3rd Topic

Matrices
Rank, Inverse and Normal form of a Matrix
Gauss-Jordan method of finding the inverse

Prepared by
Prof. Sunil
Department of Mathematics & Scientific Computing
NIT Hamirpur (HP)

Rank of a Matrix:
Definition: A matrix is said to be of rank r, when
(i) it has at least one non-zero minor of order r, and
(ii) every minor of order higher than r vanishes.
Briefly, the rank of a matrix is the largest order of any non-vanishing minor of the
matrix.
Another definition:
A number r is said to be the rank of a matrix A if it possesses the following
two properties:
1. There is atleast one square sub-matrix of A of order r, whose determinant
is not equal to zero,
2. If the matrix A contains any square sub-matrix of order (r +1), then the
determinant of every square sub-matrix of A of order (r +1) should be zero.

Remarks: If a matrix has a non-zero minor of order r, its rank is ≥ r .

If all minors of a matrix of order (r + 1) are zero, its rank is ≤ r .

Minor: In linear algebra, a minor of a matrix A is the determinant of some smaller

square matrix, cut down from A by removing one or more of its rows or columns. Minors
obtained by removing just one row and one column from square matrices (first minors)
Matrices: Rank, Inverse, Normal form 2
Visit: https://www.sites.google.com/site/hub2education/
are required for calculating matrix cofactors, which in turn are useful for computing both
the determinant and inverse of square matrices.

Elementary transformations of a matrix:

The following operations, three of which refer to rows and three to columns
are known as elementary transformations:
I. The interchange of any two rows (columns).
II. The multiplication of any row (column) by a non-zero number.
III. The addition of a constant multiple of the elements of any row (column)
to the corresponding elements of any other row (column).
Notation:
The elementary row transformation will be denoted by the following symbols:
(i) R i ↔ R j for the interchange of the ith and jth rows.

(ii) Ri → kRi for multiplication of the ith row by k.

(iii) R i → R i + pR j for addition to the ith row, p times the jth row.

The corresponding column transformation will be denoted by writing C in place of R.

Elementary transformations do not change either the order or a rank of a matrix. While
the value of the minors may get changed by the transformation (i) and (ii), their zero or
non-zero character remains unaffected.
Equivalent matrix:
Definition: Two matrices A and B are said to be equivalent if one can be obtained from
the other by a sequence of elementary transformations. Two equivalent matrices have the
same order and the same rank. The symbol ~ is used for equivalence.
 1 2 3
Example: Determine the rank of the following matrices: 1 4 2 .
2 6 5

1 2 3 
Sol.: Let A = 1 4 2 .
2 6 5
Matrices: Rank, Inverse, Normal form 3
Visit: https://www.sites.google.com/site/hub2education/

1 2 3 
Operating R 2 → R 2 − R1 and R 3 → R 3 − 2R1 , we get A ∼  0 2 −1 .
 0 2 −1

1 2 3 
Operating R 3 → R 3 − R 2 , we get A ∼  0 2 −1 .
 
 0 0 0 

Obviously, the 3rd order minor of A vanishes.

Also its 2nd order minors formed by its 2nd and 3rd rows are all zero. But another 2nd order
1 3
minor is = −1 ≠ 0 .
0 −1

∴ ρ(A) = 2 .
Now since the rank of a matrix is the largest order of any non-vanishing minor of the
matrix
Hence, the rank of the given matrix is 2.

Elementary matrices:
Definition: An elementary matrix is that, which is obtained from a unit matrix, by
subjecting it to any of the elementary transformations.
1 0 0
Examples of elementary matrices obtained from I3 = 0 1 0
0 0 1

1 0 0 1 0 0  1 p 0
are R 23 = 0 0 1 = C 23 ; kR 2 = 0 k 0 ; R1 + pR 2 = 0 1 0 .
   
0 1 0 0 0 1  0 0 1

Theorem:
Statement: Every elementary row (column) transformations of a matrix A can be
obtained by pre-multiplying (post-multiplying) A by the corresponding elementary
matrix.
Matrices: Rank, Inverse, Normal form 4
Visit: https://www.sites.google.com/site/hub2education/

 a1 b1 c1 
Remarks: Consider the matrix A = a 2 b2 c 2  .
 a 3 b3 c3 

1 0 0  a1 b1 c1   a1 b1 c1 
Then R 23 × A = 0 0 1 a 2 b2 c 2  =  a 3 b3 c3  .
 
0 1 0  a 3 b3 c3  a 2 b2 c 2 

So a pre-multiplication by R 23 has interchanged the 2nd and 3rd rows of A. Similarly pre-

multiplication by kR 2 will multiply the 2nd row of A by k and pre-multiplication by

R1 + pR2 will result in the addition of p times the 2nd row of A to its 1st row.
Thus the pre-multiplication of A by elementary matrices results in the
corresponding elementary row transformation of A.
Similarly, it can also be seen that post-multiplication will perform the elementary column
transformations.

Normal form of a matrix:

Every non-zero matrix A of rank r, can be reduced by a sequence of
I r 0
elementary transformations, to the form  called the normal form of A.
0 0
(i)
Remarks:
(i) The rank of the matrix A is r if and only if it can be reduced to the normal form (i).
(ii) Since each elementary transformation can be affected by pre-multiplication or post-
multiplication with a suitable elementary matrix and each elementary matrix is non-
singular, therefore, we have the following result:
Corresponding to every matrix A of rank r, there exist non-singular
I r 0
matrices P and Q such that PAQ =  .
0 0
If A be a m× n matrix, then P and Q are square matrices of orders m and n,
respectively.
Matrices: Rank, Inverse, Normal form 5
Visit: https://www.sites.google.com/site/hub2education/

1 1 2

Example: For the matrix A = 1 2 3  , find non-singular matrices P and Q such
0 − 1 − 1

that PAQ is in the normal form.

