Calculus and Matrix Algebra

Characteristic equation
Let A be a given square matrix and λ be a scalar. Then the equation det(A–λI) = 0, where I is an
identity matrix (or) |A–λI | = 0 is called the characteristic equation of the given matrix A.
Simply, |A–λI | = 0 is called the characteristic equation of the given matrix A.

Problems
41 10
| − λ |=0
1. Find the characteristic equation of the given matrix 3 2 0 1
( )( )
Solution:

Let
(10 01 ) .
The characteristic equation of A is |A–λI | = 0.

|( 43 12 )−(0λ 0λ )|=0 |4−λ 1 |=0

(where 3 2−λ is an identity matrix)
( 4−λ )( 2−λ ) −3=0
8−4 λ−2 λ+ λ2 −3= 0
λ2 −6 λ +5=0 (by scalar multiplication)

2
λ −6λ+5=0 (by matrix subtraction)
(5 – λ)(2 – λ) – 4 = 0
10 – 5λ – 2λ + λ2 – 4 = 0
λ2 – 7λ + 6 = 0
The characteristic equation is λ2 – 7λ + 6 = 0.

( 3 2−λ)(x ) (0)
4−λ 1 x 0
= 1 ( 1)
2. Find the characteristic equation of the given matrix 2

Solution:

Let
( )( ) ( ) .
4−1 1 x 1 0
=
3 2−1 x 2 0

The characteristic equation of A is |A–λI | = 0.

(33 11 )( xx )=(00 )
1

3 x1 + x 2=0
3 x1 + x 2=0
x 2=− 3 x 1

(−3k k )
(1 – λ)(4 – λ) – (– 6) = 0
4 – λ – 4λ + λ2 + 6 = 0
λ2 – 5λ + 10 = 0
The characteristic equation is λ2 – 5λ + 10 = 0.

Easy method to find the characteristic equation of 3×3 matrix:

Let A be a square matrix of order 3. That is,

( )( ) ( )
−1 1 x 1 = 0
3 −3 x 2 0
The characteristic equation of A (3×3) matrix is λ3 – D1λ2 + D2λ – D3 = 0.
where D1 = Sum of the principal (or) leading diagonal elements (i.e., Trace of A)
= a11+ a22+ a33
D2 = Sum of the minors of the principal diagonal elements
= minor of a11+ minor of a22+ minor of a33
D3 = det (A) = |A|

3. Find the characteristic equation of the given matrix

( )k
k
Solution:

Let
X 1= 1
1 ()
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.
D1 = –2 + 0 + 3 = 1
 X 1 = (−
1
3)

X 1 =( 1 )
1
( )
1 0 0
0 3 −1
0 −1 3 .
The characteristic equation of A is λ3 – λ2 – 20λ + 12 = 0.

3 2
4. Find the characteristic equation of the given matrix
λ −D 1 λ +D 2 λ−D 3 =0
Solution:

3 −1 1 0 1 0
D2=| |+| |+| |=(9−1)+3+3=8+6=14
Let −1 3 0 3 0 3
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.
D1 = 3 + (–3) + 7 = 7

3 2
λ −7 λ +14 λ−8=0
D2 = 16

3 2
λ −7 λ +14 λ−8=0
D3 = 12

λ3 – 7λ2 + 16λ – 12 = 0
The characteristic equation of A is λ3 – 7λ2 + 16λ – 12 = 0.
Eigen values:
The roots of the characteristic equation |A – λI | = 0 are called as Eigen values or characteristic
roots (or values) or latent roots (or values).

Eigen vectors:
Let A be a square matrix and λ be an Eigen value. Then there exists a non-zero column matrix

( ) ( )( ) ( )
1 0 0 λ 0 0 x1 0
0 3 −1 − 0 λ 0 x2 = 0
0 −1 3 0 0 λ x 0
3 which satisfies (A – λI)X = 0 is called as an Eigen vector or characteristic vector or

latent vector of an Eigen value λ.

( )( ( )
1- λ 0 0 x1 0
0 3−λ −1 x2 = 0 (1)
0 −1 3−λ x 0
Note: The Eigen vectors never zero 3 . That is, there should be at least one value.
Working rule to find Eigen vector:
Let A be a square matrix.
i. Find the characteristic equation |A–λI | = 0.
ii. Solve the characteristic equation |A–λI | = 0 and get the Eigen values.
iii. To find Eigen vector, solve (A–λI)X = 0 for different Eigen values λ.

Problems

( )( ) ( )
1-1 0 0 x 1 0
0 3−1 −1 x 2 = 0
0 −1 3−1 x 0
5. Find the Eigen values and Eigen vectors of the given matrix 3

Solution:

( )( ) ( ) .
0 0 0 x1 0
0 2 −1 x 2 = 0
0 −1 2 x 0
Let 3

Step 1: Find characteristic equation

The characteristic equation of A is |A–λI | = 0.
0 x 1 +0 x 2 +0 x 3 =0 x1
|
=2
x
2 −1 −1 0 0 2
| | | |
x
=3
|
=k (say)
−1 2 2 0 0 −1
0 x 1 +2 x 2 − x 3 =0 x1 x2 x 3
= = =k
4−1 0 0
x1 x2 x3
0 x 1− x 2 +2 x 3 =0 (where
= = =k
3 0 0
x1=3k ;x2 =0; x3=0 is an identity matrix)

( )
3 k
0
0 (by scalar multiplication)

( )
1
X 1= 0
0 (by matrix subtraction)
(4 – λ)(2 – λ) – 3 = 0
8 – 4λ – 2λ + λ2 – 3 = 0
λ2 – 6λ + 5 = 0
The characteristic equation is λ2 – 6λ + 5 = 0.
Step 2: Find Eigen values
5
λ2 – 6λ + 5 = 0
λ2 – 5λ – λ + 5 = 0
λ(λ – 5) – 1(λ – 5) = 0 1 5

(λ – 1)(λ – 5) = 0

λ = 1 and λ = 5
The Eigen values are 1 and 5.
Step 3: To find Eigen vectors, solve (A–λI)X = 0

−x 1 +0 x 2+0 x 3=0
( )
0
1
1

Case (i):
When λ = 1, we get from (1)

( )
0
X 2= 1
1

( )( ) ( )
1- 4 0 0 x1 0
0 3−4 −1 x 2 = 0
0 −1 3−4 x 0
3
3x1 + x2 = 0

3x1 + x2 = 0

2 = – 3x1
x

−3x1 +0x 2+0x 3=0

0x1−x2−x3=0
Therefore, the Eigen vector corresponding to λ = 1 is 0x1−x2−x3=0 which is the general form where k is
a constant. Here we get infinite number of Eigen vectors by giving different values for k.

The Eigen vector is

−3x1+0x2+0x3=0 (Put k = 1).
Case (ii):
When λ = 5, we get from (1)

( )
0
1
−1

( )
0
X 3= 1
−1

x1 + x2 = 0 ⇒ x2 = x1
–

x1 – 3x2 = 0 ⇒ x2 = x1
3

()
1
X 1= 0
By giving x2 = k, we have x1 = k. Therefore, the Eigen vector corresponding to λ = 5 is 0 .

()
0
X 2= 1
The Eigen vector is 1 (Put k = 1).
Conclusion:
Characteristic equation Eigen values Eigen vectors
 ( )
x1
λ=1 X 1= y 1
z1

λ2 – 6λ + 5 = 0
()
x2
λ=5 X 2= y 2
z2

()
x3
X 3= y3
z3
6. Find the Eigen values and Eigen vectors of the given matrix
Solution:

( )
X1 X2 X3

x1 x2 x3
√ x + y +z √ x + y +z √ x + y +z
2

1
2

1
2

1
2

2
2

2
2

2
2

3
2

3
2

y1 y2 y3
N=

√ x1 + y 1 +z1 √ x 2+ y 2 +z2 √ x3 + y 3 +z3

2 2 2 2 2 2 2 2 2

z1 z2 z3

Let √ x + y +z √ x + y +z √ x + y +z .
2 2 2 2 2 2 2 2 2
1 1 1 2 2 2 3 3 3

Step 1: Find characteristic equation

The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.
D1 = 1 + 3 + 3 = 7

( )
1 1 3
A= 1 5 1
3 1 1

3 −1 0 −1 0 3
D3=1⋅| |−0⋅| |+0⋅| |=1⋅( 9−1 )+0+0=8
−1 3 0 3 0 −1
λ3 – 7λ2 + 14λ – 8 = 0
The characteristic equation of A is λ3 – 7λ2 + 14λ – 8 = 0.
Step 2: Find Eigen values
1 1 –7 14 –8
0 1 –6 8
1 –6 8 0
Therefore, λ = 1 is a root.
8
λ2 – 6λ + 8 = 0
λ2 – 4λ – 2λ + 8 = 0
λ(λ – 4) – 2(λ – 4) = 0 2 4

(λ – 2)(λ – 4) = 0

λ = 2 and λ = 4
The Eigen values are 1, 2 and 4.
Step 3: To find Eigen vectors, solve (A–λI)X = 0

3 2
λ −7 λ +36=0
λ2 − 4 λ −12=0
2
λ − 4 λ −2 λ +8 =0
λ ( λ −4 ) −2 ( λ −4 ) =0
( λ −2 ) ( λ − 4 ) =0

(( ) ( ))( ) ( )
1 0 0 1 0 0 x1 0
0 3 −1 − λ 0 1 0 x2 = 0
0 −1 3 0 0 1 x3 0

Case (i):
When λ = 1, we get from (1)

( ) ( )( ) ( )
1 0 0 λ 0 0 x1 0
0 3 −1 − 0 λ 0 x 2 = 0
0 −1 3 0 0 λ x 0
3

( )( ) ( )
1-λ 0 0 x1 0
0 3−λ −1 x2 = 0 ( 1)
0 −1 3−λ x 0
3

x1 + 0x2 + 0x3 = 0
0

0x1 + 2x2 – x3 = 0
x1 – x2 + 2x3 = 0
0

Consider the last two equations and applying cross rule method, we have
 ( )( ) ( )
0 0 0 x1 0
0 2 −1 x2 = 0
0 −1 2 x3 0
0x1+0x2+0x3=0
0x1+2x2−x3=0
0x −x +2x =0
Therefore, the Eigen vector corresponding to λ = 1 is 1 2 3 which is the general form where k is a
constant. Here we get infinite number of Eigen vectors by giving different values for k.
x1 x2 x3
= = =k ( say )
|2 −1| |−1 0| |0 2 |
−1 2 2 0 0 −1
x1 x2 x3
= = =k
4−1 0 0
x1 x2 x3
= = =k
3 0 0
The Eigen vector is x1=3k ;x2=0;x3 =0 (Put k = 1/3).
Case (ii):
When λ = 2, we get from (1)

( )
3 k
0
0

( )
1
X 1= 0
0
x1+ 0x2 + 0x3 = 0
–

0x1 + x2 – x3 = 0
x1 – x2 + x3 = 0
0

From – x1+ 0x2 + 0x3 = 0, we get x1 = 0.

