Parul University

Faculty of Engineering & Technology

Department of Applied Sciences and Humanities
1st Year B.Tech Programme (All Branches)
Mathematics – 1 (203191102)
Unit – 3 MATRICES (Lecture Note)

MATRIX:

A Matrix is a rectangular array of numbers (or functions) enclosed in brackets. These number
or functions are called entries or elements of the matrix.

For example:

4 − 2 1  sin x cos x 
0 3 5 , − cos x sin x  are matrices.
   

Trace of a matrix:

If A is a square matrix, the trace of A, denoted by tr(A), is defined to be the sum of entries on
the main diagonal of A. The trace of A is undefined If A is not a square matrix.

For example:

 4 5
A=  , tr(A)= 4+6=10
10 6

Symmetric matrix:- For any square matrix A , if A = A then it is known as symmetric

T

matrix.

1 2 3 1 2 3
 
Example:- (1) A = 2 5 6  A =  2 5 6  Here, we can see that A = A so A is
T T
   
 3 6 5   3 6 5 
symmetric matrix.

Skew-symmetric matrix:- For any square matrix A , if A = − A then it is known as Skew

T

symmetric matrix.

0 3 5 0 −3 −5 0 3 5
 
Example:- (1) A = −3 0 −2  A = 3 0 2 = −  −3 0 −2  = − A
T  
     
 −5 2 0  5 −2 0   −5 2 0 

Here, we can see that A = − A so A is skew-symmetric matrix.

T

Singular and non-singular matrix:-For any square matrix A , if A  0, then it is known as

non-singular matrix and if A = 0, then it is known as singular matrix.

2 1
Example:- (1) If A =    A = 8 − 8 = 0  Singular matrix.
8 4
 1 2
(2) If 𝐴 = [ ] ⇒ |𝐴| = 4 − 6 = −2 ≠ 0 ⇒ non-singular matrix
3 4
Orthogonal Matrix: The matrix is said to be an orthogonal matrix if the product of a matrix
and its transpose gives an identity value.
Example: Determine if A is an orthogonal matrix.

−1 0
𝐴=[ ]
0 1

System of linear equation

Linear Equations: Any straight line in the xy-plane can be represented algebraically by
equation of the form ax+by=c, where a & b are real numbers.

A system of linear equation is a collection of one or more linear equations involving the same
variables.

A linear system of m linear equations in n variables: An arbitrary system of m linear equations

in n variables x1, x2, ….. , xn is a set of equations of the form
n

a
j =1
ij x j = bi (i = 1,2,3,..., m, j = 1,2,3,....n)

A system of linear equations has either

1. No solutions, or
2. Exactly one solutions, or
3. Infinitely many solutions
Geometrical representation:

x − 2 y = −1 .......... l x − 2 y = −1 ....... l
− x + 3y = 3 ........ m − x + 2 y = 3 ....... m
Exactly one solution No solution
x − 2 y = −1 ....... l
− x + 2 y = 1 ....... m
Infinitely many solutions

NOTE: (i) The system is said to be consistent if we get infinitely many solution or Unique
solution.
(ii) The system is said to be inconsistent if we get No solution.

• Condition of Consistency for non-homogeneous system:

(1) If there is a zero row to left of the augmentation bar but the last entry of this row is non-
zero then the system has no solution.
(2) If at least one of the columns on the left of the augmentation bar has zero element pivot
entry, then the system has infinitely many solutions.
(3) Otherwise the system has unique solution.

Augmented matrix: A system of m equations in n unknowns can be abbreviated by writing

only the rectangular array of numbers.

 a11 a12 .... a1n b1 

 
 a 21 a 22 .... a 2 n b2 
This is known as augmented matrix.
 ................. 
 
a m1 a m 2 .... a mn b1 

For example: Find the augmented matrix for each of the following system of linear equations:

2 x1 + + 2 x3 = 1 2 0 2 1

3x1 − x 2 + 4 x3 = 7 then augmented matrix is given by 3 − 1 4 7 .
 
6 x1 + x 2 − x3 = 0 6 1 − 1 0

ROW - ECHELON FORMS OF A MATRIX

(Gauss Elimination Method)

In the definitions that follow, a nonzero row or column in a matrix means a row or column that
contains at least one nonzero entry, a leading entry of a row refers to the leftmost nonzero
entry( in a nonzero row).
 Definition:
A rectangular matrix is in row-echelon form (or echelon form) if it has the following three
properties:
1. All nonzero rows are above any rows of all rows.
2. Each leading entry of a row is in a column to the right of the leading entry of the row
above it.
3. All entries in a column below a leading entry are zeros:
If a matrix in echelon form satisfies the following additional conditions, then it is in
reduced row- echelon form (or reduced echelon).
4. The leading entry in each nonzero row is 1.
5. Each column that contains a leading 1 has zeros everywhere else in that column.