1 1 2  1 0 0 1 0 0
Sol.: We write A = I A I ⇒ 1 2 3  = 0 1 0 A 0 1 0 .

0 − 1 − 1 0 0 1 0 0 1 

We shall affect every elementary row (column) transformation of the product by

subjecting the pre-factor (past factor) of A to the same.
Operating C 2 → C 2 − C1 , C3 → C3 − 2C1 , we get

1 0 0  1 0 0 1 − 1 − 2
1 1 1  = 0 1 0 A 0 1 0  .

0 − 1 − 1 0 0 1 0 0 1 

1 0 0   1 0 0 1 − 1 − 2
Operating R 2 → R 2 − R1 , we get 0 1 1  = − 1 0 0 A 0 1
 0  .
0 − 1 − 1  0 0 1 0 0 1 

1 0 0  1 0 0 1 − 1 − 1
Operating C3 → C3 − C 2 , we get 0 1 0 = − 1 1 0 A 0 1 − 1 .
0 − 1 0  0 0 1 0 0 1 

1 0 0  1 0 0 1 − 1 − 1
Operating R 2 → R 2 + R 2 , we get 0 1 0 = − 1 1 0 A 0 1 − 1 ,
     
0 0 0  1 1 1 0 0 1 

I 0
which is the required normal form  2 .
0 0

 1 0 0 1 − 1 − 1
Hence P = − 1 1 0 , Q = 0 1 − 1 and ρ( A ) = 2 . Ans.
 

 1 1 1  0 0 1 
Matrices: Rank, Inverse, Normal form 6
Visit: https://www.sites.google.com/site/hub2education/

Gauss-Jordan method of finding the inverse:

Johann Carl Friedrich Gauss Wilhelm Jordan

30 April 1777 – 23 February 1855 1 March 1842 – 17 April 1899
German mathematician and scientist German geodesist
It is named after Carl Friedrich Gauss and Wilhelm Jordan, because it is a
modification of Gaussian elimination as described by Jordan in 1887. However, the
method also appears in an article by Clasen published in the same year. Jordan and
Clasen probably discovered Gauss–Jordan elimination independently.

Statement: Those elementary row transformations, which reduce a given square

matrix A to the unit matrix, when applied to unit matrix I, give the
inverse of A.
Proof: Let the successive row transformations, which reduce A to I, result from pre-
multiplication by the elementary matrices R1, R 2 , .........., R i so that

R iRi −1.................R 2R1A = I

∴ R iR i −1.................R 2R1AA−1 = IA−1

⇒ R i R i −1.................R 2R1I = A−1 . [∵ AA −1

=I ]
Hence the result.
Matrices: Rank, Inverse, Normal form 7
Visit: https://www.sites.google.com/site/hub2education/

For practical evaluation of A −1 , the two matrices A and I are written side by side and the
same row transformations are performed on both. As soon as A is reduced to I, the other

matrix represents A −1 .
Example: Using the Gauss-Jordan method, find the inverse of the matrix
1 1 3 
1 3 − 3 .

 − 2 − 4 − 4

Sol.: Writing the same matrix side by side with the unit matrix of order 3, we have
1 1 3 : 1 0 0
1 3 − 3 : 0 1 0 .

− 2 − 4 − 4 : 0 0 1

1 1 3 : 1 0 0
Operating R 2 → R 2 − R1 and R 3 → R 3 + 2R1 , we get ~ 0 2 − 6 : − 1 1 0 .
0 − 2 2 : 2 0 1 

 
1 1 3 : 1 0 0
1 1  1 1 
Operate R 2 → R 2 and R 3 → R 3 , we get ~ 0 1 − 3 : − 0 .
2 2 0 − 1 1 : 2 2 
 1
1 0
 2 

 3 1 
 − 0
2 2
1 0 6 : 
1 1
Operating R1 → R1 − R 2 and R 3 → R 3 + R 2 , we get ~ 0 1 −3 : − 0
 2 2 
0 0 −2 : 1 1 1 

 2 2 2 
.
3  1
Operate R1 → R1 + 3R 3 , R 2 → R 2 − R 3 and R 3 →  − R 3 , we get
2  2
Matrices: Rank, Inverse, Normal form 8
Visit: https://www.sites.google.com/site/hub2education/

 3 
1 0 0 : 3 1
2 
 5 1 3
~ 0 1 0 : − − − .
 4 4 4
 0 0 1 : 1 1 1
 − − − 
4 4 4

 3 
 3 1
2 
 5 1 3
Hence, the inverse of the given matrix is − − −  . Ans.
 4 4 4
− 1 1 1 
 4 − 4 − 4 

Problems for finding rank of a matrix:

Q.No.1.: Prove that the row equivalent matrices have the same rank.
Sol.: Let A be any m× n matrix. Let B be a matrix row equivalent to A.
Since B is obtainable from A by a finite chain of E-row operations and every E-row
operation is equivalent to pre-multiplication by the corresponding E-matrix, there exist
E-matrices E1, E2 ,............,Ek each of the type m× m such that

B = Ek Ek −1..............E2E1A ⇒ B = PA ,
where P = Ek E k −1..............E2E1 is a non-singular matrix of the type m× m .

 p11 p12 .... p1m   R1 

p p 2 m   R 2 
 21 p 22 ....
Let us write B = PA =  .... .... .... ....   ....  , (i)
  
 .... .... .... ....   .... 
p m1 p m 2 .... p mm  R m 

where the matrix A has been expressed as a matrix of its row sub-matrices
R1, R 2 ,............, R m .
From the product of the matrices on R. H. S. of (i) we observe that the rows of the matrix
B are
p11R1 + p12R 2 + .........+ p1mR m ,
Matrices: Rank, Inverse, Normal form 9
Visit: https://www.sites.google.com/site/hub2education/

p 21R1 + p 22 R 2 + ......... + p 2 m R m ,
...................................................
...................................................
pm1R1 + pm2R 2 + .........+ pmmR m .
Thus, we see that the rows of B are all linear combinations of the rows
R1, R 2 ,............, R m of A.
Therefore, every member of the row space of B is also a member of the row space of A.