From the last equations, we get x2 = x3.
 −x1+0x2+0x3=0
−x1+0x2+0x3=0
0x1+x2−x3=0
0x −x +x =0
Therefore, the Eigen vector corresponding to λ = 2 is 1 2 3 . That is,
Case (iii):
When λ = 4, we get from (1)

( )
0
1
1

( )
0
X 2= 1
1
3x1+ 0x2 + 0x3 = 0
–

0x1 – x2 – x3 = 0
x1 – x2 – x3 = 0
0

From – 3x1+ 0x2 + 0x3 = 0, we get x1 = 0.

From the last equations, we get x2 = – x3.

−3x1+0x2+0x3=0
−3x1+0x2+0x3=0
0x1−x2−x3=0
0x −x −x =0
Therefore, the Eigen vector corresponding to λ = 4 is 1 2 3 . That is,
Conclusion:
Characteristic equation Eigen values Eigen vectors

()
λ=1 0
X 3= 1
−1

3 2
λ −7λ +36=0
λ3 – 7λ2 + 14λ – 8 = 0 λ=2
 ()
λ=4 −1
X 1= 0
1

()
−1
X 2= 1
7. Find the Eigen values and Eigen vectors of the given matrix −1
Solution:

()
1
X 3= 2
Let 1 .
Step 1: Find characteristic equation
The characteristic equation of A is |A–λI | = 0.

( )
X1 X2 X3

−1 x2 x3
√ (−1 ) + 0 +1 √x + y +z √x + y +z
2 2 2 2 2 2 2 2 2

2 2 2 3 3 3

0 y2 y3
N=

√ (−1 ) + 0 +1
2 2 2
√x + y +z
2

2
2

2
2

2 √x + y +z
2

3
2

3
2

1 z2 z3
√ (−1 ) + 0 +1 √ x 2+ y 2+ z2 √ x 3+ y 3 + z3
2 2 2 2 2 2 2 2 2

( 5
1
4
2 )
A= ( 5
1
4
2 )
(2–λ)(2–λ)(2–λ) = 0 (or) (2–λ)3= 0
The characteristic equation of A is (2–λ)(2–λ)(2–λ) = 0.
Step 2: Find Eigen values
(2–λ)(2–λ)(2–λ) = 0
λ = 2, λ = 2 and λ = 2.
The Eigen values are 2, 2 and 2.
Step 3: To find Eigen vectors, solve (A–λI)X = 0
 |5− λ 4 |=0
1 2− λ
Case (i):
When λ = 2, we get from (1)
( 5− λ ) ( 2−λ )−4=0
10−2 λ−5 λ+ λ 2 −4=0
λ2 −7 λ+ 6=0

2
λ −7λ+6=0
x1+ x2 + 0x3 = 0
0

0x1 + 0x2 + x3 = 0
x1 + 0x2 + 0x3 = 0
0

Consider the first two equations and applying cross rule method, we have
x1
x x
= 2 = 3 =k ( say )
1 0 0 0 0 1
| | | | | |
0 1 1 0 0 0 when taking the coefficients, the order should be x2, x3, x1, x2.
x1 x2 x3
= = =k
1 0 0
x1 = k; x2 = 0; x3 = 0

Therefore, the Eigen vector corresponding to λ = 2 is

( )( )
| 1 2 −λ 10 |=0
−3 4 01 which is the general form where k is a
constant. Here we get infinite number of Eigen vectors by giving different values for k.

The Eigen vector is

−3 4 0 λ ( )( )
| 1 2 − λ 0 |=0
(Put k = 1).
The second and third Eigen vector is also same as X1.
Conclusion:
Characteristic equation Eigen values Eigen vectors

λ=2
(1−λ ) ( 4−λ)−(−6)=0
4−λ−4λ+λ2+6=0
λ2−5λ+10=0

2
λ −5λ+10=0
λ=2
(2–λ) = 0
3

( )
a 11 a12 a13
λ=2
A= a 21 a22 a23
a 31 a32 a33

Properties of Eigen values:

Let λ1, λ2,…, λn be the Eigen values corresponding to a square matrix A of order n. That is,

3 2
λ −D1 λ +D2 λ−D3=0
Property 1:
Sum of Eigen values is equal to the sum of the diagonal elements. That is,
λ1+ λ2+ ··· + λn = a11+ a22+ ··· + ann

Problems:

( )
−2 2 1
5 0 2
1. Find the sum of the Eigen values of 0 2 3
Solution:
Sum of the Eigen values = Sum of the principal diagonal elements. Here, the diagonal elements
are 2, 1 and 13.
Sum of the Eigen values = 2 + 1 + 13 = 16

( )
−2 2 1
A= 5 0 2
2. If 2 and 8 are the Eigen values of 0 2 3 , then find the third Eigen value.
Solution:
Let λ1, λ2, λ3 be the Eigen values. Then λ1+ λ2+ λ3 = 6 + 3 + 3 = 12.
Given that λ1 = 2 and λ2 = 8.
2 + 8+ λ3 = 12. λ3 = 12 – 10 = 2. Therefore, the third Eigen value λ3 = 2.

Property 2:
Product of Eigen values is equal to its determinant value. That is, λ1× λ2× ··· × λn = |A|.
Problems:

3 2
1. The product of two Eigen values of the matrix
λ −D 1 λ +D 2 λ−D 3 =0 is 16, Find the third
Eigen value.
Solution:
Let λ1, λ2, λ3 be the Eigen values.
Given that λ1×λ2 = 16. (1)

D2=|02|+−2 1|+−2 2|=(0−4)+(−6−0)+(0−10)=−4−6−10=−20

But λ × λ × λ = 23 0 3 5 0
1 2 3

02 5250
D3=−2⋅| |−2| |+| |=−2⋅(0−4 )−2⋅(15−0)+(10−0)=8−30+10=−12
23 0302
λ1× λ2× λ3 = 32 (2)
Substitute (2) in (1) we get, 16 × λ3 = 32. Therefore, λ3 = 32/16 = 2. λ3 = 2.
2. For a square matrix A of order 3, the trace and determinant are given to 20 and 0
respectively. Find the other Eigen values if one of the Eigen values is 15.
Solution:
Given that A is a square matrix of order 3. Let λ1, λ2, λ3 be the Eigen values of A since order is 3.
Also, given that trace of A is 20, that is, λ1+ λ2+ λ3 = 20, determinant of A is 0, that is, λ1× λ2× λ3 =
0 and λ1 = 15.
From λ1+ λ2+ λ3 = 20 and λ1 = 15, we get λ2+ λ3 = 20 – 15 = 5. This implies that λ2 = 5 – λ3.
From λ1× λ2× λ3 = 0, we get λ2× λ3 = 0 since λ1 = 15 ≠ 0.
2
λ2 ×λ 3 =0⇒ ( 5−λ3 ) λ3 =0 ⇒ 5 λ3 =λ3 ⇒ λ3 =5
Therefore, λ2 = (5 – 5) = 0.
Hence, the Eigen values are 15, 0 and 5.

Property 3:
Every square matrix and its transpose have the same Eigen values. That is, A and AT have the
same Eigen values.
Property 4:
If λ1, λ2,…, λn are Eigen values of A, then the inverse A–1 has the Eigen values 1/λ1, 1/λ2, …, 1/λn.
Problem:

32
1. If 2 and 3 are the Eigen values of
λ −λ +(−20)λ−(−12)=0 , then find the Eigen values of A–1.
Solution:
Let λ1, λ2, λ3 be the Eigen values of A. Then λ1+ λ2+ λ3 = 3 – 3 + 7 = 7.
Given that λ1 = 2 and λ2 = 3.
2 + 3+ λ3 = 7. λ3 = 7 – 5 = 2. Therefore, the Eigen values of A–1 are ½, ½ and 1/3.
Property 5:
If λ1, λ2,…, λn are Eigen values of A, then the matrix A + kI has the Eigen values k + λ1, k + λ2, …,
k + λn.
Property 6:
If λ1, λ2,…, λn are Eigen values of A, then the matrix A – kI has the Eigen values k – λ1, k – λ2,…,
k – λn.
Property 7:
2 2 2
If λ1, λ2,…, λn are Eigen values of A, then the matrix A2 has the Eigen values λ1 , λ 2 ,…, λ n .
In general, if λ is an Eigen value of A, then λn is an Eigen value of An.
Property 8:
If λ1, λ2,…, λn are Eigen values of A, then the matrix kA has the Eigen values kλ1, kλ2,…, kλn.

Problem:
2. If 2 and 3 are the Eigen values of a square matrix with order 2, then find the Eigen
values of A2 and A–1.
Solution:
Eigen values of A2 are 22, 32 (or) 4, 9.
Eigen values of A–1 are ½, 1/3.

Diagonalization of Matrix:
A square matrix A is said to be diagonalizable if D = NTAN, where N is a normalized modal
matrix and D is a diagonal matrix.
Diagonalization is the process of transforming a matrix into a diagonal matrix.
Note:
1. Diagonalization is possible only for real symmetric matrices (i.e., A = AT).
2. Both N and D are not unique.
3. N is an orthogonal matrix (i.e., NTN = NNT = I, where I is an identity matrix).

Procedure for Diagonalization of the Matrix by orthogonal transformation or orthogonal

reduction:
Let A be a real symmetric matrix.
i. Find the characteristic equation |A–λI | = 0.
ii. Solve the characteristic equation |A–λI | = 0 and get the Eigen values.
iii. To find Eigen vector, solve (A–λI)X = 0 for different Eigen values λ.
iv. Find the normalized modal matrix by using Eigen vectors.
 λ −λ −20λ+12=0 ( )( )
32 3 10 5 3 10 5
−2 −3 −4 A= −2 −3 −4
;
3 5 7 ; 3 5 7

Let the Eigen vectors be

Hence, the normalized modal matrix, N is as follows.

