MODULE 1

LINEAR ALGEBRA
Linear Algebra: Illustration of a few Engineering Examples with Application
of Linear Algebra (Lecture by the Respective Domain Faculty members) -
Eigenvalues and Eigenvectors of a real matrix – Characteristic equation –
Properties of Eigenvalues and Eigenvectors – Cayley-Hamilton theorem –
Diagonalization of matrices – Reduction of a quadratic form to canonical
form by orthogonal transformation – Nature of quadratic forms.

Linear Algebra, mathematical discipline that deals with vectors and matrices and, more
generally, with vector spaces and linear transformations. Unlike other parts of mathematics
that are frequently invigorated by new ideas and unsolved problems, linear algebra is very
well understood. Its value lies in its many applications, from mathematical physics to
modern algebra and coding theory.
Recapitulations:
Matrix: Matrix is an array of numbers in row and columns. A matrix with ‘𝑚’ rows and ‘𝑛’
columns is called as 𝑚 × 𝑛 matrix.
Square Matrix: If number of rows and columns are equal, then the matrix is called as square
matrix.
1 2 3
Ex. 𝐴 = [4 5 6]
7 8 9
The elements 1, 5, 9 are called as leading diagonal elements.
Unit Matrix: A matrix which has unity for all its diagonal elements is called as unit matrix or
identity matrix of order ‘𝑛’ and is denoted by 𝐼.
1 0 0
Ex. 𝐼 = [0 1 0]
0 0 1
Eigenvalues and Eigenvectors: Let A be a square matrix, if there exist a scalar 𝜆 and a non-
zero column matrix 𝑋 such that (𝐴 − 𝜆𝐼)𝑋 = 0, then 𝜆 is called an Eigenvalue of 𝐴 and 𝑋 is
called Eigenvector corresponding to Eigenvalue of 𝐴.
Characteristic Equations:
• The characteristic equation of 3 × 3 matrix A is 𝜆3 − 𝑠1 𝜆2 + 𝑠2 𝜆 − 𝑠3 = 0,
where 𝑠1 = sum of leading diagonal elements,
𝑠2 = sum of minor of leading diagonal elements,
1
 𝑠3 = 𝑑𝑒𝑡(𝐴) = |𝐴|.
• The characteristic equation of 2 × 2 matrix 𝐴 is 𝜆2 − 𝑠1 𝜆 + 𝑠2 = 0,
where 𝑠1 = sum of leading diagonal elements,
𝑠2 = 𝑑𝑒𝑡(𝐴) = |𝐴|.
1 1 3
1. Find the Eigenvalues and Eigenvectors of the matrix [1 5 1].
3 1 1
1 1 3
Solution. Let 𝐴 = [1 5 1].
3 1 1
The Characteristic equation of 𝐴 is 𝜆3 − 𝑠1 𝜆2 + 𝑠2 𝜆 − 𝑠3 = 0 --------------- (B)
Now, 𝑠1 = Sum of leading diagonal elements = 1 + 5 + 1 = 7
𝑠2 = Sum of minor of leading diagonal elements
5 1 1 3 1 1
=| |+| |+| | = (5 − 1) + (1 − 9) + (5 − 1) = 0
1 1 3 1 1 5
𝑠3 = 𝑑𝑒𝑡 (𝐴) = |𝐴|
1 1 3
= |1 5 1| = 1(5 − 1) − 1(1 − 3) + 3(1 − 15) = −36.
3 1 1
(B) gives
𝜆3 − 7𝜆2 + 0𝜆 − (−36) = 0
𝜆3 − 7𝜆2 + 36 = 0
After simplification, we get 𝜆 = 3, 6, −2.
Therefore, Eigenvalues are 3, 6, −2.
To find Eigenvector:
(𝐴 − 𝜆𝐼)𝑋 = 0

1 1 3 1 0 0 𝑥1 0
𝑥
[[1 5 1] − 𝜆 [0 1 0]] [ 2 ] = [0]
3 1 1 0 0 1 𝑥3 0
1−𝜆 1 3 𝑥1 0
[ 1 5−𝜆 𝑥
1 ] [ 2 ] = [0]
3 1 1 − 𝜆 𝑥3 0
i.e.,

