MATRICES

1. ELEMENTARY OPERATIONS OR ELEMENTARY

TRANSFORMATIONS

Elementary Row Operations:

The elementary row operations are as follows:

1. Interchange of any two rows of a matrix. Generally, the interchange of 𝑖-th and 𝑗-th rows is denoted by
the symbol 𝑅𝑖𝑗 .
2. The multiplication of every element of any row by a non-zero constant. Generally the multiplication of
every element of 𝑖-th row by a constant 𝑘 ≠ 0 is denoted by the symbol 𝑅𝑖 (𝑘).
3. The addition to the elements of a row, the product of the corresponding elements of any other row by
any non-zero constant. Generally, the addition to the elements of 𝑖-th row, the product of the
corresponding elements of 𝑗-th row by a constant 𝑘 ≠ 0 is denoted by the symbol 𝑅𝑖𝑗 (𝑘).
Elementary Column Operations. There are three elementary column operations which are similar to above
three elementary row operations. Write column for row in above three operations. The elementary column
operations are respectively denoted by 𝐶𝑖𝑗 , 𝐶𝑖 (𝑘) and 𝐶𝑖𝑗 (𝑘). Equivalent Matrices :

Let A be a matrix. The matrix B obtained from A by applying a finite number of elementary operations (row or
column or both) is called a matrix equivalent to A and is written as A B. Clearly the orders of Aand Bare same.

2. MINOR OF A MATRIX
Let 𝐴 be any matrix (square or rectangular). From this matrix 𝐴, delete all columns and rows leaving a certain 𝑝
columns and 𝑝 rows. Now if 𝑝 > 1, then the elements which have been left, constitute a square matrix of order
𝑝. The determinant of this square matrix is called a minor of 𝐴 of order 𝑝.

3. RANK OF A MATRIX
Let 𝐴 be any matrix. A number 𝑟 is called the rank of the matrix 𝐴 if it obeys the following two properties :
(i) there is at least one minor of 𝐴 of order 𝑟 which does not vanish
(ii) every minor of 𝐴 of order higher than 𝑟 vanish.

Rank of zero matrix is zero.

The rank of unit matrix 𝐼𝑛 of order 𝑛 is 𝑛.
The rank of a non-singular matrix 𝐴 of order 𝑛 is 𝑛 since |𝐴| ≠ 0.
The rank of a matrix 𝐴 is usually denoted by the symbol 𝜌(𝐴).

Nullity of a Matrix :
Definition. Let Abe a square matrix of order 𝑛 and if the rank of 𝐴 is 𝑟, then 𝑛 − 𝑟 is called the nullity of the
matrix Aand is usually denoted by 𝑁(𝐴).

Thus 𝑁(𝐴) = 𝑛 − 𝑟.
 SOLVED PROBLEM
Problem 1. If 𝐴 is any matrix then prove that 𝜌(𝐴′ ) = 𝜌(𝐴).
Solution. We know that the transpose 𝐴′ of 𝐴 is obtained by the mutual exchange of rows and columns. Again
we know that the value of a determinant does not change by the mutual exchange of rows and columns, hence
|𝐴| = |𝐴′ |.

Therefore, it follows that the values of the minors of 𝐴 and 𝐴′ are same. Hence

𝜌(𝐴′ ) = 𝜌(𝐴)

Problem 2. If 𝐴 is 𝑛 × 1 non-zero matrix and 𝐵 is 1 × 𝑛 non-zero matrix, then prove that 𝑟 ′ 𝐴𝐵) = 1.

Solution. Let
𝑎11
𝑎21
𝐴=[ … ] , 𝐵 = [𝑏11 𝑏12 … 𝑏1𝑛 ]1×𝑛
𝑎𝑛1 𝑛×1

Clearly |𝐴𝐵| = 0 and all the minors of order 2 and higher vanish. But 𝐴𝐵 is non-zero-matrix. ∴ 𝜌(𝐴𝐵) = 1.
Problem 3. Prove that the points (𝑥1 , 𝑦1 ), (𝑥2 , 𝑦2 ) and (𝑥3 , 𝑦3 ) are collinear if the rank of the matrix
𝑥1 𝑦1 1
[𝑥2 𝑦2 1] is less than 3 .
𝑥3 𝑦3 1

𝑥1 𝑦1 1
Solution. Points (𝑥1 , 𝑦1 )(𝑥2 , 𝑦2 ) and (𝑥3 , 𝑦3 ) will be collinear if the determinant |𝑥2 𝑦2 1| = 0
𝑥3 𝑦3 1

Hence the rank of the given matrix is less than 3.

Problem 4. If 𝐴 is 𝑛-rowed square matrix of rank (𝑛 − 1), then show that adj. 𝐴 ≠ 0.
𝑎11 𝑎12 … 𝑎1𝑛
𝑎21 𝑎22 … 𝑎2𝑛
Solution. Let 𝐴=[ … … … …]
𝑎𝑛1 𝑎𝑛2 … 𝑎𝑛𝑛

𝐶11 𝐶21 … 𝐶𝑛1

𝐶 𝐶22 … 𝐶𝑛2
adj. 𝐴 = [ 12 ]
… … … …
𝐶1𝑛 𝐶2𝑛 … 𝐶𝑛𝑛

Since the rank of 𝐴 is given to be 𝑛 − 1, hence at least one minor of 𝐴 of order 𝑛 − 1 will not vanish, therefore, it
follows that at least one co-factor of 𝐴 will not be zero i.e., at least one element of adj. 𝐴 is not zero.
∴ adj. 𝐴 ≠ 0.

