MODULE:1

Eigen Vector
Matrix Methods to Linear Differential Equations.
𝑥1
Matrix The vector 𝑋 = (𝑥2 ) satisfying the equation (𝐴 − 𝜆𝐼)𝑋 = 0 is called an
𝑥3
A matrix is a rectangular array of numbers arranged in ′𝑚′ horizontal rows
Eigen vector.
and ′𝑛′ vertical columns.
The numbers are called entries of the matrix and such a matrix is said to be Working rule to find the Characteristic Equation
dimension 𝑚 × 𝑛 (denoted by 𝑚 by 𝑛). If 𝐴 is a square matrix of order 3, then its characteristic equation is
𝑎11 𝑎12 ⋯ 𝑎1𝑛 |𝐴 − 𝜆𝐼| = 0
𝑎21 𝑎22 ⋯ 𝑎12
𝐴=( ) 𝑎11 − 𝜆 𝑎12 𝑎13
⋮ ⋮ ⋮
𝑎𝑚1 𝑎𝑚2 ⋯ 𝑎𝑚𝑛 i. e. , | 𝑎21 𝑎22 − 𝜆 𝑎23 | = 0
𝑎31 𝑎32 𝑎33 − 𝜆
Here 𝐴 is a matrix of order 𝑚𝑛. It has 𝑚 rows and 𝑛 columns. Each of 𝑚𝑛
Expand the determinant and we get the characteristic equation (cubic
numbers is called an element of the matrix.
equation)
Eigen Values and Eigen Vectors (Or)
Let 𝐴 be the given square matrix. If there exist a non-zero column vector 𝑋 The characteristic equation can be written as𝜆3 − 𝑆1 𝜆2 + 𝑆2 𝜆 − 𝑆3 = 0,
and a scalar 𝜆 such that 𝐴𝑋 = 𝜆𝑋, then 𝜆 is called an Eigen value and 𝑋 is where
called an Eigen vector corresponding to the Eigen values 𝜆. 𝑆1 = sum of the principal diagonal elements
𝑆2 = sum of the minors of principal diagonal elements
Characteristic Equation 𝑆3 = determinant of 𝐴(or)|𝐴|
The equation |𝐴 − 𝜆𝐼| = 0 is called the characteristic equation of the matrix
𝐴. Note
If 𝐴 is a square matrix of order 2, then its characteristic equation can be
Eigen Value written as 𝜆2 − 𝑆1 𝜆 + 𝑆2 = 0
Solving the characteristic equation |𝐴 − 𝜆𝐼| = 0 we get the values of 𝜆 and Where 𝑆1 = Sum of main diagonal elements
these values are called Eigen values or Latent roots or Characteristic roots. 𝑆2 = Determinant of 𝐴 or |𝐴|

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 i. e. , λ = −2, (λ − 3)(λ − 6) = 0
∴ The eigenvalues are λ = −2, 3, 6

Illustrative Examples To find the eigenvector

𝑥1
Example 1 : Find the eigenvalues and eigenvectors of the matrix 𝐴 = 𝑥
(𝐴
Let − 𝜆𝐼)𝑋 = 0, where 𝑋 = ( 2 )
1 1 3 𝑥3
(1 5 1) 1−λ 1 3 𝑥1 0
3 1 1 ( 1 5−λ 1 ) (𝑥 2 ) = (0)
1 1 3 3 1 1−λ 𝑥 3 0
Solution :Given A = (1 5 1)
(1 − 𝜆)𝑥1 + 1𝑥2 + 3𝑥3 = 0
3 1 1
1𝑥1 + (5 − 𝜆)𝑥2 + 1𝑥3 = 0} − − − (1)
The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where 3𝑥1 + 1𝑥2 + (1 − 𝜆)𝑥3 = 0
𝑆1 = sum of the principal diagonal elements
Case (i) :λ = −2
=1+5+1=7
(1) becomes, 3𝑥1 + 1𝑥2 + 3𝑥3 = 0
𝑆2 = sum of the minors of principal diagonal elements
1𝑥1 + 7𝑥2 + 1𝑥3 = 0
5 1 1 3 1 1
=| |+| |+| | Solving by the rule of cross-multiplication, we have
1 1 3 1 1 5
= (5 − 1) + (1 − 9) + (5 − 1) = 4 − 8 + 4 = 0 1 3 3 1
1 1 3
S3 = |A| = |1 5 1| 7 1 1 7
3 1 1 𝑥1 𝑥2 𝑥3
= =
= 1(5 − 1) − 1(1 − 3) + 3(1 − 15) 1 − 21 3 − 3 21 − 1
= 1(4) − 1(−2) + 3(−14) 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
= = ⇒ = =
−20 0 20 −1 0 1
= 4 + 2 − 42 = −36
−1
∴ The characteristic equation is λ3 − 7λ2 + 0λ + 36 = 0 𝑋1 = ( 0 )
1
Solving by synthetic division method, we get
Case (ii) : λ = 3
–2 1 –7 0 36
(1) becomes, −2𝑥1 + 1𝑥2 + 3𝑥3 = 0
0 –2 18 –36
1𝑥1 + 2𝑥2 + 1𝑥3 = 0
1 –9 18 0
Solving by the rule of cross-multiplication, we have
i. e. , λ = −2, λ2 − 9λ + 18 = 0

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 1 3 −2 1 0 1 1
Solution : Given A = (1 0 1)
1 1 0
2 1 1 2
The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where
𝑥1 𝑥2 𝑥3
= = 𝑆1 = sum of the principal diagonal elements
1 − 6 3 + 2 −4 − 1
𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3 =0+0+0=0
= = ⇒ = =
−5 5 −5 −1 1 −1
𝑆2 = sum of the minors of principal diagonal elements
−1
𝑋2 = ( 1 ) 0 1 0 1 0 1
=| |+| |+| |
−1 1 0 1 0 1 0

Case (iii) : λ = 6 = (0 − 1) + (0 − 1) + (0 − 1) = −1 − 1 − 1 = −3

(1) becomes, −5𝑥1 + 1𝑥2 + 3𝑥3 = 0 S3 = |A|

0 1 1
1𝑥1 − 𝑥2 + 1𝑥3 = 0
= |1 0 1|
Solving by the rule of cross-multiplication, we have 1 1 0
1 3 −5 1 = 0(0 − 1) − 1(0 − 1) + 1(1 − 0)
=1+1=2
−1 1 1 −1 ∴ The characteristic equation is λ3 − 0λ2 − 3λ − 2 = 0
𝑥1 𝑥2 𝑥3
= = Solving by synthetic division method, we get
1+3 3+5 5−1
𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3 –1 1 0 –3 –2
= = ⇒ = =
4 8 4 1 2 1 0 –1 1 2
1
𝑋3 = (2) 1 –1 –2 0
1 i. e. , λ = −1, λ2 − λ − 2 = 0
−1 −1 1
i. e. , λ = −1, (λ + 1)(λ − 2) = 0
Therefore, the eigen vectors are 𝑋1 = ( 0 ), 𝑋2 = ( 1 ), 𝑋3 = (2)
1 −1 1 ∴ The eigenvalues are λ = −1, −1, 2
To find the eigenvector
Example 2 : Find the eigenvalues and eigenvectors of the matrix A = 𝑥1
0 1 1 𝑥
Let (𝐴 − 𝜆𝐼)𝑋 = 0, where 𝑋 = ( 2 )
(1 0 1) 𝑥3
1 1 0

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 −λ 1 1 𝑥1 0 (1) becomes, − 2𝑥1 + 1𝑥2 + 1𝑥3 = 0
( 1 −λ 1 ) (𝑥2 ) = (0)
1 1 −λ 𝑥3 0 1𝑥1 − 2𝑥2 + 1𝑥3 = 0
−𝜆𝑥1 + 1𝑥2 + 1𝑥3 = 0 Solving by the rule of cross-multiplication, we have
1𝑥1 − 𝜆𝑥2 + 1𝑥3 = 0 } − − − (1) 1 1 −2 1
1𝑥1 + 1𝑥2 − 𝜆𝑥3 = 0
Case (i) :λ = −1 −2 1 1 −2
(1) becomes, 𝑥1 + 𝑥2 + 𝑥3 = 0 𝑥1 𝑥2 𝑥3
= =
𝑥1 + 𝑥2 + 𝑥3 = 0 1+2 1+2 4−1
𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
𝑥1 + 𝑥2 + 𝑥3 = 0 = = ⇒ = =
3 3 3 1 1 1
All the three equations reduce to one and the same equation 𝑥1 + 𝑥2 + 1
𝑥3 = 0. 𝑋3 = (1)
1
Therefore, take any one equation and solve. i.e., 𝑥1 + 𝑥2 + 𝑥3 = 0. −1 −1 1
Let 𝑥3 = 0 Therefore, the eigen vectors are 𝑋1 = ( 1 ), 𝑋2 = ( 0 ), 𝑋3 = (1)
𝑥1 𝑥2 0 1 1
𝑥1 + 𝑥2 = 0 ⇒ 𝑥1 = −𝑥2 ⇒ = Note : Though two of the eigenvalues are equal, the eigen vectors
−1 1
𝑥1 𝑥2 𝑥3 𝑋1 , 𝑋2 , 𝑋3 are found to be linearly independent.
⇒ = =
−1 1 0
−1
Example 3 : Find the eigenvalues and eigenvectors of the following matrix
𝑋1 = ( 1 )
0 2 −2 2
A = (1 1 1)
Similarly, for Case (ii)
1 3 −1
Let 𝑥2 = 0 2 −2 2
𝑥1 𝑥3 Solution :Given A = (1 1 1)
𝑥1 + 𝑥3 = 0 ⇒ 𝑥1 = −𝑥3 ⇒ = 1 3 −1
−1 1
𝑥1 𝑥2 𝑥3 The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where
⇒ = =
−1 0 1 𝑆1 = sum of the principal diagonal elements
−1
𝑋2 = ( 0 ) =2+1−1=2
1 𝑆2 = sum of the minors of principal diagonal elements
Case (iii) :λ = 2