2
 (1 − 𝜆)𝑥1 + 𝑥2 + 3𝑥3 = 0
𝑥1 + (5 − 𝜆)𝑥2 + 𝑥3 = 0 } ---------- (A)
3𝑥1 + 𝑥2 + (1 − 𝜆)𝑥3 = 0
Case i). If 𝜆 = 3,
(A) becomes
(−2)𝑥1 + 𝑥2 + 3𝑥3 = 0 -------------(1)
𝑥1 + (2)𝑥2 + 𝑥3 = 0 ------------- (2)
3𝑥1 + 𝑥2 + (−2)𝑥3 = 0 ------------- (3)
Solving these equations
𝑥1 𝑥2 𝑥3
= =
1 − 6 3 − (−2) −4 − 1

𝑥1 𝑥2 𝑥3
= =
−5 5 −5
−5 −1
Eigenvector for 𝜆 = 3 is 𝑋1 = [ 5 ] or [ 1 ].
−5 −1
Case ii). If 𝜆 = 6,
(A) becomes
(−5)𝑥1 + 𝑥2 + 3𝑥3 = 0 ------------- (4)
𝑥1 + (−1)𝑥2 + 𝑥3 = 0 ------------- (5)
3𝑥1 + 𝑥2 + (−5)𝑥3 = 0 ------------- (6)
Solving these equations
𝑥1 𝑥2 𝑥3
= =
1 − (−3) 3 − (−5) 5 − 1

𝑥1 𝑥2 𝑥3
= =
4 8 4
4 1
Eigenvector for 𝜆 = 6 is 𝑋2 = [ 8] or [ 2]
4 1
Case iii). If 𝜆 = −2,
(A) becomes

3
 3𝑥1 + 𝑥2 + 3𝑥3 = 0 ------------- (7)
𝑥1 + (7)𝑥2 + 𝑥3 = 0 ------------- (8)
3𝑥1 + 𝑥2 + (3)𝑥3 = 0 ------------- (9)
Solving these equations
𝑥1 𝑥2 𝑥3
= =
1 − 21 3 − (3) 21 − 1

𝑥1 𝑥2 𝑥3
= =
−20 0 20
−20 −1
Eigenvector for 𝜆 = −2 is 𝑋3 = [ 0 ] or [ 0 ].
20 1
8 −6 2
2. Find the Eigenvalues and Eigenvectors of the matrix [−6 7 −4].
2 −4 3
8 −6 2
Solution. Let 𝐴 = [−6 7 −4].
2 −4 3
The Characteristic equation of 𝐴 is 𝜆3 − 𝑠1 𝜆2 + 𝑠2 𝜆 − 𝑠3 = 0 --------------- (B)
Now, 𝑠1 = Sum of leading diagonal elements = 8 + 7 + 3 = 18.
𝑠2 = Sum of minor of leading diagonal elements
7 −4 8 2 8 −6
=| |+| |+| |
−4 3 2 3 −6 7
= (21 − 16) + (24 − 4) + (56 − 36) = 45
𝑠3 = 𝑑𝑒𝑡 (𝐴) = |𝐴|
8 −6 2
= |−6 7 −4| = 8(21 − 16) − (−6)(−18 − (−8)) + 2(24 − 14) = 0.
2 −4 3
(B) gives
𝜆3 − 18𝜆2 + 45𝜆 − (0) = 0
𝜆3 − 18𝜆2 + 45𝜆 = 0
After simplification, we get 𝜆 = 3, 0, 15.
Therefore, Eigenvalues are 3, 0, 15.
To find Eigenvector:
4
 (𝐴 − 𝜆𝐼)𝑋 = 0

8 −6 2 1 0 0 𝑥1 0
𝑥
[[−6 7 −4] − 𝜆 [0 1 0]] [ 2 ] = [0]
2 −4 3 0 0 1 𝑥3 0
8−𝜆 −6 2 𝑥1 0
[ −6 7−𝜆 𝑥 =
−4 ] [ 2 ] [0]
2 −4 3 − 𝜆 𝑥3 0
i.e.,
(8 − 𝜆)𝑥1 − 6𝑥2 + 2𝑥3 = 0
−6𝑥1 + (7 − 𝜆)𝑥2 − 4𝑥3 = 0}------------- (A)
2𝑥1 − 4𝑥2 + (3 − 𝜆)𝑥3 = 0
Case i). If 𝜆 = 0,
(A) becomes
(8)𝑥1 − 6𝑥2 + 2𝑥3 = 0 ------------- (1)
−6𝑥1 + (7)𝑥2 − 4𝑥3 = 0 ------------- (2)
2𝑥1 − 4𝑥2 + (3)𝑥3 = 0 ------------- (3)
Solving these equations
𝑥1 𝑥2 𝑥3
= =
24 − 14 −12 − (−32) 56 − 36
𝑥1 𝑥2 𝑥3
= =
10 20 20
10 1
Eigenvector for 𝜆 = 0 is 𝑋1 = [ 20] or [2].
20 2
Case ii). If 𝜆 = 3,
(A) becomes
5𝑥1 − 6𝑥2 + 2𝑥3 = 0 ------------- (4)
−6𝑥1 + (4)𝑥2 − 4𝑥3 = 0 ------------- (5)
2𝑥1 − 4𝑥2 + (0)𝑥3 = 0 ------------- (6)
Solving these equations
𝑥1 𝑥2 𝑥3
= =
24 − 8 −12 − (−20) 20 − 36