OR

Properties of a Matrix in Row-Echelon Form

1) If there are any “all-0” rows, then they must be at the bottom of the matrix. Aside from these
“all-0” rows,
2) Every row must have a “1” (called a “leading 1”) as its leftmost non-0 entry.
3) The “leading 1”s must “flow down and to the right.” More precisely: The “leading 1” of a
row must be in a column to the right of the “leading 1”s of all higher rows.

REDUCED ROW-ECHELON (RRE) FORM FOR A MATRIX

(Gauss Jordan Elimination Methods)
A matrix is in reduced row echelon form (also called row canonical form) if it satisfies the
following conditions:

• It is in row echelon form.

• Every leading coefficient is 1 and is the only nonzero entry in its column.
Example 1: Example 2 :
Solve the system of equation
Solve the following system by gauss-
-2b + 3c = 1
Elimination method 3a + 6b – 3c = -2
2x + 2y + 2z = 0, 6a + 6b + 3c = 5 by Gauss elimination.
-2x + 5y +2z =1, Solution:
8x + y + 4z = -1 Interchanging first and second equation we get
3a + 6b – 3c = -2 (1)
Solution: 0a -2b + 3c = 1 (2)
 2 2 2 0 6a + 6b + 3c = 5 (3)
− 2 5 2 1 
  3 − 6 − 3 − 2
 8 1 4 − 1 0 − 2 3 1 

R2 + (1)R1 6 6 3 5 
R3 + (-4)R1
Operate R3 – 2R1
2 2 2 0
0 7  3 − 6 − 3 − 2
 4 1  0 − 2 3 1 
0 − 7 − 4 − 1 
0 − 6 3 9 
R3 + (1)R2
Operate R3/3
2 2 2 0 3 − 6 − 3 − 2
0 7 4 1  0 − 2 3
   1 
0 0 0 0 0 − 2 3 3 
Back substitution: from the second we find Operate R3 – R2
y= 1/7 – 4/7z. from this and first equation 3 − 6 − 3 − 2
we get x = -1/7 – 3/7z. since, z remain 0 − 2 3
arbitrary, we have infinitely many  1 
solutions. 0 0 0 2 
Which is an echelon form. This shows that the
Example 3 : system has no solution.
Solve the following system by gauss
elimination method. Example 4: Consider the following system
3 2 1
+ − = 9, x+ y+z =6
x y z
x + 2 y + 3z = 10 .
2 1 2
− + = 10, x + 2 y + z = 
x y z
−1 3 4 For what values of  and  do the system has (i) unique
+ + = 30, solution (ii) no solution (iii) infinitely many solutions.
x y z
Solution:
Solution: The Augmented matrix is
1 1 1
Let = u, = v, = w 1 1 1 6 
x y z
-u + 3v + 4w = 30 (1)
A B = 1 2 3 10 R2 → R2 − R1 , R3 → R3 − R1
3u + 2v – 1w = 9 (2) 1 2   
2u – 1v + 2w = 10 (3) 1 1 1 6 
− 1 3 4 30 
= 0 1 2 4  R3 → R3 − R2
 3 2 −1 9  
  0 1  − 1  − 6
 2 − 1 2 10 
1 1 1 6 
R3 + 2R1 
= 0 1 2 4 
R2 + 3R1 
0 0  − 3  − 10
− 1 3 4 30
 0 11 11 (i) If  -3=0 and  -10  0, that is if  =3 and   10
 99
then the system does not have any solution.
 0 5 10 79 (ii)If  -3  0 then the system has a unique solution. That
R2 – R3 is   3 and  can possess any real value.
− 1 3 4 30 (iii)If  -3=0 and  -10 =0, that is if  =3 and  =10
0 6 1 29 then the system has infinite solutions.

 0 5 10 70
6R3 – 5R2
− 1 3 4 30 
 0 6 1 29 
 
 0 0 55 275
Back substitution gives
w=5, v=4, u=2
1 1 1
 x = , y = , z = is required unique
2 4 5
solution.
 ❖GAUSS-JORDAN METHOD (Reduced Row-Echelon Form):
Example: Solve the following system using Gauss-Jordan Method:

x+ y+z =3 2 x + y + 3z = 16
(1) x + 2 y − z = 4 (2) 3 x + 2 y + w = 16
x + 3y + 2z = 4 2 x + 2 z − 5w = 5
Solution:The augmented matrix is Solution: The augmented matrix is
1 1 1 3 2 1 3 0 16
R → R2 − R1 ,
A B  = 1 2 − 1 4 2 
A B  = 3 2 0 1 16 R1 → 1 R1
R → R3 − R1
1 3 2 4 3 2 0 12 − 5 5  2

1 1 1 3 1 1 / 2 3 / 2 0 8 
= 0 1
R → R1 − R2 ,
− 2 1 1 = 3 2 0 1 16  R2 → R2 − 3R1 ,
 R → R3 − 2 R2   R → R − 2R
0 2 5 1 3 2 0 12 − 5 5 
3 3 1

