
⎢⎣0 1 14 ⎥⎦

from which c1 = −1. Thus, ( A) S = (−1, 1, − 1, 3).

solved by adding the equations to get c2 = 1,

⎡0 0 ⎤
,
13. Solving c1 A1 + c2 A2 + c3 A3 + c4 A4 = ⎢
⎣0 0 ⎥⎦
⎡a b ⎤
c1 A1 + c2 A2 + c3 A3 + c4 A4 = ⎢
, and
⎣ c d ⎥⎦
c1 A1 + c2 A2 + c3 A3 + c4 A4 = A gives the systems

(c) Solve c1u1 + c2 u 2 = w, or in terms of

components, (c1 , c1 + 2c2 ) = (a, b). It is
clear that c1 = a and solving a + 2c2 = b
for c2 gives c2 =

b−a
. Thus,
2

⎛ b−a⎞
( w ) S = ⎜ a,
⎟.

2 ⎠
9. (a) Solve c1v1 + c2 v 2 + c3 v3 = v, or in terms of
components,
(c1 + 2c2 + 3c3 , 2c2 + 3c3 , 3c3 ) = (2, − 1, 3).
Equating components gives
c1 + 2c2 + 3c3 = 2
2c2 + 3c3 = −1 which can easily be
3c3 = 3

c1
c1 + c2
c1 + c2 + c3
c1 + c2 + c3 + c4

=0
=0
,
=0
=0

c1
c1 + c2
c1 + c2 + c3
c1 + c2 + c3 + c4

=1
=0
.
=1
=0

c1
c1 + c2
c1 + c2 + c3
c1 + c2 + c3 + c4

⎡1
⎢1
It is easy to see that det ⎢
⎢1
⎣⎢1

solved by back-substitution to get c1 = 3,
c2 = −2, c3 = 1. Thus, ( v) S = (3, − 2, 1).

0
1
1
1

0
0
1
1

=a
=b
, and
=c
=d

0⎤
0⎥
⎥ = 1 ≠ 0, so
0⎥
1⎦⎥

{A1, A2 , A3 , A4 } spans M 22 . The third system
is easy to solve by inspection to get c1 = 1,

(b) Solve c1v1 + c2 v 2 + c3 v3 = v, or in terms of
components,
(c1 − 4c2 + 7c3 , 2c1 + 5c2 − 8c3 , 3c1 + 6c2 + 9c3 )

c2 = −1, c3 = 1, c4 = −1, thus,
A = A1 − A2 + A3 − A4 .

= (5, −12, 3).
Equating components gives

15. Solving the systems c1p1 + c2p 2 + c3p3 = 0,
c1p1 + c2p 2 + c3p3 = a + bx + cx 2 , and
c1p1 + c2 p 2 + c3p3 = p gives the systems
90

SSM: Elementary Linear Algebra

c1
c1 + c2
c1 + c2 + c3
c1
c1 + c2
c1 + c2 + c3

Section 4.5

= 0 c1
=a
= 0 , c1 + c2
= b , and
= 0 c1 + c2 + c3 = c
=7
= −1 .
=2

True/False 4.4
(a) False; {v1 , … , v n } must also be linearly
independent to be a basis.
(b) False; the span of the set must also be V for the
set to be a basis.

⎡1 0 0 ⎤
It is easy to see that det ⎢1 1 0 ⎥ = 1 ≠ 0, so

⎣1 1 1 ⎦
{p1, p 2 , p3} spans P2 . The third system is easy

(c) True; a basis must span the vector space.
(d) True; the standard basis is used for the
coordinate vectors in R n .

to solve by inspection to get c1 = 7, c2 = −8,

(e) False; the set {1 + x + x 2 + x3 + x 4 ,

c3 = 3, thus, p = 7p1 − 8p 2 + 3p3 .

x + x 2 + x3 + x 4 , x 2 + x3 + x 4 , x3 + x 4 , x 4 } is
a basis for P4 .

17. From the diagram, j = u 2 and
u1 = (cos 30°)i + (sin 30°) j =

3
1
i + j.
2
2

Section 4.5

3
1
i + u 2 for i in terms of u1
2
2
2
1
and u 2 gives i =
u1 −
u2 .
3
3

Exercise Set 4.5

Solving u1 =

(a)

(

3, 1) is

⎡ 1 1 −1 0 ⎤
1. The matrix ⎢ −2 −1 2 0 ⎥ reduces to

⎣ −1 0 1 0 ⎦
⎡ 1 0 −1 0 ⎤
⎢0 1 0 0 ⎥ . The solution of the system is

⎣0 0 0 0 ⎦
x1 = t , x2 = 0, x3 = t , which can be written as
( x1 , x2 , x3 ) = (t , 0, t ) or ( x1 , x2 , x3 ) = t (1, 0, 1).
Thus, the solution space has dimension 1 and a
basis is (1, 0, 1).

3i + j.

1
⎛ 2

u1 −
u 2 ⎟ + u2
3i + j = 3 ⎜
3 ⎠
⎝ 3
= 2u1 − u 2 + u 2
= 2u1
The x ′y ′-coordinates are (2, 0).
(b) (1, 0) is i which has x ′y ′-coordinates

⎡ 1 −4 3 −1 0 ⎤
3. The matrix ⎢
reduces to
⎣ 2 −8 6 −2 0⎥⎦
⎡ 1 −4 3 −1 0 ⎤
⎢⎣0 0 0 0 0 ⎥⎦ . The solution of the system
is x1 = 4r − 3s + t , x2 = r , x3 = s, x4 = t ,
which can be written as
( x1 , x2 , x3 , x4 ) = (4r − 3s + t , r , s, t ) or

1 ⎞
⎛ 2
,−

⎟.
3⎠
⎝ 3
(c) Since (0, 1) = j = u 2 , the x ′y ′-coordinates
are (0, 1).
(d) (a, b) is ai + bj.
1
⎛ 2

u1 −
u 2 ⎟ + bu 2
ai + bj = a ⎜
3 ⎠
⎝ 3
a ⎞
2

=
au1 + ⎜ b −
⎟ u2
3⎠
3

a ⎞
⎛ 2
a, b −
The x ′y ′-coordinates are ⎜
⎟.
3⎠
⎝ 3

( x1 , x2 , x3 , x4 )
= r (4, 1, 0, 0) + s(−3, 0, 1, 0) + t (1, 0, 0, 1).
Thus, the solution space has dimension 3 and a
basis is (4, 1, 0, 0), (−3, 0, 1, 0), (1, 0, 0, 1).

91

u 2 . y. 0) + s(0. A basis is (2. so the system has only the trivial ⎢ ⎥ ⎣0 0 1 0 ⎦ solution. u 2 . 0. 1. u3} is linearly independent so it is one possible basis. 4). u1. and (0. any two of the vectors (0. 4t) = t(2. The matrix ⎢ 1 0 5 0 ⎥ reduces to ⎢ ⎥ ⎣0 1 1 0⎦ ⎡ 1 0 0 0⎤ ⎢0 1 0 0 ⎥ . (d) The vectors can be parametrized as a = r. 1). 0). (0. 13. Thus. Then ( p1 + p2 )(1) = p1 (1) + p2 (1) = 0 so p1 + p 2 is in W. 4). r + s. 1. The equation c1v1 + c2 v 2 + k1u1 + k2 u 2 + k3u3 + k4 u 4 = 0 leads to the system c1 − 3c2 + k1 =0 −4c1 + 8c2 + k2 =0 . z) = (2t. u 4} clearly spans R 4 . u3 = (0. 0 ⎟ + s ⎜ − . 1. s) = r (1. ⎝3 ⎠ ⎝ 3 ⎠ ⎛2 ⎞ ⎛ 5 ⎞ A basis is ⎜ . 0. 0. 0). since a basis is the set {A1. dimension 2 92 . 0 ⎟ . There are n(n + 1) entries on or n + (n − 1) + + 1 = 2 above the main diagonal. ⎡ 1 −3 ⎢ −4 8 The matrix ⎢ − 2 4 ⎢ ⎢⎣ −3 6 ⎡ 1 0 −2 0 0 ⎢ 0 1 −1 0 0 ⎢ to ⎢ 0 0 0 1 0 ⎢ ⎢0 0 0 0 1 ⎣ (c) In vector form. 1. (a) The dimension is n. 1) can be used. r . 0. (c) A basis for W is {−1 + x. which can be written in vector form as ( x. y = r. 1). y. v 2 . so the space has n(n + 1) . 0. c) = (r . 1). Similarly. −1. 0. 1. A basis is (1. which has dimension 0 and no basis. 0.Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra n(n + 1) entries on 2 or above the main diagonal. 0). 7. 0. 0. there are 11. 1). −1. 1. − 43 0⎥ ⎥ 2 0⎥ 3 ⎦ From the reduced matrix it is clear that u1 is a linear combination of v1 and v 2 so it will not be in the basis. Then the set {v1 . b = r + s. r . An } where the only nonzero entry in Ai is aii = 1. so the space has n(n + 1) . 0. 2c1 − 4c2 + k3 =0 −3c1 + 6c2 + k4 = 0 (b) Solving the equation for x gives x = y. z = s. 1. A basis is (1. 0). 1). Let u1 = (1. (b) In a symmetric matrix. 0). c = s or (a. (0. and that {v1. 0. 0). y. the elements above the main diagonal determine the elements below the main diagonal. z = s. 0. s ⎟ 3 ⎝3 ⎠ ⎛2 ⎞ ⎛ 5 ⎞ = r ⎜ . −t. 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 0⎤ 0⎥ ⎥ reduces 0⎥ 0 ⎥⎦ −1 0 ⎤ − 13 0⎥ ⎥ . 1. 1. z ) = ⎜ r − s. A2 . u3 . 1. which can be 3 3 written in vector form as 5 ⎛2 ⎞ ( x. 0). 1. … . s) = r (1. (c) As in part (b). (a) Solving the equation for x gives 2 5 x = y − z. (0. so parametric equations are 3 3 2 5 x = r − s. the line is (x. 0) + s(0. − x + x 2 } so the dimension is 2. u 4 = (0. 1⎟ . ⎜ − . u 2 = (0. y = r. 1. u 4 and either u 2 or u3 will produce a linearly independent set. 1⎟ . (a) Let p1 = p1 ( x) and p 2 = p2 ( x) be elements of W. so the entries on and above the main diagonal determine all the entries. b. dimension 2 ⎡2 1 3 0⎤ 5. ⎝3 ⎠ ⎝ 3 ⎠ (b). 0. 9. Thus. z ) = (r . (kp1 )(1) = k ( p1 (1)) = k ⋅ 0 = 0 so kp1 is in W. 0. so parametric equations are x = r. v 2 . W is a subspace of P2 .

then linear independence is sufficient for 5 vectors to be a basis. ⎢ ⎥ ⎣ 1⎦ ⎡0 1 ⎤ ⎡ 0 1⎤ ⎫ ⎢⎣1 0 ⎥⎦ . any basis of R17 will contain 17 linearly independent vectors.6(c). Thus. 3. 3c1 − 3c2 + cc3 = 0 (i) True. c). 1 0⎤ ⎡1 0 ⎢0 1 − 1 1 ⎥ . 0 1 2 7 1 7 3⎤ ⎡2 − 28 ⎥ . then by Theorem 4. at least 17 vectors are required. adding any 5 5 vector v3 = (a. while if it contains more than n vectors. An } contains n 2 + 1 matrices. since P2 has dimension three. (h) True. it can be reduced to a basis.6 1. (j) False. c) with 9a − 3b − 5c ≠ 0 will Section 4. the set of n × n matrices has dimension 2 n 2 . …. it is a basis. ⎡ 1 0 a 0⎤ The matrix ⎢ −2 5 b 0⎥ reduces to ⎢ ⎥ ⎣ 3 −3 c 0⎦ 0⎤ a ⎡1 0 ⎢0 1 2a+ 1b 0 ⎥ and the 5 5 ⎢ ⎥ ⎢0 0 − 9 a + 3 b + c 0 ⎥ 5 5 ⎣ ⎦ homogeneous system has only the trivial 9 3 solution if − a + b + c ≠ 0. since the dimension for R5 is 5. to span R17 . Thus PB→ S = ⎢ 7 1⎥ ⎢⎣ 17 14 ⎦ ⎡2 [ w ]S = PB→ S [ w ]B = ⎢ 7 ⎢⎣ 17 (b) True. (a) Since {u1.5 (a) True. (c) False. The equation c1v1 + c2 v 2 + c3 v3 = 0 leads to the system c1 + ac3 = 0 −2c1 + 5c2 + bc3 = 0 . Let v3 = (a. 2 2⎦ ⎣ a ⎡ ⎤ =⎢ 1 ⎥. 3 1 0⎤ reduces to 8 0 1⎥⎦ . [ w ]s = ⎢ ⎥ . by definition. True/False 4. the only three-dimensional subspace of P2 is P2 itself. A. while if it contains fewer than n vectors. if the set contains n vectors. 1) or [p]S = ⎢ −3⎥ . (f) True.6 15. if the set contains n vectors. it can be enlarged to a basis.SSM: Elementary Linear Algebra Section 4. ⎢0 −1⎥ . 1 ⎥ ⎣1⎦ ⎢ 3 ⎥ 14 ⎦ ⎣ 14 ⎦ ⎡ 1 0⎤ Thus PB→ S = ⎢ 1 1 ⎥ and ⎣− 2 2 ⎦ [ w ]S = PB → S [ w ]B ⎡ 1 0⎤ ⎡ a ⎤ = ⎢ 1 1⎥ ⎢ ⎥ ⎣− 2 2 ⎦ ⎣b ⎦ (e) True. u 2 } is the standard basis B for ⎡ 3⎤ R 2 .6 Exercise Set 4. ⎦ ⎣ ⎦ ⎩⎣ basis for M 22 . 1 ⎣− 2 a + 2 b ⎦ (g) True. it is a basis. − 3. while the set {I . ⎡ 4⎤ (p) S = (4. since the dimension of R5 is 5. b. ⎢⎣ −1 0 ⎥⎦ ⎬ form a ⎭ 93 3⎤ − 28 ⎥ 1⎥ 14 ⎦ 3⎤ 5 − 28 1 ⎡ ⎤ ⎥ ⎡⎢ ⎤⎥ = ⎢ 28 ⎥ .5. then spanning is sufficient for 5 vectors to be a basis. (a) Since S is the standard basis B for P2 . the set ⎧ ⎡1 0 ⎤ ⎡ 1 0 ⎤ ⎨ ⎢ 0 1 ⎥ . b. ⎡1 0 1 0 ⎤ (c) ⎢ reduces to ⎣1 2 0 1⎥⎦ (d) True. A2 . ⎣ −7 ⎦ ⎡ 2 (b) ⎢ ⎣ −4 ⎡1 ⎢ ⎢⎣0 and 3 create a basis for R .

12).Chapter 4: General Vector Spaces ⎡1 (b) ⎢ 1 ⎢ ⎣0 ⎡1 ⎢ ⎢0 ⎢ ⎣⎢0 SSM: Elementary Linear Algebra ⎡ −1 ⎢ 1 [ B]E = PS → E [ B ]S = ⎢ ⎢ 0 ⎢⎣ 0 ⎡ 15⎤ ⎢ −1⎥ =⎢ ⎥ ⎢ 6⎥ ⎣⎢ 3⎦⎥ 1 0 1 0 0⎤ 0 1 0 1 0 ⎥ reduces to ⎥ 1 1 0 0 1⎦ 1 1 − 1⎤ 0 0 2 2 2⎥ 1 −1 1 ⎥ . 94 . ⎡ 1 2 3⎤ 5.2. [ w ]B = PE → B [ w ]E 2⎤ ⎡1 10 5 ⎥ ⎡ 3⎤ ⎢ = 1 ⎢⎣ 5 − 51 ⎥⎦ ⎢⎣ −5⎥⎦ ⎡ − 17 ⎤ = ⎢ 10 ⎥ ⎢ 8⎥ ⎣ 5⎦ [ w ]B′ = PB→ B′ [ w ]B ⎡ 0 − 5 ⎤ ⎡ − 17 ⎤ 2 ⎥ ⎢ 10 ⎥ =⎢ ⎢⎣ −2 − 13 ⎥⎢ 8 ⎥ 2 ⎦⎣ 5 ⎦ ⎡ −4 ⎤ =⎢ ⎥ ⎣ −7 ⎦ (b) The basis in Exercise 3(a) is the standard ⎡3⎤ basis B for P2 . Thus. 1 0 2 2 2 1 1⎥ 0 1 − 12 2 2 ⎦⎥ 1 − 1⎤ ⎡ 1 2 2⎥ ⎢ 2 1 ⎥ and PB→ S = ⎢ 12 − 12 2 ⎢ 1 1 1⎥ ⎢⎣ − 2 2 2 ⎥⎦ [p]S = PB → S [p]B 1 − 1⎤ ⎡ 1 2 2 ⎥ ⎡ 2⎤ ⎢ 2 1 ⎥ ⎢ −1⎥ = ⎢ 12 − 12 2 ⎢ ⎥ ⎢ 1 1 1 ⎥ ⎣ 1⎦ − 2 2 ⎦⎥ ⎣⎢ 2 ⎡ 0⎤ = ⎢ 2⎥ ⎢ ⎥ ⎣ −1⎦ 0 0 1 0 0 ⎤ ⎡ −8 ⎤ 0⎥ ⎢ 7⎥ ⎥⎢ ⎥ 0⎥ ⎢ 6⎥ 1⎥⎦ ⎢⎣ 3⎥⎦ ⎡15 −1⎤ and B = ⎢ . ⎡ −1 1 0 ⎢ 1 1 0 PS → E = ⎢ ⎢ 0 0 1 ⎣⎢ 0 0 0 1 1 0 0 0⎤ 0⎥ ⎥ where E is the 0⎥ 1⎦⎥ standard basis for M 22 . (a) ⎢ reduces to ⎣ 2 −1 3 −1⎥⎦ ⎡ 1 0 13 10 ⎢ ⎢⎣0 1 − 25 ⎡ 13 − 12 ⎤ ⎥ so PB′→ B = ⎢ 10 ⎢⎣ − 52 0 ⎥⎦ − 12 ⎤ ⎥. Thus. ⎣ 6 3⎥⎦ ⎡ 2 4 1 −1⎤ 7. 0⎥⎦ ⎡ 1 −1 2 4 ⎤ (b) ⎢ reduces to ⎣3 −1 2 −1⎥⎦ ⎡1 0 0 − 5⎤ ⎡ 0 − 5⎤ 2 ⎥ so P 2⎥. (c) From Theorem 4. PS → B = ⎢0 2 3⎥ ⎢ ⎥ ⎣ 0 0 3⎦ ⎡2 4 1 0⎤ (c) ⎢ reduces to ⎣ 2 −1 0 1⎥⎦ where B is the standard basis for R3 .2. so [q]B = [q]S = ⎢ 0 ⎥ and ⎢ ⎥ ⎣4⎦ q = 3 + 4 x2 . (a) From Theorem 4. ⎡1 0 1 10 ⎢ ⎢⎣0 1 15 2⎤ 5⎥ − 15 ⎥ ⎦ ⎡1 so PE → B = ⎢ 10 ⎢⎣ 15 2⎤ 5⎥ − 15 ⎥ ⎦ where E is the standard basis for R 2 . 10. ⎡ 1 2 3⎤ ⎡ 6 ⎤ ⎡16 ⎤ [ w ]B = PS → B [ w ]S = ⎢0 2 3⎥ ⎢ −1⎥ = ⎢10 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 0 3⎦ ⎣ 4 ⎦ ⎣12 ⎦ and w = (16.6. ⎢ = B → B′ ⎢ 13 ⎥ − − 2 ⎢⎣0 1 −2 − 13 ⎥ ⎢ 2⎦ 2⎦ ⎣ or (p) S = (0.6. Thus. − 1). 2.

it is ⎡ 2 1⎤ sufficient to compute det ⎢ = 6. b). g1 and g 2 form a basis for V. Since ⎣ 0 3⎥⎦ this determinant is nonzero. ⎢⎣ − 23 12 ⎥⎦ [ w ]B′ = PE → B′ [ w ]E ⎡ − 1 1 ⎤ ⎡ 3⎤ = ⎢ 23 2 ⎥ ⎢ ⎥ ⎢⎣ − 2 12 ⎥⎦ ⎣ −5⎦ ⎡ −4 ⎤ =⎢ ⎥ ⎣ −7 ⎦ ⎡− 7 ⎤ ⎢ 2⎥ = ⎢ 23 ⎥ ⎢ 2⎥ ⎣ 6⎦ ⎡ 3 (c) ⎢ 1 ⎢ ⎣ −5 ⎡1 ⎢ ⎢0 ⎢ ⎣0 1 −1 2 2 1⎤ 1 0 1 −1 2 ⎥ reduces to ⎥ −3 2 1 1 1⎦ 5⎤ 0 0 3 2 2⎥ 1 0 −2 −3 − 12 ⎥ so ⎥ 0 1 5 1 6⎦ 5⎤ ⎡ 3 2 2⎥ ⎢ PB→ B′ = ⎢ −2 −3 − 1 ⎥ .6 ⎡1 0 − 1 2 ⎢ ⎢⎣0 1 − 32 [ w ]B′ = PB → B′ [ w ]B 5⎤ ⎡ 3 2 2 ⎥ ⎡ 9⎤ ⎢ = ⎢ −2 −3 − 1 ⎥ ⎢ −9 ⎥ 2 ⎢ ⎥ ⎢ ⎥ 1 6 ⎦ ⎣ −5⎦ ⎣ 5 1⎤ 2⎥ 1⎥ 2⎦ ⎡− 1 1 ⎤ so PE → B′ = ⎢ 2 2 ⎥ . 1)} ⎡ 2 0⎤ PB′→ B = ⎢ . 2 ⎢ ⎥ 1 6⎦ ⎣ 5 ⎡ 3 9. (a) ⎢ 1 ⎢ ⎣ −5 ⎡1 ⎢ ⎢0 ⎢ ⎣0 1 −1 1 0 0 ⎤ 1 0 0 1 0 ⎥ reduces to ⎥ −3 2 0 0 1⎦ 1⎤ 0 0 1 12 2⎥ 1 0 −1 12 − 12 ⎥ so ⎥ 0 1 1 2 1⎦ ⎡ 1 ⎢ PE → B′ = ⎢ −1 ⎢ ⎣ 1 1 2 1 2 2 1⎤ 2⎥ − 12 ⎥ . (0. ⎣ 1 3⎥⎦ ⎡ 2 0 1 0⎤ (c) ⎢ reduces to ⎣ 1 3 0 1⎥⎦ ⎡ 1 so PB→ B′ = ⎢ 2 ⎢⎣ − 16 95 0⎤ ⎥. standard basis for R3 .SSM: Elementary Linear Algebra ⎡ 1 −1 1 0 ⎤ (d) ⎢ reduces to ⎣3 −1 0 1⎥⎦ Section 4. (a) The span of f1 and f2 is the set of all linear combinations af1 + bf 2 = a sin x + b cos x and this vector can be represented by (a. 1⎥ 3⎦ 1 ⎡1 0 2 ⎢ ⎢⎣0 1 − 16 0⎤ ⎥ 1⎥ 3⎦ . ⎥ 1⎦ [ w ]B′ = PE → B′ [ w ]E 1⎤ ⎡ 1 1 2 2 ⎥ ⎡ −5 ⎤ ⎢ = ⎢ −1 1 − 1 ⎥ ⎢ 8 ⎥ 2 2 ⎢ ⎥ ⎢ ⎥ 1⎦ ⎣ −5⎦ ⎣ 1 2 ⎡2 2 1 1 0 0⎤ (b) ⎢ 1 −1 2 0 1 0 ⎥ reduces to ⎢ ⎥ ⎣ 1 1 1 0 0 1⎦ 3 1 − 5⎤ ⎡1 0 0 2 2 2⎥ ⎢ 3 ⎥ so 1 1 ⎢0 1 0 − 2 − 2 2 ⎢ ⎥ 0 2⎦ ⎣0 0 1 −1 1 − 5⎤ ⎡ 3 2 2⎥ ⎢ 2 3 ⎥ where E is the PE → B = ⎢ − 1 − 1 2 2 2 ⎢ ⎥ 0 2⎦ ⎣ −1 ⎡− 7 ⎤ ⎢ 2⎥ = ⎢ 23 ⎥ ⎢ 2⎥ ⎣ 6⎦ 11. Since g1 = 2f1 + f 2 and g 2 = 3f2 . 0). [ w ]B = PE → B [ w ]E 1 − 5⎤ ⎡ 3 2 2 ⎥ ⎡ −5 ⎤ ⎢ 2 3 1 1 = ⎢− −2 ⎥ ⎢ 8⎥ 2 2 ⎢ ⎥ ⎢ ⎥ 0 2 ⎦ ⎣ −5 ⎦ ⎣ −1 ⎡ 9⎤ = ⎢ −9 ⎥ ⎢ ⎥ ⎣ −5⎦ (b) Since B can be represented as {(1.

(a) PB→ S = ⎢ 2 5 3⎥ ⎢ ⎥ ⎣ 1 0 8⎦ ⎡ 1 2 1 0⎤ ⎡ 1 0 −3 2 ⎤ reduces to ⎢ (d) ⎢ ⎣0 1 2 −1⎥⎦ ⎣ 2 3 0 1⎥⎦ ⎡ −3 2 ⎤ so PS → B1 = ⎢ where S is the ⎣ 2 −1⎥⎦ ⎡ 1 2 3 1 0 0⎤ (b) ⎢ 2 5 3 0 1 0⎥ reduces to ⎢ ⎥ ⎣ 1 0 8 0 0 1⎦ ⎡ 1 0 0 −40 16 9 ⎤ ⎢0 1 0 13 −5 −3⎥ so ⎢ ⎥ 0 0 1 5 −2 −1⎦ ⎣ ⎡ −40 16 9 ⎤ PS → B = ⎢ 13 −5 −3⎥ . 0). ⎡ 1 0 ⎤ ⎡ 2 ⎤ ⎡ 1⎤ ⎥ [h]B′ = PB → B′ [h ]B = ⎢ 2 = ⎢⎣ − 16 13 ⎥⎦ ⎢⎣ −5⎥⎦ ⎢⎣ −2 ⎥⎦ ⎡ 1 2 1 1⎤ ⎡ 1 0 3 5⎤ 15. 1)} for R 2 . ⎡ 2⎤ [ w ]B1 = ⎢ ⎥ . 1) for R 2 . (0. (a) ⎢ reduces to ⎢ ⎣0 1 −1 −2 ⎥⎦ ⎣ 2 3 3 4⎥⎦ ⎡ 3 5⎤ .Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra ⎡ 2⎤ (d) [h]B = ⎢ ⎥ since B is the standard basis ⎣ −5 ⎦ for V. so PS → B2 = ⎢ ⎣ −3 1⎥⎦ ⎡ 4 −1⎤ ⎡ 2 ⎤ ⎡ 3⎤ [ w ]B2 = PS → B2 [ w ]S = ⎢ = ⎣ −3 1⎥⎦ ⎢⎣ 5⎥⎦ ⎢⎣ −1⎥⎦ [ w ]B = PB 1 2 → B1 [ w ]B 2 ⎡ 3 5 ⎤ ⎡ 3⎤ =⎢ ⎣ −1 −2⎥⎦ ⎢⎣ −1⎥⎦ ⎡ 4⎤ =⎢ ⎥ ⎣ −1⎦ ⎡ 3⎤ (e) [ w ]S = ⎢ −5⎥ ⎢ ⎥ ⎣ 0⎦ [ w ]B = PS → B [ w ]S ⎡ −40 16 9 ⎤ ⎡ 3⎤ = ⎢ 13 −5 −3⎥ ⎢ −5⎥ ⎢ ⎥⎢ ⎥ ⎣ 5 −2 −1⎦ ⎣ 0 ⎦ ⎡ −200⎤ = ⎢ 64⎥ ⎢ ⎥ ⎣ 25⎦ 96 . ⎢ ⎥ ⎣ 5 −2 −1⎦ standard basis {(1. ⎣ −1 −3⎥⎦ ⎡ 1 2 3⎤ 13. so PB2 → B1 = ⎢ ⎣ −1 −2 ⎥⎦ ⎡ 1 1 1 2⎤ ⎡ 1 0 2 5⎤ reduces to ⎢ (b) ⎢ ⎣0 1 −1 −3⎥⎦ ⎣3 4 2 3⎥⎦ ⎡ 2 5⎤ so PB1 → B2 = ⎢ . ⎣ −1⎦ [ w ]B2 = PB1 → B2 [ w ]B1 ⎡ 2 5⎤ ⎡ 2 ⎤ =⎢ ⎣ −1 −3⎥⎦ ⎢⎣ −1⎥⎦ ⎡ −1⎤ =⎢ ⎥ ⎣ 1⎦ (d) [ w ]B = PS → B [ w ]S ⎡ −40 16 9 ⎤ ⎡ 5⎤ = ⎢ 13 −5 −3⎥ ⎢ −3⎥ ⎢ ⎥⎢ ⎥ ⎣ 5 −2 −1⎦ ⎣ 1⎦ ⎡ −239⎤ = ⎢ 77 ⎥ ⎢ ⎥ ⎣ 30⎦ [ w ]S = PB →S [ w ]B ⎡ 1 2 3⎤ ⎡ −239 ⎤ = ⎢ 2 5 3⎥ ⎢ 77 ⎥ ⎢ ⎥⎢ ⎥ ⎣ 1 0 8⎦ ⎣ 30 ⎦ ⎡ 5⎤ = ⎢ −3⎥ ⎢ ⎥ ⎣ 1⎦ ⎡ 1 1 1 0⎤ ⎡ 1 0 4 −1⎤ reduces to ⎢ (e) ⎢ ⎥ 3 4 0 1 ⎣ ⎦ ⎣0 1 −3 1⎥⎦ ⎡ 4 −1⎤ . Since w is the standard basis vector (0.

y e2 θ x v2 From the diagram. [ w ]B1 = PS → B1 [ w ]S [ w ]B2 θ e1 1 − 5⎤ ⎡ 3 2 2 ⎥ ⎡ −5 ⎤ ⎢ 2 3 1 1 = ⎢− −2 ⎥ ⎢ 8⎥ 2 2 ⎢ ⎥ ⎢ ⎥ 0 2⎦ ⎣ −5⎦ ⎣ −1 ⎡ 9⎤ = ⎢ −9 ⎥ ⎢ ⎥ ⎣ −5 ⎦ = PB1 → B2 [ w ]B1 x From the diagram. sin ⎜ 2θ − ⎟ ⎟ 2 2 ⎠⎠ ⎝ ⎠ ⎝ ⎝ = (sin 2θ . (a) y v1 standard basis for R3 . − cos 2θ ) Thus by Theorem 4. 2 2 2 ⎢ ⎥ 1⎦ ⎣ 1 2 [ w ]B2 = PS → B2 [ w ]S ⎡2 (b) ⎢ 1 ⎢ ⎣1 ⎡1 ⎢ ⎢0 ⎢ ⎣0 2 1 1 0 0⎤ −1 2 0 1 0 ⎥ reduces to ⎥ 1 1 0 0 1⎦ 3 1 − 5⎤ 0 0 2 2 2⎥ 3 ⎥ .6. (a) ⎢ 1 ⎢ ⎣ −5 ⎡1 ⎢ ⎢0 ⎢ ⎣0 Section 4.SSM: Elementary Linear Algebra ⎡ 3 17. sin 2θ ). so 1 0 − 12 − 12 2 ⎥ 0 1 −1 0 2⎦ 1 − 5⎤ ⎡ 3 2 2⎥ ⎢ 2 3 ⎥ where S is the PS → B = ⎢ − 1 − 1 1 2 2 ⎢ 2 ⎥ 0 2⎦ ⎣ −1 1⎤ ⎡ 1 1 2 2 ⎥ ⎡ −5 ⎤ ⎢ = ⎢ −1 1 − 1 ⎥ ⎢ 8 ⎥ 2 2 ⎢ ⎥ ⎢ ⎥ 1⎦ ⎣ −5⎦ ⎣ 1 2 ⎡− 7 ⎤ ⎢ 2⎥ = ⎢ 23 ⎥ ⎢ 2⎥ ⎣ 6⎦ 19. it is clear that [ v1 ]S = (cos 2θ . so the angle between ⎝2 ⎠ v 2 and the positive x-axis is π π⎞ ⎛ − ⎜ π − 2θ − ⎟ = 2θ − and 2 2⎠ ⎝ ⎛ π⎞ π ⎞⎞ ⎛ ⎛ [ v 2 ]S = ⎜ cos ⎜ 2θ − ⎟ .6 1 −1 2 2 1⎤ 1 0 1 −1 2 ⎥ reduces to ⎥ −3 2 1 1 1⎦ 5⎤ 0 0 3 2 2⎥ 1 0 −2 −3 − 12 ⎥ so ⎥ 0 1 5 1 6⎦ PB → B 1 2 ⎡ 3 (c) ⎢ 1 ⎢ ⎣ −5 ⎡1 ⎢ ⎢0 ⎢ ⎣0 5⎤ ⎡ 3 2 2⎥ ⎢ = ⎢ −2 −3 − 1 ⎥ . 2 ⎢ ⎥ 1 6⎦ ⎣ 5 1 −1 1 0 0 ⎤ 1 0 0 1 0 ⎥ reduces to ⎥ −3 2 0 0 1⎦ 1⎤ 0 0 1 12 2⎥ 1 0 −1 12 − 12 ⎥ so ⎥ 0 1 1 2 1⎦ 1⎤ ⎡ 1 1 2 2⎥ ⎢ PS → B = ⎢ −1 1 − 1 ⎥ .2. the angle between the positive y-axis and v 2 is 5⎤ ⎡ 3 2 2 ⎥ ⎡ 9⎤ ⎢ = ⎢ −2 −3 − 1 ⎥ ⎢ −9 ⎥ 2 ⎢ ⎥ ⎢ ⎥ 1 6 ⎦ ⎣ −5 ⎦ ⎣ 5 ⎡− 7 ⎤ ⎢ 2⎥ = ⎢ 23 ⎥ ⎢ 2⎥ ⎣ 6⎦ ⎛π ⎞ 2 ⎜ − θ ⎟ = π − 2θ . sin 2θ ⎤ ⎡cos 2θ PB→ S = ⎢ sin 2 cos 2θ ⎥⎦ θ − ⎣ 97 .

⎢ ⎥ ⎢ ⎥ ⎣ 1⎦ ⎣ 4⎦ ⎡0⎤ ⎡ 1⎤ c3 = ⎢ 7 ⎥ . (1.2. 2). e2 . so ⎢ ⎥ ⎣0 0 1 1⎦ ⎡ 5⎤ ⎡ 1⎤ ⎡ −1⎤ ⎡1⎤ ⎢ 1⎥ = ⎢9 ⎥ − 3 ⎢ 3⎥ + ⎢1⎥ . x3 = t . then [ w ]B = P[ w ]B′ and [ w ]C = Q[ w ]B . 1⎤ ⎡1 0 0 ⎢0 1 0 −3⎥ . e n . 98 . 5e 2} for R3 . 5⎠ ⎝5 5 5⎠ ⎩⎝ 5 5 ⎛ 2 2 1 ⎞⎫ ⎜ − . . 1. P is the transition matrix from B = {(1.6 (a) True (b) True. (d) True (e) False. ⎡ 2 ⎤ ⎡ 1⎤ ⎡ −1⎤ ⎡ 1⎤ Thus ⎢ 0 ⎥ = ⎢ 1⎥ + (t − 1) ⎢ 1⎥ + t ⎢ −1⎥ . e3} and B = {2e3 .6. 5 1⎥ 5 ⎦⎥ Thus. − ⎟ . A would have to be invertible. 3e1 . so the system A ⎢ x2 ⎥ = b ⎢ ⎥ ⎢ ⎥ ⎣0 0 0 0 ⎦ ⎣ x3 ⎦ has the solution x1 = 1. . (a) ⎢ reduces to ⎢ . ⎡ 1 1 0 1 0 0⎤ (b) ⎢ 1 0 2 0 1 0 ⎥ reduces to ⎢ ⎥ ⎣0 2 1 0 0 1⎦ 4 ⎡1 0 0 5 ⎢ 1 ⎢0 1 0 5 ⎢ 2 0 0 1 − 5 ⎣⎢ 1 5 − 51 2 5 ⎡ 1 3 −2 ⎤ ⎡ 1 0 1⎤ 3. c2 = ⎢ 5⎥ . Row vectors: r1 = [ 2 −1 0 1]. − 52 ⎤ ⎥ 2⎥. c4 = ⎢ −1⎥ ⎢ ⎥ ⎢ ⎥ ⎣2⎦ ⎣ 7⎦ 23. in particular [ei ]B = ei for the standard basis vectors e1 . The transition matrix from B′ to C is QP. 0. ⎟⎬ ⎝ 5 5 5 ⎠⎭ ⎡ 1 0 1 0⎤ ⎡ 1 1 2 −1⎤ (b) ⎢ 1 0 1 0⎥ reduces to ⎢0 1 1 0⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 2 1 3 2⎦ ⎣0 0 0 1⎦ so the system Ax = b is inconsistent and b is not in the column space of A. 0). (a) By Theorem 4. e2 . ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 0 ⎦ ⎣ −1⎦ ⎣ −1⎦ ⎣ 1⎦ (f) False. If w is a vector in the space and [ w ]B′ is the coordinate vector of w relative to B ′. ⎟ .1. by Theorem 4. [ w ]C = Q( P[ w ]B′ ) = QP[ w ]B′ . B must be the standard basis for R n . so ⎥ 4 6 10 − ⎣ ⎦ ⎣0 1 −1⎥⎦ ⎡ −2 ⎤ ⎡ 1⎤ ⎡ 3⎤ ⎢⎣ 10 ⎥⎦ = ⎢⎣ 4 ⎥⎦ − ⎢⎣ −6⎥⎦ . r3 = [1 4 2 7] ⎡2⎤ ⎡ −1⎤ Column vectors: c1 = ⎢ 3⎥ .6. so Exercise Set 4. (0. (c) True ⎡ 1 −1 1 2 ⎤ (d) ⎢ 1 1 −1 0 ⎥ reduces to ⎢ ⎥ ⎣ −1 −1 1 0 ⎦ ⎡ x1 ⎤ ⎡ 1 0 0 1⎤ ⎢0 1 −1 −1⎥ .7 21. 2.7 1. ⎜ . The transition matrix from C to B′ is (QP)−1 = P −1Q −1. … . r2 = [3 5 7 −1]. then ⎡0 3 0⎤ PB→ S = ⎢ 0 0 5⎥ which is not a diagonal ⎢ ⎥ ⎣2 0 0⎦ matrix. ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢⎥ ⎣ −1⎦ ⎣ 1⎦ ⎣ 1⎦ ⎣1⎦ True/False 4. 1)} to S. x2 = t − 1. since. not just square. consider the bases S = {e1 . P is the transition matrix from S to ⎧⎛ 4 1 2⎞ ⎛1 1 2⎞ B = ⎨⎜ .Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra Section 4. − . ⎡ 1 −1 1 5 ⎤ (c) ⎢9 3 1 1⎥ reduces to ⎢ ⎥ ⎣ 1 1 1 −1⎦ 27.

x2 = t . x = ⎢ ⎥ + t ⎢ ⎥ . 5 5 5 7 4 3 x2 = + s − t . and ⎥ ⎣ 2 −6 2 ⎦ ⎣0 0 0 ⎥⎦ the solution of Ax = b is x1 = 1 + 3t . or 0 0 0 0 ⎡6⎤ ⎡7⎤ ⎡ 1⎤ ⎢5⎥ ⎢5⎥ ⎢ 5⎥ 7⎥ 3 4 ⎢ vector form. ⎣ 0⎦ ⎣ 1⎦ The general form of the solution of Ax = 0 is ⎡3⎤ x = t ⎢ ⎥. (a) A basis for the row space is r1 = [1 0 2]. x3 = t . x4 = t . x2 = r . 99 . ⎢ ⎥ ⎢ ⎥ ⎣0 ⎦ ⎣0⎦ ⎡ 1 −2 1 2 −1⎤ ⎢0 0 0 0 0 ⎥ ⎢ ⎥ . or in 5 5 5 ⎡ 1 1 2 5⎤ (b) ⎢ 1 0 1 −2 ⎥ reduces to ⎢ ⎥ ⎣ 2 1 3 3⎦ ⎡ 1 0 1 −2 ⎤ ⎢0 1 1 7 ⎥ . x2 = 7 − t . r2 = [0 0 1]. x3 = s.SSM: Elementary Linear Algebra ⎡1 ⎢0 (e) ⎢ ⎢1 ⎢⎣0 2 1 2 1 0 2 1 2 1 1 3 2 ⎡1 ⎢0 ⎢ ⎢0 ⎢⎣0 0 1 0 0 0 0 1 0 0 −26 ⎤ 0 13⎥ ⎥ . 0 ⎢ ⎥ ⎢ 1⎥ ⎢ 0 ⎥ ⎣⎢ 0 ⎦⎥ ⎣⎢ 0 ⎦⎥ ⎣⎢ 1⎦⎥ ⎡4⎤ ⎡ 1⎤ ⎡ 2⎤ ⎡ 0⎤ ⎡ 1⎤ ⎢ 3⎥ ⎢0⎥ ⎢ 1⎥ ⎢ 2⎥ ⎢ 1⎥ ⎢ ⎥ = −26 ⎢ ⎥ + 13 ⎢ ⎥ − 7 ⎢ ⎥ + 4 ⎢ ⎥ . ⎣ 1⎦ 0 − 75 1 − 45 − 15 3 5 6⎤ 5⎥ 7⎥ 5 and the solution of ⎥ 0 0⎥ 0 0 ⎥⎦ 6 7 1 Ax = b is x1 = + s + t . x = 5 + s ⎢ 5 ⎥ + t ⎢ − 5 ⎥ . ⎢ ⎥ ⎢ ⎥ ⎢ 1⎥ ⎢ 0 ⎥ ⎢⎣ 0⎥⎦ ⎢⎣ 1⎥⎦ ⎡ −2 ⎤ ⎡ −1⎤ in vector form. or in vector form 4⎤ 3⎥ ⎥ reduces to 5⎥ 7 ⎥⎦ ⎡ −1⎤ ⎡2⎤ ⎡ −1⎤ ⎡ −2 ⎤ ⎢ 0⎥ ⎢ 1⎥ ⎢ 0⎥ ⎢ 0⎥ x = ⎢ ⎥ + r ⎢ ⎥ + s⎢ ⎥ +t ⎢ ⎥. ⎢ 5⎥ ⎢ 1⎥ ⎢ 2⎥ ⎢ 1⎥ ⎢ 3⎥ ⎣⎢7 ⎦⎥ ⎣⎢ 0 ⎦⎥ ⎣⎢ 1⎦⎥ ⎣⎢ 2 ⎦⎥ ⎣⎢ 2 ⎦⎥ 1 4⎤ ⎡ 1 2 −3 ⎢ −2 1 2 1 −1⎥ (d) ⎢ ⎥ reduces to 3⎥ ⎢ −1 3 −1 2 ⎣⎢ 4 −7 0 −5 −5⎦⎥ ⎡ 1 −3 1⎤ ⎡ 1 −3 1⎤ 5. x4 = t . (a) ⎢ reduces to ⎢ . The ⎢ ⎥ ⎢ ⎥ ⎣ 0 ⎦ ⎣ 1⎦ general form of the solution of Ax = 0 is ⎡ −1⎤ x = t ⎢ −1⎥ .7 x3 = s. ⎡1 ⎢ ⎢0 ⎢ ⎢0 ⎢⎣0 ⎡ 1⎤ ⎡3⎤ or in vector form. and the solution of ⎢ ⎥ ⎣0 0 0 0 ⎦ Ax = b is x1 = −2 − t . c2 = ⎢ 1⎥ . ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ 0⎥ ⎢ 1⎥ ⎢ 0 ⎥ ⎢⎣ 0 ⎥⎦ ⎢⎣ 1⎥⎦ ⎢⎣ 0⎥⎦ The general form of the solution of Ax = 0 is ⎡7⎤ ⎡ 1⎤ ⎢5⎥ ⎢ 5⎥ 3 4 x = s ⎢ 5 ⎥ + t ⎢− 5 ⎥ . ⎢ 0⎥ ⎢0⎥ ⎢ 1⎥ ⎢ 0 ⎥ ⎢⎣ 0 ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎢⎣ 1⎥⎦ The general form of the solution of Ax = 0 is ⎡2⎤ ⎡ −1⎤ ⎡ −2 ⎤ ⎢ 1⎥ ⎢ 0⎥ ⎢ 0⎥ x = r ⎢ ⎥ + s ⎢ ⎥ +t ⎢ ⎥. so 0 −7 ⎥ 1 4 ⎥⎦ Section 4. x = ⎢ 7 ⎥ + t ⎢ −1⎥ . ⎢ ⎥ ⎣ 1⎦ 7. A basis for the column ⎡ 1 −2 1 2 −1⎤ ⎢ 2 −4 2 4 −2 ⎥ (c) ⎢ ⎥ reduces to 1⎥ ⎢ −1 2 −1 −2 ⎢⎣ 3 −6 3 6 −3⎥⎦ ⎡ 1⎤ ⎡2⎤ space is c1 = ⎢0 ⎥ . and the solution of ⎢0 0 0 0 0 ⎥ ⎣⎢0 0 0 0 0 ⎦⎥ Ax = b is x1 = −1 + 2r − s − 2t .

r4 = [0 0 0 1]. r2 = [0 1 −3 0]. r2 = [0 1 4 3]. (a) A basis for the row space is r1 = [1 0 2]. r2 = [0 1 0 0]. − ⎟ . r3 = [0 0 1 −3]. 1. ⎢ 0⎥ ⎣⎢ 0 ⎦⎥ ⎡ −3⎤ ⎢ 1⎥ c2 = ⎢ ⎥ . c4 = ⎢ −3⎥ . ⎢ 0⎥ ⎣⎢ 0 ⎦⎥ (c) A basis for the row space is r1 = [1 2 4 5]. r2 = [0 1 0 0]. ⎢0 ⎥ ⎣⎢0 ⎦⎥ ⎡ 1⎤ ⎢0 ⎥ A basis for the column space is c1 = ⎢ ⎥ . ⎢ 1⎥ ⎢0⎥ ⎢ −7 ⎥ ⎢⎣ 0 ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎢⎣ 1⎥⎦ ⎡2⎤ ⎡ −1⎤ ⎡ 5⎤ ⎢ 1⎥ ⎢ 4⎥ ⎢ 3⎥ c 2 = ⎢ ⎥ . ⎢0 ⎥ ⎣⎢0 ⎦⎥ ⎡ 1⎤ ⎢0 ⎥ A basis for the column space is c1 = ⎢ ⎥ . c3 = ⎢ 1⎥ . c3 = ⎢ ⎥ . ⎢0 ⎥ ⎢⎣0 ⎥⎦ ⎡2⎤ ⎡ 4⎤ ⎡ 5⎤ ⎢ 1⎥ ⎢ −3⎥ ⎢ 0⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ c2 = ⎢ 0 ⎥ . 1. 1. r3 = [0 0 1 −3]. ⎢0⎥ ⎢ 0⎥ ⎢ 1⎥ ⎢⎣ 0 ⎥⎦ ⎢⎣ 0⎥⎦ ⎢⎣ 0 ⎥⎦ (d) A basis for the row space is r1 = [1 2 −1 5]. r2 = [0 1 4 3]. r4 = [0 0 0 1]. −5. r3 = [0 0 1 −7]. c 4 = ⎢ ⎥ . 2⎠ ⎝ 100 . c4 = ⎢ −3⎥ . c3 = ⎢ ⎥ . A basis for the column ⎡ 1⎤ ⎡2⎤ ⎢ ⎥ space is c1 = 0 . ⎡ 1⎤ ⎢0 ⎥ ⎢ ⎥ A basis for the column space is c1 = ⎢0 ⎥ . 1⎞ ⎛ ⎜ 0. ⎢0 ⎥ ⎣⎢0 ⎦⎥ ⎡ −3⎤ ⎢ 1⎥ c2 = ⎢ ⎥ . ⎡ 1 1 −4 −3⎤ 11. c3 = ⎢ 1⎥ . r4 = [0 0 0 1]. ⎡ 1⎤ ⎢0 ⎥ A basis for the column space is c1 = ⎢ ⎥ . c2 = ⎢ 1⎥ . (0. r3 = [0 0 1 −7]. ⎡ 1⎤ ⎢0 ⎥ A basis for the column space is c1 = ⎢ ⎥ .Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra (b) A basis for the row space is r1 = [1 −3 0 0]. ⎢0 ⎥ ⎣⎢0 ⎦⎥ ⎡ −1⎤ ⎡2⎤ ⎡ 5⎤ ⎢ 4⎥ ⎢ 1⎥ ⎢ 3⎥ c 2 = ⎢ ⎥ . (a) A row echelon form of ⎢ 2 0 2 −2 ⎥ ⎢ ⎥ ⎣ 2 −1 3 2 ⎦ ⎡ 1 1 −4 −3⎤ is ⎢0 1 −5 −2 ⎥ . Thus a basis for the ⎢ ⎥ 1 − 12 ⎦⎥ ⎣⎢0 0 r2 = [0 0 1]. 0. ⎢ ⎥ ⎢ ⎥ ⎣0 ⎦ ⎣0⎦ subspace is (1. c 4 = ⎢ ⎥ . r2 = [0 1 −3 0]. −4. ⎢0 ⎥ ⎢⎣0 ⎥⎦ ⎡2⎤ ⎡ 4⎤ ⎡ 5⎤ ⎢ 1⎥ ⎢ −3⎥ ⎢ 0⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ c2 = ⎢ 0 ⎥ . −2). (b) A basis for the row space is r1 = [1 −3 0 0]. r4 = [0 0 0 1]. ⎢0⎥ ⎢ 1⎥ ⎢ −7 ⎥ ⎢⎣ 0 ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎢⎣ 1⎥⎦ 9. (c) A basis for the row space is r1 = [1 2 4 5]. (d) A basis for the row space is r1 = [1 2 −1 5]. ⎢0⎥ ⎢ 0⎥ ⎢ 1⎥ ⎢⎣ 0 ⎥⎦ ⎢⎣ 0⎥⎦ ⎢⎣ 0 ⎥⎦ ⎡ 1⎤ ⎢0 ⎥ ⎢ ⎥ A basis for the column space is c1 = ⎢0 ⎥ . −3).

2.7 ⎡ −1 1 −2 0 ⎤ (b) A row echelon form of ⎢ 3 3 6 0 ⎥ is ⎢ ⎥ ⎣ 9 0 0 3⎦ 0⎤ ⎡ 1 −1 2 ⎢0 1 0 0 ⎥ . 1. 1. it is ⎡0 0 0 ⎤ clear that one such matrix is ⎢0 1 0 ⎥ . c2 = ⎢0 ⎥ . y) is in the null space of C. subspace is (1. 1). the matrices A = ⎢ and B = ⎢ ⎣0 0 ⎥⎦ ⎣ 2 6 ⎥⎦ have the same row space but different column spaces. 0). 0. 0. (c) False. 0.1. Thus the row space of B will have fewer than n vectors in its basis. That is. 0. elementary row operations do not change the null space of a matrix. y = 0. the column vectors of A that correspond to the column vectors of R containing the leading 1’s form a basis for the column space of A. this is the contrapositive of Theorem 4. Thus a basis for the ⎢ 1⎥ ⎢⎣0 0 1 − 6 ⎥⎦ (b) det(A) = det(B) = 5 A and B are invertible so the systems Ax = 0 and Bx = 0 have only the trivial solution. − ⎟ . 1. However. ⎢ ⎥ ⎢ ⎥ ⎣0 ⎦ ⎣0 ⎦ which is all linear combinations of y and x. ⎡ 1 0 0⎤ 15. i. (b) False. (a) If the null space of a matrix A is the given line. (0.7. see (e) for an example. ⎢ ⎥ ⎣0 0 0 ⎦ so the general solution of the system Ax = 0 ⎡0 ⎤ is x = 0. (b) Reversing the reasoning in part (a). the null space of C is the line 3x + y = 0. (i) True. so the null space of D is the entire xy-plane. (d) False. ⎡ 1 1 0 0⎤ ⎢0 1 1 1⎥ ⎢ ⎥ . (h) False. not both of which are zero. so in reduced row echelon form it has at least one row of zeros.. the row vectors of A may be linearly dependent if A is not in row echelon form. (a) The row echelon form of A is ⎢0 1 0 ⎥ . 1⎞ ⎛ ⎜ 0.SSM: Elementary Linear Algebra Section 4. 6⎠ ⎝ ⎡ 1 1 ⎢ 0 0 (c) A row echelon form of ⎢ ⎢ −2 0 ⎢⎣ 0 −3 0 1 2 0 0⎤ 1⎥ ⎥ is 2⎥ 3⎥⎦ True/False 4. −1. (0. i. 1. 0). 0.e. 1.. and the column ⎡ 1⎤ ⎡0 ⎤ space is the span of c1 = ⎢ 1⎥ . elementary row operations do not change the row space of a matrix. then the equation Ax = 0 is equivalent ⎡ 3 −5⎤ ⎡ x ⎤ ⎡0 ⎤ to ⎢ = and a general form ⎣0 0 ⎥⎦ ⎢⎣ y ⎥⎦ ⎢⎣0 ⎥⎦ ⎡3a −5a ⎤ where a and b are real for A is ⎢ ⎣ 3b −5b ⎥⎦ numbers. the system Ax = b may not be consistent. 1). the null spaces of A and B are both the origin. ⎢ ⎥ ⎣0 0 1⎦ (j) False. 0.e. 1). (0. ⎡x⎤ The equation Dx = 0 has all x = ⎢ ⎥ as ⎣ y⎦ solutions.7 (a) True. (g) True. Thus a basis for the ⎢0 0 1 1⎥ ⎣⎢0 0 0 1⎦⎥ subspace is (1. the row space of A will have n vectors in its basis since being invertible means that A is row equivalent to I n . ⎡1 3⎤ ⎡1 3 ⎤ (e) False. (f) True. since B is 17. the xy-plane. by the definition of column space. assuming that A and B are both n × n matrices. Thus the null ⎢ ⎥ ⎣ 1⎦ space of A is the z-axis. singular. (0. 0). it is not row equivalent to I n . 101 . The second row of C is a multiple of the first and it is clear that if 3x + y = 0. z = t or t ⎢0 ⎥ . then (x. 1.

and since nullity(A) = 3 − 3 = 0. which has the general solution b1 = r . the system is Section 4. there are 2 leading variables. so rank (A) = 2. (c) Since A is 5 × 3. Since nullity(A) = 5 − 2 = 3. and since nullity(A) = 9 − 2 = 7. b2 = s. so the smallest possible value of nullity(A) is 2. the system is consistent. (f) Since rank( A = rank ⎡⎣ A b ⎤⎦ . ⎢0 1 17 7 7 ⎢ ⎥ 0 0⎦ ⎣0 0 ⎡ 1 −3 −2 ⎤ ⎢2 1 3⎥ AT = ⎢ ⎥ and the reduced row 4 5 9⎥ ⎢ ⎣⎢ 0 2 2 ⎦⎥ ⎡1 ⎢0 echelon form of A is ⎢ ⎢0 ⎣⎢0 T 0 1 0 0 (b) Since rank( A) < rank ⎡⎣ A b ⎤⎦ . (a) Since A is 4 × 4.8 consistent. (b) Since A is 3 × 5. Since nullity(A) = 4 − 2 = 2. and the largest possible value of rank(A) is 4. inconsistent. b4 = 2r − s. The augmented matrix is ⎢1 1 b3 ⎥ . so the smallest possible value of nullity(A) is 0. Since nullity(A) = 3 − 2 = 1. consistent. there are 4 parameters in the general solution. (b) Since rank(A) = 1. there is 1 parameter. there are 0 parameters in the general solution. rank(A) + nullity(A) = 4. (a) Since rank (A) = 2. 5. rank( AT ) = rank( A) = 2. The reduced row echelon form of A is ⎡1 0 − 6 − 4⎤ 7 7⎥ ⎢ 2 ⎥ . the system is inconsistent. ⎡1 −3 b1 ⎤ ⎢1 −2 b ⎥ 2⎥ ⎢ 9. (d) Since rank(A) = 2. there are 2 parameters in the general solution. there are 3 leading variables. there are 2 leading variables. which is ⎢1 −4 b4 ⎥ ⎢1 5 b ⎥ 5⎦ ⎣ 3b2 − 2b1 ⎤ ⎡1 0 ⎢0 1 b2 − b1 ⎥ ⎢ ⎥ row equivalent to ⎢0 0 b3 − 4b2 + 3b1 ⎥ . there are 2 parameters.8 1. the system is consistent. and since nullity(A) = 4 − 0 = 4. there are 2 parameters. 1⎤ 1⎥ ⎥ . (e) Since rank( A) < rank ⎡⎣ A b ⎤⎦ . the system is (c) Since rank(A) = 2. 102 . ⎢0 0 b4 + b2 − 2b1 ⎥ ⎢0 0 b − 8b + 7b ⎥ 5 2 1⎦ ⎣ the system is consistent if and only if =0 3b1 − 4b2 + b3 −2b1 + b2 + b4 =0 + b5 = 0 7b1 − 8b2 (e) Since rank(A) = 3. rank(A) + nullity(A) = 5. rank(A) + nullity(A) = 3. and since nullity(A) = 3 − 1 = 2. so 0⎥ 0 ⎦⎥ (d) Since rank( A) = rank ⎡⎣ A b ⎤⎦ . Since nullity(A) = 5 − 3 = 2. there are 0 parameters in the general solution. the system is consistent. there are 3 parameters. Thus. there are 2 leading variables. Since nullity(A) = 3 − 1 = 2. (a) Since rank(A) = rank [ A b ] . there are 2 parameters. b5 = −7 r + 8s. (g) Since rank( A) = rank ⎡⎣ A b ⎤⎦ .Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra 7. and since nullity(A) = 2 − 2 = 0. (c) Since rank( A) = rank ⎡⎣ A b ⎤⎦ . there is 1 leading variable. so the smallest possible value of nullity(A) is 0. b3 = −3r + 4s. and the largest possible value of rank(A) is 3. Exercise Set 4. there are 7 parameters in the general solution. and the largest possible value of rank(A) is 3. the system is 3.

the rank will never be 1. Let A = ⎢ and B = ⎢ . (c) Since A has size 3 × 3. (h) False.. (c) The transformation is linear. the row space and column space must both be 2-dimensional. (a) The transformation is linear. the nullity of a nonzero m × n matrix is at most n − 1. 3. (f) False. the row space and null space cannot both have dimension 1 since dim(row space) + dim(null space) = 3. R3 and codomain R 2 . (b) The number of parameters is at most 5. 103 . (e) True. since dim(row space) = dim(column space). since a matrix cannot have more leading 1’s than rows. with domain (c) False. Since the third column will always have a nonzero entry. adding it does not affect the rank. Then ⎥ 0 0 ⎣ ⎦ ⎣ 2 4 ⎥⎦ 1. 2.9 ⎡0 1⎤ ⎡ 1 2⎤ 19. domain R 2 and codomain R3 . Section 4. then neither are the column vectors. (b) The transformation is nonlinear. A must be square. (d) The number of parameters is at most 3 since the solution cannot have more parameters than A has columns. but since A = ⎢ ⎣ 0 0 ⎥⎦ ⎡ 5 10⎤ and B 2 = ⎢ . the domain of TA is R 2 and the codomain of TA is R3 . True/False 4. (a) Since A has size 3 × 2. i. T(1. if the added column is linearly dependent on the previous columns.9 Exercise Set 4. the domain of TA is (a) False. (i) True.SSM: Elementary Linear Algebra Section 4. (a) The number of leading 1’s is at most 3. 3(1)) = (−1. since the solution cannot have more parameters than A has columns. since the row space of A has the same dimension as the column space. No. its nullity is at least 1. which is at most 3.9 11. if the row vectors are linearly independent then nullity(A) = 0 and rank(A) = n = the number of rows. rank(A) = rank(AT ) for all matrices. (g) False. The rank is 2 if r = 2 and s = 1. then Ax = b is consistent for every vector b. the relationship is that V ⊥ is a subspace of W ⊥ . a vector w in W ⊥ need not be orthogonal to every vector in V. if the rows are linearly dependent. ⎣10 20⎥⎦ (b) Since A has size 2 × 3. R3 . the domain of TA is rank( A2 ) = 0 ≠ rank( B 2 ) = 1. 3) 5. R3 and the codomain of TA is R3 . the row vectors are not necessarily linearly independent.e. with (d) False. so since the matrix is square. since rank(A) + nullity(A) = 3 and nullity(A) = 1. 13. 17. ⎡0 0⎤ rank(A) = rank(B) = 1. The domain of T is R 2 and the codomain of T is (b) True. since rank(A) + nullity(A) = the number of columns. then the rank is at least 1 less than the number of rows. R 6 and the codomain of TA is R1 . the dimension of the row space is always equal to the dimension of the column space. −(−2). planes in 3-space. if the nullity of A is zero. and if they are not. But. (c) The number of leading 1’s is at most 3. with domain R3 and codomain R3 . the domain of TA is 2 R3 and the codomain of TA is R 2 . (j) False.8 (d) Since A has size 1 × 6. −2) = (1 − 2.

kz ) ≠ kT ( x. (a) ⎢0 1 0⎥ ⎢ −5⎥ = ⎢ −5⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 0 −1⎦ ⎣ 3⎦ ⎣ −3⎦ = (ky 2 . 3). −5. −5. z ) ⎡ 1 0 0⎤ ⎡ 2 ⎤ ⎡ 2 ⎤ 15. since T(0. since for k ≠ 1. ⎡0 0 0⎤ . 4) = (3. 2. 4) (b) T is not a matrix transformation since T(0. 1. 0. w2 = 4(−1) − (2) + 4 = −2 w3 = 3(−1) + 2(2) − (4) = −3 so T(−1. 7. 0) = (0. 0⎥ 0 ⎥⎦ 104 . −3) = (2(2) − (1) + (−3). (a) The matrix is ⎢ ⎥. ⎢ 1 3⎥ ⎣⎢ 1 −1⎦⎥ ⎡ 1 0 0 ⎤ ⎡ −2 ⎤ ⎡ −2 ⎤ (b) ⎢0 0 0 ⎥ ⎢ 1⎥ = ⎢ 0⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 0 1⎦ ⎣ 3⎦ ⎣ 3⎦ The result is (−2. ⎡ 3 −4 0 ⎤ . ⎡0 ⎢0 ⎢ (c) The matrix is ⎢0 ⎢0 ⎢⎣0 0 0 0 0 0 1 1 0 0 −1 ⎡ 0 0 0 ⎤ ⎡ −2 ⎤ ⎡ 0 ⎤ (c) ⎢0 1 0 ⎥ ⎢ 1⎥ = ⎢ 1⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 0 1⎦ ⎣ 3⎦ ⎣ 3⎦ The result is (0. 1. 3). (a) AT x = ⎢ = ⎣ 0 1⎥⎦ ⎢⎣ 4⎥⎦ ⎢⎣ 4 ⎥⎦ T(−1. 1. ⎡ 1 0 0⎤ ⎡ 2⎤ ⎡2⎤ (b) ⎢0 −1 0 ⎥ ⎢ −5⎥ = ⎢ 5⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 0 1⎦ ⎣ 3⎦ ⎣ 3⎦ The result is (2. TA = ⎢ ⎣ 0 0 0 ⎥⎦ (c) T is a matrix transformation. 0) (d) T is not a matrix transformation. ⎡ 0 1⎤ ⎢ −1 0 ⎥ 11. ⎢ 4 −1 1⎥ ⎢ 2 ⎥ = ⎢ −2 ⎥ . (a) T is a matrix transformation. y. ⎡ 3 5 −1⎤ 9. −2. −2. y. 0) = (−1. 4) = (−(−1) + 4. The result is (2. 0. y. 3). z )) = (k 2 y 2 . 2. 0. 4) = (5. 0⎥ 0 ⎥⎦ ⎡ −1 1⎤ ⎡ −1⎤ ⎡ 5 ⎤ 13. 0) ≠ 0. 0) ≠ 0. 1 + (−3). ⎢ ⎥ ⎣ 3 2 −1⎦ For (x. z) = (−1. −3). ⎢ ⎥ ⎣ −1 0 0 0 ⎦ 0 0 0 0 0 1⎤ 0⎥ ⎥ 0⎥ . ⎡ 7 2 −1 1⎤ (b) The matrix is ⎢ 0 1 1 0 ⎥ . 0). w1 = 3(−1) + 5(2) − 4 = 3 ⎡ −1 0 0⎤ ⎡ 2 ⎤ ⎡ −2 ⎤ (c) ⎢ 0 1 0⎥ ⎢ −5⎥ = ⎢ −5⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 0 0 1⎦ ⎣ 3⎦ ⎣ 3⎦ The result is (−2. 3). By inspection. T (k ( x. 0⎤ 0⎥ ⎥ 0⎥ . ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 3 2 −1⎦ ⎣ 4 ⎦ ⎣ −3⎦ ⎡ 1 0 0 ⎤ ⎡ −2 ⎤ ⎡ −2 ⎤ 17. (a) ⎢0 1 0 ⎥ ⎢ 1⎥ = ⎢ 1⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 0 0 0 ⎦ ⎣ 3⎦ ⎣ 0 ⎦ The result is (−2. TA = ⎢ ⎣ 2 0 −5⎥⎦ ⎡ 2 −1 1⎤ ⎡ 2 ⎤ ⎡ 0 ⎤ (b) AT x = ⎢ 0 1 1⎥ ⎢ 1⎥ = ⎢ −2 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 0 0 0 ⎦ ⎣ −3⎦ ⎣ 0 ⎦ T(2. 5. T = ⎢ 4 −1 1⎥ . ⎡ 3 5 −1⎤ ⎡ −1⎤ ⎡ 3⎤ Also. −3). with domain R 4 and codomain R 2 .Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra ⎡0 ⎢1 ⎢ (d) The matrix is ⎢0 ⎢0 ⎢⎣ 1 (d) The transformation is nonlinear. 4). (e) T is not a matrix transformation. kz ).