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 =|
1 1
|+|
2 2
|+|
2 −2
| Case (i) : λ = 2
3 −1 1 −1 1 1
(1) becomes, 0𝑥1 − 2𝑥2 + 2𝑥3 = 0
= (−1 − 3) + (−2 − 2) + (2 + 2)
1𝑥1 − 1𝑥2 + 1𝑥3 = 0
= −4 − 4 + 4
Solving, by the rule of cross-multiplication, we have
= −4
−2 2 0 −2
S3 = |A|
2 −2 2
= |1 1 1|
−1 1 1 −1
1 3 −1 𝑥1 𝑥2 𝑥3
= =
= 2(−1 − 3) + 2(−1 − 1) + 2(3 − 1) −2 + 2 2 − 0 0 + 2
𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
= −8 − 4 + 4 = −8 = = ⇒ = =
0 2 2 0 1 1
∴ The characteristic equation is λ3 − 2λ2 − 4λ + 8 = 0 0
Solving by synthetic division method, we get 𝑋1 = (1)
1
2 1 −2 −4 8 0
0 2 0 −8 Also, 𝑋2 = (1)
1
1 0 −4 0 Case (ii) :λ = −2
2
i. e. , λ = 2, λ − 4 = 0 (1) becomes, 4𝑥1 − 2𝑥2 + 2𝑥3 = 0
i. e. , λ = 2, (λ + 2)(λ − 2) = 0 1𝑥1 + 3𝑥2 + 1𝑥3 = 0
∴ The eigenvalues are λ = 2, 2, −2 Solving, by the rule of cross-multiplication, we have
To find the eigenvector −2 2 4 −2
𝑥1
Let (𝐴 − 𝜆𝐼)𝑋 = 0, where 𝑋 = (𝑥2 )
𝑥3 3 1 1 3
𝑥1 𝑥1 𝑥2 𝑥3
2−λ −2 2 0 = =
( 1 1−λ 1 ) ( 𝑥2 ) = ( 0) −2 − 6 2 − 4 12 + 2
1 3 −1 − λ 𝑥3 0 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
= = ⇒ = =
(2 − 𝜆)𝑥1 − 2𝑥2 + 2𝑥3 = 0 −8 −2 14 −4 −1 7
1𝑥1 + (1 − 𝜆)𝑥2 + 1𝑥3 = 0 } − − − (1) −4
1𝑥1 + 3𝑥2 + (−1 − 𝜆)𝑥3 = 0 𝑋3 = (−1)
7
Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 0 0 −4 𝑖. 𝑒., 𝜆 = 2, (𝜆 − 2)(𝜆 − 2) = 0
Therefore, the eigen vectors are 𝑋1 = (1), 𝑋2 = (1), 𝑋3 = (−1)
1 1 7 ∴ The Eigen Values are 𝜆 = 2, 2, 2

Note : Two eigenvalues are equal and the eigenvectors are linearly To find the eigen vector
𝑥1
dependent.
Let (𝐴 − 𝜆𝐼)𝑋 = 0, where 𝑋 = (𝑥2 )
𝑥3
Example 4 : Find the eigenvalues and eigenvectors of the matrix 𝐴 = 2−λ 1 0 𝑥1 0
2 1 0 ( 0 2−λ 1 ) (𝑥2 ) = (0)
(0 2 1) 0 0 2 − λ 𝑥3 0
0 0 2 (2 − λ)𝑥1 + 1. 𝑥2 + 0. 𝑥3 = 0
2 1 0 0. 𝑥1 + (2 − λ)𝑥2 + 1. 𝑥3 = 0 } − − − (1)
Solution :Given 𝐴 = (0 2 1) 0. 𝑥1 + 0. 𝑥2 + (2 − λ). 𝑥3 = 0
0 0 2
Case (i) :λ = 2
The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where
(1) becomes, 0. 𝑥1 + 1. 𝑥2 + 0. 𝑥3 = 0
𝑆1 = sum of the principal diagonal elements
0. 𝑥1 + 0. 𝑥2 + 1. 𝑥3 = 0
=2+2+2=6
Solving, by the rule of cross-multiplication, we have
𝑆2 = sum of the minors of principal diagonal elements
2 1 2 0 2 1 1 0 0 1
=| |+| |+| | = (4 − 0) + (4 − 0) + (4 − 0) = 12
0 2 0 2 0 2
S3 = |𝐴| 0 1 0 0
2 1 0 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
= = ⇒ = =
= |0 2 1| 1−0 0−0 0−0 1 0 0
0 0 2 1
= 2(4 − 0) − 1(0 − 0) + 0(0 − 0) = 8 𝑋1 = (0) Corresponding to the eigen value 𝜆 = 2.
0
∴ The characteristic equation is λ3 − 6λ2 + 12λ − 8 = 0
1
Solving by synthetic division method, we get ∴ The only eigen vector is 𝑋 = (0) corresponding to the eigen values 𝜆 =
0
2 1 –6 12 –8
2.
0 2 –8 8
1 –4 4 0
2
𝑖. 𝑒., 𝜆 = 2, 𝜆 − 4𝜆 + 4 = 0

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 4 1 3𝑥1 + 1𝑥2 = 0
Example 5 :Find the eigen values and eigen vectors of 𝐴 = ( )
3 2 𝑥1 𝑥2
4 1 Let 3𝑥1 + 1𝑥2 = 0 ⇒ 3𝑥1 = −𝑥2 ⇒ =
Solution :Given 𝐴 = ( ) −1 3
3 2
−1
𝑋2 = ( )
The characteristic equation is 𝜆2 − 𝑆1 𝜆 + 𝑆2 = 0 3
𝑆1 = sum of the principal diagonal elements 1 −1
∴ The eigen vectors are 𝑋1 = ( )&𝑋2 = ( )
1 3
= 4 + 2 = 6
𝑆2 = det 𝐴 = |𝐴| Properties of Eigen Values
4 1 Property : 1 A square matrix 𝐴 and its transpose 𝐴𝑇 have the same eigen
=| |= 8– 3= 5
3 2
∴ The characteristic equation is 𝜆2 − 6𝜆 + 5 = 0 values.

i.e.,(𝜆 − 1)(𝜆 − 5) = 0 Property : 2 The sum of the eigen values of a matrix 𝐴 is equal to the sum

∴ The Eigen Values are 𝜆 = 1, 5 of the principal diagonal elements of 𝐴. (i.e. Trace of 𝐴)

To find the Eigen Vector Property : 3Product of the eigen values of a matrix 𝐴 is equal to the
𝑥1 determinant of 𝐴.
Let (A – λI)X = 0 where 𝑋 = (𝑥 )
2
Property : 4 If 𝜆1 , 𝜆2 , 𝜆3 , … , 𝜆𝑛 are the eigen values of a matrix 𝐴, then
4−λ 1 𝑥1 0
⇒ ( ) (𝑥 ) = ( ) (i) 𝑘𝜆1 , 𝑘𝜆2 , 𝑘𝜆3 , … , 𝑘𝜆𝑛 are the eigen values of the matrix 𝑘𝐴,
3 2−λ 2 0
(4 − λ)𝑥1 + 1𝑥2 = 0 where 𝑘 is a non-zero scalar.
} − − − (1)
3𝑥1 + (2 − λ)𝑥2 = 0 (ii) 𝜆1 𝑝 , 𝜆2 𝑝 , 𝜆3 𝑝 , … , 𝜆𝑛 𝑝are the eigen values of the matrix 𝐴𝑝 , where
Case (i) :𝜆 = 5 𝑝 is a positive integer.
(1) becomes, − 𝑥1 + 1𝑥2 = 0 1 1 1 1
(iii) , , ,…, are the eigen values of the inverse matrix 𝐴−1,
𝜆1 𝜆2 𝜆3 𝜆𝑛
3𝑥1 − 3𝑥2 = 0
provided 𝜆𝑟 ≠ 0. (i.e., A is nonsingular)
Let − 𝑥1 + 1𝑥2 = 0 ⇒ −𝑥1 = −𝑥2
𝑥1 𝑥2 Property : 5 If 𝜆1 , 𝜆2 , … , 𝜆𝑛 are the eigen values of a matrix 𝐴, then 𝜆1 ±
⇒ =
1 1 𝑘, 𝜆2 ± 𝑘, … , 𝜆𝑛 ± 𝑘 are the eigenvalues of the matrix 𝐴 ± 𝑘𝐼.
1 Property : 6 The eigen values of a real symmetric matrix are real.
𝑋1 = ( )
1
Property : 7 If zero is a characteristic root of matrix then the matrix is
Case (ii) :𝜆 = 1
(1) becomes, singular.
3𝑥1 + 1𝑥2 = 0
Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 8 −6 2
Example : 3If 3 and 15 are the two eigen values of 𝐴 = (−6 7 −4),
Properties of Eigen Vectors 2 −4 3
Property : 1 The eigenvectors corresponding to distinct eigenvalues of a find |𝐴| without expanding the determinant.
real symmetric matrix are orthogonal. 8 −6 2
Solution :Given𝐴 = (−6 7 −4)
2 −4 3
Illustrative Examples
Sum of the eigen values
Example : 1 Find the sum and product of the eigenvalues of 𝐴 =
= Sum of the main diagonal elements [by property]
1 0 0
(0 3 −1) If 𝜆 is the third eigen value of 𝐴, then
0 −1 3 3 + 15 + 𝜆 = 8 + 7 + 3
1 0 0
Solution :Given𝐴 = (0 i. e. , 18 + 𝜆 = 18𝜆 = 0
3 −1)
0 −1 3 |𝐴| = Product of eigen values of 𝐴
Sum of the eigen values = (3). (15). (0) = 0
= Sum of the main diagonal elements [by property]
= 1+3+3 = 7 4 6 6
Example : 4Two eigen values of 𝐴 = ( 1 3 2 ) are equal and they
Product of the eigen values = determinant of A = |A| −1 −5 −2
= 1(9 – 1)– 0 – 0 = 8 are double the third. Find the eigen values of 𝐴2.
4 6 6
3 1 4 Solution :Given𝐴 = ( 1 3 2)
Example : 2Find the eigen values 𝐴2 and 𝐴−1 given matrix 𝐴 = (0 2 6) −1 −5 −2
0 0 5 Let the third eigen value be 𝜆. Therefore, the other two eigen values are
3 1 4
Solution :Given matrix 𝐴 = (0 2 6) 2𝜆, 2𝜆
0 0 5 i. e., 𝜆 + 2𝜆 + 2𝜆 = 4 + 3 − 2
Since 𝐴 is a triangular matrix, the eigen values of 𝐴 are 2, 3, 5 (the principal ⇒ 5𝜆 = 5 ⇒ 𝜆 = 1
diagonal elements of A). ∴ The eigen values of 𝐴 are 1, 2, 2
∴ The eigen values of 𝐴 are 2 , 3 , 5 . i.e.,9, 4, 25
2 2 2 2
∴ The eigen values of 𝐴2 are 1, 4, 4
1 1 1
The eigen values of𝐴−1 are , ,
2 3 5

