UNIT I

MATRICES

Problem 1. Find the eigen values and eigen vectors of the matrix
2 2 3
A 2
1 6
1 2 0
Solution:
The characteristic equation is | A - I | = 0.

i.e.,

2-
2
2
1-
1

3
6

0-

i.e., (-2 - ) [-(1 - ) -12] - 2[-2 - 6] -3[-4 + 1 - ] = 0

i.e., (-2 - ) [2 - -12] + 4 + 12 + 9 + 3 = 0
i.e., 3 + 2 - 21 - 45 = 0
3
2
Now, (-3) + (-3) - 21(-3) - 45 = -27 + 9 + 63 45 = 0
-3 is a root of equation (1).
Dividing 3 + 2 - 21 - 45 by + 3

3 1 1 21 45
0 3 6
45
1 2 15
0
Remaining roots are given by
2 - 2 - 15 = 0
i.e.,
( + 3) ( - 5) = 0
i.e.,
= -3, 5.
The eigen values are -3, -3, 5
2 2 - 3 x1 0
1 - - 6 x 2 0
The eigen vectors of A are given by 2
- 1
- 2 - x3 0
Case 1 = -3
2
- 3
2 3
1 2 - 3

1 3 - 6 ~ 2 4 - 6
Now 2
- 1
- 1 - 2 3
-2
3

1 2 - 3
~ 0 0 0
0 0 0
x1 + 2x2 - 3x3 = 0

(1)

Matrices

Put
x2 = k1, x3 = k2
Then x1 = 3k2 - 2k1

3k 2 2k 1

The general eigen vectors corresponding to = -3 is

k1

k 2

3
When k1 = 0, k2 = 1, we get the eigen vector 0
1
2
When k1 = 1, k2 = 0, we get the eigen vector 1
0
3
2

Hence the two eigen vectors corresponding to = -3 are 0 and 1 .

0
1
These two eigen vectors corresponding to = -3 are linearly independent.
Case 2 = 5
2 5 2 - 3 7 2 - 3
2
1 - 5 - 6 ~ 2 - 4 - 6

- 1
- 2 - 5 - 1 - 2 - 5
1 2 5
~ 0 8 16
0
0
0
-x1 - 2x2 - 5x3 = 0
-8x2 - 16x3 = 0
A solution is x3 = 1, x2 = -2, x1 = -1
1
Eigen vector corresponding to = 5 is 2 .
1
1 1 2
Problem 2. Find the characteristic equation of 2 1 3 and verify Cayley 3 2 3
Hamilton Theorem. Hence find the inverse of the matrix.

Matrices

1 1 2
Solution: Let A 2 1 3 Characteristic eqn. of A is
3 2 3
3
2
1 1 3 9 9 1 26 0
i.e 3 2 19 26 0
By Cayley-Hamilton theorem A3 A2 19 A 26I 0 .
Verification:

2 7
1 1 2 1 1 2 9

A A. A 2 1 3 2 1 3 5
9 10
3 2 3 3 2 3 10 7 21

2 7 1 1 2 16 21 45
9

3
2
A A .A 5
9 10 2 1 3 43 16 67
10 7 21 3 2 3 67
45 104

Substituting in the characteristic equation
2 7 19 19 38 26 0 0 0 0 0
16 21 45 9

9 10 38 19
57 0 26 0 0 0 0
43 16 67 5
67 45 104 10 7 21 57 38 57 0 0 26 0 0 0

Hence verified.
Now to find the inverse of the matrix A, premultiply the characteristic equation by A1
A2 A 19 I 26 A1 0
1
A1 19 I A A2
26
19 0 0 1 1 2 9
2 7
9 5 5
1


1

0 19 0 2 1 3 5
9 10 3 9 7

26
26

0 0 19 3 2 3 10 7 21
7 5 1
2

1 0 3
Problem 3. Given A 2 1 1 , use Cayley-Hamilton Theorem to find the inverse of A
1 1 1
and also find A4