Similarly, by writing A = P−1B and giving the same reasoning we can prove that every
member of the row space of A is also a member of the row space of B.
Therefore the row space of A and B are identical.
Thus we see that elementary row operations do not alter the row space of a matrix
remains invariant under E-rows transformations.
Note: From the above theorem we also conclude that pre-multiplication by a non-singular
matrix does not alter the row rank of a matrix.
0 1 − 3 − 1
1 0 1 1 
Q.No.2.: Determine the rank of the following matrices: 
3 1 0 2
 
1 1 −2 0 

0 1 − 3 − 1
1 0 1 1 
Sol.: Let A =  .
3 1 0 2
 
1 1 −2 0 

0 1 − 3 − 1
1 0 0 0 
Operating C3 → C3 − C1, C4 → C 4 − C1 , we get A =  .
3 1 − 3 − 1
 
1 1 − 3 − 1

0 1 − 3 − 1
1 0 0 0 
Operating R 3 → R 3 − R1, R 4 → R 4 − R1 , we get A =  .
3 0 0 0
 
1 0 0 0
Matrices: Rank, Inverse, Normal form 10
Visit: https://www.sites.google.com/site/hub2education/

0 1 − 3 − 1
1 0 0 0 
Operating R 3 → R 3 − 3R 2 , R 4 → R 4 − R 2 , we get A =  .
0 0 0 0
 
0 0 0 0

0 1 0 0
1 0 0 0
Operating C3 → C3 + 3C2 , C 4 → C4 + C 2 , we get A =  .
0 0 0 0
 
0 0 0 0
Obviously, the 4th order minor of A is zero. Also every third order minor of A is zero.
0 1 
But, of all the 2nd order minors, only   = −1 ≠ 0 . ∴ ρ( A ) = 2 .
1 0
Hence, the rank of the given matrix is 2.
 3 − 1 2
Q.No.3: Determine the rank of the matrix − 6 2 4 .
 − 3 1 2

 3 − 1 2
Sol.: Let A = − 6 2 4 .
 
 − 3 1 2

3 − 1 2 
 1 
Operating R 3 → 2R 3 − R 2 , R 2 →  − R 2  , we get A = 3 − 1 − 2 .
 2 
0 0 0 

 3 − 1 2
Operating R 2 + 2R1 , we get A = 0 0 8  .
0 0 4

3 − 1 2 
Operating R 2 → R 2 − R1 , we get A = 0 0 8  .
0 0 0

∴ ρ( A ) = 2

Hence, the rank of the given matrix is 2.

Matrices: Rank, Inverse, Normal form 11
Visit: https://www.sites.google.com/site/hub2education/

 2 − 1 3 1 1 3 4 3
Q.No.4: Determine the rank of the matrix (i) 1 4 − 2 1 , (ii) 3 9 12 3 .
 
5 2 4 3 1 3 4 1

 2 − 1 3 1
Sol.: (i) Let A = 1 4 − 2 1 .
 
5 2 4 3

 1 − 1 3 1
Operating C1 → C1 − C4 , we get A = 0 4 − 2 1 .
 
 2 2 4 3

1 − 1 3 1
Operating R 3 → R 3 − 2R1 , we get A = 0 4 − 2 1
0 4 − 2 1

1 − 1 3 1 
Operating R 3 → R 3 − R 2 , we get A = 0 4 − 2 1  .
0 0 0 0

1 −1 3 −1 3 1
Now since all 3× 3 matrices are singular ∵ 0 4 − 2 = 0 and 4 −2 1 = 0.
0 0 0 0 0 0

−1 3
Now ∵ ≠ 0 . Hence, the rank of the given matrix is 2.
4 −2

1 3 4 3
(ii). Let A = 3 9 12 3 .
1 3 4 1

1 3 4 3
Operating R 3 → 3R 3 − R 2 , R 2 → R 2 , we get A = 1 3 4 1 .
1
3
0 0 0 0

1 3 4 3 
Operating R 2 → R 2 − R1 , we get A = 0 0 0 − 2 .
0 0 0 0 

Hence, the rank of the given matrix is 2.

Matrices: Rank, Inverse, Normal form 12
Visit: https://www.sites.google.com/site/hub2education/

1 2 3 0
2 4 3 2
Q.No.5: Determine the rank of the matrix  .
3 2 1 3
 
6 8 7 5

1 2 3 0
2 4 3 2
Sol.: Let A =  .
3 2 1 3
 
6 8 7 5

Operating R 2 → R 2 − 2R 1 , R 3 → R 3 − 3R 1 , R 4 → R 4 − 6R 1 , we get

1 2 3 0
0 0 − 3 2
A= .
0 − 4 − 8 3
 
0 − 4 − 11 5

1 2 3 0
0 0 − 3 2 
Operating R 4 → R 4 − R 3 , we get A =  .
0 − 4 − 8 3 
 
0 0 − 3 2 
1 2 3 0
0 0 − 3 2 
Operating R 4 → R 4 − R 2 , we get A =  .
0 − 4 − 8 3 
 
0 0 0 0

2 3 0
As 4 × 4 matrix is singular. But 3× 3 matrix like 0 − 3 2 is non-singular.
− 4 −8 3

So the rank of given matrix is 3.