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 1 1 3 6 −2 2
Example : 5If 3 and 6 are the two eigen values of 𝐴 = (1 5 1), write Example : 7The product of two eigen values of 𝐴 = (−2 3 −1) is 16.
3 1 1 2 −1 3
down all the eigen values of 𝐴−1 and 3𝐴. Find the third eigen value.
1 1 3 6 −2 2
Solution : Given 𝐴 = (1 5 1) Solution :Given 𝐴 = (−2 3 −1)
3 1 1 2 −1 3
Let the third eigen value be 𝜆, then by the property Let the third eigen value be 𝜆 ⇒Product of the eigen values = 16𝜆
3+6+𝜆 =1+5+1 Also, by the property the product of the eigen values = det. 𝐴 = |𝐴| = 32
⇒ 𝜆 = 7 − 9 = −2 ∴ 16𝜆 = 32 ⇒ 𝜆 = 2
∴ The eigen values of 𝐴 are 3, 6, – 2 ∴ The third eigen value = 2
1 1 −1
∴ The eigen values of 𝐴−1 are , , and the eigen values of 3𝐴 7 4 −4
3 6 2
Example : 8If – 9 is the one of eigen value of 𝐴 = (4 −8 −1)then find
are 9, 18, – 6
4 −1 −8
the other eigen values.
Example : 6Find the eigen values of 𝐴 – 2𝐼 and 𝐴−1 given that two eigen
7 4 −4
6 −2 2 Solution : Given 𝐴 = (4 −8 −1)
values of 𝐴 = (−2 3 −1) are 2 and 8. 4 −1 −8
2 −1 3
Let 𝜆1 = −9
6 −2 2
Solution :Given 𝐴 = (−2 3 −1) By the property of eigen values, 𝜆1 +𝜆2+𝜆3 = 7 − 8 − 8
2 −1 3 ⇒ −9+𝜆2+𝜆3 = −9
Let the third eigen value be 𝜆, then by the property
⇒ 𝜆2 +𝜆3 = 0 ⇒ 𝜆2 = −𝜆3
2+8+𝜆 =6+3+3
𝜆1 𝜆2 𝜆3 = │𝐴│ [by property]
𝜆 = 12 − 10 = 2
(−9)(−𝜆3)𝜆3 = 441
∴ The eigen values of 𝐴 are 2, 2, 8.
441
∴ The eigen values of 𝐴 − 2𝐼 are 0, 0, −6 and the eigen values of 𝐴−1 are (𝜆3 )2 = = 49
9
1 1 1
, , ⇒ 𝜆3 = ±7
2 2 8
∴ 𝜆1 = −9, 𝜆2 = −7, 𝜆3 = 7

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
Problems for Practise
1. Find the sum and product of the eigen values of the matrix 𝐴 =
Note : 3 By Cayley-Hamilton theorem, we can express any higher powers
1 1 2
(1 2 2) of the matrix as the linear combination of 𝐴2 , 𝐴 & 𝐼.
1 2 3
3 0 0
2. Find the eigen value of A5, where 𝐴 = (5 4 0) Example : 1 Verify Cayley-Hamilton theorem for the matrix 𝐴=
3 6 1 1 3 7
2 3 (4 2 3) and also use it to find 𝐴−1.
3. Find the eigen values of 𝐴 = ( ) corresponding to the eigenvector 1 2 1
0 4
1 1 3 7
( ) Solution :Given𝐴 = (4 2 3)
0
1 2 1
Answers
The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where
1. Sum = 6, Product = 1
𝑆1 = sum of the principal diagonal elements
2. Eigen value of 𝐴5are 35 , 45 , 15
=1+2+1=4
3. 𝜆 = 2
𝑆2 = sum of the minors of principal diagonal elements
2 3 1 7 1 3
Cayley-Hamilton Theorem =| |+| |+| |
2 1 1 1 4 2
Statement : Every square matrix satisfies its own characteristic equation. = (2 − 6) + (1 − 7) + (2 − 12) = −4 − 6 − 10 = −20
This means that if 𝜆𝑛 + 𝑐1 𝜆𝑛−1 + 𝑐2 𝜆𝑛−2 + ⋯ + 𝑐𝑛 = 0 is the characteristic S3 = |A|
𝑛 𝑛−1 𝑛−2
equation of a square matrix 𝐴 of order 𝑛, then 𝐴 + 𝑐1 𝐴 + 𝑐2 𝐴 + ⋯+ 1 3 7
= |4 2 3|
𝑐𝑛 𝐼 = 0
1 2 1
= 1(2 − 6) − 3(4 − 3) + 7(8 − 2)
Note : 1 When 𝜆 is replaced by 𝐴 in the characteristic equation, the
= −4 − 3 + 42
constant term 𝑐𝑛 should be replaced by 𝑐𝑛 𝐼 to get the result of Cayley-
= 35
Hamilton theorem, where 𝐼 is the unit matrix of order 𝑛. Also, 0 in the RHS
∴ The characteristic equation is 𝜆3 − 4𝜆2 − 20𝜆 − 35 = 0.
is a null matrix of order 𝑛.
The Cayley-Hamilton theorem states that
Note : 2 Using Cayley-Hamilton theorem we can find 𝐴 −1
and higher 𝐴3 − 4𝐴2 − 20𝐴 − 35𝐼 = 0 − − − (1)which is to be verified.
positive integral powers of 𝐴 can be computed.

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 1 3 7 1 3 7 20 23 23 1 20 23 23 4 12 28 20 0 0
Now, 𝐴2 = (4 2 3) (4 2 3) = (15 22 37) = [(15 22 37) − (16 8 12) − ( 0 20 0 )]
35
1 2 1 1 2 1 10 9 14 10 9 14 4 8 4 0 0 20
1 3 7 20 23 23 135 152 232 1 −4 11 −5
𝐴3 = 𝐴. 𝐴2 = (4 2 3) (15 22 37) = (140 163 208) = (−1 −6 25)
35
1 2 1 10 9 14 60 76 111 6 1 −10
Substituting these values in (1), we get,
1 2 2
LHS = 𝐴3 − 4𝐴2 − 20𝐴 − 35𝐼 Example : 2 If 𝐴 = (2 1 2) , Prove that 𝐴3 − 3𝐴2 − 9𝐴 − 5𝐼 = 0.
135 152 232 20 23 23 1 3 7 2 2 1
= (140 163 208) − 4 (15 22 37) − 20 (4 2 3) Hence find 𝐴4and 𝐴−1.
60 76 111 10 9 14 1 2 1
1 2 2
1 0 0
Solution :Given 𝐴 = (2 1 2)
−35 (0 1 0)
2 2 1
0 0 1
135 152 232 80 92 92 20 60 140 The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where
= (140 163 208) − (60 88 148) − (80 40 60 ) 𝑆1 = sum of the principal diagonal elements
60 76 111 40 36 56 20 40 20
=1+1+1=3
35 0 0
− ( 0 35 0 ) 𝑆2 = sum of the minors of principal diagonal elements
0 0 35 1 2 1 2 1 2
=| |+| |+| |
0 0 0 2 1 2 1 2 1
= (0 0 0)
= (1 − 4) + (1 − 4) + (1 − 4) = −9
0 0 0
= 0 = RHS S3 = |A|
1 2 2
Thus, Cayley-Hamilton theorem is verified.
= |2 1 2|
To find 𝑨−𝟏 2 2 1
Pre-multiplying (1) by 𝐴−1, we get = 1(1 − 4) − 2(2 − 4) + 2(4 − 2) = −3 + 4 + 4 = 5

𝐴2 − 4𝐴 − 20𝐼 − 35𝐴−1 = 0 ∴ The characteristic equation is 𝜆3 − 3𝜆2 − 9𝜆 − 5 = 0.

1 2 By Cayley-Hamilton theorem, we have 𝐴3 − 3𝐴2 − 9𝐴 − 5𝐼 = 0 − −(1)
𝐴−1 = (𝐴 − 4𝐴 − 20𝐼)
35 To prove 𝐴3 − 3𝐴2 − 9𝐴 − 5𝐼 = 0
1 20 23 23 1 3 7 1 0 0 1 2 2 1 2 2 9 8 8
= [(15 22 37) − 4 (4 2 3) − 20 (0 1 0)] 𝐴2 = 𝐴. 𝐴 = (2 1 2) (2 1 2) = (8 9 8)
35
10 9 14 1 2 1 0 0 1 2 2 1 2 2 1 8 8 9
Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 9 8 8 1 2 2 41 42 42 9 8 8 1 2 2 1 0 0
𝐴3 = 𝐴2 . 𝐴 = (8 9 8) (2 1 2) = (42 41 42) 5𝐴−1 = (8 9 8) − 3 (2 1 2) − 9 (0 1 0)
8 8 9 2 2 1 42 42 41 8 8 9 2 2 1 0 0 1
∴ RHS = 𝐴3 − 3𝐴2 − 9𝐴 − 5𝐼 9 8 8 3 6 6 9 0 0
= (8 9 8) − (6 3 6) − (0 9 0)
41 42 42 9 8 8 1 2 2 1 0 0
8 8 9 6 6 3 0 0 9
= (42 41 42) − 3 (8 9 8) − 9 (2 1 2) − 5 (0 1 0)
42 42 41 8 8 9 2 2 1 0 0 1 −3 2 2
= ( 2 −3 2)
41 42 42 27 24 24 9 18 18 5 0 0
2 2 −3
= (42 41 42) − (24 27 24) − (18 9 18) − (0 5 0)
42 42 41 24 24 27 18 18 9 0 0 5 1 −3 2 2
𝐴−1 = ( 2 −3 2)
0 0 0 5
2 2 −3
= (0 0 0)
0 0 0
Example : 3 Using Cayley-Hamilton theorem, Find the value of 𝐴8 − 𝐴7 +
= 0 = RHS
1 2 −2
∴ 𝐴3 − 3𝐴2 − 9𝐴 − 5𝐼 = 0 − − − (1) 5𝐴6 − 𝐴5 + 𝐴4 − 𝐴3 + 6𝐴2 + 𝐴 − 2𝐼, if 𝐴 = (2 5 −4)
To find 𝑨𝟒 3 7 −5
Multiplying (1) by 𝐴, we get Solution :

𝐴4 − 3𝐴3 − 9𝐴2 − 5𝐴 = 0 The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where