5
 𝑥1 𝑥2 𝑥3
= =
16 8 −16
16 2
Eigenvector for 𝜆 = 3 is 𝑋1 = [ 8 ] or [ 1 ].
−16 −2
Case iii). If 𝜆 = 15,
(A) becomes
−7𝑥1 − 6𝑥2 + 2𝑥3 = 0 ------------- (7)
−6𝑥1 − 8𝑥2 − 4𝑥3 = 0 ------------- (8)
2𝑥1 − 4𝑥2 − 12𝑥3 = 0 ------------- (9)
Solving these equations
𝑥1 𝑥2 𝑥3
= =
24 − (−16) −12 − (28) 56 − 36
𝑥1 𝑥2 𝑥3
= =
40 −40 20
40 2
Eigenvector for 𝜆 = 15 is 𝑋1 = [ −40] or [−2].
20 1

Practice Problems
1. Find the Eigenvalues and Eigenvectors of the matrices:
4 3 1 4
(a) [ ] (b) [ ]
2 9 3 2

7 −2 0
2. Find the Eigenvalues and Eigenvectors of the matrix[−2 6 −2].
0 −2 5

Properties of Eigenvalues
1. Any square matrix 𝐴 and its transpose 𝐴𝑇 have the same Eigenvalues.
2. The Eigenvalues of a triangular matrix are just the diagonal elements of the matrix.
3. The Eigenvalues of an idempotent matrix are either zero or unity.
4. The sum of the Eigenvalues of a matrix is its trace i.e., the sum of the elements of the
principal diagonal.
5. The product of the Eigenvalues of a matrix 𝐴 is equal to its determinant.
1
6. If 𝜆 is an eigenvalue of a matrix 𝐴, then is the Eigenvalue of 𝐴−1 .
𝜆
6
 1
7. If 𝜆 is an Eigenvalue of an orthogonal matrix, then is also its Eigenvalue.
𝜆
8. If 𝜆1 , 𝜆2 , … , 𝜆𝑛 are the eigenvalues of a matrix 𝐴, then 𝐴𝑚 has the Eigenvalues
𝜆1𝑚 , 𝜆𝑚 𝑚
2 , … , 𝜆𝑛 .
9. The Eigenvalues of a real symmetric matrix are real numbers.
10.The Eigenvectors corresponding to distinct Eigenvalues of a real symmetric matrix are
orthogonal.
11.The similar matrices have same Eigenvalues.
12.The Eigenvectors of a matrix A are not unique.

Problems
3 1 4
1. Find the sum and product of the Eigenvalues of [0 2 6] .
0 0 5

Solution. Sum of Eigenvalues = Sum of main diagonals

= 3 + 2 + 5 = 10.
Product of Eigenvalues = Determinants of the matrix
3 1 4
= |0 2 6| = 3(10 − 0) − 1(0 − 0) + 4(0 − 0) = 30
0 0 5

1 1 3
2. If −2 and 6 are the two Eigenvalues of 𝐴 = [1 5 1]. Find the third Eigenvalue of
3 1 1
𝐴.

1 1 3
Solution. Let 𝐴 = [1 5 1].
3 1 1
Let 𝜆1 , 𝜆2 , 𝜆3 are the Eigenvalues of 𝐴.
We know that, Sum of Eigenvalues = Sum of main diagonals.
𝜆1 + 𝜆2 + 𝜆3 = 𝑎11 + 𝑎22 + 𝑎33
−2 + 6 + 𝜆3 = 1 + 5 + 1
𝜆3 = 3.
Therefore, the Eigenvalues of 𝐴 are −2, 6, 3.
7
Practice Problems
2 2 1
1. Two Eigenvalues of the matrix 𝐴 = [1 3 1] are equal to 1 each. Find the third
1 2 2
Eigenvalue.
2. If 2, −1, −2 are the Eigenvalues of the matrix 𝐴, then what is the Eigenvalues of the
matrix 𝐴2 − 2𝐼.
Cayley-Hamilton Theorem
Every square matrix satisfies its own characteristic equation.
Uses. To calculate 1). the positive integral powers of 𝐴 and
2). the inverse of a non-singular square matrix 𝐴.
Problems
1 3 7
1. Verify Cayley-Hamilton theorem for the matrix [4 2 3] and also use it to find 𝐴−1 .
1 2 1
1 3 7
Solution. Given 𝐴 = [4 2 3]
1 2 1