1 0 3 2 1 1 / 2 3 / 2 0 8 
= 0 1 − 2 1  R → 1 R 
= 0 1 / 2 − 9 / 2 1 − 8  R → 2R
 3 3   2 2
0 0 5 − 1 5 0 − 1 9 − 5 − 11
1 0 3 2  1 1 / 2 3 / 2 0 8 
  R → R1 − 3R3 , = 0 1 − 9 2 − 16
= 0 1 −2 1  1 
0 R → R2 + 2 R3 0 − 1 9 − 5 − 11
0 1 −1  2
 5
R1 → R1 − 1 / 2 R2 ,
1 0 0 13 
 5 R3 → R3 + R2
= 0 1 0 3 
 5 1 0 6 − 1 16 
0 0 1 −1  = 0 1 − 9 2 − 16  R → −1 / 3R
5   3 3
 The solution is unique because 0 0 0 − 3 − 27
13 3 −1
x = ,y = ,z = . 1 0 6 − 1 16 
R1 → R1 + R3 ,
= 0 1 − 9 2 − 16
5 5 5
  R → R − 2R
0 0 0 1 9  2 2 3

1 0 6 0 25 
= 0 1 − 9 0 − 34
 
0 0 0 1 9 
Back Substitution: w=9, y-9z=-34, x+6z=25.
z is treated as independent variable , therefore
we suppose z=k, then w=9, y=-34+9k & x=25-
6k.
 The solution set is {(25-6k, -34+9k, k, 9)/ k
 R}.

❖ HOMOGENEOUS EQUATIONS:
A system of linear equations in terms of x1 , x 2 ,....., x n having the matrix form AX=O, where
A is m x n coefficient matrix, X is n x 1 column matrix, O is a m x 1 zero column matrix is
called a system of homogeneous equations.
 x+ y+z =0
x+ y =0
For example:(i) x + 2 y − z = 0 (ii)
x + 2y = 0
x + 3y + 2z = 0
Homogeneous equations are never inconsistent. They always have the solution “all
variables = 0”. The solution (0, 0,…., 0) is often called the trivial solution. Any other
solution is called nontrivial solution.

Example1: Consider the following system: Example2: Consider the following system
4x + 3y − z = 0 − 2 x + 2 y − 3z = 0
3x + 4 y + z = 0 2x + y − 6z = 0
5x + y − 4z = 0 − x − 2 y + 2z = 0
Solution: 3x + y + 4 z = 0
 4 3 − 1 0 Solution:
 3 4 1 0

1
 R1 → R1  − 2 2 − 3 0
5 1 − 4 0 4  2 1 − 6 0
 1
1 3 / 4 − 1 / 4 0  − 1 − 2 2 0 R1 → − 2 R1
= 3 4  R2 → R2 − 3R1 ,  
 1 0  R → R − 5R  3 1 4 0
5 1 − 4 0 3 3 1
 1 − 1 3 / 2 0
2 −  R2 → R2 − 2 R1 ,
1 3/ 4 − 1 / 4 0 1 6 0
0 7 / 4  =  R → R3 + R1
7 / 4 0 4  − 1 − 2 2 0 3
  R2 → R2   R → R4 − 3R1
= 0 − 11 / 4 − 11 / 4 0 7 3 1 4 0 4
 
1 0 − 1 0 1 − 1 3 / 2 0 
0 1 1 0  0 3 − 9 0
   1
= 0 0 0 0  0 − 3 7 / 2 0  R2 → R2
=  3
x − z = 0, y + z = 0, 0 = 0.  0 4 − 1 / 2 0 
The last equation does not give any information about 1 − 1 3 / 2 0 
the equations. let z = k  y = −k and x = k . 0 1 − 3 0  R1 → R1 + R2 ,
 the solution set is {( k ,− k , k ) / k  R}   R → R3 + 3R2
0 − 3 7 / 2 0  3
=  R4 → R4 − 4 R2
0 4 − 1 / 2 0 
1 0 − 3 / 2 0 
0 1 −3 0
 2
0 0 − 11 / 2 0 R3 → − R3
=  11
 0 0 23 / 2 0 
1 0 − 3 / 2 0 
0 1 − 3 0 R1 → R1 + 3 / 2 R3 ,
=  R2 → R2 + 3R3
0 0 1 0
  R4 → R4 − 23 / 2 R3
0 0 23 / 2 0
1 0 0 0 
0 1 0 0 
 
0 0 1 0 
= 
0 0 0 0 
The required solution is x=0, y= 0, z=0.
 Rank of the matrix: Let A be an m x n matrix, the rank of A is number of nonzero rows
in reduced-row-echelon form of A and is denoted by rank (A) or  ( A).