(a) ⎢0 cos 30° − sin 30°⎥ ⎢ 1⎥ = ⎢ 0 2 − 2 ⎥ ⎢ 1⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎢ ⎥ 3 ⎥ ⎣ 2⎦ ⎣0 sin 30° cos 30° ⎦ ⎣ 2 ⎦ ⎢ 0 1 ⎥ ⎣ 2 2 ⎦ ⎡ −2 ⎤ ⎢ 3 −2 ⎥ =⎢ 2 ⎥ ⎢ 1+ 2 3 ⎥ ⎣⎢ 2 ⎦⎥ ⎛ The result is ⎜ −2. 0⎤ ⎡1 0 0 0 ⎡1 ⎤ ⎡ −2 ⎤ ⎢ ⎡ −2 ⎤ 3 1 ⎥⎢ ⎥ 21. 2 2 ) . 2). (a) ⎢0 cos(−30°) − sin(−30°) ⎥ ⎢ 1⎥ = ⎢0 2 ⎥ 1 2 ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ 2⎥ − ° − ° 0 30 30 2 sin( ) cos( ) 3 1 ⎣ ⎦ ⎣ ⎦ ⎢0 − ⎣ ⎦ 2 2 ⎥⎦ ⎣ ⎡ −2 ⎤ ⎢ 3+2 ⎥ =⎢ 2 ⎥ ⎢ −1+ 2 3 ⎥ ⎢⎣ 2 ⎥⎦ ⎛ ⎞ The result is ⎜ −2. ⎝ 3 − 2 1+ 2 3 ⎞. 1. ⎝ 2 2 ⎠ 105 . ⎟ 2 2 ⎠ ⎡ 1 0 ⎡ cos 45° 0 sin 45° ⎤ ⎡ −2 ⎤ ⎢ 2 ⎢ ⎥ ⎢ ⎥ (b) 0 1 0 1 =⎢ 0 1 ⎢ ⎥⎢ ⎥ ⎢ 1 − ° ° sin cos 45 0 45 2 0 ⎣ ⎦ ⎣ ⎦ ⎢− ⎣ 2 ⎡ 0 ⎤ =⎢ 1 ⎥ ⎢ ⎥ ⎢⎣ 2 2 ⎥⎦ 1 ⎤ 2 ⎥ ⎡ −2 ⎤ 0 ⎥ ⎢ 1⎥ 1 ⎥ ⎢⎣ 2 ⎥⎦ 2 ⎥⎦ The result is ( 0.SSM: Elementary Linear Algebra Section 4.9 0 ⎤ ⎡1 0 0 0 ⎤ ⎡ −2 ⎤ ⎢ ⎡1 ⎡ −2 ⎤ 3 1⎥ 19. 3 + 2 . ⎡cos 90° − sin 90° 0 ⎤ ⎡ −2 ⎤ ⎡ 0 −1 0 ⎤ ⎡ −2 ⎤ (c) ⎢ sin 90° cos 90° 0⎥ ⎢ 1⎥ = ⎢ 1 0 0 ⎥ ⎢ 1⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥⎢ ⎥ 0 1 ⎦ ⎣ 2 ⎦ ⎣ 0 0 1⎦ ⎣ 2 ⎦ ⎣ 0 ⎡ −1⎤ = ⎢ −2 ⎥ ⎢ ⎥ ⎣ 2⎦ The result is (−1. −1 + 2 3 ⎟ . −2. .

sin 180° = 0 The matrix is ⎡ 2 2 2 2 (1 − (−1)) − 1 (0) ⎢ 3 (1 − (−1)) + (−1) 3 3 3 ⎢ 2 2 (1 − ( −1)) + ( −1) ⎢ 23 23 (1 − (−1)) + 13 (0) 3 ⎢ ⎢ 2 1 (1 − (−1)) − 2 (0) 2 1 (1 − (−1)) + 2 (0) 3 3 3 3 ⎣ 3 3 v v = () ( )( ) ( )( ) (2. ⎟ = (a. b. 2). unless T is the zero operator in which case the image is 0. Since cos 2θ = cos 2 θ − sin 2 θ and sin 2θ = 2 sin θ cos θ. ⎡cos 2θ 31. the matrix can be written as ⎢ ⎣ sin 2θ has the effect of rotating x through the angle 2θ. 0 ) . ⎛ 2 2 1⎞ = ⎜ . 106 . It is not a matrix transformation since T (0) = x0 ≠ 0. 1) 2 2 2 ( )( ) () ( )( ) ( 23 )( 13 ) (1 − (−1)) + ( 23 ) (0)⎤⎥ ⎡ 94 (2) − 1 ⎥ ⎢ ( 23 )( 13 ) (1 − (−1)) − ( 32 ) (0) ⎥⎥ = ⎢⎢ 94 (2) 2 2 1 (1 − ( −1)) + ( −1) ⎥ ⎢⎣ 9 ( 2) (3) ⎦ ⎡− 1 ⎢ 9 = ⎢ 89 ⎢ 4 ⎢⎣ 9 8 9 − 19 4 9 ⎤ ⎥ 4 ( 2) − 1 ⎥ 9 ⎥ 2 ( 2) 1 ( 2) − 1 ⎥⎦ 9 9 4 ( 2) 9 2 ( 2) 9 2 ( 2) 9 4⎤ 9⎥ 4⎥ 9 ⎥ − 79 ⎥ ⎦ 29. (a) The result is twice the orthogonal projection of x on the x-axis. c) ⎝ 3 3 3⎠ 2 + 2 +1 cos 180° = −1.Chapter 4: General Vector Spaces ⎡ 1 ⎡ cos( −45°) 0 sin(−45°) ⎤ ⎡ −2 ⎤ ⎢ 2 ⎥ ⎢ 1⎥ = ⎢ 0 (b) ⎢ 0 1 0 ⎢ ⎥⎢ ⎥ ⎢ ⎣ − sin(−45°) 0 cos(−45°) ⎦ ⎣ 2 ⎦ ⎢ 1 ⎣ 2 ⎡ −2 =⎢ 1 ⎢ ⎣⎢ 0 SSM: Elementary Linear Algebra 0 − 1 0 1 ⎤ 2 ⎥ ⎡ −2 ⎤ 0 ⎥ ⎢ 1⎥ 1 ⎥ ⎢⎣ 2 ⎥⎦ 2 ⎥⎦ 2⎤ ⎥ ⎥ ⎦⎥ The result is ( −2 2 . 35. (b) The result is twice the reflection of x about the x-axis. 1. 2. The image of the line under the operator T is the line in R n which has vector representation x = T (x0 ) + tT ( v). The result of the transformation is rotation through the angle θ followed by translation by x0 . ⎡cos(−90°) − sin(−90°) 0⎤ ⎡ −2 ⎤ ⎡ 0 1 0 ⎤ ⎡ −2 ⎤ (c) ⎢ sin(−90°) cos(−90°) 0⎥ ⎢ 1⎥ = ⎢ −1 0 0 ⎥ ⎢ 1⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥⎢ ⎥ 0 0 1 ⎦ ⎣ 2 ⎦ ⎣ 0 0 1⎦ ⎣ 2 ⎦ ⎣ ⎡1 ⎤ = ⎢2⎥ ⎢ ⎥ ⎣2⎦ The result is (1. − sin 2θ ⎤ which cos 2θ ⎥⎦ 33. 25. 2. .

⎡ 3 0 ⎤ ⎡1 1⎤ ⎡ 3 3⎤ ⎢⎣ 2 4 ⎥⎦ ⎢⎣1 −1⎥⎦ = ⎢⎣ 6 −2 ⎥⎦ . there is no angle θ for which 1 cos θ = sin θ = . ⎥ 25⎦ The standard matrix for TA ⎡2 BA = ⎢ 5 ⎢ ⎣6 ⎡ 5 = ⎢10 ⎢ ⎣ 45 ⎡1 AB = ⎢ 4 ⎢ ⎣5 ⎡ −8 = ⎢ −5 ⎢ ⎣ 44 −3 0 1 −1 −8 3 0⎤ −3⎥ ⎥ 4⎦ 0⎤ ⎡ 0 0 ⎤ ⎡0 ⎥ = 1 ⎥ ⎢⎣ 0 1⎥⎦ ⎢ 0 ⎣ 2⎦ 0⎤ 1⎥ 2⎦ (c) The standard matrix for reflection about the ⎡ 1 0⎤ x-axis is ⎢ and the standard matrix ⎣0 −1⎥⎦ ⎡ 3 0⎤ for a dilation with factor k = 3 is ⎢ .9 (a) False. ⎣ 2 4 ⎥⎦ R3 . 5. so T cannot be a matrix operator. ⎥ −11 45⎦ 107 .10 Exercise Set 4. x2 )) = (3x1 + 3x2 . The standard matrix for T1 T2 is (c) False. x1 − 4 x2 ) T2 (T1 ( x1 . (a) The standard matrix for a rotation of 90° is ⎡cos 90° − sin 90°⎤ ⎡ 0 −1⎤ ⎢⎣ sin 90° cos 90°⎥⎦ = ⎢⎣ 1 0⎥⎦ and the standard matrix for reflection about the line ⎡0 1⎤ y = x is ⎢ . x2 )) = (5 x1 + 4 x2 . (b) False. (g) False. every matrix transformation will have this property by the homogeneity property. when a = 1 it is the matrix for reflection about the x-axis and when a = −1 it is the matrix for reflection about the y-axis. then the codomain of TA is (b) The standard matrix for T2 T1 is Rm . For instance if T: R n → R1 is given by T ( v) = v then T(0) = 0. ⎣0 3⎥⎦ ⎡ 3 0⎤ ⎡ 1 0⎤ ⎡ 3 0⎤ ⎢⎣0 3⎥⎦ ⎢⎣ 0 −1⎥⎦ = ⎢⎣ 0 −3⎥⎦ TB is −2 0 ⎤ ⎡ 2 −3 3⎤ 1 −3⎥ ⎢ 5 0 1⎥ ⎥⎢ ⎥ 2 4⎦ ⎣ 6 1 7⎦ −3 1⎤ −15 −8⎥ . then the domain of TA is ⎡ 3 0⎤ the standard matrix for T2 is ⎢ . T(0) = 0 is not a sufficient condition.10 ⎡1 1⎤ 3. the only matrix transformation with this property is the zero transformation (consider x = y ≠ 0). 6 x1 − 2 x2 ) (e) False. The standard matrix for TB TA is 3⎤ ⎡ 1 −2 1⎥ ⎢ 4 1 ⎥⎢ 7⎦ ⎣5 2 21⎤ 4⎥ . if A is m × n. 2 (b) The standard matrix for orthogonal ⎡0 0 ⎤ projection on the y-axis is ⎢ and the ⎣0 1⎥⎦ standard matrix for a contraction with factor ⎡ 1 0⎤ 1 ⎥. then T(0) = 0 + b = b ≠ 0. but T is not a matrix transformation. ⎣ 1 0 ⎥⎦ ⎡0 1⎤ ⎡0 −1⎤ ⎡ 1 0 ⎤ ⎢⎣ 1 0 ⎥⎦ ⎢⎣ 1 0 ⎥⎦ = ⎢⎣0 −1⎥⎦ (f) True. (d) True (c) T1 (T2 ( x1 . (a) The standard matrix for T1 is ⎢ and ⎣1 −1⎥⎦ True/False 4.SSM: Elementary Linear Algebra Section 4. if A is 2 × 3. ⎡1 1⎤ ⎡ 3 0 ⎤ ⎡5 4⎤ ⎢⎣1 −1⎥⎦ ⎢⎣ 2 4 ⎥⎦ = ⎢⎣ 1 −4⎥⎦ . k = is ⎢ 2 2 ⎢⎣ 0 12 ⎥⎦ (i) True. (h) False. since b ≠ 0.10 ⎡1 ⎢2 ⎢⎣ 0 1. Section 4.

2 ⎥⎦ 1⎤ 0 ⎥⎥ 1 ⎥⎦ ⎡ 1 0 0⎤ (c) The standard matrix for orthogonal projection on the xy-plane is ⎢0 1 0 ⎥ and the standard matrix for ⎢ ⎥ ⎣0 0 0 ⎦ ⎡ −1 0 0 ⎤ reflection about the yz-plane is ⎢ 0 1 0 ⎥ . 9. (a) [T1 ] = ⎢ and [T2 ] = ⎢ ⎣0 0 ⎥⎦ ⎣ 0 1 ⎥⎦ ⎡0 0 ⎤ so T1 T2 = T2 T1. (a) The standard matrix for reflection about the yz-plane is ⎢ 0 1 0 ⎥ and the standard matrix for orthogonal ⎢ ⎥ ⎣ 0 0 1⎦ ⎡ 1 0 0⎤ projection on the xz-plane is ⎢0 0 0 ⎥ .Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra ⎡ −1 0 0 ⎤ 7. [T1 ][T2 ] = [T2 ][T1 ] = ⎢ ⎣0 0 ⎥⎦ 108 . ⎢ ⎥ ⎣0 0 1⎦ ⎡ 1 0 0 ⎤ ⎡ −1 0 0⎤ ⎡ −1 0 0 ⎤ ⎢0 0 0 ⎥ ⎢ 0 1 0⎥ = ⎢ 0 0 0 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 0 1⎦ ⎣ 0 0 1⎦ ⎣ 0 0 1⎦ ⎡ 1 ⎡ cos 45° 0 sin 45° ⎤ ⎢ 2 (b) The standard matrix for a rotation of 45° about the y-axis is ⎢ 0 1 0 ⎥=⎢ 0 ⎢ ⎥ ⎢ ⎣ − sin 45° 0 cos 45° ⎦ ⎢ − 1 ⎣ 2 ⎡ 2 ⎢ the standard matrix for a dilation with factor k = 2 is ⎢ 0 ⎢⎣ 0 ⎡ 2 ⎢ ⎢ 0 ⎣⎢ 0 0 2 0 ⎡ 1 0 ⎤⎢ 2 ⎥ 0 ⎥⎢ 0 ⎢ 2 ⎦⎥ ⎢ − 1 ⎣ 2 0 1 0 1 ⎤ 2⎥ ⎡1 0 ⎥=⎢0 ⎢ 1 ⎥ ⎢ −1 ⎣ 2 ⎦⎥ 0 2 0 0 2 0 0 1 0 1 ⎤ 2⎥ 0 ⎥ and 1 ⎥ 2 ⎦⎥ 0 ⎤ ⎥ 0 ⎥. ⎢ ⎥ ⎣ 0 0 1⎦ ⎡ −1 0 0 ⎤ ⎡ 1 0 0 ⎤ ⎡ −1 0 0 ⎤ ⎢ 0 1 0⎥ ⎢ 0 1 0⎥ = ⎢ 0 1 0 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 0 0 1⎦ ⎣ 0 0 0⎦ ⎣ 0 0 0 ⎦ ⎡1 0 ⎤ ⎡0 0⎤ .

[T2 T1 ] = ⎢ ⎥ ⎣ sin(θ1 + θ 2 ) cos(θ1 + θ 2 ) ⎦ [T1 T2 ] = [T1 ][T2 ] ⎡cos θ1 − sin θ1 ⎤ ⎡cos θ 2 − sin θ 2 ⎤ =⎢ cos θ1 ⎥⎦ ⎢⎣ sin θ 2 cos θ 2 ⎥⎦ ⎣ sin θ1 ⎡ cos θ1 cos θ 2 − sin θ1 sin θ 2 − cos θ1 sin θ 2 − sin θ1 cos θ 2 ⎤ =⎢ ⎥ ⎣sin θ1 cos θ 2 + cos θ1 sin θ 2 − sin θ1 sin θ 2 + cos θ1 cos θ 2 ⎦ ⎡cos(θ1 + θ 2 ) − sin(θ1 + θ 2 ) ⎤ =⎢ ⎥ ⎣ sin(θ1 + θ 2 ) cos(θ1 + θ 2 ) ⎦ = [T2 T1 ] Thus T1 T2 = T2 T1.10 ⎡ cos(θ1 + θ 2 ) − sin(θ1 + θ 2 ) ⎤ (b) From Example 1. (a) Not one-to-one because the second row (column) is all 0’s. so the determinant is 0. (b) One-to-one because the determinant is −1. w2 ) = ⎜ w1 − w2 . (e) One-to-one because the determinant is 1. (f) One-to-one because the determinant is −1. (g) One-to-one because the determinant is k 3 ≠ 0. (c) [T1 ] = ⎢ ⎥ cos θ ⎥⎦ ⎣0 0 ⎦ ⎣ sin θ ⎡ 1 0 ⎤ ⎡cos θ − sin θ ⎤ [T1 ][T2 ] = ⎢ cos θ ⎥⎦ ⎣0 0 ⎥⎦ ⎢⎣ sin θ ⎡cos θ − sin θ ⎤ =⎢ 0 ⎥⎦ ⎣ 0 ⎡cos θ − sin θ ⎤ ⎡ 1 0 ⎤ [T2 ][T1 ] = ⎢ cos θ ⎥⎦ ⎢⎣0 0 ⎥⎦ ⎣ sin θ ⎡cos θ 0 ⎤ =⎢ ⎣ sin θ 0 ⎥⎦ T1 T2 ≠ T2 T1 11.SSM: Elementary Linear Algebra Section 4. 1 2 1 ⎡1 −2 ⎤ ⎡ 3 − 3 ⎤ ⎢ ⎥ [T −1 ] = ⎢ = 1⎥ 3 ⎣1 1⎥⎦ ⎢ 1 3⎦ ⎣3 2 1 1 ⎞ ⎛1 T −1 ( w1 . ⎡ 1 2⎤ 13. (d) One-to-one because the determinant is k 2 ≠ 0. (c) One-to-one because the determinant is −1. (a) [T ] = ⎢ ⎣ −1 1⎥⎦ det[T] = 1 + 2 = 3 ≠ 0. ⎡ 1 0⎤ ⎡ cos θ − sin θ ⎤ and [T2 ] = ⎢ . so T is one-to-one. w1 + w2 ⎟ 3 3 3 ⎠ ⎝3 109 .

Chapter 4: General Vector Spaces

SSM: Elementary Linear Algebra

⎡ 4 −6 ⎤
(b) [T ] = ⎢
3⎥⎦
⎣ −2
det[T] = 12 − 12 = 0, so T is not one-to-one.
⎡ 0 −1⎤
(c) [T ] = ⎢
⎣ −1 0 ⎥⎦
det[T] = 0 − 1 = −1 ≠ 0, so T is one-to-one.
1 ⎡0 1⎤ ⎡ 0 −1⎤
[T −1 ] =
=
−1 ⎢⎣ 1 0 ⎥⎦ ⎢⎣ −1 0 ⎥⎦
T −1 ( w1 , w2 ) = (− w2 , − w1 )

⎡ 3 0⎤
(d) [T ] = ⎢
⎣ −5 0 ⎥⎦
det[T] = 3(0) − (−5)(0) = 0, so T is not one-to-one.
15. (a) The inverse is (another) reflection about the x-axis in R 2 .
(b) The inverse is rotation through the angle −

(c) The inverse is contraction by a factor of

π
in R 2 .
4

1
in R 2 .
3

(d) The inverse is (another) reflection about the yz-plane in R3 .
(e) The inverse is dilation by a factor of 5 in R3 .
17. Let u = ( x1 , y1 ), v = ( x2 , y2 ), and k be a scalar.
(a) T (u + v) = T ( x1 + x2 , y1 + y2 )
= (2( x1 + x2 ) + ( y1 + y2 ), ( x1 + x2 ) − ( y1 + y2 ))
= ((2 x1 + y1 ) + (2 x2 + y2 ), ( x1 − y1 ) + ( x2 − y2 ))
= T (u ) + T ( v )
T (ku) = T (kx1 , ky1 )
= (2kx1 + ky1, kx1 − ky1 )
= k (2 x1 + y1 , x1 − y1 )
= kT (u)
T is a matrix operator.

(b) T (ku) = T (kx1 , ky1 )
= (kx1 + 1, ky1 )
≠ k ( x1 + 1, y1 )
T is not a matrix operator, which can also be shown by considering T(u + v).

110

SSM: Elementary Linear Algebra

Section 4.10

(c) T (u + v) = T ( x1 + x2 , y1 + y2 )
= ( y1 + y2 , y1 + y2 )
= T (u) + T ( v)
T (ku) = T (kx1 , ky1 )
= (ky1 , ky1 )
= k ( y1 , y1 )
= kT (u)
T is a matrix operator.
(d) T (u + v) = T ( x1 + x2 , y1 + y2 )

( 3 x1 + x2 , 3 y1 + y2 )
≠ ( 3 x1 + 3 x2 , 3 y1 + 3 y2 )
=

T is not a matrix operator, which can also be shown by considering T(ku).
19. Let u = ( x1 , y1 , z1 ), v = ( x2 , y2 , z2 ), and k be a scalar.
(a) T (u + v) = T ( x1 + x2 , y1 + y2 , z1 + z2 )
= (0, 0)
= T (u) + T ( v)
T (ku) = T (kx1, ky1, kz1 )
= (0, 0)
= k (0, 0)
= kT (u)
T is a matrix transformation.
(b) T (u + v) = T ( x1 + x2 , y1 + y2 , z1 + z2 )
= (3( x1 + x2 ) − 4( y1 + y2 ), 2( x1 + x2 ) − 5( z1 + z2 ))
= ((3x1 − 4 y1 ) + (3x2 − 4 y2 ), (2 x1 − 5 z1 ) + (2 x2 − 5 z2 ))
= T (u ) + T ( v )
T (ku) = T (kx1 , ky1 , kz1 )
= (3kx1 − 4ky1 , 2kx1 − 5kz1 )
= k (3x1 − 4 y1, 2 x1 − 5 z1 )
= kT (u)
T is a matrix operator.

⎡ −1 0 ⎤ ⎡ 1 0 ⎤ ⎡ −1 0 ⎤
21. (a) ⎢
=
⎣ 0 1⎥⎦ ⎢⎣0 0 ⎥⎦ ⎢⎣ 0 0 ⎥⎦
⎡ 1 0 ⎤ ⎡0 1⎤ ⎡ 0 1⎤
(b) ⎢
=
⎣0 −1⎥⎦ ⎢⎣ 1 0 ⎥⎦ ⎢⎣ −1 0 ⎥⎦
⎡0 0 ⎤ ⎡0 1⎤ ⎡ 3 0 ⎤ ⎡ 0
(c) ⎢
=
⎣0 1⎥⎦ ⎢⎣ 1 0 ⎥⎦ ⎢⎣0 3⎥⎦ ⎢⎣ 1
= ⎡0
⎢⎣ 3

0⎤ ⎡ 3 0⎤
0 ⎥⎦ ⎢⎣ 0 3⎥⎦
0⎤
0⎥⎦

23. (a) TA (e1 ) = (−1, 2, 4), TA (e2 ) = (3, 1, 5), and TA (e3 ) = (0, 2, − 3)

111

Chapter 4: General Vector Spaces

SSM: Elementary Linear Algebra

(b) TA (e1 + e2 + e3 )
= (−1 + 3 + 0, 2 + 1 + 2, 4 + 5 − 3)
= (2, 5, 6)

(b) True; use c1 = c2 = 1 to get the additivity

(c) TA (7e3 ) = 7(0, 2, − 3) = (0, 14, − 21)

(c) True; if T is a one-to-one matrix transformation
with matrix A, then Ax = 0 has only the trivial
solution. That is T(x − y) = 0 ⇒ A(x − y) = 0
⇒ x = y, so there are no such distinct vectors.

property and c1 = k , c2 = 0 to get the
homogeneity property.

25. (a) Yes; if u and v are distinct vectors in the
domain of T1 which is one-to-one, then
T1 (u) and T1 ( v) are distinct. If T1 (u) and

(d) False; the zero transformation is one example.

T1 ( v) are in the domain of T2 which is also
one-to-one, then T2 (T1 (u)) and T2 (T1 ( v))

(e) False; the zero transformation is one example.

are distinct. Thus T2 T1 is one-to-one.

(f) False; the zero transformation is one example.

(b) Yes; if T1 ( x, y) = ( x, y, 0) and

Section 4.11

T2 ( x, y, z ) = ( x, y), then T1 is one-to-one

Exercise Set 4.11

but T2 is not one-to-one and the

1. (a) Reflection about the line y = −x maps (1, 0)
to (0, −1) and (0, 1) to (−1, 0), so the
⎡ 0 −1⎤
standard matrix is ⎢
.
⎣ −1 0 ⎥⎦

composition T2 T1 is the identity
transformation on R 2 which is one-to-one.
However, when T1 is not one-to-one and T2
is one-to-one, the composition T2 T1 is not
one-to-one, since T1 (u) = T1 ( v) for some

(b) Reflection through the origin maps (1, 0) to
(−1, 0) and (0, 1) to (0, −1), so the standard
⎡ −1 0 ⎤
matrix is ⎢
.
⎣ 0 −1⎥⎦

distinct vectors u and v in the domain of T1
and consequently T2 (T1 (u)) = T2 (T1 ( v)) for
those same vectors.
27. (a) The product of any m × n matrix and the
n × 1 matrix of all zeros (the zero vector in

(c) Orthogonal projection on the x-axis maps
(1, 0) to (1, 0) and (0, 1) to (0, 0), so the
⎡ 1 0⎤
standard matrix is ⎢
.
⎣0 0 ⎥⎦

R n ) is the m × 1 matrix of all zero, which is
the zero vector in R m .

(d) Orthogonal projection on the y-axis maps
(1, 0) to (0, 0) and (0, 1) to (0, 1), so the
⎡0 0 ⎤
standard matrix is ⎢
.
⎣0 1⎥⎦

(b) One example is T ( x1, x2 ) = ( x12 + x22 , x1 x2 )
29. (a) The range of T must be a proper subset of
R n , i.e., not all of R n . Since det(A) = 0,

3. (a) A reflection through the xy-plane maps
(1, 0, 0) to (1, 0, 0), (0, 1, 0) to (0, 1, 0), and
(0, 0, 1) to (0, 0, −1), so the standard matrix
⎡ 1 0 0⎤
is ⎢0 1 0 ⎥ .

⎣0 0 −1⎦

there is some b1 in R n such that Ax = b1
is inconsistent, hence b1 is not in the range
of T.
(b) T must map infinitely many vectors to 0.
True/False 4.10
2

(b) A reflection through the xz-plane maps
(1, 0, 0) to (1, 0, 0), (0, 1, 0) to (0, −1, 0),
and (0, 0, 1) to (0, 0, 1), so the standard
⎡ 1 0 0⎤
matrix is ⎢0 −1 0 ⎥ .

⎣0 0 1⎦

2

(a) False; for example T: R → R given by
T ( x1, x2 ) = ( x12 + x22 , x1 x2 ) is not a matrix
transformation, although
T(0) = T(0, 0) = (0, 0) = 0.
112

SSM: Elementary Linear Algebra

Section 4.11

(b) The geometric effect is expansion by a
factor of 5 in the y-direction and reflection

(c) A reflection through the yz-plane maps
(1, 0, 0) to (−1, 0, 0), (0, 1, 0) to (0, 1, 0),
and (0, 0, 1) to (0, 0, 1), so the standard
⎡ −1 0 0 ⎤
matrix is ⎢ 0 1 0 ⎥ .

⎣ 0 0 1⎦

(c) The geometric effect is shearing by a factor
of 4 in the x-direction.

⎡ 1 0⎤ ⎡ 12 0 ⎤ ⎡ 12 0 ⎤
13. (a) ⎢

⎥=⎢

⎣0 5⎥⎦ ⎣ 0 1⎦ ⎣ 0 5⎦

5. (a) Rotation of 90° about the z-axis maps
(1, 0, 0) to (0, 1, 0), (0, 1, 0) to (−1, 0, 0),
and (0, 0, 1) to (0, 0, 1), so the standard
⎡0 −1 0 ⎤
matrix is ⎢ 1 0 0 ⎥ .

⎣0 0 1⎦

⎡ 1 0⎤ ⎡ 1 0⎤ ⎡ 1 0⎤
=
(b) ⎢
⎣ 2 1⎥⎦ ⎢⎣ 0 5⎥⎦ ⎢⎣ 2 5⎥⎦
⎡cos180° − sin 180°⎤ ⎡ 0 1⎤
(c) ⎢
⎣ sin 180° cos180°⎥⎦ ⎢⎣ 1 0 ⎥⎦
⎡ −1 0⎤ ⎡ 0 1⎤
=⎢
⎣ 0 −1⎥⎦ ⎢⎣ 1 0 ⎥⎦
⎡ 0 −1⎤
=⎢
⎣ −1 0⎥⎦

(b) Rotation of 90° about the x-axis maps
(1, 0, 0) to (1, 0, 0), (0, 1, 0) to (0, 0, 1), and
(0, 0, 1) to (0, −1, 0), so the standard matrix
⎡ 1 0 0⎤
is ⎢0 0 −1⎥ .

⎣0 1 0 ⎦

−1

⎡0 1⎤
⎡0 1⎤
, the inverse
=⎢
15. (a) Since ⎢

1
0

⎣ 1 0 ⎥⎦
transformation is reflection about y = x.

(c) Rotation of 90° about the y-axis maps
(1, 0, 0) to (0, 0, −1), (0, 1, 0) to (0, 1, 0),
and (0, 0, 1) to (1, 0, 0), so the standard
⎡ 0 0 1⎤
matrix is ⎢ 0 1 0 ⎥ .

⎣ −1 0 0 ⎦

⎡k 0⎤
(b) Since for 0 < k < 1, ⎢
⎣ 0 1⎥⎦

−1

⎡ 1 0⎤
= ⎢k
⎥ and
⎣ 0 1⎦

−1 ⎡ 1 0 ⎤
⎡ 1 0⎤
⎢⎣0 k ⎥⎦ = ⎢ 0 1 ⎥ , the inverse
k⎦

transformation is an expansion along the
same axis.

⎡ − 3 0 ⎤ ⎡ 0 ⎤ ⎡ 0 ⎤ ⎡ − 3 0 ⎤ ⎡ 1 ⎤ ⎡ − 3⎤
7. ⎢
=
,
,
=
⎣ 0 1⎥⎦ ⎢⎣0 ⎥⎦ ⎢⎣0 ⎥⎦ ⎢⎣ 0 1⎥⎦ ⎢⎣ 0 ⎥⎦ ⎢⎣ 0 ⎥⎦
⎡ −3 0 ⎤ ⎡0 ⎤ ⎡0 ⎤ ⎡ −3 0 ⎤ ⎡1⎤ ⎡ −3⎤
⎢⎣ 0 1⎥⎦ ⎢⎣1 ⎥⎦ = ⎢⎣1 ⎥⎦ , ⎢⎣ 0 1⎥⎦ ⎢⎣1⎥⎦ = ⎢⎣ 1⎥⎦
The image is a rectangle with vertices at (0, 0),
(−3, 0), (0, 1), and (−3, 1).

⎡ −1 0 ⎤
(c) Since ⎢
⎣ 0 1⎥⎦

y

−1

⎡ −1 0 ⎤
=⎢
and
⎣ 0 1⎥⎦

−1

⎡ 1 0⎤
⎡ 1 0⎤
⎢⎣0 −1⎥⎦ = ⎢⎣ 0 −1⎥⎦ the inverse
transformation is a reflection about the same
coordinate axis.

x

9. (a) Shear by a factor of k = 4 in the y-direction
⎡ 1 0⎤
has matrix ⎢
.
⎣ 4 1⎥⎦

⎡1 k⎤
(d) Since k ≠ 0, ⎢
⎣0 1⎥⎦
−1

−1

⎡ 1 −k ⎤
=⎢
and
1⎥⎦
⎣0

⎡ 1 0⎤
⎡ 1 0⎤
⎢⎣ k 1⎥⎦ = ⎢⎣ −k 1⎥⎦ , the inverse
transformation is a shear (in the opposite
direction) along the same axis.

(b) Shear by a factor of k = −2 in the x-direction
⎡ 1 −2 ⎤
has matrix ⎢
.
1⎥⎦
⎣0
11. (a) The geometric effect is expansion by a
factor of 3 in the x-direction.
113

Chapter 4: General Vector Spaces

SSM: Elementary Linear Algebra

⎡ t⎤
17. The line y = 2x can be represented by ⎢ ⎥ .
⎣ 2t ⎦

23. (a) To map (x, y, z) to (x + kz, y + kz, z), the
⎡1 0 k⎤
standard matrix is ⎢0 1 k ⎥ .

⎣0 0 1⎦

⎡ 1 3⎤ ⎡ t ⎤ ⎡ t + 6t ⎤ ⎡ 7t ⎤
=
=
(a) ⎢
⎣0 1⎥⎦ ⎢⎣ 2t ⎥⎦ ⎢⎣ 0 + 2t ⎥⎦ ⎢⎣ 2t ⎥⎦
The image is x = 7t, y = 2t, and using
1
2
t = x, this is the line y = x.
7
7

(b) Shear in the xz-direction with factor k maps
(x, y, z) to (x + ky, y, z + ky). The standard
⎡ 1 k 0⎤
matrix is ⎢0 1 0 ⎥ .

⎣0 k 1⎦
Shear in the yz-direction with factor k maps
(x, y, z) to (x, y + kx, z + kx).
⎡ 1 0 0⎤
The standard matrix is ⎢ k 1 0⎥ .

⎣ k 0 1⎦

⎡ 1 0 ⎤ ⎡ t ⎤ ⎡t ⎤
(b) ⎢
1⎥⎢ ⎥ = ⎢ ⎥
⎣0 2 ⎦ ⎣ 2t ⎦ ⎣t ⎦
The image is x = t, y = t which is the line
y = x.
⎡0 1⎤ ⎡ t ⎤ ⎡ 2t ⎤
(c) ⎢
=
⎣ 1 0⎥⎦ ⎢⎣ 2t ⎥⎦ ⎢⎣ t ⎥⎦
The image is x = 2t, y = t, which is the line
1
y = x.
2

True/False 4.11
(a) False; the image will be a parallelogram, but it
will not necessarily be a square.
(b) True; an invertible matrix operator can be
expressed as a product of elementary matrices,
each of which has geometric effect among those
listed.

⎡ −1 0 ⎤ ⎡ t ⎤ ⎡ −t ⎤
(d) ⎢
=
⎣ 0 1⎥⎦ ⎢⎣ 2t ⎥⎦ ⎢⎣ 2t ⎥⎦
The image is x = −t, y = 2t, which is the line
y = −2x.

(c) True; by Theorem 4.11.3.

3⎤
⎡1
⎡cos 60° − sin 60°⎤ ⎡ t ⎤ ⎢ 2 − 2 ⎥ ⎡ t ⎤
(e) ⎢
=
1 ⎥ ⎢⎣ 2t ⎥⎦
⎣ sin 60° cos 60°⎥⎦ ⎢⎣ 2t ⎥⎦ ⎢ 3
2 ⎦
⎣2
1− 2 3 ⎤

t
=
⎢ 2 ⎥
⎢ 3+2 t ⎥
⎣ 2 ⎦

⎡ 1 0⎤
(d) True; ⎢
⎣0 −1⎥⎦
⎡ −1 0 ⎤
⎢⎣ 0 1⎥⎦

3+2
1− 2 3

x or y = −

⎡ 1 0⎤
=⎢
,
⎣0 −1⎥⎦

⎡ −1 0 ⎤
⎡0 1⎤
=⎢
, and ⎢

0
1
⎣ 1 0⎥⎦

−1

⎡0 1⎤
=⎢
.
⎣ 1 0 ⎥⎦

⎡1 1⎤
(e) False; ⎢
maps the unit square to the
⎣1 −1⎥⎦

1− 2 3
3+2
t, y =
t,
2
2
2
x, this is the line
and using t =
1− 2 3
The image is x =

y=

−1

−1

parallelogram with vertices (0, 0), (1, 1), (1, −1),
and (2, 0) which is not a reflection about a line.

⎡ 1 −2 ⎤
(f) False; ⎢
maps the unit square to the
1⎥⎦
⎣2
parallelogram with vertices (0, 0), (1, 2), (−2, 1),
and (−1, 3) which is not a shear in either the xor y-direction.

8+5 3
x.
11

⎡ 3 1⎤ ⎡ x ⎤ ⎡ 3x + y ⎤ ⎡ x′ ⎤
19. (a) ⎢
=
=
⎣6 2 ⎥⎦ ⎢⎣ y ⎥⎦ ⎢⎣6 x + 2 y ⎥⎦ ⎢⎣ y ′⎥⎦
Thus, the image ( x′, y ′) satisfies y ′ = 2 x′
for every (x, y) in the plane.

(g) True; this is an expansion by a factor of 3 in the
y-direction.

(b) No, since the matrix is not invertible.

Section 4.12
Exercise Set 4.12
1. (a) A is a stochastic matrix.
114

9 . The system (I − P)q = 0 is ⎡ 1 −1 0⎤ ⎡ q ⎤ 2 ⎡ 0⎤ ⎢ 2 ⎥ 1 1 − 1 ⎥ ⎢q ⎥ = ⎢ ⎥ . ⎡0.55⎤ = 0. (d) A is not a stochastic matrix because the third column is not a probability vector due to the negative entry.1⎤ ⎡1 ⎤ ⎡0. Since q 3 3 must be a probability vector.5 0.1⎤ ⎡0. Each column of P is a probability vector and ⎡3 1 1⎤ ⎢8 2 6⎥ 7 ⎥ .85.9 ⎥⎦ ⎢⎣0 ⎥⎦ ⎢⎣ 0. so P is a regular stochastic P 2 = ⎢ 13 83 18 ⎢7 1 4⎥ ⎢⎣ 24 8 9 ⎥⎦ matrix. ( ) ( ) (c) P is a regular stochastic matrix since ⎡ 21 1 ⎤ P 2 = ⎢ 25 5 ⎥ .5⎥⎦ ⎢⎣0.5 0.5⎤ ⎡ 0. so the general solution of the ⎢ ⎥ 0⎦ ⎣0 0 4 4 system is q1 = s. The system (I − P)q = 0 is ⎡ 3 − 2⎤ ⎡ q ⎤ 3 ⎥ 1 = ⎡0⎤ .8 ⎥⎦ The probability is 0.455⎥⎦ ⎡0. ⎡0.545⎤ = 0. q3 = s.45⎥⎦ ⎢⎣ 0.5455 0. The steady-state vector is q = ⎢ 4 3 ⎥ = ⎢ 11 ⎢ 3 11 ⎥ ⎢ ⎥ 3 ⎢ 3 ⎥ ⎣⎢ 11 ⎦⎥ ⎣ 11 ⎦ 0. ⎢ 9 ⎥ ⎢9⎥ ⎣ 17 ⎦ ⎣ 17 ⎦ so s = ( ) (c) A is a stochastic matrix.6 ⎤ ⎡0. q2 = s. so s = .54545⎤ x 4 = Px 3 = ⎢ = ⎣0. (b) P is not a regular stochastic matrix. 11 3 1 = q1 + q2 + q3 = s.5 ⎡ 0. 115 .4545⎥⎦ ⎡0.4 ⎥⎦ ⎢⎣0. x1 = Px0 = ⎢ ⎣0. The steady-state vector is 17 ⎡8 9 ⎤ ⎡ 8 ⎤ q = ⎢ 9 17 ⎥ = ⎢ 17 ⎥ .5 3. since the entry in row 1. q2 = s. Since q 9 17 must be a probability vector 1 = q1 + q2 = s 9 11.5454 ⎤ ⎡0. 11 3 ⎡4 3 ⎤ ⎡ 4 ⎤ ⎢ 3 11 ⎥ ⎢ 11 ⎥ 4 ⎥.45455⎥⎦ ⎡ 0.45455⎥⎦ 5.4 ⎥⎦ ⎢⎣0.4546 ⎥⎦ ⎢⎣0. (b) The entry 0. ⎢ 4 3 2 ⎥ ⎢ q ⎥ ⎢⎣ 0 ⎥⎦ ⎢− ⎣ 2⎦ 3⎦ ⎣ 4 ⎡ 3 − 2 0⎤ ⎡ 1 − 8 0⎤ 3 9 ⎢ 4 ⎥ reduces to ⎢ ⎥ so the 2 0⎥ ⎢ − 34 0 0 0 ⎢ ⎣ ⎦⎥ 3 ⎣ ⎦ 8 general solution is q1 = s.4545⎥⎦ ⎢⎣ 0.5 0. P is a regular stochastic matrix.1 = p12 represents the probability that something in state 2 moves to state 1. 4 4⎥ ⎢⎣ 25 5⎦ 7.6 ⎤ ⎡ 0. column 2 of P k will be 0 for all k ≥ 1.5⎤ x 4 = P 4 x0 = ⎢ ⎣ 0.9 ⎥⎦ ⎢⎣0. (a) The entry 0.SSM: Elementary Linear Algebra Section 4.12 (b) A is not a stochastic matrix since the sums of the entries in each column are not 1.54545⎤ =⎢ ⎣ 0.5455⎤ ⎡ 0. (a) P is a regular stochastic matrix.6 ⎤ ⎡0.5 x 2 = Px1 = ⎢ ⎣ 0.8.45⎥⎦ 0.4545 0.8 0.5⎤ ⎡0.2 0. ⎢ − 14 0 2 2 3 ⎢ ⎥ ⎢ ⎥ ⎢ 1 ⎥ q 0⎦ 1 ⎣ ⎣ 3⎦ 0 ⎢⎣ − 4 3 ⎥⎦ The reduced row echelon form of I − P is ⎡1 0 − 4⎤ 3⎥ ⎢ ⎢0 1 − 43 ⎥ .2 0.6 ⎤ ⎡0.8 0.455⎥⎦ ⎢⎣0.5⎥⎦ ⎡ 0.85⎥⎦ The probability is 0.5⎥⎦ ⎢⎣ 0.2 = p11 represents the probability that something in state 1 stays in state 1.55⎤ ⎡ 0.2 ⎤ (c) ⎢ = ⎣ 0.5455⎤ x 3 = Px 2 = ⎢ = ⎣0.4 ⎥⎦ ⎢⎣ 0.545⎤ ⎡0.5 0.5 9.4 ⎥⎦ ⎢⎣0. Since P is a stochastic matrix with all positive entries.15⎤ (d) ⎢ = ⎣ 0. ⎡0.

933 81.000) = 46.55⎤ ⎡0.05 0.625 ⎥⎦ ⎢⎣0.97 ⎥⎦ ⎡100 ⎤ x0 = ⎢ ⎥ (in thousands). s = = 0.97 ⎥⎦ ⎢⎣ 25⎥⎦ ⎢⎣33.97 ⎥⎦ ⎢⎣ 25⎥⎦ ⎢⎣ 40.000) = 78. ⎣ 25⎦ ⎡0.95 0.55⎤ 13.03⎤ ⎡ q1 ⎤ ⎡0 ⎤ .858 ⎤ (c) P3 ⎢ ⎥ = ⎢ = ⎣1 ⎦ ⎣ 0.07 0.243 84.142 that the air will be bad.95 0.078 0.05 0.8 ⎥⎦ ⎢⎣0.125 in the suburbs.03⎤ ⎡100⎤ ⎡91.757 40.375(125. ⎣ 0.000 will be distributed with 0.6(0.067 43.111 ⎡ 0.05 0.889⎤ P5 x0 = ⎢ ≈ ⎣0.77 ⎤ ⎡1 ⎤ ⎡ 0.757 ⎥⎦ 4 ⎡0.23⎥⎦ ⎢⎣0 ⎥⎦ ⎢⎣0.45⎥⎦ ⎢⎣ 0. (b) The system (I − P)q = 0 is ⎢ ⎢ ⎥= ⎣ −0.375⎤ 1 = q1 + q2 = 1. ⎡0.63 that the air will be good.63⎤ (d) P ⎢ ⎥ = ⎢ = ⎣ 0.250 33. The transition matrix is P = ⎢ .05 −0.8⎦ ⎣0.03⎥⎦ ⎣ q2 ⎦ ⎢⎣0 ⎥⎦ ⎡1 −0.95 0.6 ⎤ so the general solution is q1 = 0.03⎤ ⎡100⎤ ⎡ 84.05 0.922 0. (a) Let state 1 be good air.95 0.858⎤ ⎡ 0 ⎤ ⎡ 0.2 ⎤ ⎡ 0.142 ⎥⎦ The probability is 0.6 ⎣ 0.142 ⎥⎦ ⎢⎣1 ⎥⎦ ⎢⎣0.Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra ⎡ 0.05 0.933 ⎤ P 4 x0 = ⎢ ≈ ⎣0.07 ⎥⎦ The probability is 0.625⎥⎦ Over the long term.625 and the steady-state vector is q = ⎢ = .37 ⎥⎦ The probability is 0.875 in the city and 0.25⎥⎦ 2 ⎡0.6s.95 0.93 0.2 ⎤ ⎡0.05 0.16 ⎥⎦ 3 ⎡0.93⎤ = (b) P 2 ⎢ ⎥ = ⎢ ⎣ 0⎦ ⎣ 0.93 that the air will be good.889 Suburbs 29.97 ⎥⎦ ⎢⎣ 25⎥⎦ ⎢⎣ 29. The initial state vector is ⎣ 0.03⎤ ⎡100⎤ ⎡81.750 91. Then the transition matrix is P = ⎢ .84 ⎤ P 2 x0 = ⎢ = ⎣0.95 0.625) ⎤ ⎡0.111⎥⎦ Year 1 2 3 4 5 City 95. ⎡ 0.160 36.03⎤ ⎡100 ⎤ ⎡95.97 ⎥⎦ ⎢⎣ 25⎥⎦ ⎢⎣ 43.05 0.05 0.03⎤ ⎡100 ⎤ ⎡ 88.840 88. 116 . q2 = s. Since The reduced row echelon form of I − P is ⎢ 0 ⎥⎦ ⎣0 1 ⎡ 0.45⎥⎦ ⎡1 ⎤ ⎡ 0.95 0.05 0. 1.067 ⎥⎦ 5 ⎡0.97 ⎥⎦ ⎢⎣ 25⎥⎦ ⎢⎣36. ⎡0 ⎤ ⎡ 0.243 ⎤ P3x0 = ⎢ ≈ ⎣0.95 0.75 ⎤ = (a) Px0 = ⎢ ⎣0. Let state 1 be living in the city. the population of 125.625(125.6s.03⎤ 15.

2 1⎤ ⎡0.2893⎤ ⎡35⎤ (c) 120q ≈ 120 ⎢ 0. −2 + 5. ⎢ 47 159 ⎥ ⎢ 53 ⎥ ⎢ 47 ⎥ ⎢ 47 ⎥ ⎣ 159 ⎦ ⎣ 159 ⎦ q2 = s. respectively. so the general solution is 47 ⎢ ⎥ 0⎦ ⎣0 0 46 66 q1 = s .16 0. q3 = s.4151. =⎢ and ⎢ ⎥ 0 . (b) True.SSM: Elementary Linear Algebra ⎡1 ⎢ 10 17. 0. this is part of the definition of a transition matrix. 100 (b) The system (I − P)q = 0 is ⎡ 9 − 1 − 3⎤ 5 5 ⎥ ⎡ q1 ⎤ ⎡ 0 ⎤ ⎢ 10 7 − 1 ⎥ ⎢ q ⎥ = ⎢0 ⎥ . Column 1: 1 − − = 10 10 5 3 3 1 Column 2: 1 − − = 10 5 10 1. (e) True Chapter 4 Supplementary Exercises 7 1 1 19.4151⎥ ≈ ⎢50⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 0. the steady-state vector is a solution which is also a probability vector. and 35 parking spaces. hence P k q = P k −1 ( Pq) = P k −1q = P k −2 ( Pq) = =q and P k q = q for every positive integer k. 8 0 ⎣ ⎦ ⎣0.84 0. 47 47 159 47 Since 1 = q1 + q2 + q3 = s. and location 3 is about 0. Since q is the steady-state vector for P. so the general solution is q1 = s.2893. ⎢ −4 5 ⎢ 2⎥ ⎢ ⎥ ⎢ 15 101 ⎥ q 4 ⎣ 3 ⎦ ⎣0 ⎦ ⎢⎣ − 10 − 2 5 ⎥⎦ The reduced row echelon form of I − P is ⎡ 1 0 − 46 ⎤ 47 ⎥ ⎢ ⎢0 1 − 66 ⎥ . (d) False.2 ⎤ . The locations should have 35. 0) = ( −3. ⎢ − 15 0 10 2 ⎢ 2⎥ ⎢ ⎥ ⎢ 1 ⎥ q 0 3 7 ⎢⎣ − 10 − 5 10 ⎥⎦ ⎣ 3 ⎦ ⎣ ⎦ The reduced row echelon form of I − P is ⎡ 1 0 −1⎤ ⎢0 1 −1⎥ . ⎢ ⎥ ⎣0 0 0 ⎦ ⎡ 23 ⎤ ⎢ 100 ⎥ = ⎢ 17 ⎥ ⎢ 50 ⎥ 43 ⎢⎣ 100 ⎥⎦ 23 . (c) True. 5 ⎥ 1 5 ⎦⎥ 1 3 1 Column 3: 1 − − = 5 10 2 ⎡7 1 1⎤ ⎢ 10 10 5 ⎥ 3 1⎥ P = ⎢ 15 10 ⎢ 1 3 32 ⎥ ⎢⎣ 10 5 10 ⎥⎦ The system (I − P)q = 0 is ⎡ 3 − 1 − 1⎤ 10 5 ⎥ ⎡ q1 ⎤ ⎡ 0 ⎤ ⎢ 10 7 − 1 ⎥ ⎢q ⎥ = ⎢ ⎥ . 50. the column vectors are probability vectors 2 ⎡0. ( ) ( ) True/False 4.2956. (a) u + v = (3 + 1. 2) (−1)u = ((−1)3. 3 ⎢1 ⎥ ⎢⎣ 3 ⎥⎦ 21. Pq = q.2956 ⎦ ⎣35⎦ Over the long term.8⎥⎦ ⎡ 0. 4 + (−2)) = (4. Thus the vector 120q gives the long term distribution of 120 cars. the probability that a car ends up at location 1 is about 0.12 (a) True. Since 1 = q1 + q2 + q3 = 3s. The transition matrix is P = ⎢ 4 ⎢ 51 ⎣⎢ 10 ⎡ 23 ⎡1 ⎤ ⎢ 100 (a) P 2 ⎢ 0⎥ = ⎢ 17 ⎢ ⎥ ⎢ 50 ⎣ 0⎦ ⎢ 43 ⎣ 100 19 50 7 20 27 100 The probability is 11 ⎤ 50 ⎥ ⎡1 ⎤ 29 ⎥ ⎢ ⎥ 0 50 ⎢ ⎥ ⎥ 1 ⎣0 ⎦ 5 ⎥⎦ Chapter 4 Supplementary Exercises 1 5 3 10 1 2 3⎤ 5⎥ 1⎥. 3. 0. ⎡1 ⎤ ⎢3⎥ 1 s = and the steady-state vector is q = ⎢ 13 ⎥ . location 2 is about 0. 0) 117 . q3 = s. s = and 47 159 the steady-state vector is ⎡ 46 47 ⎤ ⎡ 46 ⎤ ⎢ 47 159 ⎥ ⎢ 159 ⎥ q = ⎢ 66 47 ⎥ = ⎢ 22 ⎥. the sum of the entries is 1 and all entries are nonnegative. q2 = s.

so the solution space ⎢ ⎥ ⎣0 0 0 ⎦ will require 1 parameter. 0) = (kmu1 . (b) If p1 and p2 are in the set. u3 ) with u2 . 0) + (mu1 . A is invertible and the solution space is the origin. … . 7. which reduced to ⎢ ⎥ ⎣1 1 1⎦ ⎡ 1 1 1⎤ ⎢0 0 0 ⎥ so the solution space will require 2 ⎢ ⎥ ⎣0 0 0 ⎦ parameters. (c) The row vectors ri will satisfy ⎧r if i is odd . is {1. and the set is a subspace of Pn . … .. 11. so p1 + p2 is in the set. Thus. ⎡1 ⎢0 (b) A = ⎢ ⎢1 ⎣⎢ 0 (c) Axioms 1−5 hold in V because they hold in R3 .e. i. A must be an invertible matrix for Av1 . then (e) For any u = (u1 . Its rank will be 2 and its nullity will be n − 2. 0. (a) If p1 and p2 are in the set. A basis is {x. (kp1 )(0) = k ( p1 (0)) = k (0) = 0.Chapter 4: General Vector Spaces SSM: Elementary Linear Algebra ⎡ 1 0 1⎤ ⎡ 1 0 1⎤ 9. u3 + v3 ) = (k (u1 + v1 ). 0. then A = ⎢ 1 −2 1⎥ .. u2 . Also (kp1 )(− x) = k ( p1 (− x)) = k ( p1 ( x)) = (kp1 )( x) so kp1 is in the set for any scalar k. x 2 . u2 + v2 . so for s ≠ 1. x3 . 0) = ku + kv Axiom 8: (k + m)u = ((k + m)u1 . ⎡1 1 s⎤ 3. 0) = (ku1. so kp1 is in the set. ⎡1 1 1⎤ If s = 1. it will be a line through the origin. which ⎢ ⎥ 1 1⎦ ⎣ −2 ⎡ 1 0 −1⎤ reduces to ⎢0 1 −1⎥ . then ( p1 + p2 )(0) = p1 (0) + p2 (0) = 0. −2. Also. The coefficient matrix is A = ⎢ 1 s 1⎥ and ⎢ ⎥ ⎣ s 1 1⎦ det( A) = −( s − 1) 2 ( s + 2). ( p1 + p2 )(− x) = p1 (− x) + p2 (− x) = p1 ( x) + p2 ( x) = ( p1 + p2 )( x) so the set is closed under addition. (a) A = ⎢ 0 1 0⎥ reduces to ⎢0 1 0 ⎥ . Thus.e. the reduced row ri = ⎨ 1 ⎩r2 if i is even echelon form will have 2 nonzero rows. 0) = (km)u 0 1 0 1 1 0 1 0 0⎤ 1⎥ ⎥ reduces to 0⎥ 1⎦⎥ ⎡ 1 0 1 0⎤ ⎢0 1 0 1⎥ ⎢ ⎥ . ⎡ 1 1 −2 ⎤ If s = −2. 0. so ⎢ ⎥ ⎢ ⎥ ⎣ 1 0 1⎦ ⎣0 0 0 ⎦ rank(A) = 2 and nullity(A) = 1. u3 ≠ 0. x 2 . Av n to be linearly independent. 0. u2 . 0. … . x 4 . 0) = ku + mu Axiom 9: k (mu) = k (mu1 . 0) + (kv1 . u3 ) = u. Similarly. A basis 1u = (1u1 . 0. x n }. 0. 0. 0) = (ku1 . x 2m } where 2m = n if n is even and 2m = n − 1 if n is odd. so rank(A) = 2 and ⎢0 0 0 0 ⎥ ⎣⎢0 0 0 0 ⎦⎥ nullity(A) = 2. (d) Axiom 7: k (u + v) = k (u1 + v1 . V is closed under scalar multiplication because the real numbers are closed under multiplication. 0. the subspace consisting of polynomials with constant term 0. Av 2 . 0) ≠ (u1. the subspace consisting of all even polynomials. 118 . (b) V is closed under addition because addition in V is component-wise addition of real numbers and the real numbers are closed under addition. i. then A = ⎢1 1 1⎥ . it is a plane through the origin. the set is a subspace of Pn .

i. below the main diagonal is determined by the a ji entry above the main diagonal. so the rank can be at most 2. the matrix will have rank 1. 4 × 4. the possible ranks are 0. ⎢ 0 0 0 ⎥ . ⎡ 0 ai 6 ⎤ det ⎢ ⎥ = −ai 6 a5 j so the rank is 2 if any of these products is nonzero. If exactly one entry in the last column is nonzero. the rank will be 0. if the matrix has 2 nonzero ⎣ a5 j a56 ⎦ entries other than a56 . i > j.SSM: Elementary Linear Algebra Chapter 4 Supplementary Exercises 13. For i < 5 and j < 6. ⎢0 0 1⎥ ⎬ . A ⎧ ⎡ 1 0 0⎤ ⎡ 0 1 0 ⎤ ⎡ 0 0 1⎤ ⎡0 0 0 ⎤ ⎡0 0 0 ⎤ ⎪ basis is ⎨ ⎢ 0 0 0⎥ .. ⎪⎩ ⎢⎣ 0 0 0⎥⎦ ⎢⎣ 0 0 0 ⎥⎦ ⎢⎣ 1 0 0 ⎥⎦ ⎢⎣0 0 0 ⎥⎦ ⎢⎣0 1 0 ⎥⎦ ⎡0 0 0 ⎤ ⎫ ⎢0 0 0 ⎥ ⎪⎬ ⎢ ⎥ ⎣0 0 1⎦ ⎪⎭ ⎧0 if i = j (b) In a skew-symmetric matrix. or 2. 1. Thus. Every 3 × 3. the entries below the main diagonal are determined ⎩−a ji if i ≠ j ⎧ ⎡ 0 1 0 ⎤ ⎡ 0 0 1⎤ ⎡0 0 0 ⎤ ⎫ ⎪ ⎪ by the entries above the main diagonal. 119 . ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎩⎪ ⎣ 0 0 0 ⎦ ⎣ −1 0 0 ⎦ ⎣0 −1 0 ⎦ ⎭⎪ 15. (a) The entry aij . and 5 × 5 submatrix will have determinant 0. A basis is ⎨ ⎢ −1 0 0 ⎥ . ⎢ 1 0 0 ⎥ . ⎢ 0 0 0 ⎥ . while if all entries are zero. ⎢0 0 1⎥ .e. aij = ⎨ . Thus. ⎢0 1 0 ⎥ .