Problem 5. Find the rank and nullity of the following matrices :

6 1 3 8
1 2 3 2
4 2 6 −1
(i) 𝐴 = [2 3 5 1], (ii) 𝐵 = [ ].
10 3 9 7
1 3 4 5
16 4 12 15

Solution. (i) Here there is no minor of order 4. Now taking the following minors of order 3,
 1 2 3 1 2 0
|2 3 5| 𝐶3 − (𝐶1 + 𝐶2 ) |2 3 0| = 0
1 3 4 1 3 0

1 3 2
|2 5 1| = 1(25 − 4) − 2(15 − 8) + 1(3 − 10) = 0.
1 4 5

Similarly it can be shown that all the remaining minors of order 3 vanish. Hence 𝜌(𝐴) < 3.
1 2
But | | = 3 − 4 = −1 ≠ 0 ∴ 𝜌(𝐴) = 2
2 3

Since 𝐴 is not a square matrix, hence nullity of 𝐴 is not defined.

6 1 3 8 6 1 3 8
4 2 6 −1 4 2 6 −1
(ii) |B| = | |=| |
10 3 9 7 0 0 0 0
16 4 12 15 0 0 0 0

[By 𝑅3 − (𝑅1 + 𝑅2 ) and 𝑅4 − (𝑅1 + 𝑅3 )]

= 0

Similarly it can be shown that all minors of order 3 vanish.

6 1
But | | = 12 − 4 = 8 ≠ 0
4 2

∴ 𝜌(𝐴) = 2.
∴ Nullity 𝑁(𝐴) = 4 − 2 = 2

4. RANK OF A MATRIX [OR RANK OF EQUIVALENT MATRICES]

Elementary transformations do not alter the rank of a matrix i.e., equivalent matrices have equal ranks.

5. ECHELON FORM OF A MATRIX

Definition. A matrix is called in Echelon form if

(i) all the non-zero rows, if any precede the zero rows,
(ii) the number of zeroes preceding the first non-zero element in a row is less than the number of such zeroes in
the next (or succeeding) row,
(iii) the first non-zero element in each row is unity (i.e., 1).

1 5 6 7 0 8
0 1 2 5 −1 2
For Problem. The matrix [ ] is in Echelon form.
0 0 0 1 2 4
0 0 0 0 0 0
Note. If a matrix is in Echelon form, the rank of the matrix is equal to the number of non-zero rows in it. For
Problem, there are three non-zero rows in the matrix of above Problem, hence the rank of above matrix is 3 .

6. NORMAL FORM (OR CANONICAL FORM) OF A MATRIX

Definition. If 𝐴 is a matrix of order 𝑚 × 𝑛 and rank𝑟, then 𝐴 can be reduced by the application of a number of
elementary transformations to any of the following forms :
 𝐼 0 𝐼
(𝑖) [ 𝑟 ] (𝑖𝑖) [ 𝑟 ] (𝑖𝑖𝑖) [𝐼𝑟 0] (𝑖𝑣) [𝐼𝑟 ]
0 0 0

where 𝐼𝑟 is a unit matrix of crder 𝑟 and 𝑂 is zero matrix of any order. These all forms are called the normal (or
canonical) forms of the matrix 𝐴. The order 𝑟 of 𝐼𝑟 is said to be the rank of 𝐴.

SOLVED PROBLEM
12 22 32 42
2
Problem 6. Find the rank of matrix 𝐴, where 𝐴 = [22 32 42 52 ].
3 42 52 62
42 52 62 72
1 4 9 16
4 9 16 25
Solution. 𝐴 = [ ]
9 16 25 36
16 15 36 49
𝑅21 (−1) 1 4 9 16
∼ 3 5 7 9
𝑅32 (−1) [5 7 9 11
]
𝑅43 (−1) 7 9 11 13

𝑅21 (−1) 1 4 9 16 𝑅32 (−1) 1 4 9 16

∼ 2 1 −2 −7 ∼ 2 1 −2 −7
𝑅32 (−1) [2 2 2 2
] [
𝑅42 (−1) 0 1 4 9
]
𝑅43 (−1) 2 2 2 2 0 0 0 0

1 4 9 16 1 4 9 16
0 −7 −20 −39 0 0 8 24
R 21 (−2) [ ] R 23 (7) [ ]
∼ 0 1 4 9 ∼ 0 1 4 9
0 0 0 0 0 0 0 0
1 4 9 16 1 4 9 16
0 1 4 9 1 0 1 4 9
R 32 [ ] R ( )[ ]
∼ 0 0 8 24 ∼3 3 0 0 1 3
0 0 0 0 0 0 0 0

which is in Echelon form and it has three non-zero rows. Hence the rank of 𝐴 is 3 .
Note. Nullity of 𝐴 is given by 𝑁(𝐴) = 4 − 3 = 1.
Problem 7. Find the rank and nullity of the following matrices:
1 2 1 2
1 3 4 3
1 3 2 2
(i) [3 9 12 9], (ii) [ ].
2 4 3 4
1 3 4 1
3 7 4 6