Solution:
The characteristic equation of A is
1
0
2 1
1

3
1 0
1

i.e., (1-) [(1 - ) (1 - ) -1] + 3[-2 - (1 - )] = 0

Matrices

i.e., (1 - )3 - (1 - ) 6 -3 + 3 = 0
i.e., 1 - 3 + 32 - 3 1 + - 9 + 3 = 0
i.e., -3 + 32 + - 9 = 0
i.e., 3 - 32 - + 9 = 0
By Cayley-Hamilton theorem,
A3-3A2 A + 9I = 0
-1
-1
To find A , multiplying by A , A2-3A - I + 9A-1 = 0

A-1 =

1
[-A2 + 3A + I]
9

3 1 0 3 4 3 6
1 0

A 2 1 1 2 1 1 3 2 4
1 1 1 1 1 1 0 2 5
9 1 0 0
4 3 6 3 0
1

1
A 3 2 4 6 3 3 0 1 0
9
0
2 5 3 3 3 0 0 1
2

3
0 3
1
3 2 7
9
3 1 1

To find A4:
We have
A3- 3A2 A + 9I = 0
i.e.,
A3 = 3A2 + A - 9I
(1)
Multiplying (1) by A, we get,
A4 = 3A3 + A2-9A

= 3(3A2 + A - 9I) + A2 - 9A
using (1)
= 10A2 - 6A - 27I
4 3 6 1 0 3
1 0 0

10 3 2 4 62 1 1 27 0 1 0
0 2 5 1 1 1
0 0 1
7 30 42
18 13 46
6 14 17

0 0 2
Problem 4. . If A 2 1 0 express A6 25 A2 122 A as a single matrix
1 1 3

Solution: To avoid higher powers of A like A6 we use Cayley Hamilton Theorem.

Characteristic equation is 3 4 2 5 2 0
By Cayley Hamilton Theorem A3 4 A2 5 A 2I 0
To find A6 25 A2 122 A we will express this in terms of smaller powers of A using the
characteristics equation. We know that (Divisor) X (Quotient) + Remainder = Dividend
Assuming A3 4 A2 5 A 2I as the divisor we get,

Matrices

A3 4 A2 11A 22I
A3 4 A2 5 A 2I

A6 0 A5 0 A4 25 A2 122 A 0 I
A6 4 A5 5 A4 2 A3
4 A5 5 A4 2 A3 25 A2 122 A
4 A5 16 A4 20 A3 8 A2
11A4 22 A3 33 A2 122 A
11A4 44 A3 55 A2 22 A
22 A3 88 A2 100 A
22 A3 88 A2 110 A 44 I
10 A 44I

A6 25 A2 122 A A3 4 A2 5 A 2I A3 4 A2 11A 22I 10 A 44I

But A3 4 A2 5 A 2I 0
A6 25 A2 122 A 0 10 A 44 I

10 A 44 I
0
0 20 44 0 0

20 10
0 0 44 0
10 10 20 0 0 44
0
44

20 54
10 10

44 0

20 54
10 10

20

0
74
20

0
74

Problem 5. If i are the eigen values of the matrix A, then prove that

i k i are the eigen values of kA where k is a nonzero scalar.

ii. im are the eigen value of Am and
1
iii.
are the eigen values of A1 .
i
Solution: Let i be the eigen values of matrix A and Xi be the corresponding eigen
vectors. Then by defn: AXi iXi......( I ) ( i.e by defn. of eigen vectors)
i. Premultiply ( I ) with the scalar k. Then
k AXi k iXi
i.e. kA X i k i Xi
k i are the eigen values of kA (comparing with ( I ) i.e by defn.)

Matrices

ii. Premultiply ( I ) with A, then

A AXi A iXi
i.e. A2 X i i AXi
i i Xi from (I)
i Xi
2

1y we can prove that A3 Xi i Xi and so on Am Xi i Xi

3

i m are the eigen values of the Am (comparing with ( I ) i.e. by defn.)

iii. Premultiply ( I ) with A1 , then

A1 AXi A1 iXi

i.e. A1 A Xi i A1 Xi
i.e. IXi i A1 Xi
i.e. A1 Xi

1
Xi
i

1
are the eigen values of A1 (comparing with ( I ) ).
i
2 0 1
Problem 6. Find the characteristic vectors of 0 2 0 and verify that they are
1 0 2

mutually orthogonal.