Example: Determine the rank of the matrix A, if 1 1 1 1 1

1 1 1 1 1 0 1 2 3 4
 4 5 6 7 8  R2 → R2 − 4 R1 
  R3 → R3 − 5R1 = 0 0 0 0 0
A= 5 6 7 8 9  
  R4 → R4 − 10 R1 0 0 0 0 0
10 11 12 13 14 R → R − 5 R 0 0 0 0 0
15 16 17 18 19 5 5 1

1 1 1 1 1
0 1 2 3 4  R → R − R  rank ( A) = 2
  3 3 2

= 0 1 2 3 4  R 4 → R 4 − R 2
 
0 1 2 3 4 R5 → R5 − R2
0 1 2 3 4

RESULT:
(1) If rank (A)  rank (A, B) then the system is inconsistent.
(2) If rank (A) = rank (A, B) =  then the system is consistent.
(a) if  < n then there are infinitely many solutions. (n is the no of unknowns ).
(b) if  =n then there is a unique solution.
Example: Find the number of parameters in the general solution of AX = O if A is a 5 x 7
matrix of rank 3.
Solution: Here rank(A) =  (A)= 3 and n= 7 then number of parameters = n -  (A) = 7-3 =
4.

Eigen values and Eigen vectors

Let A be n x n matrix, then there exists a real number  and a nonzero n-vector X such that
AX =  X then,  is called as the eigen value or characteristic value or proper roots of the
matrix A, and X is called as eigen vector or characteristic vector or real vector corresponding
to eigen value  of the matrix A.

Note: An eigen vector is never the zero vector.

1.The matrix [A -  In] is known as the characteristic matrix of A.

2.The determinant of (A -  In) after expansion gives the polynomial in  , it is known as the
characteristic polynomial of the matrix A of order n x n and is of degree n.

3. A -  In =0 is called the characteristic equation of matrix A.

4. The root of the characteristic equation is known as characteristic value or eigenvalue of

the matrix.

5. The set of all characteristic roots (eigen values) of the matrix A is called the spectrum of A.
 6. Let A be n x n matrix and  be an eigen value for A. Then the set E  = { X / AX = X } is
called the eigen space of  .

Result: 1. The eigen values of a diagonal matrix are its diagonal elements.

2. The sum of eigen values of an n x n matrix is its trace and their product is A .

3. For the upper triangular (lower triangular) n x n matrix A, the eigen values are its
diagonal elements.

1 2 − 3
Example 1: If 0 2 3  then find the eigen values of AT , A9, 5A & A-1.
0 0 2 

Solution:The eigen values for AT are 1, 2, 2. The eigen values for A9 are 1, (29), (29).

The eigen values for 5A are 5, 10, 10. The eigen values for A-1 are 1, ½, ½.

Type 1:When the eigen values are nonrepeated, whether the matrix is symmetric or non-
symmetric.

− 2 − 8 − 12
Example: Find the eigen values and eigen vector of the matrix A=  1 4 4 
 0 0 1 
Solution:The characteristic equation is A -  In =0 1 8 / 3 4 0
− 2 −  − 8 − 13 = 0 1 0 0 R1 → R1 − 8 / 3R2
1 4− 4 =0 0 0 0 0
0 0 1−  1 0 4 0
 3 − S 1  2 + S 2  − A = 0 = 0 1 0 0
S1= trace(A)= -2+4+1 = 3 0 0 0 0
S2= sum of minors of diagonal entries= 4-2+0 = 2. We suppose z=k, y=0, x+4z=0
A = 1( -8+ 8) = 0 Z= k , y = 0, x = -4k
Characteristic equation is  3 − 32 + 2 = 0  x   − 4k   − 4
     
  (2 − 3 + 2) = 0  y =  0  = k  0 
  ( − 1)( − 2) = 0  z   k   1 
 1 = 0,  2 = 1, 3 = 2 Therefore eigen vector space for 2 = 1
When 1 = 0  x   − 4
 y =  0 
− 2 − 8 − 12 0    
A − I O =  1 4 4 0 R1 → −1 / 2 R1  z   1 
 0 0 1 0 When 3 = 2
1 4 6 0 − 2 − 2 − 8 − 12 0
= 1 4 4 0 R2 → R2 − R1  A − I O  =  1 4−2 4 0
0 0 1 0  0 0 1 − 2 0
 1 4 6 0 − 4 − 8 − 12 0
= 0 0 − 2 0 =  1 2 4 0 R1 → −1 / 4 R1
0 0 1 0  0 0 − 1 0
Therefore we suppose x+4y+6z = 0, -2z = 0, y = k. 1 2 3 0
= 1 2 4 0 R2 → R2 − R1
z = 0, y = k, x = -4k.
Therefore eigen vector space is
0 0 − 1 0
 x   − 4k   − 4
 y =  k  = k  1  1 2 0
3
     
 z   0   0  = 0 0 1 0 R3 → − R3
Therefore eigen vector space for 1 = 0 0 0− 1 0
 x   − 4 1 2 3 0
R → R1 − 3R3
 y =  1  = 0 0 1 0 1
    R → R2 − R3
 z   0  0 0 1 0 2
When 2 = 1 1 2 0 0
− 2 − 1 − 8 − 12 0 = 0 0 0 0
A − I O =  1 4 − 1 4 0 0 0 1 0
 0 0 1 − 1 0 We suppose z=0, y=k, x+2y=0
Z= 0 , y = k, x = -2k
− 3 − 8 − 12 0
 x   − 2k   − 2
=  1 3 4 0 R1 → −1 / 3R1  y =  k  = k  1 
 0 0 0 0      
 z   0   0 
1 8 / 3 4 0
Therefore eigen vector space for 3 = 2
= 1 3 4 0 R2 → R2 − R1
 x   − 2
0 0 0 0  y =  1 
   