 𝐴4 = 3𝐴3 + 9𝐴2 + 5𝐴 𝑆1 = sum of the principal diagonal elements

41 42 42 9 8 8 1 2 2 =1+5−5=1
= 3 (42 41 42) + 9 (8 9 8) + 5 (2 1 2) 𝑆2 = sum of the minors of principal diagonal elements
42 42 41 8 8 9 2 2 1
5 −4 1 −2 1 2
123 126 126 81 72 72 5 10 10 =| |+| |+| |
= (126 7 −5 3 −5 2 5
123 126) + (72 81 72) + (10 5 10)
126 126 123 72 72 81 10 10 5 = (−25 + 28) + (−5 + 6) + (5 − 4)
209 208 208 =3+1+1=5
= (208 209 208)
208 208 209 S3 = |A|
To find 𝑨−𝟏 1 2 −2
= |2 5 −4|
Multiplying (1) by 𝐴−1, we get 3 7 −5
𝐴2 − 3A − 9𝐼 − 5𝐴−1 = 0 = 1(−25 + 28) − 2(−10 + 12) − 2(14 − 15)
5𝐴−1 = 𝐴2 − 3A − 9𝐼 =3−4+2=1
Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
∴ The characteristic equation is 𝜆3 − 𝜆2 + 5𝜆 − 1 = 0. Problems for Practise
3 2
By Cayley-Hamilton theorem, we have 𝐴 − 𝐴 + 5𝐴 − 𝐼 = 0 − − − (1) 1 2 −2
1. Verify Cayley-Hamilton theorem for 𝐴 = (2 5 −4 ). Hence express
By using the idea of polynomial division,
3 7 −5
𝐴5 + 𝐴 𝐴7 in terms of 𝐴2 , 𝐴 & 𝐼

𝐴3 − 𝐴2 + 5𝐴 − 𝐼 𝐴8 − 𝐴7 + 5𝐴6 − 𝐴5 + 𝐴4 − 𝐴3 + 6𝐴2 + 𝐴 − 2𝐼 3 −1
2. Verify Cayley-Hamilton theorem for the matrix 𝐴 = ( )
−1 5
𝐴8 − 𝐴7 + 5𝐴6 − 𝐴5 1 4
3. Use Cayley-Hamilton theorem to find the inverse of 𝐴 = ( )
2 3
𝐴4 − 𝐴3 + 6𝐴2 + 𝐴
1 3 7
− 2𝐼 4. Verify Cayley-Hamilton theorem for the matrix 𝐴 = (4 2 3) and
𝐴4 − 𝐴3 + 5𝐴2 − 𝐴 1 2 1
also use it to find 𝐴−1
Remainder 𝐴2 + 2𝐴 − 2𝐼
2 −1 2
5. Verify that the matrix 𝐴 = (−1 2 −1) satisfies its characteristic
1 −1 2
∴ 𝐴8 − 𝐴7 + 5𝐴6 − 𝐴5 + 𝐴4 − 𝐴3 + 6𝐴2 + 𝐴 − 2𝐼 = 𝐴2 + 2𝐴 − 2𝐼 − −(1)
equation and hence find 𝐴4
1 2 −2 1 2 −2 −1 −2 0
Now, 𝐴2 = 𝐴. 𝐴 = (2 5 −4) (2 5 −4) = ( 0 1 −4) 6. Use Cayley-Hamilton theorem to find the inverse of 𝐴=
3 7 −5 3 7 −5 2 6 −9 −1 0 3
−1 −2 0 1 2 −2 1 0 0 ( 8 1 −7)
𝐴2 + 2𝐴 − 2𝐼 = ( 0 1 −4) + 2 (2 5 −4) − 2 (0 1 0) −3 0 8
2 6 −9 3 7 −5 0 0 1
−1 −2 0 2 4 − 4 2 0 0 Answers
=( 0 1 −4) + (4 10 − 8 ) − (0 2 0)
1. To verify 𝐴3 − 𝐴2 + 5𝐴 − 𝐼 = 0, 𝐴7 = 67𝐴2 − 73𝐴 + 13𝐼
2 6 −9 6 14 −10 0 0 2
−1 2 − 4 2. To verify 𝐴2 − 8𝐴 + 14𝐼 = 0
=( 4 9 −12) −3
1 4
8 20 −21 3. Characteristic Equation is 2 − 4 − 5 = 0, 𝐴−1 = ( )
2 −1
5

∴ From (1)  −4 11 −5
1
−1 2 − 4 4. To Verify 𝐴3 − 4𝐴2 − 20𝐴 − 35𝐼 = 0, 𝐴−1 = 35 (−1 −6 25)
𝐴8 − 𝐴7 + 5𝐴6 − 𝐴5 + 𝐴4 − 𝐴3 + 6𝐴2 + 𝐴 − 2𝐼 = ( 4 9 −12) 6 1 −10
8 20 −21

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 124 −123 162 Calculation of powers of a matrix
5. To Verify A3 – 6A2 + 8A – 3I = 0, 𝐴4 = ( −95 96 −123 )
95 −95 124 If a square matrix 𝐴 has linearly independent eigen vectors, then by the
6. Characteristic Equation is 𝐴3 − 8𝐴2 + 8𝐴 − 𝐼 = 0 above definition,
𝐷 = 𝑀−1 𝐴𝑀
∴ 𝐴 = 𝑀𝐷𝑀−1
Orthogonal Matrix
𝐴2 = (𝑀𝐷𝑀−1 )(𝑀𝐷𝑀−1 )
A square matrix 𝐴 is called orthogonal if
= 𝑀𝐷(𝑀−1 𝑀)𝐷𝑀−1
𝐴𝐴𝑇 = 𝐴𝑇 𝐴 = 𝐼
= 𝑀𝐷 2 𝑀−1
Also, 𝐴𝐴−1 = 𝐴−1𝐴 = 𝐼
Similarly, 𝐴3 = 𝑀𝐷 3 𝑀−1
Hence, for an orthogonal matrix, 𝐴−1 = 𝐴𝑇
Proceeding, we get 𝐴𝑘 = 𝑀𝐷 𝑘 𝑀−1
Pair wise Orthogonal
Illustrative Examples
If 𝑋1 𝑋2𝑇 = 0, 𝑋2 𝑋3𝑇 = 0 and𝑋3 𝑋1𝑇 = 0, then the eigen vectors 𝑋1 , 𝑋2 , 𝑋3 are
2 2 −7
said to be pair wise orthogonal. Example:1Diagonalise the matrix 𝐴 = (2 1 2 ) by similarity
0 1 −3
Similarity Transformation transformation and hence find 𝐴4.
If 𝐴 and 𝐵 are two square matrices of order 𝑛 then 𝐵 is said to be similar to 2 2 −7
−1 Solution :Given A = (2 1 2)
𝐴, if there exists a non singular matrix 𝑃 such that 𝐵 = 𝑃 𝐴𝑃
0 1 −3
The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where
Diagonalisation of a Matrix
𝑆1 = sum of the principal diagonal elements
Definition : The process of finding a matrix 𝑀 such that 𝑀−1 𝐴𝑀 = 𝐷,
=2+1−3=0
where 𝐷 is a diagonal matrix is called diagonalisation of the matrix 𝐴.
𝑆2 = sum of the minors of principal diagonal elements
Since, 𝑀−1 𝐴𝑀 = 𝐷, 𝐴 and 𝐷 are similar matrices and therefore 𝐴 and 𝐷
1 2 2 −7 2 2
=| |+| |+| |
have the same eigen values. It is clear that the diagonal matrix 𝐷 has eigen 1 −3 0 −3 2 1
values of 𝐴 as its diagonal elements. = (−3 − 2) + (−6 − 0) + (2 − 4)
The transformation 𝑀−1 𝐴𝑀 = 𝐷 is called a similarity transformation. = −5 − 6 − 2 = −13
S3 = |A|

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 2 2 −7 2 −7 1 2
= |2 1 2|
0 1 −3
0 2 2 0
= 2(−3 − 2) − 2(−6 − 0) − 7(2 − 0)
𝑥1 𝑥2 𝑥3
= −10 + 12 − 14 = =
4 − 0 −14 − 2 0 − 4
= −12 𝑥1 𝑥2 𝑥3
= =
4 −16 −4
∴ The characteristic equation is λ3 − 0λ2 − 13λ + 12 = 0
𝑥1 𝑥2 𝑥3
Solving by synthetic division method, we get  = =
1 −4 −1
1 1 0 −13 12 1
𝑋1 = (−4)
0 1 1 −12 −1
1 1 −12 0 Case (ii) :λ = 3
i. e. , λ = 1, λ2 + λ − 12 = 0 (1) becomes, −1𝑥1 + 2𝑥2 − 7𝑥3 = 0
i. e. , λ = 1, (λ − 3)(λ + 4) = 0 2𝑥1 − 2𝑥2 + 2𝑥3 = 0
∴ The eigenvalues are λ = 1,3, −4 Solving, by the rule of cross-multiplication, we have
To find the eigenvector 2 −7 −1 2
𝑥1
𝑥
Let (𝐴 − 𝜆𝐼)𝑋 = 0, where 𝑋 = ( 2 ) −2 2 2 −2
𝑥3
𝑥1 𝑥2 𝑥3
2−λ 2 −7 𝑥1 0 = =
4 − 14 −14 + 2 2 − 4
( 2 1−λ 𝑥
2 ) ( 2 ) = ( 0) 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
0 1 −3 − λ 𝑥3 0 = = ⇒ = =
−10 −12 −2 5 6 1
(2 − 𝜆)𝑥1 + 2𝑥2 − 7𝑥3 = 0
5
2𝑥1 + (1 − 𝜆)𝑥2 + 2𝑥3 = 0 } − − − (1) 𝑋2 = (6)
0𝑥1 + 1𝑥2 + (−3 − 𝜆)𝑥3 = 0 1
Case (i) :λ = 1 Case (iii) :λ = −4
(1) becomes, 1𝑥1 + 2𝑥2 − 7𝑥3 = 0 (1) becomes, 6𝑥1 + 2𝑥2 − 7𝑥3 = 0
2𝑥1 + 0𝑥2 + 2𝑥3 = 0 2𝑥1 + 5𝑥2 + 2𝑥3 = 0
Solving, by the rule of cross-multiplication, we have Solving, by the rule of cross-multiplication, we have