2 0 1

Solution: A = 0 2 0 Characteristic equation is 3 6 2 11 6 0

1 0 2

Solving: 1, 2,3
Consider the matrix equation A I X 0
Case (i) when 1;
1x1 0 x2 1x3 0 1
1 0 1 x1 0

0 1 0 x2 0 i.e. 0 x1 1x2 0 x3 0 2 equation (1) & (3) are identical.
1 0 1 x 0
1x1 0 x2 1x3 0 3

3
Solving (1) and (2) using the rule of cross multiplication
1
x3
x1
x2
x1 x2 x3

i.e. X 1 0
0 1 0 1 0 1
1 0
1
1

Case (ii) when 2;

Matrices

0 x1 0 x2 1x3 0
1 0 1 x1 0

0 1 0 x2 0 i.e. 0 x1 0 x2 0 x3 0
1 0 1 x 0
1x1 0 x2 0 x3 0

x3 0
i.e. x2 is arbitrary say k
x1 0

0 0

X 2 k i.e 1 .
0 0

Case (ii) when 3;
x1 0 x2 1x3 0
1 0 1 x1 0

Solving (1) and (2)
0 1 0 x2 0 i.e. 0 x1 1x2 0 x3 0
1 0 1 x 0
1x1 0 x2 1x3 0

3
1
x1 x2 x3

X3 0
1 0
1
1

Thus the eigen values are 1,2,3 and the correspondent eigen vectors are
1 0
1


T
0 , 1 and 0 . To check orthogonallity, X1 X 2 0
1 0
1


X 2T X 3 0
X 1T X 3 0
X1, X 2 , X 3
are mutually orthogonal.

6 6 5
Problem 7. Find the latent vectors of 14 13 10
7 6 4

Solution: Characteristic equation is 1 0 1, 1, 1

3

When 1 (repeated 3 times) we have to find 3 corresponding latent vectors.

7 x1 6 x2 5 x3 0
7 6 5 x1 0

14 12 10 x2 0 i.e. 14 x1 12 x2 10 x3 0 All three equation are identical
7 6 5 x 0
7 x1 6 x2 5 x3 0

3
.i.e. we get only one equation, but we have to find three vectors that are linearly
independent.
0
x2 x3

Assume x1 0 6 x2 5 x3 0 i.e. 6 x2 5 x3 i.e. X 1 5
5
6
6

Matrices

5
x1 x3

Assume x2 0 7 x2 5 x3 0 i.e. 7 x1 5 x3i.e.. X 2 0
5 7
7

6
x1 x2

And assume x2 0 7 x2 6 x3 0 i.e. 7 x1 6 x2 0i.e.. X 3 7
6 7
0

X1, X2 and X3 are linearly independent.

Problem 8.

1 1 1
Find the eigen vectors of the matrix A 0 2 1
4 4 3

Solution:

1 -
The characteristic equation of A is 0
4

1
2-
4

1
1 0
3 -

i.e., (1 - ) [(2 - ) (3 - ) - 4] -1[0 + 4] +1[0 + 4(2 - )]

i.e., (1 - )(2 - 5 + 6 - 4) 4 + 8 - 4
i.e., (1 - )(2 - 5 + 2) + 4 - 4
i.e., (1 - )(2 - 5 + 2 + 4)
i.e., ( -1)(2 - 5 + 6)
i.e., ( -1)( - 2)( - 3)
The eigen values of A are = 1, 2, 3.

=0
=0
=0
=0
=0
=0

1
1 x 1 0
1 -

1 x 2 0
The eigen vectors are given by 0 2 -
4
4 3 - x 3 0
=1
0 1 1 4 4 2
0 1 1 ~ 0 1 1

4 4 2 0 0 0
-4x1 + 4x2 + 2x3 = 0
x2 + x3 = 0
A solution is, x3 = 2, x2 = -2, x1= -1
Case 1

1
Eigen vector X1 = 2
2

Matrices

=2
1 1 1 1 1 1
0 0 1 ~ 0 0 1

4 4 1 0 0 0
-x1 + x2 + x3 = 0
x3 = 0
A solution is, x3 = 0, x2 = 1, x1 = 1
Case 2

1
Eigen vector X2 = 1
0

=3
2 1 1 2 1 1
0 1 1 ~ 0 1 1

4 4 0 0
0 0
-2x1 + x2 + x3 = 0
-x2 + x3 = 0
A solution is, x3 = 1, x2 = 1, x1 = 1
Case 3