1 8 / 3 4 0  z   0 
= 0 1 / 3 0 0 R2 → 3R2
0 0 0 0

❖ Algebraic multiplicity: Let A be n x n matrix, let  be an eigen value for A. If 

occurs times (k  1) then k is called the algebraic multiplicity of  , and the number
of basis vectors is called Geometric multiplicity.
Type 2: When the roots are repeated and the matrix is non-symmetric.
 − 2 2 − 3
Example: Find eigen values and eigen vectors of the matrix. A=  2 1 − 6 . Also determine
 − 1 − 2 0 
algebraic and geometric multiplicity.

Solution: The characteristic equation is A − I = 0

 −2− 2 −3
2 1−  −6
−1 −2 0−
= (−2 −  )[(1 −  )(− ) − (−2)(−6)] − 2[ 2(− ) − (−1)(−6)] − 3[2(−2) − (−1)(1 −  )]
= (−2 −  )[− + 2 − 2] − 2[−2 − 6] − 3[−4 + 1 −  ]
= −3 − 2 + 21 + 45
= −(3 + 2 − 21 − 45)
 −(3 + 2 − 21 − 45) = 0
 3 + 2 − 21 − 45 = 0
 ( + 3)(2 − 2 − 15) = 0
 ( + 3)( + 3)( − 5) = 0
 1 = 5,  2 = −3, 3 = −3
Algebraic Multiplicity of  = −3 is 2 and of  = 5 is 1.
We solve the following homogeneous system:
− 2 −  2 − 3   x1  0
 [ A − I ] X =  2 1 −  − 6   x 2  = 0
 − 1 − 2 0 −    x3  0

Case I: When 1 = 5 Case II :When 2 = −3, 3 = −3

− 2 −  2 − 3  0  − 2 −  2 − 3  0 
 [ A − I O] =  2 1 −  − 6  0 
 [ A − I O ] =  2 1 −  − 6  0
 − 1 − 2 0 −   0  − 1 − 2 0 −   0
 − 7 2 − 3 0 1 2 − 3 0
=  2 − 4 − 6 0 R1  R3  4 − 6 0
= 2
 − 1 2 − 5 0 − 1 − 2 3 0
 − 1 − 2 − 5 0 R2 → R2 − 2 R1
=  2 − 4 − 6 0 R1 → − R1 R3 → R3 + R1
− 7 2 − 3 0 1 2 − 3 0 
1 2 5 0 = 0 0 0 0
  0 0 0 0
=  2 − 4 − 6 0
− 7 2 − 3 0 Let
R2 → R2 − 2 R1 x3 = k1 , x 2 = k 2 , x1 + 2 x 2 − 3x3 = 0,
.
R3 → R3 + 7 R1 x1 = −2k1 + 3k 2
Therefore eigen space is for 2 = −3, 3 = −3 is {k1
1 2 5 0
0 − 8 − 16 0 (-2, 1, 0) + k2 (3, 0, 1) / k  R}
=  
0 16 32 0 Geometric multiplicity of  = −3 is 2 and of
R 2 → −1 / 8 R 2  = 5 is 1.
1 2 5 0
R → R1 − 2 R2
= 0 1 2 0 1
R → R3 − 16 R2
0 16 32 0 3
 1 0 1 0 
= 0 1 2 0
0 0 0 0
x3 = k , x 2 + 2 x3 = 0  x2 = −2k , x1 + x3 = 0
Let
 x1 = −k
Therefore eigen space is for 1 = 5 is {k (-1, -
2, 1)/ k  R}

Type 3: When eigen values are repeated and matrix is symmetric.

0 1 1 
Example: Find eigen values and eigen vectors of the matrix. A= 1 0 1 . Also determine algebraic
1 1 0
and geometric multiplicity.