( )
X1 X2 X3

x1 x2 x3
√ x1 + y 1 + z 1 √ x2 + y 2 + z 2 √ x 3 + y 3 + z3
2 2 2 2 2 2 2 2 2

y1 y2 y3
N=

√x + y +z √x + y +z √x + y + z
2 2 2 2 2 2 2 2 2
1 1 1 2 2 2 3 3 3

z1 z2 z3
√x + y +z √x + y +z √x + y + z
2 2 2 2 2 2 2 2 2
1 1 1 2 2 2 3 3 3

v. Find NT
vi. Find AN
vii. Find D = NTAN. Then D is the required diagonal matrix.

D2=|−3 −4|+3 5|+3 10 |=(−21−(−20)+(21−15)+(−9−(−20)

5 7 3 7 −2 −3
8. Diagonalize the given matrix =(−21+20)+6+(−9+20)=−1+6+1 =16 by orthogonal transformation.

Solution:
−3 −4 −2 −4 −2 −3
D3=3⋅| |−10⋅| |+5⋅| |=3⋅(−21−(−20) −10⋅(−14−(−12) +5⋅(−10−(−9)
57 37 35
=3⋅(−21+20)−10⋅(−14+12)+5⋅(−10+9)
Let =3⋅(−1)−10⋅(−2)+5⋅(−1)=−3+20−5=12 .
Step 1: Find characteristic equation
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.

D1 = 1 + 5 + 1 = 7

3 2
λ −7 λ +16 λ−12=0
( )
x 1
X = x 2
⋮
x n

λ3 – 7λ2 – (–36) = 0
The characteristic equation of A is λ3 – 7λ2 + 36 = 0.
Step 2: Find Eigen values
3 1 –7 0 36
0 3 – 12 –36
1 –4 – 12 0
Therefore, λ = 3 is a root.
–12
λ2 – 4λ – 12 = 0
λ2 – 6λ + 2λ – 12 = 0
λ(λ – 6) + 2(λ – 6) = 0 2 –6

(λ + 2)(λ – 6) = 0
λ = – 2 and λ = 6
The Eigen values are – 2, 3 and 6.
Step 3: To find Eigen vectors, solve (A–λI)X = 0

( ) ( )
| 4 1 − λ 1 0 |=0
3 2 0 1

(0 1 )
1 0

|( 4
2 ) (0 )
1 − λ 0 |=0
3 λ
Case (i):
When λ = – 2, we get from (1)

4− λ 1
| |=0
3 2− λ
( 4−λ )( 2−λ ) −3=0
8−4 λ−2 λ+ λ2 −3= 0
λ2 −6 λ +5=0

2
λ −6λ+5=0
x1 + x2 + 3x3 = 0
3

x1 + 7x2 + x3 = 0
x1 + x2 + 3x3 = 0
3

Consider the first two equations and applying cross rule method, we have

(( ) ( ))( ) ( )
4 1
3 2
−λ
1 0
0 1
x1 0
=
x2 0
Therefore, the Eigen vector corresponding to λ = – 2 is
( ) ()
4−λ 1 x1 0
=
3 2−λ x2 0
(1)
which is the general form where
k is a constant. Here we get infinite number of Eigen vectors by giving different values for k.
The Eigen vector is
( )( ) ( )
4−1 1 x1 = 0
3 2−1 x2 0
(Put k = 1/20).
Case (ii):
When λ = 3, we get from (1)

( 3
3
1
1 )( ) ( )
x1
x2
=
0
0
3 x1 + x 2=0
3 x1 + x 2=0
x 2=− 3 x 1
x1 + x2 + 3x3 = 0
–2

x1 + 2x2 + x3 = 0
x1 + x2 – 2x3 = 0
3

Consider the first two equations and applying cross rule method, we have

X 1=
1
−3 ( )
The Eigen vector is
( )( ) ( )
4−5 1 x1 0
=
3 2−5 x2 0
(Put k = 1/5).
Case (iii):
When λ = 6, we get from (1)

( −1 1 x 1 0
=
3 −3 x 2 0 )( ) ( )
−x 1+x 2=0⇒ x 2=x 1
3x 1−3 x 2=0⇒ x 2=x 1
x1 + x2 + 3x3 = 0
–5

x1 – x2 + x3 = 0
x1 + x2 – 5x3 = 0
3

Consider the first two equations and applying cross rule method, we have

X 1=
1
1
2
( )
The Eigen vector is
λ −6λ+5=0 (Put k = 1/4).
Conclusion:
Characteristic equation Eigen values Eigen vectors
 λ=–2
()
X 1= 1
1

( )
λ=3 1 0 0
λ – 7λ + 36 = 0
3 2
0 3 −1
0 −1 3

( )
λ=6 1 0 0
A= 0 3 −1
0 −1 3

Check orthogonal

()
−1
X T1 X 2 =( −1 0 1 ) 1 =1+0−1=0
−1

()
1
X T2 X 3 =( −1 1 −1 ) 2 =−1+2−1=0
1

()
1
X T1 X 3 =( −1 0 1 ) 2 =−1+0+1=0
1
Therefore, Eigen vectors are orthogonal to each other. Hence, the normalized modal matrix, N is
as follows.
 3 2
λ −D1 λ +D2 λ−D3=0
3 −1 1 0 1 0
D2=| |+| |+| |=(9−1 )+3+3=8+6=14
−1 3 0 3 0 3
3 −1 1 0 10
D3=1⋅| |−0⋅| |+0⋅| |=(9−1)−0⋅(3−0)+0⋅(3−0)=8
−1 3 0 3 03
3 2
λ −7 λ +14 λ−8=0
3 2
λ −7λ +14 λ−8=0 (Don’t change the order. That is, don’t take NA)
 λ2 −6 λ +8=0
λ2 − 4 λ−2 λ+8=0
λ ( λ−4 )−2 ( λ−4 )=0
( λ−2 ) ( λ− 4 ) =0

( ) ( )( ) ( )
1 0 0 1 0 0 x1 0
0 3 −1 −λ 0 1 0 x2 = 0
0 −1 3 0 0 1 x3 0

( )( ))( ) ( )
1 0 0 λ 0 0 x1 0
0 3 −1 − 0 λ 0 x2 = 0
0 −1 3 0 0 λ x3 0

( )
−2 0 0
T
N AN = 0 3 0
0 0 6 =D (i.e., a diagonal matrix)
Therefore, A is reduced to a diagonal matrix by means of orthogonal transformation.

( )( ) ( )
1-1 0 0 x1 0
0 3−1 −1 x2 = 0
0 −1 3−1 x 0
9. Diagonalize the given matrix 3 by orthogonal transformation.

Solution:

( )( ) ( )
0 0 0 x1 0
0 2 −1 x 2 = 0
0 −1 2 x 0
Let 3 .
Step 1: Find characteristic equation
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.

D1 = 1 + 2 + 1 = 4
 x1 x2 x3
= = =k ( say )
|2 −1 | |−1 0| |0 2 |
−1 2 2 0 0 −1
x1 x2 x3
= = = k
4 −1 0 0
x1 x2 x3
= = =k
3 0 0
x 1=3 k ; x 2 =0 ; x3 = 0

( )
3 k
0
0

λ3 – 4λ2 + 3λ = 0
The characteristic equation of A is λ3 – 4λ2 + 3λ = 0.
Step 2: Find Eigen values
λ3 – 4λ2 + 3λ = 0
λ (λ2 – 4λ + 3) = 0 3
λ = 0 (or) λ2 – 4λ + 3 = 0
λ2 – 3λ – λ + 3 = 0 1 3
λ(λ – 3) – 1(λ – 3) = 0
(λ – 1)(λ – 3) = 0

λ = 1 and λ = 3
The Eigen values are 0, 1 and 3.
Step 3: To find Eigen vectors, solve (A–λI)X = 0
− x 1 +0 x 2 +0 x 3 =0
0 x1 + x 2− x 3 =0
0 x1 − x 2 + x 3 =0

−x 1+0 x 2+0 x 3=0

( )
0
1
1
Case (i):
When λ = 0, we get from (1)

( )
0
X 2= 1
1
( )( ) ( )
1- 4 0 0 x1 0
0 3−4 −1 x 2 = 0
0 −1 3−4 x 0
3

1 – x2 + 0x3 = 0
x

– x1 + 2x2 + x3 = 0
x1 + x2 + x3 = 0
0

Consider the first two equations and applying cross rule method, we have

−3 x1 +0 x 2 +0 x 3= 0
0 x1 − x 2 − x 3= 0
0 x1 − x 2 − x 3= 0

Therefore, the Eigen vector corresponding to λ = 0 is

−3x1+0x2+0x3=0 which is the general form where k is
a constant. Here we get infinite number of Eigen vectors by giving different values for k.

( )
0
1
The Eigen vector is −1 (Put k = –1).
Case (ii):
When λ = 1, we get from (1)

( )
0
X 3= 1
−1
 3 2
λ −7λ +14λ−8=0
x1 – x2 + 0x3 = 0
0

– x1 + x2 + x3 = 0
x1 + x2 + 0x3 = 0
0

From 0x1 + x2 + 0x3 = 0, we get x2 = 0.

Put x2 = 0 in – x1 + x2 + x3 = 0, we get – x1 + x3 = 0 ⇒ x1 = x3
Put x3 = k in x1 = x3, we get x1 = k
Therefore, we get x1 = k, x2 = 0 and x3 = k.

()
x2
X 2= y2
z2
The Eigen vector is (Put k = 1).
Case (iii):
When λ = 3, we get from (1)

( )
x3
X 3= y3
z3

( )
X1 X2 X3

x1 x2 x3
√x + y √x + y √x + y
2 2 2 2 2 2 2 2 2

1 1
+ z1 2 2
+ z2 3 3
+ z3
y1 y2 y3
N=

√x + y √x + y √x + y
2 2 2 2 2 2 2 2 2

1 1
+ z1 2 2
+ z2 3 3
+ z3
z1 z2 z3
√x + y √x + y √x + y
2 2 2 2 2 2 2 2 2
1 1 + z1 2 2 + z2 3 3 + z3

x1 – x2 + 0x3 = 0
–2

– x1 – x2 + x3 = 0
x1 + x2 – 2x3 = 0
0
Consider the first two equations and applying cross rule method, we have

( )
1 1 3
A= 1 5 1
3 1 1
3 2
The Eigen vector is
λ −D 1 λ +D 2 λ − D 3 =0 (Put k = –1).
Conclusion:
Characteristic equation Eigen values Eigen vectors

λ=0 51 1 15
D3=1⋅| |−1⋅ |+3⋅ |=1⋅(5−1) ⋅(1−3)+⋅(1−5)=4−1⋅( 2)+3⋅(−14)= +2−4=−36
1 31 31
3 2
λ −7λ −(−36)=0
λ=1
λ – 4λ + 3λ = 0
3 2

3 2
λ −7λ +36=0
λ=3
Check orthogonal

()
1
X T1 X 2 =( 1 1 −1 ) 0 =1+0−1=0
1

()
1
X T2 X 3 =( 1 0 1 ) −2 =1+0−1=0
−1

()
1
X T1 X 3 =( 1 1 −1 ) −2 =1−2+1=0
−1
Therefore, Eigen vectors are orthogonal to each other. Hence, the normalized modal matrix, N is
as follows.