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 2 −7 6 2 14 10 2
(1) Cofactor of 𝑀 = ( −7 5 −6 )
− 28 −10 26
5 2 2 5
14 10 2 𝑇
𝑥1 𝑥2 𝑥3 𝑇
= = Now, adj 𝑀 = (cofactor of 𝑀) = ( −7 5 −6 )
4 + 35 −14 − 12 30 − 4 − 28 −10 26
𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3 14 −7 −28
= = ⇒ = =
39 −26 26 3 −2 2 adj 𝑀 = (10 5 − 10 )
3 2 −6 26
𝑋3 = (−2) 14 −7 −28
1
2 Therefore, 𝑀−1 = (10 5 − 10 )
70
1 5 3 2 −6 26
Hence the modal matrix is 𝑀 = (−4 6 −2). Now, 𝐷 = 𝑀−1 𝐴𝑀
−1 1 2
1 2 2 −7 1 5 3 1 15 −12
We know that, 𝑀−1 = adj 𝑀, where adj 𝑀 = (cofactor of 𝑀)𝑇 𝐴𝑀 = (2 1 2 ) (−4 6 −2) = (−4 18 8 )
|𝑀|
0 1 − 3 −1 1 2 −1 3 −8
|𝑀| = 1(12 + 2) − 5(−8 − 2) + 3(−4 + 6)
1 14 −7 −28 1 15 −12
= 14 + 50 + 6 = 70 𝑀−1 𝐴𝑀 = (10 5 − 10 ) (−4 18 8 )
70
2 −6 26 −1 3 −8
To find cofactor of 𝑴
𝑎11 −𝑎12 𝑎13 1 70 0 0 1 0 0
= ( 0 210 0 )=(0 3 0)
Let 𝑎
Cofactor of 𝑀 = (− 21 𝑎22 −𝑎23 ) − − − (1) 70
0 0 −280 0 0 −4
𝑎31 −𝑎32 𝑎33
𝐴4is given by
6 −2 −4 −2
𝑎11 =| | = 12 + 2 = 14 𝑎12 =| | = −8 − 2 = −10 𝐴4 = 𝑀𝐷 4 𝑀−1
1 2 −1 2
1 0 0 1 14 −7 −28
−4 6 5 3 𝐷 4 𝑀−1 = ( 0 81
𝑎13 = | | = −4 + 6 = 2 𝑎21 = | | = 10 − 3 = 7 0 ) x (10 5 − 10 )
−1 1 1 2 70
0 0 256 2 −6 26
1 3 1 5 1 14 −7 −28
𝑎22 = | | = 2+3 = 5 𝑎23 = | | = 1+5 = 6
−1 2 −1 1 = (810 405 − 810 )
70
512 −1536 6656
5 3 1 3
𝑎31 = | | = −10 − 18 = −28 𝑎32 = | | = −2 + 12 = 10
6 −2 −4 −2 1 5 3 1 14 −7 −28
𝑀𝐷 4 𝑀−1 = (−4 6 −2) x (810 405 − 810 )
70
1 5 −1 1 2 512 −1536 6656
𝑎33 = | | = 6 + 20 = 26
−4 6

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 1 5600 −2590 15890 ILLUSTRATIVE EXAMPLES
= (3780 5530 −18060)
70 8 −6 2
1820 −2660 12530
Example : 1 Reduce the matrix 𝐴 = (−6 7 −4) to a diagonal form by
80 −37 227 2 −4 3
𝐴4 = (54 79 −258)
26 −38 179 orthogonal reduction.
Solution :
Diagonalisation of a Real Symmetric Matrix by Orthogonal The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where
Reduction 𝑆1 = sum of the principal diagonal elements
𝑥1
= 8 + 7 + 3 = 18
Definition : If 𝑋 = ( 2 ) is a column vector, then the normalized vector of
𝑥
𝑥3 𝑆2 = sum of the minors of principal diagonal elements
𝑥1
7 −4 8 2 8 −6
√𝑥12 +𝑥22 +𝑥32
=| |+| |+| |
−4 3 2 3 −6 7
𝑥2
= (21 − 16) + (24 − 4) + (56 − 36)
𝑋= √𝑥12 +𝑥22 +𝑥32
= 5 + 20 + 20 = 45
𝑥3
2 2 2 S3 = |A|
(√𝑥1 +𝑥2 +𝑥3 )
8 −6 2
Let 𝐴 be a real symmetric matrix. Since 𝐴 is a real symmetric matrix, the = |−6 7 −4|
eigen vectors of 𝐴 are not only linearly independent but also pairwise 2 −4 3
orthogonal. = 8(21 − 16) + 6(−18 + 8) + 2(24 − 14) = 40 − 60 + 20 = 0
∴ The characteristic equation is λ3 − 18λ2 + 45λ = 0
Let 𝑁 be the matrix whose columns are the normalized eigen vectors of 𝐴.
i. e. , λ(λ2 − 18λ + 45) = 0
−1 𝑇
Then 𝑁 is an orthogonal matrix and by property 𝑁 =𝑁 .
λ(λ − 3)(λ − 15) = 0
The matrix 𝑁 is called the normalized modal matrix. Then the i. e. , λ = 0, 3, 15
−1 𝑇
transformation 𝐷=𝑁 𝐴𝑁 = 𝑁 𝐴𝑁 is known as orthogonal ∴ The eigenvalues are λ = 0, 3, 15
transformation or orthogonal reduction.Thus, the square matrix 𝐴 can be To find the eigenvector
𝑇
diagonalised by an orthogonal reduction 𝑁 𝐴𝑁 = 𝐷. 𝑥1
Let (𝐴 − 𝜆𝐼)𝑋 = 0, where 𝑋 = (𝑥2 )
𝑥3

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 8 − λ −6 2 𝑥1 0 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
= = ⇒ = =
( −6 7 − λ −4 ) (𝑥2 ) = (0) 16 8 −16 2 1 −2
2 −4 3 − λ 𝑥3 0 2
(8 − 𝜆)𝑥1 − 6𝑥2 + 2𝑥3 = 0 𝑋2 = ( 1 )
−6𝑥1 + (7 − 𝜆)𝑥2 − 4𝑥3 = 0} − − − (1) −2
2𝑥1 − 4𝑥2 + (3 − 𝜆)𝑥3 = 0 Case (iii) :λ = 15
Case (i) :λ = 0 (1) becomes, −7𝑥1 − 6𝑥2 + 2𝑥3 = 0
(1) becomes, 8𝑥1 − 6𝑥2 + 2𝑥3 = 0 −6𝑥1 − 8𝑥2 − 4𝑥3 = 0
−6𝑥1 + 7𝑥2 − 4𝑥3 = 0 Solving, by the rule of cross-multiplication, we have
Solving, by the rule of cross-multiplication, we have −6 2 −7 −6
−6 2 8 −6
−8 −4 −6 −8
7 −4 −6 7 𝑥1 𝑥2 𝑥3
= =
𝑥1 𝑥2 𝑥3 24 + 16 −12 − 28 56 − 36
= = 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
24 − 14 −12 + 32 56 − 36 = = ⇒ = =
𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3 40 −40 20 2 −2 1
= = ⇒ = = 2
10 20 20 1 2 2
𝑋3 = (−2)
1
1
𝑋1 = (2)
2 1 2 2
The eigen vectors are 𝑋1 = (2), 𝑋2 = ( 1 ), 𝑋3 = (−2)
Case (ii) :λ = 3 2 −2 1
(1) becomes, 5𝑥1 − 6𝑥2 + 2𝑥3 = 0 Clearly, 𝑋1 , 𝑋2 , 𝑋3 are pairwise orthogonal.
−6𝑥1 + 4𝑥2 − 4𝑥3 = 0 Consider the matrix 𝑁 whose columns are normalized eigen vectors.
Solving, by the rule of cross-multiplication, we have 1 2 2
−6 2 5 −6 3 3 3
2 1 2
∴ 𝑁= −
3 3 3
4 −4 −6 4 2 2 1
𝑥1 𝑥2 𝑥3 (3 − 3 3 )
= =
24 − 8 −12 + 20 20 − 36

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 1 2 2 1 2 2 −2 1 −7 0 36
3 3 3 3 3 3
2 1 2 8 −6 2 2 1 2 0 −2 18 −36
𝑇
𝑁 𝐴𝑁 = − (−6 7 −4) −
3 3 3 2 −4 3 3 3 3 1 −9 18 0
2 2 1 2 2 1
(3 − (3 − 3 3 ) i. e. , λ = −2, λ2 − 9λ + 18 = 0
3 3 )
0 0 0 i. e. , λ = −2, (λ − 3)(λ − 6) = 0
= (0 3 0)=𝐷
∴ The eigenvalues are λ = −2, 3, 6
0 0 15
Hence 𝐴 is reduced to a diagonal matrix by means of orthogonal reduction. To find the eigenvector
𝑥1
𝑥
Let (𝐴 − 𝜆𝐼)𝑋 = 0, where 𝑋 = ( 2 )
Example : 2 Find an orthogonal matrix which reduces the matrix 𝐴 = 𝑥3
1 1 3 1−λ 1 3 𝑥1 0
(1 5 1) to a diagonal matrix. ( 1 5−λ 𝑥
1 ) ( 2 ) = (0)
3 1 1 3 1 1 − λ 𝑥3 0
Solution : (1 − 𝜆)𝑥1 + 1𝑥2 + 3𝑥3 = 0
3 2
The characteristic equation is λ − S1 λ + S2 λ − S3 = 0, where 1𝑥1 + (5 − 𝜆)𝑥2 + 1𝑥3 = 0} − − − (1)
3𝑥1 + 1𝑥2 + (1 − 𝜆)𝑥3 = 0
𝑆1 = sum of the principal diagonal elements
Case (i) :λ = −2
=1+5+1=7
(1) becomes, 3𝑥1 + 1𝑥2 + 3𝑥3 = 0
𝑆2 = sum of the minors of principal diagonal elements
5 1 1 3 1 1 1𝑥1 + 7𝑥2 + 1𝑥3 = 0
=| |+| |+| |
1 1 3 1 1 5 Solving, by the rule of cross-multiplication, we have
= (5 − 1) + (1 − 9) + (5 − 1) 1 3 3 1
=4−8+4=0
S3 = |A| 7 1 1 7
1 1 3 𝑥1 𝑥2 𝑥3
= =
= |1 5 1| 1 − 21 3 − 3 21 − 1
3 1 1 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
= = ⇒ = =
= 1(5 − 1) − 1(1 − 3) + 3(1 − 15) = 4 + 2 − 42 = −36 −20 0 20 −1 0 1
∴ The characteristic equation is λ3 − 7λ2 + 36 = 0 −1
𝑋1 = ( 0 )
Solving by synthetic division method, we get 1
Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 𝑥1 𝑥2 𝑥3
Case (ii) :λ = 3 = =
0 − 1 −1 − 1 −1 − 0
(1) becomes, −2𝑥1 + 1𝑥2 + 3𝑥3 = 0 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
= = ⇒ = =
1𝑥1 + 2𝑥2 + 1𝑥3 = 0 −1 −2 −1 1 2 1
Solving, by the rule of cross-multiplication, we have 1
𝑋3 = (2)
1 3 −2 1 1
Consider the matrix 𝑁 whose columns are normalized eigen vectors
2 1 1 2 1 1 1
𝑥1 𝑥2 𝑥3
= = √2 √3 √6
1 − 6 3 + 2 −4 − 1 1 2
𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3 ∴ 𝑁= 0 −
= = ⇒ = = √3 √6
−5 5 −5 −1 1 −1 1 1 1
−1 −
( √2 √3 √6)
𝑋2 = ( 1 )
−1 Now,
Case (iii) : The third eigen vector is given by pairwise orthogonal 1 1 1 1 1
0 −
√2 √2 √2 √3 √6
conditions 1 1 3
1 1 1 1 2
𝑋1𝑇 𝑋3 = 0 ; 𝑋2𝑇 𝑋3 = 0 𝑁 𝑇 𝐴𝑁 = − (1 5 1) 0 −
√3 √3 √3 3 1 1 √3 √6
𝑥1 1 2 1 1 1 1
−
Let 𝑋3 = (𝑥2 ) (√6 √6 √6 ) ( √2 √3 √6)
𝑥3
−2 0 0
𝑥1 = ( 0 3 0)
∴ 𝑋1𝑇 𝑋3 = 0 (−1 0 1) (𝑥2 ) = 0 − 𝑥1 + 0𝑥2 + 𝑥3 = 0 0 0 6
𝑥3
=𝐷
𝑥1
&𝑋2𝑇 𝑋3 = 0 (−1 1 −1) (𝑥2 ) = 0 − 𝑥1 + 𝑥2 − 𝑥3 = 0
𝑥3 Problems for Practise
Solving, by the rule of cross-multiplication, we have 2 1 −1
1. Diagonalise the matrix 𝐴 = ( 1 1 −2) by means of orthogonal
0 1 −1 0
−1 −2 1
transformation.
1 −1 −1 1