9 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎡λ − 10 (b) ⎢ = − λ + 2 ⎥⎦ ⎢⎣ x2 ⎥⎦ ⎢⎣0 ⎥⎦ 4 ⎣ 7 ⎤ ⎡λ + 2 (d) det ⎢ = (λ + 2)(λ − 2) + 7 ⎣ −1 λ − 2 ⎥⎦ = λ2 + 3 ⎡ −6 9 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ . Ax = ⎢ 2 3 2 ⎥ ⎢ 2⎥ = ⎢10 ⎥ = 5x ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 1 0 4 ⎦ ⎣ 1 ⎦ ⎣ 5⎦ x corresponds to the eigenvalue λ = 5. ⎡λ − 1 0 ⎤ (f) det ⎢ = (λ − 1)2 = λ 2 − 2λ + 1 λ − 1⎥⎦ ⎣ 0 The characteristic equation is ⎡ λ −3⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ (c) ⎢ ⎢ ⎥= ⎣ −4 λ ⎥⎦ ⎣ x2 ⎦ ⎢⎣ 0 ⎥⎦ λ 2 − 2λ + 1 = 0. ⎥⎢ ⎥ = 12 ⎦ ⎣ x2 ⎦ ⎣⎢0 ⎦⎥ . ⎡ −4 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ . or 2 ⎡3⎤ ⎡3⎤ ⎡ x1 ⎤ ⎡ 32 s ⎤ 2 2 ⎢ x ⎥ = ⎢ ⎥ = s ⎢ ⎥ . λ = 3 gives ⎢ ⎢ ⎥= ⎣ −8 4 ⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ ⎡ 4 0 1⎤ ⎡1 ⎤ ⎡ 5⎤ 1. ⎡ λ −3⎤ (c) det ⎢ = λ 2 − 12 ⎥ − 4 λ ⎣ ⎦ The characteristic equation is λ 2 − 12 = 0. ⎡ x1 ⎤ ⎡0 ⎤ ⎡0 ⎤ ⎡0 ⎤ ⎢ ⎥ = ⎢ ⎥ = s ⎢1 ⎥ . (a) ⎢ = − λ + 1⎥⎦ ⎢⎣ x2 ⎥⎦ ⎢⎣0 ⎥⎦ 8 ⎣ Section 5. or 2 ⎡1⎤ ⎡1⎤ ⎡ x1 ⎤ ⎡ 12 s ⎤ 2 2 ⎢ x ⎥ = ⎢ ⎥ = s ⎢ ⎥ . x2 = s. general solution is x1 = The characteristic equation is λ 2 = 0. x2 = s. or λ 2 − 8λ + 16 = 0. 9 ⎤ ⎡ λ − 10 (b) det ⎢ = (λ − 10)(λ + 2) + 36 4 − λ + 2 ⎥⎦ ⎣ = λ 2 − 8λ + 16 The characteristic equation is ⎡1 0 ⎤ ⎡ −4 0 ⎤ ⎢⎣ −8 0 ⎥⎦ reduces to ⎢⎣0 0 ⎥⎦ . (a) det ⎢ ⎣ −8 λ + 1⎥⎦ = λ 2 − 2λ − 3 The characteristic equation is λ 2 − 2λ − 3 = 0. the general solution is x1 = 0.1 ⎡ 0 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ . the ⎢ 0⎦ ⎣0 3 s. x2 = s. Since λ = −1 gives ⎢ ⎢ ⎥= ⎣ −8 0 ⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ general solution is x1 = 0 ⎤ ⎡λ − 3 = (λ − 3)(λ + 1) 3. so ⎢ ⎥ is a basis for ⎣ 2⎦ ⎣ s ⎦ ⎣1⎦ ⎣1 ⎦ the eigenspace corresponding to λ = 4.Chapter 5 Eigenvalues and Eigenvectors 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎡λ − 3 5. so ⎢ ⎥ is a basis for ⎣ 2⎦ ⎣ s ⎦ ⎣1⎦ ⎣1 ⎦ the eigenspace corresponding to λ = 3. λ = 4 gives ⎢ ⎢ ⎥= ⎣ −4 6 ⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ ⎡ −6 9 ⎤ reduces to Since ⎢ ⎣ −4 6 ⎥⎦ The characteristic equation is λ 2 + 3 = 0. ⎡ 12 λ = 12 gives ⎢ ⎣ −4 120 −3 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ . so ⎢1 ⎥ is a basis for the ⎣ ⎦ ⎣ ⎦ ⎣ x2 ⎦ ⎣ s ⎦ eigenspace corresponding to λ = −1. the ⎢ 0⎦ ⎣0 1 s. ⎡λ 0 ⎤ (e) det ⎢ = λ2 ⎣ 0 λ ⎥⎦ ⎡1 − 3⎤ 2 ⎥ .1 Exercise Set 5. ⎡ 0 0⎤ reduces to Since ⎢ ⎣ −8 4 ⎥⎦ ⎡1 − 1⎤ 2 ⎥ .

so the eigenvalues are λ = −4. every vector in R 2 is a solution. ( ) are λ = 0. 0 ⎤ −2 ⎡λ 0 ⎢ −1 λ −1 0 ⎥ 9. so the only (real) eigenvalue is λ = −8.SSM: Elementary Linear Algebra Section 5. 121 .1 ⎡1 − 3 ⎤ −3 ⎤ 12 ⎥ . 2. (f) The eigenvalues satisfy λ3 − 2λ 2 − 15λ + 36 = 0 or (λ + 4)(λ − 3)2 = 0. so the eigenvalues or (λ + 8)(λ 2 + 1) = 0. the general solution is 0 ⎥⎦ 3 x1 = − s. (d) The eigenvalues satisfy λ3 − λ 2 − λ − 2 = 0 or (λ − 2)(λ 2 + λ + 1) = 0. (a) The eigenvalues satisfy ⎡ 12 Since ⎢ ⎣ −4 λ3 − 6λ 2 + 11λ − 6 = 0 or (λ − 1)(λ − 2)(λ − 3) = 0. (c) The eigenvalues satisfy λ3 + 8λ 2 + λ + 8 = 0 λ = − 12 gives ⎡ − 12 −3 ⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ . λ = 1 gives ⎢ ⎢ ⎥= ⎣0 0 ⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ Clearly. − 2 . reduces to ⎢ ⎥ 12 ⎦ ⎢⎣0 0 ⎥⎦ 3 the general solution is x1 = s. 12 7. (b) The eigenvalues satisfy λ3 − 2λ = 0 or 3 s⎤ ⎡ 3 ⎤ ⎡ 3 ⎤ ⎡x ⎤ ⎡ or ⎢ 1 ⎥ = ⎢ 12 ⎥ = s ⎢ 12 ⎥ . 3. 3 ⎤ 12 ⎥ . (d) There are no eigenspaces since there are no real eigenvalues. so the eigenvalues are λ = 1. λ = 0 gives ⎢ ⎢ ⎥= ⎣0 0 ⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ Clearly. so ⎢ 12 ⎥ is = ⎢ ⎢x ⎥ ⎣ 2 ⎦ ⎣⎢ s ⎦⎥ ⎣⎢ 1 ⎦⎥ ⎣⎢ 1 ⎦⎥ a basis for the eigenspace corresponding to λ = − 12. (a) det ⎢ ⎥ − λ + 0 1 2 0 ⎥ ⎢ λ − 1⎦⎥ 0 ⎣⎢ 0 0 ⎡λ 0⎤ ⎡ x1 ⎤ ⎡ 0⎤ (e) ⎢ ⎢ ⎥= ⎣ 0 λ ⎥⎦ ⎣ x2 ⎦ ⎢⎣ 0⎥⎦ −2 ⎤ ⎡λ 0 = (λ − 1) det ⎢ −1 λ −1 ⎥ ⎢ ⎥ ⎣ 0 −1 λ + 2⎦ = (λ − 1){λ[λ(λ + 2) − 1] − 2(1 − 0)} ⎡0 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ . so ⎡1 ⎤ ⎡ 0 ⎤ ⎢⎣0 ⎥⎦ . ⎢⎣1 ⎥⎦ is a basis for the eigenspace corresponding to λ = 1. so ⎡1 ⎤ ⎡ 0 ⎤ ⎢⎣0 ⎥⎦ . 2 . ⎢⎣1 ⎥⎦ is a basis for the eigenspace corresponding to λ = 0. so ⎢ 12 ⎥ is a ⎣ x2 ⎦ ⎢⎣ s ⎥⎦ ⎢⎣ 1 ⎥⎦ ⎢⎣ 1 ⎥⎦ basis for the eigenspace corresponding to λ = 12. or 12 (e) The eigenvalues satisfy λ3 − 6λ 2 + 12λ − 8 = 0 or (λ − 2)3 = 0. so the only eigenvalue is λ = 2. ⎡λ − 1 0 ⎤ ⎡ x1 ⎤ ⎡ 0⎤ (f) ⎢ = λ − 1⎥⎦ ⎢⎣ x2 ⎥⎦ ⎢⎣ 0⎥⎦ ⎣ 0 ⎡0 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ . so the only (real eigenvalue is λ = 2. 3. = (λ − 1)(λ3 + 2λ 2 − λ − 2) = λ 4 + λ3 − 3λ 2 − λ + 2 The characteristic equation is λ 4 + λ3 − 3λ 2 − λ + 2 = 0. every vector in R 2 is a solution. 3 s⎤ ⎡− 3 ⎤ ⎡− 3 ⎤ ⎡ x1 ⎤ ⎡ − 12 ⎥ = s ⎢ 12 ⎥ . x2 = s. x2 = s. ⎢ ⎥⎢ ⎥ = − 12 ⎦ ⎣ x2 ⎦ ⎢⎣ 0 ⎥⎦ ⎣ −4 ⎡ − 12 −3 ⎤ Since ⎢ ⎥ reduces to − 12 ⎦ ⎣ −4 ⎡1 ⎢ ⎢⎣0 )( λ λ + 2 λ − 2 = 0.

0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ −2 ⎡λ 0 ⎢ −1 λ 0 ⎥ ⎢ x2 ⎥ ⎢0 ⎥ −1 11. x3 = s. ⎢0 ⎥ ⎣⎢1 ⎦⎥ ⎡ −2 0 −2 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎢ −1 −2 −1 0 ⎥ ⎢ x2 ⎥ ⎢0 ⎥ λ = −2 gives ⎢ ⎥ ⎢ ⎥ = ⎢ ⎥. ⎢ 0 −1 0 0 ⎥ ⎢ x3 ⎥ ⎢0 ⎥ ⎣⎢ 0 0 0 −3⎦⎥ ⎢⎣ x4 ⎥⎦ ⎣⎢0 ⎦⎥ ⎡ −2 0 −2 0 ⎤ ⎢ −1 −2 −1 0 ⎥ Since ⎢ ⎥ reduces to ⎢ 0 −1 0 0 ⎥ ⎣⎢ 0 0 0 −3⎦⎥ ⎡1 ⎢0 ⎢ ⎢0 ⎣⎢0 0 1 0 0 1 0 0 0 0⎤ 0⎥ ⎥ . x s 1 0 ⎢ 3⎥ ⎢ ⎥ ⎢1 ⎥ ⎢ ⎥ ⎢ ⎥ 0 1 ⎢⎣ x4 ⎥⎦ ⎣⎢ t ⎦⎥ ⎢⎣ 0 ⎥⎦ ⎣⎢ ⎦⎥ ⎣⎢ ⎦⎥ ⎡0 ⎤ ⎢0 ⎥ ⎢ ⎥ is a basis for the eigenspace corresponding to λ = 1. the general solution is x1 = 2 s. ⎢ x3 ⎥ ⎢ s ⎥ ⎢ 1⎥ ⎢ 1⎥ ⎢⎣ x4 ⎥⎦ ⎣⎢ 0 ⎦⎥ ⎣⎢ 0⎦⎥ ⎣⎢ 0 ⎦⎥ ⎡ −1 0 −2 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎢ −1 −1 −1 0 ⎥ ⎢ x2 ⎥ ⎢0 ⎥ λ = −1 gives ⎢ ⎥ ⎢ ⎥ = ⎢ ⎥. (a) ⎢ ⎥⎢ ⎥ = ⎢ ⎥ 0 ⎥ ⎢ x3 ⎥ ⎢0 ⎥ ⎢ 0 −1 λ + 2 0 λ − 1⎥⎦ ⎢⎣ x4 ⎥⎦ ⎢⎣0 ⎥⎦ ⎢⎣ 0 0 ⎡ 1 0 −2 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎢ −1 1 −1 0 ⎥ ⎢ x2 ⎥ ⎢0 ⎥ λ = 1 gives ⎢ ⎥ ⎢ ⎥ = ⎢ ⎥. the general solution is x1 = − s. x2 = 0. 0 0 0⎥ 0 0 0 ⎦⎥ ⎡ x1 ⎤ ⎡ 2 s ⎤ ⎡ 2 ⎤ ⎡ 0⎤ ⎡2⎤ ⎢ x ⎥ ⎢ 3s ⎥ ⎢3⎥ ⎢ ⎥ ⎢ ⎥ 3 0 x4 = t or ⎢ 2 ⎥ = ⎢ ⎥ = s ⎢ ⎥ + t ⎢ ⎥ . x2 = 3s. so ⎢ ⎥ is a basis for the eigenspace corresponding to λ = −2. ⎢ 0 −1 1 0 ⎥ ⎢ x3 ⎥ ⎢0 ⎥ ⎢⎣ 0 0 0 −2 ⎥⎦ ⎢⎣ x4 ⎥⎦ ⎢⎣0 ⎥⎦ 122 .Chapter 5: Eigenvalues and Eigenvectors SSM: Elementary Linear Algebra 9 0 0 ⎤ ⎡ λ − 10 0 0 ⎤ 0 0 ⎤ ⎡λ + 2 ⎡ −4 ⎢ −4 λ+2 0 0 ⎥ ⎢ ⎥ ⎢ = λ − − 10 9 (b) det ⎢ ( ) det det λ+2 0 7 0 λ+2 7 ⎥ ⎥ 0 7 ⎥ λ+2 ⎢ ⎥ ⎢ ⎥ ⎢ 0 −1 λ − 2 ⎦ −1 λ − 2 ⎦ ⎣ 0 ⎣0 0 −1 λ − 2 ⎥⎦ ⎢⎣ 0 = (λ − 10)(λ + 2)[(λ + 2)(λ − 2) + 7] − 9(−4)[(λ + 2)(λ − 2) + 7] = (λ 2 − 8λ − 20 + 36)(λ 2 − 4 + 7) = (λ 2 − 8λ + 16)(λ 2 + 3) = λ 4 − 8λ3 + 19λ 2 − 24λ + 48 The characteristic equation is λ 4 − 8λ3 + 19λ 2 − 24λ + 48 = 0. so ⎢ ⎥ . x3 = s. 1⎥ 0 ⎦⎥ ⎡ x1 ⎤ ⎡ − s ⎤ ⎡ −1⎤ ⎡ −1⎤ ⎢ x ⎥ ⎢ 0⎥ ⎢ ⎥ ⎢ 0⎥ 0 x4 = 0 or ⎢ 2 ⎥ = ⎢ ⎥ = s ⎢ ⎥ . ⎢ 0 −1 3 0 ⎥ ⎢ x3 ⎥ ⎢0 ⎥ ⎣⎢ 0 0 0 0 ⎥⎦ ⎢⎣ x4 ⎥⎦ ⎣⎢0 ⎦⎥ ⎡ 1 0 −2 0 ⎤ ⎢ −1 1 −1 0 ⎥ Since ⎢ ⎥ reduces to ⎢ 0 −1 3 0 ⎥ ⎣⎢ 0 0 0 0 ⎥⎦ ⎡1 ⎢0 ⎢ ⎢0 ⎣⎢0 0 −2 0 ⎤ 1 −3 0 ⎥ ⎥ .

λ−4 1 −2 λ − 1 = (λ − 4)(λ − 1) + 2 (a) det(λI − A) = ⎡ 1 0 2 0⎤ ⎢ 0 1 −1 0 ⎥ ⎢ ⎥ . and 29 = 512. The line y = x is invariant ⎡ −2 1⎤ ⎡ x ⎤ ⎡ 0 ⎤ = . the ⎥ ⎣0 0 ⎥⎦ ⎣0 0⎦ general solution is y = 0. the eigenspace corresponding to λ = 4. eigenvalues of A are 1 = 1. so ⎢ 1 ⎥ is a basis for ⎢ ⎥= =s ⎢0⎥ ⎢0⎥ ⎢ x3 ⎥ ⎢ 0 ⎥ ⎢0⎥ ⎢0⎥ ⎣⎢ x4 ⎦⎥ ⎢⎣ 0 ⎥⎦ ⎣ ⎦ ⎣ ⎦ x1 = 23. ⎜ ⎟ = 512 ⎝2⎠ 9 λ −1 = λ2 + 1 1 λ Thus 9 09 = 0. = ⎣0 0 ⎥⎦ ⎢⎣ y ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎡0 −3⎤ ⎡0 1⎤ Since ⎢ reduces to ⎢ . 9 1 ⎛1⎞ . and the line y = 0 is invariant under A. x2 = s. x3 = 0. Thus the 2 A−1 ( Ax) = A−1 (λx) = λ( A−1x) so A−1x = 1 is an eigenvalue of A−1 . = ⎣ −2 2 ⎥⎦ ⎢⎣ y ⎥⎦ ⎢⎣ 0⎥⎦ ⎡ −1 1⎤ ⎡ 1 −1⎤ ⎢⎣ −2 2 ⎥⎦ reduces to ⎢⎣0 0⎥⎦ . the general solution is 0 1⎥ 0 0 ⎥⎦ (c) det(λI − A) = λ − 2 −3 = (λ − 2)2 0 λ−2 ⎡ λ − 2 −3 ⎤ ⎡ x ⎤ ⎡ 0 ⎤ = ⎢⎣ 0 λ − 2 ⎥⎦ ⎢⎣ y ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎡0 −3⎤ ⎡ x ⎤ ⎡ 0 ⎤ λ = 2 gives ⎢ . but also 13. 0 6 −7 ⎥ ⎢ x3 ⎥ ⎢ 0 ⎥ 0 −1 2 ⎦⎥ ⎣⎢ x4 ⎦⎥ ⎣⎢ 0 ⎦⎥ (b) det(λI − A) = 9 0 0⎤ 6 0 0⎥ ⎥ reduces to 0 6 −7 ⎥ 0 −1 2 ⎦⎥ Since the characteristic equation λ 2 + 1 = 0 has no real solutions. λ = 2 gives ⎢ ⎣ −2 1⎥⎦ ⎢⎣ y ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎡ −2 1⎤ ⎡ 2 −1⎤ ⎢⎣ −2 1⎥⎦ can be reduced to ⎢⎣ 0 0 ⎥⎦ . under A. x2 = s. Since A is upper triangular. If a line is invariant under A. the general solution is ⎢0 0 0 1⎥ ⎢⎣0 0 0 0 ⎥⎦ x1 = −2s.1 ⎡ −1 0 −2 0 ⎤ ⎢ −1 −1 −1 0 ⎥ Since ⎢ ⎥ reduces to ⎢ 0 −1 1 0 ⎥ ⎢⎣ 0 0 0 −2 ⎥⎦ 15. so a general solution is x = y. 3 s. λ . with λ corresponding eigenvector x. x4 = 0 or = λ 2 − 5λ + 6 = (λ − 3)(λ − 2) 1 ⎤ ⎡ x ⎤ ⎡0⎤ ⎡λ − 4 ⎢⎣ −2 λ − 1⎥⎦ ⎢⎣ y ⎥⎦ = ⎢⎣0 ⎥⎦ ⎡ −1 1⎤ ⎡ x ⎤ ⎡ 0⎤ λ = 3 gives ⎢ . . 2. The line y = 2x is invariant under A. We have that Ax = λx. ⎡ x1 ⎤ ⎡ −2s ⎤ ⎡ −2 ⎤ ⎡ −2 ⎤ ⎢x ⎥ ⎢ s⎥ ⎢ ⎥ ⎢ 1⎥ 1 ⎢ 2⎥ = ⎢ ⎥ = s ⎢ ⎥ . there are no lines that are invariant under A. A−1 ( Ax) = Ix = x. so a general solution is 2x = y. x4 = 0 or 2 ⎡3⎤ ⎡3⎤ ⎡ x1 ⎤ ⎡ 3 s ⎤ ⎢2⎥ ⎢2⎥ ⎢x ⎥ ⎢ 2 ⎥ 2 s 1 ⎢ ⎥ ⎢ ⎥ . 123 1 x. 0 0⎤ ⎥ 1 0 ⎥ . then it can be expressed in terms of an eigenvector of A.SSM: Elementary Linear Algebra Section 5. x3 = s. 9 0 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎡λ − 10 ⎢ −4 0 0 ⎥ ⎢ x2 ⎥ ⎢0 ⎥ λ+2 (b) ⎢ ⎥⎢ ⎥ = ⎢ ⎥ 0 7 ⎥ ⎢ x3 ⎥ ⎢0 ⎥ λ+2 ⎢ 0 0 −1 λ − 2 ⎦⎥ ⎢⎣ x4 ⎥⎦ ⎣⎢0 ⎦⎥ ⎣⎢ 0 ⎡ −6 ⎢ −4 λ = 4 gives ⎢ ⎢ 0 ⎣⎢ 0 ⎡ −6 ⎢ −4 Since ⎢ ⎢ 0 ⎣⎢ 0 ⎡1 − 3 2 ⎢ 0 ⎢0 ⎢0 0 ⎢0 0 ⎣ 9 0 0 ⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ 6 0 0 ⎥ ⎢ x2 ⎥ ⎢ 0 ⎥ ⎥ ⎢ ⎥ = ⎢ ⎥. 0. the eigenvalues of A 1 are the diagonal entries: 1. so ⎢ ⎥ is a basis ⎢ x3 ⎥ ⎢ s ⎥ ⎢ 1⎥ ⎢ 1⎥ ⎢⎣ x4 ⎥⎦ ⎣⎢ 0 ⎦⎥ ⎣⎢ 0 ⎦⎥ ⎣⎢ 0⎦⎥ for the eigenspace corresponding to λ = −1.

The matrix is not diagonalizable since it has only 2 distinct eigenvectors. ⎡ 0 0⎤ For λ = 2. 11. (b) False. and 2. then Ax = λx for all λ. Since ⎢ ⎥ ⎣0 0 0 ⎦ rank is a similarity invariant. the only eigenvalue is λ = 2.5. ⎣ −1 0 ⎥⎦ hence nullity 1. (c) True. We have that Ax = λx. where 2 has algebraic multiplicity 2. 3. since p(0) = 02 + 1 ≠ 0. the eigenspace corresponding to λ contains the vector x = 0. The matrix is not diagonalizable. (g) False. hence nullity 1. Note that showing that the geometric multiplicity of either eigenvalue is less than its algebraic multiplicity is sufficient to show that the matrix is not diagonalizable. (sA)x = s(Ax) = s(λx) = (sλ)x so sλ is an eigenvalue of sA with corresponding eigenvector x. 2. so they have different ranks. Since the determinant is a similarity invariant and det(A) = 2 − 3 = −1 while det(B) = −2 − 0 = −2. or 3. then 02 = 0 is an eigenvalue of A2 .1 (a) False. whereas A can have eigenvalues that are not 1. 2 I − A = ⎢ 0 0 0 ⎥ which has rank ⎢ ⎥ ⎣ 0 −1 0 ⎦ 2. the eigenspace for λ = 0 can have dimension 1 or 2. then det(λI − A) = 0. For instance. the eigenvalues are 3 and 2. 2. 2. the reduced row echelon form of an invertible n × n matrix A is I n which has only one eigenvalue. True/False 5. 3I − A = ⎢0 1 0 ⎥ which has rank ⎢ ⎥ ⎣0 −1 1⎦ 2. the eigenspace for λ = 1 has dimension 1. the vector x must be nonzero to be an eigenvector⎯if x = 0. (e) True.1. Section 5. which is not an eigenvector of A. Since the matrix is lower triangular. Thus.2 Exercise Set 5. if λ is an eigenvalue of A. A and B are not similar matrices. λ = 0 is not an eigenvalue of A and A is invertible by Theorem 5. so (λI − A)x = 0 has nontrivial solutions. 2 I − A = ⎢ ⎥ which has ⎢ 0 0 −1 −2 ⎥ ⎣⎢0 0 0 −1⎦⎥ rank 3. 7. which has rank 1. and 3. hence nullity 1. ⎡0 0 0 ⎤ For λ = 3. 3I − A = ⎢ ⎥ which has ⎢ 0 0 0 −2 ⎥ ⎢⎣0 0 0 0 ⎥⎦ rank 3. and the eigenspace for λ = 2 can have dimension 1. (d) False. Since the geometric multiplicity of an eigenvalue is less than or equal to its algebraic multiplicity. the ⎡3 0 ⎤ matrix A = ⎢ has the eigenvalues λ = 3 ⎣8 −1⎥⎦ and λ = −1. ⎡ 1 0 0⎤ 3. the matrix has only 1 distinct eigenvector and is not diagonalizable. so the eigenspace corresponding to λ = 3 has dimension 1. ⎡ −1 0 0 ⎤ For λ = 2. Since the matrix is lower triangular with diagonal entries of 3.2 1. if 0 is an eigenvalue of A. if 0 is an eigenvalue of A. so the eigenspace corresponding to λ = 3 has dimension 1. since it has only 2 distinct eigenvectors. ⎡ 1 1 0 −1⎤ ⎢0 1 −1 1⎥ For λ = 3. A reduces to ⎢0 1 0 ⎥ while B reduces to ⎢ ⎥ ⎣0 0 1⎦ ⎡ 1 2 0⎤ ⎢0 0 1⎥ . 9. (f) False.Chapter 5: Eigenvalues and Eigenvectors SSM: Elementary Linear Algebra 25. Since s is a scalar. 124 . hence nullity 1. then A is singular so the set of column vectors of A cannot be linearly independent. with 2s on the diagonal. so the eigenspace corresponding to λ = 2 has dimension 1. the eigenvalues are 2 and 3 each with algebraic multiplicity 2. hence nullity 1. ⎡0 1 0 −1⎤ ⎢0 0 −1 1⎥ For λ = 2. so the eigenspace corresponding to λ = 2 has dimension 1. 2 I − A = ⎢ . A and B are not similar matrices. Since the matrix is upper triangular with diagonal entries 2. 5. λ = 1. so A2 is singular.

3 ⎣ x2 ⎦ ⎣1 ⎦ ⎡ 0 0⎤ Since ⎢ reduces to ⎣ −6 2 ⎥⎦ ⎡ 1 0 2 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ 2 = 3 gives ⎢0 0 0 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ which has ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 0 0 ⎦ ⎣ x3 ⎦ ⎣0 ⎦ general solution x1 = −2s. ⎡ 1 0 2 ⎤ ⎡ 2 0 −2 ⎤ ⎡ 1 −2 0 ⎤ P −1 AP = ⎢0 0 1⎥ ⎢ 0 3 0⎥ ⎢ 0 0 1⎥ ⎢ ⎥⎢ ⎥⎢ ⎥ 3⎦ ⎣ 0 1 0⎦ ⎣0 1 0 ⎦ ⎣ 0 0 0 ⎤ ⎡λ1 0 = ⎢0 λ 2 0 ⎥ ⎢ ⎥ λ2 ⎦ ⎣0 0 ⎡2 0 0⎤ = ⎢0 3 0⎥ ⎢ ⎥ ⎣ 0 0 3⎦ ⎡ −2 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ . Since each eigenvalue must have nonzero geometric multiplicity. x2 = s or ⎢ 1 ⎥ = s ⎢ ⎥ . x3 = s. x2 = s. λ 2 = −1 gives ⎢ ⎢ ⎥= ⎣ −6 0 ⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ ⎡ −2 0 ⎤ ⎡ 1 0⎤ Since ⎢ reduces to ⎢ . det(λI − A) = λ + 1 −4 2 3 λ−4 0 3 −1 λ − 3 = x3 − 6 x 2 + 11x − 6 = ( x − 1)( x − 2)( x − 3) The eigenvalues are λ1 = 1.2 13. each with algebraic multiplicity 1. A has eigenvalues λ1 = 1 and λ 2 = −1 (the diagonal entries. ⎡ 0 0⎤ ⎡ x1 ⎤ ⎡ 0⎤ λ1 = 1 gives ⎢ . P = [p1 p 2 ] = ⎢ 3 ⎥ diagonalizes A. the ⎢ ⎥ ⎢ ⎥ ⎣0 0 −1⎦ ⎣0 0 0 ⎦ general solution is x1 = s. and p = ⎢ 0 ⎥ . ⎣ 1 1⎦ ⎡ 3 0 ⎤ ⎡ 1 0 ⎤ ⎡ 13 0 ⎤ P −1 AP = ⎢ ⎢ ⎥ ⎣ −3 1⎥⎦ ⎢⎣ 6 −1⎥⎦ ⎣ 1 1⎦ 0⎤ ⎡λ =⎢ 1 ⎥ 0 λ 2⎦ ⎣ ⎡ 1 0⎤ =⎢ ⎣0 −1⎥⎦ 17. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 0 −1⎦ ⎣ x3 ⎦ ⎣ 0⎦ 125 . λ 2 = 2. the geometric multiplicity of each eigenvalue is also 1 and A is diagonalizable. and λ3 = 3. ⎢ ⎥⎢ ⎥ ⎢ ⎥ 0 λ − 3⎦ ⎣ x3 ⎦ ⎣ 0 ⎦ ⎣ 0 ⎡0 0 2⎤ ⎡ x1 ⎤ ⎡ 0⎤ λ1 = 2 gives ⎢0 −1 0⎥ ⎢ x2 ⎥ = ⎢ 0⎥ . x3 = 0 or ⎡ x1 ⎤ ⎡1 ⎤ ⎡1 ⎤ ⎢ x ⎥ = s ⎢0 ⎥ and p = ⎢0 ⎥ is a basis for the 1 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎣0 ⎦ ⎣0 ⎦ ⎣ x3 ⎦ eigenspace corresponding to λ1 = 2. Thus P = [p1 p 2 p3 ] = ⎢0 0 1⎥ ⎢ ⎥ 1 0⎦ ⎣0 diagonalizes A. (λI − A)x = 0 is 0 2 ⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ ⎡λ − 2 ⎢ 0 λ−3 0 ⎥ ⎢ x2 ⎥ = ⎢ 0 ⎥ . ⎡ x1 ⎤ ⎡0 ⎤ ⎡ −2 ⎤ ⎡ −2 ⎤ ⎡ 0 ⎤ ⎢ x ⎥ = s ⎢ 0 ⎥ + t ⎢1 ⎥ . λ−4 ⎢ ⎥⎢ ⎥ ⎢ ⎥ −1 λ − 3⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎣ 3 15. x2 = 0. x2 = t . or ⎢ 1 ⎥ = s ⎢ 3 ⎥ . and ⎣1 ⎦ ⎣ x2 ⎦ ⎡0 ⎤ p 2 = ⎢ ⎥ is a basis for the eigenspace ⎣1 ⎦ corresponding to λ 2 = −1. ⎡ 1 0⎤ Thus. ⎢ ⎥= ⎣ −6 2⎥⎦ ⎣ x2 ⎦ ⎢⎣ 0⎥⎦ ⎡1 − 1⎤ 3 ⎥ .SSM: Elementary Linear Algebra Section 5. the general ⎢ 0 0 ⎣ ⎦ ⎡1 ⎤ x 1 ⎡ ⎤ solution is x1 = s. since A is triangular). p = ⎢1 ⎥ 2 3 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 1⎦ ⎣ 0⎦ ⎣0 ⎦ ⎣ 1⎦ ⎣ x3 ⎦ is a basis for the eigenspace corresponding to ⎡ 1 −2 0 ⎤ λ 2 = 3. the general ⎥ − 6 0 ⎣0 0 ⎥⎦ ⎣ ⎦ ⎡x ⎤ ⎡0 ⎤ solution is x1 = 0. (λI − A)x = 0 is ⎢ ⎢ ⎥= ⎣ −6 λ + 1⎦⎥ ⎣ x2 ⎦ ⎣⎢0 ⎦⎥ ⎡0 0 2 ⎤ ⎡0 1 0 ⎤ Since ⎢0 −1 0 ⎥ reduces to ⎢0 0 1⎥ . since A is triangular). or ⎡1⎤ and p1 = ⎢ 3 ⎥ is a basis for the eigenspace ⎣1⎦ corresponding to λ1 = 1. A has eigenvalues λ1 = 2 and λ 2 = 3 (the diagonal entries. ⎡λ − 1 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ . (λI − A)x = 0 is 2 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎡ λ + 1 −4 ⎢ 3 0 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ .

Thus λ = 0 has 1 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎣ −3 ⎦ ⎣ 0 ⎦ ⎣ x3 ⎦ geometric multiplicity 2. the eigenvalues are λ1 = 0 (with algebraic multiplicity 2) and λ 2 = 1. ⎥⎢ ⎥ ⎢ ⎥ 0 0 ⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎡1 0 reduces to ⎢0 1 ⎢ ⎣0 0 general solution is x1 = 0. ⎢0 1 −1⎥ ⎢⎣0 0 0 ⎥⎦ ⎡ 0 0 0⎤ Since ⎢ 0 0 0 ⎥ reduces to ⎢ ⎥ ⎣ −3 0 −1⎦ 2 s. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ −3 0 λ − 1⎦ ⎣ x3 ⎦ ⎣ 0 ⎦ ⎡ 0 0 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ1 = 0 gives ⎢ 0 0 0 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . x3 = 4t which is ⎢ x2 ⎥ = t ⎢ 3⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 4⎦ ⎣ x3 ⎦ ⎡ 4 −4 2 ⎤ Since ⎢ 3 −1 0 ⎥ reduces to ⎢ ⎥ ⎣ 3 −1 0 ⎦ 0 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ 1 0 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . ⎡ x1 ⎤ ⎡1⎤ ⎡1⎤ ⎢ x ⎥ = s ⎢1⎥ and p = ⎢1⎥ is a basis for the 2 1 ⎢ ⎥ ⎢⎥ ⎢⎥ ⎣1⎦ ⎣1⎦ ⎣ x3 ⎦ eigenspace corresponding to λ1 = 1. x2 = s. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ −3 0 −1⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎡3 −4 2 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ 2 = 2 gives ⎢3 −2 0 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . x2 = s. the ⎢ ⎥ ⎢ ⎥ ⎣ 3 −1 −2 ⎦ ⎣0 0 0 ⎦ general solution is x1 = s. p 2 = ⎢ 1⎥ are a basis for the ⎢ ⎥ ⎢ ⎥ ⎣ −3⎦ ⎣0 ⎦ eigenspace corresponding to λ1 = 0. a ⎢0 0 0 ⎥ ⎣⎢0 0 0 ⎥⎦ ⎡1 2 1⎤ p3 ] = ⎢1 3 3⎥ ⎢ ⎥ ⎣1 3 4 ⎦ Thus P = [p1 p 2 126 ⎡ 1 0 0⎤ p3 ] = ⎢ 0 1 0 ⎥ ⎢ ⎥ ⎣ −3 0 1⎦ . x2 = t . x2 = 3t . or diagonalizes A and 0 ⎤ ⎡ 1 0 0⎤ ⎡λ1 0 P −1 AP = ⎢0 λ 2 0 ⎥ = ⎢0 2 0 ⎥ . x3 ⎡ 1 λ 2 = 1 gives ⎢ 0 ⎢ ⎣ −3 ⎡ 1 0 0⎤ Since ⎢ 0 1 0 ⎥ ⎢ ⎥ ⎣ −3 0 0 ⎦ 0⎤ 0⎥ . x2 = s. Since A is triangular with diagonal entries of 0. x3 = s or 4 4 ⎡ x1 ⎤ ⎡1 ⎤ x1 = t . ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 3 −1 −2 ⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎡ 2 −4 2 ⎤ ⎡ 1 0 −1⎤ Since ⎢ 3 −3 0 ⎥ reduces to ⎢0 1 −1⎥ .Chapter 5: Eigenvalues and Eigenvectors SSM: Elementary Linear Algebra ⎡ 2 −4 2 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ1 = 1 gives ⎢ 3 −3 0 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . x3 = s. and A is diagonalizable. x2 = 3t . 0. x2 = 0. 0 ⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ ⎡λ 0 (λI − A)x = 0 is ⎢ 0 λ 0 ⎥ ⎢ x2 ⎥ = ⎢ 0 ⎥ . ⎡ 1⎤ ⎡0 ⎤ p1 = ⎢ 0⎥ . ⎡1 ⎤ and p3 = ⎢ 3 ⎥ is a basis for the eigenspace ⎢ ⎥ ⎣4⎦ corresponding to λ3 = 3. x2 = t . λ 2 = 1 must also have geometric multiplicity 1. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣3 −1 −1⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎡ 3 −4 2 ⎤ Since ⎢3 −2 0⎥ reduces to ⎢ ⎥ ⎣3 −1 −1⎦ ⎡1 0 − 2⎤ 3⎥ ⎢ . a ⎢ ⎥ 0⎦ ⎣0 0 1 3 general solution is x1 = s. x3 = r or 3 x1 = s. x3 = −3s which is ⎡ x1 ⎤ ⎡2⎤ ⎡ 2⎤ ⎢ x ⎥ = t ⎢ 3⎥ and p = ⎢ 3 ⎥ is a basis for the 2 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 3⎦ ⎣3⎦ ⎣ x3 ⎦ eigenspace corresponding to λ 2 = 2. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 3 −1 0 ⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎡1 0 − 1⎤ 4⎥ ⎢ ⎢0 1 − 34 ⎥ . ⎡ x1 ⎤ ⎡ 1⎤ ⎡0 ⎤ ⎢ x ⎥ = s ⎢ 0 ⎥ + t ⎢ 1⎥ . x3 = 3t which is the general solution is x1 = 1 general solution is x1 = − r . Thus P = [p1 p 2 ⎡1 0 1⎤ 3⎥ ⎢ . 1. x3 = s 3 or x1 = 2t . ⎢ ⎥ ⎢ ⎥ λ 3 ⎦ ⎣ 0 0 3⎦ ⎣0 0 19. a ⎥ 0⎦ = s or ⎡ x1 ⎤ ⎡0 ⎤ ⎡0 ⎤ ⎢ x ⎥ = s ⎢0 ⎥ and p = ⎢0 ⎥ is a basis for the 3 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 1⎦ ⎣ 1⎦ ⎣ x3 ⎦ eigenspace corresponding to λ 2 = 1. ⎡ 4 −4 2 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ3 = 3 gives ⎢ 3 −1 0 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ .

SSM: Elementary Linear Algebra Section 5. x4 = 0 or ⎢ 2 ⎥ = s ⎢ ⎥ + t ⎢ ⎥ . both with algebraic multiplicity 2. 0 0 5 −5 0 0 0 0 ⎡5 ⎢0 Since ⎢ ⎢0 ⎣⎢0 0⎤ 5⎥ ⎥ reduces to 0⎥ 0 ⎦⎥ 0⎤ 0⎥ ⎥. a general solution is x1 = s. det(λI − A) = 0 0 3 0 127 . and ⎡ −2 0 ⎢ 0 −2 =⎢ ⎢ 0 0 ⎢⎣ 0 0 ⎡0 ⎤ ⎡1 ⎤ ⎢1 ⎥ ⎢0⎥ p1 = ⎢ ⎥ . A is diagonalizable. Since A is triangular with entries −2. ⎢ x3 ⎥ ⎢1 ⎥ ⎢ 0 ⎥ ⎢⎣ x4 ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎢⎣ 1⎥⎦ Thus λ 2 = 3 has geometric multiplicity 2. and 3 on the diagonal. 3. λ−3 0 0 ⎥ ⎢ x3 ⎥ ⎢0 ⎥ ⎢ 0 λ − 3⎥⎦ ⎢⎣ x4 ⎥⎦ ⎢⎣0 ⎥⎦ 0 0 ⎢⎣ 0 ⎡0 ⎢0 λ1 = −2 gives ⎢ ⎢0 ⎣⎢0 ⎡0 ⎢0 Since ⎢ ⎢0 ⎣⎢0 ⎡0 ⎢0 ⎢ ⎢0 ⎢⎣0 0 0 0 0 1 0 0 0 ⎡0 ⎤ ⎡ 0⎤ ⎢1 ⎥ ⎢ −1⎥ p3 = ⎢ ⎥ . x3 = 0. ⎢ x3 ⎥ ⎢0 ⎥ ⎢ 0⎥ ⎢⎣ x4 ⎥⎦ ⎢⎣0 ⎥⎦ ⎢⎣ 0⎥⎦ Thus λ1 = −2 has geometric multiplicity 2. x4 = t . ⎢ ⎥⎢ ⎥ ⎢ ⎥ −15 λ + 2 ⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎣ 0 ⎡ −1 −7 1⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ1 = −2 gives ⎢ 0 −3 0⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . and λ3 = 1. −2. (λI − A)x = 0 is 0 0 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎡λ + 2 ⎢ 0 λ + 2 −5 5 ⎥ ⎢ x2 ⎥ ⎢0 ⎥ ⎢ ⎥ ⎢ ⎥ = ⎢ ⎥. p 2 = ⎢ ⎥ are a basis for the ⎢0 ⎥ ⎢0⎥ ⎢⎣0 ⎥⎦ ⎢⎣ 0 ⎥⎦ eigenspace corresponding to λ1 = −2. ⎡ 1 0 0 0⎤ ⎢ 0 1 1 −1⎥ P = [p1 p 2 p3 p 4 ] = ⎢ ⎥ ⎢0 0 1 0⎥ ⎣⎢ 0 0 0 1⎦⎥ diagonalizes A. a general solution is x1 = 0. or 21. x3 = s. ⎢ ⎥ ⎢ ⎥ ⎣ 0 0 λ 2 ⎦ ⎣0 0 1⎦ ⎡ 1 0 0 0⎤ ⎢0 1 −1 1⎥ ⎢ ⎥ . 0 ⎥ ⎢ x3 ⎥ ⎢ 0⎥ 0 ⎦⎥ ⎢⎣ x4 ⎥⎦ ⎣⎢ 0⎦⎥ ⎡5 ⎢0 λ 2 = 3 gives ⎢ ⎢0 ⎣⎢0 0 0 5 −5 0 0 0 0 23. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 0 −15 0⎦ ⎣ x3 ⎦ ⎣0 ⎦ 0 ⎤ ⎡ x1 ⎤ ⎡ 0⎤ 5⎥ ⎢ x2 ⎥ ⎢ 0⎥ ⎥ ⎢ ⎥ = ⎢ ⎥. and λ 2 = 3. 0⎥ 3⎥⎦ 1 λ + 1 −7 0 0 λ −1 0 −15 λ + 2 = λ3 + 2λ 2 − λ − 2 = (λ + 2)(λ + 1)(λ − 1) Thus. and 0⎤ ⎡λ1 0 0 ⎢0 λ 0 0⎥ −1 1 ⎥ P AP = ⎢ ⎢ 0 0 λ2 0 ⎥ ⎢⎣ 0 0 0 λ 2 ⎥⎦ 0 0 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ 0 −5 5⎥ ⎢ x2 ⎥ ⎢0 ⎥ ⎥ ⎢ ⎥ = ⎢ ⎥. ⎢0 0 0 0 ⎥ ⎢⎣0 0 0 0 ⎥⎦ x2 = s − t . p 4 = ⎢ ⎥ are a basis for the ⎢1 ⎥ ⎢ 0⎥ ⎢⎣ 1⎥⎦ ⎣⎢0 ⎦⎥ eigenspace corresponding to λ 2 = 3. 0 −5 0 ⎥ ⎢ x3 ⎥ ⎢0 ⎥ 0 0 −5⎦⎥ ⎣⎢ x4 ⎦⎥ ⎣⎢0 ⎦⎥ 0 0 0⎤ 0 −5 5⎥ ⎥ reduces to 0 −5 0 ⎥ 0 0 −5⎦⎥ 0⎤ 1⎥ ⎥ . the eigenvalues of A are λ1 = −2. (λI − A)x = 0 is 1 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎡ λ + 1 −7 ⎢ 0 λ −1 0 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . the eigenvalues are λ1 = −2 and ⎡ x1 ⎤ ⎡0⎤ ⎡ 0⎤ ⎢x ⎥ ⎢1 ⎥ ⎢ −1⎥ 2 ⎢ ⎥ = s ⎢ ⎥ +t ⎢ ⎥. λ 2 = −1. 0⎥ 0 ⎥⎦ ⎡ x1 ⎤ ⎡1 ⎤ ⎡ 0 ⎤ ⎢x ⎥ ⎢0 ⎥ ⎢1 ⎥ x2 = t .2 diagonalizes A and ⎡λ1 0 0 ⎤ ⎡0 0 0 ⎤ P −1 AP = ⎢ 0 λ1 0 ⎥ = ⎢0 0 0 ⎥ . Since the geometric and algebraic multiplicities are equal for both eigenvalues.

x3 = s or ⎡1 ⎤ p3 = ⎢1⎥ is a basis for the eigenspace ⎢ ⎥ ⎣5⎦ corresponding to λ3 = 1. ⎢ ⎥⎢ ⎥ ⎢ ⎥ λ − 3⎦ ⎣ x3 ⎦ ⎣ 0 ⎦ 1 ⎣ 0 ⎡ −2 1 0 ⎤ ⎡ x1 ⎤ ⎡ 0⎤ λ1 = 1 gives ⎢ 1 −1 1⎥ ⎢ x2 ⎥ = ⎢ 0⎥ . x2 = s. ⎢ ⎥ ⎣1 0 5⎦ ⎡ x1 ⎤ ⎡1 ⎤ ⎡1 ⎤ ⎢ x ⎥ = s ⎢ 2⎥ and p = ⎢ 2 ⎥ is a basis for the 1 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎣1 ⎦ ⎣1 ⎦ ⎣ x3 ⎦ eigenspace corresponding to λ1 = 1. x2 = t . so p = ⎢0 ⎥ is a basis for the 1 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎣1 ⎦ ⎣1 ⎦ ⎣ x3 ⎦ eigenspace corresponding to λ1 = −2. x3 = s or 5 5 ⎡ x1 ⎤ ⎡1 ⎤ x1 = t . x2 = 0. or ⎡ x1 ⎤ ⎡1 ⎤ ⎡1 ⎤ ⎢ x ⎥ = s ⎢0 ⎥ . x3 = 0. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 0 1 −2 ⎦ ⎣ x3 ⎦ ⎣ 0⎦ ⎡ −2 1 0 ⎤ ⎡ 1 0 −1⎤ Since ⎢ 1 −1 1⎥ reduces to ⎢0 1 −2⎥ . x3 = s. a ⎢ ⎥ ⎢ ⎥ ⎣ 0 1 −2 ⎦ ⎣0 0 0⎦ general solution is x1 = s. ⎡0 −7 1⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ λ 2 = −1 gives ⎢0 −2 0 ⎥ ⎢ x2 ⎥ = ⎢ 0 ⎥ . ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 1 0 ⎦ ⎣ x3 ⎦ ⎣0 ⎦ 128 . x2 = 0. (λI − A)x = 0 is 1 0 ⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ ⎡λ − 3 ⎢ 1 λ−2 1 ⎥ ⎢ x2 ⎥ = ⎢ 0 ⎥ . ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 −15 1⎦ ⎣ x3 ⎦ ⎣ 0 ⎦ ⎡0 −7 1⎤ ⎡0 1 0 ⎤ Since ⎢0 −2 0 ⎥ reduces to ⎢0 0 1 ⎥ . or 11 A ⎡ x1 ⎤ ⎡1 ⎤ ⎡1 ⎤ ⎢ x ⎥ = s ⎢0 ⎥ so p = ⎢0 ⎥ is a basis for the 2 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎣0 ⎦ ⎣0 ⎦ ⎣ x3 ⎦ eigenspace corresponding to λ 2 = −1. 237 −2047 ⎤ =⎢ 0 1 0⎥ ⎢ ⎥ ⎣ 0 10. 245 −2048⎦ λ −3 1 0 25. det(λI − A) = 1 λ−2 1 0 1 λ −3 = λ3 − 8λ 2 + 19λ − 12 = (λ − 1)(λ − 3)(λ − 4) The eigenvalues of A are λ1 = 1. a ⎢ ⎥ ⎢ ⎥ ⎣0 −15 1⎦ ⎣0 0 0 ⎦ general solution is x1 = s. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 0 −15 3⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎡1 0 − 1⎤ ⎡ 2 −7 1⎤ 5⎥ ⎢ Since ⎢ 0 0 0⎥ reduces to ⎢0 1 − 1 ⎥ . ⎡ 2 −7 1⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ3 = 1 gives ⎢ 0 0 0 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . λ 2 = 3.Chapter 5: Eigenvalues and Eigenvectors SSM: Elementary Linear Algebra and D = P −1 AP ⎡ 0 −5 = ⎢1 4 ⎢ 1 ⎣0 ⎡ λ1 0 = ⎢ 0 λ2 ⎢ ⎣0 0 ⎡ −2 0 = ⎢ 0 −1 ⎢ ⎣ 0 0 ⎡ −1 −7 1⎤ ⎡ 1 0 −1⎤ Since ⎢ 0 −3 0 ⎥ reduces to ⎢0 1 0 ⎥ . and λ3 = 4. x2 = 2s. ⎡0 1 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ 2 = 3 gives ⎢1 1 1 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . P = [p1 p 2 11 −1 1⎤ ⎡ −1 7 −1⎤ ⎡ 1 1 1⎤ −1⎥ ⎢ 0 1 0 ⎥ ⎢0 0 1⎥ ⎥⎢ ⎥⎢ ⎥ 0 ⎦ ⎣ 0 15 −2 ⎦ ⎣ 1 0 5⎦ 0⎤ 0⎥ ⎥ λ3 ⎦ 0⎤ 0⎥ ⎥ 1⎦ ⎡1 1 1 ⎤ p3 ] = ⎢0 0 1⎥ diagonalizes A. a 5 ⎢ ⎥ ⎢ ⎥ ⎣ 0 −15 3⎦ 0⎦ ⎣0 0 1 1 general solution is x1 = s. x3 = 5t which is ⎢ x2 ⎥ = t ⎢1⎥ so ⎢ ⎥ ⎢ ⎥ ⎣5 ⎦ ⎣ x3 ⎦ = PD P ⎡(−2)11 0 0⎤ ⎢ ⎥ −1 11 = P⎢ 0 (−1) 0 ⎥P ⎢ 0 0 111 ⎥⎦ ⎣ ⎡ −2048 0 0 ⎤ = P⎢ 0 −1 0 ⎥ P −1 ⎢ ⎥ 0 0 1⎦ ⎣ ⎡ −1 10. a ⎢ ⎥ ⎢ ⎥ ⎣ 0 −15 0 ⎦ ⎣0 0 0 ⎦ general solution is x1 = s.

⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣0 1 1 ⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎡1 1 0 ⎤ ⎡ 1 0 −1⎤ Since ⎢1 2 1 ⎥ reduces to ⎢0 1 1⎥ .2 (a) True. P −1 Ak P −1 =  ) A" ( PP −1 ) AP P −1 A( PP −1 ) A( PP ⎡ x1 ⎤ ⎡ 1⎤ ⎡ 1⎤ ⎢ x ⎥ = s ⎢ 0 ⎥ and p = ⎢ 0 ⎥ is a basis for the 2 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎣ −1⎦ ⎣ −1⎦ ⎣ x3 ⎦ eigenspace corresponding to λ 2 = 3. 31. 1⎡ a + d − (a − d )2 + 4bc ⎤⎦ .1. Thus the dimensions of the eigenspaces are: λ = 1: dimension 1 λ = 3: dimension 2 λ = 4: dimension 3 ⎡ 1 1 1⎤ p3 ] = ⎢ 2 0 −1⎥ ⎢ ⎥ ⎣ 1 −1 1⎦ diagonalizes A and ⎡ λ1 0 P −1 AP = D = ⎢ 0 λ 2 ⎢ ⎣0 0 0 ⎤ ⎡ 1 0 0⎤ 0 ⎥ = ⎢0 3 0⎥ . 1 1⎤ ⎡1 6 3 6⎥ ⎢ P −1 = ⎢ 12 0 − 12 ⎥ ⎢1 1 1⎥ ⎢⎣ 3 − 3 3 ⎥⎦ n True/False 5. or Thus P = [p1 p 2 with k factors of A −1 −1 −1 = ( P1P2 )−1 A( P1P2 ). the dimension is 1. since B = P −1 AP. the dimension is 1 or 2. which must be less than or equal to the algebraic multiplicity of the eigenvalue.SSM: Elementary Linear Algebra Section 5. (a) The dimension of the eigenspace corresponding to an eigenvalue is the geometric multiplicity of the eigenvalue.2 ⎡0 1 0 ⎤ ⎡1 0 Since ⎢1 1 1 ⎥ reduces to ⎢0 1 ⎢ ⎥ ⎢ ⎣0 1 0 ⎦ ⎣0 0 general solution is x1 = s. so is D k . −b ⎤ ⎡ −b one possibility is P = ⎢ ⎥ where ⎣ a − λ1 a − λ 2 ⎦ 1 λ1 = ⎡ a + d + (a − d )2 + 4bc ⎤ and ⎦ 2⎣ λ2 = k Since D is a diagonal matrix. use P = I. Thus. ⎡1 1 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ3 = 4 gives ⎢1 2 1 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . B = P1−1 AP1 for n −1 A = PD P n ⎡ 1 1 1⎤ ⎡1 ⎢ = ⎢ 2 0 −1⎥ ⎢ 0 ⎢ ⎥ ⎣ 1 −1 1⎦ ⎢⎣ 0 0 3n 0 ⎡ 0 ⎤ ⎢6 ⎥ 1 0 ⎥ ⎢2 ⎢ 4n ⎥⎦ ⎢ 13 ⎣ 1 1 3 0 − 13 some invertible matrix P1. then B −1 = ( P −1 AP)−1 = P −1 A−1P so A−1 and B −1 are similar. C = P2−1 BP2 for some invertible matrix P2 . Since B is similar to 1⎤ 6⎥ − 12 ⎥ 1⎥ 3 ⎥⎦ C. 2⎣ (c) True. for λ = 3. x2 = 0. for λ = 1. then P −1 AP = D. x3 = s. for λ = 4. a ⎢ ⎥ ⎢ ⎥ ⎣0 1 1 ⎦ ⎣0 0 0 ⎦ general solution is x1 = s. ⎥ ⎢ ⎥ λ3 ⎦ ⎣0 0 4⎦ (c) Since only λ = 4 can have an eigenspace with dimension 3. If A is diagonalizable. For the 1⎤ 0 ⎥ . Using the result in Exercise 20 of Section 5. (b) If A is diagonalizable then each eigenvalue has the same algebraic and geometric multiplicity. the ⎥ 0⎦ x3 = − s or same P. = ( P AP)( P AP)" ( P AP ) k factors of P −1 AP =D 33. or 3. Then C = P2−1 BP2 = P2−1 ( P1−1 AP1 ) P2 = ( P2−1P1−1 ) A( P1 P2 ) 27. so Ak is also diagonalizable. the eigenvalue must be 4. 2. since A is similar to B. 129 . Since the product of invertible matrices is invertible. the dimension is 1. (b) True. ⎡ x1 ⎤ ⎡ 1⎤ ⎡ 1⎤ ⎢ x ⎥ = s ⎢ −1⎥ and p = ⎢ −1⎥ is a basis for the 3 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 1⎦ ⎣ 1⎦ ⎣ x3 ⎦ eigenspace corresponding to λ3 = 4. P = P1 P2 is invertible and A is similar to C. x2 = − s.

Re(4i). Since the algebraic and geometric multiplicities are the same for each eigenvalue. 5 + 3i) = i(2 − i) + 2i(2i) + 3(5 + 3i) = i(2 + i) + 2i(−2i) + 3(5 − 3i) = 2i − 1 + 4 + 15 − 9i = 18 − 7i v ⋅ w = (4. then every eigenvalue also has geometric multiplicity 1. 0. 2 − i. 2i. 2i. 12 + 6i) = (7 − 6i. 2i. 6i) ⋅ (4. so A−1 is also diagonalizable. 1 + i) = −2(4) − 4(2i) + 6i(1 − i) = −8 − 8i + 6i + 6 = −2 − 2i = k (u ⋅ v ) (f) True. − 4 − 8i. 2i. 5 + 3i) = 4(2 − i) − 2i(2i) + (1 + i)(5 + 3i) = 4(2 + i) − 2i(−2i) + (1 + i)(5 − 3i) = 8 + 4i − 4 + 5 + 2i + 3 = 12 + 6i v ⋅ u = 4(−i) − 2i(−2i) + (1 + i)(3) = −4i − 4 + 3 + 3i = −1 − i = u⋅v u ⋅ ( v + w) = (i. 1 + i ) = (2 + i. (g) True. − 2i. 0. i x = −i(3(1 + i. u ⋅ v = (i. then ix = 3v + u. 3) ⋅ (4. 1 − i) = i(4) + 2i(−2i) + 3(1 + i) = 4i + 4 + 3 + 3i = 7 + 7i ⎡ −5i 4 ⎤ ⎡ 5i 4 ⎤ 7. 6i)) = −i(6 + 7i. 2i. 4. A = ⎢ ⎥ = ⎢ 2 + i 1 − 5i ⎥ 2 i 1 5 i − + ⎣ ⎦ ⎣ ⎦ Re(4) ⎤ ⎡ 0 4 ⎤ ⎡ Re(−5i) Re(A) = ⎢ = ⎣ Re(2 − i) Re(1 + 5i) ⎥⎦ ⎢⎣ 2 1 ⎥⎦ w ⋅ u = (u ⋅ w) = u ⋅ w = 18 − 7i (u ⋅ v) − w ⋅ u = 7 + 7i − (18 − 7i) = −11 + 14i = −11 − 14i 130 . (e) True. (h) True.3 Exercise Set 5. 4) + (3 − 4i. 1) u = = 2 2 2 2 − i + 4i + 1 + i ( 22 + 12 = 5 + 16 + 2 = 23 ) +( 2 42 ) +( 2 12 + 12 ) 2 5. 12) + (3 + 4i.Chapter 5: Eigenvalues and Eigenvectors SSM: Elementary Linear Algebra Im(4) ⎤ ⎡ −5 0⎤ ⎡ Im(−5i) Im( A) = ⎢ = ⎣ Im(2 − i) Im(1 + 5i) ⎥⎦ ⎢⎣ −1 5⎥⎦ det( A) = −5i(1 + 5i) − 4(2 − i) = −5i + 25 − 8 + 4i = 17 − i tr(A) = −5i + (1 + 5i) = 1 (d) False. − 2i. 1 + i) = (−2. DT = ( P −1 AP)T = PT AT ( P −1 )T . − 2i. 2 − i. 6 − 12i) 13. 2i. 3) ⋅ (2 − i. if A is n × n and has n linearly independent eigenvectors. then the geometric multiplicity of each eigenvalue must be equal to its algebraic multiplicity. 3)) ⋅ (4. Im(4i). A is diagonalizable. if every eigenvalue of A has algebraic multiplicity 1. 3) ⋅ (4. − 4. then P −1 A−1P = ( P −1 AP)−1 = D −1. Im(1 + i)) = (−1. 8 − 4i. hence A is diagonalizable. 2 + i. 1 + i) = i( 4) + 2i(−2i) + 3(1 + i) = 4i + 2i(2i) + 3(1 − i) = 4i − 4 + 3 − 3i = −1 + i u ⋅ w = (i. If ix − 3v = u. Re(1 + i)) = (2. Section 5. 6 − 3i. if P −1 AP = D. 4i. − 6i)) = −i((3 + 3i. 2i. 11. − 4i. 1 + i) ⋅ (2 − i. 6 + 4i) = i(6 + i) + 2i(0) + 3(6 − 4i) = 6i − 1 + 18 − 12i = 17 − 6i = (−1 + i) + (18 − 7i) = u⋅v + u⋅w k(u ⋅ v) = 2i(−1 + i) = −2i − 2 = −2 − 2i (ku) ⋅ v = (2i(i. u ⋅ v = (i. 1 − i) Re(u) = (Re(2 − i). − 2i. 2i. 3) ⋅ (6 − i. and 1 x = (3v + u) = −i(3v + u). u = ( 2 − i. 1) Im(u) = (Im(2 − i). so AT is also diagonalizable. a matrix P that diagonalizes A is not unique⎯it depends on the order of the eigenvalues in the diagonal matrix D and the choice of basis for each eigenspace.3 1.

3⎥⎦ ⎣2 17. a Since ⎢ 0 ⎥⎦ ⎣0 ⎣ −1 1 − i ⎥⎦ general solution is x1 = (1 − i) s. tr(A) = 4 + 0 = 4 det(A) = 4(0) − 1(−5) = 5 The characteristic equation of A is 21. a ⎥ 3 − 3i ⎦ ⎣ 3 ⎣0 0 ⎥⎦ general solution is x1 = (−1 + i) s. ⎡λ − 4 5 ⎤ ⎡ x1 ⎤ ⎡ 0⎤ (λI − A) = 0 is ⎢ . x2 = s or 25. λ 2 − 10λ + 34 = 0 which has solutions λ = 5 ± 3i. x2 = s or 19. (λI − A)x = 0 is ⎢ 1 − λ − 3⎥⎦ ⎢⎣ x2 ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎣ ⎡ −1 − i 2 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ1 = 4 − i gives ⎢ . ⎝1⎠ 4 ⎡ x1 ⎤ ⎡1 − i ⎤ ⎡1 − i ⎤ ⎢ ⎥ = s ⎢ −1 ⎥ . ⎢ ⎥= ⎣ −1 1 − i ⎦⎥ ⎣ x2 ⎦ ⎣⎢0 ⎦⎥ ⎡ −1 − i 2 ⎤ ⎡1 −1 + i ⎤ reduces to ⎢ . ⎡ −3 − 3i −6 ⎤ ⎡1 1 − i ⎤ Since ⎢ reduces to ⎢ .SSM: Elementary Linear Algebra Section 5. which has solutions λ = 2 ± i. x2 = 2t which is ⎡2 + i ⎤ x 2 = x1 = ⎢ is a basis for the eigenspace ⎣ 1 ⎥⎦ corresponding to λ 2 = λ1 = 2 + i. ⎡1 1 + 1 i ⎤ 5 ⎤ ⎡ 4 − 2i 2 ⎥. a reduces to ⎢ ⎢⎣ −4 −4 − 2i ⎥⎦ 0 ⎦ ⎣0 1 ⎞ ⎛ general solution is x1 = ⎜ −1 − i ⎟ s. For λ = 3 − 2i. tr(A) = 8 + 2 = 10 det(A) = 8(2) − (−3)(6) = 34 The characteristic equation of A is ⎡ x1 ⎤ ⎡1 − i ⎤ ⎡1 − i ⎤ ⎢ x ⎥ = s ⎢ 1 ⎥ so x1 = ⎢ 1 ⎥ is a basis for the ⎣ ⎦ ⎣ ⎦ ⎣ 2⎦ eigenspace corresponding to λ1 = 4 − i. tr(A) = −1 + 7 = 6 det(A) = −1(7) − 4(−5) = 13 The characteristic equation of A is ⎡ 1 −2 + i ⎤ . The angle from the positive xaxis to the ray that joins the origin to the point ⎛1⎞ π (1. 2 ⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ ⎡λ − 5 = . x2 = s or 2 ⎠ ⎝ x1 = (−2 − i)t . 23. = ⎢⎣ 3 3 − 3i ⎥⎦ ⎢⎣ x2 ⎥⎦ ⎢⎣0 ⎥⎦ ⎡1 + i ⎤ x 2 = x1 = ⎢ is a basis for the eigenspace ⎣ 1 ⎥⎦ corresponding to λ 2 = λ1 = 4 + i. 1) is φ = tan −1 ⎜ ⎟ = . − 3 ) is φ = tan −1 ⎜ − 3 ⎟ = − . Here a = b = 1. ⎡ −2 ⎤ ⎡ −1⎤ Re(x) = ⎢ ⎥ and Im(x) = ⎢ ⎥ . tr(A) = 5 + 3 = 8 det(A) = 5(3) − 1(−2) = 17 The characteristic equation of A is λ 2 − 8λ + 17 = 0 which has solutions λ = 4 ± i.3 15. ⎢ ⎥= ⎣ −1 2 − i ⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ 5 ⎤ ⎡ −2 − i reduces to Since ⎢ 1 2 − − i ⎥⎦ ⎣ general solution is x1 = (2 − i) s. (λI − A)x = 0 is ⎡ −3 − 3i −6 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ . ⎡ 1⎤ ⎡ −1⎤ Re(x) = ⎢ ⎥ and Im(x) = ⎢ ⎥ . ⎣ 2⎦ ⎣ 0⎦ ⎡ −2 −1⎤ Thus A = PCP −1 where P = ⎢ and ⎣ 2 0 ⎥⎦ ⎡ 3 −2 ⎤ C=⎢ . so x = ⎢ −1 ⎥ is a basis for the ⎣ ⎦ ⎣ ⎦ ⎣ x2 ⎦ eigenspace corresponding to λ = 5 − 3i. Here a = 1. det(λI − A)x = 0 is 5 ⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ ⎡ 4 − 2i . a ⎢⎣0 0 ⎥⎦ x2 = s or λ 2 − 6λ + 13 = 0 which has solutions λ = 3 ± 2i. b = − 3. ⎡ x1 ⎤ ⎡ −2 − i ⎤ ⎡ −2 − i ⎤ ⎢ x ⎥ = t ⎢ 2 ⎥ so x = ⎢ 2 ⎥ is a basis for ⎣ ⎦ ⎣ ⎦ ⎣ 2⎦ the eigenspace corresponding to λ = 3 − 2i. For λ = 5 − 3i. Since = ⎢⎣ −4 −4 − 2i ⎥⎦ ⎢⎣ x2 ⎥⎦ ⎢⎣ 0 ⎥⎦ ⎡ x1 ⎤ ⎡2 − i ⎤ ⎡2 − i ⎤ ⎢ x ⎥ = s ⎢ 1 ⎥ so x1 = ⎢ 1 ⎥ is a basis for the ⎣ ⎦ ⎣ ⎦ ⎣ 2⎦ eigenspace corresponding to λ1 = 2 − i. ⎢ ⎥= ⎣ −1 λ ⎦⎥ ⎣ x2 ⎦ ⎣⎢ 0⎦⎥ λ = 1− i 3 = 1+ 3 = 2 5 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎡ −2 − i λ1 = 2 − i gives ⎢ . − 1 ⎣ ⎦ ⎣ 0⎦ 2 2 λ = 1 + 1i = 1 + 1 = 2 131 . 3 ⎝ 1 ⎠ λ 2 − 4λ + 5 = 0. The angle from the positive x-axis to the ray that joins the origin to π ⎛ ⎞ the point (1.