The Characteristic equation of 𝐴 is 𝜆3 − 𝑠1 𝜆2 + 𝑠2 𝜆– 𝑠3 = 0.

Now, s1 = Sum of the main diagonal elements = 1 + 2 + 1 = 4.

s2 =Sum of the minors of the main diagonal elements.

= (2 − 6) + (1 − 7) + (2 − 12) = −4 − 6 − 10 = −20.
s3= |𝐴| = 1(2 − 6) − 3(4 − 3) + 7(8 − 2) = −4 − 3 + 42 = 35

The characteristic equation is λ3 − 4λ2 − 20𝜆 − 35 = 0.

To prove that: λ3 − 4λ2 − 20𝜆 − 35 = 0

1 3 7 1 3 7 20 23 23
2
𝐴 = [4 2 3] [4 2 3] = [15 22 37]
1 2 1 1 2 1 10 9 14

20 23 23 1 3 7 135 152 232

3 2
𝐴 = 𝐴 𝐴 = [15 22 37] [4 2 3] = [140 163 208]
10 9 14 1 2 1 60 76 111

8
 135 152 232 20 23 23
3 2
𝐴 − 4𝐴 − 20𝐴 − 35𝐼 = [140 163 208] − 4 [15 22 37]
60 76 111 10 9 14
1 3 7 35 0 0
−20 [4 2 3] − [ 0 35 0 ]
1 2 1 0 0 35
0 0 0
=[0 0 0].
0 0 0
Therefore, Cayley-Hamilton theorem is verified.

Consider A3 − 4A2 – 20A– 35I = 0 ………………………(i)

Multiplying (i) by 𝐴−1 , we get
𝐴2 − 4𝐴 − 20𝐼 − 35𝐴−1 = 0
1
𝐴−1 = [𝐴2 − 4𝐴 − 20𝐼].
35

20 23 23 1 3 7 35 0 0
−1 1
𝐴 = [[15 22 37] − 4 [4 2 3] − 20 [ 0 35 0 ]]
35
10 9 14 1 2 1 0 0 35
20 23 23 4 12 28 20 0 0
1
= [[15 22 37] − [16 8 12] − [ 0 20 0 ]]
35
10 9 14 4 8 4 0 0 20
−4 11 −5
1
= [−1 −6 25 ].
35
6 1 −10
2 1 1
2. Find the characteristic equation of the matrix, 𝐴 = [0 1 0] and hence compute
1 1 2
𝐴 . Also, find the matrix represented by 𝐴 − 5𝐴7 + 7𝐴6 − 3𝐴5 + 𝐴4 − 5𝐴3 +
−1 8

8𝐴2 − 2𝐴 + 𝐼.
2 1 1
Solution. Given 𝐴 = [0 1 0]
1 1 2
The Characteristic equation of 𝐴 is 𝜆3 − 𝑠1 𝜆2 + 𝑠2 𝜆 − 𝑠3 = 0.
𝑠1 = 2 + 1 + 2 = 5
1 0 2 1 2 1
𝑠2 = | |+| |+| | = (2 − 0) + (4 − 1) + (2 − 0) = 7
1 2 1 2 0 1
2 1 1
𝑠3 = |𝐴| = |0 1 0| = 2(2 − 0) − 1(0 − 0) + 1(0 − 1) = 3
1 1 2
Therefore, the characteristic equation is 𝜆3 − 5𝜆2 + 7𝜆 − 3 = 0.
9
 By Cayley-Hamilton theorem 𝐴3 − 5𝐴2 + 7𝐴 − 3𝐼 = 0---------------(i)
Multiplying (i) by 𝐴−1 , we get
1
𝐴2 − 5𝐴 + 7𝐼 − 3𝐴−1 = 0 or 𝐴−1 = [𝐴2 − 5𝐴 + 7𝐼]
3