2
λ − 4 λ−12=0
2
λ − 4 λ−2 λ+8=0
λ ( λ−4 )−2 ( λ−4 )=0
( λ−2 )( λ− 4 ) =0

( ) ( )( ) ( )
1 0 0 1 0 0 x1 0
0 3 −1 −λ 0 1 0 x2 = 0
0 −1 3 0 0 1 x3 0
( ) ( )( ) ( )
1 0 0 λ 0 0 x1 0
0 3 −1 − 0 λ 0 x2 = 0
0 −1 3 0 0 λ x 0
3

( )( ) ( )
1- λ 0 0 x1 0
0 3− λ −1 x2 = 0 (1)
0 −1 3− λ x 0
3

( )( ) ( )
1-1 0 0 x1 0
0 3−1 −1 x2 = 0
0 −1 3−1 x 0
3 (Don’t change the order. That is, don’t take NA)

( )( ) ( )
0 0 0 x1 0
0 2 −1 x 2 = 0
0 −1 2 x 0
3

0 x 1 +0 x 2 +0 x 3 =0
0 x 1 +2 x 2 − x 3 =0
0 x 1− x 2 +2 x 3 =0
x1 x2 x3
= = =k ( say )
2 −1 −1 0 0 2
| | | | | |
−1 2 2 0 0 −1
x1 x2 x3
= = =k
4−1 0 0
x1 x2 x3
= = =k
3 0 0
x 1=3k ; x 2 =0; x3=0
 ( )
3 k
0
0
Therefore, A is reduced to a diagonal matrix by means of orthogonal transformation.

Quadratic form
A homogeneous polynomial of second degree in any number of variables is a called quadratic
form.
A homogeneous polynomial of second degree means each and every term in any expression
should have degree two.
Examples:
2 2
1. x 1 +5 x1 x 2 + 2 x 2 is a quadratic form in the variables x1 and x2.
2. x2 + y2 + z2 – xy + 2yz – 3xz is a quadratic form in the variables x, y and z.
A general quadratic form in three variables x, y and z is given by
f(x, y, z) = a11x2 + a12xy + a13xz + a21yx + a22y2 + a23yz + a31zx + a32zy + a33z2
This quadratic form can be written as

( )( ) ()
a11 a12 a13 x x
T
f ( x , y , z )=( x y z ) a21 a22 a23 y = X AX y
z
where X = z and A is called the matrix of
a31 a32 a33

the quadratic form.

The matrix from quadratic form can be written as
x1 x2 x3

( )
Coeff x 1 x 2 Coeff x 1 x 3
Coeff x 21
2 2
x1
Coeff x 2 x 1 Coeff x 2 x 3
A= x 2 Coeff x 22
2 2
x 3 Coeff x x Coeff x 3 x 2
3 1
Coeff x 23
2 2
x1x2 = x2x1; x1x3 = x3x1; x2x3 = x3x2 (or)
 x y z

( )
Coeff xy Coeff xz
Coeff x 2
2 2
x
Coeff yx Coeff yz
A= y Coeff y 2
2 2
z
Coeff zx Coeff zy
Coeff z 2
2 2
xy = yx; xz = zx; yz = zy

Problems:
1. Write down the matrix for the given quadratic form x2 + y2 + z2 – 6xy.
Solution:
Given quadratic form is x2 + y2 + z2 – 6xy.
x y z

( )
Coeff xy Coeff xz
Coeff x 2
2 2
x
Coeff yx Coeff yz
A= y Coeff y 2
2 2
z
Coeff zx Coeff zy
Coeff z 2
2 2

( )( )
1 −6/2 0 1 −3 0
A= −6/2 1 0 = −3 1 0
0 0 1 0 0 1

2. Write down the matrix for the given quadratic form 10x2 + 2y2 + 5z2 – 4xy + 6yz – 10xz.
Solution:

Given quadratic form is 10x + 2y + 5z – 4xy + 6yz – 10xz.

2 2 2

x y z

( )
Coeff xy Coeff xz
Coeff x 2
2 2
x
Coeff yx Coeff yz
A= y Coeff y 2
2 2
z
Coeff zx Coeff zy
Coeff z 2
2 2

( )( )
10 −4 /2 −10 /2 10 −2 −5
A= −4 /2 2 6/2 = −2 2 3
−10 /2 6/2 5 −5 3 5
Procedure for finding quadratic form from given symmetric matrix:

( )
a 11 a12 a13
A= a 21 a22 a23
a 31 a32 a33
Let be a symmetric matrix.

The quadratic form corresponding to any symmetric matrix is

( )( )
a11 a12 a13 x
T
X AX=( x y z ) a21 a22 a23 y
a31 a32 a33 z

( )
−2 2 −1
2 0 5
1. Transform the given symmetric matrix
−1 5 3 into a quadratic form
Solution:

( )
−2 2 −1
A= 2 0 5
Let −1 5 3 .
The quadratic form corresponding to any symmetric matrix is

( )( )
a11 a12 a13 x
T
X AX=( x y z ) a21 a22 a23 y
a31 a32 a33 z

( )( )
−2 2 −1 x
T
X AX=( x y z ) 2 0 5 y
That is, −1 5 3 z
Using matrix multiplication we have,
 ()
x
T
X AX=( −2 x +2 y−z 2 x+5 z −x +5 y+3 z ) y
z
=(−2 x+2 y−z ) x+ (2 x+5 z ) y + (−x+5 y+3 z ) z
=(−2 x 2 +2 xy −xz ) + ( 2 xy+5 yz ) + (−xz+5 yz+3 z 2 )
=−2 x2 +3 z2 +4 xy +10 yz−2 xz
The quadratic form corresponding to the given matrix is – 2x2 + 3z2 + 4xy + 10yz – 2xz.

( )
2 1 −1
1 1 −2
2. Transform the given symmetric matrix
−1 −2 1 into a quadratic form
Solution:

( )
2 1 −1
A= 1 1 −2
Let −1 −2 1 .
The quadratic form corresponding to any symmetric matrix is

( )( )
a11 a12 a13 x
T
X AX=( x y z ) a21 a22 a23 y
a31 a32 a33 z

( )( )
2 1 −1 x
T
X AX=( x y z ) 1 1 −2 y
That is, −1 −2 1 z
Using matrix multiplication we have,

()
x
T
X AX=( 2 x + y −z x + y −2 z −x−2 y +z ) y
z
=( 2 x + y −z ) x + ( x+ y −2 z ) y + (−x−2 y+z ) z
=( 2 x2 +xy−xz ) + ( xy + y 2 −2 yz ) + (−xz−2 yz +z 2 )
=2 x 2 + y 2 +z 2 +2 xy−4 yz−2 xz
The quadratic form corresponding to the given matrix is 2x2 + y2 + z2 + 2xy – 4yz – 2xz.
Nature of quadratic form:
Let A be a symmetric matrix and XTAX be the quadratic form corresponding to the matrix A. Let
λ1, λ2 and λ3 be the Eigen values of A and XTAX is said to be
i. Positive definite: if all the Eigen values λ1, λ2 and λ3 are positive.
ii. Negative definite: if all the Eigen values λ1, λ2 and λ3 are negative.
iii. Positive semi-definite: if at least one Eigen value is zero and the remaining are
positive.
iv. Negative semi-definite: if at least one Eigen value is zero and the remaining are
negative.
v. Indefinite: if some Eigen values are positive and some Eigen values are negative.

1. Discuss the nature of the quadratic form 3x2 + 3y2 – 5z2 – 2xy – 6yz – 6xz.
Solution:

Given quadratic form is 3x + 3y – 5z – 2xy – 6yz – 6xz.

2 2 2

x y z

( )
Coeff xy Coeff xz
Coeff x 2
2 2
x
Coeff yx Coeff yz
A= y Coeff y 2
2 2
z
Coeff zx Coeff zy
Coeff z 2
2 2

( )( )
3 −2/2 −6 /2 3 −1 −3
A= −2/2 3 −6 /2 = −1 3 −3
−6 /2 −6/2 −5 −3 −3 −5
Step 1: Find characteristic equation
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.

D1 = 3 + 3 – 5 = 1
3 −3 3 −3 3 −1
D2 =| |+| |+| |=(−15−9)+(−15−9)+(9−1)=−24−24+8=−40
−3 −5 −3 −5 −1 3
3 −3 −1 −3 −1 3
D3 =3⋅| |+ (−1 )⋅| |+ (−3 )⋅| |=3⋅(−15−9)−(−1 )⋅(5−9)+ (−3 )⋅( 3−(−9 ) )
−3 −5 −3 −5 −3 −3
=3⋅(−24 )−4−36=−72−4−36=−112
λ3 – λ2 – 40λ – (–112) = 0
λ3 – λ2 – 40λ + 112 = 0
The characteristic equation of A is λ3 – λ2 – 40λ + 112 = 0.
Step 2: Find Eigen values
4 1 –1 –40 112
0 4 12 –112
1 3 – 28 0
Therefore, λ = 4 is a root.
–28
λ2 + 3λ – 28 = 0
λ2 + 7λ – 4λ – 28 = 0
λ(λ + 7) – 4(λ + 7) = 0 7 –4

(λ + 7)(λ – 4) = 0
λ = – 7 and λ = 4
The Eigen values are 4, 4 and – 7.

The Eigen values of A are –1, 4 and 8. The nature of the given canonical form is indefinite since
it has both positive and negative Eigen values.