1
Eigen vector X3 = 1
1

2 2 0
Problem 9. Diagonalise the matrix 2 5 0 using orthogonal transformation.
0 0 3

3
2
Solution: Characteristic equation is 10 27 18 0
Solving we get the eigen value as 1,3,6
2
0
1



When 1, X 1 1 ; When 3, X 2 0 ; When 6, X 3 2
0
0
1



2
1
5 0
5

1

2
Normalizing each vector, we get
, 0 and
5
5

0
0

Matrices

2
0
5

Normalized Modal Matrix, N 1

0
5

1
0

Then by the orthogonal transformation,

2
1
2
0

5
5
5

2 2 0

N AN 0
0
1 2 5 0 0

2
1
0 0 0 3 1
5
5

N AN D 1 , 2 , 3

10

2
5
5

2
. N N
0

1
0
5

0
0
1

0
1,

2
0
5

2 . On simplifying, we get
5
0

1 0 0

which is diagonal matrix with eigen values along the

D 1,3, 6 0 3 0
0 0 6

diagonal (in order).

6 2 2
Problem 10. Reduce 2 3 1 to a diagonal matrix by orthogonal reduction.
2 1 3

3
Solution: Characteristic equation is 12 2 36 32 0 8, 2, 2
When 8
2 2 2 x1 0

2 5 1 x2 0
2 1 5 x 0

3
2 x1 2 x2 2 x3 0
i.e
2 x1 5x2 1x3 0
2 x1 1x2 5x3 0
2
x1 x2 x3

X 1 1
Solving any two equations
2 1 1
1

When 2 (repeated twice)
4 2 2 x1 0

2 1 1 x2 0 i.e 2 x1 2 x2 2 x3 0 . All the equations are identical.
2 1 1 x 0

Matrices

11

0
x2 x3

To get one of the vectors, assume x1 0 x2 x3 0 i.e. X 2 1
1
1
1

a

T
X1 X 2 0 . Therefore X 1 and X 2 are orthogonal. Now assume X 3 b to be mutually
c

orthogonal with X1 and X2.

X 1T X 3 0 i.e. 2 1 1 b 0 i.e.2a b c 0
c
b c

a

i.e
a
2 2 2

and X 2T X 3 0 i.e. 0 1 1 b 0 i.e.0a b c 0

1

X3 1 .
1

After normalizing these 3 mutually orthogonal vectors, we get the normalized Modal
2
1
0

6
3

1
1
Matrix N 1
6
2
3

1
1
1

6
2
3

Diagonalizing we get
2

T
D N AN 0

on simplifying we get

6
6 6 2 2
6

1
1 2 3 1 1

2
2
6

1
3

1
1
1
3
3
6

D D 1 , 2 , 3
1

8 0

0 2
0 0

D 8,

0
2

2, 2

1
3
1
3
1

6
1
2
1
3

Matrices

12

3 1 1
Problem 11. Diagonalise the matrix A 1 3 -1
1 -1 3
Solution:

1
3- 1

The characteristic equation of A is 1 3- -1 0

-1 3-
1
i.e., (-1)(2 - 8 + 16) = 0
The eigen values of A are = 1, 4, 4.

1 x1 0
3- 1

The eigen vectors are given by 1 3- -1 x 2 0

1
-1 3- x 3 0
Case 1

=1

1
Eigen vector X1 = 1
1
Case 2 = 4
0
Eigen vector X2 = 1
1
a

Now assume X 3 b to be mutually orthogonal with X1 and X2.
c

T
X 1 X 3 0 i.e. a b c 0
a b c
i.e
T
2 1 1
and X 2 X 3 0 i.e. b c 0
2

X3 1 .
1

1 0 2
Hence the modal matrix M 1 1 1
1 1 1

Matrices

13

1
0

1
The Normalized Modal Matrix is N 1
3
2

1
1

3
2

Diagonalizing, we get
1
1

3
3

1
D N T AN 0
2

2
1

6
6

1
6
1
6

0
3 3 1 1
3

1 1 3 1 1
1

2
3
2

1
3

1
1
1

6
3
2

1
6
1
6

1 0 0
0 4 0 = D(1, 4, 4)
0 0 4

Problem 12. Reduce the Quadratic From 10 x12 2 x22 5x32 6 x2 x3 10 x3 x1 4 x1x2 into
canonical form by orthogonal reduction. Hence find the nature, rank, index and the
signature of the Q.F. Find also a nonzero set of values of X which will make the Q.F.
vanish.