Solution. (i) Here

1 3 4 3 𝑅21 (−3) 1 3 4 3
[3 9 12 9] ~ [0 0 0 0]
1 3 4 1 𝑅31 (1) 0 0 0 −2
1 1 3 4 3
𝑅
1 3 4 3
𝑅3 (− ) [0 0 0 0] 23 [0 0 0 1]
2 ~
0 0 0 1 0 0 0 0

which is in Echelon form and it has two non-zero rows.

Hence the rank of matrix = 2. The nullity of the given matrix is not defined.
(ii)
 1 2 1 2 𝑅21 (+1) 1 2 1 2
1 3 2 2 ∼ 0 1 1 0
𝐴=[ ]
2 4 3 4 𝑅31 (−2) [0 0 1 0
]
3 7 4 6 𝑅41 (−3) 0 1 1 0
1 2 1 2
𝑅42 (−1) 0 1 1 0
[ ]
~ 0 0 1 0
0 0 0 0

which is in Echelon form, it has three non-zero rows.

∴ 𝜌(𝐴) = 3 and 𝑁(𝐴) = 4 − 𝜌(𝐴) = 4 − 3 = 1.

Problem 8. Find the rank of

1 𝑎 𝑏 0 −2 −1 −3 −1
0 𝑐 𝑑 1 1 2 −3 −1
(i) [ ], (ii) [ ].
1 𝑎 𝑏 0 1 0 1 1
0 𝑐 𝑑 1 0 1 1 −1
1 𝑎 𝑏 0
0 𝑐 𝑑 1
Solution. (i) Let 𝐴=[ ]
1 𝑎 𝑏 0
0 𝑐 𝑑 1
1 𝑎 𝑏 0
0 𝑐 𝑑 1
∼[ ] [𝑏𝑦 𝑅3 → 𝑅3 − 𝑅1 and 𝑅4 → 𝑅4 − 𝑅2 ]
0 0 0 0
0 0 0 0

which is triangular form.

∴ 𝜌(𝐴) = Number of non-zero rows

=2

Note. Nullity of 𝐴 = 4 − 𝜌(𝐴) = 4 − 2 = 2.

−2 −1 −3 −1 1 0 1 1
1 2 −3 −1 𝑅13 1 2 −3 −1
(ii) [ ] [ ]
1 0 1 1 ~ −2 −1 −3 −1
0 1 1 −1 0 1 1 −1
1 1
𝑅21 (−1) 1 0 1 1
𝑅2 ( ) 0
0 0 1
~ 0 2 −4 −2 2 1 −2 −1
[ ] ~ [0 −1 −1 ]
0 −1 −1 1 1
𝑅31 (2) 𝑅43 (1) 0
0 0 1 −1 0 0 0
1 0 1 1 𝐶31 (−1) 1 0 0 1
𝑅32 (1)
0 1 −2 −1 ~ 0 1 −2 −1
~ [ ] [ ]
0 0 −3 0 0 0 −3 0
𝑅41 (−1)
0 0 1 0 0 0 0 0

𝑅32 (2) 1 0 0 0 1 0 0 0
~ [0 1 0 0 𝑅3 (− 1) 0 1 0 0 𝐼 0
] 3 [0 ]∼[3 ]
𝑅42 (1) 0 0 −3 0 ~ 0 1 0 0 0
0 0 0 0 0 0 0 0

which is the normal form of the given matrix. Hence the rank of the given matrix = 3.
 1 2 3
Problem 9. Find the rank of the matrix [1 4 2].
2 6 5
1 2 3
Solution. Let 𝐴 = [1 4 2]
2 6 5
1 2 3
∼ [0 2 −1] [By 𝑅2 → 𝑅2 − 𝑅1 and 𝑅3 → 𝑅3 − 2𝑅1 ]
0 2 −1
1 2 3
~ [0 2 −1] [by 𝑅3 → 𝑅3 − 𝑅2 ]
0 0 0

which is in triangular form.

Now 𝜌(𝐴) = number of non-zero rows

=2

Note. Nullity of 𝐴 = 𝑁(𝐴)

= order of matrix 𝐴 − 𝜌(𝐴)

= 3 − 2 = 1.