2 3 − 1 − 1 
1 − 1 − 2 − 4 
Q.No.6: Determine the rank of the matrix  .
3 1 3 − 2
 
6 3 0 − 7
Matrices: Rank, Inverse, Normal form 13
Visit: https://www.sites.google.com/site/hub2education/

2 3 − 1 − 1
1 − 1 − 2 − 4 
Sol.: Let A =  .
3 1 3 − 2
 
6 3 0 − 7

2 3 − 1 − 1
1 − 1 − 2 − 4 
Operating R 4 → R 4 − 3R1 , R 3 → R 3 − 3R 2 , we get A =   ,
0 4 9 10 
 
 0 − 6 3 − 4
2 3 − 1 − 1
0 − 5 − 3 − 7 
Operating R 2 = 2R 2 − R1 , we get A =  .
0 4 9 10 
 
0 − 6 3 4

2 3 − 1 0 
0 − 5 − 3 − 10
Operating C 4 → C 4 + C3 , we get A =  .
0 4 9 19 
 
0 6 3 −1 

2 3 − 1 − 1
0 − 11 − 3 − 7
Operating C2 → C 2 − 6C4 , C3 → C3 + 3C 4 , we get A =  .
0 22 9 10 
 
0 0 3 − 4

2 3 − 1 − 1
0 − 11 − 3 − 7
Operating R 3 → R 3 + 2R 2 , we get A =  .
0 0 3 − 4
 
0 0 3 − 4

2 3 − 1 − 1
0 − 11 − 3 − 7 
Operating R 4 → R 4 − R 3 , we get A =  .
0 0 3 − 4
 
0 0 0 0 

As 4 × 4 matrix is singular. But 3× 3 matrix is non-singular.

So the rank of the matrix is 3. Ans.
Matrices: Rank, Inverse, Normal form 14
Visit: https://www.sites.google.com/site/hub2education/

5 6 7 8
6 7 8 9
Q.No.7: Determine the rank of the matrix  .
11 12 13 14
 
16 17 18 19
5 6 7 8
6 7 8 9
Sol.: Let A =  .
11 12 13 14
 
16 17 18 19
5 6 7 8
6 6 8 9
Operating R 3 → R 3 − R1, R 4 → R 4 − R 2 , we get A =  .
6 6 6 6
 
10 10 10 10
5 6 7 8
1 1 1 1
Operating R 2 → R 2 − R1 , we get A =  .
6 6 6 6
 
10 10 10 10
3 2 1 8
0 0 0 1 
Operating C1 → C1 − C 4 , C 2 → C 2 − C 4 , C3 → C3 − C 4 , we get A =  .
0 0 0 6
 
0 0 0 10

3 2 1 8
0 0 0 1 
Operating R 4 → R 4 − 10R 2 , R 3 → R 3 − 6R 2 , we get A =  .
0 0 0 0
 
0 0 0 0
Hence the rank of the given matrix is 2. Ans.
Q.No.8.: Find the rank of matrix
2 3 − 1 − 1 
 2 3 4 − 1 1 − 1 − 2 − 4 
 1 2 3 4    .
(i).   (ii).  5 2 0 − 1 (iii). 3 1
 − 2 0 5 7  3 − 2 
 − 4 5 12 − 1  
6 3 0 − 7

 1 2 3 4
Sol.: (i). Here A =   is a 2 × 4 matrix.
− 2 0 5 7
Matrices: Rank, Inverse, Normal form 15
Visit: https://www.sites.google.com/site/hub2education/

∴ρ(A) ≤ 2 , the smaller of 2 and 4.

1 2
The second order minor =4≠0 ∴ρ(A) = 2 .
−2 0

 2 3 4 − 1
(ii). Here A =  5 2 0 − 1 is a 3× 4 matrix.
 
 − 4 5 12 − 1

∴ ρ(A) ≤ 3 .

− 1 3 4 2 
Operating C14 , we get A = − 1 2 0 5 
 − 1 5 12 − 5

− 1 3 4 2
Operating R 2 − R 1 , R 3 − R 1 , we get A = 0 − 1 − 4 3 

 
 0 2 8 − 6

− 1 0 − 8 11 
Operating R 1 + 3R 2 , R 3 + 2R 2 , we get A =  0 − 1 − 4 3  .
 0 0 0 − 0

All the first order minors are zero but the second order minor
−1 0
=1≠ 0. ∴ρ(A) = 2 .
0 −1

2 3 − 1 − 1 
1 − 1 − 2 − 4 
(iii). Here A =   is a 4 × 4 matrix. ∴ρ(A) ≤ 4
3 1 3 − 2
 
6 3 0 − 7

1 − 1 − 2 − 4 
2 3 − 1 − 1 
Operating R12, we get A =  .
3 1 3 − 2
 
6 3 0 − 7
Matrices: Rank, Inverse, Normal form 16
Visit: https://www.sites.google.com/site/hub2education/

1 0 0 0
2 5 3 7 
Operating C 2 + C1 , C 3 + 2C1 , C 4 + 4C1 , , we get A =  .
3 4 9 10
 
6 9 12 17 

1 0 0 0
0 5 3 7 
Operating R 2 − 2R 1 , R 3 − 3R 1 , R 4 − 6R 1 , we get A =  .
0 4 910 
 
0 9 12 17 

1 0 0 0
0 1 − 6 − 3
Operating R 2 − R 3 , R 4 − 2R 3 , we get A =  .
0 4 9 10 
 
0 1 − 6 − 3

1 0 0 0
0 1 0 0 
Operating C 3 + 6C 2 , C 4 + 3C 2 , we get A =  .
0 4 33 22
 
0 1 0 0

1 0 0 0
0 1 0 0 
Operating R 3 − 4R 2 , R 4 − R 2 , we get A =  .
0 0 33 22 
 
0 0 0 0

1 0 0 0
0 1 0 0 
1
Operating C 3 , we get A =  .
33 0 0 1 22
 
0 0 0 0

1 0 0 : 0
0 1 0 : 0
  3
I : O 3×2 

Operating C 4 − 22C 3 , we get A =  0 0 1 : 0  =  .... .... ....  .