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 10 −2 −5 𝑥 2 + 2𝑦 2 − 3𝑧 2 + 4𝑥𝑦 − 6𝑦𝑧 + 8𝑧𝑥,
2. Diagonalise the matrix 𝐴 = ( −2 2 3 ) by means of orthogonal
−5 3 5 3𝑥12 + 5𝑥22 + 3𝑥32 − 2𝑥1 𝑥2 + 2𝑥1 𝑥3 − 2𝑥2 𝑥3
transformation.
General form
2 0 4
3. Diagonalise the matrix 𝐴 = (0 6 0) by means of orthogonal The general form of a quadratic form in n variables 𝑥1 , 𝑥2 , … 𝑥𝑛 is denoted
4 0 2
by 𝑄 = ∑𝑛𝑗=1 ∑𝑛𝑖=1 𝑎𝑖𝑗 . 𝑥𝑖 . 𝑥𝑗 , where 𝑎𝑖𝑗 = 𝑎𝑗𝑖 .
transformation.
The matrix of the coefficients 𝐴 = [𝑎𝑖𝑗 ] is a symmetric matrix, called the
Answers
matrix of the quadratic form. The general form 𝑄 = ∑𝑛𝑗=1 ∑𝑛𝑖=1 𝑎𝑖𝑗 . 𝑥𝑖 . 𝑥𝑗
2 1
0
√6 √3 where 𝑎𝑖𝑗 = 𝑎𝑗𝑖 can always be written as 𝑋 𝑇 𝐴𝑋, where
1 −1 1
1. 𝑁 = √2 √6 √3
, 𝐷 = (−1, 1, 4)
1 1 −1
𝑥1
𝑥2
( √2 √6 √3 ) 𝐴 = (𝑎𝑖𝑗 ), 𝑋 = ( … ) and 𝑋 𝑇 = (𝑥1 𝑥2 … . . 𝑥𝑛 )
1 1 −3 …
√42 √3 √14
𝑥𝑛
−5 1 1
2. 𝑁 = , 𝐷 = (0, 3, 14) For example,
√42 √3 √14
4 1 2
𝑄 = 𝑥12 − 4𝑥1 𝑥2 + 3𝑥22 , can be written
(√42 √3 √14)
1 −2 𝑥1
1 1
0 𝑄 = (𝑥1 𝑥2 ) ( )( )
√2 √2 −2 3 𝑥2
3. 𝑁 = ( 0 0 1) , 𝐷 = (−2, 6, 6) = 𝑋 𝑇 𝐴𝑋
−1 1
0 𝑥1 1 −2
√2 √2
Where 𝑋 = (𝑥 ) and 𝐴 = ( )
2 −2 3
Note: 1 To find the symmetric matrix A of a quadratic form,
Quadratic forms
 the coefficients 𝑎𝑖𝑗 are the coefficients of 𝑥𝑖2 in the quadratic form
Definition :A homogeneous polynomial of the second degree in any
 half of coefficient 𝑥𝑖 𝑥𝑗 is placed in each of the positions 𝑎𝑖𝑗 and 𝑎𝑗𝑖 of
number of variables is called a quadratic form.
A.
For example, Quadratic forms of two variables:
Note: 2 (short cut method)
𝑥 2 + 4𝑥𝑦 − 2𝑦 2,𝑥12 − 3𝑥1 𝑥2 + 2𝑥22
For three variables 𝑥, 𝑦, 𝑧 first write
Quadratic forms of three variables:

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 𝑥 𝑦 𝑧 In the linear transformation, 𝑋 = 𝑃𝑌, the matrix 𝑃 is chosen such that 𝐵 =
(𝑥 𝑦 𝑧) and then
𝑥 𝑦 𝑧 𝑃𝑇 𝐴𝑃 = diag(𝑑1 , 𝑑2 , … 𝑑𝑛 ), a diagonal matrix, then the quadratic form
𝑥 𝑦 𝑧 multiply first row by 𝑥 𝑑1 𝑦12 + 𝑑2 𝑦22 + ⋯ + 𝑑𝑛 𝑦𝑛2, which is called the canonical form or sum of
(𝑥 𝑦 𝑧) multiply second row by 𝑦
squares form of 𝑸.
𝑥 𝑦 𝑧 multiply third row by 𝑧
𝑥 2 𝑥𝑦 𝑥𝑧 Nature of the quadratic form
That is, (𝑥𝑦 𝑦 2 𝑦𝑧)
𝑥𝑧 𝑦𝑧 𝑧2 Let 𝑄 = 𝑋 𝑇 𝐴𝑋 be a quadratic form in 𝑛variables 𝑥1 , 𝑥2 , … 𝑥𝑛 .
Similarly we can write for 𝑥1 , 𝑥2 , 𝑥3
Rank of the quadratic form (𝒓)

Examples If the canonical form of 𝑄 consists only ‘𝑟’ square terms. Then, the rank of
(i) The matrix of the quadratic form 𝑥 2 − 𝑦 2 − 𝑧 2 + 4𝑥𝑦 + 6𝑥𝑧 − 2𝑦𝑧is 𝐴is 𝑟.
1 2 3
(2 −1 −1) Index of the quadratic form (𝒔)
3 −1 −1
The number of positive square terms in the canonical form is called the
(ii) The matrix of the quadratic form 𝑥12 + 𝑥22 + 3𝑥32 + 3𝑥1 𝑥2 + 3𝑥1 𝑥3 +
index of the quadratic form. Therefore, the number of positive square
1 3⁄ 3⁄
2 2 terms is index (𝑠).
3𝑥2 𝑥3is 3⁄2 1 3⁄2
3 3⁄ Signature of the quadratic form
( ⁄2 2 3 )
(iii) Similarly, we reverse the process that is the quadratic form whose The difference of number of positive and negative square terms is called
2 −1 1 the signature of the quadratic form.
matrix 𝐴 = (−1 2 −1) is 2𝑥 2 + 2𝑦 2 + 2𝑧 2 − 2𝑥𝑦 − 2𝑦𝑧 + 2𝑥𝑧
1 −1 2
Nature of the quadratic form
Canonical form of a quadratic form The quadratic form 𝑄 = 𝑋 𝑇 𝐴𝑋 in 𝑛 variables is said to be
𝑇
A real quadratic form 𝑄 = 𝑋 𝐴𝑋 be reduced as sum or difference of the (i) Positive definite if 𝑟 = 𝑛 and 𝑠 = 𝑛
squares of the new variables, then the resulting quadratic expression is
(or) all the eigen values of A are positive numbers.
called the Canonical form of the given quadratic form.