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

2
Now, 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 1 0 0
𝐴2 − 5𝐴 + 7𝐼 = [0 1 0] − 5 [0 1 0 ] + 7 [0 1 0]
4 4 5 1 1 2 0 0 1
2 −1 −1
=[ 0 3 0]
−1 −1 2
2 −1 −1
−1 1
Hence from (ii), 𝐴 = [ 0 3 0 ].
3
−1 −1 2

Now 𝐴8 − 5𝐴7 + 7𝐴6 − 3𝐴5 + 𝐴4 − 5𝐴3 + 8𝐴2 − 2𝐴 + 𝐼

= 𝐴5 (𝐴3 − 5𝐴2 + 7𝐴 − 3𝐼) + 𝐴(𝐴3 − 5𝐴2 + 7𝐴 − 3𝐼) + 𝐴2 + 𝐴 + 𝐼
= 𝐴2 + 𝐴 + 𝐼
5 4 4 2 1 1 1 0 0 8 5 5
= [0 1 0] + [0 1 0] + [0 1 0] = [0 3 0] .
4 4 5 1 1 2 0 0 1 5 5 8

Practice Problem:
2 −1 2
1. Verify the matrix A= [−1 2 −1] satisfies its characteristic equation, and use it
1 −1 2
−1 4
to find 𝐴 & 𝐴 .
Reduction to Diagonal Form
If a square matrix 𝐴 of order 𝑛 has 𝑛 linearly independent Eigenvectors, then a matrix 𝑃 can
be found such that 𝑃−1 𝐴𝑃 is a diagonal matrix.
Power of a matrix: Diagonalisation of a matrix is quite useful for obtaining powers of a
matrix.
𝜆1 𝑛 0 0
i.e., 𝐴𝑛 = 𝑃𝐷𝑛 𝑃−1 where 𝐷𝑛 = [ 0 𝜆2 𝑛 0 ]
0 0 𝜆3 𝑛
10
 1 1 3
1. Reduce the matrix 𝐴 = [1 5 1] to the diagonal form.
3 1 1
Soln. The Eigenvalues of 𝐴 (found in Ex 1. Page 2) are -2, 3, 6 and the Eigenvectors are
(−1,0,1), (1, −1,1), (1,2,1).
Writing these Eigenvectors as the three columns, the required transformation matrix is
−1 1 1
𝑃 = [ 0 −1 2].
1 1 1
−1 1 1 𝑎1 𝑏1 𝑐1
−1
To find 𝑃 , |𝑃| = | 0 −1 2| = |𝑎2 𝑏2 𝑐2 | (say)
1 1 1 𝑎3 𝑏3 𝑐3
−1 2
𝐴1 = | | = −3, 𝐵1 = 2, 𝐶1 = 1, 𝐴2 = 0, 𝐵2 = −2, 𝐶2 = 2, 𝐴3 = 3, 𝐵3 = 2, 𝐶3 = 1.
1 1
Also |𝑃| = 𝑎1 𝐴1 + 𝑏1 𝐵1 + 𝑐1 𝐶1 = −1(−3) + 1(2) + 1(1) = 6
𝐴1 𝐴2 𝐴3 −3 0 3
−1 1 1
𝑃 = [𝐵1 𝐵2 𝐵3 ] = [ 2 −2 2]
|𝑃| 6
𝐶1 𝐶2 𝐶3 1 2 1
−2 0 0
−1
Thus 𝐷 = 𝑃 𝐴𝑃 = [ 0 3 0].
0 0 6

−1 2 −2 1 0 0
2. If the diagonal form of the matrix 𝐴 = [ 1 2 1 ] is [0 √5 0 ] , then find
−1 −1 0 0 0 −√5
4
𝐴 .
Solution: The characteristic equation of 𝐴 is
−1 − 𝜆 2 −2
| 1 2−𝜆 1 |=0
−1 −1 −𝜆
𝜆3 − 𝜆2 − 5𝜆 + 5 = 0.
Solving, we get 𝜆1 = 1, 𝜆2 = √5, 𝜆3 = −√5 as the Eigenvalues of 𝐴.
When 𝜆 = 1, the corresponding Eigenvector is given by
−2𝑥 + 2𝑦 − 2𝑧 = 0, 𝑥 + 𝑦 + 𝑧 = 0, −𝑥 − 𝑦 − 𝑧 = 0.
𝑥 𝑦 𝑧
Solving the first two equations, we get = = giving the Eigenvector (1, 0, −1),
4 0 −4

11
 when 𝜆 = √5, the corresponding Eigenvector is given by

(−1 − √5)𝑥 + 2𝑦 − 2𝑧 = 0, 𝑥 + (2 − √5)𝑦 + 𝑧 = 0, −𝑥 − 𝑦 − √5𝑧 = 0.