Problem 10. Find the normal form of the matrix and hence find its rank

8 1 3 6
𝐴=[ 0 3 2 2]
−8 −1 −3 4
8 1 3 6 1 1 1 1 3 3
Solution. Here A = [ 0 3 2 2] 𝐶1 ( ) , 𝐶4 ( ) [ 0 3 2 1]
8 2
−8 −1 −3 4 −1 −1 −3 2

𝐶21 (−1) 1 0 0 0 1 0 0 0
~
𝐶31 (−3) [ 0 3 2 1] 𝑅 (1) [0 3 2 1]
31
𝐶41 (−3) −1 0 0 5 0 0 0 5
~
1 ∼
𝐶2 ( ) 1 0 0 0 1 0 0 0
3 [0 1 1 1] 32 (−1) [0 1 0
𝐶 0]
1 0 5 𝐶42 (−1) 0 0 0
𝐶3 ( ) 0 0 5
2
~
~ 1 0 0 0 1 0 0 0
𝐶34 [0 1 0 0] 𝐶 ( 1 ) [ 0 1 0 0]
3
0 0 5 0 5 0 0 1 0

∼ [𝐼3 , 0]. Hence 𝜌(𝐴) = 3

Problem 11. Find the rank of the matrices 𝐴, 𝐵, 𝐴 + 𝐵, 𝐴𝐵 and 𝐵𝐴

1 1 −1 −1 −2 −1
where 𝐴 = [2 −3 4] , 𝐵 = [ 6 12 6]
3 −2 3 5 10 5
 1 1 −1 ~ 1 0 0
Solution. (i) Here 𝐀 = [2 −3 4] 21 (−1) [2 −5 6]
𝐶
3 −2 3 𝐶31 (1) 3 −5 6

~ 1 0 0 𝐶2 (−1/5) 1 0 0
𝑅21 (−2) [0 −5 6] 1 [ 0 1 1]
𝑅31 (−3) 0 −5 6 𝐶3 (6) 0 1 1

~ 1 0 0 ~ 1 0 0
𝐶32 (−1) [0 1 0] 𝑅 (−1) [0 1
32
0]
0 1 0 0 0 0

∴ 𝜌(𝐀) = 2.

−1 −2 −1 ∼ −1 −2 −1
(ii) 𝐁 = [ 6 12 6] 𝑅2 (1/6) [ 1 2 1]
5 10 5 𝑅3 (1/5) 1 2 1
~ −1 −1 −1 1 1 1
1 [ 1 ~
𝐶2 ( ) 1 1 ] 𝑅1 (−1) 1 1 1]
[
2 1 1 1 1 1 1
∼ 1 0 0 ∼ 1 0 0
𝐶21 (−1) [1 0 0] 21 (−1) [0 0
𝑅 0]
𝐶31 (−1) 1 0 0 𝑅31 (−1) 0 0 0

∴ 𝜌(𝐁) = 1.

(iii)

1 1 −1 −1 −2 −1
𝐀 + 𝐁 = [2 −3 4] + [ 6 12 6]
3 −2 3 5 10 5
0 −1 −2 ~ 0 −1 −2
= [8 9 10] 𝑅 (1/8) [8 9 10]
3
8 8 8 1 1 1
~
~ 1 1 1 1 0 0
𝑅13 [ 8 9 10] 21 (−1) [8
𝐶 1 2]
0 −1 −2 𝐶31 (−1) 0 −1 −2
~ 1 0 0 ∼ 1 0 0
𝑅21 (−8) 0[ 1 2] 𝐶 (−2) [0 1 0]
32
0 −1 −2 0 −1 0
1 0 0
∼ [0 1 0]
𝑅32 (1)
0 0 0

∴ 𝝆(𝐀 + 𝐁) = 2

1 1 −1 −1 −2 −1 0 0 0
𝐀𝐁 = [2 −3 4] × [ 6 12 6] = [0 0 0]
3 −2 3 5 10 5 0 0 0

∴ 𝜌(𝐀𝐁) = 0.

−1 −2 −1 1 1 −1 −8 7 −10
(iv) 𝐁𝐀 = [ 6 12 6] × [2 −3 4] = [ 48 −42 60]
5 10 5 3 −2 3 40 −35 50
 ∼
𝐶1 (−1/8) 1 1 1 ∼ 1 1 1 ∼ 1 1 1
𝐶2 (1/7) [−6 −6 −6] 𝑅2 (−1/6) [1 1 1] 𝑅21 (−1) [0 0 0]
1 −5 −5 −5 𝑅3 (−1/5) 1 1 1 𝑅31 (−1) 0 0 0
𝐶3 (− )
10

∴ 𝜌(𝐵𝐴) = 1.

Problem 12. Find the normal form of the following matrices :

1 2 1 0
2 3 4 5
3 2 1 2
3 4 5 6
(i) 2 −1 2 5, (ii) [ ].
4 5 6 7
5 6 3 2
[1 9 10 11 12
3 −1 −3]

Also find their ranks.