Procedure for finding canonical form from given quadratic form:

i. Convert the given quadratic form into matrix using the following expression.
x y z

( )
Coeff xy Coeff xz
Coeff x 2
2 2
x
Coeff yx Coeff yz
A= y Coeff y 2
2 2
z
Coeff zx Coeff zy
Coeff z 2
2 2

ii. Find the characteristic equation |A–λI | = 0.

iii. Solve the characteristic equation |A–λI | = 0 and get the Eigen values.
iv. To find Eigen vector, solve (A–λI)X = 0 for different Eigen values λ.
v. Find the normalized modal matrix by using Eigen vectors.
 λ −λ −20λ+12=0 ( )( )
32 3 10 5 3 10 5
−2 −3 −4 A= −2 −3 −4
;
3 5 7 ; 3 5 7

Let the Eigen vectors be

Hence, the normalized modal matrix, N is as follows.

( )
X1 X2 X3

x1 x2 x3
√x + y +z √x + y +z √x + y + z
2 2 2 2 2 2 2 2 2
1 1 1 2 2 2 3 3 3

y1 y2 y3
N=

√x + y +z
2
1
2
1
2
1 √x + y +z
2
2
2
2
2
2 √x + y + z
2
3
2
3
2
3

z1 z2 z3
√x + y +z √x + y +z √x + y + z
2 2 2 2 2 2 2 2 2
1 1 1 2 2 2 3 3 3

vi. Find NT
vii. Find AN
viii. Find D = NTAN

()
y1
Y = y2
y3
ix. Find the canonical form C.F. = YTDY where
Index and Signature of the Canonical form:
Index (s): The number of positive square terms in the canonical form
Rank (r): The number of square terms in the canonical form
Signature: 2s − r (or) the number of positive terms and negative terms in the canonical form

2 2
1. State the nature and rank of the canonical form − y 1 +5 y 3
Solution:
The matrix of the quadratic form
 y1 y2 y3

( )
Coeff y 1 y 2 Coeff y 1 y 3
Coeff y 21
2 2
y1
Coeff y 2 y 1 Coeff y 2 y 3
A= y 2 Coeff y 22
2 2
y3
Coeff y 3 y 1 Coeff y 3 y 2
Coeff y 23
2 2
y1y2 = y2y1; y1y3 = y3y1; y2y3 = y3y2

( )
−1 0 0
A= 0 0 0
0 0 5
The Eigen values are 0, – 1 and 5. The nature of the given canonical form is indefinite since it
has both positive and negative Eigen values.
The number of square terms in the canonical form is 2. Hence, Rank is 2.

2. Reduce the quadratic form into its canonical form by using orthogonal transformation
8x2 + 7y2 + 3z2 – 12xy − 8yz + 4xz.
Solution:
Given quadratic form is 8x2 + 7y2 + 3z2 – 12xy − 8yz + 4xz.
Step: 1 Find the matrix of the quadratic form
The matrix of the quadratic form
x y z

( )
Coeff xy Coeff xz
Coeff x 2
2 2
x
Coeff yx Coeff yz
A= y Coeff y 2
2 2
z
Coeff zx Coeff zy
Coeff z 2
2 2

( )( )
8 −12/2 4/2 8 −6 2
A= −12/2 7 −8 /2 = −6 7 −4
4 /2 −8 /2 3 2 −4 3

( )
8 −6 2
A= −6 7 −4
2 −4 3
Step 1: Find characteristic equation
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.
D1 = 8 + 7 + 3 = 18
7 −4 8 2 8 −6
D2 =| |+| |+| |=( 21−16 ) + ( 24−4 ) + ( 56−36 )=5+20+20=45
−4 3 2 3 −6 7
7 −4 −6 −4 −6 7
D3 =8⋅| |−(−6 )⋅| |+2⋅| |=8⋅( 21−16 ) +6⋅(−18−(−8 ) ) +2⋅( 24−14 )
−4 3 2 3 2 −4
=8⋅5+6⋅(−18+8 ) +2⋅10=40−60+20=0 λ3
– 18λ2 + 45λ = 0
The characteristic equation of A is λ3 – 18λ2 + 45λ = 0.
Step 2: Find Eigen values
λ3 – 18λ2 + 45λ = 0
λ (λ2 – 18λ + 45) = 0 45
λ = 0 (or) λ2 – 18λ + 45 = 0
λ2 – 15λ – 3λ + 45 = 0 3 15
λ(λ – 15) – 3(λ – 15) = 0
(λ – 3)(λ – 15) = 0

λ = 3 and λ = 15
The Eigen values are 0, 3 and 15.
Step 3: To find Eigen vectors, solve (A–λI)X = 0
λ3 – 18λ2 + 45λ = 0

( ) ( )( ) ( )
8 −6 2 λ 0 0 x1 0
−6 7 −4 − 0 λ 0 x2 = 0
2 −4 3 0 0 λ x3 0

( )( ) ( )
8−λ −6 2 x1 0
−6 7−λ −4 x 2 = 0 (1)
2 −4 3−λ x 0
3

Case (i):
When λ = 0, we get from (1)

( )( ) ( )
8 −6 2 x 1 0
−6 7 −4 x 2 = 0
2 −4 3 x 0
3
 x1 – 6x2 + 2x3 = 0
8

– 6x1 + 7x2 – 4x3 = 0

2x1 – 4x2 + 3x3 = 0

Consider the first two equations and applying cross rule method, we have
x1 x x
= 2 = 3 =k ( say )
−6 2 2 8 8 −6
| | | | | |
7 −4 −4 −6 −6 7
x1 x x
= 2 = 3 =k
24−14 (−12− (−32 ) ) 56−36
x1 x2 x3
= = =k
10 (−12+32 ) 20
x1 x2 x3
= = =k
10 20 20
x1 = 10k; x2 = 20k; x3 = 20k

( )
10k
20k
λ = 0 is 20k which is the general form where k is

Therefore, the Eigen vector corresponding to

a constant. Here we get infinite number of Eigen vectors by giving different values for k.

()
1
X 1= 2
The Eigen vector is 2 (Put k = 1/10).

Case (ii):
When λ = 3, we get from (1)
( )( ) ( )
8−3 −6 2 x1 0
−6 7−3 −4 x 2 = 0
2 −4 3−3 x 0
3

( )( ) ( )
5 −6 2 x 1 0
−6 4 −4 x 2 = 0
2 −4 0 x 0
3

x1 – 6x2 + 2x3 = 0
5

– 6x1 + 4x2 – 4x3 = 0

2x1 – 4x2 + 0x3 = 0

Consider the first two equations and applying cross rule method, we have
x1 x2 x3
= = =k ( say )
|−6 2 | |2 5 | |5 −6 |
4 −4 −4 −6 −6 4
x1 x x
= 2 = 3 =k
24−8 (−12−(−20 ) ) 20−36
x 1 x2 x3
= = =k
16 (−12+20 ) −16
x 1 x2 x 3
= = =k
16 8 −16
x1 = 16k; x2 = 8k; x3 = –16k

()
2
X 2= 1
Therefore, the Eigen vector corresponding to λ = 3 is −2 .
Case (iii):
When λ = 15, we get from (1)

( )( ) ( )
8−15 −6 2 x1 0
−6 7−15 −4 x 2 = 0
2 −4 3−15 x 0
3
( )( ) ( )
−7 −6 2 x1 0
−6 −8 −4 x 2 = 0
2 −4 −12 x 0
3

x1 – 6x2 + 2x3 = 0
–7

– 6x1 – 8x2 – 4x3 = 0

2x1 – 4x2 – 12x3 = 0

Consider the first two equations and applying cross rule method, we have

x1 x2 x3
= = =k ( say )
−6 2 2 −7 −7 −6
| | | | | |
−8 −4 −4 −6 −6 −8
x1 x x
= 2 = 3 =k
( 24−(−16 ) ) (−12−28 ) 56−36
x1 x x
= 2 = 3 =k
24+16 −40 20
x1 x2 x3
= = =k
40 −40 20
x1 = 40k; x2 = – 40k; x3 = 20k

()
2
X 3 = −2
Therefore, the Eigen vector corresponding to λ = 15 is 1 .
Conclusion:
Characteristic Eigen values Eigen vectors
equation

()
λ=0 1
X 1= 2
2

()
λ=3 2
λ3 – 18λ2 + 45λ = 0
X 2= 1
−2
 ()
λ = 15 2
X 3 = −2
1

Check orthogonal

()
2
X T1 X 2 =( 1 2 2 ) 1 =2+2−4=0
−2

()
2
X T2 X 3 =( 2 1 −2 ) −2 =4−2−2=0
1

()
2
X T1 X 3 =( 1 2 2 ) −2 =2−4+2=0
1
Therefore, Eigen vectors are orthogonal to each other. Hence, the normalized modal matrix, N is
as follows.

( )
X1 X2 X3

1 2 2
√ 12 +22 +22 √ 2 +1 + (−2 )
2 2 2
√ 2 + (−2 ) +1
2 2 2

2 1 −2
N=
√ 12 +22 +22 √ 2 +1 + (−2 ) √ 2 + (−2 ) +1
2 2 2 2 2 2

2 −2 1
√ 1
2
+2
2
+2
2
√ 2 +1 + (−2 )
2 2 2
√ 2 + (−2 ) +1
2 2 2

( )( )( )
1 2 2 1 2 2 1 2 2
√1+ 4+ 4 √ 4+1+ 4 √ 4+ 4+1 √9 √9 √ 9 3 3 3
2 1 −2 2 1 −2
N= = = 2 1 −2
√1+ 4+ 4 √ 4+1+ 4 √ 4+ 4+1 √9 √9 √9 3 3 3
2 −2 1 2 −2 1 2 −2 1
√1+ 4+ 4 √ 4+1+ 4 √ 4+ 4+1 √9 √9 √9 3 3 3
 ( )
1 2 2
3 3 3

( )( )
1 2 2
2 1 −2 1
N= = 2 1 −2
3 3 3 3
2 −2 1
2 −2 1
3 3 3

( )( )
1 2 2
T 1
N = 2 1 −2
3
2 −2 1

( )( )( )
8 −6 2 1 2 2
1
AN = −6 7 −4 2 1 −2
3
2 −4 3 2 −2 1 (Don’t change the order. That is, don’t take NA)

( )( ) ( )( )
8−12+4 16−6−4 16+12+2 0 6 30
1 1
AN = −6+14−8 −12+7+8 −12−14−4 = 0 3 −30
3 3
2−8+6 4−4−6 4+8+3 0 −6 15

( )( )( ) ( ) ( )
2 1 2 2 0 6 30 2 0 6+6−12 30−60+30
T 1 1
N AN = 2 1 −2 0 3 −30 = 0 12+3+12 60−30−30
3 3
2 −2 1 0 −6 15 0 12−6−6 60+60+15

( )( )() ( )( )
2 0 0 0 2 0 0 0 0 0 0
T 1 1 2
N AN = 0 27 0 = ⋅3 ⋅ 0 3 0 = 0 3 0
3 3
0 0 135 0 0 15 0 0 15

( )
0 0 0
T
N AN = 0 3 0 =D
0 0 15 (i.e., a diagonal matrix)

Therefore, A is reduced to a diagonal matrix by means of orthogonal transformation.