 
.... .... .... : .... O1×3 : O1×1 
 0 0 0 : 0 

∴ ρ(A) = 3 .
Matrices: Rank, Inverse, Normal form 17
Visit: https://www.sites.google.com/site/hub2education/

Problems for inverse of a matrix by Gauss-Jordan method:

Q.No.1.: Use Gauss-Jordan method to find inverse of the following matrices:
 2 1 − 1 8 4 3  2 1 2  1 1 2
(i) 0 2 1  , (ii) 2 1 1 , (iii) 2 2 1 , (iv) 1 2 3 .
       
5 2 − 3 1 2 1 1 2 2 2 3 1

 2 1 − 1
Sol.: (i). Given matrix is 0 2 1  .
 
5 2 − 3

Writing the same matrix side by side with the unit matrix of order 3., we have
2 1 − 1 : 1 0 0
0 2 1 : 0 1 0 .
 
5 2 − 3 : 0 0 1 

2 1 − 1 : 1 0 0 
Operating R 3 → 2R 3 − 5R1 , we get 0 2 1 : 0 1 0  .
0 − 1 − 1 : − 5 0 2

2 1 − 1: 1 0 0
Operating R 3 → 2R 3 − R 2 , we get 0 2 1 : 0 1 0 .
0 0 − 1: − 10 1 4

 2 1 0 : 11 − 1 − 4
Operating R 2 → R 2 + R 3 , R1 → R1 − R 3 , we get 0 2 0 : − 10 2 4  .
0 0 − 1 : − 10 1 4 

1
Operating R1 → 2R1 − R 2 , R 2 → R 2 , R 3 → (− 1)R 3 , we get
2
4 0 0 : 32 − 4 − 12
0 1 0 : − 5 1 2  .

0 0 1 : 10 − 1 − 4 

1 0 0 : 8 − 1 − 3
Operating R1 → R1 , we get 0 1 0 : − 5 1 2  .
1
4
0 0 1 : 10 − 1 − 4
Matrices: Rank, Inverse, Normal form 18
Visit: https://www.sites.google.com/site/hub2education/

 8 − 1 − 3
Hence, the inverse of the given matrix is − 5 1 2  . Ans.
 10 − 1 − 4

8 4 3
(ii). Given 2 1 1 .
 
1 2 1

Writing the same matrix side by side with the unit matrix of order 3., we have
8 4 3 : 1 0 0
 2 1 1 : 0 1 0 .
 
1 2 1 : 0 0 1

8 4 3 : 1 0 0
Operating R 3 → 2R 3 − R 2 , R 2 → 4R 2 − R1 , we get 0 0 1 : − 1 4 0 .

 
0 3 1 : 0 − 1 2

8 4 3 : 1 0 0
Operating R 2 → R 2 − R 3 , we get 0 − 3 0 : − 1 5 − 2 .

 
0 3 1 : 0 − 1 2 

8 4 3 : 1 0 0
 1  1 5 2
Operating R 3 → R 3 + R 2 , R 2 →  − R 2 , we get 0 1 0 : − .
 3 0 0 1 : −31 43 3
 0

8 4 0 : 4 - 12 0
 1 5 2
Operating R1 → R1 − 3R 3 , we get 0 1 0 : − .
0 0 1 : −31 43 3 
 0

 8 16 8
8 0 0 : 3 − −
3 3
 1 5 2 

Operating R1 → R1 − 4R 2 , we get 0 1 0 : − .
 3 3 3 
0 0 1 : − 1 4 0 
 
Matrices: Rank, Inverse, Normal form 19
Visit: https://www.sites.google.com/site/hub2education/

 1 2 1
1 0 0 : 3 − 3 − 3 
1  1 5 2 
Operating R1 → R1 , we get 0 1 0 : − 
8  3 3 3 
0 0 1 : − 1 4 0 
 

1 2 1
3 −3 − 
3
1 5 2 
Hence the inverse of given matrix is  − .
3 3 3 
− 1 4 0 
 

 1 − 2 − 1
=  1 − 5 2  . Ans.
−1 1
∴A
9
 − 1 4 0 

 2 1 2
(iii). Let A = 2 2 1 .
1 2 2

According to Gauss-Jordan method, we have A = IA .

2 1 2 1 0 0
⇒ 2 2 1  = 0 1 0 A
1 2 2 0 0 1

1 2 2 0 0 1
Operating R 3 → R 3 − R1 , we get 2 2 1 = 0 1 0 A .
2 1 2 1 0 0

1 2 2  0 0 1 
Operating R 3 → R 3 − R 2 , R 2 → R 2 − 2R1 , we get 0 − 2 − 3 = 0 1 − 2 A .
  