Solution: The characteristic equation is A − I = 0

− 1 1
1 − 1
1 1 −
= − (2 − 1) − 1(− − 1) + 1(1 +  )
= −3 + 3 + 2 = −(3 − 3 − 2)
 −(3 − 3 − 2) = 0
 3 − 3 − 2 = 0
 ( − 2)( + 1) 2 = 0
 1 = 2,  2 = −1, 3 = −1
Algebraic Multiplicity of  = −1 is 2 and of  = 2 is 1.
CaseI: 1 = 2 CaseI: 2 = −1, 3 = −1
−  1 1  0  −  1 1  0 

 [ A − I O] =  1 −  1  0 
 [ A − I O] =  1 −  1  0
 1 1 −   0  1 1 −   0
− 2 1 1 0 1 1 1 0
R → R2 − R1

=  1 − 2 1 0 R2  R1 = 1 1 1 0 2
R → R3 − R1
 1 1 − 2 0 1 1 1 0 3
 1 − 2 1 0 1 1 1 0
R → R2 + 2 R1
= − 2 1 1 0 2 = 0 0 0 0
R → R3 − R1
 1 1 − 2 0 3 0 0 0 0
1 − 2 1 0
= 0 − 3 3 0 R2 → −1 / 3R2 Let x3 = k1 , x 2 = k 2 , x1 + x 2 + x3 = 0, x1 = −k1 − k 2 .
0 3 − 3 0 Therefore eigen space is for
 2 = −1, 3 = −1 is {k1 (-1, 0, 1) + k2 (-1, 1, 0) / k 
R}
 1 − 2 1 0 Geometric Multiplicity of  = −1 is 2 and of
R → R1 + 2 R2  = 2 is 1.
= 0 1 − 1 0 1
R → R3 − 3R2
0 3 − 3 0 3
1 0 − 1 0
= 0 1 − 1 0
0 0 0 0
Let
x3 = k , x 2 − x3 = 0,  x 2 = k , x1 − x3 = 0, x1 = k .
Therefore eigen space is for 1 = 2 is {k (1,1,1) /
k  R}

 1 2 2
Example: Determine algebraic and geometric multiplicity of  0 2 1  .
 
− 1 2 2
Answer:  = 1, 2, 2 therefore algebraic multiplicity of  = 2 is 2 and geometric multiplicity is 1. For  = 1
A.M. is 1 and G.M. is 1.

❖ Every square matrix can be decomposed as a sum of symmetric and skew-symmetric

matrices.
Proof: Let A be m x n matrix.
1 1
Let B = ( A + AT ) and C = ( A − AT ) be two matrices.
2 2
Obviously A= B+C.

     
T
1  1 T 1 1
Now, B =  ( A + AT ) = ( A + AT ) = ( A) T + ( AT ) T = ( AT + A) = B
2  2 2 2
Therefore, B = B , B is symmetric.
T

     
T
1  1 T 1 1
C =  ( A − AT ) = ( A − AT ) = ( A) T − ( AT ) T = ( AT − A) = −C
2  2 2 2
Therefore, − C = C , C is skew- symmetric.
T

Therefore A is a sum of symmetric and skew-symmetric matrices.

Caley –Hamilton Theorem: Every square matrix satisfies its characteristic equation i.e. The
theorem states that, for a square matrix A of order n, if A −  I = 0 .
1 4
Example (i): Using Caley-Hamilton theorem find inverse of 2 3
 .
Solution: The characteristics equation is
A − I = 0
1−  4
 =0
2 3−
 (1 −  )(3 −  ) − 8 = 0
  2 − 4 − 5 = 0
By caley-Hamilton theorem
A2 − 4 A − 5 I = 0
 A2 A−1 − 4 AA−1 − 5IA−1 = 0
 A − 4 I − 5 A−1 = 0
 5 A−1 = 4 I − A
1 4 1 0  1  1 4 
 A−1 = (4 I − A) =  −
5 5 0 1  5  2 3 
 4 / 5 0  1/ 5 4 / 5   3 / 5 −4 / 5  1  3 −4 
= − = =
 0 4 / 5  2 / 5 3 / 5   −2 / 5 1/ 5  5  −2 1 

2 1 1
0 1 0
 
Example (ii): Find the characteristics equation of the matrix 1 1 2  and hence prove that
 
8 5 5 
A − 5 A + 7 A − 3 A + A − 5 A + 8 A − 2 A + I = 0 3 0 
8 7 6 5 4 3 2

5 5 8  .
Solution: The characteristics equation is
2− 1 1
A − I = 0 1−  0 =0
1 1 2−
  3 − 5 2 + 7 − 3 = 0
By Caley-Hamilton Theorem
 A3 − 5 A2 + 7 A − 3I = 0 ..............(1)
Now,
A7 + 7 A6 − 3 A5 + A4 − 5 A3 + 8 A2 − 2 A + I
= A5 ( A3 − 5 A 2 + 7 A − 3I ) + A( A3 − 5 A 2 + 7 A − 3I ) + A 2 + A + I
= A2 + A + I u sin g (1)
2 1 1 2 1 1 2 1 1 1 0 0
 A + A + I = 0 1 0 0 1 0 + 0 1 0 + 0 1 0
2