(b) True. i. 6i.Chapter 5: Eigenvalues and Eigenvectors SSM: Elementary Linear Algebra ⎡ 1 −1⎤ Thus. (a) The coefficient matrix for the system is ⎡1 4 ⎤ A=⎢ . A = PCP −1 where P = ⎢ and ⎣ −1 0 ⎥⎦ ⎡ 5 −3⎤ C=⎢ . λ 2 − Tr( A)λ + det( A) = 0 is the characteristic equation of a 2 × 2 matrix A. ⎡1 −2 ⎤ diagonalizes Thus P = [p1 p 2 ] = ⎢ ⎣1 1⎥⎦ 0 ⎤ ⎡5 0⎤ ⎡λ A and P −1 AP = D = ⎢ 1 ⎥=⎢ ⎥. ⎢ ⎥= ⎣ −2 −4 ⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ ⎡ −2 −4 ⎤ ⎡1 2 ⎤ reduces to ⎢ a Since ⎢ ⎥ − − 2 4 ⎣0 0 ⎥⎦ ⎣ ⎦ general solution is x1 = −2s. True/False 5. ⎡ x1 ⎤ ⎡1⎤ ⎡1⎤ ⎢ x ⎥ = s ⎢1⎥ and p1 = ⎢1⎥ is a basis for the ⎣⎦ ⎣⎦ ⎣ 2⎦ eigenspace corresponding to λ1 = 5. k ) = 2i(−i) + i(−6i) + 3i(k ) = 2 + 6 + 3ik = 8 + 3ik If u ⋅ v = 0. (a) u ⋅ v = (2i. ⎡ 5 0⎤ u the “diagonal system” u′ = Du = ⎢ ⎣ 0 −1⎥⎦ u ′ = 5u1 or 1 which has the solution u2′ = −u2 (e) False. − 1. then kA has the same eigenvalues as A with the same algebraic multiplicities but tr(kA) = ktr(A) ≠ tr(A). this is only true if the eigenvalues of A satisfy λ = 1. (c) False. 1 − i) = k (1) + k (−1) + (1 + i)(1 + i) = (1 + i)2 = 2i Thus. ⎡ x1 ⎤ ⎡ −2 ⎤ ⎡ −2 ⎤ ⎢ x ⎥ = ⎢ 1⎥ and p 2 = ⎢ 1⎥ is a basis for ⎣ ⎦ ⎣ 2⎦ ⎣ ⎦ the eigenspace corresponding to λ 2 = −1. Then y = Pu gives the ⎣⎢c2 e ⎦⎥ solution 5x 5x −x ⎡1 −2 ⎤ ⎡ c1e ⎤ ⎡c1e − 2c2 e ⎤ y=⎢ ⎢ − x ⎥ = ⎢ 5x ⎥ and ⎥ ⎣1 1⎦ ⎣⎢c2 e ⎦⎥ ⎣⎢ c1e + c2 e − x ⎦⎥ 132 . k ≠ 0. so there must be an even number. ⎢ ⎥= ⎣ −2 λ − 3⎥⎦ ⎣ x2 ⎦ ⎢⎣ 0 ⎥⎦ 8 so k = i. ⎣3 5⎥⎦ Section 5. λ 0 2 ⎦ ⎣ 0 −1⎦ ⎣ The substitutions y = Pu and y ′ = Pu ′ give (d) True. real symmetric matrices have real eigenvalues. non-real complex eigenvalues occur in conjugate pairs.4 1. 1 + i) ⋅ (1. 8 8 = i 3i 3 ⎡λ − 1 −4 ⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ (λI − A)x = 0 is ⎢ . x2 = s. at least one must be real. 1. 3 ⎡ 4 −4 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ1 = 5 gives ⎢ . or (b) u ⋅ v = (k . (f) False. ⎡ c e5 x ⎤ u = ⎢ 1 − x ⎥ . Since a real 5 × 5 matrix will have 5 eigenvalues. ⎢ ⎥= ⎣ −2 2 ⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ ⎡ 4 −4 ⎤ ⎡ 1 −1⎤ reduces to ⎢ a Since ⎢ ⎥ − 2 2 ⎣0 0⎥⎦ ⎣ ⎦ general solution is x1 = s. 3i) ⋅ (i. there is no complex scalar k for which u ⋅ v = 0.3 ⎡ −2 −4 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ 2 = −1 gives ⎢ . complex eigenvalues and eigenvectors occur in conjugate pairs. if A is a matrix with real entries and complex eigenvalues and k is a real scalar. then 3ik = −8 or k = − det (λI − A) = λ 2 − 4λ − 5 = (λ − 5)(λ + 1) The eigenvalues of A are λ1 = 5 and λ 2 = −1. x2 = s or (a) False.4 Exercise Set 5. ⎣ 2 3 ⎥⎦ 27. k .

SSM: Elementary Linear Algebra Section 5. conditions is 1 y2 = 0 ⎡ −1 0 −1⎤ ⎡ x1 ⎤ ⎡ 0⎤ λ3 = 3 gives ⎢ 2 2 0 ⎥ ⎢ x2 ⎥ = ⎢ 0⎥ . ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 2 0 2 ⎦ ⎣ x3 ⎦ ⎣ 0⎦ ⎡ −1 0 −1⎤ Since ⎢ 2 2 0 ⎥ reduces to ⎢ ⎥ ⎣ 2 0 2⎦ ⎡ 1 0 1⎤ ⎢0 1 −1⎥ . x2 = s. ⎢ ⎥ ⎣0 0 0 ⎦ ⎡ x1 ⎤ ⎡ 1⎤ x2 = − s. so the solution satisfying the given initial y =0 . Then y = Pu ⎢ 3x ⎥ ⎣ c3e ⎦ gives the solution reduces to ⎡1 0 1⎤ 2⎥ ⎢ . Thus P = [p1 p 2 diagonalizes A and ⎡ λ1 0 −1 ⎢ P AP = D = 0 λ 2 ⎢ ⎣0 0 0 ⎤ ⎡1 0 0⎤ 0 ⎥ = ⎢0 2 0⎥ . ⎡ −2 λ 2 = 2 gives ⎢ 2 ⎢ ⎣ 2 ⎡ −2 0 −1⎤ Since ⎢ 2 1 0 ⎥ ⎢ ⎥ ⎣ 2 0 1⎦ ⎡ 0 −1 1⎤ p3 ] = ⎢ 1 2 −1⎥ ⎢ ⎥ ⎣ 0 2 −1⎦ 0 −1⎤ ⎡ x1 ⎤ ⎡ 0⎤ 1 0 ⎥ ⎢ x1 ⎥ = ⎢ 0⎥ . λ 2 = 2. and λ3 = 3.4 the solution of the system is y1 = c1e5 x − 2c2 e− x y2 = c1e5 x + c2 e− x 1 x1 = − s. the general solution is ⎢0 1 −1⎥ ⎢⎣0 0 0 ⎥⎦ 133 . ⎢ ⎥⎢ ⎥ ⎢ ⎥ 0 λ − 1⎦ ⎣ x3 ⎦ ⎣ 0⎦ ⎣ 2 ⎡ −3 0 −1⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ1 = 1 gives ⎢ 2 0 0 ⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . ⎥⎢ ⎥ ⎢ ⎥ 0 1⎦ ⎣ x3 ⎦ ⎣ 0⎦ ⎡ c ex ⎤ ⎢ 1 ⎥ has the solution u = ⎢c2 e2 x ⎥ . 3. x2 = s. (a) The coefficient matrix for the system is ⎡ 4 0 1⎤ A = ⎢ −2 1 0 ⎥ . ⎢ ⎥ ⎢ ⎥ ⎣ 2 0 0⎦ ⎣0 0 0 ⎦ the general solution is x1 = 0. . x3 = s or x1 = −t . ⎥ ⎢ ⎥ λ3 ⎦ ⎣0 0 3⎦ The substitutions y = Pu and y ′ = Pu ′ give the “diagonal system” u1′ = u1 ⎡1 0 0 ⎤ ⎢ ⎥ ′ u = Du = 0 2 0 u or u2′ = 2u2 which ⎢ ⎥ u3′ = 3u3 ⎣0 0 3⎦ ⎡ x1 ⎤ ⎡0 ⎤ ⎡0 ⎤ x3 = 0 or ⎢ x2 ⎥ = s ⎢1 ⎥ and p1 = ⎢1 ⎥ is a ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣0 ⎦ ⎣0 ⎦ ⎣ x3 ⎦ basis for the eigenspace corresponding to λ1 = 1. (λI − A)x = 0 is 0 −1 ⎤ ⎡ x1 ⎤ ⎡ 0⎤ ⎡λ − 4 ⎢ 2 λ − 1 0 ⎥ ⎢ x2 ⎥ = ⎢ 0⎥ . ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 2 0 0 ⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎡ −3 0 −1⎤ ⎡1 0 0 ⎤ Since ⎢ 2 0 0 ⎥ reduces to ⎢0 0 1 ⎥ . the general solution is x1 = s. x3 = 2t which is x2 ⎥ = t ⎢ 2 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 2⎦ ⎣ x3 ⎦ ⎡ −1⎤ and p 2 = ⎢ 2 ⎥ is a basis for the eigenspace ⎢ ⎥ ⎣ 2⎦ corresponding to λ 2 = 2. c − 2c2 = 0 (b) The initial conditions give 1 c1 + c2 = 0 which has the solution c1 = c2 = 0. ⎢ ⎥ ⎣ −2 0 1⎦ det(λI − A) = λ3 − 6λ 2 + 11λ − 6 = (λ − 1)(λ − 2)(λ − 3) The eigenvalues of A are λ1 = 1. x3 = − s or ⎢ x2 ⎥ = s ⎢ −1⎥ and ⎢ ⎥ ⎢ ⎥ ⎣ −1⎦ ⎣ x3 ⎦ ⎡ 1⎤ p3 = ⎢ −1⎥ is a basis for the eigenspace ⎢ ⎥ ⎣ −1⎦ corresponding to λ3 = 3. 2 ⎡ x1 ⎤ ⎡ −1⎤ ⎢ x2 = 2t .

⎢ ⎥ ⎣6 −11 6 ⎦ −1 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎡λ (λI − A)x = 0 is ⎢ . Since ⎢ ⎥= ⎣ −6 −3⎥⎦ ⎣ x2 ⎦ ⎢⎣ 0 ⎥⎦ (b) The initial conditions give −c2 + c3 = −1 c1 + 2c2 − c3 = 1 2c2 − c3 = 0 ⎡ −2 −1⎤ ⎢⎣ −6 −3⎥⎦ reduces to ⎡1 1⎤ 2 ⎥ . y1 = −c2 e2 x + c3e3 x y2 = c1e x + 2c2 e2 x − c3e3 x . ⎡ c e3 x ⎤ which has the solution u = ⎢ 1 −2 x ⎥ . ⎣3 −2⎦⎥ ⎣⎢ c2 e−2 x ⎦⎥ ⎣⎢3c1e3 x − 2c2 e−2 x ⎦⎥ 7. y2 = y ′ = y1′ and y3 = y ′′ = y2′ . ⎣6 1⎥⎦ Thus y1 = y = c1e3 x + c2 e−2 x is the solution of the original differential equation. so the solution satisfying the given y1 = e2 x − 2e3 x initial conditions is y2 = e x − 2e2 x + 2e3 x . ⎣ 0 λ 2 ⎦ ⎣0 −2 ⎦ The substitutions y = Pu and y ′ = Pu ′ give the which has the solution c1 = 1. 1 0⎤ ⎡0 A = ⎢0 0 1⎥ . c2 = −1. x2 = s or x1 = t . This gives the system det(λI − A) = λ 2 − λ − 6 = (λ − 3)(λ + 2) The eigenvalues of A are λ1 = 3 and λ 2 = −2. 3 ⎡x ⎤ ⎡1 ⎤ ⎡1 ⎤ x2 = 3t which is ⎢ 1 ⎥ = t ⎢ ⎥ . so p1 = ⎢ ⎥ is a x 3 ⎣ ⎦ ⎣ 3⎦ ⎣ 2⎦ basis for the eigenspace corresponding to λ1 = 3. y3 = 2c2 e2 x − c3e3 x ⎡ −2 −1⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ λ 2 = −2 gives ⎢ . y3 = −2e 2 x + 2e3 x 5. 2 ⎡x ⎤ ⎡ 1⎤ ⎡ 1⎤ x2 = −2t which is ⎢ 1 ⎥ = t ⎢ ⎥ so p 2 = ⎢ ⎥ x 2 − ⎣ ⎦ ⎣ −2 ⎦ ⎣ 2⎦ is a basis for the eigenspace corresponding to ⎡ 1 1⎤ λ 2 = −2. c3 = −2. then y2′ = y ′′ = y ′ + 6 y = y2 + 6 y1 which yields the system y1′ = y2 y2′ = 6 y1 + y2 The coefficient matrix of this system is ⎡0 1⎤ A=⎢ . Then for g ( x) = f ( x)e − ax we have g ′( x) = f ′( x)e− ax + f ( x)(− ae −ax ) = af ( x)e −ax − af ( x)e− ax =0 ⎡ 3 0⎤ “diagonal system” u′ = Du = ⎢ u or ⎣ 0 −2 ⎥⎦ u1′ = 3u1 u2′ = −2u2 Hence g ( x) = f ( x)e −ax = c where c is a constant or f ( x) = ceax . Use the substitutions y1 = y. ⎢ ⎥= ⎣ −6 2 ⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ ⎡1 − 1⎤ ⎡ 3 −1⎤ 3 ⎥ . Thus P = [p1 p 2 ] = ⎢ ⎣3 −2 ⎥⎦ diagonalizes A and 0 ⎤ ⎡ 3 0⎤ ⎡λ P −1 AP = D = ⎢ 1 ⎥=⎢ ⎥. Then ⎢⎣c2 e ⎥⎦ y = Pu gives the solution 3x 3x −2 x ⎤ ⎡ 1 1⎤ ⎡ c1e ⎤ ⎡ c1e + c2 e y=⎢ = ⎢ ⎥ ⎢ ⎥. x2 = s or x1 = t . the reduces to Since ⎢ ⎢ ⎣ −6 2 ⎥⎦ 0⎦ ⎣0 1 general solution is x1 = s.Chapter 5: Eigenvalues and Eigenvectors SSM: Elementary Linear Algebra x ⎤ ⎡ ⎡0 −1 1⎤ ⎢ c1e ⎥ y = ⎢ 1 2 −1⎥ ⎢ c2 e2 x ⎥ ⎢ ⎥ ⎣0 2 −1⎦ ⎢ c e3 x ⎥ ⎣ 3 ⎦ ⎡ − c e 2 x + c e3 x ⎤ 2 3 ⎢ ⎥ = ⎢c1e x + 2c2 e 2 x − c3e3 x ⎥ ⎢ ⎥ 2x 3x ⎣ 2c2 e − c3e ⎦ and the solution of the system is ⎡ 3 −1⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ1 = 3 gives ⎢ . If y1 = y and y2 = y ′. 9. Let y = f(x) be a solution of y ′ = ay. ⎢ ⎥= ⎣ −6 λ − 1⎥⎦ ⎣ x2 ⎦ ⎢⎣0 ⎥⎦ 134 . the general ⎢ 0 0 ⎣ ⎦ 1 solution is x1 = − s.

3 ⎢ ⎥ ⎢ ⎥ ⎣ −6 11 −3⎦ 0⎦ ⎣0 0 1 1 the general solution is x1 = s. x2 = s. x and y may differ by any constant vector b. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ −6 11 λ − 6 ⎦ ⎣ x3 ⎦ ⎣0 ⎦ Thus P = [p1 p 2 ⎡1 1 1 ⎤ p3 ] = ⎢1 2 3⎥ ⎢ ⎥ ⎣1 4 9 ⎦ diagonalizes A and ⎡ λ1 0 −1 P AP = D = ⎢ 0 λ 2 ⎢ ⎣0 0 ⎡ 1 −1 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ1 = 1 gives ⎢ 0 1 −1⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . ⎢ 3x ⎥ u3′ = 3u3 ⎣ c3e ⎦ Then y = Pu gives the solution x ⎤ ⎡ ⎡1 1 1 ⎤ ⎢ c1e ⎥ y = ⎢1 2 3⎥ ⎢c2 e 2 x ⎥ ⎢ ⎥ ⎣1 4 9 ⎦ ⎢ c e3 x ⎥ ⎣ 3 ⎦ ⎡ c e x + c e 2 x + c e3 x ⎤ 2 3 ⎢ 1 ⎥ = ⎢ c1e x + 2c2 e2 x + 3c3e3 x ⎥ . ⎥ ⎢ ⎥ λ3 ⎦ ⎣0 0 3⎦ The substitutions y = Pu and y ′ = Pu ′ give the ⎡1 0 0 ⎤ “diagonal system” u′ = Du = ⎢0 2 0⎥ u or ⎢ ⎥ ⎣0 0 3⎦ ⎡ x1 ⎤ ⎡1⎤ ⎡1⎤ ⎢ x ⎥ = s ⎢1⎥ so p = ⎢1⎥ is a basis for the 1 ⎢ 2⎥ ⎢⎥ ⎢⎥ ⎣1⎦ ⎣1⎦ ⎣ x3 ⎦ eigenspace corresponding to λ1 = 1. just as not every system Ax = b has a solution. reduces to Since ⎢ − 0 1 ⎢ 2 ⎣ 0 2 −1⎥⎦ ⎢ ⎥ 0⎦ ⎣0 0 1 1 s. if A is a square matrix with distinct real eigenvalues. x2 = 3t . λ 2 = 2. not every system of differential equations has a solution. but the systems y ′ = Ay and u′ = Pu do not have the same solutions. ⎡ c ex ⎤ u1′ = u1 ⎢ 1 ⎥ u2′ = 2u2 which has the solution u = ⎢c2 e2 x ⎥ . x3 = s or 0 ⎤ ⎡1 0 0⎤ 0 ⎥ = ⎢0 2 0⎥ . x2 = 2t . det(λI − A) = λ3 − 6λ 2 + 11λ − 6 = (λ − 1)(λ − 2)(λ − 3) The eigenvalues of A are λ1 = 1. and λ3 = 3. 4 2 x3 = s or x1 = t . ⎡ 3 −1 0 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ λ3 = 3 gives ⎢ 0 3 −1⎥ ⎢ x2 ⎥ = ⎢0 ⎥ . Then A is similar to a diagonal matrix P which can be used to solve the system y ′ = Ay.4 ⎡ x1 ⎤ ⎡1 ⎤ ⎡1 ⎤ ⎢ x ⎥ = t ⎢ 3⎥ so p = ⎢3⎥ is a basis for the 3 ⎢ 2⎥ ⎢ ⎥ ⎢ ⎥ ⎣9 ⎦ ⎣9 ⎦ ⎣ x3 ⎦ eigenspace corresponding to λ3 = 3. consider the case where A is diagonalizable but not diagonal. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ −6 11 −5⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎡ 1 −1 0 ⎤ ⎡ 1 0 −1⎤ Since ⎢ 0 1 −1⎥ reduces to ⎢0 1 −1⎥ . (e) False. 0 ⎤ ⎡ x1 ⎤ ⎡ 0⎤ ⎡ λ −1 ⎢ (λI − A)x = 0 is 0 λ −1 ⎥ ⎢ x2 ⎥ = ⎢ 0⎥ . ⎡ x1 ⎤ ⎡1 ⎤ ⎡1 ⎤ ⎢ x ⎥ = t ⎢ 2⎥ so p = ⎢ 2 ⎥ is a basis for the 2 2 ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 4⎦ ⎣4⎦ ⎣ x3 ⎦ eigenspace corresponding to λ 2 = 2. ⎢ x 2x 3x ⎥ ⎣c1e + 4c2 e + 9c3e ⎦ ⎡ 2 −1 0 ⎤ ⎡ x1 ⎤ ⎡ 0⎤ λ 2 = 2 gives ⎢ 0 2 −1⎥ ⎢ x2 ⎥ = ⎢ 0⎥ . ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ −6 11 −4 ⎦ ⎣ x3 ⎦ ⎣ 0⎦ ⎡1 0 − 1⎤ 4⎥ ⎢ ⎡ 2 −1 0 ⎤ 1⎥. x2 = s.SSM: Elementary Linear Algebra Section 5.4 (a) False. x2 = s. it is diagonalizable. (c) True (d) True. 9 3 x3 = s or x1 = t . x3 = 9t which is (b) False. x3 = 4t which is the general solution is x1 = Thus y1 = y = c1e x + c2 e2 x + c3e3 x is the solution of the original differential equation. True/False 5. ⎢ ⎥ ⎢ ⎥ ⎣ −6 11 −5⎦ ⎣0 0 0 ⎦ the general solution is x1 = s. 135 . ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ −6 11 −3⎦ ⎣ x3 ⎦ ⎣0 ⎦ ⎡1 0 − 1⎤ ⎡ 3 −1 0 ⎤ 9⎥ ⎢ Since ⎢ 0 3 −1⎥ reduces to ⎢0 1 − 1 ⎥ .

hence A2 − 5 A = 0 or A2 = 5 A. Suppose that Ax = λx for some nonzero ⎡ c1 x1 + c2 x2 + " + cn xn ⎤ ⎡ x1 ⎤ ⎢c x + c x + " + c x ⎥ ⎢x ⎥ n n⎥ x = ⎢ 2 ⎥ . cos2 θ < 1 hence b2 − 4ac < 0. then S12 = D. S = PS1 P −1 (b) The matrix rotates vectors in R 2 through the angle θ. ⎡ d1 0 " ⎢0 d " 2 3. x3 = ⎢ 2 ⎥ . (a) If D = ⎢ # ⎢# ⎢⎣ 0 0 " ⎡ d1 0 " ⎢ 0 d2 " S=⎢ ⎢ # # ⎢ 0 0 " ⎣⎢ 1 ⎡1 1 1 ⎤ ⎡1 0 0 ⎤ ⎡⎢ 1 −1 2 ⎤⎥ = ⎢0 1 2⎥ ⎢ 0 2 0 ⎥ ⎢ 0 1 −1⎥ ⎢ ⎥⎢ ⎥ ⎣0 0 2⎦ ⎣ 0 0 3 ⎦ ⎢⎣ 0 0 12 ⎥⎦ ⎡1 1 0 ⎤ = ⎢0 2 1 ⎥ ⎢ ⎥ ⎣0 0 3⎦ 0⎤ 0⎥ ⎥ then # ⎥ d n ⎥⎦ S 2 = A. (c) A has eigenvalues λ1 = 1. λ 2 = 4. Let S = PS1P −1 . # ⎥ ⎥ d n ⎦⎥ 9. then 150 ⎤ ⎡375 750 ⎤ = 50 ⎥⎦ ⎢⎣125 250 ⎥⎦ 750 ⎤ ⎡1875 3750 ⎤ = 250 ⎥⎦ ⎢⎣ 625 1250 ⎥⎦ 11. so the matrix has no real eigenvalues or eigenvectors. ⎢ ⎥ ⎢ ⎥ ⎣0⎦ ⎣2⎦ ⎡1 1 1 ⎤ Thus P = ⎢ 0 1 2 ⎥ diagonalizes A. ⎡3 6 ⎤ ⎡15 30⎤ A2 = 5 A = 5 ⎢ = ⎣ 1 2 ⎥⎦ ⎢⎣ 5 10⎥⎦ ⎡15 30 ⎤ ⎡ 75 150 ⎤ A3 = 5 A2 = 5 ⎢ = ⎣ 5 10 ⎥⎦ ⎢⎣ 25 50 ⎥⎦ (b) If A is a diagonalizable matrix with nonnegative eigenvalues. λ3 = 9 λ = c1 + c2 + " + cn = tr( A) or not all the xi are ⎡1 ⎤ with corresponding eigenvectors x1 = ⎢ 0 ⎥ . ⎢ ⎥ ⎣0 0 3⎦ 2 b − 4ac = 4 cos θ − 4 = 4(cos θ − 1). as required. det(λI − A) = λ − 3 −6 = λ 2 − 5λ −1 λ − 2 The characteristic equation of A is λ 2 − 5λ = 0. 0 ⎤ ⎥ 0 ⎥ . By part (a) there is a matrix S1 ⎡ 75 A4 = 5 A3 = 5 ⎢ ⎣ 25 ⎡375 A5 = 5 A4 = 5 ⎢ ⎣125 such that S12 = D. ⎢ ⎥ ⎣0⎦ equal and λ must be 0. there is a matrix P such that P −1 AP = D and ⎡ λ1 0 " 0 ⎤ ⎢0 λ " 0 ⎥ 2 D=⎢ ⎥ is a diagonal # # ⎥ ⎢# ⎣⎢ 0 0 " λ n ⎦⎥ matrix with nonnegative entries on the main diagonal. (a) det(λI − A) = λ − cos θ − sin θ sin θ λ − cos θ = λ 2 − (2 cos θ )λ + 1 Thus the characteristic equation of A is λ 2 − (2 cos θ )λ + 1 = 0. ⎢ ⎥ ⎣0 0 9 ⎦ ⎡1 0 0 ⎤ If S1 = ⎢ 0 2 0 ⎥ .Chapter 5: Eigenvalues and Eigenvectors SSM: Elementary Linear Algebra ⎡1 ⎤ ⎡1 ⎤ x 2 = ⎢ 1 ⎥ . Since 0 < θ < π. then either x1 = x2 = " = xn . Since Ax = ⎢ 1 1 2 2 # # ⎢ ⎥ ⎢ ⎥ ⎢⎣ c1 x1 + c2 x2 + " + cn xn ⎥⎦ ⎢⎣ xn ⎥⎦ S 2 = ( PS1P −1 )( PS1 P −1 ) = PS12 P −1 = PDP −1 = A. Since 0 < θ < π. and ⎢ ⎥ ⎣0 0 2⎦ Chapter 5 Supplementary Exercises 1. no vector is transformed into a vector with the same or opposite direction. 136 . For this equation 2 2 ⎡1 0 0 ⎤ P −1 AP = D = ⎢ 0 4 0 ⎥ .

e. eigenvalue of A. The matrix P = ⎢ 1 ⎢ ⎣ −1 ⎡0 −1 and P AP = D = ⎢ 0 ⎢ ⎣0 1 0⎤ −1 1⎥ will diagonalize A ⎥ 1 1⎦ 0 0⎤ 1 0 ⎥ . i. ⎡ 0 15. so all the eigenvalues of A must also be 0. then then det(λI − An ) = det(λI ) = λ k and An has characteristic polynomial λ k = 0. if Ax1 = λx1. Since A3 = A. thus ⎥ 0 −1⎦ A = PDP −1 1 1 ⎡ 0 1 0 ⎤ ⎡0 0 0 ⎤ ⎡⎢ 1 2 − 2 ⎤⎥ = ⎢ 1 −1 1⎥ ⎢0 1 0 ⎥ ⎢ 1 0 0⎥ ⎢ ⎥⎢ ⎥⎢ 1 1 ⎥ 1 1 1 0 0 1 − − ⎣ ⎦⎣ ⎦ ⎣0 2 2⎦ 0 0⎤ ⎡ 1 ⎢ −1 − 1 − 1 ⎥ =⎢ 2 2⎥ ⎢ 1 − 1 − 1⎥ 2 2⎦ ⎣ 137 . the λ3 will be an eigenvalue of A3 corresponding to the same eigenvector. But if λ1 is an must satisfy λ3 = λ and the only possibilities are −1.. 17. so the A3 x1 = λ3 x1. If A is a k × k nilpotent matrix where An = 0. 0. and 1. then λ1n is an eigenvalue of An .SSM: Elementary Linear Algebra Chapter 5 Supplementary Exercises 13. the eigenvalues eigenvalues of An are all 0. If λ is an eigenvalue of A.

− 1) (f) ku. w = (3(−1) + (−2)(6)) + (4(−1) + 5(6)) = −15 + 26 = 11 u. (4. Then AT = ⎢ 5. − 2). 1) − 3(3. (−16. (4. 0 = (4. 5). v = (0. w = u. Let A = ⎢ ⎥ ⎣1 1⎥⎦ ⎣1 1⎦ ⎡5 3 ⎤ AT A = A2 = ⎢ . (3. − 2).1 1. (a) u. w = (3(4) + (−2)(5)) + (3(−1) + (−2)(6)) = 2 − 15 = −13 138 . v = −4(2) = −8 u + v.Chapter 6 Inner Product Spaces Section 6. 0) = 4(0) + 5(0) = 0 (e) d (u.1 (d) Exercise Set 6. u = 4(3) + 5(−2) = 2 (b) (c) 0. 5) = −12(4) + 8(5) = −8 k u. w = (7. v = vT AT Au ⎡ 5 3⎤ ⎡ 2 ⎤ = [ −1 1] ⎢ ⎣3 2 ⎥⎦ ⎢⎣1 ⎥⎦ ⎡2⎤ = [ −2 −1] ⎢ ⎥ ⎣1 ⎦ = −4 − 1 = −5 (b) v. 6) = 7(−1) + 3(6) = 11 u. (−1. 2) = (−8. (0. v) = u − v = (−2. − 5) = (−8)2 + (−5)2 = 89 3. w + v. v = (−12. w = 3v. 11) = 3(3) + (−2)(11) = −13 u. ⎣3 2 ⎥⎦ v. 3). kv = (3. w = wT AT A(u + v) ⎡ 5 3⎤ ⎡ 1 ⎤ = [0 −1] ⎢ ⎣3 2⎥⎦ ⎢⎣ 2 ⎥⎦ ⎡1 ⎤ = [ −3 −2] ⎢ ⎥ ⎣2⎦ = −3 − 4 = −7 u − kv = (1. w = wT AT Av ⎡5 3⎤ ⎡ −1⎤ = [0 −1] ⎢ ⎣ 3 2 ⎥⎦ ⎢⎣ 1⎥⎦ ⎡ −1⎤ = [ −3 −2] ⎢ ⎥ ⎣ 1⎦ = 3−2 =1 (c) u + v. v = 3(4) + (−2)(5) = 2 v. 0). w = 1 ⋅ 0 + 1⋅ (−1) + 3 ⋅ 0 + 2 ⋅ (−1) = −3 (d) v = u. v + w = (3. 8). v + u. w = 9 ⋅ 0 + 6 ⋅ (−1) = −6 (c) u + v. − 20) = 3(−16) + (−2)(−20) = −8 (e) ⎡ 2 1⎤ ⎡ 2 1⎤ = A so . v = 32 + 22 = 13 = (−2)2 + (−1)2 = 5 (a) u. 5) = 0(4) + 0(5) = 0 v. (a) u. v = 1 ⋅ 3 + 1⋅ 2 = 5 (b) kv. w + v.

v = vT AT Av ⎡5 3⎤ ⎡ −1⎤ = [ −1 1] ⎢ ⎣3 2 ⎥⎦ ⎢⎣ 1⎥⎦ ⎡ −1⎤ = [ −2 −1] ⎢ ⎥ ⎣ 1⎦ = 2 −1 =1 v = ⎡ 3 11. since 5⎦ 0⎤ . v = tr(uT v) = 4 + 52 = 56 = 47 ⎡ 3 0⎤ ⎡9 0 ⎤ 9. (a) u v = ⎢ = ⎣ −2 8⎥⎦ ⎢⎣ 1 1⎥⎦ ⎢⎣10 2 ⎥⎦ T ⎡ 3 3⎤ (b) A − B = ⎢ ⎣ −5 −2 ⎥⎦ d ( A. v − w v−w 2 A = 13. B) = A − B = A − B. ⎣ 0 6 ⎥⎦ ⎡ −1⎤ ⎡ 0 ⎤ ⎡ −1⎤ (e) v − w = ⎢ ⎥ − ⎢ ⎥ = ⎢ ⎥ ⎣ 1⎦ ⎣ −1⎦ ⎣ 2⎦ v − w. v = 1 = 1 ⎡2 (b) The matrix is A = ⎢ ⎣0 ⎡4 0⎤ AT A = A2 = ⎢ . (a) AT = ⎢ = A. B) = A − B = A − B. (a) The matrix is A = ⎢ ⎣ 0 ⎡ 3 0⎤ AT A = A2 = ⎢ . so AT A = A2 = ⎢ . ⎣0 2⎥⎦ ⎣0 4 ⎥⎦ ⎡v ⎤ ⎡u ⎤ For u = ⎢ 1 ⎥ and v = ⎢ 1 ⎥ . (b) ⎡6 −1⎤ 15. v = 9(−3)(1) + 4(2)(7) = −27 + 56 = 29 (b) = 62 + (−1)2 + 82 + (−2)2 = (d ( v. w ))2 = 12 = 1 = 105 ⎡ 3 4⎤ ⎡ −1 3⎤ ⎡ 1 13⎤ 7. A − B = 1 =1 (f) u. v = tr(u v) = 1 + 2 = 3 ⎡ 1 −3⎤ ⎡ 4 6 ⎤ ⎡ 4 −18⎤ (b) uT v = ⎢ = ⎣ 2 5⎥⎦ ⎢⎣ 0 8⎥⎦ ⎢⎣ 8 52 ⎥⎦ = 32 + 32 + (−5)2 + (−2)2 u. A − B T u. p = [ p(−1)]2 + [ p(0)]2 + [ p(1)]2 + [ p(2)]2 = (−2)2 + 02 + 22 + 102 = 108 =6 3 139 . u. since 6 ⎥⎦ A. u ⎣v2 ⎦ ⎣ 2⎦ 17.SSM: Elementary Linear Algebra (d) Section 6.1 v. q = p(−1)q(−1) + p(0)q(0) + p(1)q(1) + p(2)q(2) = (−2)(2) + (0)(1) + (2)(2) + (10)(5) = −4 + 0 + 4 + 50 = 50 p = ⎡9 0 ⎤ ⎡ u1 ⎤ = [v1 v2 ] ⎢ ⎢ ⎥ ⎣0 4 ⎥⎦ ⎣u2 ⎦ ⎡u ⎤ = [9v1 4v2 ] ⎢ 1 ⎥ ⎣ u2 ⎦ = 9u1v1 + 4u2 v2 p. (a) A − B = ⎢ ⎣ 8 −2⎥⎦ d ( A. A = (−2)2 + 52 + 32 + 62 = 74 A = 0 by inspection. v − w = ( v − w)T AT A( v − w ) ⎡5 3⎤ ⎡ −1⎤ = [ −1 2 ] ⎢ ⎣ 3 2 ⎥⎦ ⎢⎣ 2 ⎥⎦ ⎡ −1⎤ = [1 1] ⎢ ⎥ ⎣ 2⎦ = −1 + 2 =1 d ( v. w ) = v − w = v − w. v = vT AT Au p. ⎣0 5⎥⎦ v. (a) 0⎤ ⎥ .

u 1/ 2 = 1 2 1 2 x + y and the unit circle is 4 16 is the ellipse shown. v) = u − v = u − v. u = u. −3) (a) d (u. y). y 4 2 –2 x –4 140 1 2 1 2 x2 y 2 x + y = 1 or + = 1 which 4 16 4 16 . u − v = (u − v)T AT A(u − v) ⎡ 2 −1⎤ ⎡ −3⎤ = [ −3 −3] ⎢ ⎣ −1 13⎥⎦ ⎢⎣ −3⎥⎦ ⎡ − 3⎤ = [ −3 −36] ⎢ ⎥ ⎣ − 3⎦ = 9 + 108 = 117 d (u. u − v = 117 = 3 13 21. u − v = (−3. u − v = 3(−3) 2 + 2(−3)2 = 45 =3 5 ⎡ 1 −1⎤ ⎡ 1 2 ⎤ ⎡ 2 −1⎤ (c) AT A = ⎢ = ⎣ 2 3⎥⎦ ⎢⎣ −1 3⎥⎦ ⎢⎣ −1 13⎥⎦ u − v. v) = u − v = u − v. v) = u − v = u − v.Chapter 6: Inner Product Spaces SSM: Elementary Linear Algebra 19. u − v = (−3)2 + (−3)2 = 18 =3 2 (b) d (u. (a) For u = (x.

y). q = ∫ (1 − x + x 2 + 5 x3 )( x − 3 x 2 )dx −1 1 = ∫ ( x − 4 x 2 + 4 x3 + 2 x 4 − 15 x5 )dx −1 1 3 5 6⎞ ⎛ 2 = ⎜ x − 4 x + x4 + 2 x − 5x ⎟ 3 5 2 ⎠ −1 ⎝ 2 29 ⎛ 1 ⎞ = − −⎜− ⎟ 15 ⎝ 15 ⎠ 28 =− 15 (b) 1 p. (a) 1 p. Let V = ⎢ . v − v. v + v. u = u. u+v 2 1 2 1 –1 2 + u−v 2 x = u + v. u + v + u − v.SSM: Elementary Linear Algebra (b) For u = (x. u − u. 141 x2 1 2 + y2 = 1 which is the 1 . then ⎣ −1 0 ⎥⎦ V . u − v − v. q = ∫ ( x − 5 x3 )(2 + 8 x 2 )dx −1 1 = ∫ (2 x − 2 x3 − 40 x5 )dx −1 1 4 6⎞ ⎛ = ⎜ x 2 − x − 20 x ⎟ 2 3 ⎠ −1 ⎝ 37 ⎛ 37 ⎞ = − −⎜− ⎟ 6 ⎝ 6 ⎠ =0 True/False 6. u + v. y 1 – 25. u + v. u + 2 v. u − v = u. u + v + v. v =2 u 2 +2 v 2 ⎡ 0 1⎤ 27. u Section 6.1 (a) True.1 1/ 2 = 2 x 2 + y 2 and the unit circle is 2 x 2 + y 2 = 1 or ellipse shown. V = 0(0) + (1)(−1) + (−1)(1) + (0)(0) = −2 < 0 so Axiom 4 fails. v = 2 u. u + v + u. u + u. u − v = u. 29. v + u. with w1 = w2 = 1. the weighted Euclidean inner product is the dot product on R 2 .

x0 = Ax0 ⋅ Ax0 = 0 ⋅ 0 = 0. B A B = 19 5 2 14 = A. B = (−3)2 + 12 + 42 + 22 2 v = 2 + 4 + (−9) = 101 = 30 142 19 10 7 . for example. 2 u v 0 = 2 2 2 v = 1 + 0 + (−3) = 10 u. (a) 2 (b) u v −10 10 20 =− 10 200 =− 1 = u v −6 6 2 1 =− 2 u. u and v are (f) True. (e) = v. i. kv = k u. 1). v = (−1)(3) + 0(8) = −3 cos θ = u v = = 50 =5 2 2 2 u = (−1) + 0 = 1 2 u. v u v = 8 4 55 2 = 55 A. A = 22 + 42 + (−1)2 + 32 73 = 30 (c) 2 2 2 u = (−1) + 5 + 2 = 30 2 2 B = B. v = 2(4) + 1(0) + 7(0) + (−1)(0) = 8 cos θ = u. Axiom 1. cos θ = (c) True. otherwise there is a nontrivial solution x0 to the system 1. using the dot product on R 2 with u = (1. u. using the dot product on R 2 with u = (0. Exercise Set 6. 0) and v = (0. (d) (d) True. then Axiom 2.. u. v = 22 + 62 + 12 + (−3) 2 v = 2 + 4 = 20 u.2 2 2 u = 1 + (−3) = 10 3.Chapter 6: Inner Product Spaces SSM: Elementary Linear Algebra u. v (b) −3 A = A. v = (−1)(2) + 5(4) + 2(−9) = 0 (b) False. u = k (k ) v. v . v 20 2 2 2 2 u = 2 + 1 + 7 + (−1) = 55 Section 6. v = 1(2) + (−3)(4) = −10 = 9 10 2 2 2 2 v = 4 +0 +0 +0 = 4 u. applying Axiom 3. v = 1(−3) + 0(−3) + 1(−3) + 0(−3) = −6 (g) False. A B. then v = 0 by Axiom (f) Ax = 0 and x0 . −1) and v = (0. B = 32 + 22 + 12 + 02 = 14 A. then Axiom 3 and Axiom 1 again gives ku. 5). v 2 2 2 2 v = (−3) + (−3) + (−3) + (−3) = 36 =6 u. if 0 = v 4. for example. v A = B = 2 v = 3 + 8 = 73 u. the matrix A must be invertible. v = 1(0) + 0(1) = 0. (a) =0 30 101 2 2 2 u = 4 + 1 + 8 = 81 = 9 cos θ = orthogonal. but x 0 ≠ 0. v = 0(0) − 1(5) = −5. kv = k kv. u. v .e. first apply Axiom 1. v = 4(1) + 1(0) + 8(−3) = −20 (e) False.2 cos θ = u v =− 2 2 2 2 u = 1 + 0 +1 + 0 = 2 cos θ = 2 u. u =k 2 u. B = 2(3) + 6(2) + 1(1) + (−3)(0) = 19 cos θ = 2 u.

w = l (1) + 5(2) + 3(3) = l + 19 Thus. −10. 0 = u. B A B = 0 =0 30 p. v = −4(8) + 2(−4) + 1(−2) = −42 = 42 2 2 2 2 2 2 u v = (−4) + 2 + 1 8 + (−4) + (−2) = 21 84 = 1764 = 42 143 .8 (c) u. (a) u. q = 1(0) + (−1)(2) + 2(1) = 0 7. u. 6). v = −3(2) + 1(−1) + 0(3) = −7 = 7 2 2 2 2 2 2 u v = (−3) + 1 + 0 2 + (−1) + 3 = 10 14 = 140 ≈ 11. (b) u. B = 2(−3) + 4(1) + (−1)(4) + 3(2) = 0 cos θ = 5. v = 2(−19) − 10(5) + 6(3) = −70 ≠ 0 Thus. (a) u. −3.SSM: Elementary Linear Algebra Section 6. v = k (k ) + k (5) + 1(6) = k 2 + 5k + 6 k 2 + 5k + 6 = (k + 2)(k + 3) u and v are orthogonal for k = −2. 0 = v. 5. 3). l = −19 and v = (−19.9 (b) u. A.2 A. v = 3(4) + 2(−1) = 10 = 10 2 2 2 2 u v = 3 + 2 4 + (−1) = 13 17 = 221 ≈ 14. v = 2(1) + 1(7) + 3k = 9 + 3k u and v are orthogonal for k = −3. 9. 11. w = 2(1) + k (2) + 6(3) = 20 + 2k Thus k = −10 and u = (2. there are no scalars such that the vectors are mutually orthogonal.

0. …. u 2 . −2t. k2 . −5. −1) or x − z = 0. u. v − v. then w must be orthogonal to every vector in span{u1 . Thus W ⊥ consists of all vectors orthogonal to every basis vector. y. u 2 . u 2 . v = w. span{u1 . 21. which is all of W. u r }. 17. u. 0. u − v − v. v = u 2 −0−0+ v 2 =2 Thus u − v = 2. u−v 2 = u − v. −1). −t) or t(1. (b) The line contains the vector u = (2. u + v. … . u 2 . z = −3t. u r } = k1u1 + k2 u 2 + + kr u r where k1. Then. z) = (t. u 2 + + kr w. … . u r }. −3t) so parametric equations for W ⊥ are x = t. −3). it is not orthogonal to the subspace. w1 = 0 by inspection. 15. (a) W ⊥ will have dimension 1. w. −2. hence v is in W ⊥ . v = 0(−1) + (−2)(−1) + 2(1) + 1(1) =5 =5 (d) 2 2 2 2 2 2 2 2 u v = 0 + (−2) + 2 + 1 (−1) + (−1) + 1 + 1 = 9 4 =6 13. so W ⊥ will consist of all scalar multiples of u or tu = (t. (c) The intersection of the planes is the line x + z = 0 in the xz plane (y = 0) which can be parametrized as (x. … . 0. k1u1 + k2u 2 + + kr u r = k1 w. u1 + k2 w. v is orthogonal to the span of the set of basis vectors. …. y = −2t. A normal to the plane is u = (1. 19. as in exercise 19. u − u. Let v ∈ span{u1. kr are arbitrary scalars. w 2 = −1(1) + 1(−1) + 0(3) + 2(0) = −2 Since u is not orthogonal to w 2 . W ⊥ is the plane with normal (1. u − v = u. u r = 0+0+ +0 =0 Thus if w is orthogonal to each vector u1 . 4). u r . 144 . If v is not orthogonal to every basis vector. thus W ⊥ is the plane through the origin with normal u.Chapter 6: Inner Product Spaces SSM: Elementary Linear Algebra u. then v clearly cannot be in W ⊥ . or 2x − 5y + 4z = 0. u − v = u. Suppose that v is orthogonal to every basis vector.

v − v. 0. .− . . ⎟. (b) (e) False. 1). for any vector w in W. ⎜ 3⎠ 2⎠ ⎝ 3 3 ⎝ 2 1 1 = +0− 6 6 =0 1 ⎞ ⎛ 1 1 ⎞ ⎛ 1 . u + v. ⎟ ⎜ ⎟. . . so ku is in W ⊥ . for any vector w in W. 1 ⎞ ⎛ 1 1 ⎞ 1 1 ⎛ 1 . 1 ⎞ 1 ⎞ ⎛ 1 1 ⎛ 1 . 3. ⎜ ⎟ =−2−2 2 2⎠ ⎝ 2 2⎠ ⎝ = −1 ≠0 The set is not orthogonal. ⎜ ⎟. v = 0 = 0. (b) True. w = u. (0. ⎜ ⎟ =−2+2 =0 2 2⎠ ⎝ 2 2⎠ ⎝ The set is orthogonal. 1 ⎞ ⎛ 1 1 ⎞ 1 1 ⎛ 1 . − ⎟ = − − 9 9 9 3⎠ ⎝ 3 3 3 ⎠ =0 2 4 2 2 1⎞ ⎛1 2 2⎞ ⎛2 ⎜ . ⎟ = − + = 0 9 9 9 3 3⎠ ⎝3 3 3⎠ ⎝3 2 2 4 2⎞ ⎛1 2 2⎞ ⎛2 1 ⎜ .3 BC = v − u. 3 ⎝3 4 2 2 1⎞ ⎛ 2 1 2⎞ ⎟. ku. if u is orthogonal to every vector of a subspace W. 0. ⎜ − ⎟ 2⎠ ⎝ 2 2⎠ ⎝ 2 1 1 = − +0+ 2 2 =0 1 ⎞ ⎛ 1 1 ⎞ ⎛ 1 1 . 2 ⎛2 ⎜ . . → 1. − ⎟. if u and v are orthogonal u. AB = v + u and → Exercise Set 6. + v2 ≠ u + v 145 . if u and v are orthogonal. Thus. so the angle at B is a right angle. ⎜− ⎟. ⎜ . w + v. ⎜ ⎟. so u + v is in W ⊥. BC = v + u. 0) = 0 + 0 = 0 The set is orthogonal. Use the Euclidean inner product on R 2 and note that u = v . W ∩ W ⊥ = {0}. v − u. then u is in W ⊥ .3 27. (c) 31. . ⎜ − ⎟ 3⎠ ⎝ 2 2⎠ ⎝ 3 3 1 1 =− +0− 6 6 2 =− ≠0 6 The set is not orthogonal. ⎜− ⎟. w = 0. v − u = v. W ⊥ is the xz-plane. 0. u + u. (0.2 (a) False. u+ v 2 = u 2 + v 2 thus u+v = u 2 (0. (f) False. (a) (c) W ⊥ is the line through the origin which is perpendicular to the yz-plane. (a) → AB. − .− . True/False 6. w = k u. . u (b) = v2− u2 =0 → → Thus AB and BC are orthogonal. ⎟ = + − = 0 9 9 9 ⎝3 3 3⎠ ⎝3 3 3⎠ The set is orthogonal.− . (d) (b) W ⊥ is the plane through the origin which is perpendicular to the y-axis. Thus W ⊥ is the line y = −x. (a) W ⊥ is the line through the origin which is perpendicular to y = x.3 → Section 6. (d) True. (c) True. Thus W ⊥ is the x-axis. 1) = 0 + 0 = 0 The set is orthogonal.− . . 0). w = k (0) = 0. ⎜ . ⎜ . Using the figure in the text. 0. (2.SSM: Elementary Linear Algebra Section 6.

⎜ 0.Chapter 6: Inner Product Spaces (c) SSM: Elementary Linear Algebra 2⎛1⎞ 1⎛ 2⎞ 2⎛ 2⎞ p2 ( x). 2) = 2 + 0 − 2 = 0 (1. 3) (6. ⎜ 6 6 6 2 2 ⎝ ⎠ ⎝ 1 1 = − +0 12 12 =0 The set is orthogonal. (b) (1. 0. 2 5. (d) 1 2 ⎞ ⎛ 1 1 ⎛ 1 . . 1 1 ⎞ ⎛ . 0. 0. (2. 0) = 0 + 0 + 0 = 0 (2. ⎜ ⎟. p3 ( x) = 0(0) + = 1 2 (0) + 1 2 (1) 1 ≠0 2 The set is not an orthonormal set. p2 ( x) = 1(0) + 0(1) + 0(1) = 0 p1 ( x). (0. 2) 5 5⎠ 5 ⎝ 2 (6. (0. 3) = −6 + 6 = 0 (−1. − 1) = 12 + 02 + (−1)2 = 2 146 . 0. 3) 3 5 3 ⎞ ⎛ 6 .− . 5.− . 2) (−1. − 1). 3) = 62 + 32 = 45 = 3 5 ⎛1⎞ ⎛ 2⎞ ⎛ 2⎞ p3 ( x) = ⎜ ⎟ + ⎜ ⎟ + ⎜ ⎟ ⎝ 3⎠ ⎝ 3⎠ ⎝ 3⎠ 1 4 4 = + + 9 9 9 = 1 =1 2⎛ 2⎞ 2⎛1⎞ 1⎛ 2⎞ p1 ( x). ⎟ . 2) ⎛ 1 . = = ⎜− ⎟ (−1. 1) = 0 + 0 + 2 2 2 ⎝ ⎠ 1 = ≠0 2 The set is not orthogonal. p3 ( x) = 1(0) + 0(0) + 0(1) = 0 p2 ( x). 1) = 0 + 0 + 0 = 0 1 1 ⎞ 1 ⎛ . The vectors ⎜ − ⎟ and ⎜ ⎟ 5 5⎠ 5⎠ ⎝ ⎝ 5 are an orthonormal set. . p3 ( x) = ⎜ ⎟ − ⎜ ⎟ + ⎜ ⎟ 3⎝ 3⎠ 3⎝ 3⎠ 3⎝ 3⎠ 2 4 2 = − + 9 9 9 =0 (6. (a) (−1. (1. (0. (0. p2 ( x) = ⎜ ⎟ − ⎜ ⎟ + ⎜ − ⎟ 3⎝ 3⎠ 3⎝ 3⎠ 3⎝ 3⎠ 4 2 2 = − − 9 9 9 =0 2⎛1⎞ 2⎛ 2⎞ 1⎛ 2⎞ p1 ( x). 2). =⎜ ⎟ 5⎠ ⎝ 5 2 ⎞ ⎛ 1 ⎛ 2 1 ⎞ . ⎟ = 0+0+0 = 0 2 2⎠ ⎝ (1. 0. 3) = (6. ⎜ 0. 0). (a) ⎛ 2⎞ ⎛ 2⎞ ⎛1⎞ p1 ( x) = ⎜ ⎟ + ⎜ − ⎟ + ⎜ ⎟ ⎝ 3⎠ ⎝ 3⎠ ⎝3⎠ 4 4 1 = + + 9 9 9 = 1 =1 2 2 2 2 2 ⎛ 2⎞ ⎛1⎞ ⎛ 2⎞ p2 ( x) = ⎜ ⎟ + ⎜ ⎟ + ⎜ − ⎟ ⎝ 3⎠ ⎝3⎠ ⎝ 3⎠ 4 1 4 = + + 9 9 9 = 1 =1 2 2 (b) p1 ( x) = 12 = 1 2 ⎛ 1 ⎞ ⎛ 1 ⎞ p2 ( x) = ⎜ ⎟ +⎜ ⎟ ⎝ 2⎠ ⎝ 2⎠ 1 1 = + 2 2 = 1 =1 ⎞ 0⎟ ⎠ 2 p3 ( x) = 12 = 1 p1 ( x). 2). 7. 0. p3 ( x) = ⎜ ⎟ + ⎜ ⎟ − ⎜ ⎟ 3⎝ 3⎠ 3⎝ 3 ⎠ 3⎝ 3⎠ 2 2 4 = + − 9 9 9 =0 The set is an orthonormal set. (6. − 1). =⎜ ⎟ ⎝3 5 3 5⎠ ⎛ 2 1 ⎞ . 0). 0. 2) = (−1) 2 + 22 = 5 2 ⎞ (−1. 5. 0. 0) = 0 + 0 + 0 = 0 (1. 0.

0⎟ = − + + 0 10 10 ⎝5 5 5⎠ ⎝ 2 2 ⎠ =0 1 1 2 2⎞ ⎛1 1 1⎞ ⎛1 1 ⎜ . 5. 1. . − The vectors ⎜ ⎟. ⎟ 1. ⎟.− =⎜ ⎟ 6 6⎠ ⎝ 6 ⎛ 1 1 1 ⎞ . ( (3 1 1 ⎛1 1 1⎞ ⎛ 1 1 ⎞ ⎜ . − = =⎜ ⎟ (1. 1. − 1) ⎛ 1 .SSM: Elementary Linear Algebra Section 6. 2⎠ ⎝ 2 1 ⎞ ⎛ 1 . 1. − ⎟ = + − 15 15 15 ⎝5 5 5⎠ ⎝3 3 3⎠ =0 1 1 2⎞ ⎛ 1 1 ⎞ ⎛1 1 ⎜ − . 12 . . The vectors ⎜ ⎟. 0. 0. ⎟.3 1 ⎞ (1. 5. . . ⎜ . 0. 0) ( − 12 . . 0. 2) (2. 0⎟ = ⎜− 2 2 ⎠ ⎝ 2 (0. 12 . 2 ⎛ 3⎞ ⎛ 4⎞ v1 = ⎜ − ⎟ + ⎜ ⎟ + 02 ⎝ 5⎠ ⎝ 5⎠ 9 16 = + 25 25 = 1 =1 2 2 ⎛ 4⎞ ⎛ 3⎞ v 2 = ⎜ ⎟ + ⎜ ⎟ + 02 ⎝ 5⎠ ⎝5⎠ 16 9 = + 25 25 = 1 =1 ⎛ 1 1 1 ⎞ . 0) = 0 + 5 + 0 = 5 (c) 2 ⎛ 1 1 ⎞ ⎛ 1⎞ ⎛1⎞ ⎜ − . 2 2 9. . ⎜ − . 0⎟. 1. − 1) 2⎠ 2 ⎝ 2 2 (2. 0) are an 2 2⎠ ⎝ orthonormal set. 2) 2 2 2 ⎞ ⎛ 2 . 1. 0⎟ 1 ⎝ 2 2 ⎠ 2 (0. 2) = (2. 0) 5 (0. − 2 6 ⎝3 1 ⎞ ⎛ 1 . . − ⎟ = − + + 0 6 6 ⎝ 2 2 ⎠ ⎝3 3 3⎠ =0 2 2 ( (5 5 ) 5) 2⎛ 1 1 ⎞ ⎜ − . 0) = = (0. and (0. =⎜ ⎟ ⎝ 3 3 3⎠ v3 = 02 + 02 + 12 = 1 v1 . . 0⎟ = ⎜ − ⎟ + ⎜ ⎟ + 0 ⎝ 2 2 ⎠ ⎝ 2⎠ ⎝ 2⎠ 1 = 2 1 = 2 ) 3) 3 2 ⎛1⎞ ⎛ 2⎞ +⎜ ⎟ +⎜− ⎟ ⎝3⎠ ⎝ 3⎠ 2 1 2⎞ . 5. 0. 2) = 22 + 02 + 22 = 8 = 2 2 (2. . ⎝ 3 3 3⎠ 1 1 2 ⎞ ⎛ 1 ⎛ 1 ⎞ . 0. 0) (0. 0 ⎟ and ⎜ . 1. 0. 1 5 ⎛1 1 1⎞ 5 5 5 = ⎜ . .− ⎟ 3 3⎠ 1 2 ⎞ ⎛ 1 . 0. v 2 = − 147 12 12 + +0 = 0 25 25 . =⎜ ⎟ 2 2⎠ ⎝2 2 1 ⎞ ⎛ 1 . . 0) ⎛1 1 1⎞ ⎛1⎞ ⎛1⎞ ⎛1⎞ ⎜ . 5. =⎜ ⎟ 2⎠ ⎝ 2 2 2 ( − 12 . 0. 0) 2⎞ ⎛1 1 ⎛1⎞ ⎜ . . 0. ⎜ ⎟ . . 0. . 0.− ⎟ = ⎜ ⎟ ⎝3 3 3⎠ ⎝ 3⎠ 6 = 9 6 = 3 1. . − 2 3 ⎛1 3 3 3 = ⎜ . 1. 1 3 ⎝5 5 5⎠ = 1 ⎛ 1 ⎞ . ⎟ = ⎜ ⎟ +⎜ ⎟ +⎜ ⎟ ⎝5 5 5⎠ ⎝5⎠ ⎝5⎠ ⎝5⎠ 3 = 25 3 = 5 1. − 1) (1.− ⎟ are ⎜− 2 2 6 6 6⎠ ⎝ ⎠ ⎝ an orthonormal set. . ⎜ . .

2) = − v1 + v 2 + 2 v3 . it must be an orthogonal basis for R 4 . v 2 = 2 − 2 − 12 + 12 = 0 v1 . (a) u = (1. − . v3 = 3 + 8 + 3 − 14 = 0 (b) u = (3. v1 = − − + 0 = − = − 35 35 35 7 4 9 5 1 − +0= − = − u. − . v3 = 0 + 0 + 4 = 4 u. v1 = −1 − 4 + 9 − 28 = −24 7 1 Thus. u1 = w. ⎜ .Chapter 6: Inner Product Spaces SSM: Elementary Linear Algebra v1. w = u1 − u 2 − u3 . (3. 4) = − v1 (b) 4 10 14 +0+ = 3 3 3 2 10 8 w. w = 0u1 + v1 . u 2 = + 0 − = − 3 3 3 4 5 1 w . − 1. v3 = 0 + 0 + = 7 7 3 1 5 3 5⎞ ⎛1 Thus. v 4 = −12 + 12 + 2 − 2 = 0 Thus. −1. −7. v 4 = 4 − 6 + 6 − 4 = 0 + 11 6 4 66 5 6 + 2 2 = 5 14 = 6 + u2 + =0 11 66 11 6 66 11 66 u3 . 7 7 7 7 7⎠ ⎝7 11. ⎟ 7 7⎠ ⎝7 3 12 15 3 u. 3 3 3 w. v3 = −3 − 8 + 3 + 8 = 0 3 w. v4 = 8 + 3 − 8 − 3 = 0 v3 . 2) 3 4 7 u. Since S contains 4 linearly independent orthogonal vectors in v 2 . v 4 = −4 + 6 + 6 + 7 = 15 −24 −33 0 15 u= v1 + v 2 + v3 + v 4 30 30 30 30 4 11 1 = − v1 − v 2 + 0 v3 + v 4 5 10 2 37 9 v1 − v 2 + 4 v3 . v3 = 0 + 0 + 0 = 0 R 4 . ⎟ = − v1 − v 2 + v3 . v 2 = − + 0 = 5 5 5 u. u 2 = v1 . v 2 = 35 35 35 7 5 5 u. v 3 = −6 + 4 − 4 + 6 = 0 v2 . (a) 2 u. u3 = + 0 − = − 3 3 3 14 8 1 Thus. − 7. v1 = − − + 0 = − 5 5 5 4 3 1 u. v 2 = −2 + 2 − 12 − 21 = −33 u. v 2 . v3 = 0 + 0 + 0 = 0 linearly independent. (a) 3 5⎞ ⎛1 (c) u = ⎜ . u3 = 11 1 6 1 v 2 . v 4 } is a set of nonzero orthogonal vectors in R 4 . (1. 5 5 13. v 2 = − + 0 = − 5 5 5 u. since S = {v1 . S is 148 1 2 + + 66 − Thus. 5 5 Thus. u1 = − w. . v1 = − − + 0 = − 5 5 5 12 21 9 u. 4) 9 28 37 u. v3 . v3 = 0 + 0 + 2 = 2 (b) v2 v3 v4 = 12 + (−2)2 + 32 + (−4)2 = 30 2 = 22 + 12 + (−4) 2 + (−3)2 = 30 2 = (−3) 2 + 42 + 12 + (−2)2 = 30 2 = 42 + 32 + 22 + 12 = 30 u.

− . . − ⎟ 18 18 ⎠ ⎝ 18 18 18 ⎠ ⎝ 3 3 3 ⎠ ⎝ 18 1 17 ⎞ ⎛ 23 11 =⎜ . 0. − ⎟ 4 4⎠ ⎝4 4 4 4⎠ ⎝4 4 4 4⎠ ⎝4 4 3 5⎞ ⎛7 5 =⎜ . − . . ⎟ + ⎜ . v3 = +0+0+ = 18 18 18 3 5 projW x = v1 + 2 v 2 + v3 18 18 12 3 ⎞ ⎛ 5 1 1⎞ ⎛ 5 5 20 ⎞ ⎛ 3 = ⎜ 0. − ⎟ + ⎜ . − ⎟ 4 4⎠ ⎝4 4 projW x = (b) v1 v2 v3 2 = 02 + 12 + (−4)2 + (−1) 2 = 18 2 = 32 + 52 + 12 + 12 = 36 2 = 12 + 02 + 12 + (−4) 2 = 18 x. . − . . v1 = 0 + 2 + 0 + 2 = 4 x. − . ⎟ + ⎜ . . . v3 = 1 − 2 + 0 + 2 = 1 1 5 1 v1 + v 2 + v3 4 4 4 5 5⎞ ⎛1 1 1 1⎞ ⎛1 1 1 1⎞ ⎛5 5 = ⎜ . − 2 ⎟ 9 9 ⎠ ⎝ 36 36 36 36 ⎠ ⎝ 2 2 ⎝ 9 ⎠ 1 23 ⎞ ⎛ 17 7 =⎜ . − ⎟+⎜ . . ⎟+⎜ . v 2 = + + 0 − = 2 2 6 6 1 4 5 x. (a) v1 v2 v3 2 = 12 + 12 + 12 + 12 = 4 2 = 12 + 12 + (−1)2 + (−1) 2 = 4 2 = 12 + (−1)2 + 12 + (−1)2 = 4 Section 6. . v 2 = 3 + 10 + 0 − 2 = 11 x. − ⎟ ⎝ 12 4 12 12 ⎠ projW x = 17. . . − ⎟ + ⎜1. . − . v1 = 1 + 2 + 0 − 2 = 1 x. − ⎟ 18 18 ⎠ ⎝ 18 6 x. v1 = 0 + 2 149 . − . . 0. (a) 1 3 +0+ = 18 18 18 1 10 1 x.3 x. v3 = 1 + 0 + 0 + 8 = 9 4 11 9 v1 + v 2 + v3 18 36 18 2 11 1 = v1 + v 2 + v3 9 36 2 2 ⎞ ⎛ 33 55 11 11 ⎞ ⎛ 1 1 ⎛ 2 8 ⎞ = ⎜ 0. v 2 = 1 + 2 + 0 + 2 = 5 x. .SSM: Elementary Linear Algebra 15. . . − .

− 1). so ⎝2 2 ⎠ ⎛ 1 1 ⎞ w 2 = x − projW x = ⎜ − . ⎝ 4 4 4 4⎠ –1 q2 q1 1 x (b) First. − . . 0) q1 = 1 = = (1. v 2 }. . v3 = − 1 + 0 + = 0 2 2 projW x = 1v1 + 2 v 2 + 0 v3 ⎛1 1 1 1⎞ = ⎜ .Chapter 6: Inner Product Spaces (b) SSM: Elementary Linear Algebra 1 1 +1+ 0 − = 1 2 2 1 1 x. v2 5 y ⎛ 12 ⎞ ⎛ 4 ⎞ v2 = ⎜ ⎟ + ⎜ ⎟ ⎝ 5 ⎠ ⎝5⎠ 160 = 25 4 10 = 5 The orthonormal basis is 1 –1 150 –1 q1 q2 1 x . =⎜ ⎟ ⎝ 10 10 ⎠ = 5 y 1 –1 (b) From Exercise 15(a). − 3) ⎛ 1 . − 1⎟ . − 1. ⎟ ⎝ 5 5⎠ 2 v1 3 ⎞ (1. v1 v1 2 v1 3+0 = (3. 0) v1 1 q2 = 2 v2 (0. − 5) v1 = u1 = (1. − 5) − (1. ⎟ 4 10 ⎝ 5 5 ⎠ 1 ⎞ ⎛ 3 . ⎟ ⎝ 5 5⎠ ⎛ 12 4 ⎞ =⎜ . − 1⎟ . − 5) = = (0. . 3 5⎞ ⎛7 5 w1 = projW x = ⎜ . 2) − (1. − 3) v1 = 12 + (−3) 2 = 10 u 2 . v1 10 ⎠ 10 ⎝ 10 = 19. 54 ) 4 10 5 ⎛ 12 4 ⎞ ⎜ . ⎟ + (1. v2 = u2 − u 2 . − ⎟ ⎝2 2 2 2⎠ x. − ⎟ . ⎛3 3 ⎞ w1 = projW x = ⎜ . v1 = q2 = v2 v2 ( 125 . (a) From Exercise 14(a). . transform the given basis into an orthogonal basis {v1 . transform the given basis into an orthogonal basis {v1 . 0) v1 = 12 + 02 = 1 21. 0) = (0. − 1) ⎝2 2 2 2⎠ 1 1⎞ ⎛3 3 =⎜ . 2) − ⎜ − . − 3) 10 ⎛ 2 6⎞ = (2. 1. . . ⎝ 2 2 ⎠ v2 = u2 − q1 = v 2 = 02 + (−5)2 = 5 The orthonormal basis is v (1. v 2 = + 1 + 0 + = 2 2 2 1 1 x. − 1. (a) First. v1 = u1 = (1. − 5) − (3. v1 v1 2 v1 2−6 = (2. . v 2 }. . − . 1. 0) 12 = (3.− = =⎜ ⎟. so ⎝4 4 4 4⎠ 3⎞ ⎛ 3 3 3 w 2 = x − projW x = ⎜ − . − ⎟ .