0 − 1 1  1 − 1 0 

1 0 − 1 0 1 − 1 
Operating R1 → R1 + R 2 , R 3 → 2R 3 − R 2 , we get 0 − 2 − 3 = 0 1 − 2 A .
0 0 5  2 − 3 − 2
Matrices: Rank, Inverse, Normal form 20
Visit: https://www.sites.google.com/site/hub2education/

 
1 0 − 1  0 1 − 1
1
Operating R 3 → R 3 , we get 0 − 2 − 3 =  0 1 − 2 A .
5   2 3 2 
0 0 1   − 
5 5 5 

2 2 3
 − 
1 0 0  5 5 5

0 − 2 0 =  6 − 4 − 4  A
Operating R1 → R1 + R 3 , R 2 → R 2 + 3R 3 , we get,   5 .
5 5
0 0 1  2 3 2 
 5 − 5 5 

 2 2 3
 − 
1 0 0  5 5 5

R2 0 1 0 = − 3 2 − 2  A
Operating R 2 → − , we get,    5 5 ⇒ I = A −1A .
2 5
0 0 1  2 3 2 
 5 − 5 5 

 2 2 3
 5 − 
5 5
 3 2 2
∴ A = −
−1
−  . Ans.
 5 5 5
 2 3 2 
 5 − 5 5 

1 1 2 
(iv). Let A = 1 2 3 .
2 3 1

1 1 2 1 0 0
According to Gauss-Jordan method, we have A = IA ⇒ 1 2 3 = 0 1 0 A .
2 3 1  0 0 1

1 1 2  1 0 0
Operating R 2 → R 2 + R1 , we get 0 − 1 − 1 = 0 − 1 0 A .

2 3 1  0 0 1 

1 1 2  1 0 0
Operating R 3 → −2R1 − R 3 , we get 0 − 1 − 1 = 0 − 1 0  A .
  
0 − 1 3   2 0 − 1
Matrices: Rank, Inverse, Normal form 21
Visit: https://www.sites.google.com/site/hub2education/

1 1 2 1 0 0
Operating R 3 → − R 3 + R 2 , we get 0 − 1 − 1  =  0 − 1 0 A .
  
0 0 − 4 − 2 − 1 1

1 1 2   1 0 0

Operating R 2 → −R 2 , we get 0 1 1 = 0   1 0 A .
  
0 0 − 4  − 2 − 1 1 

 
1 1 2  1 0 0 
R
Operating R 3 → 3 , we get 0 1 1 =  0 1 0 A .
−4   1 1 1
0 0 1  − 
2 4 4

 
1 0 1  1 −1 0 
Operating R1 → R1 − R 2 , we get, 0 1 1 =  0 1 0 A .
  1 1 1
0 0 1  − 
2 4 4

1 5 1 
1 0 0  −
2 4 4 
Operating R1 → R1 − R 3 , we get, 0 1 1 =  0 1 0 A .
1 1 1
0 0 1  − 
 2 4 4 

 1 5 1 
 −
1 0 0  2 4 4 
  1 3 1 
Operating R 2 → R 2 − R 3 , we get, 0 1 0 = − A .
 2 4 4 
0 0 1  1 1 1
 2 −
4 4 

 1 5 1 
 2 −
4 4 
 1 3 1 
A −1 = −  . Ans.
 2 4 4 
 1 1
− 
1
 2 4 4

0 1 2
Q.No.2.: Find the inverse of A = 1 2 3 by elementary row operations.
 3 1 1 

Sol.: Writing the given matrix side by side with unit matrix I3, we get
Matrices: Rank, Inverse, Normal form 22
Visit: https://www.sites.google.com/site/hub2education/

0 1 2 : 1 0 0
[A : I 3 ] = 1 2 3 : 0 1 0
3 1 1 : 0 0 1 

1 2 3 : 0 1 0
Operating R12 , we get = 0 1 2 : 1 0 0
3 1 1 : 1 0 1

1 2 3 : 0 1 0

Operating R 3 − 3R 1 , we get = 0 1 2 : 1 0 0
0 − 5 − 8 : 0 − 3 1 

1 0 − 1 : − 2 1 0
Operating R 1 − 2R 2 , R 3 + 5R 2 , we get = 0 1 2 : 1 0 0
0 0 2 : 5 − 3 1 

 
1 1 0 − 1 : − 2 1 0
Operating R 3 , we get = 0 1 2 : 1 0 0
2  5 3 1
0 0 1 : − 
 2 2 2

 1 1 1
1 0 0 : 2 − 2 2
Operating R 1 + R 2 , R 2 − 2R 3 , we get = 0 1 0 : − 4 3
 5 3 1
[
− 1 = I 3 : A −1 ]
0 0 1 : − 
 2 2 2 

1 1 1
 2 −2 − 
2
∴ A −1 = − 4 3 −1  .
5 3 1 
 − 
 2 2 2 

Problems on normal form:

 3 − 3 4
Q.No.1.: If A = 2 − 3 4 find A −1 . Also find two non-singular matrices P and Q
0 − 1 1 

such that PAQ = I, where I is the unit matrix and verify that A−1 = QP .
Matrices: Rank, Inverse, Normal form 23
Visit: https://www.sites.google.com/site/hub2education/

 3 − 3 4
Sol.: Here A = 2 − 3 4 .
0 − 1 1 

Part I: To find A −1 , A11 = 1, A12 = −2, A13 = −2 ,

A 21 = −1, A 22 = 3, A 23 = 3 ,

A31 = 0, A32 = −4, A33 = −3 .