1 2 1 0 1 1 2 0
3 2 1 2 ∼ 3 1 2 2
Solution. (i) 𝐀 = 1 −1 2 5 𝐶 2 2 −1 5
23
5 6 3 2 5 3 6 2
[1 3 −1 −3] [1 −1 3 −3]

1 0 0 0 ~ 1 0 0 0
∼ 3 −2 −4 2
𝐶21 (−1) 2 𝑅21 (−3) 0 −2 −4 2
0 −5 5 𝑅31 (−2) 0 0 −5 5
𝐶31 (−2) 5 −2 −4 2 0 −2 −4 2
[1 𝑅41 (−5) [
−2 1 −3] 0 −2 1 −3]
𝑅51 (−1)
1 0 0 0 1 0 0 0
~ 0 −2 −4 2 0 −2 −4 2
∼
𝑅42 (−1) 0 0 −5 5 𝑅53 (1) 0 0 −5 5
𝑅52 (−1) 0 0 0 0 0 0 0 0
[0 0 5 −5] [0 0 0 0]
~ 1 0 0 0
1 1 0 0 0 ~ 0 1 0 0
𝑅2 (− ) 0 1 2 −1 𝐶32 (−2)
2 [ ] 0 0 1 −1
1 0 0 1 −1 𝐶 (1)
42 0 0 0 0
𝑅3 (− ) 0 0 0 0
5 [0 0 0 0]
1 0 0 0
~ 0 1 0 0 𝐈 𝟎
𝐶41 (1) [0 ]∼[ 𝟑 ]
0 1 0 𝟎 𝟎
0 0 0 0

∴ 𝜌(𝐀) = 3

(ii) Let 𝐴 denote the given matrix, then

2 3 4 5
1 1 1 1
𝐴−[ ]
0 −1 −2 −3
1 −2 −5 −8

[by 𝑅2 → 𝑅2 − 𝑅1 , 𝑅3 → 𝑅3 − 2𝑅1 , and 𝑅4 → 𝑅4 − 4𝑅1 ]

 1 2 3 4
1 1 1 1
~[ ]
0 −1 −2 −3
0 −3 −6 −9

[𝑏𝑦 𝑅1 → 𝑅1 − 𝑅2 , 𝑅4 → 𝑅4 − 𝑅2 ]

1 2 3 4
0 −1 −2 −3
∼[ ] [𝑏𝑦 𝑅2 → 𝑅2 − 𝑅1 , 𝑅4 → 𝑅4 − 3𝑅3 ]
0 −1 −2 −3
0 0 0 0
1 2 3 4
0 −1 −2 −3
~[ ] [𝑏𝑦 𝑅3 → 𝑅3 − 𝑅2 ]
0 0 0 0
0 0 0 0
1 0 0 0
0 −1 −2 −3
∼[ ]
0 0 0 0
0 0 0 0

[by 𝐶2 → 𝐶2 − 2𝐶1 , 𝐶3 → 𝐶3 − 3𝐶1 , 𝐶4 → 𝐶4 − 4𝐶1 ]

1 0 0 0
0 −1 0 0
∼[ ]
0 0 0 0
0 0 0 0

[by 𝐶3 → 𝐶3 − 2𝐶2 , 𝐶4 → 𝐶4 − 3𝐶2 ]

1 0 0 0
0 1 0 0
∼[ ] [𝑏𝑦 𝐶2 → (−1)𝐶2 ]
0 0 0 0
0 0 0 0
𝐼2 ⋮ 0
∼ [⋯ ⋯⋯ ] which is in normal form.
𝑂 ⋮ 𝑂

∴ 𝜌(𝐀) = 2

0 1 −3 −1
1 0 1 1
Problem 13. Find the rank of the matrix 𝐴, where 𝐴 = [ ].
3 1 0 2
1 1 −2 0
1 0 1 1 𝑅 (−3) 1 0 1 1
𝑅 0 1 −3 −1 31 0 1 −3 −1
Solution. Here 12 [ ] ~ [ ],
𝐴~ 3 1 0 2
𝑅41 (−1)
0 1 −3 −1
1 1 −2 0 0 1 −3 −1

𝐶31 (−1) 1 0 0 0 𝑅 (−3) 1 0 0 0

~ 0 1 −3 −1 32~ 0 1 −3 −1
[ ] [ ]
𝐶41 (−1) 0 1 −3 −1 𝑅 (−1) 0 0 0 0
0 1 −3 −1 42 0 0 0 0
 𝐶32 (3) 1 0 0 0
𝐼 0
~ [0 1 0 0
]~[ 2 ]
𝐶42 (1) 0 0 0 0 0 0
0 0 0 0
which is the normal form of the matrix 𝐴.

∴ 𝜌(𝐴) = 2

Problem 14. Find non-singular matrices P and 𝑄 such that PAQ is in the normal form where

1 −1 2 −1
𝐴 = [4 2 −1 2]
2 2 −1 0

Also find the rank of the matrix 𝐴.

Solution. Here to evaluate 𝑃 we shall apply row transformations on 𝐼3 ( ∵ 𝐴 has three rows) and to evaluate 𝑄,
we shall apply column transformations on 𝐼4 ( ∵ 𝐴 has four columns). Hence to evaluate 𝑃𝐴𝑄, we write as
follows

1 0 0 0
1 −1 2 −1 1 0 0
0 1 0 0
𝐴 = 𝐼3 𝐴𝐼4 , i.e., [4 2 −1 2] = [ 0 1 0] 𝐴 [ ]
0 0 1 0
2 2 −2 0 0 0 1
0 0 0 1