Solving 2nd and 3rd equations. We get
𝑥 𝑦 𝑧
= =
6−2√5 −1+√5 1−√5
𝑥 𝑦 𝑧
= =
√5−1 1 −1

Giving the Eigenvector (√5 − 1, 1, −1).

Similarly, the Eigenvector corresponding to 𝜆 = −√5, is (√5 + 1, −1, 1).

Writing the three Eigenvectors as the three columns, we get the transformation

1 √5 − 1 √5 + 1
matrix as 𝑃=[ 0 1 −1 ]
−1 −1 1
1 √5 − 1 √5 + 1 𝑎1 𝑏1 𝑐1
−1
To find 𝑃 , |𝑃| = [ 0 1 −1 ] = |𝑎2 𝑏2 𝑐2 | (say
−1 −1 1 𝑎3 𝑏3 𝑐3
1 −1
𝐴1 = | | = 0, 𝐵1 = 1, 𝐶1 = 1, 𝐴2 = −2√5, 𝐵2 = 2 + √5, 𝐶2 = 2 − √5, 𝐴3 =
−1 1
−2√5, 𝐵3 = 1, 𝐶3 = 1.
Also |𝑃| = 𝑎1 𝐴1 + 𝑏1 𝐵1 + 𝑐1 𝐶1 = 1(0) + (√5 − 1)(1) + (√5 + 1)(1) = 2√5
−1 −1 0
1 1 1 1
1 𝐴1 𝐴2 𝐴3 1 0 −2√5 −2√5 +
1 ] = 2√5 2 √5 2√5
−1
𝑃 = [𝐵 𝐵2 𝐵3 ] = [1 2 + √5
|𝑃| 1 2√5 1 1 1 1
𝐶1 𝐶2 𝐶3 1 2 − √5 1 − +
[2√5 2 √5 2√5]
Hence the diagonal matrix is
𝜆1 0 0 1 0 0
𝐷 = [0 𝜆2 0 ] = [0 √5 0 ].
0 0 𝜆3 0 0 −√5
𝜆1 4 0 0 14 0 0 1 0 0
Therefore, 𝐷4 = [ 0 𝜆2 4 0 ]=[0 (√5)4 0 ] = [0 25 0 ].
0 0 𝜆3 4 0 0 (−√5)4 0 0 25

12
 0 −1 −1
1 √5 − 1 √5 + 1 1 0 0 1 1
+
1 1
4 4 −1
Hence 𝐴 = 𝑃𝐷 𝑃 =[ 0 1 −1 ] [0 25 0 ] [2√5 2 √5 2√5]
−1 −1 1 0 0 25 1 − 1 + 1 1
2√5 2 √5 2√5

25 24 24
=[ 0 25 0 ].
0 −24 1

Quadric Forms
A homogeneous polynomial of the second degree in any number of variables is
called a quadratic form.
Ex. 2𝑥12 + 3𝑥22 − 𝑥32 + 4𝑥1 𝑥2 + 5𝑥1 𝑥3 − 6𝑥2 𝑥3 is a quadratic form in three variables.
𝑎11 ⋯ 𝑎1𝑛
The symmetric matrix 𝐴 = [𝑎𝑖𝑗 ] = [ ⋮ ⋱ ⋮ ] is called the matrix of the quadratic
𝑎𝑛1 ⋯ 𝑎𝑛𝑛
form 𝑄.
The matrix corresponding to the quadratic form is co-eff
1 1
𝑐𝑜 − 𝑒𝑓𝑓 𝑥12 𝑐𝑜 − 𝑒𝑓𝑓 𝑥1 𝑥2 𝑐𝑜 − 𝑒𝑓𝑓 𝑥1 𝑥3
2 2
1 1
𝑐𝑜 − 𝑒𝑓𝑓 𝑥2 𝑥1 𝑐𝑜 − 𝑒𝑓𝑓 𝑥22 𝑐𝑜 − 𝑒𝑓𝑓 𝑥2 𝑥3
2 2
1 1
[2 𝑐𝑜 − 𝑒𝑓𝑓 𝑥3 𝑥1 𝑐𝑜 − 𝑒𝑓𝑓 𝑥3 𝑥2 𝑐𝑜 − 𝑒𝑓𝑓 𝑥32 ]
2
Note: 𝑐𝑜 − 𝑒𝑓𝑓 𝑥1 𝑥2 = 𝑐𝑜 − 𝑒𝑓𝑓 𝑥2 𝑥1
𝑐𝑜 − 𝑒𝑓𝑓 𝑥2 𝑥3 = 𝑐𝑜 − 𝑒𝑓𝑓 𝑥3 𝑥2
𝑐𝑜 − 𝑒𝑓𝑓 𝑥1 𝑥3 = 𝑐𝑜 − 𝑒𝑓𝑓 𝑥3 𝑥1 .