1 1 2 1 1 2 1 1 2 0 0 1

5 4 4 2 1 1 1 0 0
= 0 1 0 + 0 1 0 + 0 1 0
4 4 5 1 1 2 0 0 1
8 5 5
= 0 3 0
5 5 8
Diagonalization of a matrix:
An n x n matrix A is diagonalizable if and only if A has n linearly independent eigenvectors.
OR
If n x n matrix A has a basis of eigenvectors, then D = P −1 AP is diagonal, with the eigenvalues
of A as the entries on the main diagonal. Here P is the matrix with these eigenvectors as column
vectors.
Also, D n = P −1 A n P and A n = PD n P −1

Example (i): Find a matrix P that diagonalizes A and determine P-1 AP

 − 1 4 − 2
A = − 3 4 0 
− 3 1 3 
 − 1 4 − 2 1 − 4 / 3 2 / 3 0
R1 → R1 + 4 / 3R2
Solution (i): A = − 3 4 0  
= 0 1 − 1 0
  R → R3 + 3R2
− 3 1 3  0 −3 3 0 3
The characteristic equation is A − I = 0 1 0 − 2 / 3 0
−1−  4 −2 = 0 1 − 1 0
−3 4− 0 =0 0 0 0 0
−3 1 3−  z = k , y − z = 0 & x − 2 / 3z = 0
 z = k , y = k , x = 2 / 3k
 3 − 62 + 11 − 6 = 0
2
 ( − 1)( − 2)( − 3) = 0  ( x, y, z ) = k ( ,1,1); k  R
3
  = 1,2,3
 ( x, y, z ) = 3k (2,3,3); k  R ( 3k = k ' )
For  = 1
E1 = {k ' (2,3,3) / k ' R}
− 1 −  4 − 2  0 

 [ A − I O ] =  − 3 4− 0  0 For  = 3
− 1 −  4 − 2  0 
 − 3 1 3 −   0 
 [ A − I O ] =  − 3 4− 0  0
 − 2 4 − 2 0
 − 3 3 −   0
=  − 3 3 0 0 R1 → −1 / 2 R1
1
 
 − 3 1 2 0  − 4 4 − 2 0
=  − 3 1 0 0 R1 → −1 / 4 R1
 1 − 2 1 0  
R2 → R2 + 3R1  − 3 1 0 0
=  − 3 3 0 0
  R → R + 3R
− 3 1 2 0 3 3 1  1 −1 1 / 2 0
R2 → R2 + 3R1
= − 3 1 0 0
1 − 2 1 0   R3 → R3 + 3R1
 0 − 3 3 0  R → −1 / 3 R − 3 1 0 0
  2 2
= 0 − 5 5 0  1 − 1 1 / 2 0
  0 − 2
 3 / 2 0 R2 → −1 / 2 R2
1 − 2 1 0
0 1 − 1 0 R1 → R1 + 2 R2
= 0 − 2 3 / 2 0

=  R → R + 5R
0 − 5 5 0 3 3 2 1 − 1 1 / 2 0
0 1 − 3 / 4 0 R1 → R1 + R2
=  R → R + 2R
0 − 2 3 / 2 0 3 3 2
 1 0 − 1 0  1 0 − 1 / 4 0
0 1 − 1 0  0 1 − 3 / 4 0 
   
= 0 0 0 0  = 0 0 
   0 0
 z = k, y − z = 0 & x − z = 0  z = k, y − 3 / 4z = 0 & x − 1/ 4z = 0
 z = k, y = k, x = k  z = k , y = 3 / 4k , x = 1 / 4k
 ( x, y, z ) = k (1,1,1); k  R 1 3
 ( x, y, z ) = k ( , ,1); k  R
E1 = {k (1,1,1) / k  R} 4 4
For  = 2  ( x, y, z ) = 4k (1,3,4); k  R ( 4k = k ' )
− 1 −  4 − 2  0  E1 = {k ' (1,3,4) / k ' R}

 [ A − I O ] =  − 3 4− 0  0 1 2 1
 − 3 1 3 −   0  P = 1 3 3
 − 3 4 − 2 0 1 3 4
= − 3 2 0 0 R1 → −1 / 3R1 1 0 0
 
− 3 1 1 0  P AP = D = 0 2 0
−1

 1 − 4 / 3 2 / 3 0 0 0 3
R2 → R2 + 3R1
= − 3 2 0 0
 R → R3 + 3R1
− 3 1 1 0 3
1 − 4 / 3 2 / 3 0
0 − 2 2 0 R2 → −1 / 2 R2

= 0 − 3 3 0


Example (ii): Find a matrix P that diagonalizes A and determine P-1 AP

1 0 
A= 10
 . Also find A and find eigenvalues of A .
2

 6 − 1 
1 0  For  = −1
Solution (ii): A =  
6 − 1 1 −  0 0
 [ A − I O ] = 
The characteristic equation is A − I = 0  6 − 1 −  0