Applying 𝑅21 (−4) and 𝑅31 (−2), we get

1 0 0 0
1 −1 2 −1 1 0 0
0 1 0 0
[0 6 −9 6] = [−4 1 0] 𝐴 [ ].
0 0 1 0
0 4 −6 2 −2 0 1
0 0 0 1

Now by 𝐶21 (1), 𝐶31 (−2), 𝐶41 (1),

1 1 −2 1
1 0 0 0 1 0 0
0 1 0 0
[0 6 −9 6] = [−4 1 0] 𝐴 [ ].
0 0 1 0
0 4 −6 2 −2 0 1
0 0 0 1
1 1
Now by 𝑅2 ( ) and 𝑅3 ( )
3 2

1 0 0 1 1 −2 1
1 0 0 0
[0 2 −3 2] = [−4/3 1/3 0 ] 𝐴 [0 1 0 0
]
0 0 1 0
0 2 −3 1 −1 0 1/2
0 0 0 1
1 0 0 0 1 0 0 1 1 −2 1
Now by 𝑅32 (−1), [0 2 −3 2] = [−4/3 1/3 0] [0 1 0 0
].
0 0 1 0
0 0 0 −1 1/3 −1/3 1/2
0 0 0 1
1 1 1
Now by 𝐶2 ( ) , 𝐶3 (− ) , 𝐶4 ( )
2 3 2
 1 2 1
1
1 0 0 2 3 2
1 0 0 0 4 1 1
0 1 1 1 − 0 0 2
0 0
[ 1] = 3 3 1
.
0 0 0 − 1 1 1 0 0 − 0
2 [ 3 − 3
3 2] 1
[0 0 0
2]

Now by 𝐶32 (−1), 𝐶42 (−1) and 𝑅3 (−2)

1 1
1 0
2 6
1 1 1
1 0 0 0 1 0 0 0 − −
[0 1 0 0] = [−4/3 1/3 0] 2 2 2
1
0 0 0 1 −2/3 2/3 −1 0 0 − 0
3
1
[0 0 0
2]
1 1
1 0
2 6
1 1 1
1 0 0 0 1 0 0 0 − −
Now by 𝐶34 , [0 1 0 0] = [−4/3 1/3 0 ] 𝐴 2 2 2
1
0 0 1 0 −2/3 2/3 −1 0 0 0 −
3
1
[0 0 2
0 ]
or 𝑁 = 𝑃𝐴𝑄, where 𝑁 is the normal form of the matrix 𝐴

Hence

1 1
1 0
2 6
1 1 1
1 0 0 0 − −
𝑃 = [−4/3 1/3 0] and 𝑄 = 2 2 2
1
−2/3 2/3 −1 0 0 0 −
3
1
[0 0 2
0]

Since 𝐴 ∼ [𝐼3 𝑂], so 𝜌(𝐴) = 3.

Note. 𝑃 and 𝑄 are not unique matrices.

1 1 1
Problem 15. Find the rank of the following matrix 𝐴 = [𝑏 + 𝑐 𝑐+𝑎 𝑎 + 𝑏].
𝑏𝑐 𝑐𝑎 𝑎𝑏
1 0 0
Solution. Here |𝐀| = | 𝑏 + 𝑐 𝑎−𝑏 𝑎−𝑐 | [By𝐶2 − 𝐶1 , 𝐶3 − 𝐶1 ]
𝑏𝑐 𝑐(𝑎 − 𝑏) 𝑏(𝑎 − 𝑐)

𝑎−𝑏 𝑎−𝑐
=| |
𝑐(𝑎 − 𝑏) 𝑏(𝑎 − 𝑐)
= 𝑏(𝑎 − 𝑏)(𝑎 − 𝑐) − 𝑐(𝑎 − 𝑏)(𝑎 − 𝑐) = −(𝑎 − 𝑏)(𝑏 − 𝑐)(𝑐 − 𝑎).
Now following three cases arise :
Case I. If 𝑎 = 𝑏 = 𝑐, then |𝐀| = 0.
1 1 1
∴ 𝐀 = [2𝑎 2𝑎 2𝑎].
𝑎2 𝑎2 𝑎2
Clearly an the minors of the 2nd order vanish.

∴ 𝜌(𝐀) < 2

But 𝐀 is non-zero matrix.

𝜌(𝐴) = 1.

Case II. If 𝑎 = 𝑏 ≠ 𝑐.

1 1 1
Then |𝐀| = 0 and 𝐀 = [𝑎 + 𝑐 𝑎+𝑐 2𝑎]
𝑎𝑐 𝑎𝑐 𝑎2
1 1
Now minor of order 2 = | | = 𝑎 − 𝑐 ≠ 0.
𝑎+𝑐 2𝑎

∴ 𝜌(𝐀) = 2

Case III. If 𝑎 ≠ 𝑏 ≠ 𝑐.
Then |𝐀| ≠ 0 ∴ 𝜌(𝐀) = 3.