Find Canonical form YT(NTAN)Y
 ( )( )
0 0 0 y1
Y ( N AN ) Y =( y 1
T T
y2 y3 ) 0 3 0 y2
0 0 15 y3

()
y1
Y ( N AN ) Y =( 0 y 1 +0 y 2 + 0 y 3 0+3 y 2 + 0
T T
0+0+ 15 y 3 ) 2 2 2
y 2 =0 y 1 + 3 y 2 +15 y 3
y3
2 2 2
Therefore, 0 y 1 +3 y 2 +15 y 3 is the required canonical form of the given quadratic form
Index (s): 2; Rank (r): 2; Signature: 2s – r = 2
Nature: Positive semi-definite since one Eigen value is zero and the remaining are positive.

Cayley-Hamilton Theorem
Statement: Every square matrix satisfies its own characteristic equation.

Uses:
i) To obtain higher powers of the matrices.
ii) To obtain the inverse of a matrix.

Verify that
A= ( 1 2
)
2 −1 satisfies its own characteristic equation (i.e., Cayley-Hamilton
theorem) and hence find A4, A-1.
Solution:

Let
A= 1 2 (
2 −1 . )
The characteristic equation of A is |A–λI | = 0.

|( 12 −12 )− λ(10 01 )|=0 |4−λ 1 |=0

(where 3 2−λ is an identity matrix)

|(
2 −1 0 λ )
)−(
1 2 λ 0
|=0
(by scalar multiplication)
|1− λ 2 |=0
2 −1− λ (by matrix subtraction)
(1 – λ)( –1 – λ) – 4 = 0
– 1 – λ + λ + λ2 – 4 = 0
λ2 – 5 = 0
The characteristic equation is λ2 – 5 = 0.
Step 1: Verification
To verify Cayley-Hamilton theorem, we have to prove that A2 – 5I = 0

A 2 −5 I= ( )( ) ( )
1 2 1 2
2 −1 2 −1
−5
1
0
0
1
= ( )( )( )( )( )
1+4 2−2 5 0
2−2 4+1
−
0 5
=
5 0 5 0 0 0
− =
0 5 0 5 0 0
A 2 −5 I= 0 0
0 0( )
Hence A satisfies its own characteristic equation.
Step 2: Find A4
Multiply A2 – 5I = 0 by A2, we get A4 – 5A2 = 0
Then A4 = 5A2

(50 05 )=(250 250 )

=5⋅

A =(
0 25 )
425 0

Step 3: Find A-1

Pre-multiply A2 – 5I = 0 by A-1, we get
A-1A2 – 5A-1 = 0
(A-1A)·A – 5A-1 = 0
I·A – 5A-1 = 0
A – 5A-1 = 0
5A-1 = A
A-1 = 1/5 · A
 1
(
A−1 = 1 2
5 2 −1 )
Therefore,

( )
1 3 0
A= 4 −1 5
Examine whether the matrix 6 −2 2 satisfies its own characteristic equation (i.e.,
Cayley-Hamilton theorem) and hence find A4.
Solution:

( )
1 3 0
A= 4 −1 5
Let 6 −2 2 .
Step 1: Find characteristic equation
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.
D1 = 1 – 1 + 2 = 2
−1 5 1 0 1 3
D2 =| |+| |+| |=( −2−(−10 ) ) + (2−0 ) + (−1−12 )=(−2+10 ) +2−13=8+2−13=−3
−2 2 6 2 4 −1

D3 =1⋅|−1 5 |−3⋅|4 5 |+0⋅|4 −1 |=1⋅(−2−(−10 ) )−3⋅( 8−30 ) =(−2+10 )−3 (−22 ) =8+66=74
−2 2 6 2 6 −2
λ3 – 2λ2 + (–3)λ – 74= 0
The characteristic equation of A is λ3 – 2λ2 – 3λ – 74 = 0.
Step 2: Verification
To verify Cayley-Hamilton theorem, we have to prove that A3 – 2A2 – 3A – 74I = 0.

( )( )( )( )
1 3 0 1 3 0 1+12+0 3−3+0 0+15+0 13 0 15
2
A = A⋅A= 4 −1 5 4 −1 5 = 4−4 +30 12+1−10 0−5+10 = 30 3 5
6 −2 2 6 −2 2 6−8+12 18+2−4 0−10+4 10 16 −6

( )( ) ( )( )
13 0 15 1 3 0 13+0+90 39+0−30 0+0+30 103 9 30
3 2
A = A ⋅A= 30 3 5 4 −1 5 = 30+12+30 90−3−10 0+15+10 = 72 77 25
10 16 −6 6 −2 2 10+64−36 30−16+12 0+80−12 38 26 68
 ( )( )( ) ( )
103 9 30 13 0 15 1 3 0 1 0 0
3 2
A −2 A −3 A−74 I= 72 77 25 −2⋅ 30 3 5 −3⋅ 4 −1 5 −74⋅ 0 1 0
38 26 68 10 16 −6 6 −2 2 0 0 1

( )( )( )( )
103 9 30 26 0 30 3 9 0 74 0 0
= 72 77 25 − 60 6 10 − 12 −3 15 − 0 74 0
38 26 68 20 32 −12 18 −6 6 0 0 74

( )( )
103−26−3−74 9−0−9−0 30−30−0−0 0 0 0
= 72−60−12−0 77−6−(−3 )−74 25−10−15−0 = 0 0 0
38−20−18−0 26−32−(−6 ) −0 68−(−12 )−6−74 0 0 0
Hence, Cayley-Hamilton theorem is verified.
Step 2: Find A4
Multiply A by A3 – 2A2 – 3A – 74I, we get A4 – 2A3 – 3A2 – 74A = 0.

A 4 =2 A3 +3 A 2 +74 A

( )( ) ( )
103 9 30 13 0 15 1 3 0
=2⋅ 72 77 25 +3⋅ 30 3 5 +74⋅ 4 −1 5
38 26 68 10 16 −6 6 −2 2

( )( )( )
206 18 60 39 0 45 74 222 0
= 144 154 50 + 90 9 15 + 296 −74 370
76 52 136 30 48 −18 444 −148 148

( )( )
206+39+74 18+0+222 60+45+0 319 240 105
= 144+90+296 154+9−74 50+15+370 = 530 89 435
76+30+444 52+48−148 136−18+148 550 −48 266

( )
319 240 105
4
A = 530 89 435
550 −48 266

( )
8 −8 2
A= 4 −3 −2
Determine A-1 and A4 for the given matrix 3 −4 1 using Cayley-Hamilton
theorem.
Solution:
 ( )
8 −8 2
A= 4 −3 −2
Let 3 −4 1 .
Step 1: Find characteristic equation
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.
D1 = 8 – 3 + 1 = 6

D2 =|−3 −2 |+|8 2|+|8 −8 |= (− 3−8 )+ ( 8−6 )+ (−24−(−32 ) )=−11+2+8=−1

−4 1 3 1 4 −3
−3 −2 4 −2 4 −3
D3 =8⋅| |−(−8 )⋅| |+2⋅| |=8⋅(−3−8 ) +8⋅( 4−(−6 ) ) +2⋅(−16−(−9 ))
−4 1 3 1 3 −4
=8⋅(−11 )−8⋅10+2⋅(−7 ) =−88+80−14=−22
λ3 – 6λ2 + (–1)λ – (–22) = 0
The characteristic equation of A is λ3 – 6λ2 – λ + 22 = 0.
Step 2: Verification
To verify Cayley-Hamilton theorem, we have to prove that A3 – 6A2 – A + 22I = 0.

( )( )( )( )
8 −8 2 8 −8 2 64−32+6 −64 +24−8 16+16+2 38 −48 34
2
A = A⋅A= 4 −3 −2 4 −3 −2 = 32−12−6 −32+9+8 8+6−2 = 14 −15 12
3 −4 1 3 −4 1 24−16+3 −24 +12−4 6 +8+1 11 −16 15

( )( )( )
38 −48 34 8 −8 2 304−192+102 −304 +144−136 76+96+34
3 2
A = A ⋅A= 14 −15 12 4 −3 −2 = 112−60+36 −112+45−48 28+30+12
11 −16 15 3 −4 1 88−64 +45 −88+48−60 22+32+15

( )
214 −296 206
= 88 −115 70
69 −100 69
 ( ) ( )( ) ( )
214 −296 206 38 −48 34 8 −8 2 1 0 0
3 2
A −6 A − A+22 I= 88 −115 70 −6⋅ 14 −15 12 − 4 −3 −2 +22⋅ 0 1 0
69 −100 69 11 −16 15 3 −4 1 0 0 1

( )( )( )( )
214 −296 206 −228 288 −204 −8 8 −2 22 0 0
= 88 −115 70 + −84 90 −72 + −4 3 2 + 0 22 0
69 −100 69 −66 96 −90 −3 4 −1 0 0 22

( )( )
214−228−8+22 −296 +288+8+0 206−204−2+0 0 0 0
= 88−84−4+0 −115+90+3+22 70−72+2+0 = 0 0 0
69−66−3+0 −100+96+4+0 69−90−1+22 0 0 0

( )
0 0 0
3 2
A −6 A − A+22 I= 0 0 0
0 0 0
Hence, Cayley-Hamilton theorem is verified.
Step 2: Find A-1
Pre-multiply A3 – 6A2 – A +22I by A-1, we get A-1 A3 – 6 A-1 A2 – A-1 A +22 A-1 I = 0.
That is, A2 – 6A – I +22 A-1 = 0.
Hence, 22 A-1 = – (A2 – 6A – I )
1 2
A−1 =− ( A −6 A−I )
22

( ) ( )( )
38 −48 34 8 −8 2 1 0 0
2
A −6 A−I = 14 −15 12 −6⋅ 4 −3 −2 − 0 1 0
11 −16 15 3 −4 1 0 0 1

( )( )( )
38 −48 34 −48 48 −12 −1 0 0
= 14 −15 12 + −24 18 12 + 0 −1 0
11 −16 15 −18 24 −6 0 0 −1

( )( )
38−48−1 −48+48+0 34−12+0 −11 0 22
= 14−24 +0 −15+18−1 12+12+0 = −10 2 24
11−18+0 −16+24 +0 15−6−1 −7 8 8

( )
−11 0 22
2
A −6 A−I = −10 2 24
−7 8 8

( )
−11 0 22
−11
A =− −10 2 24
22
−7 8 8
Step 3: Find A4
Multiply A3 – 6A2 – A +22I by A, we get A4 – 6 A3 – A2 +22 A = 0.