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
(ii) Negative definite if 𝑟 = 𝑛 and 𝑠 = 0  The quadratic form is positive semi definite if 𝐷𝑛 ≥ 0 and atleast one
(or) all the eigen values of A are negative numbers. 𝐷𝑖 = 0.
(iii) Positive semi definite if 𝑟 < 𝑛 and 𝑠 = 𝑟  The quadratic form is negative semi definite if (−1)𝑛 𝐷𝑛 ≥ 0 and
(or) all the eigen values of A are positive numbers and atleast one atleast one 𝐷𝑖 = 0.
eigen value is zero.  The quadratic form is indefinite in all other cases.
(iv) Negative semi definite if 𝑟 < 𝑛 and 𝑠 = 0
(or) all the eigen values of A are negative numbers and at least one Reduction of a quadratic form by means of orthogonal reduction
eigen value is zero. Consider a quadratic form 𝑄 = 𝑋 𝑇 𝐴𝑋 in 𝑛variables 𝑥1 , 𝑥2 , … 𝑥𝑛 .
(v) Indefinite in all other cases  Let 1 , 2 , … 𝑛 be the eigen values of A and 𝐸1 , 𝐸2 , … 𝐸𝑛 be the
(or) if 𝐴 has both positive and negative eigen values. normalized eigen vectors of 𝐴.
 Consider the matrix 𝑁 whose columns are normalized eigen vectors of
Test for nature of a quadratic form through principal minors
𝐴, called normalized modal matrix.
Let 𝐴 = [𝑎𝑖𝑗 ] be the matrix of the quadratic form in 𝑛 variables. Then 𝐴 is a
 Clearly N is an orthogonal matrix.
square symmetric matrix of order 𝑛.  Then the transformation 𝑋 = 𝑁𝑌 where 𝑌 is a column vector of new
Let 𝐷1 = |𝑎11 | = 𝑎11 variables 𝑌1 , 𝑌2 , … 𝑌𝑛 reduces the given quadratic form 𝑄 into 𝑌 𝑇 𝐷𝑌
𝑎11 𝑎12
𝐷2 = |𝑎 | where D is a diagonal matrix with the eigen values of 𝐴 as diagonal
21 𝑎22
𝑎11 𝑎12 𝑎13 elements.
𝐷3 = |𝑎21 𝑎22 𝑎23 |  The resulting Q.F 𝑌 𝑇 𝐷𝑌 takes the form 1 𝑦12 + 2 𝑦22 + ⋯ +𝑛 𝑦𝑛2,
𝑎31 𝑎32 𝑎33
which is the required canonical form.
… . . , 𝐷𝑛 = |𝐴|
 This is applied only when the eigen vectors of 𝐴 are pairwise
Where 𝐷1 , 𝐷2 , 𝐷3 , … 𝐷𝑛 are the principal minors of 𝐴.
orthogonal.
 The quadratic form is positive definite if 𝐷1 , 𝐷2 , 𝐷3 , … 𝐷𝑛 are all positive.
i,e. 𝐷𝑛 > 0, for all 𝑛.
ILLUSTRATIVE EXAMPLES
 The quadratic form is negative definite if 𝐷1 , 𝐷3 , 𝐷5 , … are all negative
Example : 1Write the matrix of the quadratic form 2𝑥12 − 2𝑥22 + 4𝑥32 +
and 𝐷2 , 𝐷4 , 𝐷6 , … are all positive. i,e. (−1)𝑛 𝐷𝑛 > 0, for all 𝑛.
2𝑥1 𝑥2 − 6𝑥1 𝑥3 + 6𝑥2 𝑥3
Solution :Matrix of the quadratic form
Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 1 1 Problems for Practise
coeff. of 𝑥12 coeff. of 𝑥1 𝑥2 coeff. of 𝑥1 𝑥3
2 2
1 1 1. Find the matrix of the quadratic form 2𝑥12 − 3𝑥1 𝑥2 + 4𝑥22
𝐴= coeff. of𝑥2 𝑥1 coeff. of 𝑥22 coeff. of𝑥2 𝑥3
2 2 1 2 5
1 1 2. Write the quadratic form corresponding to the matrix (2 4 6)
(2 coeff. of 𝑥3 𝑥1 coeff. of𝑥3 𝑥2 coeff. of𝑥32 )
2 5 6 3
2 1 −3 1
3 2 0
=( 1 −2 3)
−3 3 4 3. Write the quadratic form corresponding to the matrix ( 1
0 6)
2
0 6 −7
Example : 2Write the matrix of the quadratic form 10𝑥12 + 2𝑥22 + 5𝑥32 + Answers
6𝑥2 𝑥3 − 10𝑥3 𝑥1 − 4𝑥1 𝑥2 −3
2
2
Solution :Matrix of the quadratic form 𝐴 = 1. (−3 )
2
4
1 1
coeff. of 𝑥12 coeff. of 𝑥1 𝑥2 coeff. of 𝑥1 𝑥3
2 2
1 1
10 −2 −5 2. 𝑥1 + 4𝑥2 2 + 3𝑥3 2 + 4𝑥1 𝑥2 + 12𝑥2 𝑥3 + 10𝑥1 𝑥3
2
coeff. of𝑥2 𝑥1 coeff. of 𝑥22 coeff. of𝑥2 𝑥3 = ( −2 2 3)
2 2
1 1 −5 3 5 3. 3𝑥1 2 − 7𝑥3 2 + 𝑥1 𝑥2 + 12𝑥2 𝑥3
coeff. of 𝑥3 𝑥1 coeff. of𝑥3 𝑥2 coeff. of𝑥32
(2 2 )
ILLUSTRATIVE EXAMPLES
Example : 3Write the quadratic form corresponding to the following
Example : 1Determine the nature of the Q.F 𝑥 2 + 2𝑦 2 + 3𝑧 2 + 2𝑥𝑦 +
0 −1 2
symmetric matrix (−1 1 4) 2𝑦𝑧 − 2𝑥𝑧without reducing it to canonical form
2 4 3 1 1 −1
Solution :Quadratic form corresponding to the symmetric matrix 𝐴 is Solution :The matrix A of the Q.F is𝐴 = ( 1 2 1)
−1 1 3
0 −1 2 𝑥1
1 1 −1
𝑋 ′ 𝐴𝑋 = (𝑥1 𝑥2 𝑥3 ) (−1 1 4) (𝑥2 ) 1 1
𝐷1 = |1| = 1; 𝐷2 = | | = 1; 𝐷3 = | 1 2 1 | = −2
2 4 3 𝑥3 1 2
−1 1 3
𝑥1
= (0𝑥1 − 𝑥2 + 2𝑥3 −𝑥1 + 𝑥2 + 4𝑥3 2𝑥1 + 4𝑥2 + 3𝑥3 ) (𝑥2 ) ∴ 𝐷1 = +𝑣𝑒; 𝐷2 = +𝑣𝑒; 𝐷3 = −𝑣𝑒
𝑥3 ∴ The given Q.F is indefinite.
= (0𝑥1 − 𝑥2 + 2𝑥3 )𝑥1 + (−𝑥1 + 𝑥2 + 4𝑥3 )𝑥2 + (2𝑥1 + 4𝑥2 + 3𝑥3 )𝑥3
= 0𝑥12 − 𝑥2 𝑥1 + 2𝑥3 𝑥1 −𝑥1 𝑥2 + 𝑥2 2 + 4𝑥3𝑥2 + 2𝑥1 𝑥3 + 4𝑥2 𝑥3 + 3𝑥3 2 Example : 2Determine the nature of the Q.F 5𝑥 2 + 26𝑦 2 + 10𝑧 2 + 4𝑦𝑧 +

= 0𝑥12 + 𝑥2 2 + 3𝑥3 2 − 𝑥2 𝑥1 −2𝑥1 𝑥2 + 4𝑥1 𝑥3 + 8𝑥2 𝑥3 6𝑥𝑦 + 14𝑧𝑥 using principal minors

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 5 3 7 ILLUSTRATIVE EXAMPLES
Solution :The matrix A of the Q.F is 𝐴 = (3 26 2 )
7 2 10 Example :1By an orthogonal transformation, reduce the Q.F 3𝑥 2 − 2𝑦 2 −
5 3 7 𝑧 2 − 4𝑥𝑦 + 12𝑦𝑧 + 8𝑥𝑧 to a canonical form. Find also its nature.
5 3
𝐷1 = |5| = 5; 𝐷2 = | | = 121; 𝐷3 = |3 26 2 | = 0
3 26 Solution :The given Q.F can be rewritten in matrix form as 𝐴=
7 2 10
∴ 𝐷1 = +𝑣𝑒; 𝐷2 = +𝑣𝑒; 𝐷3 = 0 3 −2 4
(−2 −2 6)
∴ The given Q.F is positive semi-definite. 4 6 −1
The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where
2
Example : 3Determine the nature of the Q.F 𝑥1 + 3𝑥2 + 6𝑥3 + 2 2
𝑆1 = sum of the principal diagonal elements
2𝑥2 𝑥3+4𝑥3 𝑥1 + 2𝑥1 𝑥2without reducing it to canonical form =3−2−1=0
1 1 2 𝑆2 = sum of the minors of principal diagonal elements
Solution :The matrix A of the Q.F is 𝐴 = (1 3 1)
2 1 6 −2 6 3 4 3 −2
=| |+| |+| |
1 1 2 6 −1 4 −1 −2 −2
1 1 = (2 − 36) + (−3 − 16) + (−6 − 4) = −34 − 19 − 10 = −63
𝐷1 = |1| = 1; 𝐷2 = | | = 2; 𝐷3 = |1 3 1| = 3
1 3
2 1 6 S3 = |A|
∴ 𝐷1 = +𝑣𝑒; 𝐷2 = +𝑣𝑒; 𝐷3 = +𝑣𝑒
3 −2 4
∴ The given Q.F is positive definite. = |−2 −2 6|
4 6 −1
Problems for Practise = 3(2 − 36) + 2(2 − 24) + 4(−12 + 8)

Determine the nature of the following Quadratic forms: = −102 − 44 − 16 = −162

1. 2𝑥1 2 + 𝑥2 2 − 3𝑥3 2 + 12𝑥1 𝑥2 −8𝑥2 𝑥3 − 4𝑥3 𝑥1 ∴ The characteristic equation is λ3 − 63λ + 162 = 0
Solving by synthetic division method, we get
2. 6𝑥1 2 + 3𝑥2 2 + 14𝑥32 +4𝑥2 𝑥3 + 18𝑥1 𝑥3 + 4𝑥1 𝑥2
3 1 0 −63 162
3. 𝑥1 2 − 2𝑥1 𝑥2 + 𝑥2 2 + 𝑥3 2
0 3 9 −162
Answers
1. Indefinite 1 3 −54 0

2. +𝑣𝑒 definite i. e. , λ = 3, λ2 + 3λ − 54 = 0

3. +𝑣𝑒 semi-definite i. e. , λ = 3, (λ − 6)(λ + 9) = 0

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
∴ The eigenvalues are λ = 3, 6, −9 −2 4 −3 −2
To find the eigenvector
−8 6 −2 −8
𝑥1
Let (𝐴 − 𝜆𝐼)𝑋 = 0, where 𝑋 = (𝑥2 ) 𝑥1 𝑥2 𝑥3
𝑥3 = =
−12 + 32 −8 + 18 24 − 4
3− −2 4 𝑥1 0 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
= = ⇒ = =
( −2 −2 −  𝑥
6 ) ( 2 ) = (0) 20 10 20 2 1 2
4 6 −1 −  𝑥3 0 2
(3 − )𝑥1 − 2𝑥2 + 4𝑥3 = 0 𝑋2 = (1)
−2𝑥1 + (−2 − )𝑥2 + 6𝑥3 = 0} − − − (1) 2
4𝑥1 + 6𝑥2 + (−1 − )𝑥3 = 0 Case (iii) :λ = −9
Case (i) :λ = 3 (1) becomes, 12𝑥1 − 2𝑥2 + 4𝑥3 = 0
(1) becomes, 0𝑥1 − 2𝑥2 + 4𝑥3 = 0 −2𝑥1 + 7𝑥2 + 6𝑥3 = 0
−2𝑥1 − 5𝑥2 + 6𝑥3 = 0 Solving, by the rule of cross-multiplication, we have
Solving, by the rule of cross-multiplication, we have −2 4 12 −2
−2 4 0 −2
7 6 −2 7
−5 6 −2 −5 𝑥1 𝑥2 𝑥3
= =
𝑥1 𝑥2 𝑥3 −12 − 28 −8 − 72 84 − 4
= = 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
−12 + 20 −8 − 0 0 − 4 = = ⇒ = =
𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3 −40 −80 80 −1 −2 2
= = ⇒ = = −1
8 −8 −4 2 −2 −1
𝑋3 = (−2)
2
2
𝑋1 = (−2)
−1 Clearly, the eigen vectors 𝑋1 , 𝑋2 , 𝑋3 are pairwise orthogonal, Since𝑋1𝑇 𝑋2 =
Case (ii) :λ = 6 0,𝑋2𝑇 𝑋3 = 0 and 𝑋1𝑇 𝑋3 = 0.
(1) becomes, −3𝑥1 − 2𝑥2 + 4𝑥3 = 0 Let 𝑁 be the matrix whose columns are normalized eigen vectors.
−2𝑥1 − 8𝑥2 + 6𝑥3 = 0 The normalized modal matrix is
Solving, by the rule of cross-multiplication, we have