Problem 16. Find the normal form of the matrix A and hence find its rank
2 3 −1 −1
1 −1 −2 −4
where 𝐴=[ ]
3 1 3 −2
6 3 0 −7

Solution. Here

2 3 −1 −1 1 −1 −2 −4
1 −1 −2 −4 𝑅12 2 3 −1 −1
𝐴=[ ] [ ]
3 1 3 −2 ~ 3 1 3 −2
6 3 0 −7 6 3 0 −7
𝑅21 (−2) 1 −1 −2 −4 𝐶21 (1) 1 0 0 0
~ 0 5 3 7 ~ 0 5 3 7
𝑅31 (−3) [0 4 9 10
] 𝐶31 (2) [0 4 9 10
]
𝑅41 (−6) 0 9 12 17 𝐶41 (1) 0 9 12 17

1 0 0 0 1 0 0 0
𝑅43 (−1) 0 5 3 7 𝑅42 (−1) 0 5 3 7
[ ] [ ]
~ 0 4 9 10 ~ 0 4 9 10
0 5 3 7 0 0 0 0
1 0 0 0 1 0 0 0
𝑅23 (−1) 0 1 −6 −3 𝑅32 (−4) 0 1 −6 −3
[ ] [ ]
~ 0 4 9 10 ~ 0 0 33 22
0 0 0 0 0 0 0 0

𝐶32 (6) 1 0 0 0 1 0 0 0
~ [0 1 0 0 𝐶3 (1/33) 0
] [
1 0 0
]
𝐶42 (3) 0 0 33 22 ~ 0 0 1 22
0 0 0 0 0 0 0 0
 1 0 0 0
𝐶43 (−22) 0 1 0 0 𝐼 0
[ ]~[ 3 ]
~ 0 0 1 0 0 0
0 0 0 1

∴ 𝜌(𝐴) = 3.

7. (A) SOLUTION OF SIMULTANEOUS LINEAR EQUATIONS BY

ELEMENTARY TRANSFORMATION AND CONSISTENCY OF EQUATIONS
A system of 𝑛 simultaneous equations in 𝑛 unknowns 𝑥1 , 𝑥2 , … , 𝑥𝑛 is given as :

𝑎11 𝑥1 + 𝑎12 𝑥2 + ⋯ + 𝑎1𝑛 𝑥𝑛 = 𝑏1

𝑎21 𝑥1 + 𝑎22 𝑥2 + ⋯ + 𝑎2𝑛 𝑥𝑛 = 𝑏2
……………………………………… (1)
………………………………………
𝑎𝑚1 𝑥1 + 𝑎𝑚2 𝑥2 + ⋯ + 𝑎𝑚𝑛 𝑥𝑛 = 𝑏𝑚 }

The above system (1) of 𝑚 equations can be represented by the following matrix equation

𝐀𝐗 = 𝐁

where

𝑎11 𝑎12 … 𝑎1𝑛 𝑥1 𝑏1

𝑎21 𝑎22 … 𝑎2𝑛 𝑥2 𝑏2
𝐀 = [𝑎𝑖𝑗 ] = … … … … ,𝐗 = … ,𝐁 = …
… … … … … …
[𝑎𝑚1 𝑎𝑚2 … 𝑎𝑚𝑛 ]𝑚×𝑛 [𝑥𝑛 ] [𝑏𝑚 ]

The matrix 𝐀 is called the 'Coefficient matrix', 𝐗 is called the 'Column matrix' of 𝑛 unknowns (or variables) and
B is called the 'Column matrix' of 𝑛.

Augmented Matrix. The matrix obtained by introducing the column matrix B of constants to the right of the
coefficient matrix 𝐀, is called the Augmented matrix for the system 𝐀𝐗 = 𝐁. The augmented matrix is denoted
by [𝐀 ⋮ 𝐁]. Thus

𝑎11 𝑎12 … 𝑎 ⋮ 𝑏1
… 𝑎1𝑛
[𝐀 ⋮ 𝐁] = 21 𝑎22
𝑎
… …
2𝑛 ⋮ 𝑏2
… … ⋮ ⋮
… … … … ⋮ ⋮
[ 𝑛1 𝑎𝑛2
𝑎 ... 𝑎 ⋮ 𝑏𝑛 ]𝑛×(𝑛+1)
𝑛𝑛

Consistent and Inconsistent Equations. The set of those values of 𝑥1 , 𝑥2 , … , 𝑥𝑛 which satisfy the system of
equations 𝐀𝐗 = 𝐁, is called the solution set of the system 𝐀𝐗 = 𝐁.

7. (B) SUMMARY (OR WORKING METHOD TO FIND THE SOLUTION QF

SIMULTANEOUS EQUATIONS)
First of all write the augmented matrix [A ⋮ 𝐁]. Now reduce [A ：𝐁] to the Echelon form by the application of
elementary row operations only. It gives the rank of [A 𝐁 ]. Now obtain 𝐴 after deleting the last column from
the Echelon form of [𝐀 ⋮ 𝐁] and it gives the rank of 𝐴. Now following two cases arise.
Case I. Rank of 𝐀 ≠ Rank of [𝐀 ⋮ 𝐁].
In this position the equations are called inconsistent i.e., they have no solution.
Case II. Rank of A = Rank of [A ⋮ B] = 𝑟 (say).
In this position the equations are called consistent i.e., they have solution.
Now following three sub-cases arise :
(i) If 𝐫 = 𝐧, then 𝑛 − 𝑟 = 0; i.e., rank of 𝐀 is equal to the number of unknowns. In this case, the equations have a
unique solution.
(ii) If 𝐫 < 𝐧, then 𝑛 − 𝑟 > 0, i.e., rank of 𝐴 is less than the number of unknowns. In this case, the equations have
infinite solutions and arbitrary values will be given to 𝑛 − 𝑟 unknowns.
(iii) If 𝐦 < 𝐧, then clearly 𝑟 ≤ 𝑚 < 𝑛. In this case also, the equations have infinite solutions and arbitrary
values will be given to 𝑛 − 𝑟 unknowns.