A 4 =6 A 3 + A 2−22 A

( )( ) ( )
214 −296 206 38 −48 34 8 −8 2
=6⋅ 88 −115 70 + 14 −15 12 −22⋅ 4 −3 −2
69 −100 69 11 −16 15 3 −4 1

( )( )( )
1284 −1776 1236 38 −48 34 −176 176 −44
= 528 −690 420 + 14 −15 12 + −88 66 44
414 −600 414 11 −16 15 −66 88 −22

( )( )
1284+38−176 −1776−48+176 1236+34−44 1146 −1648 1226
= 528+14−88 −690−15+66 420+12+44 = 454 −639 476
414+11−66 −600−16+88 414+15−22 359 −528 407

( )
1146 −1648 1226
4
A = 454 −639 476
359 −528 407

( )
1 −1 4
A= 3 2 −1
Verify Cayley-Hamilton theorem for the given matrix 2 1 −1 and hence find its
inverse A-1.
Solution:

( )
1 −1 4
A= 3 2 −1
Let 2 1 −1 .

Step 1: Find characteristic equation

The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.
D1 = 1 + 2 –1 = 2

D2 =|2 −1|+|1 4 |+|1 −1 |=(−2−(−1 ) ) + (−1−8 ) + ( 2−(−3 ) ) =−1−9+5=−5

1 −1 2 −1 3 2
 2 −1 3 −1 3 2
D3 =1⋅| |− (−1 )⋅| |+4⋅| |=(−2− (− 1 ) )+1⋅(−3−(−2 ) ) +4⋅( 3−4 )
1 −1 2 −1 2 1
=−1−1−4=−6
λ3 – 2λ2 – 5λ – (–6) = 0
The characteristic equation of A is λ3 – 2λ2 – 5λ + 6 = 0.
Step 2: Verification
To verify Cayley-Hamilton theorem, we have to prove that A3 – 2A2 – 5A + 6I = 0.

( )( )( )( )
1 −1 4 1 −1 4 1−3+8 −1−2+4 4 +1−4 6 1 1
2
A = A⋅A= 3 2 −1 3 2 −1 = 3+6−2 −3+4−1 12−2+1 = 7 0 11
2 1 −1 2 1 −1 2+3−2 −2+2−1 8−1+1 3 −1 8

( )( )( )( )
6 1 1 1 −1 4 6+3+2 −6+2+1 24−1−1 11 −3 22
3 2
A = A ⋅A= 7 0 11 3 2 −1 = 7+0+22 −7+0+11 28+0−11 = 29 4 17
3 −1 8 2 1 −1 3−3+16 −3−2+8 12+1−8 16 3 5

( )( )( )( )
11 −3 22 6 1 1 1 −1 4 1 0 0
3 2
A −2 A −5 A +6 I= 29 4 17 −2⋅ 7 0 11 −5⋅ 3 2 −1 +6⋅ 0 1 0
16 3 5 3 −1 8 2 1 −1 0 0 1

( )( )( )( )
11 −3 22 −12 −2 −2 −5 5 −20 6 0 0
= 29 4 17 + −14 0 −22 + −15 −10 5 +0 6 0
16 3 5 −6 2 −16 −10 −5 5 0 0 6

( )( )
11−12−5+6 −3−2+5+0 22−2−20+0 0 0 0
= 29−14−15+0 4 +0−10+6 17−22+5+0 = 0 0 0
16−6−10+0 3+2−5+0 5−16+5+6 0 0 0

( )
0 0 0
3 2
A −2 A −5 A +6 I= 0 0 0
0 0 0
Hence, Cayley-Hamilton theorem is verified.
Step 2: Find A-1
Pre-multiply A3 – 2A2 – 5A +6I by A-1, we get A-1 A3 – 2A-1 A2 – 5A-1 A +6 A-1 I = 0.
That is, A2 – 2A – 5I +6 A-1 = 0.
Hence, 6 A-1 = – (A2 – 2A – 5I )
1
A−1 =− ( A 2−2 A−5 I )
6

( )( )( )
6 1 1 1 −1 4 1 0 0
2
A −2 A−5 I= 7 0 11 −2⋅ 3 2 −1 −5⋅ 0 1 0
3 −1 8 2 1 −1 0 0 1

( )( )( )
6 1 1 −2 2 −8 −5 0 0
= 7 0 11 + −6 −4 2 + 0 −5 0
3 −1 8 −4 −2 2 0 0 −5

( )( )
6−2−5 1+2+0 1−8+0 −1 3 −7
= 7−6+0 0−4−5 11+2+0 = 1 −9 13
3−4+0 −1−2+0 8+2−5 −1 −3 5

( )
−1 3 −7
2
A −2 A−5 I= 1 −9 13
−1 −3 5

( )
1 1 7
−
6 2 6

( )
−1 3 −7
1 1 3 13
A−1 =− 1 −9 13 = − −
6 6 2 6
−1 −3 5
1 1 5
−
6 2 6

Application Problems

Find A-1 of
A= 1 −1
2 4 ( ) (
−15 −4
and multiply with the key matrix 78
)
22 to get the coded

message
Solution:

Let
A= 1 −1( )
2 4 . Let the key matrix be
B= −15 −4
78 22 ( )
Step 1: Find A-1
|A| = 4 – (– 2) = 6.

adj A= (4−2 11 )
= (
6 −2 1 )
−1 1 4 1
A

B= ⋅( )( )= ⋅( )= ⋅(
6 108 30 6 18 5 18 5 )
) = ⋅( )=(
1 4 1 −15 −4 1 −60+78 −16+22 1 18 6 6 3 1 3 1
A−1
6 −2 1 78 22 6 30+78 8+22

(183 15 )
The matrix corresponding to the matrix is . Hence, the coded message is “CARE”.

( )
107 87 230
A= 68 49 139
Students received a coded message 99 81 218 from the University to open result

( )
3 2 6
B= 1 1 2
file. It can be cracked by the product of A and the inverse of 2 2 5 . Decode the
password to open the result file.
Solution:
Step 1: Find Characteristic equation

( )
3 2 6
B= 1 1 2
Let 2 2 5
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.
D1 = 3 + 1 + 5 = 9
D 2 =|1 2 |+|3 6 |+|3 2 |=( 5−4 ) + ( 15−12 ) + ( 3−2 )=1+3+1=5
2 5 2 5 1 1
1 2 1 2 1 1
D3 =3⋅| |−2⋅| |+6⋅| |=3⋅( 5−4 ) −2⋅( 5−4 )+6⋅( 2−2 )=3−2+0=1
2 5 2 5 2 2
λ3 – 9λ2 + 5λ – 1 = 0
The characteristic equation of B is λ3 – 9λ2 + 5λ – 1 = 0.
Step 2: Find AB-1
Pre-multiply B3 – 9B2 + 5B – I by B-1, we get B-1 B3 – 9B-1 B2 + 5B-1 B – B-1 I = 0.
That is, B2 – 9B + 5I – B-1 = 0.
Hence, B-1 = (B2 – 9B + 5I )

( )( )( )
23 20 52 3 2 6 1 0 0
−1 2
B =B −9 B+5 I= 8 7 18 −9⋅ 1 1 2 +5⋅ 0 1 0
18 16 41 2 2 5 0 0 1

( )( )( )
23 20 52 −27 −18 −54 5 0 0
=8 7 18 + −9 −9 −18 +0 5 0
18 16 41 −18 −18 −45 0 0 5

( )( )
23−27+5 20−18+0 52−54+0 1 2 −2
= 8−9+0 7−9+5 18−18+0 = −1 3 0
18−18+0 16−18+0 41−45+5 0 −2 1

( )
1 2 −2
−1
B = −1 3 0
0 −2 1

( )( )( )
107 87 230 1 2 −2 107−87+0 214+261−460 −214+0+230
−1
AB = 68 49 139 −1 3 0 = 68−49+0 136+147−278 −136 +0+139
99 81 218 0 −2 1 99−81+0 198+243−436 −198+0+218

( )
20 15 16
−1
AB = 19 5 3
18 5 20
The coded message is “TOPSECRET”
 ( )
2 1 1
A= 0 1 0
Verify Cayley-Hamilton theorem for the matrix 1 1 2 and hence evaluate the
matrix equation, A8 – 5A7+ 7A6 – 3A5 + A4 – 5A3 – 8A2 + 2A – I = 0.
Solution:

( )
2 1 1
A= 0 1 0
Let 1 1 2 .
Step 1: Find characteristic equation
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.
D1 = 2 + 1 + 2 = 5
1 0 2 1 2 1
D2 =| |+| |+| |=( 2−0 )+ ( 4−1 ) + ( 2−0 )=2+3+2=7
1 2 1 2 0 1

D3 =2⋅|1 0 |−1⋅|0 0|+1⋅|0 1 |=2⋅( 2−0 )−0+1 ( 0−1 ) =4−1=3

1 2 1 2 1 1
λ3 – 5λ2 + 7λ – 3 = 0
The characteristic equation of A is λ3 – 5λ2 + 7λ – 3 = 0.
Step 2: Verification
To verify Cayley-Hamilton theorem, we have to prove that A3 – 5A2 + 7A – 3I = 0.

