ADU4303 – Unit I Session 03: Rank and Inverse of a Matrix

Using elementary transformation

Session 3

Rank and Inverse of a Matrix Using elementary

transformation
Session outline

3.1 Introduction
3.2 Rank of a Matrix
3.3 Elementary Transformations (Operations) on a Matrix
3.4 Inverse by Gauss-Jordan method
Review Questions

3.1 Introduction

In this session, we shall learn to determine the rank of a matrix and obtain the inverse of a
matrix using elementary transformations.

Learning Outcomes

By the end of this session you should be able to

▪ find the rank and inverse of a matrix using elementary row operations.

3.2 Rank of a Matrix

Given a matrix A not necessarily square, the rank of a matrix A is the positive integer
r such that there exists at least one r -rowed square sub-matrix with non-vanishing
r

determinant while the determinant of every (r + 1) or more rowed sub matrices is zero.

Thus the rank of a matrix is the largest order of a non-zero minor of that matrix. The rank
of A is denoted by r ( A ) .

3.2.1 Example
 2 1 -1 
Determine the rank of the following matrix A =  0 3

-2 .
 
 2 4 -3 

Solution

Now, the rank of A  3 since A is of 3rd order.

Copyright © The Open University of Sri Lanka 37

ADU4303 – Unit I Session 03: Rank and Inverse of a Matrix
Using elementary transformation

A = 2 ( −9 + 8 ) − 1 ( 0 + 4 ) − 1 ( 0 − 6 ) = 0
Since A = 0 implies rank of A  3 i.e., r ( A)  2
nd
Consider the determinant of 2 order sub matrices

2 1
Since = 6  0,
0 3

 r ( A ) = 2.

Result: Rank of A and A T is same.

Note 1: Rank of a null matrix is zero.

Note 2: For a rectangular matrix A of order m  n , rank of A  min ( m , n )
i .e., rank cannot exceed the smaller of m and n .
Note 3: For a square matrix of order n, if the rank = n then A  0 i.e., A is non-
singular.
Note 4: For any square matrix of order n , if the rank  n , then A = 0 i.e., A is
singular.

3.3 Elementary Row Transformations (Operations) on a Matrix

Ri j : Interchange of the ith and j th rows.

Ri ( k ) : Multiplication of every element of i th row by a non-zero scalar k .
Ri j ( k ) : Addition to the elements of ith row, of k times the corresponding elements of the
j th row.

In a similar way, elementary column transformations (operations) are denoted by

C i j , C i ( k ) , C i j ( k ) where the row in the above definitions is replaced by column.

3.3.1 Equivalent matrices

Two matrices A and B are said to be equivalent, if one matrix say A can be obtained
from B by a sequence of elementary transformations. The equivalence is denoted by
A ~ B, .

3.3.2 Row-equivalence

Two matrices A and B are said to be row-equivalent if A can be reduced to B by a

sequence of elementary row transformations or vice versa.

Copyright © The Open University of Sri Lanka 38

ADU4303 – Unit I Session 03: Rank and Inverse of a Matrix
Using elementary transformation

Result: Equivalent matrices have the same order and same rank because elementary
transformations do not alter (effect) its order and rank.

3.3.3 Reduced row-echelon form and row-echelon form

A matrix is said to be in reduced row-echelon form if it satisfies all four of the

following conditions.

1. If there are any rows of all zeros then they are at the bottom of the matrix.
2. If a row does not consist of all zeros then its first non-zero entry (i.e. the left most
non-zero entry) is a 1. This 1 is called a leading 1.
3. In any two successive rows, neither of which consists of all zeroes, the leading 1 of
the lower row is to the right of the leading 1 of the higher row.
4. If a column contains a leading 1 then all the other entries of that column are zero.

A matrix is said to be in row-echelon form if it satisfies items 1 – 3 of the reduced row-

echelon form definition.

An alternate form of the above definitions is as follows:

A matrix is said to be in row echelon form if

(i) all nonzero rows (rows with at least one nonzero element) are above any rows
of all zeroes, and
(ii) the leading coefficient (the first nonzero number from the left, sometimes
called the pivot) of a nonzero row is always strictly to the right of the leading
coefficient of the row above it.

This is an example of a matrix in row echelon form:

1 2 3
 
0 1 5
 
 0 0 1 

A matrix is said to be in reduced row echelon form if it satisfies the additional condition
that every leading coefficient is 1 and is the only nonzero entry in its column, like in this
example:

1 4 0 −3 0
 
0 0 1 2 0
 
0 1 
 0 0 0

Notice from the above definitions, that a matrix that is in reduced row-echelon form is
also in row-echelon form while a matrix in row-echelon form may or may not be in
reduced row-echelon form.

Copyright © The Open University of Sri Lanka 39

ADU4303 – Unit I Session 03: Rank and Inverse of a Matrix
Using elementary transformation

The number of non-zero rows in an Echelon form is the rank.