10 2 5

Solution: Matrix of the given Q.F. is A 2 2 3 , which is a real and symmetric

5 3 5

3
2
matrix. The characteristic equation is 17 42 0
Solving, we get 0, 3, 14
1
1
3



When 0, X 1 5 ; When 3, X 2 1 ; When 14, X 3 1
2
4
1



and X1 , X 2 , X 3 are mutually orthogonal since X1 , X 2 0, X 2 X 3 0 andX 3 X1 0
Normalizing these vectors we get the normalized model matrix
1

3
1

42
3
14

1
1
N 5

42
3
14

1
2

42
3
14

Matrices

14

Diagonalising we get D N T AN
D 12 , 3 in order

D 0, 3, 14
0 0 0

D 0 3 0 (i.e. the eigen values in order along the principal

0 0 14

i.e

diagonal).
Now to reduce the Q.F to C.F (.i.e Canonical form)

y1

Consider the orthogonal transformation X = NY where Y y2
y
3
Then the Q.F. X T AX becomes NY A NY Y T N T AN Y
T

= Y T DY since
0

y1 y2 y3 0
0

N T AN D
0 0 y1

3 0 y2

0 14
y3

0 y1 3 y2 14 y3
2

Thus = 0 y1 3 y2 14 y3 is the Canonical form of the given Q.F. And the equations of
this transformation are got from X= NY.
1

3
1

42
3
14 y1
x1

5
1
1
x

NY

y2
2

42
3
14

x
3
4
y3
1
2

42
3
14

1
1
3
y1
y2
y3
42
3
14
5
1
3
x2
y1
y2
y3
42
3
14
4
1
3
x3
y1
y2
y3
42
3
14

x1

To get the non-zero set of values of x which make the Q.F zero we assume values
for y1 , y2 and y3 such that the C.F. vanishes.

Matrices

15

i.e 0 y1 3 y2 14 y3 will vanish if y2 0, y3 0and y1 is any arbitrary value (for

simplicity sake, assume y1 as the denominator of the coeff. of y1 in the equations) let
2

y1 42
1
1
3
x1
42
(0)
0
42
3
14
i.e. x1 1 0 0 1

III 1 y x2 5 0 0 5
and x3 4 0 0 4

Thus the set of values of x i.e 1, 5, 4 will reduce the given Q.F. to zero.
To find the rank, index, signature and nature using canonical form:
2
2
2
C.F. is 0 y1 3 y2 14 y3
rank is 2 (no. of terms in C.F)
Index is 2 (no. of positive terms)
Signature of Q.F. = ( no. of positive terms) (no. of negative terms) = 2
Nature of the Q.F. is positive semi definite.

Problem 13. Reduce the Q.F. 2 xy 2 yz 2 zx into a form of sum of squares. Find the
rank, index and signature of it. Find also the nature of the Q.F.
0 1 1

Solution: Matrix of the Q.F. is A 1 0 1

1 1 0

3
Characteristic equation is 3 2 0 solving 2, 1, 1
1

When 2, X 1 1
1

When 1 (repeated twice) we get identical equations as x1 x2 x3 0
x
x
x1 0 x2 x3 0 i.e. x2 x3 i.e. 2 3
1 1
0
Assume

X 2 1
1

which is orthogonal with X 1.

a

Now to find X 3 orthogonal with both X1 and X 2 assume X 3 b
c

Matrices

16
if X 2T X 3 0, a b c 0

if X 2T X 3 0, 0a b c 0

a b
c

2 1 1
2
2


X 3 1 i.e. 1
1
1


which is orthogonal with X1 and X 2 .
i.e.

1
0
3

3
2
6

1
1
Normalising these vectors we get N 1
and D N AN
3
2
6

1
2

3
2
6

2 0 0

= D 1 , 2 , 3 0 1 0 .Consider the orthonormal transformation X = NY

0 0 1

such that Q.F.is reduced to C.F.