1−  0  2 0 0
= 
=0 6 0 0
6 −1− 
y=k, 6x =0 = 2x
 (1 −  )(−1 −  ) − 0 = 0
x=0, y=k
  = 1,−1 (x, y) = { k (0, 1) / k  R }
For  = 1 1 0  1 0
1 −  0 0 P=   P −1 =  
 [ A − I O ] =   3 1 − 3 1
 6 − 1 −  0
1 0 
0 0 0  P −1 AP =  =D
=  0 − 1
6 − 2 0 
x=k, 6x-2y=0
x=k, y=3k
(x, y) = { k (1, 3) / k  R }
 1 0 
P −1 AP = D =    A = PDP
−1

 0 − 1
10 −1 1 0 110 0   1 0
A = PD P = 
10
  10   
3 1  0 (−1)  − 3 1
1 0 1 0  1 0
=   
3 1 0 1 − 3 1
1 0
A10 =  
0 1 
Eigenvalues of A2 are : 12=1 and (-1)2=1.

Quadratic Forms:
A homogeneous polynomial of second degree in real variables x1 , x 2 , x3 ,....., x n is called
Quadratic form.
For example:
(i) ax 2 + 2hxy + by 2 is a quadratic form in the variables x and y
(ii) 2 x1 x 2 + 2 x 2 x3 + 2 x3 x1 + x32 is a quadratic form in the variables x1 , x 2 , x3 .
A quadratic on Rn is a function Q define on Rn whose value at a vector x in Rn can be computed
in n variables x1 , x 2 , x3 ,....., x n by an expression of the form.

n n
Q( x) = x T Ax =  aij xi x j
i =1 j =1

Here A is known as the coefficient matrix. Where A is n x n symmetric matrix and is called
matrix of the quadratic form.

Matrix Representation of Quadratic Forms:

A quadratic form can be represented as a matrix product.
For example:
a h   x 
(i) ax 2 + 2hxy + by 2 = x y    
h b   y 
0 1 1  x1 
(ii) 2 x1 x 2 + 2 x 2 x3 + 2 x3 x1 + x = x1
2
3 x2 x3  1 0 1  x 2 
1 1 1  x3 
Example:
(i) Find a real symmetric matrix C of the quadratic form
Q = x1 + 3x 2 + 2 x3 + 2 x1 x 2 + 6 x 2 x3 .
2 2 2

1 2 0 
Solution: The coefficient matrix of Q is A = 0 3 6
0 0 2
 1 2 0 1 0 0  1 1 0
1 T  1  
So, C = symmetric matrix =  ( A + A ) = 1 0 3 6 + 2 3 0  = 1 3 3
2  2       
     0 3 2 
 0 0 2   0 6 2   

(ii) Express the following quadratic forms in matrix notation Q = x 2 − 4 xy + y 2 .

 1 − 2  x 
Solution: x 2 − 4 xy + y 2 = x y    
− 2 1   y 
Transformation (Reduction) of Quadratic form to canonical form OR Diagonalizing
Quadratic Forms:
Procedure to Reduce Quadratic form to canonical form:
1. Identity the real symmetric matrix associated with the quadratic form.
2. Determine the eigenvalues of A.
3. The required canonical form is given by
Q( x) = y T Dy = 1 y1 + 2 y 2 + 3 y 3 + ...... + n y n
2 2 2 2
........(1)
where 1 , 2 , 3 ,......., n are the eigenvalues of A and D = P T AP . The matrix P is said
to orthogonally diagonalize the quadratic form.
And equation (1) is known as canonical form.
4. Form modal matrix P (where x = Py) containing the n eigenvectors of A as n column
vectors.
Example: Reduce the quadratic form Q = 3x + 3z + 4 xy + 8 xz + 8 yz into canonical
2 2

form .
 3 2 4
 
Solution: A = 2 0 4
4 4 3
Eigenvalues for A is 3, -4/3, -1.
The canonical form of the given quadratic form is
3 0 0   y1 
y T By =  y1 y 2 y 3  0 − 4 / 3 0   y 2  = 3 y1 − 4 / 3 y 2 − y 3
 2 2 2

0 0 − 1  y 3 
Nature of quadratic form Q :
a. Positive definite if Q(x) > 0 for all x  0,
b. Negative definite if Q(x) < 0 for all x  0,
c. Indefinite if Q(x) assumes both positive and negative values.
d. Positive semidefinite if Q(x)  0 for all x.
e. Negative semidefinite if Q(x)  0 for all x.
OR
a. Positive definite if and only if the eigenvalues of A are positive,
b. negative definite if and only if the eigenvalues of A are positive,
c. Indefinite if and only if A has both positive and negative eigenvalues.
d. Positive semi-definite if and only if A has only non-negative eigenvalues.
e. Indefinite if and only if A has only non-positive eigenvalues.
Example: Describe the nature of quadratic forms.
1. Q = 3x1 + x 2 + x32 + 4 x1 x 2 + 4 x 2 x3
2 2

2. Q = 2 x1 x 2 + 2 x 2 x3 + 2 x 2 x1