2. . .v v 4 = u 4 − 4 1 v1 − 4 2 v 2 − 4 3 v3 2 2 2 v1 v2 v3 = (1. . − 1. . − 1. v 2 . 0. 0) − ⎜1. 0 ⎟ − ⎜ . . 0 ⎟ 6 5 5 5 ⎠ ⎝ 5 2 = (1. 0) v1 = 02 + 22 + 12 + 02 = 5 v2 = u2 − u 2 . 1. 1) − 1 0+0+0+0 1+ 0 + 0 + 0 ⎛ 1 2 ⎞ 2 + 0 + 0 −1⎛ 1 1 ⎞ v1 − − − . − 1) − 1 1 ⎞ ⎛ 8 4 ⎞ ⎛1 = (1. v 2 v2 2 v2 1− 5 + 0 + 0 ⎛ 0+4+0+0 1 2 ⎞ (0. 0. 1. 2. . 0 ⎟ ⎝ 5 5 ⎠ 1 2 ⎞ ⎛ = ⎜1. − 1) − ⎜ 0. 0. v1 = u1 = (0. transform the given basis into an orthogonal basis {v1 . 0) 5 ⎛ 4 2 ⎞ = (1. − 1⎟ 6 5 5 5 5 ⎠ ⎝ ⎠ 2 2 ⎝ 5 2 1 1 ⎞ ⎛ 1 1 1 1⎞ ⎛5 = (1.v u . 1 0 ⎜ ⎟ ⎜ . ⎟ 6 3 ⎠ ⎝ 10 10 5 5 ⎠ ⎝6 8 4⎞ ⎛ 4 4 =⎜ . − . 2. 2. − . 0. 1) − ⎜ . v1 v1 v1 − 2 u3 . 0. . v3 . − . − 1. − 1⎟ ⎝2 2 ⎠ 2 2 ⎛1⎞ ⎛1⎞ v3 = ⎜ ⎟ + ⎜ ⎟ + (−1)2 + (−1) 2 ⎝2⎠ ⎝2⎠ 5 = 2 10 = 2 u . − .v u . . . . − . 0 ⎟ ⎝ 5 5 ⎠ ⎝ 2 10 5 ⎠ ⎛1 1 ⎞ = ⎜ . 0) − 2 2 30 30 ⎛ 1⎞ ⎛2⎞ v 2 = 12 + ⎜ − ⎟ + ⎜ ⎟ + 02 = = 25 5 ⎝ 5⎠ ⎝5⎠ v 3 = u3 − u3 . 0 ⎟ − ⎜ − . . First. . 0) + ⎜ 0. 1. − 1. 2. 0. v1 v1 2 v1 0−2+0+0 (0. . 0. v 4 }. . 0.3 23. 0 ⎟ 5 5 ⎠ ⎝ = (1. ⎟ ⎝ 15 15 15 5 ⎠ 2 2 2 ⎛ 4 ⎞ ⎛ 4 ⎞ ⎛ 8 ⎞ ⎛4⎞ v4 = ⎜ ⎟ + ⎜ ⎟ + ⎜ − ⎟ + ⎜ ⎟ ⎝ 15 ⎠ ⎝ 15 ⎠ ⎝ 15 ⎠ ⎝ 5 ⎠ 240 = 225 4 = 15 2 151 . − .SSM: Elementary Linear Algebra Section 6.

⎜ ⎟. 0⎟ ⎝3 3 ⎠ 152 . 1. 1. 1) = (1. − ⎟ 6⎠ ⎝6 6 6⎠ ⎝6 6 ⎛2 1 ⎞ = ⎜ . − ⎟ 2⎠ ⎝2 2 v2 = v2 . . − 15 . ⎟ − ⎜ . . − 1.Chapter 6: Inner Product Spaces The orthonormal basis is q1 = ( SSM: Elementary Linear Algebra 2 1 v1 (0.− ⎜ ⎟.v u .v v3 = u3 − 3 1 v1 − 3 2 v 2 2 2 v1 v2 1 +0+0 1+ 0 + 0 1⎞ ⎛1 1 = (1. 0) − (1. . 0 ⎛ 5 v2 1 . 1. 0⎟. 0⎟. 25 . 1) 2 1⎞ ⎛1 1 =⎜ . .− . q2 = = =⎜ 30 v2 30 ⎝ 30 q3 = v3 = v3 q4 = v4 = v4 ( 5 1 . v3}. 1) v1 = v1 . − 158 . 0) ⎛ ⎞ .− . v2 2 2 2 ⎛1⎞ ⎛1⎞ ⎛ 1⎞ = ⎜ ⎟ + 2 ⎜ ⎟ + 3⎜ − ⎟ ⎝2⎠ ⎝2⎠ ⎝ 2⎠ ⎛1⎞ = 6⎜ ⎟ ⎝4⎠ 6 = 2 u . 0. 30 ⎠ 2 ) =⎛ −1 1 1 2 2 ⎞ . 1. 1) − 2 ⎜ . .v v 2 = u 2 − 2 1 v1 2 v1 1(1) + 2(1)(1) + 3(0)(1) (1. 54 ) = ⎛ 4 15 1 1 2 3 ⎞ . 1 . v1 = 12 + 2(1)2 + 3(1) 2 = 1+ 2 + 3 = 6 u . 0) − (1. 0.− ⎟ 6 6 ⎝2 2 2⎠ 4 1⎞ ⎛1 1 1⎞ ⎛1 1 = (1. v1 5 5 ⎠ 5 ⎝ ) 1. 1. 0) − 6 1 = (1. 15 15 ⎠ ⎝ 15 15 25. v1 = u1 = (1. v 2 . transform the given basis into an orthogonal basis {v1 . . = = ⎜ 0.− . 0) − ⎜ . 1. 2 2 10 2 ⎞ . − . 154 . 1. . First. 10 10 ⎠ ⎝ 10 10 ( 154 . 2.

− = 2 2 =⎜ ⎟. 2) − ⎜ . 3) − (1. 1. u 2 }. q1 u2 . − . 3) − ⎜ . ⎜ 6 6 ⎝ ⎠ 27. 2. − . Let W = span{u1 . 2) ⎛ 1 . (a) A = ⎢ ⎣ 2 3⎥⎦ Since det(A) = 3 + 2 = 5 ≠ 0. 1. ⎟ ⎝ 14 14 7 ⎠ 3 1⎞ ⎛1 =⎜ . 1. = =⎜ ⎟ v1 5⎠ 5 ⎝ 5 q2 = v2 (−2. 0⎟. = = ⎜− ⎟ v2 5 5⎠ 5 ⎝ u1 . v1 6 6⎠ 6 ⎝ 6 q2 = q3 = ( = 9 9 ⎛5 1 4⎞ = 2(1. 2 v1 = u1 = (1. . . q 2 153 1 4 = 5 5 1 6 =− + = 5 5 5 2 3 = + = 5 5 5 5 + . ⎟ ⎝ 14 14 7 ⎠ w 2 = w − w1 ⎛ 13 31 20 ⎞ = (1. =⎜ q1 = 1 = ⎟.− . 0 2. v1 v1 = (2. ⎟ ⎝3 3 3⎠ 4⎞ ⎛5 1 =⎜ . 1) ⎛ 1 . 2) v1 = 12 + 22 = 5 v2 = u2 − 2 2 u 2 . − 1) − ⎜ . ⎟ ⎝ 14 14 7 ⎠ ) 1. v3 2 2 ⎛2⎞ ⎛ 1⎞ = ⎜ ⎟ + 2 ⎜ − ⎟ + 3(0) 2 ⎝3⎠ ⎝ 3⎠ 4 2 = + 9 9 6 = 3 The orthonormal basis is 1 1 ⎞ v (1. Let u1 = (1. q1 = u 2 . 0. − ⎟ ⎝3 3 3⎠ {v1 . 6 v2 6 6⎠ ⎝ 6 v3 = v3 ( 2 − 13 . − ⎟ 14 ⎝ 3 3 3 ⎠ 3 6⎞ ⎛ 15 = (2. 1) − ⎜ . 3 6 3 ) =⎛ 2 1 ⎞ . 1) v1 = 12 + 12 + 12 = 3 ⎛5⎞ ⎛ 1⎞ ⎛ 4⎞ v2 = ⎜ ⎟ + ⎜ − ⎟ + ⎜ − ⎟ ⎝ 3⎠ ⎝ 3⎠ ⎝ 3⎠ 42 = 9 42 = 3 5 − 2 − 12 1+ 2 + 3 v1 + 3 3 3 v 2 42 3 2 v 2 = (−2)2 + 12 = 5 q1 = v1 2 ⎞ (1. − 1) − ⎡ 1 −1⎤ 29. A is invertible. Apply the Gram-Schmidt process with normalization to u1 and u 2 . v1 2 2 + 0 −1 (1. 1. 2) = (−2. v1 w.SSM: Elementary Linear Algebra v3 = Section 6. 1) ⎛ 2 1 ⎞ . so it has a QR-decomposition. . − . 2) = (−1. 2. − 1 1 1 ⎞ v2 ⎛ 1 2 . 1) v2 = u2 − u 2 . v 2 } is an orthogonal basis for W. 1.3 w1 = projW w w. 3). − ⎟ 7⎠ ⎝ 14 14 ⎛ 13 31 20 ⎞ =⎜ . . . 2) and u 2 = (−1. 1) 1 + 12 + 12 ⎛1 1 1⎞ = (2. v1 = u1 = (1. 0. v1 v1 2 v1 −1 + 6 (1. v 2 v1 + v2 = 2 2 v1 v2 v3 . − . 3) − 5 = (−1.

⎟ q1 = 3 v1 ⎝3 3 3⎠ = (2. q 2 = 234 234 234 26 = 3 26 26 = 3 =3 2 1 4 + + = 3 3 3 3 The QR-decomposition of A is ⎡ 1 2⎤ ⎡u . 2) and u 2 = (1. 1) and v1 = u1 = (1. 1. 0. v1 = u1 = (1. =⎜ ⎟ 234 234 ⎠ ⎝ 234 1 4 4 u1 . 1. 2) ⎛ 1 2 2⎞ = =⎜ . 1) − ⎜ . 1. the column vectors of A are linearly independent. so A has a QR-decomposition. ⎟ 9 9⎠ ⎝9 ⎛ 8 11 7 ⎞ =⎜ . 4) − 1 u 2 . = = ⎜− ⎟ v2 3 3 3⎠ 3 ⎝ u1 . 1 ⎥⎢ 0 ⎢ 5⎦ ⎣ 5 5 ⎣ ⎦ ⎡ 1 2⎤ (b) A = ⎢0 1⎥ ⎢ ⎥ ⎣ 1 4⎦ By inspection. 1) v1 = 12 + 02 + 12 = 2 v2 = u2 − u 2 . q1 = 2 2 2 +0+ +0+ 1 2 4 2 q2 = = = 2 2 v2 v2 ( 89 . 1. 1) 2 = (2. 0. Apply the Gram-Schmidt process with normalization to u1 and u 2 . the column vectors of A are linearly independent. 0. 4). Apply the Gram-Schmidt process with normalization to u1 and u 2 . − 2. − 2.q u 2 . 1. so A has a QR-decomposition. q 2 ⎦ ⎢ ⎥ ⎣ 1 4 ⎣ ⎦ ⎡ 1 − 1 ⎤ 3⎥ ⎢ 2 1 ⎥ ⎡ 2 3 2⎤ =⎢ 0 ⎥ 3⎥⎢ 0 ⎢ 3⎦ 1 ⎥⎣ ⎢ 1 ⎢⎣ 2 3 ⎥⎦ u2 . q1 ⎤ ⎢0 1⎥ = [q1 q 2 ] ⎢ 1 1 ⎥ 0 u2 . 1. Let u1 = (1. q 2 = − 2 ⎛ 8 ⎞ ⎛ 11 ⎞ ⎛ 7 ⎞ v2 = ⎜ ⎟ + ⎜ ⎟ + ⎜ ⎟ ⎝9⎠ ⎝ 9 ⎠ ⎝9⎠ 234 = 81 234 = 9 v1 (1. 1) ⎛ 1 1 1 ⎞ . 0. 3) = (−1. q1 ⎤ ⎡ 1 −1⎤ ⎥ ⎢⎣ 2 3⎥⎦ = [q1 q 2 ] ⎢ 0 u2 . q1 = u 2 . − 2. − 2. = =⎜ ⎟ v1 2⎠ 2 ⎝ 2 q2 = v2 (−1. q1 u 2 .Chapter 6: Inner Product Spaces SSM: Elementary Linear Algebra ⎡ 1 1⎤ (c) A = ⎢ −2 1⎥ ⎢ ⎥ ⎣ 2 1⎦ By inspection. 2) 9 2 2⎞ ⎛1 = (1. 1) ⎛ 1 . 1. v1 v1 2 v1 2 2+0+4 (1. . 0. . 1. 2) v1 = 12 + (−2)2 + 22 = 3 v2 = u2 − v1 2 v1 1− 2 + 2 = (1. − . v1 2 154 . 97 ) 234 9 11 7 ⎞ ⎛ 8 . q 2 ⎦ ⎣ ⎡1 − 2 ⎤ 5⎤ 5⎥⎡ 5 = ⎢ 25 ⎥. q1 = − + = 3 3 3 3 8 11 7 + + u2 . ⎟ ⎝9 9 9⎠ u 2 = (2. q1 = + + = 3 3 3 3 1 2 2 1 u 2 . 0. The QR-decomposition of A is ⎡ u1 . 1) − (1. 1). − . . Let u1 = (1. 1) v 2 = (−1)2 + 12 + 12 = 3 q1 = v1 1 ⎞ (1. 119 . 4) − (3.

A is invertible. 2) − 2 0+0+2 (1. 1) = (−1. and u3 = (2. 1. 1) v1 = 12 + 02 + 12 = 2 v2 = u2 − u 2 . . 1) = (0. v 2 v2 2 v2 2+0+0 −2 + 1 + 0 (1. 0) − (1. q1 ⎤ ⎢ −2 1⎥ = [q1 q 2 ] ⎢ 1 1 ⎥ u 0 ⎢ ⎥ 2 . v1 v1 v1 − 2 u3 . and u3 . 1) ⎛ 1 1 1 ⎞ . q1 = 0 + 0 + 2 2 2 = 2 = 2 155 . 1. 4. 0). 0. u 2 = (0. v1 v1 v1 = (0. 2). q2 ⎦ ⎣ ⎣ 1 1⎦ 8 ⎤ ⎡ 1 234 ⎥ ⎡ ⎢ 3 3 13 ⎤ 11 ⎥ ⎢ ⎥ = ⎢ − 23 234 ⎥ ⎢ 26 ⎥ ⎢ 0 3 ⎦ 7 ⎥⎣ ⎢ 2 ⎢⎣ 3 234 ⎥⎦ ⎡1 0 2 ⎤ (d) A = ⎢0 1 1 ⎥ ⎢ ⎥ ⎣1 2 0 ⎦ Since det(A) = −4 ≠ 0. Apply the Gram-Schmidt process with normalization to u1 . v1 = u1 = (1. 1. 1.SSM: Elementary Linear Algebra Section 6. 0. 0) − 2 2 2 ⎛2⎞ ⎛4⎞ ⎛ 2⎞ v3 = ⎜ ⎟ + ⎜ ⎟ + ⎜ − ⎟ = ⎝3⎠ ⎝3⎠ ⎝ 3⎠ 24 2 6 = 9 3 q1 = v1 1 ⎞ (1. 1) 2 v 2 = (−1)2 + 12 + 12 = 3 v 3 = u3 − u3 . 0. 2) − (1. so it has a QR-decomposition. 1. 1) − (−1.q u 2 . 0. 1) ⎛ 1 . . . 1. . 0. = =⎜ ⎟ v1 2 2⎠ 2 ⎝ q2 = v2 (−1. 1) + ⎜ − . ⎟ ⎝ 3 3 3⎠ 2⎞ ⎛2 4 =⎜ . 1. 1. 0.− = 3 3 =⎜ q3 = ⎟ 2 6 v3 6 6⎠ ⎝ 6 3 u1 . 0. 0. = = ⎜− ⎟ v2 3 3 3⎠ 3 ⎝ ( ) 2. − 2 v3 2 1 ⎞ ⎛ 1 3 . − ⎟ 3⎠ ⎝3 3 = (2.3 The QR-decomposition of A is ⎡ 1 1⎤ ⎡ u . 1. u 2 . q1 = 1 2 1 +0+ u 2 . 1) 2 3 ⎛ 1 1 1⎞ = (2. 1). Let u1 = (1.

0) 2 ⎛3 3 ⎞ = (2. and u3 . q 2 u3 . 1. 0) v1 = 12 + 12 + 02 = 2 v2 = u2 − u 2 . and u3 = (1. q2 0 2 3 0 u3 . 3) − ⎜ . − . u 2 = (2. q1 ⎡1 0 2 ⎤ ⎢0 1 1 ⎥ = [q1 q 2 q3 ] ⎢ 0 ⎢ ⎢ ⎥ ⎢⎣ 0 ⎣1 2 0 ⎦ ⎡ 1 − ⎢ 2 =⎢ 0 ⎢ ⎢ 1 ⎣⎢ 2 1 3 1 3 1 3 − 1 ⎤ ⎡ 6⎥⎢ 2 ⎥⎢ 6⎥⎢ 1 ⎥⎢ 6 ⎦⎥ ⎣ 2 0 0 u 2 . v1 = u1 = (1. 3) − 2 2 19 ⎛1⎞ ⎛ 1⎞ v 2 = ⎜ ⎟ + ⎜ − ⎟ + 32 = 2 ⎝2⎠ ⎝ 2⎠ 156 . q 3 ⎤ ⎥ ⎥ ⎥⎦ 2⎤ ⎥ − 1 ⎥. 3⎟ 2 ⎠ ⎝2 = (2. q3 = 3 + 2 2 + 3 2 u3 . 1. q 2 = − SSM: Elementary Linear Algebra 3 1 3 = 3 +0 = − 2 1 3 4 + +0 = 6 6 6 The QR-decomposition of A is ⎡ u1. Let u1 = (1. 1. u 2 . q1 u2 . 1. . v1 v1 2 v1 2 +1+ 0 (1. q1 = 2 +0+0 = 2 2 1 u2 . 1.Chapter 6: Inner Product Spaces u3 . 3). 0 ⎟ ⎝2 2 ⎠ 1 ⎞ ⎛1 = ⎜ . q 2 = 0 + u3 . A is invertible. 1. 0). so it has a QR-decomposition. 3 4 ⎥ ⎥ 6 ⎦ ⎡1 2 1⎤ (e) A = ⎢1 1 1⎥ ⎢ ⎥ ⎣ 0 3 1⎦ Since det(A) = −1 ≠ 0. Apply the Gram-Schmidt process with normalization to u1 . 1. q1 u3 . 1).

q 3 2⎤ ⎥ 3 2⎥ 19 ⎥ 1 ⎥ 19 ⎥⎦ 157 ⎤ ⎥ ⎥ ⎥⎦ . ⎟ ⎝ 19 19 19 ⎠ 2 2 2 1 1 ⎛ 3⎞ ⎛ 3⎞ ⎛1⎞ v3 = ⎜ − ⎟ + ⎜ ⎟ + ⎜ ⎟ = = 19 ⎝ 19 ⎠ ⎝ 19 ⎠ ⎝ 19 ⎠ 19 q1 = v1 1 (1. q3 = − + + = 19 19 19 19 The QR-decomposition of A is ⎡ u1. 1. 3 ⎛ 2 v2 2 3 2⎞ 2 = 2 =⎜ . q 2 = u3 . 0) − 2 2 ⎜ . 1 − 19 19 19 1 19 1 2 2 2 1 2 + + + 2 2 2 19 2 1 2 1 2 − ⎜ ⎝ 3 19 . .3 u3 . − . q2 ⎢ 0 ⎢ ⎥ 0 3 1 ⎢ 0 ⎣ ⎦ ⎣ 0 ⎡ 1 ⎢ 2 ⎢ =⎢ 1 2 ⎢ ⎢0 ⎣ 2 2 19 − 2 2 19 3 2 19 − 3 ⎤⎡ 2 19 ⎥ ⎢ 3 ⎥⎢ 0 19 ⎥ ⎢ ⎥ 1 ⎥⎢ 0 ⎢ 19 ⎦ ⎣ 3 2 19 2 0 u3 . q 2 = ( ) 2 3 . 3 . 1. 0) ⎛ 1 ⎞ . 1. 0) − ⎜ . v 2 v2 2 v2 1 − 1 +3 1+1+ 0 1 ⎞ ⎛1 (1. 1. 3 19 . 0⎟ = =⎜ v1 2 ⎠ 2 ⎝ 2 q2 = 1. q1 u3 . 1) − v1 − 2 Section 6. q 2 u3 . − 1. 1. − . q1 = u2 . 1) − (1. q1 = u 2 .− . q1 ⎡1 2 1⎤ ⎢ ⎢1 1 1⎥ = [q1 q 2 q3 ] u2 . q1 u 2 . 3⎟ 19 2 2 ⎠ ⎝2 2 3 18 ⎞ ⎛ 3 = (1. v1 v1 = (1. 1 ⎞ ⎟ 19 ⎠ +0 = 2 2 1 − ) = ⎛− +0 = 3 2 +0 = 2 2 2 19 2 + 9 2 + 3 2 19 = 19 = 3 2 2 2 19 2 19 19 19 3 3 1 1 u3 . q1 = u3 . ⎟ 19 v2 2 19 19 ⎠ ⎝ 2 19 q3 = ( v3 = v3 u1 . . ⎟ ⎝ 19 19 19 ⎠ ⎛ 3 3 1⎞ = ⎜− .SSM: Elementary Linear Algebra v 3 = u3 − u3 .

. v i = v i . q i = u i .. v2 2 1⎛ 1 ⎞ = ∫ ⎜ − + x ⎟ dx 0⎝ 2 ⎠ 1⎛ 1 ⎞ = ∫ ⎜ − x + x 2 ⎟ dx 0⎝ 4 ⎠ q2 = = 1 1 1 ⎞ ⎛1 = ⎜ x − x 2 + x3 ⎟ 2 3 ⎠0 ⎝4 1 = 12 v2 v2 − 12 + x 1 2 3 ⎛ 1 ⎞ = 2 3⎜− + x⎟ ⎝ 2 ⎠ = 3 (−1 + 2 x) 158 . u n }.. x. 1 and 12 vi vi = 1 vi 1 1 =x − + −x 3 2 1 = − x + x2 6 vi . v3}. so the column vectors of A are not linearly independent and A does not have a QR-decomposition. v1 = p1 = 1 2 1 ⎛ 1 ⎞ = x 2 − 3 (1) − 12 ⎜ − + x ⎟ 1 ⎝ 2 1 ⎠ Since each vector vi is nonzero. so ui = vi + k1v1 + k2 v 2 + + ki −1 vi −1 . v1 1 v3 = 1 2 1 v1 = x − (1) = − + x 1 2 = v2 . v1 = ∫ x 2 ⋅1 dx = 0 1 1 ⎞ ⎛ 1 = ⎜ − x 3 + x3 ⎟ 4 ⎠0 ⎝ 6 1 = 12 p .. x 2} into an orthogonal basis {v1 . 2. v i = v i . Thus. where qi = i is the vi ui . v1 = ∫ x ⋅1 dx = 0 v2 = p2 − v2 2 p 2 . v 2 = ∫ x 2 ⎜ − + x ⎟ dx 0 ⎝ 2 ⎠ 1⎛ 1 2 ⎞ = ∫ ⎜ − x + x3 ⎟ dx 0⎝ 2 ⎠ 31.Chapter 6: Inner Product Spaces SSM: Elementary Linear Algebra ⎡ 1 0 1⎤ ⎢ −1 1 1⎥ (f) A = ⎢ ⎥ = [c1 c 2 c3 ] ⎢ 1 0 1⎥ ⎢⎣ −1 1 1⎥⎦ By inspection. The diagonal entries of R are ui . each diagonal entry of R is nonzero. ui . Thus. u 2 . vi −1 . v 2 . v1 = ∫ 12 dx = x 10 = 1 0 1 4 1 1 ⎞ ⎛ 1 = ⎜ x − x 2 + x3 − x 4 + x5 ⎟ 6 9 2 5 ⎠0 ⎝ 36 1 = 180 v1 = 1 = 1 1 p 2 . First transform the basis 2 2 1 1 = v1 ... 2 v3 33. n. . qi for v i = 1. 1 p3 . v1 1 1 1 3 x = 3 0 3 1 ⎛ 1 ⎞ p3 .v p .. v1 v1 2 1 2 x 2 1 = 0 = v3 . v3 1⎛ 1 ⎞ = ∫ ⎜ − x + x 2 ⎟ dx 0⎝ 6 ⎠ 1⎛ 1 1 4 ⎞ = ∫ ⎜ − x + x 2 − 2 x3 + x 4 ⎟ dx 0 ⎝ 36 3 3 ⎠ S = {p1. vi is ui minus a linear combination of the vectors v1.. p 2 .. v 2 . p3} = {1.v v3 = p3 − 3 1 v1 − 3 2 v 2 2 2 v1 v2 normalization of a vector vi that is the result of applying the Gram-Schmidt process to {u1 . c3 − c1 = 2c2 .. v i 1 1 = 12 2 3 v2 = 1 2 1 1 = 180 6 5 The orthonormal basis is v 1 q1 = 1 = = 1. .

(f) True. (e) False. Section 6. a nonzero finite-dimensional inner product space will have finite basis which can be transformed into an orthonormal basis with respect to the inner product via the GramSchmidt process with normalization. (b) False. (a) A = ⎢ 2 3⎥ ⎢ ⎥ ⎣ 4 5⎦ ⎡1 ⎡ 1 2 1 1⎤ ⎢ 2 (b) AT A = ⎢ 0 1 1 1⎥ ⎢ ⎢ ⎥ 1 ⎣ −1 −2 0 −1⎦ ⎢⎢ 1 ⎣ ⎡ 7 4 −6 ⎤ = ⎢ 4 3 −3⎥ ⎢ ⎥ ⎣ −6 −3 6 ⎦ ⎡ 1 −1⎤ ⎡ 1 2 4⎤ ⎢ ⎡ 21 25⎤ AT A = ⎢ 2 3⎥ = ⎢ ⎥ − 1 3 5 ⎢ ⎥ ⎣ ⎦ 4 5 ⎣ 25 35 ⎥⎦ ⎣ ⎦ ⎡ 2⎤ ⎡ 1 2 4 ⎤ ⎢ ⎥ ⎡ 20⎤ AT b = ⎢ −1 = ⎣ −1 3 5⎥⎦ ⎢ 5⎥ ⎢⎣ 20⎥⎦ ⎣ ⎦ The associated normal system is ⎡ 21 25⎤ ⎡ x1 ⎤ ⎡ 20 ⎤ ⎢⎣ 25 35 ⎥⎦ ⎢ x ⎥ = ⎢⎣ 20 ⎥⎦ ⎣ 2⎦ 0 −1⎤ 1 −2 ⎥ ⎥ 1 0⎥ 1 −1⎦⎥ ⎡ 6⎤ ⎡ 1 2 1 1⎤ ⎢ ⎥ ⎡ 18⎤ 0 A b=⎢ 0 1 1 1⎥ ⎢ ⎥ = ⎢ 12⎥ ⎢ ⎥ ⎢ 9⎥ ⎢ ⎥ ⎣ −1 −2 0 −1⎦ ⎢ 3⎥ ⎣ −9⎦ ⎣ ⎦ The associated normal system is T 159 . 2) and (−1.SSM: Elementary Linear Algebra q3 = = Section 6. ⎣ 2⎦ ⎡ 1 0 5⎤ ⎡ 3 −2 14 ⎤ reduces to ⎢0 1 1 ⎥ . (c) True. the vectors (1.4 ⎡ 2 −1 0 ⎤ ⎢ 3 1 2⎥ (b) A = ⎢ ⎥ ⎢ −1 4 5⎥ ⎢⎣ 1 2 4 ⎥⎦ v3 v3 1 6 − x + x2 1 6 5 ⎡ 2 −1 0⎤ ⎡ 2 3 −1 1⎤ ⎢ 3 1 2⎥ AT A = ⎢ −1 1 4 2⎥ ⎢ ⎥ ⎢ ⎥ ⎢ −1 4 5⎥ 0 2 5 4 ⎣ ⎦ ⎢ 1 2 4⎥ ⎣ ⎦ − 15 1 5 ⎡ ⎤ = ⎢ −1 22 30⎥ ⎢ ⎥ ⎣ 5 30 45⎦ ⎡ −1⎤ ⎡ 2 3 −1 1⎤ ⎢ ⎥ ⎡ −1⎤ 0 T A b = ⎢ −1 1 4 2 ⎥ ⎢ ⎥ = ⎢ 9 ⎥ ⎢ ⎥ ⎢ 1⎥ ⎢ ⎥ ⎣ 0 2 5 4 ⎦ ⎢ 2 ⎥ ⎣ 13⎦ ⎣ ⎦ The associated normal system is ⎡15 −1 5⎤ ⎡ x1 ⎤ ⎡ −1⎤ ⎢ −1 22 30 ⎥ ⎢ x2 ⎥ = ⎢ 9⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 5 30 45⎦ ⎣ x3 ⎦ ⎣ 13⎦ ⎛1 ⎞ = 6 5 ⎜ − x + x2 ⎟ ⎝6 ⎠ 2 = 5 (1 − 6 x + 6 x ) True/False 6. so ⎢⎣ −2 6 −7 ⎥⎦ 2⎦ ⎣ T (d) True. 3) in R 2 are linearly independent but not orthogonal. a nontrivial subspace of R3 will have a basis. (a) A A = ⎢ −1 1⎥ = ⎢ ⎥ 1 1 2 ⎢ ⎥ ⎣ ⎦ −1 2 ⎣ −2 6⎥⎦ ⎣ ⎦ ⎡ 7⎤ ⎡1 −1 −1⎤ ⎢ ⎥ ⎡14 ⎤ AT b = ⎢ 0 = ⎣1 1 2 ⎥⎦ ⎢ −7 ⎥ ⎢⎣ −7 ⎥⎦ ⎣ ⎦ The associated normal system is ⎡ 3 −2 ⎤ ⎡ x1 ⎤ ⎡14 ⎤ ⎢⎣ −2 6 ⎥⎦ ⎢ x ⎥ = ⎢⎣ −7 ⎥⎦ .3 (a) False. 2 Exercise Set 6. x2 = . the vectors must be nonzero for this to be true. every invertible n × n matrix has a QR-decomposition.4 1 the least squares solution is x1 = 5. for example. which can be transformed into an orthonormal basis with respect to the Euclidean inner product. ⎡ 1 1⎤ ⎡1 −1 −1⎤ ⎢ ⎡ 3 −2 ⎤ 3.4 ⎡ 1 −1⎤ 1. projW x is a vector in W.

so e is AT e = ⎢ ⎢ ⎥= ⎣1 1 2 ⎥⎦ ⎢ 2 ⎥ ⎢⎣ 0⎥⎦ ⎣ −3⎦ orthogonal to the column space of A. x3 = 9. ⎡ 7 4 ⎢ 4 3 ⎢ ⎣ −6 −3 ⎡ 7 4 ⎢ 4 3 ⎢ ⎣ −6 −3 ⎡ 24 8 12⎤ ⎢⎣ 8 6 8⎥⎦ reduces to 5.Chapter 6: Inner Product Spaces SSM: Elementary Linear Algebra ⎡ 2 1⎤ ⎡ 2 4 −2 ⎤ ⎢ ⎡ 24 8⎤ 7. ⎣ 2⎦ −6 ⎤ ⎡ x1 ⎤ ⎡ 18⎤ −3⎥ ⎢ x2 ⎥ = ⎢ 12 ⎥ . ⎥⎢ ⎥ ⎢ ⎥ 6 ⎦ ⎣ x3 ⎦ ⎣ −9 ⎦ −6 18⎤ −3 12 ⎥ reduces to ⎥ 6 −9 ⎦ ⎡ 1 0 0 12 ⎤ ⎢0 1 0 −3⎥ so the least squares ⎢ ⎥ ⎣0 0 1 9 ⎦ solution is x1 = 12. 6⎥ 5⎦ so 1 . (a) e = b − Ax ⎡ 7 ⎤ ⎡ 1 1⎤ ⎡ 5 ⎤ = ⎢ 0 ⎥ − ⎢ −1 1⎥ ⎢ 1 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ −7 ⎦ ⎣ −1 2⎦ ⎣ 2 ⎦ 11 ⎡ 7 ⎤ ⎡⎢ 2 ⎤⎥ = ⎢ 0⎥ − ⎢− 9 ⎥ 2 ⎢ ⎥ ⎣ −7 ⎦ ⎣⎢ −4 ⎦⎥ ⎡ 3⎤ ⎢ 2⎥ = ⎢ 9⎥ ⎢ 2⎥ ⎣ −3⎦ ⎡ 3⎤ ⎡1 −1 −1⎤ ⎢ 92 ⎥ ⎡ 0⎤ . T ⎡ 14 42 4 ⎤ ⎢⎣ 42 126 12 ⎥⎦ reduces to 160 ⎡1 3 2⎤ 7 ⎥ . e = ⎜ ⎟ +⎜− ⎟ +0 = 25 5 5 5 ⎝ ⎠ ⎝ ⎠ ⎡ 6 ⎤ ⎡ 3⎤ ⎢ 0 ⎥ ⎢ 3⎥ = ⎢ ⎥−⎢ ⎥ ⎢ 9 ⎥ ⎢9 ⎥ ⎣⎢ 3⎦⎥ ⎣⎢0 ⎦⎥ ⎡ 3⎤ ⎢ −3⎥ =⎢ ⎥ ⎢ 0⎥ ⎣⎢ 3⎦⎥ ⎡ 1 3⎤ ⎡ 1 −2 3⎤ ⎢ ⎡ 14 42 ⎤ (b) AT A = ⎢ −2 −6⎥ = ⎢ ⎥ ⎣3 −6 9 ⎦ ⎢ 3 9⎥ ⎣ 42 126 ⎥⎦ ⎣ ⎦ ⎡ 1⎤ ⎡ 1 −2 3⎤ ⎢ ⎥ ⎡ 4 ⎤ AT b = ⎢ 0 = ⎣3 −6 9 ⎥⎦ ⎢ 1⎥ ⎢⎣12 ⎥⎦ ⎣ ⎦ The associated normal system is ⎡ 14 42 ⎤ ⎡ x1 ⎤ ⎡ 4 ⎤ ⎢⎣ 42 126 ⎥⎦ ⎢ x ⎥ = ⎢⎣12 ⎥⎦ . 5 ⎢⎣ 65 ⎥⎦ The error vector is e = b − Ax ⎡ 3⎤ ⎡ 2 1 ⎤ ⎡ 1 ⎤ ⎥ = ⎢ 2 ⎥ − ⎢ 4 2 ⎥ ⎢ 10 ⎢ ⎥ ⎢ ⎥⎢ 6 ⎥ − 1 2 1 5 ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ 7 ⎡ ⎤ ⎡ 3⎤ ⎢ 5 ⎥ = ⎢ 2 ⎥ − ⎢ 14 ⎥ ⎢ ⎥ ⎢5⎥ ⎣ 1⎦ ⎣ 1 ⎦ ⎡ 8⎤ ⎢ 5⎥ = ⎢− 4 ⎥ ⎢ 5⎥ ⎣ 0⎦ so the least squares error is x2 = 0 −1⎤ ⎡ 12 ⎤ 1 −2 ⎥ ⎢ ⎥ ⎥ −3 1 0⎥ ⎢ ⎥ 9 1 −1⎦⎥ ⎣ ⎦ 2 2 80 4 ⎛8⎞ ⎛ 4⎞ 2 = 5. so e is ⎢ ⎥ ⎢ 0⎥ ⎢ ⎥ ⎣ −1 −2 0 −1⎦ ⎢ 3⎥ ⎣0 ⎦ ⎣ ⎦ orthogonal to the column space of A. 10 ⎡1⎤ 6 or x = ⎢ 10 ⎥ . so ⎢ ⎣0 0 0⎦ . ⎡6 ⎤ ⎡ 1 ⎢0 ⎥ ⎢ 2 (b) e = b − Ax = ⎢ ⎥ − ⎢ ⎢9 ⎥ ⎢ 1 ⎣⎢ 3⎦⎥ ⎣⎢ 1 ⎡1 0 ⎢ ⎢⎣0 1 the least squares solution is x1 = 1⎤ 10 ⎥ . (a) AT A = ⎢ 4 2⎥ = ⎢ ⎥ 1 2 1 ⎣ ⎦ ⎢ −2 1⎥ ⎣ 8 6 ⎥⎦ ⎣ ⎦ ⎡ 3⎤ ⎡ 2 4 −2⎤ ⎢ ⎥ ⎡12 ⎤ AT b = ⎢ 2 = 1⎥⎦ ⎢ ⎥ ⎢⎣ 8⎥⎦ ⎣1 2 ⎣ 1⎦ The associated normal system is ⎡ 24 8⎤ ⎡ x1 ⎤ ⎡12 ⎤ ⎢⎣ 8 6⎥⎦ ⎢ x ⎥ = ⎢⎣ 8⎥⎦ . x2 = −3. ⎣ 2⎦ ⎡ 3⎤ ⎡ 1 2 1 1⎤ ⎢ ⎥ ⎡0 ⎤ −3 A e=⎢ 0 1 1 1⎥ ⎢ ⎥ = ⎢0 ⎥ .

is the subspace of R 4 spanned by v1 . The error vector is ⎡ 1⎤ ⎡ 1 3⎤ ⎛ ⎡ 2 ⎤ ⎡ −3⎤ ⎞ e = b − Ax = ⎢ 0⎥ − ⎢ −2 −6 ⎥ ⎜ ⎢ 7 ⎥ + t ⎢ ⎥ ⎟ ⎜ ⎢ ⎥ ⎢ ⎥ 0 ⎣ 1⎦ ⎠⎟ ⎣ 1⎦ ⎣ 3 9 ⎦ ⎝ ⎣ ⎦ ⎛⎡ 2⎤ ⎞ ⎡ 1⎤ ⎜ ⎢ 7 ⎥ ⎡ 0⎤ ⎟ = ⎢ 0⎥ − ⎜ ⎢ − 74 ⎥ + t ⎢ 0⎥ ⎟ ⎢ ⎥ ⎜⎢ ⎥ ⎢ ⎥⎟ ⎣ 1⎦ ⎜ ⎢ 6 ⎥ ⎣ 0⎦ ⎟ ⎝⎣ 7⎦ ⎠ ⎡5⎤ ⎢7⎥ = ⎢ 74 ⎥ . so the least squares ⎢0 1 1 6 ⎢ ⎥ 0 0 0 0 ⎣ ⎦ 7 7 solutions are x1 = − − t . x2 = − t . T ( A A) 161 −1 ⎡ 3 −2 2 ⎤ = ⎢⎢ −2 2 −1⎥⎥ 5 ⎢⎣ 2 −1 3 ⎥⎦ . 7 ⎡ 2 ⎤ ⎡ −3⎤ x2 = t or x = ⎢ 7 ⎥ + t ⎢ ⎥ for t a real ⎣ 0 ⎦ ⎣ 1⎦ number. ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 4 10 14 ⎦ ⎣ x3 ⎦ ⎣ 7 ⎦ ⎡2 ⎢1 9.4 the least squares solutions are x1 = ⎡− 7 ⎤ −1 ⎢ 6⎥ ⎡ ⎤ x3 = t or x = ⎢ 7 ⎥ + t ⎢ −1⎥ for t a real ⎢ 6 ⎥ ⎢⎣ 1⎥⎦ ⎣ 0⎦ number.SSM: Elementary Linear Algebra Section 6. and v3 . 1 0⎥ 1 −1⎥⎦ ⎡2 ⎡ 2 1 1 1⎤ ⎢ 1 A A = ⎢ 1 0 1 1⎥ ⎢ ⎢ ⎥ ⎢1 ⎣ −2 −1 0 −1⎦ ⎢ 1 ⎣ ⎡ 7 4 −6 ⎤ = ⎢ 4 3 −3⎥ ⎢ ⎥ ⎣ −6 −3 6 ⎦ T ⎡ 5 −1 4 −7 ⎤ ⎢ −1 11 10 14⎥ reduces to ⎢ ⎥ ⎣ 4 10 14 7 ⎦ ⎡1 0 1 − 7⎤ 6⎥ ⎢ 7 ⎥ . ⎢ 49 ⎥ ⎢⎣ − 6 ⎥⎦ The least squares error is 2 − 3t . (a) Let A = ⎢ ⎢1 ⎢⎣ 1 1 −2 ⎤ 0 −1⎥ ⎥. 2 ⎡ −1 2 0 ⎤ ⎡ −1 3 2 ⎤ (c) AT A = ⎢ 3 1 1⎥ ⎢ 2 1 3⎥ ⎢ ⎥⎢ ⎥ ⎣ 2 3 1⎦ ⎣ 0 1 1⎦ ⎡ 5 −1 4 ⎤ = ⎢ −1 11 10 ⎥ ⎢ ⎥ ⎣ 4 10 14 ⎦ ⎡ −1 2 0 ⎤ ⎡ 7 ⎤ ⎡ −7 ⎤ AT b = ⎢ 3 1 1⎥ ⎢ 0 ⎥ = ⎢ 14 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 2 3 1⎦ ⎣ −7 ⎦ ⎣ 7 ⎦ The associated normal system is ⎡ 5 −1 4 ⎤ ⎡ x1 ⎤ ⎡ −7 ⎤ ⎢ −1 11 10 ⎥ ⎢ x2 ⎥ = ⎢ 14 ⎥ . ⎢1⎥ ⎢⎣ 7 ⎥⎦ The least squares error is 2 2 ⎛5⎞ ⎛4⎞ ⎛1⎞ e = ⎜ ⎟ +⎜ ⎟ +⎜ ⎟ ⎝7⎠ ⎝7⎠ ⎝7⎠ 42 = 49 1 = 42 . 7 2 2 2 ⎛ 7 ⎞ ⎛ 7 ⎞ ⎛ 49 ⎞ e = ⎜ ⎟ +⎜ ⎟ +⎜− ⎟ ⎝3⎠ ⎝6⎠ ⎝ 6 ⎠ 294 = 4 7 = 6. The error vector is e = b − Ax ⎛ 7 ⎞ ⎡ 7 ⎤ ⎡ −1 3 2 ⎤ ⎜ ⎢⎡ − 6 ⎥⎤ ⎡ −1⎤ ⎟ = ⎢ 0 ⎥ − ⎢ 2 1 3⎥ ⎜ ⎢ 7 ⎥ + t ⎢ −1⎥ ⎟ 6 ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ −7 ⎦ ⎣ 0 1 1⎦ ⎜⎜ ⎣⎢ 0 ⎦⎥ ⎣ 1⎦ ⎟⎟ ⎝ ⎠ ⎛ ⎡ 14 ⎤ ⎞ ⎡ 7 ⎤ ⎜ ⎢ 3 ⎥ ⎡0 ⎤ ⎟ = ⎢ 0 ⎥ − ⎜ ⎢ − 76 ⎥ + t ⎢0 ⎥ ⎟ ⎢ ⎥ ⎜⎢ ⎥ ⎢ ⎥⎟ ⎣ −7 ⎦ ⎜ ⎢ 7 ⎥ ⎣ 0 ⎦ ⎟ ⎝⎣ 6⎦ ⎠ 7 ⎡ ⎤ ⎢ 3⎥ = ⎢ 76 ⎥ . W. 6 6 2 1 −2 ⎤ 0 −1⎥ ⎥ 1 0⎥ 1 −1⎥⎦ det( AT A) = 3 ≠ 0. so the column vectors of A are linearly independent and the column space of A. v 2 .

− . ⎟ 5 5 5⎠ ⎝ 5 ⎤ ⎡ −2 ⎤ − 31 5 ⎥⎡ 1 1 3 0⎤ ⎢ ⎥ 0 79 ⎥ ⎢ − 10 −2 −1 −2 1⎥ ⎢ ⎥ 2 ⎢ ⎥ ⎥ 1 3⎦ ⎢ ⎥ 29 ⎣ −3 −1 ⎢⎣ 4 ⎥⎦ 10 ⎥⎦ 162 . . 9. 2. 1⎥ ⎢ 3 −2 1 3⎥⎦ ⎢⎣ 0 ⎡ 1 −2 −3⎤ ⎡ 1 1 3 0⎤ ⎢ 1 −1 −1⎥ AT A = ⎢ −2 −1 −2 1⎥ ⎢ ⎥ 3 1⎥ ⎢ ⎥ ⎢ −2 − − 3 1 1 3 ⎣ ⎦ ⎢0 1 3⎦⎥ ⎣ ⎡ 11 −9 −1⎤ = ⎢ −9 10 8⎥ ⎢ ⎥ ⎣ −1 8 20⎦ det( AT A) = 10 ≠ 0 so the column vectors of A are linearly independent and the column space of A. 5) ⎡ 1 −2 −3⎤ ⎢ 1 −1 −1⎥ (b) Let A = ⎢ ⎥. W. 86 − 31 ⎤ ⎡ 68 5 5⎥ ⎢ 5 219 − 79 ⎥ ( AT A)−1 = ⎢ 86 5 10 10 ⎢ 31 ⎥ 79 29 ⎢⎣ − 5 − 10 10 ⎥⎦ projW u = A( AT A)−1 AT u 86 ⎡ 1 −2 −3⎤ ⎡ 68 5 5 ⎢ 1 −1 −1⎥ ⎢ 86 219 ⎢ =⎢ ⎥ 10 1⎥ ⎢ 5 ⎢ 3 −2 31 79 1 3⎥⎦ ⎢⎣ − 5 − 10 ⎢⎣0 ⎡ − 12 ⎤ ⎢ 5⎥ ⎢ − 4⎥ = ⎢ 125 ⎥ ⎢ 5⎥ ⎢ 16 ⎥ ⎢⎣ 5 ⎥⎦ 4 12 16 ⎞ ⎛ 12 projW u = ⎜ − . and v3 . is the subspace of R 4 spanned by v1 .Chapter 6: Inner Product Spaces SSM: Elementary Linear Algebra projW u = A( AT A)−1 AT u ⎡ 2 1 −2 ⎤ ⎡ 3 −2 2⎤ ⎡6 ⎤ ⎡ 2 1 1 1⎤ ⎢ ⎥ ⎢ 1 0 −1⎥ ⎢ 3 ⎥ =⎢ ⎥ ⎢ −2 2 −1⎥ ⎢ 1 0 1 1⎥ ⎢ ⎥ 1 1 0 9 ⎢ ⎥ ⎢ ⎥ ⎢ 2 −1 5 ⎥ ⎣ −2 −1 0 −1⎦ ⎢ ⎥ 3⎦ ⎢⎣ 1 1 −1⎥⎦ ⎣ ⎢⎣6 ⎥⎦ ⎡7 ⎤ ⎢ 2⎥ =⎢ ⎥ ⎢ 9⎥ ⎢⎣ 5⎥⎦ projW u = (7. v 2 .

0. ⎡ 1 0⎤ Thus A = ⎢0 0 ⎥ and ⎢ ⎥ ⎣0 1⎦ ⎡1 ⎡ 1 0 0⎤ ⎢ AT A = ⎢ 0 ⎣ 0 0 1⎥⎦ ⎢ 0 ⎣ ⎡ 1 0 0⎤ AT = ⎢ . ( AT A)−1 = 1 ⎡10 15⎤ 35 ⎢⎣15 26 ⎥⎦ [ P] = A( AT A)−1 AT ⎡ 1 0⎤ ⎛ 1 ⎡10 15 ⎤ ⎞ ⎡ 1 0 −5⎤ = ⎢ 0 1⎥ ⎜ ⎢ ⎢ ⎥ ⎝ 35 ⎣15 26 ⎥⎦ ⎟⎠ ⎢⎣0 1 3⎥⎦ ⎣ −5 3⎦ ⎡ 10 15⎤ 1 ⎢ ⎡ 1 0 −5 ⎤ = 15 26 ⎥ 35 ⎢ −5 3⎥ ⎢⎣0 1 3⎥⎦ ⎣ ⎦ ⎡ 10 15 −5⎤ 1 ⎢ = 15 26 3⎥ 35 ⎢ −5 3 34 ⎥ ⎣ ⎦ 163 . 1. (a) If x = s and y = t. w1 = (1. 1). Thus A = ⎢ 1 0 ⎥ and AT = ⎢ ⎢ ⎥ ⎣ 0 0 1⎥⎦ 0 1 ⎣ ⎦ det( AT A) = 0. w 2 } where 15. 0. t. w 2 } where w1 = (0. ⎣ 0 0 1⎥⎦ ⎡ 1 0⎤ ⎡1 (b) A = ⎢ 0 1⎥ . 1. w 2 = (0. AT = ⎢ ⎢ ⎥ ⎣0 ⎣ −5 3⎦ ⎡ 1 ⎡ 1 0 −5⎤ ⎢ AT A = ⎢ 0 ⎣ 0 1 3⎥⎦ ⎢ −5 ⎣ 0⎤ ⎡1 0 ⎤ 0⎥ = ⎢ 0 1 ⎥⎦ ⎥ 1⎦ ⎣ [ P] = A( AT A)−1 AT ⎡ 1 0⎤ ⎡1 0 ⎤ ⎡ 1 0 0 ⎤ = ⎢0 0⎥ ⎢ ⎢ ⎥ ⎣ 0 1 ⎥⎦ ⎢⎣0 0 1⎥⎦ ⎣0 1⎦ ⎡ 1 0⎤ ⎡ 1 0 0⎤ = ⎢0 0⎥ ⎢ ⎢ ⎥ ⎣ 0 0 1⎥⎦ ⎣0 1⎦ ⎡1 0 0 ⎤ = ⎢0 0 0 ⎥ ⎢ ⎥ ⎣0 0 1 ⎦ 0 −5 ⎤ . then a point on the plane is (s. 1).4 0 ⎤ ⎡ −1 3 2 ⎤ 1⎥ ⎢ 2 1 3⎥ ⎥⎢ ⎥ 1⎦ ⎣ 0 1 1⎦ 4⎤ 10 ⎥ ⎥ 14 ⎦ (b) W = the yz-plane is two-dimensional and one basis for W is {w1. ⎡0 0 ⎤ ⎡ 0 1 0⎤ . −5) + t(0. 3). so A does not have linearly independent column vectors. so A does not have linearly independent column vectors. 1. (a) W = the xz-plane is two-dimensional and one basis for W is {w1. 0) and w 2 = (0. 0. 13. 1 3⎥⎦ 0⎤ ⎡ 26 −15⎤ 1⎥ = ⎢ ⎥ ⎣ −15 10 ⎥⎦ 3⎦ det( AT A) = 35. 0). ⎡0 0⎤ ⎡ 0 1 0⎤ ⎢ ⎡1 0 ⎤ AT A = ⎢ 1 0⎥ = ⎢ ⎥ 0 0 1 ⎢ ⎥ ⎣ ⎦ 0 1 ⎣0 1 ⎥⎦ ⎣ ⎦ ⎡ 2 −1 3 ⎤ ⎡ 2 0 −1 4 ⎤ ⎢ 0 1 1⎥ (b) AT A = ⎢ −1 1 0 −5⎥ ⎢ ⎥ ⎢ ⎥ ⎢ −1 0 −2 ⎥ − 3 1 2 3 ⎣ ⎦ ⎢ 4 −5 3 ⎥ ⎣ ⎦ ⎡ 21 −22 20 ⎤ = ⎢ −22 27 −17 ⎥ ⎢ ⎥ 23⎦ ⎣ 20 −17 [ P] = A( AT A)−1 AT ⎡0 0⎤ ⎡1 0 ⎤ ⎡0 1 0 ⎤ = ⎢ 1 0⎥ ⎢ ⎢ ⎥ ⎣ 0 1 ⎥⎦ ⎢⎣0 0 1⎥⎦ ⎣0 1⎦ ⎡0 0⎤ ⎡0 1 0⎤ = ⎢ 1 0⎥ ⎢ ⎢ ⎥ ⎣ 0 0 1⎥⎦ ⎣0 1⎦ ⎡0 0 0 ⎤ = ⎢0 1 0 ⎥ ⎢ ⎥ ⎣0 0 1⎦ det( AT A) = 0. 3) are a basis for W. −5s + 3t) = s(1. 0. (a) A A = ⎢ 3 ⎢ ⎣ 2 ⎡ 5 = ⎢ −1 ⎢ ⎣ 4 T 2 1 3 −1 11 10 Section 6. w1 = (1.SSM: Elementary Linear Algebra ⎡ −1 11. 0. − 5) and w 2 = (0.

7 17. the standard matrix for the orthogonal projection of R n onto the row space of A is [ P] = AT [( AT )T AT ]−1 ( AT )T = AT ( AAT )−1 A. AT b = 0. then AT A is invertible and the least squares solution of Ax = b is the solution of AT Ax = AT b. 1) = (1.3. When s = 1. Thus. only square matrices have inverses. 7 35 35 7 7 ⎠ ⎝ ⎠ ⎝ 7 2 2 2 3 35 ⎛ ⎛ 8 ⎞⎞ ⎛ 25 ⎞ ⎛ 5 ⎞⎞ ⎛ The distance between P0 (1. and the column space of AT is the row space of A. Thus since P − Q ≥ 0. 1) is on line m. t) = (1. ⎜ ⎟ = ⎜− . . 2s − 1. ⎟. the distance is 5(1) − 3(−2) + 4 52 + (−3)2 + 12 = 15 35 = 3 35 . so x is a solution of AT Ax = 0.Chapter 6: Inner Product Spaces SSM: Elementary Linear Algebra ⎡ x0 ⎤ ⎡ 10 15 −5⎤ ⎡ x0 ⎤ 1 ⎢ (c) [ P] ⎢ y0 ⎥ = 15 26 3⎥ ⎢ y0 ⎥ ⎢ ⎥ 35 ⎢ ⎥⎢ ⎥ ⎣ −5 3 34 ⎦ ⎣⎢ z0 ⎦⎥ ⎣⎢ z0 ⎦⎥ ⎡ 10 x0 + 15 y0 − 5 z0 ⎤ 1 ⎢ 15 x0 + 26 y0 + 3z0 ⎥ = ⎥ 35 ⎢ ⎣⎢ −5 x0 + 3 y0 + 34 z0 ⎦⎥ ⎡ 10 x0 +15 y0 −5 z0 ⎤ 35 ⎢ ⎥ = ⎢ 15 x0 + 26 y0 +3 z0 ⎥ 35 ⎢ ⎥ ⎢ −5 x0 +3 y0 +34 z0 ⎥ 35 ⎣⎢ ⎦⎥ ⎛ 2 x − 3 y0 − z0 15 x0 + 26 y0 + 3 z0 −5 x0 + 3 y0 + 34 z0 ⎞ The orthogonal projection is ⎜ 0 . so is AT . (c) True. but AT A can be invertible when A is not a square matrix. 4) and W is d = ⎜1 − ⎜ − ⎟ ⎟ + ⎜ −2 − ⎜ − ⎟ ⎟ + ⎜ 4 − ⎟ = . − 2. AT A is an n × n matrix. 7 35 35 ⎝ ⎠ (d) The orthogonal projection of P0 (1. If A has linearly independent column vectors. − . . − 2. 19. 4) on W is 5 25 ⎞ ⎛ 2(1) − 3(−2) − 4 15(1) + 26(−2) + 3(4) −5(1) + 3(−2) + 34(4) ⎞ ⎛ 8 . 1. the point (t. 1) is on line l. the point (s. if A is invertible. True/False 6. 21. (d) True (e) False. ⎟. t. when t = 1. the system AT Ax = AT b may be consistent. (b) False. By inspection. AT will have linearly independent column vectors. 1. 7 7 ⎠ ⎝ 7 ⎠⎠ ⎝ ⎝ ⎝ 7 ⎠⎠ ⎝ Using Theorem 3. but since b is orthogonal to the column space of A.4. x = 0 since AT A is invertible. so the product AT A is also invertible. these are the values of s and t that minimize the distance between the lines. (f) True 164 . Thus.4 (a) True.

the least squares solution may involve a parameter. ⎜ .SSM: Elementary Linear Algebra Section 6. 7). the columns are linearly independent if and only if the second column has at least two different entries. ⎢ ⎥ ⎣ 0 1 2 ⎥⎦ 1 2 ⎣ ⎦ ⎡1 0 ⎤ ⎡ 1 1 1⎤ ⎢ ⎡3 3⎤ MT M = ⎢ . 3 ⎟ . ⎣b⎦ ⎢⎣ 48 ⎥ 7 ⎦ ⎡ 0⎤ ⎢ −10 ⎥ y=⎢ ⎥ ⎢ −48⎥ ⎣⎢ −76 ⎦⎥ 5 48 + X . so the 21 7 5 48 required curve is y = + . 1⎟ . ⎢ ⎥ ⎢1 1 ⎥ ⎣ 1⎦ ⎣ 6⎦ ⎡1 1⎤ ⎡1 1 1⎤ ⎢ 1 ⎥ ⎡ 3 32 ⎤ T ⎥. M = ⎢ ⎢1 x3 ⎢1 x 4 ⎣ x12 ⎤ ⎡1 ⎥ x22 ⎥ ⎢1 ⎥=⎢ x32 ⎥ ⎢1 ⎢1 x42 ⎥⎦ ⎣ 2 4⎤ 3 9⎥ ⎥ 5 25⎥ 6 36 ⎥⎦ ⎡5⎤ ⎡a ⎤ v* = ⎢ ⎥ = ( M T M )−1 M T y = ⎢ 21 ⎥ . then AT A is invertible.5 ⎡1 0⎤ ⎡ 1 1 1⎤ 1. The two column vectors of M are linearly independent if and only neither is a nonzero multiple of the other. With the substitution X = . 5.. M T = ⎢ . .e. the problem x becomes to find a line of the form y = a + b ⋅ X ⎛1 ⎞ that best fits the data points (1. i. 2 2 = ⎡1 x 1 ⎢ ⎢1 x2 3.5 ⎡ a0 ⎤ ⎡ 2⎤ v* = ⎢ a1 ⎥ = ( M T M )−1 M T y = ⎢ 5⎥ ⎢ ⎥ ⎢ ⎥ ⎣ −3⎦ ⎣ a2 ⎦ (g) False. so AT Ax = AT b has a unique solution. xn are distinct. (h) True. ⎝6 ⎠ ⎡1 1⎤ ⎡7 ⎤ ⎢ 1⎥ M = ⎢1 3 ⎥ .. Section 6.5 Exercise Set 6. Since all the entries in the first column are equal. if A has linearly independent column The desired quadratic is y = 2 + 5 x − 3x 2 . if and only if at least two of the numbers x1 . ⎝3 ⎠ ⎛1 ⎞ ⎜ . M = ⎢1 1⎥ . 21 7 x The line in terms of X is y = ⎡1 2 4 ⎤ ⎡ 1 1 1 1⎤ ⎢ 1 3 9⎥ M T M = ⎢2 3 5 6⎥ ⎢ ⎥ ⎢ ⎥ 1 5 25 ⎣ 4 9 25 36 ⎦ ⎢⎢1 6 36 ⎥⎥ ⎣ ⎦ 74 ⎤ ⎡ 4 16 = ⎢16 74 376 ⎥ ⎢ ⎥ ⎣74 376 2018⎦ ⎡ 1989 −1116 135⎤ 1 ⎢ ( M T M )−1 = −1116 649 −80 ⎥ ⎥ 90 ⎢ 135 −80 10 ⎦ ⎣ y 10 x 10 165 .. vectors. M M = ⎢ 1 1 ⎥ ⎢1 3 ⎥ = ⎢ 3 41 1 ⎣ 3 6 ⎦ ⎢1 1 ⎥ ⎢⎣ 2 36 ⎥⎦ ⎣ 6⎦ 1 −54 ⎤ 41 ⎡ ( M T M )−1 = ⎢ 42 ⎣ −54 108⎥⎦ 1 ⎡ 5 − 3 ⎤ 1 ⎡ 5 − 3⎤ = 15 − 9 ⎢⎣ −3 3⎥⎦ 6 ⎢⎣ −3 3⎥⎦ v* = ( M T M )−1 M T y ⎡0⎤ 1 ⎡ 5 −3⎤ ⎡ 1 1 1⎤ ⎢ ⎥ 2 6 ⎣⎢ −3 3⎦⎥ ⎣⎢0 1 2 ⎦⎥ ⎢7 ⎥ ⎣ ⎦ ⎡− 1 ⎤ = ⎢ 72 ⎥ ⎢⎣ 2 ⎥⎦ 1 7 The desired line is y = − + x. 1 1⎥ = ⎣0 1 2 ⎥⎦ ⎢1 2 ⎥ ⎢⎣3 5⎥⎦ ⎣ ⎦ ( M T M ) −1 = 1 11. y = ⎢ 3⎥ .. x2 .