1. Write the matrix of the quadratic form 2𝑥12 − 2𝑥22 + 4𝑥32 + 2𝑥1 𝑥2 − 6𝑥1 𝑥3 + 6𝑥2 𝑥3 .
Soln. Matrix of the quadratic form is
1 1
𝑐𝑜 − 𝑒𝑓𝑓 𝑥12 𝑐𝑜 − 𝑒𝑓𝑓 𝑥1 𝑥2 𝑐𝑜 − 𝑒𝑓𝑓 𝑥1 𝑥3
2 2
1 1
𝑄= 𝑐𝑜 − 𝑒𝑓𝑓 𝑥2 𝑥1 𝑐𝑜 − 𝑒𝑓𝑓 𝑥22 𝑐𝑜 − 𝑒𝑓𝑓 𝑥2 𝑥3
2 2
1 1
[2 𝑐𝑜 − 𝑒𝑓𝑓 𝑥3 𝑥1 𝑐𝑜 − 𝑒𝑓𝑓 𝑥3 𝑥2 𝑐𝑜 − 𝑒𝑓𝑓 𝑥32 ]
2

13
 1 1
2 (2) (−6)
2 2 2 1 −3
1 1
= (2) −2 (6) = [ 1 −2 3 ].
2 2
1 1 −3 3 4
[2 (−6) 2
(6) 4 ]

Nature of a Quadratic Form.

Let 𝑄 = 𝑋′𝐴𝑋 be a quadratic form in a variables x1 , x2 , … … … xn .
A real quadratic form 𝑋′𝐴𝑋 in a variable is said to be
(i) Positive definite if all the Eigenvalues of A > 0.
(ii) negative definite if all the Eigenvalues of A < 0.
(iii) Positive semidefinite if all the Eigenvalues of A ≥ 0 and at least one
Eigenvalue=0.
(iv) Negative semidefinite if all the Eigenvalues of A ≤ 0 and at least one
Eigenvalue=0.
(v) Indefinite if some of the Eigenvalues of A are positive and others negative.

1. Reduce the quadratic form 6𝑥 2 + 3𝑦 2 + 3𝑧 2 − 4𝑥𝑦 − 2𝑦𝑧 + 4𝑧𝑥 into canonical

form by an orthogonal transformation. Also discuss its nature.
Soln. Given 6𝑥 2 + 3𝑦 2 + 3𝑧 2 − 4𝑥𝑦 − 2𝑦𝑧 + 4𝑧𝑥
The matrix of the quadratic form is

1 1
𝑐𝑜 − 𝑒𝑓𝑓 𝑥 2 𝑐𝑜 − 𝑒𝑓𝑓 𝑥𝑦 𝑐𝑜 − 𝑒𝑓𝑓 𝑥𝑧
2 2
1 1 6 −2 2
𝐴= 𝑐𝑜 − 𝑒𝑓𝑓 𝑦𝑥 𝑐𝑜 − 𝑒𝑓𝑓 𝑦 2 𝑐𝑜 − 𝑒𝑓𝑓 𝑦𝑧 = [−2 3 −1]
2 2 2 −1 3
1 1 2
[ 2 𝑐𝑜 − 𝑒𝑓𝑓 𝑧𝑥 2
𝑐𝑜 − 𝑒𝑓𝑓 𝑧𝑦 𝑐𝑜 − 𝑒𝑓𝑓 𝑧 ]

The characteristic equation of A is 𝜆3 − 𝑠1 𝜆2 + 𝑠2 𝜆 − 𝑠3 = 0.

𝑠1 = 6 + 3 + 3 = 12
3 −1 6 2 6 −2
𝑠2 = | |+| |+| |
−1 3 2 3 −2 3
= (9 − 1) + (18 − 4) + (18 − 4) = 36
6 −2 2
𝑠3 = |𝐴| = |−2 3 −1| = 6(9 − 1) + 2(−6 + 2) + 2(2 − 6) =32.
2 −1 3
The characteristic equation is 𝜆3 − 12𝜆2 + 36𝜆 − 32 = 0.