3.3.4 Normal form N of a matrix

Normal form N of a matrix A of rank r is one of the forms

 Ir 0 I 
N = Ir ,  ,  Ir 0,  r 
 0 0 0
where I r is an identity matrix of order r . By the application of both elementary row and
column operations, a matrix of rank r can be reduced to normal form. Then the rank of A
is r .

3.3.5 Example

Determine the rank of the following matrix

 2 1 -1 
 
A= 0 3 -2 .
 
 2 4 − 3 

Solution

Apply elementary row operations on A

2 1 −1 2 1 −1
   
R31( − 1) 0 3 −2 R32( − 1) 0 3 −2
   
~
 0 3 − 2  ~
 0 0 0 
The number of non-zero rows is two. So the rank of A is two.

3.3.6 Example

1 2 3 1 2 3
   
If A= 3 4 5 and B = 2 1 0 ,
   
 4 5 6   0 1 2 

find (i) rank of A , (ii) rank of B, (iii) rank of A + B , (iv) rank of AB and (v) rank of
BA .

Copyright © The Open University of Sri Lanka 40

ADU4303 – Unit I Session 03: Rank and Inverse of a Matrix
Using elementary transformation

Solution

R
21( − 3 )
1 2 3 1 2 3 1 2 3 1 2 3
A= 3 4 5 0
~
−2 −4  R  0 −2 −4  R  0 1 2
3 1
4  31( − 4 )  0
R
−3

−6 
32  −    2 − 
0   2   0

 5 6   2  0 0 0 0
~ ~ ~

r ( A ) = 2 since the number of non-zero rows is 2,

1 2 3 1 2 3 1 2 3 1 2 3
       
B= 2 1 0 R 0 −3 −6 R   0 −3 −6 R   0 1 2
  21( − 2 )   32 1    2 − 1   
 0 1 2   0 1 2   3  0
   0 0   3   0 0 0 
~
~ ~

 r ( B ) = 2,

R
2 4 6  5
21 −  2 4 6 2 4 6 2 4 6
   2
     
A+ B = 5 5 5 ~
0 −5 − 10 R  2  0 −5 − 10 R  1  0 1 2
    32  −    2 −   
 4 6 8  R31( −2 )  0 −2 − 4  ~  5   0 0 0  ~ 5   0 0 0 
~

 r ( A + B ) = 2,
5 7 9  R21( − 2 )  5 7 9 1 1 1  R21( −5 )  1 1 1 
  ~     ~  
AB = 11 15 19 1 1 1 R21 5 7 9 0 2 4
  R   ~   R  
31( − 3 ) 31(1)
14 19 24  ~  − 1 −2 − 3   − 1 −2 − 3  ~  0 −1 − 2 
1 1 1  1 1 1
   
R 0 1 2 R32 (1) 0 1 2 ,
1
2    ~  
~
2
 0 −1 − 2   0 0 0 
 rank ( AB ) = 2

Copyright © The Open University of Sri Lanka 41

ADU4303 – Unit I Session 03: Rank and Inverse of a Matrix
Using elementary transformation

19 25 31  −3 −3 − 3  R1 − 1   1 1 1  R21( −5 )  1 1 1 

     3   ~  
BA = 5 8 11 R13( −2 ) 5 8 11 ~ 5 8 11 0 3 6 ,
  ~    R  
31( − 1)
11 14 17   11 14 17  R32 ( −2 )  1 −2 − 5  ~  0 −3 − 6 
~

1 1 1  1 1 1
   
R 0 1 2 R32 ( 3 ) 0 1 2 ,
1
2    ~  
~
3
 0 −3 − 6   0 0 0 
 r ( BA ) = 2 .

Note: Rank of product  rank of either.

3.3.7 Example

Find rank of matrix A where

 -2 -1 −3 -1
 
1 2 3 -1
A= .
 1 0 1 1
 
 0 1 1 -1

Solution

 1 2 3 -1 1 2 3 -1  R  1  1 2 3 -1 1 2 3 -1

  R21( 2 )   2  3    R32 ( −1)  
-2 -1 −3 -1 ~ 0 3 3 -3 0 1 1 -1 ~ 0 1 1 -1
R21     ~    ,
~  1 0 1 1  R31( −1)  0 -2 -2 2  R  1 0 1 1 -1 R42 ( −1)  0 0 0 0
  ~   3 −    ~  
0 1 1 -1 0 1 1 -1  ~ 2  0 1 1 -1 0 0 0 0

 rank ( A ) = 2

3.3.8 Example

1 1 − 1
If A =  2 b

2 then determine the values of b such that the rank of A is 2.
 
 9 9 b 

Copyright © The Open University of Sri Lanka 42

ADU4303 – Unit I Session 03: Rank and Inverse of a Matrix
Using elementary transformation

Solution
R
21( − 2 )
1 1 − 1 1 1 −1 
  ~  
2 b 2 0 b−2 4
  R  
 9 9 b  31( − 9 )  0 0 b + 9 
~

Cases:
(i) if b = 2, A = 0, rank of A = 2
(ii) if b = − 9, no. of non-zero rows is 2, rank of A = 2 .