2 g1 1 = ∫ 12 dx = 1 0 1 d12 + d 22 + not the individual terms di . 1] is spanned by (a) False. the sum + d n2 v 2 . Let g1 = v1 = 1. b1 = −2. g1 x. (c) False. g1 g2 = v2 − 1 0 1 2π π ∫0 x sin kx dx 2π 2 1 cos kx − k kπ 0 2 =− k =− 1 x. if the points are not collinear. there is only a unique least squares straight line fit if the data points do not lie on a vertical line. k ⎝ k ⎠ sin 2 x ⎛ f ( x) ≈ (1 + π ) − 2 ⎜ sin x + + 2 ⎝ 1 1 1 2 x = 2 0 2 + 1 1 = − (1 − e + e x ) 2 1− e 1 1− e 1 x = − − e 2 1− e 1− e 1 1 x =− + e 2 e −1 sin nx ⎞ ⎟ n ⎠ 166 .6 ak = = 1 2π π ∫0 1 2π π ∫0 1 2π π ∫0 1 = ∫ (1 − e + e x ) 2 dx 2π (1 + x)dx = 0 1 1⎛ x2 ⎞ = 2 + 2π ⎜x+ ⎟ π⎝ 2 ⎠0 = ∫ (1 − 2e + e2 + 2e x − 2ee x + e 2 x )dx 0 1 1 ⎛ ⎞ = ⎜ x − 2ex + e2 x + 2e x − 2e x +1 + e2 x ⎟ 2 ⎝ ⎠0 3 1 2 = − + 2e − e 2 2 1 = − (1 − e)(3 − e) 2 g1 and g 2 are an orthogonal basis for W. With f(x) = 1 + x: a0 = sin kx dx + = e −1 g1 2 g1 e −1 (1) = ex − 1 = 1− e + ex (d) True = v 2 . g 2 = ∫ x(1 − e + e x )dx 0 3 1 − e 2 2 1 = (3 − e) 2 (a) a0 = 2 + 2π . (a) The space W of continuous functions of the True/False 6.5 form a + be x over [0. a1 = a2 = 0.Chapter 6: Inner Product Spaces SSM: Elementary Linear Algebra 3.6 Exercise Set 6. Thus f ( x) ≈ 1 1 (3 − e) 2 projW x = 2 (1) + (1 − e + e x ) 1 − 1 (1 − e)(3 − e) (b) a1 = a2 = an = 0 2 2 ⎛ sin kx ⎞ bk = − so bk sin kx = −2 ⎜ ⎟. The least squares approximation of x is the orthogonal projection of x on W. g 2 g + g projW x = 2 1 2 2 g1 g2 (1 + x) cos kx dx cos kx dx + 2π 1 2π π ∫0 x cos kx dx 1 sin kx + 0 kπ 0 =0 1 2π bk = ∫ (1 + x)sin kx dx = π 0 1 2π π ∫0 2 g2 1. g1 = ∫ x dx = 0 1 x. Section 6. v1 = 1 and v 2 = e x . (b) True. x. b2 = −1 = 1 (2 + 2π ) − 2 sin x − sin 2 x 2 = (1 + π ) − 2 sin x − sin 2 x. g1 = ∫ e x dx = e x 0 is minimized. there is no solution to the system.

g 2 g2 2 2 − f1 . g1 − g1 2 − 2 1 2 x dx 0 =∫ f1 . 0 < x < π 9. g1 2 f1 . Let f ( x) = ⎨ . 1] is 1 = ∫ x dx = x 2 −1 2 g2 g2 + ⎛ 1 ⎞ f1. 2 2 3 1 v3 . v 2 = x. then the mean square error is f1 − f2 2 −1 . g1 g3 = v3 − 1 1 4 x 4 1 f1 − projW f1 = f1 − f2 and f2 are orthogonal. g 2 . g1 = ∫ sin π x dx = − 1 = 3 Thus. then in a clear extension of the process in Exercise 3(b). π ≤ x ≤ 2π 1 2π 1 π a0 = ∫ f ( x) dx = ∫ dx = 1 g1 = v 2 = x 1 2 1 3 = x 3 −1 3 π 167 0 π 0 6 π2 2 2 . Recall that f1 − f 2 = = = = v3 . g 2 3 =x 3 0 f1 . f1 − f2 − f2 . f 2 f1 .SSM: Elementary Linear Algebra Section 6. Now decompose f2 in terms of g1 and g 2 from part (a). g1 − 2 g2 f1 . g1 = ∫ x 2 dx = =0 2 (2) = 1− 0 − π 2 3 2 − 0 = 1− ⎧1. and v3 = x . f2 2 = f1 2 = f1 2 − = f1 2 f1 2 g1 f1 . g1 1 2 g3 2 2 v 2 . ⎩0. Let g1 = v1 = 1.5 (b) Let f1 = x and f2 = projW f1. g1 g1 =0 v3 . g 2 = ∫ x3 dx = −1 f1 − f 2 . g3} is an orthogonal basis for W. = ∫ 12 dx = x 1−1 = 2 −1 g2 = v2 − 1 2 3 spanned by v1 = 1. the mean square error is 1 2 −1 g3 (b) Let f2 = projW f1. g 2 f1 . f1 − f1. g3 = ∫ ⎜ − + x 2 ⎟ sin π x dx = 0 −1 ⎝ 3 ⎠ − 12 (1 − e)(3 − e) 1 g2 1 5. g 2 2 f1 − f 2 3 1 ( 12 (3 − e) ) 2 − 2 f1 − f 2 2 2 − = f1 g2 2 f1 1 g1 = ∫ x 2 dx = −1 π 1 2 −1 π 2 f1 . The least squares approximation of f1 = sin π x is the orthogonal projection of f1 on W. f1 − f 2 f1. g 2 = ∫ x sin π x dx = 1 1 3−e = − + 3 4 2(1 − e) 1 2 − 2e + 1 + e = + 12 2(1 − e) 13 1+ e = + 12 2(1 − e) g1 g1 f1 . g3 3 projW f1 = 0g1 + π x + 0g3 = x form a0 + a1 x + a2 x 2 over [−1. f1 − f 2 v3 . (a) The space W of continuous functions of the 2 2 f1 . g 2 g1 + f1 . g3 g3 1 = ∫ sin 2 π x dx = 1 −1 f1 − f 2 1 −1 1 cos π x Thus. f1 − f2 f1. v 2 . g1 g1 −1 =0 π 2 2 − f1 . 1 2 1 (2) = − f1 . g1 projW f1 = 2 g2 1 2 2 g2 2 − f1 . f1 − f2 f1. g 2 g1 − 2 −1 = x 2 − 3 (1) − 0 2 1 2 =− +x 3 {g1 .

. x2 . then it must be the case that v1 = v4 = 0 and V must have zeros on the main diagonal. similarly. Let u = v. v + 1)2 ≤ u 2 v 2 or n terms . u 2 = x2 and. u 2 = v2 . so we want x2 = 2 x3 . For V to be in the orthogonal complement of the subspace of all diagonal matrices. it must be the case that v1 = v4 = 0 and v2 = −v3 . +∑ 2 k =1 kπ = ⎡ u u2 ⎤ ⎡ v1 3. u1 = v1. u ⋅ v 2 = (1 + u. v2 . 5. x1. so 1 2 ⎛ ⎞ .. u3 = − x1 + x3 x1 = x4 = 0. 168 2. u 2 = − x2 . v = (0. so x2 = 0 and Since u 2 = u3 = 1 and we want x = 1. u3 = v3 . an ⎛ 1 . u 4 = 0. Thus. x. u1 = v. 2π (d) False. v3 .6 and U . 2 x2 . v2 ⎤ .. x. since x. 0) with a ≠ 0. then u2 = u3 = 0 True/False 6. a. v3 . x = ( x1. u 4 = 0. u 4 = v4 If v. u1 = x1 + x2 − x3 x. (a) Let v = (v1. Recall that if U = ⎢ 1 ⎥ and V = ⎢v u u 4⎦ ⎣ 3 ⎣ 3 then U . then v1 = v4 = 0 and v = (0. for V to be in the orthogonal complement of the subspace of all symmetric matrices. 0⎟. x3 ). and x = 0. ⎠ 7. u3 = 0 ⇒ − x1 + x3 = 0. Since the angle θ between u and v satisfies cos θ = ( a1 . a. . 0). v. the area between the graphs is the error. x2 . Then x. u1 = x2 and x. u3 = x3 . V = u1v1 + u4 v4 . u v then v making equal angles with u 2 and u3 n 2 ≤ (a1 + means that v2 = v3 . 0 . 5 5 ⎠ ⎝ 1 π sin kx dx π ∫0 1 (1 − (−1)k ) kπ So the Fourier series is 1 ∞ 1 (1 − (−1)k ) sin kx. not the mean square error.. . (a) False. then x2 = ± f ( x)sin kx dx 5. (b) True (c) True (b) If U is a symmetric matrix. then the cosine of the angle between x and u 2 is cos θ 2 = x. u 2 = −2 x1 − x2 + 2 x3 (b) As in part (a). V = u1v1 + u2 v2 + u3v3 + u4 v4 . x = ± ⎜ 0. Then x = x12 + x12 = 2 x12 = x1 cos θ3 = x. thus V must be skew-symmetric. . so x1 = x3 . 1 If x = 1. v=⎜ ⎝ a1 ) and 1 ⎞ ⎟ . u1 = x. 0.Chapter 6: Inner Product Spaces 2π 1 ak = π ∫0 bk = 1 = π 2π ∫0 f ( x) cos kx dx = SSM: Elementary Linear Algebra 1 π π x = x22 + 4 x22 = 5 x22 = x2 ∫0 cos kx dx = 0 5 . By the Cauchy-Schwarz an ⎠ Inequality. v4 ). 1 = 1. x1 ).. v2 .. v. (e) True Chapter 6 Supplementary Exercises 1. Let x = ( x1. v4 ⎥⎦ (a) If U is a diagonal matrix. 1 = ∫ 12 dx = 2π ≠ 1 0 Since u1 and u4 can take on any values. V = u1v1 + u2 (v2 + v3 ) + u4 v4 . then u2 = u3 and U . . In order for the angle ⎛1 + a1 ) ⎜ + ⎝ a1 + 1 an ⎞ ⎟. between v and u3 to be defined v ≠ 0. v.

. Then ui = k . 0.. Recall that u can be expressed as the linear combination u = a1 v1 + + an v n where ai = u. u2 ). = = ⎣ 25 35 ⎥⎦ ⎢⎣ 2 ⎥⎦ ⎢⎣95⎥⎦ ⎢⎣5s + 2⎥⎦ The resulting equations have solutions s = 17 and s = 18. approaches π 2 1 n . and cos2 α1 + k (k (b) As n approaches ∞. be the edges of the ‘cube’ in R n and u = (k. w 2 = w1 . . If w is in W. let = u. w2 so that w 2 . n) = 1 n ⎡ 1 −1⎤ ⎡ 1 2 4⎤ 17.. Since v is in V. 0. To show that (W ⊥ )⊥ = W .6.. we have. w 2 = 0. 0. let v be in (W ⊥ ) ⊥ .... Thus ui .. w 2 = w1 ... 0). the ⎢ ⎥ ⎢ ⎥ ⎣0 0 1⎦ ⎣ 3 −2 0 ⎦ system is inconsistent and there is no such weighted inner product. k. 11. w 2 = 0. Thus ⎡ 21 25⎤ ⎡1 ⎤ ⎡ 71⎤ ⎡ 4 s + 3⎤ then ⎢ . u ( vi 15..6 ⎛ u. by the Projection Theorem.. ui . AT A = ⎢ ⎣ 25 35 ⎥⎦ ⎡1 ⎤ ⎡ 1 2 4 ⎤ ⎢ ⎥ ⎡ 4 s + 3⎤ AT b = ⎢ 1 = ⎣ −1 3 5⎥⎦ ⎢ s ⎥ ⎢⎣5s + 2 ⎥⎦ ⎣ ⎦ The associated normal system is ⎡ 21 25⎤ ⎡ x1 ⎤ ⎡ 4s + 3⎤ ⎢⎣ 25 35 ⎥⎦ ⎢ x ⎥ = ⎢⎣5s + 2 ⎥⎦ .. approaches 0. . v2 ) in R 2 . that v = w1 + w 2 where w1 is in W and w 2 is in W ⊥ . . respectively. w 2 + w 2 . ⎢ ⎥ ⎣ −1 3 5⎥⎦ 4 5 ⎣ ⎦ ⎡ 21 25⎤ . so θ . w 2 = w2. . so that v is in W. 2 = a(3)2 + b(−1)2 = 9a + b = 1. vi ⎞ cos αi = ⎜ ⎜ u v ⎟⎟ i ⎠ ⎝ and the vectors are 2 1 ⎞ ⎛ 1 . v. 2⎠ ⎝ 2 ⎛a ⎞ =⎜ i ⎟ ⎝ u ⎠ 9. But v. so that w is in (W ⊥ ) ⊥ . . v = au1v1 + bu2 v2 be a weighted inner v 2 = a(1)2 + b(2)2 = a + 4b = 1. 0). For u = (u1 . so no such value of s exists. If u = (1. A = ⎢ 2 3⎥ . v = (v1 . k . k ) = ai2 = 1) W ⊆ (W ⊥ )⊥ .. u 2 = (0. u 2 = (0. then w is orthogonal ⎡ 1 4⎤ ⎡1 ⎤ ⎡a ⎤ This leads to the system ⎢9 1⎥ ⎢ ⎥ = ⎢1 ⎥ . k. u = k 2 . u = k n . 0. . w 2 = w1 + w 2 . By definition.. To show that (W ⊥ ) ⊥ ⊆ W . 2) and v = (3. AT = ⎢ .. (a) Let u1 = (k . ±⎜ ⎟. cos θ = 2 Therefore product. n. ⎢ ⎥ ⎣b ⎦ ⎢ ⎥ ⎣ 3 −2 ⎦ ⎣0 ⎦ ⎡ 1 4 1⎤ ⎡ 1 0 0⎤ Since ⎢9 1 1⎥ reduces to ⎢0 1 0 ⎥ . k) be the diagonal. 13. 0. ⎣ 2⎦ If the least squares solution is x1 = 1 and x2 = 2. and k2 a12 + a22 + to every vector in W ⊥ . 0. so ui u + an2 Thus W ⊆ (W ⊥ )⊥ . 169 . then u 2 2 (W ⊥ ) ⊥ ⊆ W . we first show that u.. −1) form an orthonormal set. . Hence w 2 = 0 and therefore v = w1. vi for i = 1..SSM: Elementary Linear Algebra If x = 1 then x1 = ± 1 Section 6. v = a(1)(3) + b(2)(−1) = 3a − 2b = 0. + cos2 α n = a12 + a22 + + an2 a12 + a22 + + an2 = 1.

AT A ⎡− 1 0 ⎢ 2 =⎢ 1 − 2 6 ⎢ 6 1 ⎢ 1 ⎢⎣ 3 3 ⎡1 0 0⎤ = ⎢ 0 1 0⎥ ⎢ ⎥ ⎣0 0 1⎦ 1 2 1 6 1 6 − 56 1 ⎤ 3⎥ 1 ⎥ 3⎥ 1 ⎥ 3 ⎥⎦ π 3 Thus. 1 π 3 = . 3 + 3 ) .Chapter 7 Diagonalization and Quadratic Forms A is orthogonal with inverse ⎡− 1 0 1 ⎤ 2⎥ ⎢ 2 1 ⎥ . 5 ⎥ 16 25 ⎥⎦ (e) For the given matrix A. (b) Since A is orthogonal. 2 (c) ⎛ 1 ⎞ r1 = 02 + 12 + ⎜ ⎟ = ⎝ 2⎠ matrix is not orthogonal. (a) For the given matrix A. AT = ⎢ 12 ⎥ 1 1 −5 ⎢2 6 6 6⎥ ⎢1 1 −5 1⎥ ⎢⎣ 2 6 6 6 ⎥⎦ 3. ( x′. ⎡ 4 −9 25 ⎢ 5 −1 T 4 A =A =⎢ 0 5 ⎢ 3 12 ⎢⎣ − 5 − 25 12 ⎤ 25 ⎥ 3⎥. (a) cos (d) For the given matrix A. ⎢− ⎥ 2⎦ ⎣ 2 Note that the given matrix is the standard matrix for a rotation of 45°.1 1. 170 1⎤ 2⎥ 1⎥ 6 ⎥ − 56 ⎥ 1⎥ 6 ⎥⎦ . 2 (f) 3 ≠ 1 so the 2 1 ⎤ ⎡− 2⎥⎢ 1 ⎥⎢ 6⎥⎢ 1 ⎥⎢ 3 ⎥⎦ ⎢⎣ 1 2 0 − 1 2 1 6 2 6 1 6 2 7 ⎛ 1 ⎞ ⎛ 1⎞ ≠1 r2 = ⎜ ⎟ +⎜− ⎟ = 12 ⎝ 3⎠ ⎝ 2⎠ The matrix is not orthogonal. ⎣ 0 1 ⎥⎦ ⎢⎣0 1 ⎥⎦ ⎢⎣ 0 1 ⎥⎦ ⎡1 0 ⎤ A is orthogonal with inverse AT = ⎢ . AT A 1 1 1⎤ ⎡1 1 ⎡1 2 2 2⎥ ⎢2 2 ⎢2 5 5 1 1 1 1 −6 −6 ⎢ ⎥⎢ 6 6 2 = ⎢2 ⎥⎢ 1 1 1 −5 1 1 ⎢2 6 6 6⎥ ⎢2 6 ⎢1 1 −5 1⎥ ⎢1 1 ⎢⎣ 2 6 6 6 ⎥⎦ ⎢⎣ 2 6 ⎡ 1 0 0 0⎤ ⎢0 1 0 0⎥ =⎢ ⎥ ⎢0 0 1 0⎥ ⎣⎢ 0 0 0 1⎦⎥ A is orthogonal with inverse 1 1 1⎤ ⎡1 2 2 2⎥ ⎢2 1 −5 1 1⎥ ⎢ 6 6 6 . ⎡ 1 0 ⎤ ⎡1 0 ⎤ ⎡ 1 0 ⎤ AT A = ⎢ = . y ′) = ( −1 + 3 3. ⎣ 0 1⎥⎦ A is orthogonal with inverse 1 ⎤ ⎡ 1 AT = ⎢ 12 12 ⎥ . AT = ⎢ 1 − 2 6 6⎥ ⎢ 6 1 1 ⎥ ⎢ 1 ⎢⎣ 3 3 3 ⎥⎦ Section 7. sin = 2 3 2 3⎤ ⎡ 1 ⎡ x′ ⎤ ⎢ 2 2 ⎥ ⎡ −2⎤ ⎡ −1 + 3 3 ⎤ = =⎢ ⎥ ⎢⎣ y ′⎥⎦ ⎢ 3 1 ⎥ ⎢⎣ 6 ⎥⎦ ⎣ 3 + 3 ⎦ − ⎣ 2 2⎦ 7. 1 ⎤⎡ 1 ⎡ 1 − 1 ⎤ 2⎥ AT A = ⎢ 12 12 ⎥ ⎢ 12 1 ⎥ ⎢− ⎥⎢ 2⎦⎣ 2 2⎦ ⎣ 2 ⎡ 1 0⎤ =⎢ .1 Exercise Set 7. ⎣ 0 1 ⎥⎦ (b) For the given matrix A.

Thus the transition matrix ⎢ ⎥ ⎣cos θ ⎦ ⎡ cos θ 0 sin θ ⎤ from B′ to B is P = ⎢ 0 1 0 ⎥. and [u3 ]B = ⎢ − sin θ ⎥ . then the transition matrix from B′ to B is ⎡ cos π 0 sin π ⎤ ⎡ 1 0 3 3⎥ ⎢ 2 ⎢ P=⎢ 0 1 0 ⎥=⎢ 0 1 ⎢ − sin π 0 cos π ⎥ ⎢ 3 0 3 3 ⎦ ⎢⎣ − 2 ⎣ 3⎤ 2 ⎥ 0⎥ 1 ⎥ 2 ⎥⎦ ⎡1 ⎤ (b) With the same notation. If B = {u1 . 3 + 1⎟ . ⎢ ⎥ ⎣ sin θ 0 cos θ ⎦ (b) Since a rotation matrix is orthogonal. 2. and ⎢ ⎥ ⎢ ⎥ ⎣ − sin θ ⎦ ⎣0 ⎦ r3 171 2 2 ⎛ 1 ⎞ ⎛ 1 ⎞ 2 2 = c2 + ⎜ ⎟ +⎜ ⎟ =c +3 3 3 ⎝ ⎠ ⎝ ⎠ . so a and b 0 2(a 2 + b2 ) ⎦⎥ ⎣⎢ 1 5⎞ ⎛ 1 5 ( x′. 2 2 ⎛ 1 ⎞ ⎛ 1 ⎞ 2 = a2 + ⎜ ⎟ +⎜− ⎟ = a +1 2 2 ⎝ ⎠ ⎝ ⎠ Thus a = 0. ( x. y. u3} is the standard basis for R3 and B ′ = {u1′ . ⎢ ⎥ ⎣ 0 − sin θ cos θ ⎦ ⎡ x′ ⎤ ⎡ x⎤ (a) ⎢ y ′⎥ = P −1 ⎢ y ⎥ ⎢ ′⎥ ⎢ ⎥ ⎣z ⎦ ⎣z⎦ ⎡ 1 0 − 3⎤ 2 ⎥ ⎡ −1⎤ ⎢ 2 =⎢ 0 1 0 ⎥ ⎢ 2⎥ ⎢ 3 ⎥ ⎢ 5⎥ 1 ⎣ ⎦ ⎢⎣ 2 0 2 ⎥⎦ ⎡− 1 − 5 3 ⎤ ⎢ 2 2 ⎥ =⎢ 2 ⎥ ⎢− 1 3 + 5 ⎥ 2⎦ ⎣ 2 Thus ⎡a + b b − a ⎤ 13. then ⎡ cos θ ⎤ ⎡0 ⎤ [u1′ ]B = ⎢ 0 ⎥ . Then ⎣ a − b b + a ⎥⎦ ⎡ 2(a 2 + b 2 ) ⎤ 0 AT A = ⎢ ⎥ . u 2 . 6. Let A = ⎢ . = = ⎢⎣ y ⎥⎦ ⎢ 3 1 ⎥ ⎢⎣ 2 ⎥⎦ ⎢ 5 3 + 1⎥ 2 ⎣ ⎦ 2⎦ ⎣2 5 ⎛5 ⎞ Thus. z ) = ⎜ − 3. Equation (2) gives 3⎤ ⎡ 1 ⎡5 ⎤ ⎡ x ⎤ ⎢ 2 − 2 ⎥ ⎡5⎤ ⎢ 2 − 3 ⎥ . Thus b = ± 6 2 11. ⎝2 2 2 2⎠ 17. − 3 + ⎟. ⎢ ⎥ θ θ⎦ sin 0 cos − ⎣ ⎡x⎤ ⎡ x′ ⎤ i. ⎢ y ⎥ = P ⎢ y ′⎥ . [u1′ ]B = ⎢ 0 ⎥ . u 2 . u′2 . Then ⎢ ⎥ ⎢ ′⎥ ⎣z⎦ ⎣z ⎦ ⎡cos θ 0 − sin θ ⎤ −1 ⎢ A=P = 0 1 0 ⎥.. ⎢ ⎥ ⎣0⎦ ⎡ 0 ⎤ ⎡ 0 ⎤ ⎢ ⎥ ′ ′ [u 2 ]B = cos θ . [u′2 ]B = ⎢1 ⎥ . u3′ } is the rotated basis. ⎝ 2 2 2 2⎠ ⎡ 1 ⎡x⎤ ⎡ x′ ⎤ ⎢ 2 (b) ⎢ y ⎥ = P ⎢ y ′⎥ = ⎢ 0 ⎢ ⎥ ⎢ ⎥ ⎢ ⎣z⎦ ⎣ z′ ⎦ ⎢− 3 ⎣ 2 0 1 0 3⎤ 2 ⎥⎡ 1⎤ 0 ⎥ ⎢ 6⎥ ⎢ ⎥ 1 ⎥ ⎣ − 3⎦ 2 ⎥⎦ 1 must satisfy a 2 + b 2 = . u′3}. and B ′ = {u1′ . The row vectors of an orthogonal matrix are an orthonormal set.1 ⎡ sin θ ⎤ [u3′ ]B = ⎢ 0 ⎥ . so ⎢ ⎥ ⎢ ⎥ ⎣ sin θ ⎦ ⎣ cos θ ⎦ the transition matrix from B′ to B is 0 0 ⎤ ⎡1 P = ⎢ 0 cos θ − sin θ ⎥ and ⎢ ⎥ ⎣ 0 sin θ cos θ ⎦ 0 0 ⎤ ⎡1 A = ⎢ 0 cos θ sin θ ⎥ . u3} is the standard basis for R3 .SSM: Elementary Linear Algebra Section 7. − 3 − ⎟. u′2 . y ′. (a) If B = {u1 . y) = ⎜ − 3. ⎝2 ⎠ 2 9.e. 2 ⎡ 1−3 3 ⎤ ⎢ 2 2 ⎥ =⎢ 6 ⎥ ⎢− 1 3 − 3 ⎥ ⎣ 2 2⎦ 1 3⎞ ⎛1 3 Thus ( x. z ′) = ⎜ − − 3. r1 2 2 2 2 2 ⎛ 1 ⎞ ⎛ 1 ⎞ 2 2 r2 = b + ⎜ ⎟ +⎜ ⎟ =b +6 ⎝ 6⎠ ⎝ 6⎠ 2 .

the column vectors must form an orthonormal set. (h) True. from which it is clear that b and c must have opposite signs. only square matrices can be orthogonal. dilations. (a) False. so A λ −1 4 −2 4 2 = λ3 − 27λ − 54 λ −1 2 −2 λ+2 = (λ − 6)(λ + 3)2 The characteristic equation is (b) Rotations about the origin.1 λ − 1 −1 −1 −1 λ − 1 −1 = λ3 − 3λ 2 = λ 2 (λ − 3) −1 −1 λ − 1 The characteristic equation is λ3 − 3λ 2 = 0 and the eigenvalues are λ = 3 and λ = 0. the eigenspace for λ = 2 is two-dimensional.3. the eigenspace for λ = 0 is two-dimensional. (f) True. a product of orthogonal matrices is orthogonal.2 . λ − 4 −2 −2 −2 λ − 4 −2 = λ3 − 12λ 2 + 36λ − 32 −2 −2 λ − 4 = (λ − 8)(λ − 2) 2 The characteristic equation is (e) True. the eigenspace for λ = −3 is two-dimensional. The eigenspace for λ = 6 is one-dimensional. reflections about any line through the origin. (g) True.2 1. c=− or 6 3 2 1 . since AT A = I for an orthogonal matrix A. λ3 − 12λ 2 + 36λ − 32 = 0 and the eigenvalues are λ = 8 and λ = 2. (b) False. for any nonzero vector x. It follows that x 1 Ax = 1 and Ax = x . the row and column vectors are not unit vectors. vector. The eigenspace for λ = 3 is one-dimensional. The eigenspace for λ = 0 is three-dimensional. only square matrices can be orthogonal. (a) The characteristic equation is λ 2 − 5λ = 0 and the eigenvalues are λ = 0 and λ = 5. and any combination of these are rigid operators. (c) True/False 7. A must be invertible. b = .1. λ − 1 −2 = λ 2 − 5λ −2 λ − 4 (e) x is a unit x λ − 4 −4 −4 λ − 4 0 0 0 0 0 0 λ 0 0 0 = λ 4 − 8λ3 = λ3 (λ − 8) 0 λ The characteristic equation is λ 4 − 8λ3 = 0 and the eigenvalues are λ = 0 and λ = 8. x = 1.Chapter 7: Diagonalization and Quadratic Forms Thus c = ± 1 SSM: Elementary Linear Algebra Section 7. Thus the only 2 1 possibilities are a = 0. b = − 2 3 Exercise Set 7. so A is orthogonal x by Theorem 7. Dilations and contractions are angle preserving operators that are not rigid. (c) All rigid operators on R 2 are angle preserving. the eigenspace for λ = 8 is one-dimensional. (d) (d) False. a = 0. and combinations of these are angle preserving. 172 . hence det( A2 ) = (det A)2 = (±1)2 = 1. reflections about lines through the origin. (a) Rotations about the origin. since Ax = x for an orthogonal matrix. Both eigenspaces are one-dimensional. 21. 3 The column vectors of an orthogonal matrix are an orthonormal set. (c) False. (b) λ3 − 27λ − 54 = 0 and the eigenvalues are λ = 6 and λ = −3. so A2 is orthogonal. contractions. The eigenspace for λ = 8 is one-dimensional. c= .

2 ⎡3⎤ ⎡3⎤ 4 ⎢5⎥ x x3 = ⎢ 0 ⎥ . v3 = 3 = ⎢ 0 ⎥ is an orthonormal ⎢ ⎥ x3 ⎢4⎥ ⎣⎢ 1 ⎦⎥ ⎣5⎦ 4 ⎡− 0 35 ⎤ ⎢ 5 ⎥ basis. 625 = (λ + 25)2 (λ − 25)2 = 0. ⎢⎥ ⎣1⎦ 1 ⎡ ⎤ ⎢ 3⎥ x v1 = 1 = ⎢ 1 ⎥ is an orthonormal basis. The Gram-Schmidt process gives the ⎢ ⎥ ⎣ 1⎦ ⎡− 1 ⎤ ⎡− 1 ⎤ ⎢ 6⎥ ⎢ 2⎥ 1 1 orthonormal basis v 2 = ⎢ ⎥ . v3 = ⎢− ⎥ . v1 = ⎢ ⎥ x1 ⎢ 3 ⎥ ⎢⎣ 1⎥⎦ ⎣ 5⎦ basis. ⎢ ⎥ ⎣ 0⎦ is an ⎡− ⎢ orthonormal basis. The characteristic equation of A is λ 4 − 1250λ 2 + 390. 6⎥ ⎢ ⎢ 2⎥ 2 ⎥ ⎢ ⎣⎢ 0 ⎦⎥ ⎣⎢ 6 ⎦⎥ 5. ⎢ ⎥ ⎣ 0⎦ − 1 ⎤ 6⎥ − 1 ⎥ orthogonally 6⎥ 2 ⎥ 0 6 ⎥⎦ 1 2 1 2 0 ⎤ ⎡0 0 0 ⎤ 0 ⎥ = ⎢0 3 0 ⎥ . ⎥ ⎢ ⎥ λ 2 ⎦ ⎣0 0 3 ⎦ 9. λ 2 = −3: A basis for the eigenspace is ⎡ 0⎤ x 2 = ⎢1 ⎥ = v 2 . The characteristic equation of A is λ3 + 28λ 2 − 1175λ − 3750 = (λ − 25)(λ + 3)(λ + 50) = 0. λ1 = −25: A basis for the eigenspace is λ3 = −50: A basis for the eigenspace is 173 . Both eigenspaces are two-dimensional. λ 2 = 10: A basis for the eigenspace is ⎡ ⎡ 3⎤ x x2 = ⎢ 2 ⎥ . x1 ⎢ 3 ⎥ ⎢1 ⎥ ⎢⎣ 3 ⎥⎦ ⎡ −1⎤ λ 2 = 3: A basis for the eigenspace is x 2 = ⎢ 1⎥ . λ1 = 3: A basis for the eigenspace is 7. ⎡− 2 ⎤ ⎡− 2 ⎤ x1 ⎢ 7 ⎥ x1 = ⎢ 3 ⎥ . v 2 = 2 = ⎢ ⎢ x2 ⎣⎢ 1 ⎦⎥ ⎢⎣ 3⎤ 7⎥ 2 ⎥ 7 ⎥⎦ ⎡1⎤ λ1 = 0: A basis for the eigenspace is x1 = ⎢1⎥ . The characteristic equation of A is λ 2 − 13λ + 30 = (λ − 3)(λ − 10) = 0. λ1 = 25: A basis for the eigenspace is ⎡1 − ⎢ 3 Thus. Thus P = ⎢ 0 1 0 ⎥ orthogonally ⎢ 3 0 4⎥ 5⎦ ⎣ 5 diagonalizes A and 0 ⎤ ⎡ 25 0 ⎡λ1 0 0⎤ P −1 AP = ⎢ 0 λ 2 0 ⎥ = ⎢ 0 −3 0⎥ . 3. Thus P = ⎢ ⎢⎣ orthogonally diagonalizes A and 0 ⎤ ⎡ 3 0⎤ ⎡λ P −1 AP = ⎢ 1 ⎥=⎢ ⎥. P = ⎢ 1 ⎢ 3 ⎢1 ⎢⎣ 3 diagonalizes A and ⎡λ1 0 P −1 AP = ⎢ 0 λ 2 ⎢ ⎣0 0 ⎡− 4 ⎤ ⎡− 4 ⎤ 3⎥ x1 ⎢ 5 ⎥ ⎢ = ⎢ 0 ⎥ is an orthonormal x1 = 0 . since x 2 = 1. The characteristic equation of A is λ3 − 6λ 2 + 9λ = λ(λ − 3)2 = 0. λ 0 2 ⎦ ⎣ 0 10 ⎦ ⎣ 2 7 3 7 3⎤ 7⎥ 2 ⎥ ⎥ 7⎦ ⎡ −1⎤ x3 = ⎢ 0⎥ . ⎢ ⎥ ⎢ ⎥ − 0 0 50 ⎦ ⎣ 0 0 λ3 ⎦ ⎣ λ−2 1 0 0 1 λ−2 0 0 0 0 λ−2 1 0 0 1 λ−2 = λ 4 − 8λ3 + 22λ 2 − 24λ + 9 = (λ − 1)2 (λ − 3)2 The characteristic equation is λ 4 − 8λ3 + 22λ 2 − 24λ + 9 = 0 and the eigenvalues are λ = 1 and λ = 3.SSM: Elementary Linear Algebra (f) Section 7. v1 = = is an x1 ⎢ 3 ⎥ ⎢⎣ 1 ⎥⎦ ⎢⎣ 7 ⎥⎦ orthonormal basis.

both AAT and AT A are symmetric. ⎢ 5⎥ ⎢ 3⎥ ⎢⎣ 5 ⎥⎦ λ 2 = 25: A basis for the eigenspace is 15. (e) True. since v1 and v 2 are from distinct eigenspaces of a symmetric matrix. v1 + 2 v1. 0 0⎤ ⎡ −25 0 ⎢ 0 25 0 0⎥ =⎢ ⎥. For instance. follow the steps on page 399. x 2 = ⎢ 4 ⎥ .2 ⎡3⎤ ⎢5⎥ 4 process gives the orthonormal basis v3 = ⎢ 5 ⎥ . to diagonalize A. Since the eigenvalues of a symmetric matrix must be real numbers. since Ax = x for an orthogonal matrix. (c) False. ⎢ 0 0 −25 0 ⎥ 0 25⎥⎦ ⎢⎣ 0 0 (f) True. it must be symmetric. ⎡3⎤ ⎡0⎤ ⎢4⎥ ⎢0⎥ x3 = ⎢ 1 ⎥ . 17. which form a basis for R n . No. ⎢5⎥ ⎢4⎥ ⎢⎣ 5 ⎥⎦ ⎡− 4 3 0 0⎤ ⎢ 5 5 ⎥ ⎢ 53 54 0 0⎥ P = [ v1 v 2 v3 v 4 ] = ⎢ ⎥ 4 3 ⎢ 0 0 − 5 5⎥ ⎢ 3 4⎥ 0 0 5 5 ⎦⎥ ⎣⎢ orthogonally diagonalizes A and ⎡λ1 0 0 0 ⎤ ⎢0 λ 0 0⎥ 2 P −1 AP = ⎢ ⎥ ⎢ 0 0 λ1 0 ⎥ ⎢⎣ 0 0 0 λ 2 ⎥⎦ (a) True. if A is an n × n orthogonally diagonalizable matrix. Thus. an orthogonal matrix is not necessarily symmetric. Yes. ⎡3 0 ⎤ the eigenvalues of the matrix ⎢ are 3 and ⎣8 −1⎥⎦ −1. for any square matrix A. The Gram-Schmidt ⎢ 0⎥ ⎢− 3 ⎥ ⎢ 0⎥ ⎢ 1⎥ ⎣ ⎦ ⎣ ⎦ ⎡− 4 ⎤ ⎢ 5⎥ 3 process gives the orthonormal basis v1 = ⎢ 5 ⎥ . they are orthogonal. If A is an orthogonal matrix then its eigenvalues have absolute value 1. a non-symmetric matrix A can have eigenvalues that are real numbers. Since eigenvectors from distinct eigenvalues are orthogonal. this means that P will be an orthogonal matrix. hence orthogonally diagonalizable. v 2 = v1 2 + 0 + v2 2 . v 2 + v 2 . so is ( P −1 AP)−1 . 174 . v1 + v 2 = v1. then A has an orthonormal set of n eigenvectors. ⎢ ⎥ ⎢0⎥ ⎢⎣ 0 ⎥⎦ ⎡0⎤ ⎢0⎥ v 4 = ⎢ 3 ⎥ . Then since A is orthogonally diagonalizable. but may be complex. (b) True. so v1 + v 2 2 = v1 + v 2 . 19. The Gram-Schmidt ⎢0⎥ ⎢4⎥ ⎢0⎥ ⎢1 ⎥ ⎣ ⎦ ⎣ ⎦ True/False 7. The Gram-Schmidt process will ensure that columns of P corresponding to the same eigenvalue are an orthonormal set. (d) True. ⎢ ⎥ ⎢ 0⎥ ⎢⎣ 0 ⎥⎦ ⎡ 0⎤ ⎢ 0⎥ v 2 = ⎢− 4 ⎥ . ( P −1 AP)−1 = P −1 A−1P since P −1 and A−1 both exist and since P −1 AP is diagonal. the only possible eigenvalues for an orthogonal symmetric matrix are 1 and −1.Chapter 7: Diagonalization and Quadratic Forms SSM: Elementary Linear Algebra ⎡− 4 ⎤ ⎡ 0⎤ ⎢ 3⎥ ⎢ 0⎥ x1 = ⎢ 1⎥ . x 4 = ⎢ 3 ⎥ .

1.3 ⎡− 2 ⎤ ⎢ 3⎥ bases for the eigenspaces are λ = 1: ⎢ 23 ⎥ . then 1 ⎥ ⎢− 1 2⎦ ⎣ 2 ⎡ 3 0 ⎤ ⎡ y1 ⎤ xT Ax = yT ( PT AP)y = [ y1 y2 ] ⎢ ⎢ ⎥ ⎣ 0 1 ⎥⎦ ⎣ y2 ⎦ 0 x 2 + y 2 + 0 xy + (7 x − 8 y) − 5 = 0 or [x ⎡0 0 ⎤ ⎡ x ⎤ ⎡ x⎤ + [7 −8] ⎢ ⎥ − 5 = 0. (a) 2 x 2 + xy + x − 6 y + 2 = 0 can be written as 2 x 2 + 0 y 2 + xy + ( x − 6 y) + 2 = 0 or ⎡ 2 −1⎤ ⎡ x1 ⎤ 5. ⎢ 1 ⎥ ⎣ 2⎦ 1 ⎤ ⎡ 1 2 2⎥ ⎢ For P = .3 ⎡2⎤ ⎢3⎥ λ = 4: ⎢ 13 ⎥ . Q = x Ax = [ x1 x2 ] ⎢ ⎢ ⎥ ⎣ −1 2 ⎥⎦ ⎣ x2 ⎦ The characteristic equation of A is T [x λ 2 − 4λ + 3 = (λ − 3)(λ − 1) = 0. ⎣ y⎦ (b) y 2 + 7 x − 8 y − 5 = 0 can be written as ⎡ 1 ⎤ the eigenspaces are λ = 3: ⎢ 12 ⎥ . so the eigenvalues of A will all be real numbers. Section 7. ⎢ 2⎥ ⎢⎣ − 3 ⎥⎦ (d) x 2 + y 2 − 25 = 0 is x 2 + y 2 = 25 which is the equation of a circle. 4. Orthonormal 175 . ⎡ 3 2 0 ⎤ ⎡ x1 ⎤ 7. so the eigenvalues are λ = 3. then A must be symmetric. and Q = 3 y12 + y22 . (a) 2 x 2 + 5 y 2 = 20 is (b) x 2 − y 2 − 8 = 0 is x 2 − y 2 = 8 or x2 y 2 − = 1 which is the equation of a 8 8 hyperbola.SSM: Elementary Linear Algebra Section 7. if A is orthogonally diagonalizable. 3 ⎥ 2 − 3 ⎥⎦ let x = Py. 2 λ − 12λ + 39λ − 28 = (λ − 1)(λ − 4)(λ − 7) = 0. so the eigenvalues are λ = 1. Q = x Ax = [ x1 x2 x3 ] ⎢ 2 4 −2 ⎥ ⎢ x2 ⎥ ⎢ ⎥⎢ ⎥ 5⎦ ⎣ x3 ⎦ ⎣ 0 −2 The characteristic equation of A is (c) 7 y 2 − 2 x = 0 is x = T 3 2 3 1 3 2 3 ⎡ 1⎤ ⎢ 3⎥ ⎢ 23 ⎥ .3 1. ⎢ 1⎥ ⎣⎢ 3 ⎦⎥ (g) True. (a) 3x12 + 7 x22 = [ x1 ⎡ 3 0 ⎤ ⎡ x1 ⎤ x2 ] ⎢ ⎢ ⎥ ⎣ 0 7 ⎥⎦ ⎣ x2 ⎦ (b) 4 x12 − 9 x22 − 6 x1 x2 = [ x1 ⎡− 2 ⎢ 3 For P = ⎢ 23 ⎢ 1 ⎢⎣ 3 ⎡ 4 −3⎤ ⎡ x1 ⎤ x2 ] ⎢ ⎢ ⎥ ⎣ −3 −9 ⎦⎥ ⎣ x2 ⎦ (c) 9 x12 − x22 + 4 x32 + 6 x1 x2 − 8 x1 x3 + x2 x3 ⎡ 9 3 −4 ⎤ ⎡ x ⎤ 1 ⎢ 1⎥ = [ x1 x2 x3 ] ⎢ 3 −1 2 ⎥ ⎢ x2 ⎥ ⎢ ⎥ ⎢ −4 1 4 ⎥⎦ ⎣ x3 ⎦ ⎣ 2 3. 11. λ = 7: ⎢2⎥ ⎢⎣ 3 ⎥⎦ Exercise Set 7. 7 2 y which is the 2 equation of a parabola. Orthonormal bases for ⎡2 y] ⎢ ⎢⎣ 12 1⎤ 2 ⎥ ⎡ x ⎤ + [1 0⎥⎦ ⎢⎣ y ⎥⎦ ⎡ x⎤ −6] ⎢ ⎥ + 2 = 0. ⎡ 2 −3⎤ ⎡ x ⎤ = 2 x 2 + 5 y 2 − 6 xy y] ⎢ ⎥ ⎢ ⎥ − 3 5 y ⎣ ⎦⎣ ⎦ 9. y] ⎢ ⎥ ⎢ ⎥ ⎣0 1 ⎦ ⎣ y ⎦ ⎣ y⎦ x2 y 2 + = 1 which is the 10 4 equation of an ellipse. 7. [ x 1⎤ 3⎥ 2⎥. then xT Ax = yT ( PT AP)y = [ y1 y2 ⎡1 0 0 ⎤ ⎡ y1 ⎤ and y3 ] ⎢ 0 4 0 ⎥ ⎢ y2 ⎥ ⎢ ⎥⎢ ⎥ ⎣ 0 0 7 ⎦ ⎣ y3 ⎦ Q = y12 + 4 y22 + 7 y32 . let x = Py. λ = 1: ⎢− ⎥ ⎣ 2⎦ ⎡ 1 ⎤ ⎢ 2 ⎥.

so the matrix is positive semidefinite. = 3. thus A is positive definite. ⎢− ⎥ 5⎦ ⎣ 5 Note that det(P) = 1. The equation can be written in matrix form as ⎡ 2 −2 ⎤ xT Ax = −8 where A = ⎢ . thus the conic is a hyperbola. ⎝ 2⎠ 19. with 23. 13. so the matrix is negative definite. The angle through which the axes have been ⎛ 1⎞ rotated is θ = tan −1 ⎜ − ⎟ ≈ −26. Thus the matrix ⎣ 4⎦ ⎡4 − 3⎤ 5 ⎥ orthogonally diagonalizes A. x2 = 0 and Q < 0 for x1 = 0.6°. The ⎣ −2 −1⎥⎦ eigenvalues of A are λ1 = 3 and λ 2 = −2. 15. Since 5 = 5 and 5 −2 = 21 are positive. λ = 3. thus Q is positive semidefinite. we reach the −2 5 same conclusion using Theorem 7. x2 ≠ 0. so P is a rotation matrix. ⎡ 2 −1 0 ⎤ (b) The eigenvalues of A = ⎢ −1 2 0 ⎥ are ⎢ ⎥ ⎣ 0 0 5⎦ λ = 1. We have Q = x12 + x22 > 0 for ( x1 . (a) The eigenvalues of the matrix ⎡ 5 −2 ⎤ are λ = 3 and λ = 7. 0). so P is a rotation matrix. 21. ⎡4⎤ corresponding eigenvectors v1 = ⎢ ⎥ and ⎣3⎦ ⎡ −3⎤ v 2 = ⎢ ⎥ respectively. The equation can be written in matrix form as ⎡11 12 ⎤ xT Ax = 15 where A = ⎢ . ⎡ 3 0 ⎤ ⎡ x′ ⎤ = −8 which can be written y ′] ⎢ ⎣ 0 −2 ⎦⎥ ⎣⎢ y ′⎦⎥ as 2 y ′2 − 3x′2 = 8. and λ = 5.Chapter 7: Diagonalization and Quadratic Forms SSM: Elementary Linear Algebra ⎡ −1 0 ⎤ (b) The eigenvalues of ⎢ are λ = −1 ⎣ 0 −2 ⎥⎦ and λ = −2. ⎡1 0 ⎤ (d) The eigenvalues of ⎢ are λ = 1 and ⎣0 0 ⎥⎦ λ = 0. with ⎡ 2⎤ corresponding eigenvectors v1 = ⎢ ⎥ and ⎣ −1⎦ ⎡1 ⎤ v 2 = ⎢ ⎥ respectively. thus Q is indefinite. so the matrix is positive definite. so the matrix is indefinite.4. Note P = ⎢5 4⎥ ⎢ 35 5⎦ ⎣ that det(P) = 1. thus we reach the same 0 0 5 conclusion using Theorem 7. (a) The eigenvalues of ⎢ are λ = 1 and ⎣0 2 ⎥⎦ λ = 2. ⎡0 0 ⎤ (e) The eigenvalues of ⎢ are λ = 0 and ⎣0 −2 ⎥⎦ λ = −2. thus Q is positive definite. The angle through which the axes have been ⎛3⎞ rotated is θ = tan −1 ⎜ ⎟ ≈ 36. and −1 2 ⎡ 20 0⎤ ⎡ x ′ ⎤ = 15 which we can write y ′] ⎢ ⎣ 0 −5⎦⎥ ⎣⎢ y ′⎦⎥ as 4 x ′2 − y ′2 = 3.3. We have x12 − x22 > 0 for x1 ≠ 0. so the matrix is negative semidefinite. Thus the matrix ⎣2⎦ 1 ⎤ ⎡ 2 P = ⎢ 15 25 ⎥ orthogonally diagonalizes A. ⎝4⎠ 2 −1 0 −1 2 0 = 15. We have Q = ( x1 − x2 )2 > 0 for x1 ≠ x2 and Q = 0 for x1 = x2 . ⎡1 0 ⎤ 17. thus the conic is a hyperbola. thus A is A=⎢ 5⎥⎦ ⎣ −2 positive definite. 176 .9°. The determinants of the principal 2 −1 submatrices are 2 = 2.3. The equation of the conic in the rotated x ′y ′-coordinate system is [ x′ ⎡ −1 0 ⎤ are λ = −1 and (c) The eigenvalues of ⎢ ⎣ 0 2 ⎥⎦ λ = 2.4. x2 ) ≠ (0. The ⎣12 4 ⎥⎦ eigenvalues of A are λ1 = 20 and λ 2 = −5. The equation of the conic in the rotated x ′y ′-coordinate system is [ x′ 25.

. because of the term 4 x1 x2 x3 . (d) True. The eigenvalues of A must be positive and equal to each other.. A must have a positive eigenvalue of multiplicity 2. if and only if x1 = x2 = " = xn . (c) True. and s x2 = 0 if and only if x1 = x . n we have ( xi − x )2 = xi2 − 2 xi x + x 2 2 n ⎞ 1 n 1 ⎛ ⎜ ∑ xj ⎟ = xi2 − 2 xi ∑ x j + n j =1 n2 ⎜⎝ j =1 ⎟⎠ n n n ⎞ 2 n 1 ⎛ 2 = xi2 − ∑ xi x j + ⎜ ∑ x j + 2 ∑ ∑ x j xk ⎟ . The quadratic form Q = 5 x12 + x22 + kx32 + 4 x1 x2 − 2 x1 x3 − 2 x2 x3 can be expressed in matrix notation as Q = xT Ax ⎡ 5 2 where A = ⎢ 2 1 ⎢ ⎣ −1 −1 5 2 −1 2 1 −1 = k − 2. . It follows n − 1 ⎢⎣ n n 2 ⎥⎦ n n − 1 ⎢⎣ n n n 2 ⎥⎦ n(n − 1) ⎡ 1 ⎢ n ⎢− 1 that s x2 = xT Ax where A = ⎢ n( n −1) ⎢ # ⎢− 1 ⎣⎢ n( n −1) − n( n1−1) " − n( n1−1) ⎤ ⎥ 1 1 ⎥ " − n n( n −1) ⎥ . x ⋅ x = x12 + x22 + " + xn2 177 . True/False 7.. −1 −1 k −1⎤ 5 2 = 1. (b) We have s x2 = n −1 x2 = x . none of the eigenvalues will be 0.. a symmetric matrix can be positive semidefinite or negative semidefinite. 2 1 ⎥ k⎦ Thus Q is positive definite if and only if k > 2. (a) For each i = 1. 2 ⎟ n j =1 n ⎜⎝ j =1 j =1 k = j +1 ⎠ 1 Thus in the quadratic form s x2 = [( x1 − x )2 + ( x2 − x )2 + " + ( xn − x )2 ] the coefficient of xi2 is n −1 1 ⎡ 2 1 ⎤ 1 1 ⎡ 2 2 2 ⎤ 2 1− + n = . ( x1 − 3x2 )2 = x12 − 6 x1 x2 + 9 x22 ... The determinants of the principal submatrices of A are 5 = 5.e. xn = x . 33. # % # ⎥ 1 ⎥ − n( n1−1) " ⎥⎦ n 1 [( x1 − x ) 2 + ( x2 − x )2 + " + ( xn − x )2 ] ≥ 0.. (f) True (g) True. 31.3 27. That is. (e) False.SSM: Elementary Linear Algebra Section 7. . and −1⎥ . Thus s x2 is a positive semidefinite form.3 (a) True (b) False. and the coefficient of xi x j for i ≠ j is n =− − − + . i.

Thus the constrained maximum is 5 ⎣1 ⎦ occurring at ( x. 0. The first partial derivatives of f are f x ( x. this is only true for symmetric matrices. The ⎢ ⎥ ⎣0 0 3⎦ eigenvalues of A are λ1 = 9. (0. ⎢ ⎥ ⎣0 ⎦ 2 . matrix notation as xT Ax where A = ⎢ ⎣ 0 −1⎥⎦ The eigenvalues of A are λ1 = 5 and λ 2 = −1. ±1). y) = (±1. Rewrite the constraint as ⎜ ⎟ + ⎜ ⎟ = 1. ⎡3 0⎤ . 0). Thus the constrained maximum is ⎣0 ⎦ 7 occurring at (x. and the constrained minimum is 3 occurring at (x. 1 ⎥ 2⎦ Thus the minimum value is z = − 2 which occurs at ( x. y. xT Ax = c has no graph. ± 1). 5x2 − y2 = 5 y 5 x2 + y 2 = 1 5. y. there will be no cross product terms if aij = −a ji for all i ≠ j. (0. y = −1. z) = (±1. Thus the constrained ⎢ ⎥ ⎢ ⎥ ⎣0 ⎦ ⎣1 ⎦ maximum is 9 occurring at (x. and the constrained ⎡ 0 y1 ] ⎢ ⎣ 2 2 ⎤ ⎡ x1 ⎤ ⎥ ⎢ ⎥. Since ⎡1 ⎤ with corresponding eigenvectors v1 = ⎢ ⎥ and ⎣0 ⎦ ⎡0 ⎤ v 2 = ⎢ ⎥ . y) = (0. 0) 5 13. ±1). matrix notation as xT Ax where A = ⎢ ⎣0 7 ⎥⎦ The eigenvalues of A are λ1 = 7 and λ 2 = 3. 0).4 x12 + y12 = 1. y) = (±1. The quadratic form 3x 2 + 7 y 2 can be written in 1 ⎤ 2⎥ .Chapter 7: Diagonalization and Quadratic Forms SSM: Elementary Linear Algebra ⎡0 ⎤ ⎡0 ⎤ v 2 = ⎢1 ⎥ . Write 1. 0). − 1) . 1) and 9. (h) True (i) False. 3. if c < 0. 0 ⎦ ⎣ y1 ⎦ λ − 2 = λ 2 − 2 = ( λ + 2 )( λ − 2 ) . (j) True (k) False. 1 ⎤ 2⎥ . v3 = ⎢0 ⎥ . 2 2 ⎛x⎞ ⎛ y ⎞ 7. y) = 3 x 2 − 3 y and f y ( x. λ 2 = 4. The quadratic form 9 x 2 + 4 y 2 + 3z 2 can be ( x (1. y) = (0. –1) T ⎡1 ⎤ with corresponding eigenvectors v1 = ⎢0 ⎥ . − 2 λ the largest eigenvalue of A is 2. 1 ⎥ 2⎦ x = 2 x1 = 2. ⎡− with corresponding unit eigenvectors ± ⎢ ⎢ ⎣ ⎡0 ⎤ with corresponding unit eigenvectors v1 = ⎢ ⎥ ⎣1 ⎦ ⎡1 ⎤ and v 2 = ⎢ ⎥ . λ3 = 3.4 now to maximize z = xy = 2 2 x1 y1 subject to Exercise Set 7. z) = (0. y = 2 y1 = 1 or x = − 2. The quadratic form 5 x 2 − y 2 can be written in z = xT Ax = [ x1 ⎡5 0⎤ . and the constrained minimum is 3 occurring at (x. To find the critical points we set f x and f y 178 . with ⎡ corresponding positive unit eigenvector ⎢ ⎢ ⎣ Thus the maximum value is ⎛ 1 ⎞⎛ 1 ⎞ z = 2 2⎜ ⎟⎜ ⎟ = 2 occurs when ⎝ 2 ⎠⎝ 2 ⎠ minimum is −1 occurring at ( x. y) = ( − 2 . then let x = 2 x1 and y = 2 y1 . 0) ⎡9 0 0 ⎤ expressed as x Ax where A = ⎢ 0 4 0 ⎥ . ⎝2⎠ ⎝ 2 ⎠ (l) False. 0. y) = −3x − 3 y 2 . 1) 5 x 2 − y 2 = −1 (–1. The problem is Section 7. The smallest eigenvalue of A is − 2.

Thus there are three critical points: (0. this matrix is positive definite and so f has a relative minimum at (0.5 17. A* = ⎢ + −i 1 i 3 0 ⎥⎦ ⎣ 179 . The eigenvalues of ⎡ −6 −3⎤ are λ = −3 and H (−1. Let ⎝5⎠ ⎝1⎠ x = 5 x1 and y = y1. thus f ⎣ 4 4⎥⎦ has a saddle point at (−2. 0) = ⎢ ⎣ 0 4⎥⎦ are λ = 2 and λ = 4. This yields the equations y = x 2 2 2 ⎛ x⎞ ⎛ y⎞ x 2 + 25 y 2 = 25. so that the problem is to and x = − y 2 . 1). The first partial derivatives of f are f x ( x. λ = 2 ± 2 5. y) f xy ( x.5 Exercise Set 7.5 4 5 − i⎤ ⎡ −2i 1. (2. Thus there are two critical points: (0. or ⎜ ⎟ + ⎜ ⎟ = 1. 1). 1) = ⎢ are ⎣ −4 4 ⎥⎦ (d) False. 1). The largest eigenvalue of A is λ = 10 which has ⎡ 1 ⎤ positive unit eigenvector ⎢ 12 ⎥ . The Hessian matrix is ⎡ f xx ( x. y) ⎤ H ( x. Similarly. then q(x) = xT Ax = xT (λx) = λ(xT x) = λ(1) = λ. Thus the ⎢ ⎥ ⎣ 2⎦ ⎛ 1 ⎞⎛ 1 ⎞ maximum value of z = 20 ⎜ ⎟⎜ ⎟ = 10 ⎝ 2 ⎠⎝ 2 ⎠ 5 which occurs when x = 5 x1 = and 2 1 y = y1 = . From this we conclude that y = y 4 and so y = 0 or y = 1. ⎡ 0 −4 ⎤ The eigenvalues of H (2. 4 ⎥⎦ ⎣ −2 x (b) True (c) True ⎡2 0⎤ The eigenvalues of the matrix H (0. −10 λ ⎡ 0 −3⎤ are The eigenvalues of H (0. y) f yy ( x. this matrix is negative definite and so f has a relative maximum at (−1.4 (a) False. 0) and (−1. To find the critical points we set f x and f y equal to zero. This yields the equations 2x(1 − y) = 0 1 and y = x 2 . 0). 1). ⎡ 0 10⎤ ⎡ x1 ⎤ Write z = xT Ax = [ x1 y1 ] ⎢ ⎢ ⎥. If x is a unit eigenvector corresponding to λ. The corresponding values of x are x = 0 and x = −1 respectively. this matrix is indefinite and so f has a saddle point at (0. y) f xy ( x. From the first. 0). y) = ⎢ ⎥=⎢ ⎥. then A will have negative eigenvalues. which are the corner points of the 2 rectangle. 15. True/False 7. and (−2. the Hessian form of the second derivative test is inconclusive. The problem is to maximize z = 4xy subject to equal to zero. ⎣ f yx ( x. 0) = ⎢ ⎣ −3 0 ⎥⎦ λ = ±3. y) = 2 x − 2 xy and f y ( x. One of these is positive and one is negative. y) ⎦ 21. if det(A) < 0. y) f yy ( x. 1) = ⎢ ⎣ −3 −6 ⎥⎦ λ = −9. y ) = ⎢ ⎥ ⎣ f yx ( x. y1 ) = 1. y) = 4 y − x 2 .SSM: Elementary Linear Algebra Section 7. 1). y) ⎤ ⎡6 x −3 ⎤ H ( x. we conclude that 4 x = 0 or y = 1. then it will have neither a maximum nor a minimum value. thus this matrix is indefinite and f has a saddle point at (2. ⎡ 2 − 2 y −2 x ⎤ =⎢ . Section 7. y) ⎦ ⎣ −3 −6 y ⎦ maximize z = 20 x1 y1 subject to ( x1 . ⎣10 0⎥⎦ ⎣ y1 ⎦ λ −10 = λ 2 − 100 = (λ + 10)(λ − 10). The Hessian matrix is ⎡ f xx ( x. (e) True. 1). the eigenvalues ⎡ 0 4⎤ of H (−2. if the only critical point of the quadratic form is a saddle point. 1) = ⎢ are λ = 2 ± 2 5. 0).

Thus −2 0 ⎥⎦ ⎣0 0 0 ⎥⎦ ⎣ −1 − i ⎡ −1+i ⎤ ⎡ −1 + i ⎤ ⎢ 3 ⎥ v1 = ⎢ is a basis for the eigenspace corresponding to λ = 3. Thus A is unitary. which reduces to ⎢ . The following computations show that the row vectors of A are orthonormal: 2 r1 = 3 4 + i 5 5 r2 = − 2 2 4 3 + i 5 5 9 16 + =1 25 25 = 2 16 9 + =1 25 25 = ⎛ 3 ⎞ ⎛ 4 ⎞ ⎛ 4 ⎞⎛ 3 ⎞ r1 ⋅ r2 = ⎜ ⎟ ⎜ − ⎟ + ⎜ i ⎟⎜ − i ⎟ ⎝ 5 ⎠ ⎝ 5 ⎠ ⎝ 5 ⎠⎝ 5 ⎠ 12 12 =− + 5 5 =0 ⎡ 3 − 54 ⎤ ⎥. and p = 1 ⎢ 1 ⎥ is a unit eigenvector. ( −i − 3 ) ⎥⎦ 13. The following computations show that the column vectors of A are orthonormal: c1 = 1 c2 = 1 2 2 4 4 = + 8 8 =1 = =1 c1 ⋅ c2 = ( 2 2 4 4 + 8 8 1 2 2 2 3 + i) + (1 − i 3 ) ( 3 + i) 1 2 2 2 + 1 2 2 (1 + i 3 ) 1 2 2 (i − (1 + i 3 ) + 3) 2 2 1 2 2 (1 + i 3 ) ⎡ 1 ( 3 − i) Thus A is unitary. A ⎣ 1 ⎥⎦ ⎣⎢ 3 ⎦⎥ 180 . and A−1 = A* = ⎢ 21 2 ⎢ (1 + i 3 ) ⎣2 2 1 2 2 1 2 2 1 2 2 ( −i − 3) = 0 (1 − i 3 ) ⎤ ⎥. 9. and A−1 = A* = ⎢ 5 ⎢ − 54 i − 53 i ⎥ ⎣ ⎦ 11.Chapter 7: Diagonalization and Quadratic Forms SSM: Elementary Linear Algebra i 2 − 3i ⎤ ⎡ 1 3. A = ⎢ −i −3 1 ⎥ ⎢ ⎥ 2 ⎦ ⎣ 2 + 3i 1 5. (b) A is not Hermitian since a22 is not real. ⎡ −1 −1 + i 0 ⎤ ⎡1 1 − i 0 ⎤ The augmented matrix of the system (3I − A)x = 0 is ⎢ . The characteristic polynomial of A is λ 2 − 9λ + 18 = (λ − 3)(λ − 6). (a) A is not Hermitian since a31 ≠ a13 . thus the eigenvalues of A are λ = 3 and λ = 6.

The augmented matrix of the system (−2I − A)x = 0 is ⎢ 0 −1 1 − i 0 ⎥ . −2 − 2i 0 ⎤ ⎡ −4 ⎡2 1 + i 0⎤ The augmented matrix of the system (2I − A)x = 0 is ⎢ .5 ⎡ 1−i ⎤ similar computation shows that p 2 = ⎢ 6 ⎥ is a unit eigenvector corresponding to λ = 6. The vectors p1 and p 2 ⎢1 ⎥ ⎢⎣ 3 ⎥⎦ are orthogonal. and the unitary matrix P = [p1 p 2 ] diagonalizes the matrix A: ⎡ −1+i 6 ⎢ P*AP = ⎢ 1−i ⎣⎢ 3 ⎡ 2 0⎤ =⎢ ⎣ 0 8 ⎥⎦ 2 ⎤ 6⎥⎡ 6 1 ⎥ ⎢⎣ 2 − 2i 3 ⎦⎥ ⎡ −1−i 2 + 2i ⎤ ⎢ 6 4 ⎥⎦ ⎢ 2 ⎣⎢ 6 1+i ⎤ 3⎥ 1 ⎥ 3 ⎦⎥ 17. The characteristic polynomial of A is λ 2 − 10λ + 16 = (λ − 2)(λ − 8). The vectors {p1. The characteristic polynomial of A is (λ − 5)(λ 2 + λ − 2) = (λ + 2)(λ − 1)(λ − 5). which reduces to ⎢ . and p1 = ⎢ 3 ⎥ is a unit eigenvector. λ 2 = 1. ⎢ ⎥ ⎣ 0 1 + i −2 0 ⎦ 0 0⎤ ⎡1 0 ⎡ 0 ⎤ which can be reduced to ⎢0 1 −1 + i 0 ⎥ . and p3 = ⎢0 ⎥ is a unit eigenvector corresponding to λ3 = 5. and the unitary matrix P = [p1 p 2 p3 ] diagonalizes the matrix A: 181 . p 2 . and p1 = ⎢ 6 ⎥ is a unit eigenvector. The vectors p1 and p 2 ⎢ 2 ⎥ ⎣⎢ 6 ⎦⎥ are orthogonal. ⎥ −2 0⎦ ⎣ 0 0 0 ⎥⎦ ⎣ −2 + 2i ⎡ −1−i ⎤ ⎡ −1 − i ⎤ Thus v1 = ⎢ is a basis for the eigenspace corresponding to λ = 2. A ⎢ 2 ⎥ ⎣ 2 ⎥⎦ ⎣⎢ 6 ⎦⎥ ⎡ 1+i ⎤ similar computation shows that p 2 = ⎢ 3 ⎥ is a unit eigenvector corresponding to λ = 8. and the unitary matrix P = [p1 p 2 ] diagonalizes the matrix A: ⎡ −1−i P*AP = ⎢ 3 ⎢ 1+i ⎢⎣ 6 ⎡3 0⎤ =⎢ ⎣ 0 6 ⎥⎦ 1 ⎤ 3⎥⎡ 4 2 ⎥ ⎢⎣1 + i 6 ⎥⎦ ⎡ −1+i 1− i⎤ ⎢ 3 5 ⎥⎦ ⎢ 1 ⎣⎢ 3 1−i ⎤ 6⎥ 2 ⎥ 6 ⎦⎥ 15. Thus v1 = ⎢1 − i ⎥ is a basis for the eigenspace corresponding to ⎢ ⎥ ⎢ ⎥ 0 0⎦ ⎣ 1 ⎦ ⎣0 0 ⎡0⎤ ⎡ 0 ⎤ ⎢ 1−i ⎥ ⎢ −1+i ⎥ λ1 = −2. and λ3 = 5. thus the eigenvalues of A are λ = 2 and λ = 8. p3} ⎢ ⎥ ⎣0 ⎦ form an orthogonal set.SSM: Elementary Linear Algebra Section 7. thus the eigenvalues of A are 0 0⎤ ⎡ −7 0 λ1 = −2. Similar computations show that p 2 = ⎢ 6 ⎥ is a unit eigenvector ⎢ 1 ⎥ ⎢ 2 ⎥ ⎣⎢ 3 ⎦⎥ ⎣⎢ 6 ⎦⎥ ⎡1 ⎤ corresponding to λ 2 = 1.