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 2⁄ 2⁄ −1⁄
3 3 3
𝑁= −2⁄3 1⁄3 −2⁄ = 6 + 3 + 3 = 12
3
−1 2⁄ 2⁄ 𝑆2 = sum of the minors of principal diagonal elements
( ⁄3 3 3)
3 −1 6 2 6 −2
The transformation 𝑋 = 𝑁𝑌 reduces the given Q.F 𝑄 = 𝑋 𝑇 𝐴𝑋 into =| |+| |+| |
−1 3 2 3 −2 3
𝑄 = (𝑁𝑌)𝑇 𝐴(𝑁𝑌) = (9 − 1) + (18 − 4) + (18 − 4) = 8 + 14 + 14 = 36
𝑇 𝑇
= 𝑌 𝑁 𝐴𝑁𝑌 S3 = |A|
𝑇 𝑇
= 𝑌 𝑁 𝐴𝑁𝑌 6 −2 2
=𝑌 𝑇 (𝑁 𝑇
𝐴𝑁)𝑌 = |−2 3 −1|
2 −1 3
3 0 0
= 𝑌 𝑇 (𝐷)𝑌 where 𝐷 = (0 6 0) = 6(9 − 1) − (−2)(−6 + 2) + 2(2 − 6)
0 0 −9 = 48 − 8 − 8 = 32
3 0 0 𝑦1
∴ The characteristic equation is 3 − 122 + 36 − 32 = 0
i. e. 𝑄 = (𝑦1 𝑦2 𝑦3 ) (0 6 0 ) (𝑦2 )
0 0 −9 𝑦3 Solving by synthetic division method, we get
𝑄 = 3𝑦12 + 6𝑦22 − 9𝑦32, which is the required canonical form. 2 1 −12 36 −32
Hence the Q.F is indefinite in nature, as the canonical form contains both 0 2 −20 32
positive and negative terms. 1 −10 16 0
i. e. , λ = 2, λ2 − 10λ + 16 = 0
Example : 2Reduce the Q.F 6𝑥 2 + 3𝑦 2 + 3𝑧 2 − 4𝑥𝑦 − 2𝑦𝑧 + 4𝑥𝑧 into
i. e. , λ = 2, (λ − 2)(λ − 8) = 0
canonical form by an orthogonal reduction. Find its nature, rank, index and
∴ The eigenvalues are λ = 8, 2, 2
signature.
To find the eigenvector
Solution :The given Q.F can be rewritten in matrix form as 𝐴 =
𝑥1
6 −2 2 Let (𝐴 − 𝜆𝐼)𝑋 = 0, where 𝑋 = (𝑥2 )
(−2 3 −1) 𝑥3
2 −1 3
6 −  −2 2 𝑥1 0
To find the eigen values ( −2 3 −  −1 ) (𝑥2 ) = (0)
The characteristic equation is λ3 − S1 λ2 + S2 λ − S3 = 0, where 2 −1 3 −  𝑥3 0
(6 − )𝑥1 − 2𝑥2 + 2𝑥3 = 0
𝑆1 = sum of the principal diagonal elements
−2𝑥1 + (3 − )𝑥2 − 𝑥3 = 0} − − − (1)
2𝑥1 − 𝑥2 + (3 − )𝑥3 = 0

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
Case (i) :λ = 8 𝑥1
Let 𝑋3 = (𝑥2 )
(1) becomes, −2𝑥1 − 2𝑥2 + 2𝑥3 = 0 𝑥3
−2𝑥1 − 5𝑥2 − 𝑥3 = 0 𝑥1
∴ 𝑋1𝑇 𝑋3 = 0 (2 −1 1) (𝑥2 ) = 0 2𝑥1 − 1𝑥2 + 1𝑥3 = 0
Solving, by the rule of cross-multiplication, we have 𝑥3
−2 2 −2 −2 𝑥1
&𝑋2𝑇 𝑋3 = 0 (0 1 1) (𝑥2 ) = 0 0𝑥1 + 1𝑥2 + 1𝑥3 = 0
𝑥3
−5 −1 −2 −5
𝑥1 𝑥2 𝑥3 Solving, by the rule of cross-multiplication, we have
= =
2 − (−10) −4 − 2 10 − 4 −1 1 2 −1
𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
= = ⇒ = =
12 −6 6 2 −1 1 1 1 0 1
2 𝑥1 𝑥2 𝑥3
𝑋1 = (−1) = =
−1 − 1 0 − 2 2 − 0
1 𝑥1 𝑥2 𝑥3 𝑥1 𝑥2 𝑥3
Case (ii):λ = 2 = = ⇒ = =
−2 −2 2 1 1 −1
(1) becomes, 4𝑥1 − 2𝑥2 + 2𝑥3 = 0 1
𝑋3 = ( 1 )
−2𝑥1 + 𝑥2 − 𝑥3 = 0 −1
2𝑥1 − 𝑥2 + 𝑥3 = 0 The normalized modal matrix 𝑁 is
All three equations are identical, we choose one equation. i.e.2𝑥1 − 𝑥2 + 2⁄ 0 1⁄
√6 √3
𝑥3 = 0 −1 1 1
𝑁= ⁄ ⁄ ⁄
Let 𝑥1 = 0 √6 √2 √3
1⁄ 1⁄ −1⁄
𝑥2 𝑥3 ( √6 √2 √3)
 − 𝑥2 + 𝑥3 = 0  𝑥2 = 𝑥3  =
1 1
The transformation 𝑋 = 𝑁𝑌 reduces the given Q.F 𝑄 = 𝑋 𝑇 𝐴𝑋 into
0
𝑋2 = (1) 𝑄 = (𝑁𝑌)𝑇 𝐴(𝑁𝑌)
1 = 𝑌 𝑇 𝑁 𝑇 𝐴𝑁𝑌
Case (iii) : The third eigen vector is given by pairwise orthogonal
= 𝑌 𝑇 𝑁 𝑇 𝐴𝑁𝑌
conditions
= 𝑌 𝑇 (𝑁 𝑇 𝐴𝑁)𝑌
𝑋1𝑇 𝑋3 = 0 ; 𝑋2𝑇 𝑋3 = 0

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India
 8 0 0 Answers
= 𝑌 𝑇 (𝐷)𝑌 where 𝐷 = (0 2 0)
2 1
0 0 2 0
√6 √3
8 0 0 𝑦1 1 −1 1
1. 𝑁 = ; 𝑄 = −𝑦12 + 𝑦22 + 4𝑦32 ;
i. e. 𝑄 = (𝑦1 𝑦2 𝑦3 ) (0 2 0) (𝑦2 ) √2 √6 √3
1 1 −1
0 0 2 𝑦3
(√2 √6 √3)
 𝑄 = 8𝑦12 + 2𝑦22 + 2𝑦32, which is the required canonical form.
Rank=3; Index=2; Signature=1; Nature-indefinite
Hence the Q.F is positive definite in nature, as the canonical form contains 1 −2
0
√5 √5
only positive terms.
2. 𝑁 = (2 0
1 );𝑄 = 𝑦12 + 3𝑦22 + 6𝑦32 ;
Rank of the Q.F is 3 (Since, the canonical form contains 3 square terms) √5 √5
0 1 0
Index of the Q.F is 3 (Since, the number of positive square terms is 3)
Rank=3; Index=3; Signature=3
Signature of the Q.F is 3 (Since, difference between positive and negative −3 1 1

square terms is 3. i.e. 3 – 0 = 3) √10 √35 √14

−5 2
3. 𝑁 = 0 ; 𝑄 = 4𝑦12 − 𝑦22 − 8𝑦32 ;
√35 √14
1 3 3
Problems for Practise
(√10 √35 √14)
2 2 2
1. Reduce the quadratic form 2𝑥1 + 𝑥2 + 𝑥3 + 2𝑥1 𝑥2 −2𝑥1 𝑥3 − 4𝑥2 𝑥3 to Rank=3; Index=1; Signature =-1
canonical form by orthogonal transformation. Also find the rank, index, 1 1 −3
√42 √3 √14
signature and nature of the Q.F. −5 1 1
4. 𝑁 = ; 𝑄 = 3𝑦22 + 14𝑦32 ;
2 2 2 √42 √3 √14
2. Reduce the quadratic form 2𝑥1 + 5𝑥2 + 3𝑥3 + 4𝑥1 𝑥2 to canonical 4 1 2

form by orthogonal transformation. Also find the rank, index and (√42 √3 √14)

Positive semi-definite.𝑥1 = 1, 𝑥2 = −5, 𝑥3 = 4

signature the Q.F.
3. Reduce the quadratic form 3𝑥1 2 − 3𝑥2 2 − 5𝑥3 2 − 2𝑥1 𝑥2 − 6𝑥2 𝑥3 −
6𝑥3𝑥1 to canonical form by orthogonal transformation. Also find the
rank, index and signature the Q.F.
4. Reduce the quadratic form 10𝑥1 2 + 2𝑥22 + 5𝑥3 2 − 4𝑥1 𝑥2 + 6𝑥2 𝑥3 −
10𝑥3 𝑥1 to a canonical form by orthogonal reduction. Find a set of
values of 𝑥1 , 𝑥2 , 𝑥3 which will make the form vanish. Examine its nature.

Department of Mathematics, (School of Advanced Sciences & Languages) Prepared by Dr.A. Manickam, VIT BHOPAL UNIVERSITY, Kottrikalan-466114, Sehore (Dt)Madhya Pradesh, India