( )( ) ( )( )
2 1 1 2 1 1 4+0+1 2+1+1 2+0+2 5 4 4
2
A = A⋅A= 0 1 0 0 1 0 = 0+0+0 0+1+0 0+0+0 = 0 1 0
1 1 2 1 1 2 2+0+2 1+1+2 1+0+4 4 4 5

( )( ) ( )( )
5 4 4 2 1 1 10+0+4 5+4+4 5+0+8 14 13 13
3 2
A = A ⋅A= 0 1 0 0 1 0 = 0+0+0 0+1+0 0+0+0 = 0 1 0
4 4 5 1 1 2 8+0+5 4+4+5 4+0+10 13 13 14
 ( )( )( )( )
14 13 13 5 4 4 2 1 1 1 0 0
3 2
A −5 A +7 A−3 I= 0 1 0 −5⋅ 0 1 0 +7⋅ 0 1 0 −3⋅ 0 1 0
13 13 14 4 4 5 1 1 2 0 0 1

( )( )( )( )
14 13 13 −25 −20 −20 14 7 7 −3 0 0
=0 1 0 +0 −5 0 + 0 7 0 + 0 −3 0
13 13 14 −20 −20 −25 7 7 14 0 0 −3

( )( )
14−25+14−3 13−20+7−3 13−20+7−3 0 0 0
=0 1−5+7−3 0 =0 0 0
13−20+7+0 13−20+7+0 14−25+14−3 0 0 0
Hence, Cayley-Hamilton theorem is verified. That is, A3 – 5A2 + 7A – 3I = 0. (1)
Now A8 – 5A7+ 7A6 – 3A5 + A4 – 5A3 – 8A2 + 2A – I
= A8 – 5A7+ 7A6 – 3A5 + A4 – 5A3 + 7A2 – 15A2 + 5A – 3A – I
= A5 (A3 – 5A2+ 7A – 3I) + A (A3 – 5A2+ 7A – 3I) – 15A2 + 5A – I
= A5 (0) + A (0) – 15A2 + 5A – I

( ) ( )( )
5 4 4 2 1 1 1 0 0
2
−15 A +5 A−I=−15⋅ 0 1 0 +5⋅ 0 1 0 − 0 1 0
4 4 5 1 1 2 0 0 1

( )( )( )
−75 −60 −60 10 5 5 −1 0 0
=0 −15 0 + 0 5 0 + 0 −1 0
−60 −60 −75 5 5 10 0 0 −1

( )( )
−75+10−1 −60+5+0 −60+5+0 −66 −55 −55
=0 −15+5−1 0 =0 −11 0
−60+5+0 −60+5+0 −75+10−1 −55 −55 −66

( ) ( )
−66 −55 −55 6 5 5
=0 −11 0 =−11⋅ 0 1 0
−55 −55 −66 5 5 6

( )
6 5 5
2
−15 A +5 A−I=−11⋅ 0 1 0
5 5 6

( )
6 5 5
8 7 6 5 4 3 2
A −5 A +7 A −3 A + A −5 A −8 A +2 A−I=−11⋅ 0 1 0
5 5 6
The optimization technique separable programming with three variables x1, x2 and x3 is
applicable only when the objective function is in the form Z = f1(x1) + f2(x2) + f3(x3) where
fi(xi) is function in xi, i = 1, 2, 3. Hence, write the objective function
2 2 2
Z =x1 +3 x2 +3 x 3 −2 x 2 x3 in canonical form by using orthogonal transformation.

Solution:
2 2 2
Given quadratic form is Z =x1 +3 x2 +3 x 3 −2 x 2 x3 .
Step: 1 Find the matrix of the quadratic form
The matrix of the quadratic form
x1 x2 x3

( )
Coeff x 1 x 2 Coeff x 1 x 3
Coeff x 21
2 2
x1
Coeff x 2 x 1 Coeff x 2 x 3
A= x 2 Coeff x 22
2 2
x3
Coeff x 3 x 1 Coeff x 3 x 2
Coeff x 23
2 2

( )( )
1 0 0 1 0 0
A= 0 3 −2/2 = 0 3 −1
0 −2/2 3 0 −1 3

( )
1 0 0
A= 0 3 −1
0 −1 3
Step 2: Find characteristic equation
The characteristic equation of A is λ3 – D1λ2 + D2λ – D3 = 0.
D1 = 1 + 3 + 3 = 7
3 −1 1 0 1 0
D2 =| |+| |+| |=( 9−1 )+ (3−0 ) + ( 3−0 ) =8+3+3=14
−1 3 0 3 0 3

D3 =1⋅| 3 −1 |−0+0=9−1=8
−1 3

λ3 – 7λ2 + 14λ – 8 = 0
The characteristic equation of A is λ3 – 7λ2 + 14λ – 8 = 0.
Step 3: Find Eigen values
1 1 –7 14 –8
0 1 –6 8
1 –6 8 0
Therefore, λ = 1 is a root.
8
λ2 – 6λ + 8 = 0
λ2 – 4λ – 2λ + 8 = 0
λ(λ – 4) – 2(λ – 4) = 0 2 4

(λ – 2)(λ – 4) = 0

λ = 2 and λ = 4
The Eigen values are 1, 2 and 4.

Step 4: To find Eigen vectors, solve (A–λI)X = 0

3 2
λ −7 λ +36=0
λ2 − 4 λ −12=0
λ2 − 4 λ −2 λ +8 =0
λ ( λ −4 ) −2 ( λ −4 ) =0
( λ −2 ) ( λ − 4 ) =0

(( ) ( ))( ) ( )
1 0 0 1 0 0 x1 0
0 3 −1 − λ 0 1 0 x2 = 0
0 −1 3 0 0 1 x3 0

Case (i):
When λ = 1, we get from (1)

( ) ( )( ) ( )
1 0 0 λ 0 0 x1 0
0 3 −1 − 0 λ 0 x 2 = 0
0 −1 3 0 0 λ x 0
3
( )( ) ( )
1-λ 0 0 x1 0
0 3−λ −1 x2 = 0 ( 1)
0 −1 3−λ x 0
3

x1 + 0x2 + 0x3 = 0
0

0x1 + 2x2 – x3 = 0
x1 – x2 + 2x3 = 0
0

Consider the last two equations and applying cross rule method, we have

( )( ) ( )
0 0 0 x1 0
0 2 −1 x2 = 0
0 −1 2 x3 0
0x1+0x2+0x3=0
0x1+2x2−x3=0
0x −x +2x =0
Therefore, the Eigen vector corresponding to λ = 1 is 1 2 3 which is the general form where k is a
constant. Here we get infinite number of Eigen vectors by giving different values for k.
x1 x x
= 2 = 3 =k ( say )
2 −1 −1 0 0 2
| | | | | |
−1 2 2 0 0 −1
x1 x2 x3
= = =k
4−1 0 0
x1 x2 x3
= = =k
3 0 0
The Eigen vector is x1=3k ;x2=0;x3 =0 (Put k = 1/3).
Case (ii):
When λ = 2, we get from (1)

( )
3 k
0
0
 ( )
1
X 1= 0
0
x1+ 0x2 + 0x3 = 0
–

0x1 + x2 – x3 = 0
x1 – x2 + x3 = 0
0

From – x1+ 0x2 + 0x3 = 0, we get x1 = 0.

From the last equations, we get x2 = x3.

−x1+0x2+0x3=0
−x1+0x2+0x3=0
0x1+x2−x3=0
0x −x +x =0
Therefore, the Eigen vector corresponding to λ = 2 is 1 2 3 . That is,

Case (iii):
When λ = 4, we get from (1)

( )
0
1
1

( )
0
X 2= 1
1
x1+ 0x2 + 0x3 = 0
–3

0x1 – x2 – x3 = 0
x1 – x2 – x3 = 0
0

From – 3x1+ 0x2 + 0x3 = 0, we get x1 = 0.

From the last equations, we get x2 = – x3.
 −3x1+0x2+0x3=0
−3x1+0x2+0x3=0
0x1−x2−x3=0
0x −x −x =0
Therefore, the Eigen vector corresponding to λ = 4 is 1 2 3 . That is,
Conclusion:
Characteristic equation Eigen values Eigen vectors

()
λ=1 0
X 3= 1
−1

3 2
λ −7λ +36=0
λ3 – 7λ2 + 14λ – 8 = 0 λ=2

()
λ=4 −1
X 1= 0
1

Check orthogonal

()
0
X T1 X 2 =( 1 0 0 ) 1 =0
1

()
0
X T2 X 3 =( 0 1 1 ) 1 =0+1−1=0
−1

()
0
X T1 X 3 =( 1 0 0 ) 1 =0
−1
Therefore, Eigen vectors are orthogonal to each other. Hence, the normalized modal matrix, N is
as follows.
 ( )
X1 X2 X3

1 0 0
√ 12 +02 +02 √ 02 +12 +12 √ 02 +12 + (−1 )
2

0 1 1
N=
√ √ √ 0 +1 + (−1 )
2 2 2 2 2 2 2 2 2
1 +0 +0 0 +1 +1
0 1 −1
√ 12 +02 +02 √ 02 +12 +12 √ 0
2
+1
2
+ ( −1 )
2

( )
1 0 0
1 1
0
N= √ 2 √ 2 =N T
1 1
0 −
√2 √2

( )( )
0 0 1
1 0 0 0 1 1
AN = 0 3 −1 √2 √2
0 −1 3 1 1
0 −
√2 √2 (Don’t change the order. That is, don’t take NA)

( )( )
1+0+0 0 0 1 0 0
3 1 3 1 2 4
0 − 0+ + 0
AN = √2 √2 √2 √2 = √2 √2
1 3 1 3 2 4
0 0− + 0− − 0 −
√2 √2 √ 2 √2 √2 √2

( )( ) ( )
0 1
0 1 0 0 1 0 0
1 1 2 4 2 2 4 4
T 0 0 0 0+ + 0+ −
N AN = √2 √2 √2 √2 = 2 2 2 2
1 1 2 4 2 2 4 4
0 − 0 − 0 0+ − 0+ +
√2 √2 √2 √2 2 2 2 2
 ( )( )
1 0 0 1 0 0
T
N AN = 0 1+1 0 = 0 2 0
0 0 2+2 0 0 4

( )
1 0 0
T
N AN = 0 2 0 =D
0 0 4 (i.e., a diagonal matrix)

Therefore, A is reduced to a diagonal matrix by means of orthogonal transformation.

Find Canonical form YT(NTAN)Y

( )( )
1 0 0 y1
Y ( N AN ) Y =( y 1
T T
y2 y3 ) 0 2 0 y2
0 0 4 y3

()
y1
Y ( N AN ) Y =( y 1 +0 y 2 + 0 y 3
T T
0+2 y 2 + 0 0+0+ 4 y 3 ) y 2 = y 21 +2 y 22 + 4 y 23
y3
2 2 2
Therefore, y 1 +2 y 2 + 4 y 3 is the required canonical form of the given quadratic form
Index (s): 3; Rank (r): 3; Signature: 2s – r = 3
Nature: Positive definite since all Eigen values are positive.