A = ⎢ i 0 1 ⎥ ⎢ ⎥ 4i ⎦ ⎣ −2 − 3i −1 21. Thus A is normal if and only if B and C commute i. We have A*A = 1 ⎡ e−iθ ⎢ 2 ⎣⎢ eiθ −ie−iθ ⎤ 1 ⎡ eiθ ⎥ ⎢ ieiθ ⎦⎥ 2 ⎢⎣ieiθ e −iθ ⎤ 1 ⎡ 1 + 1 ⎥= ⎢ −ie−iθ ⎦⎥ 2 ⎢⎣ e2iθ − e2iθ e−2iθ − e −2iθ ⎤ ⎡1 0 ⎤ −1 ⎥=⎢ ⎥⎦ . (a) If B = ( A + A*). This is just one possibility. Thus if A is unitary and Ax = λx where x ≠ 0. A unitary matrix A has the property that Ax = x for all x in C n . and a32 ≠ −a23 ⎡ 0 −1 + i ⎤ 23. 2 2 2 2 (c) AA* = ( B + iC )( B − iC ) = B 2 − iBC + iCB + C 2 and A*A = B 2 + iBC − iCB + C 2 . If A is unitary. It depends on the particular choice of the unit eigenvectors. Thus AA* = A*A if and only if −iBC + iCB = iBC − iCB. i 2 − 3i ⎤ ⎡ 0 19. (a) a21 = −i ≠ i = −a12 and a31 ≠ −a13 (b) a11 ≠ −a11 . thus A* = A 0 1 ⎣ 1+1 ⎦⎥ and A is unitary. C* = C. thus the eigenvalues of A are ⎥ i i + 1 ⎣ ⎦ λ = 2i and λ = −i. then B* = ( A + A*)* 2 2 1 = ( A * + A **) 2 1 = ( A * + A) 2 = B. we must have λ x = Ax = x and so λ = 1. 182 . thus A* is also unitary. 27. (b) We have B + iC = 1 1 1 1 ( A + A*) + ( A − A*) = A and B − iC = ( A + A*) − ( A − A*) = A * .Chapter 7: Diagonalization and Quadratic Forms SSM: Elementary Linear Algebra 1 ⎤ 0 1⎤ ⎡ 0 1+i ⎡0 0 0 ⎤ ⎢ 1−i −1+i 3 3 ⎥ ⎡5 ⎢ 0⎥⎥ P*AP = ⎢ 0 −1−i 2 ⎥ ⎢0 −1 −1 + i ⎥ ⎢ 3 6 ⎥ 6 6 ⎥ ⎢ 0 −1 − i ⎢ ⎥ 2 0 ⎦⎢ 1 0⎥ ⎣ 0 0 ⎦⎥ 6 ⎣⎢ 3 ⎦ ⎣⎢1 ⎡ −2 0 0 ⎤ = ⎢ 0 1 0⎥ ⎢ ⎥ ⎣ 0 0 5⎦ Note: The matrix P is not unique. 1 1 29. Similarly. or 2iCB = 2iBC. 33. a31 ≠ −a13 . then A−1 = A * and so ( A*) −1 = ( A−1 )* = ( A*)*.. 31.e. CB = BC. The characteristic polynomial of A = ⎢ is λ 2 − iλ + 2 = (λ − 2i)(λ + i).

5 5. so A−1 = A*. 4 A =A =⎢ 0 5 5 ⎢ 3 ⎥ 12 16 ⎢⎣ − 5 − 25 25 ⎥⎦ 39. then Px = (uu*)x = (uuT )x = u(uT x) = (x ⋅ u)u. 7. normal matrices that are not Hermitian are also unitarily diagonalizable. ⎡ 37. HH * = ( I − 2uu*)( I − 2uu*) = I − 2uu * −2uu * +4uu * uu * Chapter 7 Supplementary Exercises ⎡3 1. ⎢ ⎥ ⎢ 1 ⎥ ⎣0 ⎦ ⎢⎣ 2 ⎥⎦ 1 ⎡− 1 0⎤ 2 ⎢ 2 ⎥ Thus P = ⎢ 0 0 1⎥ orthogonally ⎢ 1 ⎥ 1 0⎥ 2 ⎣⎢ 2 ⎦ ⎡0 0 0 ⎤ diagonalizes A. If P = uu*. so ⎢ ⎥ ⎣0 0 1⎦ ⎡ 4 − 9 12 ⎤ 25 25 ⎥ ⎢ 5 −1 T 3⎥. If H = I − 2uu*. The characteristic equation of A is λ3 − 3λ 2 + 2λ = λ(λ − 2)(λ − 1). λ = 1: ⎢1 ⎥ . (c) True. then H* = (I − 2uu*)* = I* − 2u**u*= I − 2uu* = H. Orthogonal bases for ⎡− 1 ⎤ ⎢ 2⎥ the eigenspaces are λ = 0: ⎢ 0 ⎥ . The xT Ax = [ x1 x2 ] ⎢ ⎢ ⎥ 4 ⎥⎦ ⎣ x2 ⎦ ⎢⎣ − 32 (d) False. so the eigenvalues are λ = 0. ⎢ ⎥ ⎣0 0 1 ⎦ (a) False. If u = 1. (b) For A = ⎢ − 25 25 ⎢ 12 53 ⎥ 16 ⎢⎣ 25 5 25 ⎥⎦ ⎡1 0 0 ⎤ AT A = ⎢ 0 1 0 ⎥ . 183 7 =0 4 . In matrix form. then ( A*) −1 = A = ( A*) *.SSM: Elementary Linear Algebra Chapter 7 Supplementary Exercises 35. then ( A2 )* = ( A*)( A*) = (− A)(− A) = A2 ≠ − A2 . and PT AP = ⎢0 2 0 ⎥ . so AT A = ⎢ ⎣ 0 1 ⎥⎦ i 2 ( 0) + i ⎤ 3⎦ i 6 and i ⎛ i ⎞ i ⎛ i ⎞ ⎜− ⎟+ ⎜ ⎟ 6⎝ 3⎝ 3⎠ 6⎠ 2 2 ⎛ i ⎞ ⎛ i ⎞ = 0+⎜ ⎟ −⎜ ⎟ ⎝ 6⎠ ⎝ 3⎠ 1 1 =− + 6 3 1 = 6 Thus the row vectors do not form an orthonormal set and the matrix is not unitary. A = ⎢ ⎢ ⎣ 1 2 i 2 − i ⎤ 2⎥ is both Hermitian and unitary. 6 3⎦ ⎣ r1 ⋅ r2 = − ⎡1 0 ⎤ . λ = 2: ⎢ 1 ⎥ ⎢⎣ 2 ⎥⎦ ⎡ 1 ⎤ ⎡0 ⎤ ⎢ 2⎥ ⎢ 0 ⎥ . i ≠ i . (a) For A = ⎢ 5 ⎢ 45 ⎣ − 45 ⎤ ⎥. if A is skew-Hermitian. thus H is Hermitian. True/False 7. 2. 3⎥ 5⎦ ⎡ 3 −1 T A =A =⎢ 5 ⎢ − 45 ⎣ 2 = I − 4uu * +4u u u * =I and so H is unitary. for r1 = ⎡ − i ⎣ 2 i ⎤ r2 = ⎡0 − i . 3⎥ 5⎦ ⎡ 4 0 − 3⎤ 5⎥ ⎢ 5 9 4 − 12 ⎥ . 1 − ⎥ 2⎦ 4⎤ 5⎥. then multiplication of x by H = I − 2uu* corresponds to reflection of x about the hyperplane u ⊥ . 1. (b) False. characteristic equation of A is λ 2 − 5λ + (e) False. the quadratic form is ⎡ 1 − 3⎤ ⎡x ⎤ 2 ⎥ 1 . if A is unitary. Thus 2 multiplication of x by P corresponds to u times the orthogonal projection of x onto W = span{u}.

Chapter 7: Diagonalization and Quadratic Forms SSM: Elementary Linear Algebra 5±3 2 or 2 λ ≈ 4. the quadratic form is positive definite. Since both eigenvalues of A are positive. which has solutions λ = 9.62. 0. (b) 3x − 11y 2 = 0 or x = 11 2 y represents a 3 parabola. (a) y − x 2 = 0 or y = x 2 represents a parabola.38. 184 .

x2 ) = c1 (1. 5. x1 − 3x2 ) T(5.1 Exercise Set 8. 3. T(kA) = (kA)B = k(AB) = kT(A) T ( A1 + A2 ) = ( A1 + A) B = A1 B + A2 B = T ( A1 ) + T ( A2 ) Thus T is a linear transformation. 0) = (c1 + c2 . F (kA) = (kA)T = kAT = kF ( A) F ( A + B) = ( A + B)T = AT + BT = F ( A) + F ( B) Thus F is a linear transformation. For x = ( x1. −3) = (−20 − 15. − 2) + ( x1 − x2 )(−4. 5 + 9) = (−35. T (−u) = −u = u = T (u) ≠ −T (u). 14). we have ( x1 . so the function is not a linear transformation.Chapter 8 Linear Transformations Section 8. (a) T (kp( x)) = ka0 + ka1 ( x + 1) + ka2 ( x + 1)2 = kT ( p( x)) T ( p( x) + q( x)) = a0 + b0 + (a1 + b1 )( x + 1) + (a2 + b2 )( x + 1)2 = a0 + a1 ( x + 1) + a2 ( x + 1)2 + b0 + b1 ( x + 1) + b2 ( x + 1)2 = T ( p( x)) + T (q( x)) Thus T is a linear transformation. c1 ) or c1 + c2 = x1 c1 = x2 which has the solution c1 = x2 . 185 . c2 = x1 − x2 . 1) + ( x1 − x2 )(1. (b) T (kp( x)) = T (ka0 + ka1 x + ka2 x 2 ) = (ka0 + 1) + (ka1 + 1) x + (ka2 + 1) x 2 ≠ kT ( p( x)) T is not a linear transformation. Let p( x) = a0 + a1 x + a2 x 2 and q( x) = b0 + b1 x + b2 x 2 . 1) + c2 (1. ( x1 . x2 ) = c1v1 + c2 v 2 .1 1. 1) = (−4 x1 + 5 x2 . x2 ) = x2T ( v1 ) + ( x1 − x2 )T ( v 2 ) = x2 (1. 9. x2 ) = x2 (1. 7. 0) = x2 v1 + ( x1 − x2 ) v 2 and T ( x1.

4) + ( x2 − x3 )(3. 0. 0 + 0 + 0 − 4. (c) R(T) is the set of all polynomials in P3 with (b) T(3. ⎢ 5 ⎥ and ⎢ 6⎥ ⎢ ⎥ ⎢ ⎥ ⎣7 ⎦ ⎣ 4⎦ form a basis for the column space of A. 10) = (10 − 10. −4) is a basis for this space. 2. 2) = (6 − 2. x3 ) = x3T ( v1 ) + ( x2 − x3 )T ( v 2 ) + ( x1 − x2 )T ( v3 ) = x3 (2. −24 + 8) = (4. 1. 21. − 1) = (−2 + 16 + 1. 6) + (−12. −4) so (1. x2 . 186 . 4. A basis for R(T) is a ⎢ ⎥ ⎣ 6 0 −9 9 ⎦ basis for the column space of A. c1 + c2 . 0) + c3 (1. − 1) The columns of A corresponding to the columns of B containing leading 1s are ⎡ 4 ⎤ ⎡ 1⎤ ⎡ −3⎤ ⎢ 2 ⎥ . x3 ) = c1 v1 + c2 v 2 + c3 v3 . 18 − 18) = (0. and (−3. 3⎤ ⎡ 1 −1 ⎢ 19 23. x2 . −4. T (2 v1 − 3v 2 + 4 v3 ) = 2T ( v1 ) − 3T ( v 2 ) + 4T ( v3 ) = (2. 14 ⎤ ⎡1 0 11 ⎥ ⎢ (b) A reduces to ⎢0 1 − 19 ⎥ . and ⎢ −4 ⎥ . −8. 9. 2. c2 = x2 − x3 . − 4. (a) T(3. 9) form a basis for R(T). 19. − 9) so (0. 0) is in ker(T). − 9. − 2. x3 ) = x3 v1 + ( x2 − x3 ) v 2 + ( x1 − x2 ) v3 (b) T can be expressed as a matrix operator from R 4 to R3 with the matrix ⎡ 4 1 −2 −3⎤ A = ⎢ 2 1 1 −4 ⎥ . 2. c1 + c2 + c3 = x1 or c1 + c2 = x2 which has the solution c1 = x3 c1 = x3 . 2) is not in ker(T). x3 ) = c1 (1. − 9) = (−6. 1) = (0 + 0 + 0 − 3. c1 ) (b). 1. The vector (1. 1) + c2 (1. (c) Neither 1 + x nor 3 − x 2 can be expressed as xp(x) with p(x) in P2 . (a) Since x + x 2 = x(1 + x). {x. (b) T(0. (4. ⎡1 ⎤ ⎡ −1⎤ thus columns 1 and 2 of A. x3} is a basis for R(T). 6) (1. constant term 0. 0) is not in ker(T). Thus. −4. 2 − 3) = (15. x1 + 3 x3 ) T (2. 1. 5. c3 = x1 − ( x2 − x3 ) − x3 = x1 − x2 . 0) = (−4 − 2. −4x) or the line y = −4x. −8. we have ( x1 . 6 − 8 + 2. Thus. 1) = (2 − 1. R(T) is the set of all vectors (x. 1.Chapter 8: Linear Transformations SSM: Elementary Linear Algebra 11. 9) so (0. (a) A row echelon form of A is 0 1 − 11 ⎥ . 11 11 x3 = s or x1 = −14t . x2 . 6). 0. x + x 2 is R(T). x2 . 1) is not in ker(T). (c) T(1. −16) so (3. 0) = (12 − 8 − 4. 10) is in ker(T). 0) = (c1 + c2 + c3 . 5 x1 − 5 x2 − x3 . x 2 . so a general 11 ⎢ ⎥ 0 0 0⎦ ⎣ 14 19 solution of Ax = 0 is x1 = − s. (a) Since −8x + 4y = −4(2x − y) and 2x − y can assume any real value. so neither are in R(T). A row ⎡ 1 1 1 −2 ⎤ 2 2 echelon form of A is B = ⎢0 1 4 −5⎥ . ( x1 . 0. x2 = 19t . 10 − 20 + 1. 0). 0) so (3. x2 = s. − 1. −40 + 40) = (0. 4) − (0. ⎢ 1⎥ . 0. 1) is not in ker(T). − 3. 0. 1) + ( x1 − x2 )(−1. 1) = (− x1 + 4 x2 − x3 . −8 + 4) = (1. so (c) T (0. 0. ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 6 ⎦ ⎣0 ⎦ ⎣ 9⎦ 13. x3 = 11t . 4. 0 − 0 + 9) = (−3. 1. 15. − 7. (a) T(5. 0. which is the range of T. 0) so (5. −4. ⎢ ⎥ ⎢⎣0 0 0 1⎥⎦ T ( x1. For x = ( x1. ⎢ ⎥ 0 ⎦⎥ ⎣⎢0 0 17. 8) = (−10. − 4 + 1.

(a) nullity(T) = 5 − rank(T) = 2 187 . (c) Since R(T ) = R3 . a plane through the origin. 27. (b) False. T must be the identity operator. so nullity (T) = 1. ⎢ ⎥ ⎣ 11⎦ which is ker(T). so nullity(T) = 5 − rank(T) = 4 ⎡ −14 ⎤ ⎢ 19 ⎥ is a basis for the null space of A. or all of R3 . every linear transformation will have T(−v) = −T(v). ker(T) is one-dimensional. the origin only. (a) True. a2 k 2 x 2 + b2 k 2 y 2 ) 4⎤ ⎡1 0 1 7 ⎥ . Let w = cv1 + c2 v 2 + " + cn v n be any vector in V. (a) nullity(A) = 7 − rank(A) = 3. (d) The column space of A is two-dimensional. (b) No. (a) The fourth derivative of any polynomial of (c) Both R(T) and ker(T) are two-dimensional. the range is the entire plane y = x. so rank(T) = 2.1 (b) dim( P4 ) = 5. the range is the entire xz-plane. If p′( x) = 0. so the solution space of Ax = 0 has dimension 3. c2 = 0 gives the homogeneity property and c1 = c2 = 1 gives the additivity property. True/False 8. y) 37. which is ker(T). so rank(T) = nullity(T) = 2. x3 = s. so T ( f ( x)) = f ( 4) ( x) has ker(T ) = P3 . (a) The kernel is the y-axis. = k 2 F ( x. F (kx. so nullity(A) = 1.1 (b) The kernel is the x-axis. (d) Both the column space and the null space of A are two-dimensional. since rank(A) = 4 while R5 has dimension 5. thus columns 1 and 2 of 7⎦ ⎣ ⎡4⎤ ⎡1 ⎤ A. 35. 43. nullity(T) = 6 − rank(T) = 3 (c) R(T) is two-dimensional. R(T) must be a subspace of R3 . y) ≠ kF ( x. 29. 7 2 x2 = − s + t . T has rank 3. the range is the entire yz-plane. Then T (w ) = c1T ( v1 ) + c2T ( v 2 ) + " + cnT ( v n ) = c1 v1 + c2 v 2 + " + cn v n =w Since w was an arbitrary vector in V. 33. ky) = (a1k 2 x 2 + b1k 2 y 2 . so ⎢ ⎥ and ⎢ ⎥ form a basis ⎢ 0⎥ ⎢ 1⎥ ⎢⎣ 7 ⎥⎦ ⎢⎣ 0⎥⎦ for the null space of A. which is the range of T. 41. degree 3 or less is 0. (b) By similar reasoning. ⎢ ⎥ and ⎢ ⎥ form a basis for the column ⎣1 ⎦ ⎣2⎦ space of A. 25. x2 = − s + 2r . so rank(A) = 2. x3 = s. thus the possibilities are a line through the origin. x4 = t or 7 x1 = − s − 4r . so rank(A) = nullity(A) = 2. (b) No. so a general (b) A reduces to ⎢ ⎢⎣0 1 1 − 72 ⎥⎦ 4 solution of Ax = 0 is x1 = − s − t . (a) A row echelon form of A is 0⎤ ⎡1 2 3 ⎢0 1 1 − 2 ⎥ . The null space of A is onedimensional. then p(x) is a constant. ⎡ −1⎤ ⎡ −4 ⎤ ⎢ −1⎥ ⎢ 2⎥ x4 = 7r . (c) The kernel is the line through the origin perpendicular to the plane y = x. T ( f ( x)) = f ( n +1) ( x) has ker(T ) = Pn . c1 = k . so ker(D) consists of all constant polynomials. (d) nullity(T) = 4 − rank(T) = 1 31.SSM: Elementary Linear Algebra Section 8.

then it would have nullity 0 (because it would be oneto-one). so ker(T) = {k(−1. ⎡a⎤ ⎛ ⎡a b ⎤ ⎞ ⎢ c ⎥ T2 ⎜ ⎢ ⎥⎟ = ⎢ ⎥. so multiplication by A is one-to-one. (a) Since nullity(T) = 0. Thus. (a) By inspection. then p(x) has constant term 0. z) is in ker(T) if both x + y + z = 0 and x − y − z = 0. 1)}. T is one-to-one. ker(T) = {k(0. (a) By inspection. ⎛ ⎡a b 11. (g) False. 1⎟ ⎬ and T is not one-to⎩ ⎝ 2 ⎠⎭ one. T does not necessarily have rank 4. T is not one-to-one. only the zero transformation has this property.2 (d) Since R(T ) = R n . (d) False. y) is in ker(T) only if x + y = 0 and x − y = 0. which is x = 0 and y + z = 0. (a) All points on the line y = −x are mapped to 0. 188 . y. (c) True. (a) T ⎜ ⎢ b d ⎜⎜ ⎢ ⎝ ⎣c e (d) By inspection. so T is not oneto-one. (i) False. and have rank no greater than the dimension of W (because its range is a subspace W). ⎡a⎤ T (ax3 + bx 2 + cx) = ⎢ b ⎥ ⎢ ⎥ ⎣c ⎦ (b) A can be viewed as a mapping from R 4 to R3 . Thus there is no such transformation. T (0) = v 0 + 0 = v 0 ≠ 0. rank(T) = n.2 (c) Since n < m. 1. −1)} and T is not one-to-one. y) = 0 if 2x + 3y = 0 or x = − y so 2 ⎧ ⎛ 3 ⎞⎫ ker(T ) = ⎨k ⎜ − . so T is one-toone. so nullity (T) = 0. thus multiplication by A is not one-toone. 1)} and T is not one-to-one. dim(V) = rank(T) + nullity(T) ≤ dim(W) which contradicts that dim(W) < dim(V). T is not one-to-one. (c) If p(x) is a polynomial and p(0) = 0. so T is not a linear transformation. ⎡ a⎤ ⎢ b⎥ c ⎤⎞ ⎢ ⎥ c e ⎥⎟ = ⎢ ⎥ ⎟ d ⎥ ⎢ ⎥ f ⎦ ⎟⎠ ⎢ ⎥ e ⎢f⎥ ⎣ ⎦ ⎡a ⎤ ⎛ ⎡a b ⎤ ⎞ ⎢ b ⎥ (b) T1 ⎜ ⎢ ⎥ ⎟ = ⎢ ⎥ and ⎝ ⎣c d ⎦ ⎠ ⎢ c ⎥ ⎣⎢ d ⎦⎥ (f) (x. ker(T) = {0}. so ⎢ ⎥ ⎣0 0⎦ nullity(A) = 1 and multiplication by A is not one-to-one. 1. Thus T is one-to-one. If there were such a transformation T. so T is one-toone. 9. (b) nullity(T) = n − (n − 1) = 1. so nullity(A) = 0 or ⎢ ⎥ ⎣0 0 ⎦ Ax = 0 has only the trivial solution. 7. (e) True 5.Chapter 8: Linear Transformations SSM: Elementary Linear Algebra ⎡ 1 0⎤ (c) A reduces to ⎢0 1⎥ . det(A + B) ≠ det(A) + det(B) in general. (h) False. 3 (b) T(x. nullity(T) = rank(T) = 2 Section 8. Then by the dimension Theorem. ker(T) = {0}. (e) (x. (f) True (b) Since ker(T) ≠ {0}. (c) (x. A reduces to ⎢0 0⎥ . so ker(T) = {k(1. so ker(T) = {0} and T is one-toone. ⎝ ⎣c d ⎦ ⎠ ⎢b ⎥ ⎣⎢ d ⎦⎥ ⎡ 1 −2 ⎤ 3. Exercise Set 8. y) is in ker(T) if x − y = 0 or x = y.

SSM: Elementary Linear Algebra Section 8. so ker(T) = {0}. − 3 y. y. since T (u). if T(a. so ker(T) ≠ {0} and T is not one-to-one. 9. 3x − 9 y) Let f = f ( x) = x 2 ( x − 1)2 . ( x − y) + y + x) = (0. No. if V consists of all matrices of ⎡a b 0⎤ the form ⎢ .3 Exercise Set 8. (c) False. for instance.3 ⎡a⎤ (d) T (a + b sin( x) + c cos( x)) = ⎢ b ⎥ ⎢ ⎥ ⎣c ⎦ Section 8. V → R4 . then f ′( x) = 2 x( x − 1)(2 x − 1) so f(0) = 0. z) = (x. 2 x) 15. y. y.3 1. 0) = (x. y. 0) = (4( x − 3 y) − 5(0). b. T ( v) = u. − 3 y + x + y ) = ( 2 x + 3 y. 3( x − 3 y) − 6(0)) = (4 x − 12 y. there can be no such transformation. x) = (0. T is not one-to-one because ker(T) ≠ {0} as T(a) = a × a = 0. if there were such a transformation T. y) = T2 ( x − y. By inspection.2 1 v. if ker(T) = {0} is one-to-one rank(T) = 4 so T is one-to-one and onto. then ⎣ c d 0 ⎥⎦ ⎡a⎤ ⎛ ⎡ a b 0⎤ ⎞ ⎢ b ⎥ T ⎜⎢ ⎥ ⎟ = ⎢ ⎥ is an isomorphism T: ⎝ ⎣ c d 0⎦ ⎠ ⎢ c ⎥ ⎣⎢ d ⎦⎥ 11. f ′(0) = 0. T(x. y. x − 2 y ) (d) (T2 D T1 )( x. (b) (T2 D T1 )( A) does not exist because T1 ( A) is not a 2 × 2 matrix. y) = T2 ( x − 3 y. dim( R ) = 2 while dim( P2 ) = 3. then 4 ⎛1 ⎞ ⎛1 ⎞ (T1 D T2 )( v) = T1 ⎜ v ⎟ = 4 ⎜ v ⎟ = v and ⎝4 ⎠ ⎝4 ⎠ 1 (T2 D T1 )( v) = T2 (4 v) = (4 v) = v. (a) (T2 D T1 )( x. T (f ) = f (0) + 2 f ′(0) + 3 f ′(1) = 0. T2 ( v) = 2 (a) False. (f) False. 0). y. (e) False. 2 x + 3 y ) 13. Yes. T(I) = P − P = 0 so ker(T) ≠ 0. (a) (T1 D T2 )( A) = T1 ( AT ) = tr ⎜ ⎢ ⎥⎟ = a + d ⎝ ⎣b d ⎦ ⎠ 19. z)) = T(x. then a = b = c = 0. y) = T2 (2 x. 0) = T(x. dim( M 33 ) = 9 while dim( P9 ) = 10. and f ′(1) = 0. (b) (T2 D T1 )( x. ⎛ ⎡a c ⎤ ⎞ 3. but since dim(ker(T )) + dim( R(T )) = dim( P4 ) = 5. T does not have an inverse. x + y) = ( 2 x + 3 y. v . y) = T2 (2 x. (b) True. u = u = 1 and T(u) is orthogonal to T(v) if u is orthogonal to v. so T (u) = T (u). then dim(ker(T)) = dim(R(T)). 189 . Then T(T(x. (a) Since det(A) = 0. 4 5. (d) True. T (u) = u. T (kf ) = kf (0) + 2kf ′(0) + 3kf ′(1) = kT (f ) T (f + g) = f (0) + g (0) + 2( f ′(0) + g ′(0)) + 3( f ′(1) + g ′(1)) = T (f ) + T ( g ) Thus T is a linear transformation. (c) (T2 D T1 )( x. 17. Yes. c) = 0. y. True/False 8. z) or T DT = T. 3 y) = ( 2 x − 3 y.

13. y) = T1 (− x. P(x) ⎡ 1 x1 − 1 x2 + 1 x3 ⎤ 2 2 ⎢ 2 ⎥ 1 1 = ⎢ − 2 x1 + 2 x2 + 12 x3 ⎥ . T1−1 ( p( x)) = T2−1 ( p( x)) = p( x − 1). (a) T(1 − 2x) = (1 − 2(0).. a0 + a1 ) so if T(p(x)) = (0. 3) has a0 = 2 and a0 + a1 = 3 or a1 = 1. ⎡ 3 3 −1⎤ A−1 = ⎢ −2 −2 1⎥ so ⎢ ⎥ ⎣ −4 −5 2 ⎦ 21. 2. (a) T1 ( x. . (a) Since T1 ( p( x)) = xp( x)... so T has an inverse. 0) = (0. xn ⎟ = ⎜ x1 . n must be nonzero. − y) T1 D T2 = T2 D T1 ⎛ ⎡ x1 ⎤ ⎞ ⎡ x1 ⎤ T −1 ⎜ ⎢ x2 ⎥ ⎟ = A−1 ⎢ x2 ⎥ ⎜⎢ ⎥⎟ ⎢ ⎥ ⎜ x ⎟ ⎣ x3 ⎦ ⎝⎣ 3⎦⎠ ⎡ 3x1 + 3x2 − x3 ⎤ = ⎢ −2 x1 − 2 x2 + x3 ⎥ . 1 − 2(1)) = (1. i = 1. x Since T2 ( p( x)) = p( x + 1). ⎢ 1 ⎥ 1 1 ⎢⎣ 2 x1 + 2 x2 − 2 x3 ⎥⎦ x (d) det(A) = 1 ≠ 0. y) = (− x. ⎛ ⎡ x1 ⎤ ⎞ ⎡ x1 ⎤ −1 ⎜ ⎢ ⎥ ⎟ −1 ⎢ ⎥ x x =A T ⎜⎢ 2⎥⎟ ⎢ 2⎥ ⎜ x ⎟ 3 ⎣ x3 ⎦ ⎝⎣ ⎦⎠ 1 ⎡ x1 + 1 x2 − 3 x3 ⎤ 8 4 ⎢ 8 ⎥ 1 1 = ⎢ 8 x1 + 8 x2 + 14 x3 ⎥ . y) = ( x.. 3) = 2 + x. 1): (T2 D T1 )(0. 1) = T1 (− sin θ . so T has an inverse. a0 + a1 ). − y) and T2 ( x.. a a a 2 n ⎝ 1 ⎠ 190 .. cosθ ) = (− sin θ .Chapter 8: Linear Transformations SSM: Elementary Linear Algebra (b) det(A) = −8 ≠ 0. ⎢ 3 ⎥ 5 1 ⎢⎣ − 8 x1 + 8 x2 + 4 x3 ⎥⎦ (T2 D T1 ) −1 ( p( x)) = (T1−1 D T2−1 )( p( x)) = T1−1 ( p( x − 1)) 1 = p( x − 1) x 17. 0). 0) but (T1 D T2 )(0. (a) For T to have an inverse.. all the ai . −1) (c) Let p( x) = a0 + a1 x. (b) T −1 ( x1 . y) (T1 D T2 )( x. (c) det(A) = −2 ≠ 0. − y) (T2 D T1 )( x. xn ) 1 1 ⎛1 ⎞ x2 . then T −1 (2. 1 p( x).. − y) = (− x. 1⎤ ⎡ 1 −1 2 2 2⎥ ⎢ 1 1 ⎥ so A−1 = ⎢ − 12 2 2 ⎢ 1 1 − 1⎥ ⎢⎣ 2 2 2 ⎥⎦ ⎛ ⎡ x1 ⎤ ⎞ ⎡ x1 ⎤ −1 ⎜ ⎢ ⎥ ⎟ −1 ⎢ ⎥ T x x =A ⎜⎢ 2⎥⎟ ⎢ 2⎥ ⎜ x ⎟ ⎣ x3 ⎦ ⎝⎣ 3⎦⎠ (d) Since T ( p( x)) = (a0 . ⎡ 1 1 − 3⎤ 4⎥ ⎢ 8 8 1 ⎥ so A−1 = ⎢ 18 18 4 ⎢ 3 5 ⎥ 1 − 4 ⎦⎥ ⎣⎢ 8 8 15. y) = (− x. so ker(T) = {0}. then a0 = a1 = 0 and p is the zero polynomial. 0) Thus T1 D T2 ≠ T2 D T1 . T −1 (2. then T ( p( x)) = (a0 . so T has an inverse. ⎢ ⎥ ⎢⎣ −4 x1 − 5 x2 + 2 x3 ⎥⎦ (b) Consider (x. x2 . y) = T2 ( x. Thus. . 1) = T2 (0. y) = (0.. .

(d) True. (c) False. it must be one-to-one. (f) True. the linear operators in Exercise 21(b) are an example. B [ x]B = ⎢ ⎢0 ⎣⎢ 0 0 0 1 0 0 0 1 0 ⎡0⎤ 0⎤ ⎡c ⎤ 0 ⎥ ⎢ 0 ⎥ ⎢ c0 ⎥ ⎥ c =⎢ ⎥ 0 ⎥ ⎢ 1 ⎥ ⎢ c1 ⎥ c 0 ⎦⎥ ⎣ 2 ⎦ ⎣⎢ c2 ⎦⎥ ⎡0⎤ ⎢c ⎥ [T (x)]B′ = [c0 x + c1 x + c2 x ]B′ = ⎢ 0 ⎥ ⎢ c1 ⎥ ⎢⎣ c2 ⎥⎦ 2 3 3. ky. T −1 does not exist. kz) (T2 D T1 )( x. kz ) and T2 ( x. kx sin θ + ky cos θ . x sin θ + y cos θ . then (T2 D T1 )( v1 ) = T2 (0) = 0 and ker(T2 D T1 ) ≠ {0}. for T to have an inverse. (a) T (u1 ) = T (1) = x = v 2 T (u 2 ) = T ( x) = x 2 = v3 T (u 3 ) = T ( x 2 ) = x 3 = v 4 ⎡0 ⎢1 Thus the matrix for T relative to B and B′ is ⎢ ⎢0 ⎢⎣0 ⎡0 ⎢1 (b) [T ]B′. z ) (T1 D T2 )( x.3 (a) True (b) False. z) = (k ( x cosθ − y sin θ ). Section 8. kz) T1 D T2 = T2 D T True/False 8. (a) T(1) = 1 T(x) = x − 1 = −1 + x T ( x 2 ) = ( x − 1)2 = 1 − 2 x + x 2 ⎡ 1 −1 1⎤ Thus the matrix for T relative to B is ⎢0 1 −2 ⎥ . z) = ( x cosθ − y sin θ . k ( x sin θ + y sin θ ). z) = (kx cos θ − ky sin θ . y. (e) False. y.4 1. z) = (kx. a linear transformation need not have an inverse.4 Exercise Set 8. ⎢ ⎥ 1⎦ ⎣0 0 191 0⎤ 0⎥ ⎥. 0⎥ 1⎥⎦ . y.4 (c) T1 ( x. y.SSM: Elementary Linear Algebra Section 8. if T1 is not one-to-one then there is some nonzero vector v1 with T1 ( v1 ) = 0.

T (x) = a0 + a1 ( x − 1) + a2 ( x − 1) 2 = a0 − a1 + a2 + (a1 − 2a2 ) x + a2 x 2 . ⎢ ⎥ ⎣ 4⎦ (c) T (2 − 3 x + 4 x 2 ) = 2 − 3(2 x + 1) + 4(2 x + 1)2 = 2 − 6 x − 3 + 16 x 2 + 16 x + 4 = 3 + 10 x + 16 x 2 192 . (a) Since A is the matrix for T relative to B. (a) T(1) = 1 T(x) = 2x + 1 = 1 + 2x T ( x)2 = (2 x + 1)2 = 1 + 4 x + 4 x 2 ⎡1 1 1 ⎤ [T ]B = ⎢0 2 4 ⎥ ⎢ ⎥ ⎣0 0 4 ⎦ ⎡ 2⎤ (b) The matrix for 2 − 3 x + 4 x 2 is ⎢ −3⎥ . and [T ( v3 )]B . ⎣ 3⎦ ⎣ 8 ⎦ ⎣ −5 ⎦ Similarly. ⎡ 1⎤ ⎡3⎤ [T ( v1 )]B = ⎢ ⎥ and [T ( v 2 )]B = ⎢ ⎥ . A = [[T ( v1 ) B [T ( v 2 )]B ] . That is. That is. ⎡ a0 − a1 + a2 ⎤ so [T (x)]B = ⎢ a1 − 2a2 ⎥ . (a) T (u1 ) = T ⎜ ⎢ ⎥ ⎟ = ⎢ −1⎥ = 0 v1 − v 2 + v 2 3 ⎢ ⎥ 2 3 ⎣ ⎦ ⎝ ⎠ ⎣ 0⎦ ⎡6 ⎤ 4 ⎛ ⎡ −2 ⎤ ⎞ ⎢ ⎥ T (u 2 ) = T ⎜ ⎢ ⎥ ⎟ = 2 = 0 v1 + v 2 + v3 4 ⎢ ⎥ 3 ⎝⎣ ⎦⎠ 0 ⎣ ⎦ ⎡ 0 0⎤ ⎢ 1 ⎥ [T ]B′.Chapter 8: Linear Transformations SSM: Elementary Linear Algebra ⎡ 1 −1 1⎤ ⎡ a0 ⎤ (b) [T ]B [ x]B = ⎢ 0 1 −2 ⎥ ⎢ a1 ⎥ ⎢ ⎥⎢ ⎥ 1⎦ ⎣ a2 ⎦ ⎣0 0 ⎡ a0 − a1 + a2 ⎤ = ⎢ a1 − 2a2 ⎥ ⎢ ⎥ a2 ⎣ ⎦ 9. respectively. ⎣9 ⎦ ⎣ 20⎦ ⎣ 29⎦ 2 For x = a0 + a1 x + a2 x . [T ( v 2 )]B . [T ( v 2 )]B = ⎢ 0 ⎥ . x 3 ⎣ ⎦ ⎣ 4⎦ ⎣ 2⎦ x = c −c then 1 1 2 . the columns of A are [T ( v1 )]B . (a) Since A is the matrix for T relative to B. Thus T (2 − 3 x + 4 x 2 ) = 3 + 10 x + 16 x 2 . ⎣ −2 ⎦ ⎡ 1⎤ ⎡ −2⎤ ⎡ 3⎤ T ( v1 ) = 1v1 − 2 v 2 = ⎢ ⎥ − ⎢ ⎥ = ⎢ ⎥ . ⎡ 3⎤ ⎡ −5 ⎤ ⎡ −2 ⎤ T ( v 2 ) = 3v1 + 5v 2 = ⎢ ⎥ + ⎢ ⎥ = ⎢ ⎥ . ⎢ ⎥ ⎣ 4⎦ 18 ⎛ ⎡1⎤ ⎞ ⎡ (d) T ⎜ ⎢ ⎥ ⎟ = ⎢ 7 ⎝ ⎣1⎦ ⎠ ⎢⎣ − 107 7 ⎡ 2 ⎤ ⎡1 1 1 ⎤ ⎡ 2⎤ ⎡ 3⎤ [T ]B ⎢ −3⎥ = ⎢ 0 2 4 ⎥ ⎢ −3⎥ = ⎢10 ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 4 ⎦ ⎣ 0 0 4 ⎦ ⎣ 4⎦ ⎣16 ⎦ 1 ⎤ 7 ⎥ ⎡1⎤ 24 ⎥ ⎢⎣1⎥⎦ 7 ⎦ ⎡ 19 ⎤ =⎢ 7⎥ ⎢ − 83 ⎥ ⎣ 7⎦ 11. ⎢ ⎥ a2 ⎣ ⎦ ⎡x ⎤ ⎡ 1⎤ ⎡ −1⎤ (c) If ⎢ 1 ⎥ = c1v1 + c2 v 2 = c1 ⎢ ⎥ + c2 ⎢ ⎥ . ⎡ 1⎤ ⎡ 3⎤ [T ( v1 )]B = ⎢ 2 ⎥ . Solving for c1 and c2 x2 = 3c1 + 4c2 ⎡ 7⎤ 1 8 ⎛ ⎡1⎤ ⎞ 5. B = ⎢ − 2 1⎥ ⎢ 8 4⎥ ⎣ 3 3⎦ gives c1 = 4 1 3 1 x1 + x2 . and ⎢ ⎥ ⎢ ⎥ ⎣ −2 ⎦ ⎣6⎦ ⎡ −1⎤ [T ( v3 )]B = ⎢ 5⎥ . − 2 ⎣5⎦ ⎣ ⎦ ⎡ 1⎤ (b) Since [T ( v1 )]B = ⎢ ⎥ . c2 = − x1 + x2 . 7 7 7 7 so ⎛⎡ x ⎤⎞ T ⎜ ⎢ 1⎥ ⎟ ⎝ ⎣ x2 ⎦ ⎠ = c1T ( v1 ) + c2T ( v 2 ) 1 ⎞ ⎡ 3⎤ ⎛ 3 1 ⎞ ⎡ −2 ⎤ ⎛4 = ⎜ x1 + x2 ⎟ ⎢ ⎥ + ⎜ − x1 + x2 ⎟ ⎢ ⎥ − 5 7 ⎠⎣ ⎦ ⎝ 7 7 ⎠ ⎣ 29 ⎦ ⎝7 18 1 ⎡ x1 + 7 x2 ⎤ ⎥ =⎢ 7 24 x ⎥ ⎢ − 107 + x 1 2 7 ⎣ 7 ⎦ 1 ⎤ ⎡ 18 7 ⎥ ⎡ x1 ⎤ =⎢ 7 ⎢ ⎥ 107 24 ⎢− 7 ⎥ ⎣ x2 ⎦ 7 ⎦ ⎣ 7.

.4 ⎡ 1⎤ (b) Since [T ( v1 )]B = ⎢ 2 ⎥ . so 8 0 0 3 0 0⎤ 0⎥ ⎥. v n } of V to {kv1.. 0⎥ 3⎦⎥ (b) [T2 D T1 ]B′. 8 1 c3 = (a0 + a1 − a2 ). ⎢ ⎥ ⎣6⎦ (d) In 1 + x 2 . a2 = 3c1 + 2c2 + 2c3 T2 ( x 2 ) = 3x3 ⎡0 ⎢3 Thus [T2 ]B′. [T2 D T1 ]B′. B′′ = ⎢ ⎢0 ⎣⎢0 Solving for c1 . B ⎡0 ⎢3 (c) ⎢ ⎢0 ⎣⎢0 T (a0 + a. and T ( v3 ) = − v1 + 5v 2 + 4 v3 = −3 x − 3x 2 − 5 + 15 x + 10 x 2 + 12 T1 (1) = 2 T1 ( x) = −3x + 28 x + 8 x 2 2 = 7 + 40 x + 15 x . . a1 = 0. #⎥ ⎢# # # ⎢⎣ 0 0 0 " k ⎥⎦ 193 . B = ⎢ ⎥. (a) (T2 D T1 )(1) = T2 (2) = 6 x = 16 + 51x + 19 x 2 . 3 1 c2 = (−5a0 + 3a1 − 3a2 ). B′′ [T1 ]B′′. a2 = 1. B = ⎢ 0 −3⎥ . ⎢ 0 −9 ⎥ ⎣⎢0 0⎦⎥ = −6 − 5 x + 5 x 2 . . and c3 gives 1 c1 = (a0 − a1 + 2a2 ). ⎡2 0⎤ Thus [T1 ]B′′.SSM: Elementary Linear Algebra Section 8. T (1 + x 2 ) 239 + 289 201 + 247 61 + 107 2 = + x+ x 24 8 12 = 22 + 56 x + 14 x 2 T ( v1 ) = v1 + 2 v 2 + 6 v3 = 3 x + 3 x 2 − 2 + 6 x + 4 x 2 + 18 + 42 x + 12 x 2 13. a0 = −c2 + 3c3 Then a1 = 3c1 + 3c2 + 7c3 . a0 = 1. T ( v 2 ) = 3v1 − 2 v3 2 = 9 x + 9 x − 6 − 14 x − 4 x (T2 D T1 )( x) = T2 (−3x) = −9 x 2 2 ⎡0 0⎤ ⎢6 0⎥ Thus. ⎢ ⎥ ⎣0 0⎦ T2 (1) = 3x (c) If a0 + a1 x + a2 x 2 = c1 v1 + c2 v 2 + c3 v3 = c1 (3x + 3x 2 ) + c2 (−1 + 3x + 2 x 2 ) T2 ( x) = 3 x 2 + c3 (3 + 7 x + 2 x 2 )... Similarly. B = [T2 ]B′. x + a2 x 2 ) = c1T ( v1 ) + c2T ( v 2 ) + c3T ( v3 ) 1 = (a0 − a1 + 2a2 )(16 + 51x + 19 x 2 ) 3 1 + (−5a0 + 3a1 − 3a2 )(−6 − 5 x + 5 x 2 ) 8 1 + (a0 + a1 − a2 )(7 + 40 x + 15 x 2 ) 8 239a0 − 161a1 + 289a2 = 24 201a0 − 111a1 + 247 a2 + x 8 61a0 − 31a1 + 107a2 2 + x 12 0 0 3 0 0⎤ ⎡0 0 ⎤ ⎡2 0⎤ ⎢ ⎥ 0 ⎢ 6 0⎥ ⎥ 0 −3⎥ = ⎢ ⎥ 0⎥ ⎢ ⎥ ⎢ 0 −9 ⎥ 0 0 ⎦ ⎢0 0 ⎥ 3⎦⎥ ⎣ ⎣ ⎦ 15. kv n } where k is a nonzero constant.. kv 2 . If T is a contraction or dilation of V. c2 . then T maps any basis B = {v1 .. v 2 . Thus the matrix for T relative to B is ⎡k 0 0 " 0 ⎤ ⎢0 k 0 " 0⎥ ⎢ ⎥ ⎢0 0 k " 0⎥ .

(a) [T2 D T1 ]B′. since the matrix for T is invertible. ⎢ ⎥ ⎣ 0 0 2⎦ ⎡ 2 −3⎤ B′ to B is P = ⎢ . (a) D(f1 ) = D(1) = 0 (b) False. Also. v1 = 2u1 + u 2 and (c) D(f1 ) = D(e2 x ) = 2e2 x D(f2 ) = D( xe 2 x ) = e 2 x + 2 xe2 x v 2 = −3u1 + 4u 2 .e. i. the matrix of S D T relative to B is [ S ]B [T ]B . B 17. B = [T3 ]B′. The matrix for T relative to B is the matrix whose columns are the transforms of the basis vectors in B in terms of the standard basis.5 ⎡1 ⎤ ⎡ −2 ⎤ 1. D(4e2 x + 6 xe2 x − 10 x 2 e 2 x ) = 14e 2 x − 8 xe2 x − 20 x 2 e2 x . the conclusion would only be true if T: V → V were a linear operator. 19. B = [T2 ]B′. (b) D(f1 ) = D(1) = 0 (e) True D (f 2 ) = D ( e x ) = e x Section 8.. Since (b) [T3 D T2 D T1 ]B′. True/False 8. Thus ⎣ 1 4⎥⎦ ⎡ 4 3⎤ PB→ B′ = P −1 = ⎢ 11 11 ⎥ . 1 2⎥ ⎢⎣ − 11 11 ⎦ [T ]B′ = PB → B′ [T ]B PB′→ B ⎡ 4 3 ⎤ ⎡1 −2 ⎤ ⎡ 2 −3⎤ = ⎢ 11 11 ⎥ ⎢ 1 2 ⎥ ⎣ 0 −1⎥⎦ ⎢⎣ 1 4 ⎥⎦ ⎢⎣ − 11 11 ⎦ ⎡ − 3 − 56 ⎤ 11 ⎥ = ⎢ 11 3⎥ 2 ⎢⎣ − 11 11 ⎦ (d) D(4e2 x + 6 xe2 x − 10 x 2 e 2 x ) ⎡ 2 1 0⎤ ⎡ 4⎤ = ⎢ 0 2 2⎥ ⎢ 6⎥ ⎢ ⎥⎢ ⎥ ⎣ 0 0 2 ⎦ ⎣ −10 ⎦ ⎡ 14 ⎤ = ⎢ −8 ⎥ ⎢ ⎥ ⎣ −20 ⎦ Thus.. ⎣0 −1⎥⎦ By inspection.5 D(f3 ) = D(e2 x ) = 2e2 x The matrix for D relative to this basis is ⎡0 0 0 ⎤ ⎢0 1 0 ⎥ . T (u1 ) = ⎢ ⎥ and T (u 2 ) = ⎢ ⎥ . B B is the standard basis for R n . this matrix is the standard matrix for T. (d) False. since B′ is the standard basis for R m . the conclusion would only be true if T: V → V were a linear operator. ⎢ ⎥ ⎣0 1 0 ⎦ (c) True.4 (a) False. i. ker(T) = {0}. the resulting transformation will give vector components relative to the standard basis. D(f 2 ) = D(sin x) = cos x D(f3 ) = D(cos x) = − sin x The matrix for D relative to this basis is ⎡0 0 0 ⎤ ⎢0 0 −1⎥ . if V = W. B′′ [T1 ]B′′. so the transition matrix from D(f3 ) = D( x 2 e2 x ) = 2 xe2 x + 2 x 2 e2 x The matrix for D relative to this basis is ⎡2 1 0⎤ ⎢ 0 2 2⎥ .Chapter 8: Linear Transformations SSM: Elementary Linear Algebra 21. if V = W. B′′′ [T2 ]B′′′. ⎢ ⎥ ⎣0 0 2 ⎦ Exercise Set 8. B′′ [T1 ]B′′.e. 194 . so 0 ⎣ ⎦ ⎣ −1⎦ ⎡ 1 −2 ⎤ [T ]B = ⎢ .

⎢ ⎥ ⎣0 0 0 ⎦ 7⎤ 2⎥ and (b) The matrix for T relative to the standard ⎡ 4 −1⎤ . we have Then for P = ⎢ ⎣ −1 −2 ⎥⎦ ⎡ 2 1⎤ ⎡2 0⎤ and P −1[T ]B P = ⎢ P −1 = ⎢ . ⎢ ⎥ ⎣ 0 0 0⎦ By inspection. basis B is [T ]B = ⎢ ⎣ −3 1⎥⎦ 1 1⎤ 1 1⎥ ⎥ 0 1⎦ The eigenvalues of [T ]B are λ = 5 + 21 2 5 − 21 with corresponding 2 ⎡ −3− 21 ⎤ ⎡ −3+ 21 ⎤ eigenvectors ⎢ 6 ⎥ and ⎢ 6 ⎥ . so [T ]B′ = PB → B′ [T ]B PB′→ B ⎡ 4 3 ⎤ ⎡ 1 − 1 ⎤ ⎡ 2 −3⎤ 2⎥ = ⎢ 11 11 ⎥ ⎢ 2 1 2 ⎥⎢ 1 1 ⎥ ⎢⎣ 1 4 ⎥⎦ ⎢⎣ − 11 11 ⎦ ⎣ 2 2⎦ ⎡ 13 − 25 ⎤ 11 2 ⎥ = ⎢ 115 2 . ⎣0 1⎥⎦ 3. 9 ⎥ ⎢ 11 2 ⎦ ⎣ 11 2 1 ⎥⎦ 11. ⎢ ⎥ ⎣ 0 0 1⎦ ⎡ 1 −1 0 ⎤ Thus PB→ B′ = P −1 = ⎢ 0 1 −1⎥ ⎢ ⎥ ⎣ 0 0 1⎦ [T ]B′ = PB→ B′ [T ]B PB′→ B ⎡ 1 −1 0 ⎤ ⎡ 1 0 0⎤ ⎡ 1 = ⎢0 1 −1⎥ ⎢ 0 1 0⎥ ⎢ 0 ⎢ ⎥⎢ ⎥⎢ ⎣0 0 1⎦ ⎣ 0 0 0⎦ ⎣ 0 ⎡ 1 0 0⎤ = ⎢0 1 1⎥ . so [T ]B = ⎢ 3 4⎥ ⎢⎣ 12 3⎦ 2 1 7 1 q1 = − p1 + p 2 and q 2 = p1 − p 2 . and v3 = u1 + u 2 + u3 . then ⎧ ⎡ 1⎤ ⎡ 1⎤ ⎫ B ′ = ⎨ ⎢ ⎥ . ⎡ 1 1⎤ . so ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣0⎦ ⎣0 ⎦ ⎣0⎦ ⎡1 0 0⎤ [T ]B = ⎢ 0 1 0⎥ . Thus P = = P=⎢ 9 P B → B′ ⎢⎣ 13 − 61 ⎥⎦ ⎢⎣ 32 and [T ]B′ = PB→ B′ [T ]B PB′→ B 7⎤ ⎡ 3 7 ⎤ ⎡ 2 − 2 ⎤ ⎡− 2 9⎥ ⎢ 9 9⎥ = ⎢4 2⎥ ⎢3 4⎥ ⎢ 1 − 1⎥ ⎢⎣ 32 1⎥⎦ ⎢⎣ 12 3⎦ ⎣ 3 6⎦ ⎡1 1⎤ =⎢ . The 1 ⎥ ⎣ sin 45° cos 45° ⎥⎦ ⎢ 1 2⎦ ⎣ 2 matrices for PB→ B′ and PB′↔ B are the same as in Exercise 1.SSM: Elementary Linear Algebra Section 8. (a) The matrix for T relative to the standard ⎡ 1 −1⎤ . Since B is the standard basis for R 2 .5 7⎤ ⎡− 2 ⎡3 −1 ⎢ 4 9 ⎥ . so the transition matrix from ⎡ 1 1 1⎤ B′ to B is P = ⎢ 0 1 1⎥ . ⎡ 1 − 1 ⎤ ⎡cos 45° − sin 45° ⎤ ⎢ 2 2⎥ = [T ]B = ⎢ . so the 9 3 9 6 transition matrix from B′ to B is ⎡ −3− 21 −3+ 21 ⎤ 6 Then for P = ⎢ 6 ⎥ . The eigenvalues basis B is [T ]B = ⎢ ⎣ 2 4 ⎥⎦ of [T ]B are λ = 2 and λ = 3 with ⎡ 1⎤ corresponding eigenvectors ⎢ ⎥ and ⎣ −1⎦ ⎡ 1⎤ ⎢⎣ −2 ⎥⎦ . we have ⎢⎣ 1 1 ⎥⎦ −3+ 21 ⎤ ⎡− 3 −1 ⎢ 21 2 21 ⎥ P =⎢ and 3+ 21 ⎥ 3 ⎢⎣ 21 2 21 ⎥⎦ 195 . 9 3 ⎡2 − 2⎤ 9⎥. ⎢ ⎥ ⎬ is a basis for which ⎩ ⎣ −1⎦ ⎣ −2 ⎦ ⎭ [T ]B′ is diagonal. ⎥ − − 1 1 ⎣ ⎦ ⎣ 0 3⎥⎦ Since P represents the transition matrix from the basis B′ to the standard basis B. ⎡1 ⎤ ⎡0 ⎤ ⎡0⎤ 5. T (p1 ) = 6 + 3( x + 1) = 9 + 3x = p1 + p 2 . v1 = u1 . T (u1 ) = ⎢ 0 ⎥ . ⎣⎢ 1 ⎦⎥ ⎣⎢ 1 ⎦⎥ and λ = 2 1 7. T (u 2 ) = ⎢1 ⎥ and T (u3 ) = ⎢ 0 ⎥ . and 3 2 2 4 T (p 2 ) = 10 + 2( x + 1) = 12 + 2 x = − p1 + p 2 . v 2 = u1 + u 2 .

. so the eigenvalues of T are λ = −4 and λ = 3. A basis for 3 the eigenspace corresponding to λ = 3 is ⎡ 5⎤ ⎢ −2 ⎥ . then T (cx) = cAx + B ≠ c( Ax + B) = cT (x). B = PB → B′ [T ]B = PB → B′ PB′→ B = I n . every matrix is similar to itself since A = I −1 AI . 2 T ( x ) = 2 − 8x − 2x Thus the matrix for T relative to the standard ⎡5 6 2⎤ basis B is [T ]B = ⎢ 0 −1 −8⎥ . then A = P −1 (Q −1CQ) P = (QP)−1 C (QP). (a) T (1) = 5 + x 2 T(x) = 6 − x 2 A−1 = ( P −1BP)−1 = P −1 B −1P. Then λ3 − 2λ 2 − 15λ + 36 = (λ + 4)(λ − 3)2 = 0. T (kv) = kv kv = k k v v ≠ kT ( v) if k ≠ 1. so the polynomial so the polynomial ⎢ ⎥ ⎣ 1⎦ 3. True/False 8. The reduced form of A shows that the general solution of Ax = 0 is x = −s.5 −1 (a) False.. No. If v is an eigenvector of T corresponding to λ. if B and B′ are different. Since [ vi ]B = ei for each i where {e1 . . n. T (e2 ).. . then v is a nonzero vector which satisfies the equation T(v) = λv or (λI − T)(v) = 0.. (a) The matrix for T relative to the standard ⎡ 1 0 1 1⎤ basis is A = ⎢ 2 1 3 1⎥ . if A = P −1BP and B = Q −1CQ.Chapter 8: Linear Transformations SSM: Elementary Linear Algebra ⎡ 5+ 21 P [ I ]B P = ⎢ 2 ⎢ 0 ⎣ 0 ⎤ ⎥ . any two of T (e1 ). T (x1 + x 2 ) = A(x1 + x 2 ) + B ≠ ( Ax1 + B) + ( Ax 2 + B) = T (x1 ) + T (x2 ). invertibility is a similarity invariant. (c) True. then 13. C = D. 17. [T ]B′. Thus v is in the kernel of λI − T. ⎢ 6 ⎥ ⎬ is a basis for ⎩⎪ ⎢⎣ 1 ⎥⎦ ⎢⎣ 1 ⎥⎦ ⎭⎪ which [T ]B′ is diagonal. Since P 5− 21 ⎥ 2 ⎦ represents the transition matrix from the basis B′ to the standard basis B. v n of V. x2 = s.. 2. then we can let x = vi for each of the basis vectors v1. 8 −2 + x + x 2 is a basis in P 2 . A reduces to ⎢0 1 0 −1⎥ . this yields Cei = Dei for i = 1. for example. (b) A basis for the eigenspace corresponding to ⎡ −2 ⎤ λ = −4 is ⎢⎢ 83 ⎥⎥ . the reduced form of A indicates that T (e3 ) and 5 − 2x + x 2 is a basis in P 2 .. (b) True. so the polynomial ⎢⎣ 1⎥⎦ Chapter 8 Supplementary Exercises 1. then ⎧⎪ ⎡ −3− 21 ⎤ ⎡ −3+ 21 ⎤ ⎫⎪ B ′ = ⎨ ⎢ 6 ⎥ . if A = P −1 BP. 15. T (e4 ) form a basis for the range. (d) True. e n } is the standard basis for R n . let [T ]B be given by the matrix PB→ B′ . Since C[ x]B = D[ x]B for all x in V. and if c ≠ 1. Since corresponding columns of C and D are all equal. . (g) True characteristic equation for [T ]B is (h) False. 196 .. respectively. ⎢ ⎥ ⎣0 0 1 0 ⎦ Since row operations don’t change the dependency relations among columns... But Cei and Dei are the ith columns of C and D. A basis for the ⎢ ⎥ ⎣ 1 0 0 1⎦ range of T is a basis for the column space of ⎡ 1 0 0 1⎤ A. The ⎢ ⎥ ⎣ 1 0 −2 ⎦ (e) True (f) False. 5. let T be the zero operator.