3.3.9 Activity

Find ranks of matrices given below.

2 3 -1 -1 
1 1 -1 1 5 2 4  
    3. 
1 -1 -2 -4

1. 2 -3 4 2. 0 1 3 1
    3 1 3 -2 
 3 -2 3   0 0 1 3   
6 3 0 -7 

2 −1 3 4
   −1 −2 − 1
4. 
0 3 4 1
  
5. 6 12 6
2 3 7 5  
   5 10 5 
2 5 11 6 

Answers
1. 2 2. 2 3. 3 4. 3 5. 1

3.4 Inverse by Gauss-Jordan method

In the Gauss-Jordan method we compute the inverse of a matrix by applying elementary

row operations. Consider the augmented matrix ( A I ) and apply elementary row
operations on both matrices A and I until A gets transformed to I .

3.4.1 Example
1 2 3
 
Find the inverse of A by Gauss-Jordan method where A =  2 4 5 .
3 6 
 5

Copyright © The Open University of Sri Lanka 43

ADU4303 – Unit I Session 03: Rank and Inverse of a Matrix
Using elementary transformation

Solution

Consider ( A | I ) and apply elementary row operations on both A and I until A gets
transformed to I .

1 2 3  1 0 0 ~ 1 2 3  1 0 0
  R 21 ( − 2 )  
2 4 5  0 1 0 0 0 3  -2 1 0
   
 3 5 6  0 0 1  R31 ( − 3 )  0 -1 1  -3 0 1 
R23 1 2 3 1 0 0
R 2 ( − 1)  
~ 0 1 3 2 0 -1
 
R3 ( − 1 )  0 0 1 3 -1 0 

R23( −3 ) 1 2 0 -5 3 0
 
~ 0 1 0 -3 3 -1
R13( −3 )  
 0 0 1 2 -1 0 
1 0 0  1 -3 2
 
R12 ( − 2 ) ~ 0 1 0  -3 3 -1
 
 0 0 1  2 -1 0 

 1 −3 2
−1  
Thus A = −3 3 −1
 
 2 −1 0 

3.4.2 Activity

Find the inverse of each of following matrices using Gauss-Jordan elimination:

2 4 3 2
1 3 3   0 1 3
  3 6 5 2  
1. 1 4 3 2.   3. 1 2 3
  2 5 2 − 3  
1 3 4     3 1 1 
4 5 14 14 

- 1 - 3 - 1
3
9 7 3  
  1 1 -1 0
4. 5 -1 4 5.  
  2 -5 2 - 3
 3 4 1   
- 1 1 0 1

Copyright © The Open University of Sri Lanka 44

ADU4303 – Unit I Session 03: Rank and Inverse of a Matrix
Using elementary transformation

Answers
 − 23 29 − 64 − 18 

 7 −3 −3   5 5 
26 7
   10 − 12 5 
1. A −1
= −1 1 0 2. A −1 =  5
2 
  −2 6
 − 1 0 1   1 5 5 
 2 −2 3 1 
 5 5 
 1 −1 1  -17 5 31 
1
−1  −1 1  
3. A = −8 6 -2 4. A = − 7 0 - 21
2  35  
 5 -3 1   23 -15 -44 
 0 2 1 3
 
1 1 −1 −2
5. A = 
−1 
 1 2 0 1
 
 −1 1 2 6

Review Questions

1. Find ranks of each of the matrices given below.

 1 2 −3 4 9
  0 1 -3 -1 
1 0 −1 1 1 1 2 3  
 
 3   
0 0 1 1

a. -1 1 0 −1  b. 1 4 2 c.
    3 1 0 2
 -1 1 0 2 9  2 6 5   
 3 1 1 -2 0
1 0 3 9 

1 2 3 0 
3 0 2 2  1 2 −1 4  
    f. 
2 4 3 2

d. -6 42 24 54 e. 2 4 3 5
    3 2 1 3 
 21 -21 0 -15   − 1 − 2 6 − 7   
6 8 7 5 

2. Find the inverse of following matrices by using Gauss-Jordan elimination.

1 1 3  4 -1 1
   
a. 1 3 -3 b. 2 0 -1
   
 -2 -4 -4   1 -1 3 

Copyright © The Open University of Sri Lanka 45

ADU4303 – Unit I Session 03: Rank and Inverse of a Matrix
Using elementary transformation

Answers

(1) a. 4 b. 2 c. 2 d. 2 e. 2 f. 3

 3
 3 1
2
   -1 2 1
−1
=
5 1 3 −1  
(2) a. A - - - b. A = -7 11 6
 4 4 4  
   -2 3 2 
1 1 1
 - - - 
 4 4 4 

Copyright © The Open University of Sri Lanka 46