14
 If 𝜆 = 1, 13 − 12(1)2 + 36(1) − 32 = −7 ≠ 0
𝜆 = 2, (2)3 − 12(2)2 + 36(2) − 32 = 0
The other two roots are 𝜆 = 8, 2.
Therefore, the Eigenvalues are 2, 2, 8. Here the roots are repeated.
6 −2 2
𝑇
𝐴 = [−2 3 −1] = 𝐴
2 −1 3
where A is symmetric matrix
To find model matrix 𝑃
2 1 2 1 1
0 −
√6 √3 √6 √6 √6
1 1 1 1 1
𝑃= − , 𝑃𝑇 = 0
√6 √2 √3 √2 √2
1 1 1 1 1 1
− −
[ √6 √2 √3] [√ 3 √3 √3]

2 1 16 2
0 0
√6 √3 √6 √3
6 −2 2 1 1 1 8 2 2
𝐴𝑃 = [−2 3 −1] − = −
2 −1 3 √6 √2 √3 √6 √2 √3
1 1 1 8 2 2
− −
[ √6 √2 √3] [ √6 √2 √3]

2 1 1 16 2
− 0
√6 √6 √6 √6 √3
1 1 8 2 2 8 0 0
𝐷 = 𝑃𝑇 𝐴𝑃 = 0 − = [ 0 2 0]
√2 √2 √6 √2 √3 0 0 2
1 1 1 8 2 2
− −
[√ 3 √3 √3] [ √6 √2 √3]

Canonical form is
𝑥 8 0 0 𝑥
[𝑥 𝑦 𝑧]𝐷 [𝑦] = [𝑥 𝑦 𝑧] [0 2 0] [𝑦] = 8𝑥 2 + 2𝑦 2 + 3𝑧 2
𝑧 0 0 2 𝑧
Nature of Quadratic form: Quadratic form is said to be positively definite (since
canonical form contains positive terms).
15
 2. Reduce the quadratic form 2 𝑥1 𝑥2 + 2 𝑥1 𝑥3 − 2𝑥2 𝑥3 to a canonical form by an
orthogonal reduction and discuss its nature.
Soln.
Given 2 𝑥1 𝑥2 + 2 𝑥1 𝑥3 − 2𝑥2 𝑥3
0 1 1
The matrix of the given quadratic form A = [1 0 −1]
1 −1 0
The characteristic equation is |A − λI| = 0,
−λ 1 1
i.e., [ 1 −λ −1] = 0 which gives λ2 − 3λ + 2 = 0
1 −1 −λ
Solving, we get λ = 1, 1, −2 as the eigen values.
Hence the given quadratic form reduces to the canonical form
λ1 x 2 + λ2 y 2 + λ3 z 2 = 0,
i.e., x 2 + y 2 − 2z 2 = 0
(ii) since some of the eigen values of A are positive and others are negative, the
given quadratic form is indefinite.
(iii) to find the matrix of transformation from |A − λI|X = 0, we get the equations
−λx + y + z = 0, x − λy + z = 0, x − y − λz = 0
When λ = −2, we get 2x + y + z = 0, x + 2y − z = 0, x − y + 2z = 0
Solving first and second equations, we get
x y z
= =
−1 1 1
∴ The corresponding eigen vector X1 = (−1, 1, 1) and its normalized form is
−1 1 1
, ,
√3 √3, √3

When λ = 1, we get −x + y + z = 0, x − y − z = 0, x − y − z = 0.
These equations are same. Take y = 0 so that x = z.
1 1
∴ The corresponding eigen vector X2 = (1, 0, 1) and it’s normalized form is , 0,
√2 √2

To find the eigen vector X3 = (l, m, n) (say)

Since X3 is orthogonal to X1 , ∴ −l + m + n = 0
Since X3 is orthogonal to X2 , ∴ l + n = 0
16
 l m n
These equations give , , ,
1 2 −1
1 2 −1
∴ The eigen vector X3 = (1, 2, −1) and normalized form is , , .
√6 √6, √6

Practice Problems
1. Reduce the quadratic form 𝑥 2 + 𝑦 2 + 𝑧 2 − 2𝑥𝑦 − 2𝑦𝑧 − 2𝑧𝑥 to canonical form
through an orthogonal transformation and discuss its nature.
2. Reduce the quadratic form 3𝑥 2 + 5𝑦 2 + 3𝑧 2 − 2𝑥𝑦 − 2𝑦𝑧 + 2𝑧𝑥 to the canonical
form.

17