The Q.F. is reduced as
T
X T AX NY A NY
Y T N T AN Y

Y T DY
2 0 0 y1

y1 , y2 , y3 , 0 1 0 y2
0 0 1 y

3
The C.F. is 2y12 y2 2 y32
rank of Q.F.is = no. of terms in C.F=3
index of Q.F. = no. of positive terms in C.F. = 1
signature of Q.F. = ( no. of positive terms) (no. of negative terms)
= 1-2 = -1
Nature of the Q.F. is indefinite.

Problem 14. Reduce the quadratic form

8x12 7 x22 3x32 12x1 x2 4 x1 x3 8x2 x3 to the

canonical form by an orthogonal transformation. Find also the rank, index, signature and
the nature of the quadratic form.

Matrices

17

Solution:

8 6 2
The matrix of the quadratic form is A 6 7 4
2 4 3
The eigen values of this matrix are 0, 3 and 15 and the corresponding eigen vectors are
1
2
2

X 1 2 ,
X2 1 ,
X 3 2 , which are mutually orthogonal.
2
2
1
2/3
1/3 2/3

The normalized modal matrix is N 2/3 1/3 2/3

2/3 2/3 1/3
0 0 0
and N AN = D 0 3 0
0 0 15
T

Now the orthogonal transformation X = NY will reduce the given quadratic form to the
canonical form 0y12 3y22 15y32 .
Also rank = 2, index = 2, signature = 2. The quadratic form is positive semi definite.

Problem 15.

Find the orthogonal transformation which reduces the quadratic form

2 x1 x2 2 x2 x3 2 x1 x3 into the canonical form. Determine the rank, index,

signature and the nature of the quadratic form.
2 x12

2 x22

2 x32

Solution:
2 1 1
The matrix of the quadratic form is A 1 2 1
1 1 2
2 - -1
1

The characteristic equation of A is

-1

2-

-1

-1 0
2-

Expanding - 6 + 9 - 4 = 0
= 1 is a root
Dividing 3 - 62 + 9 - 4 by -1,
3

1 6
0

1 5

The remaining roots are given by

2-5 + 4 = 0
2 - 5 + 4 = ( - 1) ( - 4) = 0
i.e., = 1, 4

Matrices

18

The eigen values of A are = 4, 1, 1

Case 1

=4

1 x 1 0
2 - 4 - 1

The eigen vectors are given by - 1 2 - 4 - 1 x 2 0

1
- 1 2 - 4 x 3 0
2 1 1 1 - 1 - 2
1 2 1 ~ 0 - 3 - 3

1 1 2 0 0 0
x1 - x2 - 2x3 = 0
-3x2 - 3x3 = 0
A solution is x3 = 1, x2 = -1, x1 = 1.
1
The corresponding eigen vector is X1 = 1
1
Case 2

=1

1 x1 0
2 - 1 - 1

The eigen vectors are given by - 1 2 - 1 - 1 x 2 0

1
- 1 2 - 1 x3 0

1 - 1 1 1 1 1
- 1 1 - 1 ~ 0 0 0

1 - 1 1 0 0 0
x1- x2 + x3 = 0
Put x3 = 0. We get x1 = x2 = 1. Let x1 = x2 = 1
1
The eigen vector corresponding to = 1 is X2 = 1
0

X1 and X2 are orthogonal as X1T X 2 = 10 + (-1) 1 + 11 = 0.

a
To find another vector X3 = b corresponding to =1 such that it is orthogonal to both
c

X1 and X2 and satisfies x1- x2 + x3 = 0

i.e.,
X1.X3 = 0,
X2.X3 = 0 and a b + c = 0
i.e.,
1.a -1.b + 1.c = 0, 1.a + 1.b + 0.c = 0 and a b + c = 0.
i.e.,
a b + c = 0 and
a+b=0
i.e.,
a = -b and
c = 2b
Put b =1, so that a = -1, c = 2

Matrices

19
1
X 3 1
2

1 1 1
The modal matrix is 1 1 1
1 0 2
1/ 3 1/ 2

Hence the normalized modal matrix is N 1 / 3 1 / 2

1/ 3
0

1/
1/

2/

The required orthogonal transformation is X = NY will reduce the given quadratic

form to the canonical form.
C.F= 4 y12 y 22 y32
Rank of the quadratic form = 3, index = 3, signature = 3. The quadratic form is positive
definite.