 1 −1 0 
∴ Adj . A =  − 2 3 − 4 .
 − 2 3 − 3

3 −3 4
Now A = 2 − 3 4 = 3(− 3 + 4 ) − 2(− 3 + 4 ) = 3 − 2 = 1 .
0 −1 1

 1 −1 0 
Adj. A 
∴A −1
= = − 2 3 − 4 . Ans.
A
− 2 3 − 3

Part II: Since A= PAQ, where P and Q are two non-singular unit matrices of order 3
each.
3 − 3 4 1 0 0 1 0 0
⇒ 2 − 3 4 = 0 1 0 A 0 1 0 .
0 − 1 1  0 0 1 0 0 1

1 0 0 1 − 1 0 1 0 0
Operating R1 → R1 − R 2 , we get 2 − 3 4 = 0 1 0 A 0 1 0 .
0 − 1 1  0 0 1 0 0 1

1 0 0 1 − 1 0 1 0 0
Operating C2 → C2 + C3 , we get 2 1 4 = 0 1 0 A 0 1 0 .
0 0 1  0 0 1 0 1 1

1 0 0  1 − 1 0 1 0 0
Operating R 2 → R 2 − 2R1 , we get 0 1 4 = − 2 3 0 A 0 1 0 .
0 0 1   0 0 1 0 1 1
Matrices: Rank, Inverse, Normal form 24
Visit: https://www.sites.google.com/site/hub2education/

1 0 0  1 − 1 0 1 0 0 
Operating C3 → C3 − 4C2 , we get 0 1 0 = − 2 3 0 A 0 1 − 4 .
0 0 1  0 0 1 0 1 − 3

 1 − 1 0 1 0 0 
  
⇒ I = PAQ , where P = − 2 3 0 and Q = 0 1 − 4 .

   
 0 0 1 0 1 − 3

Part III: Verification: A−1 = Q P .

1 0 0   1 − 1 0  1 − 1 0 
Now RHS = QP = 0 1 − 4 − 2 3 0 = − 2 3 − 4 .
0 1 − 3  0 0 1  − 2 3 − 3

 1 −1 0 
Now LHS = A −1 =  − 2 3 − 4 .
 − 2 3 − 3

∴ L.H.S. = R.H.S .

Hence A−1 = QP .
Q.No.2.: Reduce the following matrices to the normal form and hence find their ranks.
1 0 2 1
8 1 3 6 0 1 −2 1
(i)  0 3 2 2 , (ii)  .
 1 −1 4 0
 − 8 − 1 − 3 4  
− 2 2 8 0

8 1 3 6

Sol.: (i). Let A =  0 3 2 2 .
 − 8 − 1 − 3 4

1 3 6
Operating C2 → C2 − C1 , C3 → C3 − C1 , C4 → C4 − C1 , we get
8 8 8
 8 0 0 0
A =  0 3 2 2  .
− 8 0 0 10

1 0 0 0 
Operating R1 → R1 , R 3 → R 3 + R1 , we get A = 0 3 2 2  .
1
8
0 0 0 10
Matrices: Rank, Inverse, Normal form 25
Visit: https://www.sites.google.com/site/hub2education/

1 0 0 0
Operating R 3 → R 3 , we get A = 0 3 2 2 .
1
10
0 0 0 1 

1 0 0 0
Operating C4 → C4 − C2 , C3 → C3 − C3 , we get A = 0 3 0 0 .
2 2
3 3
0 0 0 1

1 0 0 0
Operating R 2 → R 2 , we get A = 0 1 0 0 .
1
3
0 0 0 1

1 0 0 0
Operating C3 ↔ C4 , we get A = 0 1 0 0 ,
0 0 1 0

which is the required normal form [I3 O] .

Hence, rank of the matrix A is 3. Ans.
1 0 2 1
0 1 −2 1
(ii). Let A =  .
 1 −1 4 0
 
− 2 2 8 0

1 0 0 0 
0 1 − 2 1 
Operating C4 → C3 − 2C1 , C4 → C4 − C1 , we get A =  .
 1 − 1 2 − 1
 
− 2 2 12 − 2
1 0 0 0
0 1 − 2 1 
Operating R 3 → R 3 − R1 , R 4 → R 4 + 2R1 , we get A =  .
0 − 1 2 − 1 
 
0 2 12 − 2
1 0 0 0 
0 1 0 0 
Operating C3 → C3 + 2C2 , C4 → C4 − C2 , we get A =  .
0 − 1 0 0 
 
0 2 16 − 4
Matrices: Rank, Inverse, Normal form 26
Visit: https://www.sites.google.com/site/hub2education/

1 0 0 0
0 1 0 0 
Operating R 3 → R 3 + R 2 , R 4 → R 4 − 2R 2 , we get A =  .
0 0 0 0
 
0 0 16 − 4

1 0 0 0
0 1 0 0
4
Operating C4 → C4 + C3 , we get A =  .
16 0 0 0 0
 
0 0 16 0

1 0 0 0
0 1 0 0
1
Operating R 4 → R 4 + R 4 , we get A =  .
16 0 0 0 0
 
0 0 1 0

1 0 0 0
0 1 0 0
Operating R 3 ↔ R 4 , we get A =  ,
0 0 1 0
 
0 0 0 0

I O
which is the required normal form  3 .
 O O
Hence, the rank of the matrix A is 3. Ans.
Q.No.3.: Find non-singular matrices P and Q such that PAQ is in the normal form for
the matrices:
1 − 1 − 1 1 2 3 − 2 
(i) 1 1 1  , (ii) 2 − 2 1 3  .
 