T is not one-to-one. L(u 4 ) = u 4 . Thus. so a basis for ker(D) is f(x) = 1. ⎜⎜ ⎢ ⎥ ⎟⎟ ⎢ ⎥ ⎢ ⎥ ⎣ 1 0 −1⎦ ⎝ ⎣ 1⎦ ⎠ ⎣ −1⎦ Note that this can also be read directly from [T (x)]B . u3 = ⎢ . and (b) Since R(T) is three-dimensional and ker(T) is one-dimensional.SSM: Elementary Linear Algebra Chapter 8 Supplementary Exercises x3 = 0. (b) A−1 = ( P −1 BP)−1 = P −1 B −1 P Thus A−1 and B −1 are similar. L(u3 ) = u 2 . rank(T) = 2 and nullity(T) = 2. 13. =⎢ d ⎥⎦ ⎣ d T(X) = 0 gives the equations a+b+c = 0 2b + d = 0 d = 0. Thus c = −a and X is in ker(T) if it has the form ⎧ ⎡ 1 0⎤ ⎫ ⎡ k 0⎤ ⎢⎣ − k 0 ⎥⎦ . rank(T) = 3 and nullity(T) = 1. For X = ⎢ . T ⎜ ⎢0 ⎥ ⎟ = ⎢0 ⎥ . Since u2 = ⎢ ⎥ ⎥ 0 0 1 0 ⎣ ⎦ ⎣0 1 ⎥⎦ ⎣ ⎦ L(u1 ) = u1. ⎡ −4 0 9 ⎤ [T ]B′ = P −1[T ]B P = ⎢ 1 0 −2 ⎥ . the matrix of L relative to the 7. The transition matrix from B′ to B is ⎡ 1 1 1⎤ P = ⎢ 0 1 1⎥ . u4 = ⎢ . Hence. (a) Since A = P −1 BP. which is the kernel of T is ⎢ ⎥ . Thus. (a) D(f + g) = ( f ( x) + g ( x))′′ = f ′′( x) + g ′′( x) and D(kf ) = (kf ( x))′′ = kf ′′( x). thus [T }B = ⎢0 1 0⎥ . and ⎜⎜ ⎢ ⎥ ⎟⎟ ⎢ ⎥ ⎜⎜ ⎢ ⎥ ⎟⎟ ⎢ ⎥ ⎝ ⎣0 ⎦ ⎠ ⎣ 0⎦ ⎝ ⎣0 ⎦ ⎠ ⎣ 1⎦ ⎛ ⎡0 ⎤ ⎞ ⎡ 1⎤ ⎡ 1 −1 1⎤ T ⎜ ⎢0 ⎥ ⎟ = ⎢ 0⎥ . 19. g(x) = x. so ker(T ) = ⎨k ⎣⎢ −1 0⎦⎥ ⎬ which is ⎩ ⎭ (b) If f is in ker(D). T ⎜ ⎢ 1⎥ ⎟ = ⎢ 1⎥ . f ( x) = e x and g ( x) = e − x form a basis for this subspace. ⎣ c d ⎥⎦ ⎡a + c b + d ⎤ ⎡ b b ⎤ T(X ) = ⎢ + 0 ⎥⎦ ⎢⎣ d d ⎥⎦ ⎣ 0 ⎡ a + b + c 2b + d ⎤ . 9. (a) The matrix for T relative to B is ⎡ 1 1 2 −2 ⎤ ⎢ 1 −1 −4 6 ⎥ [T ]B = ⎢ ⎥. (c) D(f) = f if and only if f = f ( x) = ae x + be− x for arbitrary constants a and b. 197 . L(u 2 ) = u3 . ⎡a b ⎤ 11. ⎛ ⎡ 1⎤ ⎞ ⎡ 1⎤ ⎛ ⎡0 ⎤ ⎞ ⎡ −1⎤ 17. then f has the form f = f ( x) = a0 + a1 x. AT and BT are similar. 0⎥ 1⎥⎦ 15. nullity(T) = 1 and rank(T ) = dim( M 22 ) − nullity(T ) = 3. Thus. ⎢ 1 2 5 −6 ⎥ 3 −2 ⎥⎦ ⎢⎣3 2 ⎡1 ⎢0 standard basis is ⎢ ⎢0 ⎢⎣0 0 0 1 0 0 1 0 0 0⎤ 0⎥ ⎥. ⎡ −1⎤ ⎢ 1⎥ of A. ⎢ ⎥ ⎣ 0 1 1⎦ (b) Since rank(T) < dim(V). x4 = s so a basis for the null space one-dimensional. ⎢ 0⎥ ⎢⎣ 1⎥⎦ ⎡1 0 ⎤ . The standard basis for M 22 is u1 = ⎢ ⎣0 0 ⎥⎦ ⎡0 1 ⎤ ⎡0 0 ⎤ ⎡0 0 ⎤ . then AT = ( P −1BP)T = PT BT ( P −1 )T = (( PT )−1 )−1 BT ( P −1 )T = (( P −1 )T )−1 BT ( P −1 )T . thus ⎢ ⎥ ⎣ 0 0 1⎦ ⎡ 1 0 −1 2 ⎤ ⎢ 0 1 3 −4 ⎥ [T ]B reduces to ⎢ ⎥ which ⎢0 0 0 0 ⎥ ⎣⎢0 0 0 0 ⎦⎥ has rank 2 and nullity 2.

the points lie on the graph. J ( x) = a1 . 23. J (1) = x 2x D( x n ) = nx n −1 This gives the matrix shown. a2 .Chapter 8: Linear Transformations SSM: Elementary Linear Algebra 21. ⎜⎢ ⎥⎟ ⎜ a ⎟ ⎝⎣ 3⎦⎠ D(1) = 0 D( x) = 1 D( x 2 ) = # 1 2 x 2 1 3 x 3 1 n +1 x n +1 This gives the (n + 2) × (n + 1) matrix ⎡0 0 0 " 0 ⎤ ⎢1 0 0 " 0 ⎥ ⎢ ⎥ 1 ⎢0 2 0 " 0 ⎥ ⎢ ⎥. x2 . 198 . and x3 gives the values 25. and a3 . # Thus ⎡ a1 ⎤ T (a1 P1 ( x) + a2 P2 ( x) + a3 P3 ( x)) = ⎢ a2 ⎥ . respectively. (c) Note that a1 P1 ( x) + a2 P2 ( x) + a3 P3 ( x) evaluated at x1 . or ⎢ ⎥ ⎣ a3 ⎦ ⎛ ⎡ a1 ⎤ ⎞ T −1 ⎜ ⎢ a2 ⎥ ⎟ = a1 P1 ( x) + a2 P2 ( x) + a3 P3 ( x). since J ( x2 ) = Pi ( xi ) = 1 and Pi ( x j ) = 0 for i ≠ j. 1 ⎢0 0 3 " 0 ⎥ ⎢# # # # ⎥ ⎢ ⎥ 1 ⎢⎣0 0 0 " n+1 ⎥⎦ J ( xn ) = (d) From the computations in part (c).

and − 2 5 . y2 = −1. x1 + 4 x2 = −1 ⎡ 1 4 ⎤ ⎡ x1 ⎤ ⎡ −1⎤ which ⎢⎣0 1⎦⎥ ⎢ x ⎥ = ⎣⎢ −1⎦⎥ is x2 = −1 ⎣ 2⎦ gives the solution to the original system: x1 = 3. so the ⎢ ⎥ 5 ⎣ −1 4 5⎦ system is ⎡ 2 0 0 ⎤ ⎡ 1 −1 −1⎤ ⎡ x1 ⎤ ⎡ −4 ⎤ ⎢ 0 −2 0 ⎥ ⎢0 1 −1⎥ ⎢ x2 ⎥ = ⎢ −2 ⎥ .Chapter 9 Numerical Methods ⎡ 1 −1 −1⎤ ⎡ 1 −1 −1⎤ → ⎢ 0 −2 2 ⎥ → ⎢ 0 1 −1⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 0 4 1⎦ ⎣ 0 4 1⎦ ⎡ 1 −1 −1⎤ ⎡ 1 −1 −1⎤ → ⎢ 0 1 −1⎥ → ⎢0 1 −1⎥ = U ⎢ ⎥ ⎢ ⎥ ⎣ 0 0 5⎦ ⎣0 0 1⎦ 1 1 The multipliers used were . Section 9.1 1. the system is ⎡ 3 −6 ⎤ ⎡ x1 ⎤ ⎡ 3 0⎤ ⎡ 1 −2 ⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ = = ⎢⎣ −2 5⎥⎦ ⎢⎣ x2 ⎥⎦ ⎢⎣ −2 1⎥⎦ ⎢⎣ 0 1⎥⎦ ⎢⎣ x2 ⎥⎦ ⎢⎣1 ⎥⎦ ⎡ 1 −2 ⎤ ⎡ x1 ⎤ ⎡ y1 ⎤ = which reduces to ⎢ and 1⎥⎦ ⎢⎣ x2 ⎥⎦ ⎢⎣ y2 ⎥⎦ ⎣0 ⎡ 3 0 ⎤ ⎡ y1 ⎤ ⎡0 ⎤ ⎢⎣ −2 1⎥⎦ ⎢ y ⎥ = ⎢⎣1 ⎥⎦ . x2 = 1. ⎣ 2⎦ The second system is 3 y1 = 0 . ⎢ ⎥⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ −1 4 5⎦ ⎣0 0 1⎦ ⎣ x3 ⎦ ⎣ 6 ⎦ ⎡ 2 0 0 ⎤ ⎡ y1 ⎤ ⎡ −4 ⎤ ⎢ 0 −2 0 ⎥ ⎢ y2 ⎥ = ⎢ −2 ⎥ is ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ −1 4 5⎦ ⎣ y3 ⎦ ⎣ 6 ⎦ = −4 2 y1 −2 y 2 = −2 which has the solution − y1 + 4 y2 + 5 y3 = 6 y1 = −2. 1. 3. 1. so the system is ⎣ −1 3⎥⎦ ⎡ 2 0 ⎤ ⎡ 1 4 ⎤ ⎡ x1 ⎤ ⎡ −2 ⎤ ⎢⎣ −1 3⎥⎦ ⎢⎣0 1⎥⎦ ⎢ x ⎥ = ⎢⎣ −2 ⎥⎦ . so the first system x1 − 2 x2 = 0 ⎡ 1 −2 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ becomes ⎢ . 5. and . −4. which 2 3 ⎡ 2 0⎤ leads to L = ⎢ . and 2 2 ⎡ 2 0 0⎤ 1 . ⎣ 2⎦ x1 − x2 − x3 = −2 ⎡ 1 −1 −1⎤ ⎡ x1 ⎤ ⎡ −2 ⎤ ⎢0 1 −1⎥ ⎢ x2 ⎥ = ⎢ 1⎥ is x2 − x3 = 1 ⎢ ⎥⎢ ⎥ ⎢ ⎥ 0 0 1 0 x x3 = 0 ⎣ ⎦⎣ 3⎦ ⎣ ⎦ which gives the solution to the original system: x1 = −1. which has −2 y1 + y2 = 1 solution y1 = 0. Reduce the coefficient matrix. Reduce the coefficient matrix. Reduce the coefficient matrix. − . ⎢ x ⎥ = ⎢1 ⎥ or ⎥ 0 1 x2 = 1 ⎣ ⎦⎣ 2⎦ ⎣ ⎦ which gives the solution to the original system: x1 = 2. 8. 5 10 ⎤ ⎡ 1 1 2⎤ ⎡ 5 ⎢ −8 −7 −9 ⎥ → ⎢ −8 −7 −9⎥ ⎢ ⎥ ⎢ ⎥ 4 26⎦ ⎣ 0 4 26 ⎦ ⎣ 0 ⎡ 1 1 2 ⎤ ⎡ 1 1 2⎤ → ⎢0 1 7 ⎥ → ⎢0 1 7 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 0 4 26 ⎦ ⎣ 0 0 −2⎦ ⎡ 1 1 2⎤ → ⎢0 1 7 ⎥ = U ⎢ ⎥ ⎣ 0 0 1⎦ 2 y1 = −2 ⎡ 2 0 ⎤ ⎡ y1 ⎤ ⎡ −2 ⎤ ⎢⎣ −1 3⎦⎥ ⎢ y ⎥ = ⎣⎢ −2 ⎦⎥ is − y + 3 y = −2 which ⎣ 2⎦ 1 2 has the solution y1 = −1. ⎡ 2 8⎤ ⎡ 1 4 ⎤ ⎡ 1 4 ⎤ ⎡ 1 4 ⎤ ⎢⎣ −1 −1⎥⎦ → ⎢⎣ −1 −1⎥⎦ → ⎢⎣ 0 3⎥⎦ → ⎢⎣0 1⎥⎦ =U 1 1 The multipliers used were . −4.1 Exercise Set 9. y3 = 0. 7. y2 = 1. ⎡ 2 −2 −2 ⎤ ⎡ 1 −1 −1⎤ ⎢ 0 −2 2 ⎥ → ⎢ 0 −2 2 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ −1 5 2 ⎦ ⎣ −1 5 2 ⎦ The multipliers used were 199 1 1 . x3 = 0. x2 = 1. In matrix form. x2 = −1. which leads to L = ⎢ 0 −2 0 ⎥ . y2 = 1.

and −2. −1. x3 = 0. y3 = 0. so 0⎥ 4 ⎦⎥ (b) To change the 2 on the diagonal of L to a 1. 0 0 2 1 − 12 ⎤ ⎥ 0 1⎥ = U 0 1⎦⎥ 1 2 1 . 3 0 3 −1 0 ⎡ −1 0 ⎢ 2 3 ⎢ ⎢ 0 −1 ⎣⎢ 0 0 to the original system: x1 = −3. which leads to L = ⎢ 4 ⎢ 0 ⎣⎢ 0 the system is 0 0 2 1 x4 = 1. x3 = 2. the first column of L is divided by 2 and the diagonal matrix has a 2 as the 1. 2 ⎡ 2 0 0⎤ which leads to L = ⎢ −2 1 0⎥ where the ⎢ ⎥ ⎣ 2 0 1⎦ 1’s on the diagonal reflect that no multiplication was required on the 2nd and 3rd diagonal entries. x2 = 1.Chapter 9: Numerical Methods SSM: Elementary Linear Algebra ⎡ 5 0 0⎤ which leads to L = ⎢ −8 1 0 ⎥ . y4 = 1. which has the =3 =7 0 −1 0 ⎤ ⎡ x1 ⎤ ⎡ −5⎤ 1 0 2 ⎥ ⎢ x2 ⎥ ⎢ 3⎥ ⎥ ⎢ ⎥ = ⎢ ⎥ is 0 1 1⎥ ⎢ x3 ⎥ ⎢ 3⎥ 0 0 1⎦⎥ ⎢⎣ x4 ⎥⎦ ⎣⎢ 1⎦⎥ x1 − x3 = −5 x2 + 2 x4 = 3 . 0⎤ ⎡ 1 0 ⎥ ⎢0 ⎥⎢ 0 ⎥ ⎢0 4 ⎦⎥ ⎣⎢0 solution y1 = −5. The multipliers used were 1 . x2 = 1. −2. 200 . Reduce the coefficient matrix. y2 = 3. ⎢ ⎥⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 0 4 −2 ⎦ ⎣0 0 1⎦ ⎣ x3 ⎦ ⎣ 4 ⎦ ⎡ 5 0 0⎤ ⎡ y1 ⎤ ⎡ 0 ⎤ ⎢ −8 1 0⎥ ⎢ y2 ⎥ = ⎢ 1 ⎥ is ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 0 4 −2⎦ ⎣ y3 ⎦ ⎣ 4 ⎦ =0 5 y1 −8 y1 + y2 = 1 which has the solution 4 y2 − 2 y3 = 4 y1 = 0. ⎡1 ⎢0 →⎢ ⎢0 ⎢⎣ 0 The multipliers used were −1. ⎡ 1 0 −1 ⎢0 3 0 →⎢ ⎢ 0 −1 2 ⎣⎢ 0 0 1 0 ⎤ ⎡ y1 ⎤ ⎡ 5⎤ 0 ⎥ ⎢ y2 ⎥ ⎢ −1⎥ ⎥ ⎢ ⎥ = ⎢ ⎥ is 0 ⎥ ⎢ y3 ⎥ ⎢ 3⎥ 4 ⎦⎥ ⎢⎣ y4 ⎥⎦ ⎣⎢ 7 ⎦⎥ =5 = −1 . 0⎤ 6⎥ ⎥ 0⎥ 5⎦⎥ 0⎤ ⎡ 1 0 −1 ⎢0 1 0 6⎥ ⎥→⎢ 0⎥ ⎢ 0 −1 2 5⎦⎥ ⎣⎢ 0 0 1 0 0 2 1 − y1 2 y1 + 3 y2 − y2 + 2 y3 y3 + 4 y4 x1 + x2 + 2 x3 = 0 ⎡ 1 1 2 ⎤ ⎡ x1 ⎤ ⎡0 ⎤ ⎢0 1 7 ⎥ ⎢ x2 ⎥ = ⎢1 ⎥ is x2 + 7 x3 = 1 ⎢ ⎥⎢ ⎥ ⎢ ⎥ x3 = 0 ⎣0 0 1⎦ ⎣ x3 ⎦ ⎣0 ⎦ which gives the solution to the original system: x1 = −1. 9. ⎡ 1 1 − 1⎤ ⎡ 2 1 −1⎤ 2 2⎥ ⎢ ⎢ −2 −1 2⎥ → −2 −1 2 ⎢ ⎥ ⎢ ⎥ ⎢⎣ 2 1 ⎣ 2 1 0⎦ 0 ⎥⎦ 0⎤ ⎡ 1 2⎥ ⎢0 ⎥→⎢ 2⎥ ⎢0 5⎥⎦ ⎢⎣ 0 0 −1 0 ⎤ 1 0 2⎥ ⎥ 0 1 1⎥ 0 1 5⎥⎦ ⎡1 ⎢0 →⎢ ⎢0 ⎣⎢ 0 0 −1 1 0 0 1 0 0 0⎤ ⎡ 1 2⎥ ⎢0 ⎥→⎢ 1⎥ ⎢ 0 4 ⎦⎥ ⎣⎢ 0 0 −1 0 ⎤ 1 0 2⎥ ⎥ =U 0 1 1⎥ 0 0 1⎦⎥ ⎡ −1 ⎢ 2 1 and . which gives the solution x3 + x4 = 3 x4 = 1 ⎡1 ⎢0 ⎢ ⎢0 ⎣⎢0 0⎤ 2⎥ ⎥ 0⎥ 5⎦⎥ 11. y2 = 1. 0 1 1⎥ ⎢ x3 ⎥ ⎢ 3⎥ 0 0 1⎦⎥ ⎣⎢ x4 ⎦⎥ ⎣⎢ 7 ⎦⎥ ⎡ −1 0 ⎢ 2 3 ⎢ ⎢ 0 −1 ⎣⎢ 0 0 ⎡1 1 − 1⎤ ⎡1 2 2⎥ ⎢ → ⎢0 0 1⎥ → ⎢0 ⎢ 0 ⎥⎦ ⎢⎣0 ⎣⎢ 2 1 1. so the system ⎢ ⎥ ⎣ 0 4 −2 ⎦ ⎡ 5 0 0 ⎤ ⎡ 1 1 2 ⎤ ⎡ x1 ⎤ ⎡ 0 ⎤ is ⎢ −8 1 0 ⎥ ⎢0 1 7 ⎥ ⎢ x2 ⎥ = ⎢ 1 ⎥ . 0 −1 1 0 0 2 0 1 1 . y3 = 3. 1 0⎤ ⎡ −1 0 ⎡ 1 0 −1 ⎢ 2 3 −2 6 ⎥ ⎢ 2 3 −2 ⎢ ⎥→⎢ ⎢ 0 −1 2 0 ⎥ ⎢ 0 −1 2 1 5⎦⎥ 1 ⎣⎢ 0 0 ⎣⎢ 0 0 0 −1 0 ⎤ ⎡ x1 ⎤ ⎡ 5⎤ 1 0 2 ⎥ ⎢ x2 ⎥ ⎢ −1⎥ ⎥ ⎢ ⎥ = ⎢ ⎥. (a) Reduce A to upper triangular form. 2. 1 entry. 2 0⎤ 0⎥ ⎥ .

then ⎢ ⎥⎢ ⎥ 2 2 1⎦ ⎣ 0 0 1⎦ − ⎣ ⎡ 1 0 0⎤ ⎡ 3 0 0 ⎤ ⎡ 1 −4 2 ⎤ A = ⎢0 1 0⎥ ⎢ 0 2 0 ⎥ ⎢0 1 0 ⎥ is the ⎢ ⎥⎢ ⎥⎢ ⎥ ⎣ 2 −2 1⎦ ⎣ 0 0 1⎦ ⎣0 0 1⎦ LDU-decomposition of A. −3. 17 17 17 ⎡ 1 0 0 ⎤ ⎡ 2 0 0 ⎤ ⎡ 1 12 − 12 ⎤ ⎥ = ⎢ −1 1 0 ⎥ ⎢ 0 1 0 ⎥ ⎢0 0 1⎥ ⎢ ⎥⎢ ⎥⎢ ⎣ 1 0 1⎦ ⎣ 0 0 1⎦ ⎢⎣0 0 1⎥⎦ is the LDU-decomposition of A.SSM: Elementary Linear Algebra Section 9. ⎡ 1 − 1 0⎤ ⎡ 3 −1 0 ⎤ 3 ⎥ ⎢0 2 1⎥ → ⎢0 2 1⎥ ⎢ ⎢ ⎥ ⎢⎣ 3 −1 1⎥⎦ ⎣ 3 −1 1⎦ ⎡ 1 0 0 ⎤ ⎡ 2 1 −1⎤ A = L2U 2 = ⎢ −1 1 0 ⎥ ⎢ 0 0 1⎥ . the system to solve is ⎢ ⎥ ⎣ 1⎦ 1 ⎡ 3 0 0 ⎤ ⎢⎡ 1 − 3 0 ⎥⎤ ⎡ x1 ⎤ ⎡ −2 ⎤ ⎢ ⎥ ⎢0 2 0 ⎥ ⎢0 1 12 ⎥ ⎢ x2 ⎥ = ⎢⎢ 4 ⎥⎥ . Since ⎢ ⎥ ⎣6 −4 1⎦ ⎡ 1 0 0⎤ ⎡ 3 0 0⎤ L1 = ⎢ 0 1 0 ⎥ ⎢ 0 2 0 ⎥ . If rows 2 and 3 of A are interchanged. then the resulting matrix has an LU-decomposition. x3 = . Reduce A to upper triangular form. Since P = P −1 . so the ⎢ ⎥ ⎢ ⎥ ⎣ 0 0 1⎦ ⎣5⎦ system P −1 Ax = P −1b is ⎡ 1 0 0⎤ ⎡ 1 2 2 ⎤ ⎡ x1 ⎤ ⎡1 ⎤ ⎢0 1 0⎥ ⎢ 0 1 4 ⎥ ⎢ x2 ⎥ = ⎢ 2⎥ . ⎢ ⎥ ⎢ ⎥ ⎣0 1 0 ⎦ ⎣ 3 −1 1⎦ Reduce PA to upper triangular form. y1 =1 ⎡ 1 0 0 ⎤ ⎡ y1 ⎤ ⎡ 1 ⎤ ⎢0 y2 =2 1 0 ⎥ ⎢ y2 ⎥ = ⎢ 2 ⎥ is ⎢ ⎥⎢ ⎥ ⎢ ⎥ − 3 5 1 5 3 y1 − 5 y2 + y3 = 5 ⎣ ⎦ ⎣ y3 ⎦ ⎣ ⎦ which has the solution y1 = 1. P −1 = ⎢ 1 0 0 ⎥ and P −1b = ⎢ 2 ⎥ . ⎢ ⎥⎢ ⎥⎢ ⎥ ⎣ 0 1 0⎦ ⎣ 3 0 1⎦ ⎣0 0 1⎦ ⎡ −2 ⎤ Since Pb = ⎢ 4⎥ . −6. x2 = − .1 A = L1 DU1 x1 + 2 x2 + 2 x3 = 1 ⎡ 1 2 2⎤ ⎡ x1 ⎤ ⎡ 1⎤ ⎢0 1 4⎥ ⎢ x2 ⎥ = ⎢ 2 ⎥ is x2 + 4 x3 = 2 ⎢ ⎥⎢ ⎥ ⎢ ⎥ 0 0 17 12 x 17 x3 = 12 ⎣ ⎦⎣ 3⎦ ⎣ ⎦ which gives the solution of the original system: 21 14 12 x1 = . ⎡ 3 −12 6 ⎤ ⎡ 1 −4 2 ⎤ ⎢0 2 0⎥ → ⎢0 2 0⎥ ⎢ ⎥ ⎢ ⎥ 6 28 13 6 28 13⎦ − − ⎣ ⎦ ⎣ ⎡ 1 −4 2 ⎤ ⎡ 1 −4 2 ⎤ → ⎢ 0 2 0⎥ → ⎢0 1 0⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 0 −4 1⎦ ⎣0 −4 1⎦ ⎡ 1 −4 2 ⎤ → ⎢0 1 0⎥ = U ⎢ ⎥ ⎣ 0 0 1⎦ 1 1 The multipliers used were . ⎢ ⎥⎢ ⎥ ⎣ 3 0 1⎦ ⎣0 0 1⎦ ⎣ x3 ⎦ ⎣ 1⎦ 3 y1 = −2 ⎡ 3 0 0 ⎤ ⎡ y1 ⎤ ⎡ −2 ⎤ ⎢0 2 0 ⎥ ⎢ y2 ⎥ = ⎢ 4 ⎥ is 2 y2 =4 ⎢ ⎥⎢ ⎥ ⎢ ⎥ 3 0 1 1 3 y1 + y3 = 1 ⎣ ⎦ ⎣ y3 ⎦ ⎣ ⎦ ⎡0 1 0⎤ ⎡1 ⎤ 15. we have ⎢ ⎥ ⎣ 3 0 1⎦ 1 ⎡ 1 0 0⎤ ⎡ 3 0 0 ⎤ ⎢⎡ 1 − 3 0 ⎥⎤ A = PLU = ⎢ 0 0 1⎥ ⎢ 0 2 0 ⎥ ⎢0 1 12 ⎥ . y2 = 2. so 2 3 ⎡ 3 0 0⎤ L = ⎢0 2 0⎥ . y3 = 12. and . ⎢ ⎥⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 3 −5 1⎦ ⎣ 0 0 17 ⎦ ⎣ x3 ⎦ ⎣ 5 ⎦ 2 which has the solution y1 = − . . 3 y3 = 3. and L2 = L1. (c) Let U 2 = DU . ⎡ 1 − 1 0⎤ ⎡ x ⎤ ⎡− 2 ⎤ 3 ⎢ ⎥ ⎢ 1 ⎥ ⎢ 3⎥ 1 12 ⎥ ⎢ x2 ⎥ = ⎢ 2 ⎥ is ⎢0 ⎢ ⎥ 0 1⎦ ⎣ x3 ⎦ ⎢⎣ 3⎥⎦ ⎣0 201 . then 17. PA = ⎢ 0 2 1⎥ . For ⎡ 1 0 0⎤ ⎡ 3 −1 0 ⎤ P = ⎢0 0 1⎥ . ⎢ ⎥⎢ ⎥ ⎣ 1 0 1⎦ ⎣ 0 0 1⎦ 13. ⎡ 1 − 1 0⎤ ⎡ 1 − 1 0⎤ 3 3 ⎢ ⎥ ⎢ ⎥ → 0 → ⎢0 1 12 ⎥ = U 2 1 ⎢ ⎥ ⎢ ⎥ ⎢⎣ 0 0 1⎥⎦ 0 1⎦ ⎣0 1 1 The multipliers used were . 2 3 ⎡ 3 0 0⎤ which leads to L1 = ⎢0 2 0 ⎥ . and 4. y2 = 2.

16228 is the dominant eigenvalue and the dominant ⎡ 1 ⎤ ⎡ 1 ⎤ eigenvector is ⎢ ≈⎢ . 202 .78446 ⎥⎦ Ax1 ⎡ 0.09902 ⎤ Ax1 ≈ ⎢ ≈ ⎣ −1 −1⎥⎦ ⎢⎣ −0.19612⎥⎦ ⎢⎣ −0.15206 ⎥⎦ ⎢⎣ −0.16201⎥⎦ ⎡ 5 −1⎤ ⎡ 0.09552 ⎤ Ax 4 = ⎢ ≈ ⎣ −1 −1⎥⎦ ⎢⎣ −0.16018⎥⎦ which x 4 approximates to three decimal place accuracy. so the eigenvalues of A are 1 1 to the original system: x1 = − .15977 ⎥⎦ ⎢⎣ −0.19612 ⎥⎦ ⎡ 5 −1⎤ ⎡ 0. a a Because the solution is unique. ⎥ ⎥ 0 16228 − . x2 = . it can be written as y ⎤ ⎡a b ⎤ ⎡ 1 0⎤ ⎡ x y ⎤ ⎡ x ⎢⎣ c d ⎥⎦ = ⎢⎣ w 1 ⎥⎦ ⎢⎣ 0 z ⎥⎦ = ⎢⎣ wx wy + z ⎥⎦ which leads to the equations x=a y=b wx = c wy + z = d Since a ≠ 0. ⎡ 5 −1⎤ ⎡1 ⎤ ⎡ 5⎤ Ax 0 = ⎢ = ⎣ −1 −1⎥⎦ ⎢⎣0 ⎥⎦ ⎢⎣ −1⎥⎦ Ax 0 1 ⎡ 5⎤ ⎡ 0.2 λ 2 − 6λ − 4 = 0.15206⎥⎦ ⎡ 5 −1⎤ ⎡ 0. 2 2 x3 = 3. 2 3 which gives the solution 3. if the matrix cannot be reduced to row echelon form without interchanging rows.16226 (c) True λ( 4) = ( Ax4 )T x4 ≈ 5. w = . the LU-decomposition is also unique. y = b. so the eigenvalues of A are Exercise Set 9. (a) If A has such an LU-decomposition.98715⎤ ⎡ 5. λ = 2 ± 10. The characteristic equation of A is Section 9. λ = 2 + 10 ≈ 5.1 (a) False.98709 ⎤ The dominant unit eigenvector is ≈ ⎢ .09391⎤ Ax 2 ≈ ⎢ ≈ ⎣ −1 −1⎥⎦ ⎢⎣ −0.98679 ⎤ x3 = ≈⎢ Ax 2 ⎣ −0.16201⎥⎦ ⎢⎣ −0. (a) λ3 = −8 is the dominant eigenvalue.16138 λ(3) = ( Ax3 )T x3 ≈ 5.98837 ⎤ ⎡ 5. True/False 9.98837 ⎤ x2 = ≈ Ax1 ⎢⎣ −0. λ(1) = ( Ax1 )T x1 ≈ 5.16223 (d) True λ( 4) approximates λ to four decimal place accuracy. λ = 3 + 13 ≈ 6.98050⎤ ⎡ 5. then it does not have an LU-decomposition. and a bc ad − bc z=d− = .15977 ⎥⎦ (b) From part (a) the LU-decomposition is b ⎤ ⎡a b ⎤ ⎡ 1 0⎤ ⎡ a ⎢⎣ c d ⎥⎦ = ⎢ c 1 ⎥ ⎢ 0 ad −bc ⎥ . then A does not have an LU-decomposition. (e) True 5.98715⎤ x4 = ≈⎢ Ax 3 ⎣ −0.60555 is the dominant eigenvalue with corresponding scaled 1.82478⎥⎦ Ax 3 ⎡ 0.98679 ⎤ ⎡ 5.24297 (b) False.Chapter 9: Numerical Methods 1 x1 − x2 3 x2 + =− 1 x3 = 2 2 x3 = 3 SSM: Elementary Linear Algebra (b) There is no dominant eigenvalue since λ1 = λ 4 = 5.98058⎤ x1 = = ≈ Ax 0 26 ⎢⎣ −1⎥⎦ ⎢⎣ −0.83631⎥⎦ Ax 2 ⎡ 0. the system has the unique c solution x = a.09596 ⎤ Ax 3 ≈ ⎢ ≈ ⎣ −1 −1⎥⎦ ⎢⎣ −0.82738⎥⎦ ⎡ 0. ⎣ −0. λ( 2) = ( Ax2 )T x2 ≈ 5. a ⎦⎥ ⎣⎢ a ⎦⎥ ⎣⎢ ⎡ 5 −1⎤ ⎡ 0. The characteristic polynomial of A is λ 2 − 4λ − 6 = 0.2 λ = 3 ± 13. ⎦ ⎣3 − 10 ⎦ ⎣ 19. if the row equivalence of A and U requires interchanging rows of A.

with ⎡ 1⎤ dominant eigenvector ⎢ ⎥ . 1 ⎣⎢ 1 ⎦⎥ ⎣ (b) λ(1) = ⎡ 1 −3⎤ ⎡1⎤ ⎡ −2⎤ Ax 0 = ⎢ = ⎣ −3 5⎥⎦ ⎢⎣1⎥⎦ ⎢⎣ 2 ⎥⎦ Ax 0 ⎡ −1⎤ x1 = =⎢ ⎥ max( Ax0 ) ⎣ 1⎦ λ(1) = Ax1 ⋅ x1 =6 x1 ⋅ x1 Ax 2 ⋅ x 2 = 6.886 ⎢ ⎥ ⎣ 0.61538⎥⎦ 1 ⎣ −3 5⎥⎦ ⎢⎣ Ax 3 ⎡ −0.5⎤ x2 = = max( Ax1 ) ⎢⎣ 1 ⎥⎦ λ ( 2) = Ax1 ⋅ x1 = 2.SSM: Elementary Linear Algebra Section 9.8⎥⎦ ⎢⎣ −2. λ(6) ≈ 10. ⎣ −1⎦ ⎡ 1 −3⎤ ⎡ −0.903 ⎢ ⎥ ⎣ 0.344⎤ x 4 ≈ ⎢ 0. λ(1) ≈ 7.8⎥⎦ ⎡ 2 −1⎤ ⎡ 1 ⎤ ⎡ 2.001 3 or 0.044 ⎥ . By Formula (10). max( A x0 ) ⎣ 1 ⎦ Ax 5 ⋅ x 5 Thus λ(5) = ≈ 2.53518⎤ eigenvector ⎢ 3 ⎥ ≈ ⎢ ⎥⎦ .53846 ⎤ ⎡ −3.2 ⎡ 2 − 13 ⎤ ⎡ −0.997 x3 ⋅ x 3 (d) The percentage error is 3 − 2.5⎤ Ax1 = ⎢ = ⎣ −1 2 ⎥⎦ ⎢⎣ −0. 9.197 ⎥ .507 ⎤ x 2 ≈ ⎢ 0.320 ⎥ .016 ⎥ .5⎤ ⎡ −3.5⎤ Ax 2 = ⎢ = ⎣ −3 5⎥⎦ ⎢⎣ 1 ⎥⎦ ⎢⎣ 6.668 ⎥ .827 ⎢ ⎥ ⎣ 0. λ(3) ≈ 10.934⎦ ⎡0. it takes 8 iterations.99993.668 ⎦ ⎡0.53846 ⎤ Ax 3 ≈ ⎢ ⎥⎦ ≈ ⎢⎣ 6. λ(5) ≈ 10.6 x 2 ⋅ x2 λ(3) = Ax 3 ⋅ x 3 ≈ 2.53488⎤ x4 = ≈ ⎥⎦ 1 max( Ax3 ) ⎢⎣ λ ( 4) = Ax 2 ⋅ x 2 ≈ 2. (a) Ax0 = ⎢ = ⎣ −1 2 ⎥⎦ ⎢⎣0 ⎥⎦ ⎢⎣ −1⎥⎦ Ax0 ⎡ 1 ⎤ x1 = = max( Ax0 ) ⎢⎣ −0.60555 x 4 ⋅ x4 ⎡ 2 −1⎤ ⎡1 ⎤ ⎡ 2⎤ 7.1%.60550 x 3 ⋅ x3 ⎡ 1 −3⎤ ⎡ −0.302 ⎤ x6 ≈ ⎢0. λ( 4) ≈ 10.800 ⎦ ⎡ 0.929 ⎥⎦ 203 . (a) Starting with x0 = ⎢ 0⎥ . x5 ⋅ x 5 x5 = Ax 3 ⋅ x 3 ≈ 6.997 = 0.8 x1 ⋅ x1 5 ⎡1 ⎤ 13.968 ⎢ ⎥ ⎣ 0.5⎥⎦ ⎢⎣ 2 ⎥⎦ Ax1 ⎡ 1 ⎤ x2 = = max( Ax1 ) ⎢⎣ −0.380 ⎤ x3 ≈ ⎢0.99180 ⎤ ≈⎢ ⎥.53846 ⎤ x3 = ≈ ⎥⎦ 1 max( Ax2 ) ⎢⎣ λ(3) = λ ( 2) = (c) The characteristic polynomial of A is ⎡ 1 −3⎤ ⎡ −1⎤ ⎡ −4 ⎤ Ax1 = ⎢ = ⎣ −3 5⎥⎦ ⎢⎣ 1⎥⎦ ⎢⎣ 8⎥⎦ Ax1 ⎡ −0.229 ⎤ x1 ≈ ⎢0.976 x 2 ⋅ x2 λ 2 − 4λ + 3 = (λ − 3)(λ − 1) = 0 so the dominant eigenvalue of A is λ = 3.948 ⎦ ⎡0.8⎤ Ax 2 = ⎢ = ⎣ −1 2 ⎥⎦ ⎢⎣ −0.904 ⎦ ⎡ 0.5⎥⎦ Ax 2 ⎡ −0.317 ⎤ x5 ≈ ⎢ 0. ⎢ ⎥ ⎣ 0⎦ ⎡0.953⎦ Ax 4 ⋅ x 4 ≈ 6.6⎥⎦ Ax 2 ⎡ 1 ⎤ x3 = ≈⎢ max( Ax2 ) ⎣ −0.5⎥⎦ ⎡ 2 −1⎤ ⎡ 1 ⎤ ⎡ 2.632 ⎢ ⎥ ⎣0.096⎥ .622 ⎢ ⎥ ⎣ 0. A5 x0 ⎡0. λ( 2) ≈ 9.

513 ⎢ 0. 1 ⎢3⎥ ⎢0⎥ ⎣ ⎦ ⎡ 0.498⎥ ⎣⎢0.909 ⎢ ⎥ ⎣ 0. λ ≈ 8. λ(8) ≈ 10.70656 ⎤ ⎢0.951⎦⎥ ⎡2⎤ ⎢3⎥ ⎢2⎥ a0 = ⎢ 3 ⎥ . λ ≈ 8.006 ⎥ .957 ⎦ ⎡1 ⎤ 1.513 ⎢ 0. ⎢ ⎥ ⎣3⎦ ⎡1 ⎤ ⎢ 0⎥ (b) Starting with x0 = ⎢ ⎥ .70225⎤ ⎢ 0.70656 ⎥ ≈⎢ a2 = ⎥ AT Aa1 ⎢ 0.03925⎥ ⎢⎣ 0 ⎥⎦ AT Aa1 204 . λ ≈ 8. λ ≈ 6. ⎢ 0⎥ ⎢⎣ 0⎥⎦ ⎡ 0 0 1⎤ ⎡ 2 ⎤ ⎡3⎤ 3.249 ⎤ ⎢ 0 ⎥ ( 2) x2 ≈ ⎢ ⎥ . λ ≈ 8. λ ≈ 8.3 ⎡ 0.953⎦⎥ ⎡ 0. so we start with ⎢1 ⎥ ⎢⎣ 0 ⎥⎦ ⎡ 0.79262 ⎥⎦ AT h1 ⎡2 ⎢1 5.382 ⎢0.503 ⎢ 0.905⎦⎥ a1 = ⎡ 0.577 ⎥⎦ ⎡0.65094 ⎦ ⎡1.266 ⎥ ⎢⎣ 0.376 ⎥ ⎣⎢ 0.107 ⎥ ( 7) x7 ≈ ⎢ ⎥ .305⎥ ⎢⎣ 0.69246 ⎦ ⎡0.954 ⎥⎦ ⎡0.60971⎤ ≈⎢ 0 ⎥ AT h1 ⎢⎣ 0.175⎤ ⎢ 0.167 ⎤ ⎢ 0.073⎥ ( 4) x4 ≈ ⎢ ⎥ .830 ⎦⎥ ⎡ 0.956⎦ Section 9.333 ⎢0.70225⎥ a1 = ≈⎢ ⎥ AT Aa0 ⎢0.294⎤ x7 ≈ ⎢ 0. AT A = ⎢ ⎢0 ⎣⎢ 0 ⎡ 0.091⎥ (5) x5 ≈ ⎢ ⎥ .945⎥⎦ ⎡2⎤ ⎢2⎥ The column sums of A are ⎢ ⎥ .908 ⎢ ⎥ ⎣ 0.062 ⎢ 0.101⎥ (6) x6 ≈ ⎢ ⎥ .290 ⎤ x8 ≈ ⎢ −0.30189 ⎤ AT h1 ≈ ⎢ 0 ⎥ ⎢ ⎥ ⎣1.65094 ⎥ h1 = ⎢ ⎥ Aa 0 ⎣ 0.511 ⎢ 0. The row sums ⎢ ⎥ ⎣2⎦ ⎡2⎤ of AT are a0 = ⎢ 0 ⎥ .245⎥ ⎣⎢ 0.234 ⎥ ⎣⎢ 0.3 Exercise Set 9.002⎥ .228⎥ ⎢⎣0.39057 ⎤ Aa 0 ≈ ⎢ 0.110 ⎥ (8) x8 ≈ ⎢ ⎥ .158⎤ ⎢0.041⎥ (3) x3 = ⎢ ⎥ .933⎥⎦ 1 2 0 0 0 0 1 0 0⎤ 0⎥ ⎥ 0⎥ 0 ⎦⎥ ⎡0.11704 ⎥ ⎢⎣ 0 ⎥⎦ AT Aa0 ⎡ 0.162 ⎤ ⎢ 0.193⎤ ⎢ 0.476 ⎢ 0.577 ⎤ ⎢ 0 ⎥ (1) x1 ≈ ⎢ ⎥ .Chapter 9: Numerical Methods SSM: Elementary Linear Algebra ⎡ 0. it takes 8 iterations.577 ⎥ ⎢⎣0. The row sums of A are h0 = ⎢ 2 ⎥ . Aa0 = ⎢ 1 0 1⎥ ⎢ 0 ⎥ = ⎢5⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎣ 1 0 1⎦ ⎣ 3 ⎦ ⎣5⎦ ⎡0. λ(7) ≈ 10. λ ≈ 8. λ ≈ 8.159 ⎤ ⎢ 0.

while sites 1 and 5 are irrelevant.58889 ⎥ a3 = T A Aa 2 ⎢0. 682 × 109 flops or 666. so it will take 0.15250 ⎤ ⎢0. 000 = 104 : 2 4 3 3 4 2 7 4 (10 ) + (10 ) − (10 ) 3 2 6 = 666.68825⎥ 1 ⎢ ⎥ ⎢ ⎥ a0 = 2 ⎥ ≈ ⎢ 0. 655. 205 . 816.58289⎥ a2 = T A Aa1 ⎢0. followed by sites 3 and 4.72861⎥ T A Aa1 ⎢ ⎥ ≈ ⎢ 0.59086 ⎥ a5 = T A Aa 4 ⎢0.07075⎥⎦ ⎡ 0. the flops for both phases is 2 3 3 3 3 2 7 3 (10 ) + (10 ) − (10 ) = 668.70710 ⎤ ⎢0.22942 ⎤ ⎢ 3⎥ ⎢ 0.4 Exercise Set 9. ⎡ 0.00436 ⎥ ⎢⎣ 0 ⎥⎦ AT Aa3 ⎡ 0.00048⎥ ⎣⎢ 0 ⎦⎥ AT Aa5 From a6 . 500. so we start with ⎢ 1⎥ ⎢⎣ 2 ⎥⎦ ⎡ 1⎤ ⎡ 0.55916 ⎥ T A Aa 0 ⎢0.70705⎥ ≈⎢ a3 = ⎥ AT Aa3 ⎢ 0. (a) For n = 1000 = 103 .6681655 × 10−1 ≈ 0.59097 ⎥ T A Aa5 ⎢ 0.01309 ⎥ ⎢⎣ 0 ⎥⎦ AT Aa 2 T ⎡0.32790 ⎥ ⎢⎣ 0.4 ⎡ 1⎤ ⎢ 3⎥ ⎢ ⎥ The column sums of A are ⎢ 2 ⎥ . (c) n = 100.08327 ⎤ ⎢ 0.22942 ⎥ ⎢⎣ 2 ⎥⎦ ⎢⎣ 0.30500 ⎥ ⎢⎣0. ⎡1 ⎢0 ⎢ T 7.59045⎥ T A Aa3 ⎢ 0.00612⎤ ⎢ 0. 3 2 6 which is 0.73455⎥ A Aa 2 ⎢ ⎥ ≈ ⎢0.SSM: Elementary Linear Algebra Section 9. ⎢ 19 ⎢ 1⎥ ⎢ 0.70711⎥ =⎢ a6 = ⎥ T A Aa5 ⎢0.01908⎥⎦ ⎡ 0.70705⎤ ⎢ 0.4 ⎡0. 000 flops or 666.6681655 gigaflops.73693⎥ T A Aa5 ⎢ ⎥ a6 = ≈ ⎢0.668 seconds which is about 18.03677 ⎥⎦ ⎡0.13531⎥⎦ 1.45883⎥⎦ ⎡0. while sites 3 and 4 are irrelevant. 000 = 105 2 5 3 3 5 2 7 5 (10 ) + (10 ) − (10 ) 3 2 6 ≈ 666.01179 ⎤ ⎢0.70711⎤ ⎢ 0.32267 ⎥ ⎢⎣ 0.067 second.5 hours. it is apparent that sites 1 and 2 are tied for the most authority.32670 ⎥ ⎢⎣ 0.71166 ⎥ T A Aa 0 ⎢ ⎥ a1 = ≈ ⎢0.00990⎥⎦ From a6 .68 seconds.70711⎤ ⎢ 0.73679 ⎥ AT Aa 4 ⎢ ⎥ ≈ ⎢0.70710 ⎥ ≈⎢ a4 = ⎥ AT Aa3 ⎢0.32796⎥ ⎢⎣ 0.682 gigaflops The time is about 66. it is apparent that site 2 has the most authority.70711⎥ ≈⎢ a5 = ⎥ T A Aa 4 ⎢0. (b) n = 10.00145⎥ ⎣⎢ 0 ⎦⎥ AT Aa 4 ⎡ 0.04370 ⎤ ⎢ 0.32766⎥ ⎢⎣0.73630⎥ T A Aa3 ⎢ ⎥ a4 = ≈ ⎢ 0.165.816655 gigaflops The time is about 66.02273⎤ ⎢ 0. A A = ⎢ 0 ⎢0 ⎢⎣ 1 0 3 2 1 0 0 2 2 1 0 0 1 1 1 0 1⎤ 0⎥ ⎥ 0⎥ 0⎥ 2⎥⎦ Section 9.45883⎥ .25416 ⎥⎦ ⎡0.

which requires n − 1 flops. We have ⎡ 1 ⎡1 −1⎤ ⎡1 ⎤ ⎢ Av1 = u1 = = σ1 2 ⎢⎣1 1⎥⎦ ⎢⎣0 ⎥⎦ ⎢ ⎣ 1 7. Exercise Set 9.67 ≈ 9.52 seconds. and σ1 = 5 and σ 2 = 0 are singular values of A. for a total of 3. λ 2 = 0. This u2 = = σ2 2 2 ⎢⎣1 1⎥⎦ ⎢⎣1 ⎥⎦ ⎢ 1 ⎥ 2 ⎣ ⎦ results in the following singular value decomposition of A: ⎡ 1 − 1 ⎤⎡ 2 0 ⎤ ⎡1 0 ⎤ T 2⎥ ⎢ A = U ΣV = 12 ⎢ ⎥ 1 ⎢ ⎥⎣ 0 2 ⎦ ⎢⎣0 1 ⎥⎦ 2⎦ ⎣ 2 requires n 2 flops. thus the singular value of A is σ1 = 5. Let C = [cij ] = AB. 1 ⎥ 2⎦ and ⎡ 1 ⎤ 1 ⎡1 −1⎤ ⎡ 0⎤ ⎢ − 2 ⎥ Av = . (a) n = 100.667 ×1012 eigenspace (which is all of R 2 ). 2 3 2 12 n ≈ (10 ) ≈ 666.667 × 1015 3. 5.67 ×105 seconds. adding three numbers requires 2 flops.e. 000 = 104 2 3 2 n ≈ × 1012 3 3 ≈ 0.. Note that adding two numbers requires 1 flop. which will 666. 5. which will take about 28.1× 109 0. Thus. each of the n 2 entries 206 .57 seconds.0014 second. and the vectors v1 = ⎢ ⎥ and ⎣0 ⎦ ⎡0 ⎤ v 2 = ⎢ ⎥ form an orthonormal basis for the ⎣1 ⎦ n 2 = 1010 = 10 × 109 The backward phase would require about 10 seconds. Computing cij requires first multiplying each of the n entries aik by the corresponding entry bkj . The eigenvalues of ⎡ 1 2 ⎤ ⎡ 1 −2 ⎤ ⎡ 5 0 ⎤ AT A = ⎢ = are λ1 = 5 1⎥⎦ ⎢⎣0 5⎥⎦ ⎣ −2 1⎥⎦ ⎢⎣ 2 and λ 2 = 5 (i.5 (b) n 2 ≈ 108 = 0.67 × 109 3 3 666. so the computer would have to execute 2(667) = 1334 gigaflops per second. values of A are σ1 = 2 and σ 2 = 2.5 1. which requires n flops. which will take about 0. The only eigenvalue of ⎡ 1 1⎤ ⎡1 −1⎤ ⎡ 2 0⎤ AT A = ⎢ = is λ = 2 ⎣ −1 1⎥⎦ ⎢⎣1 1⎥⎦ ⎢⎣ 0 2⎥⎦ ⎡1 ⎤ (multiplicity 2).67 gigaflops are required. n = 10. ⎢ ⎥ ⎢ ⎥ ⎣0 ⎦ ⎣ 0 0 0⎦ (c) This is the same as part (a). the forward phase would require about 6. 000 = 104 (a) n 2 (2n − 1) = 2n3 − n2 flops. (b) n = 10. The singular ≈ 6. = 6. 000 = 105 2 3 2 n ≈ × 1015 3 3 ≈ 0.67 × 102 × 109 About 667 gigaflops are required. λ = 5 is an eigenvalue of multiplicity 2). Multiplying each of the n 2 entries of A by c 1 ⎤ 2⎥ .52 seconds.1 gigaflop is required. and in general.67 × 105 ×109 Thus. n − 1 flops are required to add n numbers. about 9. The characteristic polynomial of ⎡1 ⎤ ⎡1 2 0⎤ AT A = ⎢ 2 ⎥ [1 2 0] = ⎢ 2 4 0⎥ is λ 2 (λ − 5). 1 9. thus the eigenvalues of AT A are λ1 = 5 and (d) 2n3 ≈ 2 ×1012 = 2000 × 109 2000 gigaflops are required. Then the n terms aik bkj must be summed.Chapter 9: Numerical Methods SSM: Elementary Linear Algebra in AB requires 2n − 1 flops. take 70 Section 9.

The only nonzero singular value of A is σ1 = 18 = 3 2. and ⎢ ⎥ ⎣ 5⎦ 2 ⎡ ⎤ ⎡ 1 ⎤ 1 1 ⎡4 6⎤ ⎢ − 5 ⎥ ⎢ − 5 ⎥ Av = . ⎢ 2⎥ ⎣⎢ − 3 ⎦⎥ We must choose the vectors u 2 and u3 so that {u1. The singular values of A are σ1 = 8 A = U ΣV T 2⎤ ⎡ 2 1 6 ⎥⎡ ⎢ 3 1 ⎤ 2 3 2 0⎤ ⎡− 1 ⎢ 2 2⎥ 2 2⎥⎢ ⎥ ⎢ . ⎡ ⎢ A possible choice is u 2 = ⎢ ⎢ ⎢⎣ 1 ⎤ 3⎥ 1 ⎥ 3⎥ 1 ⎥ 3 ⎦⎥ 1 ⎤ 2⎥ orthonormal basis for R3 . and we have ⎡ −2 2 ⎤ ⎡ − 1 ⎢ Av1 = u1 = −1 1⎥ ⎢ σ1 3 2 ⎢ 2 −2 ⎥ ⎢ ⎣ ⎦⎣ 1 1 ⎤ 2⎥ 1 ⎥ 2⎦ and ⎡0⎤ ⎡ 1 0⎤ ⎢ 1 ⎥ 1 ⎢ 0 ⎡ ⎤ u2 = 1 1⎥ ⎢ ⎥ = ⎢ 2 ⎥ . 11. u3} is an ⎢ 6⎥ ⎢ 1 ⎥ ⎢⎣ 6 ⎥⎦ ⎡ 2⎤ ⎢ 3⎥ = ⎢ 13 ⎥ . with corresponding unit ⎡ ⎢ ⎡ 1 0⎤ 1 1 ⎢ ⎡1 ⎤ ⎢ ⎥ = Av = u1 = 1 1 σ1 1 3 ⎢ −1 1⎥ ⎢⎣ 0 ⎥⎦ ⎢ ⎣ ⎦ ⎢− ⎣⎢ ⎡− 1 ⎤ ⎡ 1 ⎤ eigenvectors v1 = ⎢ 12 ⎥ and v 2 = ⎢ 12 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 2⎦ ⎣ 2⎦ respectively. This results in the following ⎢ 3 ⎥ ⎢ ⎥ − 62 ⎥ ⎣⎢ ⎦ singular value decomposition: 7. It depends on the choice of the and σ 2 = 2.SSM: Elementary Linear Algebra Section 9. The singular values of A are σ1 = 3 and σ 2 = 2. This is just one possibility. We have ⎡ 1 ⎡4 6⎤ ⎢ Av1 = ⎢ u1 = 8 ⎣ 0 4 ⎦⎥ ⎢ σ1 ⎣ ⎡ 2 ⎤ = ⎢ 15 ⎥ . u 2 . We choose 2 ⎢ −1 1⎥ ⎣1 ⎦ ⎢ 1 ⎥ ⎣ ⎦ ⎢⎣ 2 ⎥⎦ ⎡ 2 ⎤ ⎢ 6⎥ u3 = ⎢ − 1 ⎥ so that {u1. u2 = = σ 2 2 2 ⎢⎣ 0 4 ⎥⎦ ⎢ 1 ⎥ ⎢ 2 ⎥ ⎣ 5⎦ ⎣ 5⎦ This results in the following singular value decomposition: 2 ⎤ ⎡ 2 − 1 ⎤ ⎡ 1 T 5 5 ⎥ ⎡8 0 ⎤ ⎢ 5 15 ⎥ ⎢ A = U ΣV = 1 2 ⎥ ⎢⎣ 0 2 ⎥⎦ ⎢ − 2 1 ⎥ ⎢ 5⎦ 5⎦ ⎣ 5 ⎣ 5 1 1 ⎤ 5⎥ 2 ⎥ 5⎦ (extended) orthonormal basis for R3 . =⎢ 1 0 − 3 ⎥ 0 0⎥ 1 ⎥ 3 ⎢ ⎢ 1 ⎢ 2 1 ⎥⎢ 0 0 ⎦⎥ ⎣ 2 2⎦ − 62 ⎥ ⎣ ⎢− 3 2 ⎣ ⎦ Note: The singular value decomposition is not unique. with corresponding unit ⎡1 ⎤ ⎡− 2 ⎤ 5⎥ ⎢ eigenvectors v1 = 2 and v 2 = ⎢ 15 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 5⎦ ⎣ 5⎦ respectively. The eigenvalues of ⎡ 1 0⎤ ⎡ 1 1 −1⎤ ⎢ ⎡3 0 ⎤ AT A = ⎢ are λ1 = 3 1 1⎥ = ⎢ ⎥ 0 1 1 ⎢ ⎥ ⎣ ⎦ −1 1 ⎣0 2 ⎥⎦ ⎣ ⎦ and λ 2 = 2. u3} is an orthonormal basis R3 . with corresponding unit ⎡1 ⎤ ⎡0 ⎤ eigenvectors v1 = ⎢ ⎥ and v 2 = ⎢ ⎥ ⎣0 ⎦ ⎣1 ⎦ respectively. This results in the following singular value decomposition: 0 ⎥ and 1 ⎥ 2 ⎥⎦ A = U ΣV T ⎡ 1 ⎢ 3 =⎢ 1 ⎢ 3 ⎢− 1 ⎢⎣ 3 207 0 1 2 1 2 − 2 ⎤ 6⎥⎡ 1 ⎥⎢ 6⎥⎢ 1 ⎥ ⎢⎣ 6 ⎥⎦ 3 0 0 0 ⎤ ⎥ ⎡1 0 ⎤ 2 ⎥ ⎢0 1 ⎥ ⎣ ⎦ 0 ⎥⎦ .5 2⎤ ⎡ 6 ⎥ ⎢ u3 = ⎢ − 2 2 ⎥ . u 2 . The eigenvalues of ⎡ 4 0⎤ ⎡ 4 6⎤ ⎡ 16 24 ⎤ AT A = ⎢ = are λ1 = 64 ⎣ 6 4⎥⎦ ⎢⎣ 0 4⎥⎦ ⎢⎣ 24 52 ⎥⎦ and λ 2 = 4. We have 9. The eigenvalues of ⎡ −2 2 ⎤ ⎡ −2 −1 2 ⎤ ⎢ ⎡ 9 −9 ⎤ AT A = ⎢ are −1 1⎥ = ⎣ 2 1 −2 ⎦⎥ ⎢ 2 −2 ⎥ ⎣⎢ −9 9⎦⎥ ⎣ ⎦ λ1 = 18 and λ 2 = 0.

(f) False. A has the singular value decomposition 2 ⎤ ⎡ 1 0 6⎥⎡ 3 ⎢ 3 0 ⎤ ⎥ ⎡1 0 ⎤ 1 1 1 ⎥⎢ ⎢ − A= 2 ⎥ ⎢0 1 ⎥ 2 6⎥⎢ 0 ⎢ 3 ⎣ ⎦ 0 ⎥⎦ 1 1 ⎥ ⎢⎣ 0 ⎢− 1 2 6 ⎦⎥ ⎣⎢ 3 where the lines indicate the relevant blocks.6 (b) True 1. 2 2⎦ ⎣ ⎢ 2⎥ ⎢⎣ − 3 ⎥⎦ (c) False.100 numbers. L=⎢ ⎣ −2 1⎥⎦ ⎣ −2 1⎥⎦ ⎢⎣ 0 2 ⎥⎦ 3. V1 has size n × k so that V1T has size k × n.6 Section 9. The reduced singular value decomposition of A ⎡ 1 ⎤ ⎡0⎤ ⎢ 3⎥ ⎢ 1 ⎥ is 3 ⎢ 1 ⎥ [1 0] + 2 ⎢ 2 ⎥ [0 1]. since AT A is a symmetric n × n matrix. ⎡2 4 6⎤ ⎡1 2 3⎤ ⎡ 1 2 3⎤ ⎢ 1 4 7 ⎥ → ⎢1 4 7 ⎥ → ⎢0 2 4⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 1 3 7⎦ ⎣1 3 7 ⎦ ⎣ 1 3 7⎦ ⎡ 1 2 3⎤ ⎡ 1 2 3⎤ ⎡ 1 2 → ⎢ 0 2 4 ⎥ → ⎢0 1 2 ⎥ → ⎢ 0 1 ⎢ ⎥ ⎢ ⎥ ⎢ ⎣0 1 4⎦ ⎣0 1 4 ⎦ ⎣ 0 0 3. if A is an m × n matrix.5. ⎡ −6 2 ⎤ ⎡ −3 1⎤ ⎡ −3 1⎤ ⎢⎣ 6 0 ⎥⎦ → ⎢⎣ 6 0⎥⎦ → ⎢⎣ 0 2 ⎥⎦ = U 1 The multipliers used were and 2. so 2 ⎡ 2 0⎤ ⎡ 2 0 ⎤ ⎡ −3 1⎤ and A = ⎢ . then AT is an n × m matrix.6 (a) True Exercise Set 9. and AT A is an n × n matrix. while A has 200(500) = 100. (g) True True/False 9. 9. A has the singular value decomposition 2 ⎤ ⎡ 2 1 6 ⎥⎡ 1 ⎤ ⎢ 3 2 3 2 0⎤ ⎡− 1 ⎢ 2 2⎥ ⎢ 2 2⎥⎢ ⎥ A=⎢ 1 0 − 3 ⎥ 0 0⎥ ⎢ 1 1 ⎥ 3 ⎢ ⎢ 2 1 ⎥⎢ 0 0 ⎦⎥ ⎣⎢ 2 2 ⎦⎥ − 62 ⎥ ⎣ ⎢− 3 2 ⎣ ⎦ where the lines show the relevant blocks. the eigenvalues of AT A are the squares of the singular values of A. (d) False. (b) True 7. From Exercise 11 in Section 9. 3 ⎢ ⎥ ⎢ 1 ⎥ ⎢− 1 ⎥ ⎢⎣ 2 ⎥⎦ ⎣⎢ 3 ⎦⎥ (c) False.000 has entries. −1. Reduce A to upper triangular form. −1. The reduced singular value expansion of A is ⎡ 2⎤ ⎢ 3⎥ 1 ⎤ 3 2 ⎢ 13 ⎥ ⎡ − 1 . Chapter 9 Supplementary Exercises 1. AT A may have eigenvalues that are 0. Reduce A to upper triangular form. Thus the reduced singular value decomposition of A is ⎡ 1 0⎤ ⎢ 3 ⎥ 0 ⎤ ⎡1 0 ⎤ 1 1 ⎥⎡ 3 T ⎢ . From Exercise 9 in Section 9. (e) True. 2 2 .5. A rank 100 approximation of A requires storage space for 100(200 + 500 + 1) = 70.Chapter 9: Numerical Methods SSM: Elementary Linear Algebra 5. −1. A would have to be symmetric to be orthogonally diagonalizable. 2 2⎦ ⎣ ⎢ 2⎥ ⎢⎣ − 3 ⎥⎦ True/False 9.5 (a) False. Thus the reduced singular value decomposition of A is ⎡ 2⎤ ⎢ 3⎥ 1 ⎤ A = U1Σ1V1T = ⎢ 13 ⎥ [3 2 ] ⎡ − 1 . . A = U1Σ1V1 = ⎥ 2⎥⎢ 0 ⎢ 3 2 ⎦ ⎢⎣ 0 1 ⎥⎦ ⎣ 1 1 ⎢− ⎥ 2 ⎦⎥ ⎣⎢ 3 3⎤ 2⎥ ⎥ 2⎦ ⎡ 1 2 3⎤ → ⎢0 1 2⎥ = U ⎢ ⎥ ⎣ 0 0 1⎦ The multipliers used were 208 1 1 .

⎢ ⎥ ⎢− 1 ⎥ ⎣ 2⎦ ⎣ 2⎦ respectively. 1 ⎥ 2 ⎦⎥ u3 so that {u1. ⎡ 2 1 ⎤ ⎡1 ⎤ ⎡ 2 ⎤ = (b) Ax0 = ⎢ ⎣ 1 2 ⎥⎦ ⎢⎣0 ⎥⎦ ⎢⎣1 ⎥⎦ Ax 0 1 ⎡ 2⎤ ⎡ 0.7042 ⎥⎦ 1 ⎤ 2⎥ 0⎥ .6983⎥⎦ x5 = λ4 the rate of convergence is likely to be quite slow.9918⎥⎦ Ax 4 ⎡0. (a) The characteristic equation of A is 7. since λ 2 − 4λ + 3 = (λ − 3)(λ − 1) = 0 so the dominant eigenvalue of A is λ1 = 3. v = ⎢ 12 ⎥ ≈ ⎢ ⎢ ⎥ ⎣ 0.7071⎥⎦ ⎣ 2⎦ the ratio Ax1 ⎡ 0. The only nonzero singular value A is σ1 = 4 = 2. ⎥ 0 1⎦ x5 = ⎡ 1 ⎤ x5 ≈ ⎢ as compared to the exact ⎣0.7328⎤ ≈⎢ Ax 2 ⎣0.7071⎥⎦ ⎡2⎤ (c) Ax0 = ⎢ ⎥ ⎣1 ⎦ Ax 0 ⎡1 ⎤ = x1 = max( Ax0 ) ⎢⎣ 0.7042⎥⎦ ⎡0.SSM: Elementary Linear Algebra ⎡2 1 so L = ⎢ 1 ⎢ 2 ⎣1 ⎡2 0 0⎤ ⎡ 1 A = ⎢ 1 2 0 ⎥ ⎢0 ⎢ ⎥⎢ ⎣ 1 1 2 ⎦ ⎣0 and Chapter 9 Supplementary Exercises 0 0⎤ 2 0 ⎥ and ⎥ 1 2⎦ 2 3⎤ 1 2⎥ . u 2 .4472 ⎥⎦ x2 = Ax 4 ⎡ 1 ⎤ ≈ max( Ax4 ) ⎢⎣ 0. A ⎡ 1 ⎤ ⎡0 ⎤ ⎢ 2⎥ possible choice is u 2 = ⎢1 ⎥ and u3 = ⎢ 0 ⎥ .8944 ⎤ = ≈ x1 = Ax 0 5 ⎢⎣1 ⎥⎦ ⎢⎣ 0.6805⎥⎦ λ1 = 8.7809 ⎤ ≈ Ax1 ⎢⎣0.7158⎤ x4 = ≈⎢ Ax 3 ⎣ 0. v≈⎢ ⎣0.7100⎤ x5 ≈ ⎢ as compared to ⎣0.7071⎤ . with corresponding positive unit eigenvector ⎡ 1 ⎤ ⎡ 0.1 = 1. 8 ⎡2 2⎤ are λ1 = 4 9. The eigenvalues of AT A = ⎢ ⎣ 2 2 ⎥⎦ and λ 2 = 0 with corresponding unit ⎡− 1 ⎤ ⎡− 1 ⎤ 2⎥ ⎢ eigenvectors v1 = and v 2 = ⎢ 12 ⎥ .8⎦ A = U ΣV T 1 ⎤ ⎡− 1 0 2 ⎥ ⎡ 2 0⎤ ⎡ − ⎢ 2 =⎢ 0 1 0⎥ ⎢ 0 0 ⎥ ⎢ ⎢ ⎢ 1 1 ⎥ ⎢ 0 0 ⎦⎥ − ⎣ ⎢⎣ − 2 0 − 2 ⎥⎦ ⎣ Ax 2 ⎡ 1 ⎤ ≈ max( Ax2 ) ⎢⎣ 0. u3} is an orthonormal basis for R3 .6247 ⎥⎦ x3 = Ax 2 ⎡0.0125 is very close to 1. The Rayleigh quotients will converge to the dominant eigenvalue λ 4 = −8. ⎣1⎦ 5.7071⎤ .7100 ⎤ ≈⎢ Ax 4 ⎣0.9918⎥⎦ ⎡1⎤ eigenvector v = ⎢ ⎥ . However.9756 ⎥⎦ x3 = 209 1 2 1 2 − 1 ⎤ 2⎥ 1 ⎥ 2⎦ .5⎥⎦ Ax1 ⎡1 ⎤ =⎢ ⎥ x2 = max( Ax1 ) ⎣ 0. ⎢ ⎥ ⎢ 1 ⎥ − ⎣0 ⎦ ⎣⎢ 2 ⎦⎥ This results in the following singular value decomposition: ⎡0.9286 ⎥⎦ Ax 3 ⎡ 1 ⎤ ≈⎢ x4 = max( Ax3 ) ⎣ 0.1. and we have ⎡− ⎡ 1 1⎤ ⎡ − 1 ⎤ ⎢ 1⎢ u1 = Av = 0 0 ⎥ ⎢ 2 ⎥ = ⎢ σ1 1 2 ⎢ 1 1⎥ ⎢ − 1 ⎥ ⎢ ⎣ ⎦ ⎣ 2 ⎦ ⎢− ⎣ We must choose the vectors u 2 and 1 Ax 3 ⎡0.

A has rank 2. thus U1 = [u1 u 2 ] and ⎡ vT ⎤ V1T = ⎢ 1T ⎥ and the reduced singular value ⎣⎢ v 2 ⎦⎥ decomposition of A is A = U1Σ1V1T 1⎤ ⎡1 2⎥ ⎢2 ⎢ 12 − 12 ⎥ ⎡ 24 0 ⎤ ⎡ 23 = ⎢1 ⎥ ⎢2 1⎥ ⎢ ⎢ 2 − 2 ⎥ ⎣ 0 12 ⎦ ⎢⎣ 3 ⎢1 1⎥ ⎣⎢ 2 2 ⎦⎥ − 13 2 3 2⎤ 3⎥ − 13 ⎥ ⎦ 210 .Chapter 9: Numerical Methods SSM: Elementary Linear Algebra 11.