3 1 1  3 0 4 1 

1 − 1 − 1
Sol.: (i) Let A = 1 1 1  .
3 1 1 

1 − 1 − 1 1 0 0 1 0 0
Since we know that A = I . A . I ⇒ 1 1 1  = 0 1 0 A 0 1 0 .
3 1 1  0 0 1 0 0 1
Matrices: Rank, Inverse, Normal form 27
Visit: https://www.sites.google.com/site/hub2education/

1 0 0 1 0 0 1 1 1 
Operating C2 → C2 + C1, C3 → C3 + C1 , we get 1 2 2 = 0 1 0 A 0 1 0 .
3 4 4 0 0 1 0 0 1 

Operating R 2 → R 2 − R1, R 3 → R 3 − 3R1 , we get

1 0 0   1 0 0 1 1 1
0 2 2  =  − 1 1 0  A 0 1 0  .
     
0 4 4 − 3 0 1 0 0 1

1 0 0  1 0 0 1 1 1
Operating R 3 → R 3 − 2R 2 , we get 0 2 2 = − 1 1 0 A 0 1 0 .
  
0 0 0 − 1 − 2 1 0 0 1

1 0 0  1 0 0 1 1 1
 1 
, we get 0 1 1 = − 0 A 0 1 0 .
R2 1
Operating R 2 →
2 2 2
0 0 0  − 1 − 2 1 0 0 1

1 0 0  1 0 0 1 1 0 
 1 
Operating C3 → C3 − C2 , we get 0 1 0 = − 0 A 0 1 − 1 .
1
2 2
0 0 0  − 1 − 2 1 0 0 1 

 1 0 0 1 1 0 
I 0  1 1  0 1 − 1 . Ans.
∴ PAQ =  2  , where P = − 0 and Q =  
0 0  − 12 −22 1 0 0 1 
 
Also rank of the matrix A is 2.
1 2 3 − 2 
(ii). Let A = 2 − 2 1 3  .
3 0 4 1 

Since we know that A = I3 A I 4 , where I3 and I 4 are the unit matrix of order 3 and 4
respectively.
1 0 0 0
1 2 3 − 2 1 0 0 
0 1 0 0
⇒ 2 − 2 1 3  = 0 1 0 A  .
0 0 1 0
3 0 4 1  0 0 1  
0 0 0 1

Operating C2 → C2 − 2C1 , C3 → C3 − 3C1 , C4 → C4 + 2C1 , we get

Matrices: Rank, Inverse, Normal form 28
Visit: https://www.sites.google.com/site/hub2education/

1 − 2 − 3 2
1 2 0 0  1 0 0  
 2 − 6 − 5 7  =  0 1 0  A 0 1 0 0
    0 0 .
1 0
1 − 6 − 5 7  0 0 1  
0 0 0 1

1 − 2 − 3 2
1 0 0 0 1 0 0 
0 1 0 0
Operating R 3 → R 3 − R 2 , we get 2 − 6 − 5 7 = 0 1 0 A  .
0 0 1 0
1 0 0 0 0 − 1 1  
0 0 0 1

 1  1 1
Operating C2 →  − C2 , C3 →  − C3 , C4 →  C4 , we get
 6  5 7
 1 3 2
1 3 5 7
 
1 0 0 0 1 0 0 0 − 1 0 0
 2 1 1 1  = 0 1 0  A  6 .
     1 
1 0 0 0 0 − 1 1 0 0 − 0
5
 1
0 0 0 
 7
Operating C 4 → C 4 − C 2 , C3 → C3 − C 2 , we get

 1 4 1
1 3 15
−
21
 1 
1 0 0 0  1 0 0  0 − 1 1

 2 1 0 0  = 0 1 0  A  6 6 6 .
     1 
1 0 0 0 0 − 1 1 0 0 − 0 
5
 1 
0 0 0 
 7 
Operating R 2 → R 2 − 2R 3 , R 3 → R 3 − R1 we get

 1 4 1
1 3 15
−
21
 1 
1 0 0 0 1 0 0  0 − 1 1

0 1 0 0  =  0 3 − 2 A  6 6 6 .
   1 
0 0 0 0 − 1 − 1 1  0 0 − 0 
5
 1 
0 0 0 
 7 
Matrices: Rank, Inverse, Normal form 29
Visit: https://www.sites.google.com/site/hub2education/

 1 4 1
1 3 15
−
21
 1 1 1 
1 0 0  0 − 
I O 
⇒2  = PAQ , where P =  0 3 − 2 and Q =  6 6 6  . Ans.
 O O  1 
 − 1 − 1 1  0 0 − 0 
5
 1 
0 0 0 
 7 
Also rank of the matrix A is 2.

Home Assignments:
Q.No.1.: Find the rank of matrix
1 2 3  1 3 4 5 
2 − 1 0 5     
(i).   (ii). 2 4 7  (iii). 1 2 6 7  .
 0 3 1 4  3 6 10 1 5 0 10

Ans.: (i). 2 (ii). 2 (iii). 3.

Q.No.2.: Find the rank of matrix
1 − 1 2 − 3
1 2 3 2  4 1 0 2 
(i). 2 3 5 1  (ii).  .
0 3 1 4
1 3 4 5  
0 1 0 2
Ans.: (i). 2 (ii). 4 .
 2 0 − 1
Q.No.3.: Use Gauss-Jordan method to find the inverse of the matrix 5 1 0  .
0 1 3 

 3 −1 1 
Ans.: 15 6 − 5 .
 5 − 2 2 

Q.No.4.: Reduce the matrices to normal form and hence find its rank
 2 1 − 3 − 6
A =  3 − 3 1 2  .
1 1 1 2 
Matrices: Rank, Inverse, Normal form 30
Visit: https://www.sites.google.com/site/hub2education/

Ans.: 3.
6 1 3 8
4 2 6 − 1
Q.No.5.: Determine the rank of the matrix A =  by reducing it to the
10 3 9 7
 
16 4 12 15 
normal form.
Ans.: 2.

*** *** *** *** ***

*** *** ***
***