Remark. The above theorem has been proved for a general case. In case of simultaneous equations put 𝑚 = 𝑛
and proceed as in above theorem.

[Link] Simultaneous Equations :

The 𝑛 equations given by 𝐀𝐗 = 𝐁 in 𝑛 unknowns are called homogeneous simultaneous equations if 𝐁 = 𝟎 i.e.,
if 𝑏1 = 𝑏2 = ⋯ = 𝑏𝑛 = 0. Thus above equations can be written as 𝐀𝐗 = 𝟎. Hence

Rank of the coefficient matrix 𝐀 is always equal to the rank of the augmented matrix [𝐀 ⋮ 𝐁 ] and thus the
homogeneous simultaneous equations are always consistent.

Following cases are possible :

(i) If 𝜌(𝐀) = 𝑟 = 𝑛 i.e., 𝑛 − 𝑟 = 0, then the equations 𝐀𝐗 = 𝟎 have a single solution given by 𝑥1 = 𝑥2 = ⋯ = 𝑥𝑛 =
0. This solution is called a trivial solution.
(ii) If 𝜌(A) = 𝑟 < 𝑛 i.e., 𝑛 − 𝑟 > 0, then arbitrary values are given to 𝑛 − 𝑟 unknowns. Thus the equations 𝐀𝐗 =
𝟎 have infinite many non-trivial solutions.

SOLVED PROBLEM
Problem 17. Examine the following equations for consistency and if consistent, find the complete solution 𝑥 +
2𝑦 − 𝑧 = 3,3𝑥 − 𝑦 + 2𝑧 = 1,2𝑥 − 2𝑦 + 3𝑧 = 2, 𝑥 − 𝑦 + 𝑧 = −1.
Solution. The matrix equation 𝐀𝐗 = 𝐁 is given by

1 2 −1 𝑥
3
2 −2 3 𝑦
[ ] [ ] = [ 2] (1)
3 −1 2 𝑧
−1
1 −1 −1
1 2 −1 ⋮ 3
[𝐀 ⋮ 𝐁] = [ 2 −2 3 ⋮ 2
]
3 −1 2 ⋮ 1
1 −1 1 ⋮ −1
1 2 −1 ⋮ 3
0 −6 5 ⋮ −4
∼[ ]
0 −7 5 ⋮ −8
0 −3 2 ⋮ −4

[by 𝑅2 → 𝑅2 − 2𝑅1 , 𝑅3 → 𝑅3 − 3𝑅1 , 𝑅4 → 𝑅4 − 𝑅1 ]

1 2 −1 ⋮ 3
0 −6 5 ⋮ −4
∼[ ] [by 𝑅3 → 6𝑅3 − 7𝑅2 , 𝑅4 → 2𝑅4 − 𝑅2 ]
0 0 −5 ⋮ −20
0 0 −1 ⋮ −4
 1 2 −1 ⋮ 3
0 −6 5 ⋮ −4
∼[ ] [𝑏𝑦 𝑅4 → 5R 4 − 𝑅3 ]
0 0 −5 ⋮ −20
0 0 0 ⋮ 0

which is in triangular form.

Clearly 𝜌(𝐀 ⋮ 𝐁) = 𝜌(𝐀) = 3 = number of unknowns.
Hence given system is consistent and has a unique solution.
Now (1) reduces to

1 2 −1 𝑥
3
0 −6 5 𝑦
[ ] [ ] = [ 4]
0 0 −5 𝑧
−20
0 0 0
𝑥 + 2𝑦 − 𝑧 =3
⇒ −6𝑦 + 5𝑧 = −4 } ⇒ 𝑥 = −1, 𝑦 = 4, 𝑧 = 4
−5𝑧 = −20

Hence the solution is

𝑥 = −1, 𝑦 = 4, 𝑧 = 4

Problem 18. Solve 𝑥 + 2𝑦 + 3𝑧 = 14,3𝑥 + 𝑦 + 2𝑧 = 11,2𝑥 + 3𝑦 + 𝑧 = 11.

Solution. The matrix equation 𝐴𝑋 = 𝐵 is given by

1 2 3 𝑥 14
[3 1 2] [𝑦] = [11] (1)
2 3 1 𝑧 11
1 2 3 ⋮ 14
[𝐴 ⋮ B] = [3 1 2 ⋮ 11]
2 3 1 ⋮ 11
𝑅21 (−3) 1 2 3 ⋮ 14
∼ [0 −5 −7 ⋮ −31]
𝑅31 (−2) 0 −1 −5 ⋮ −17
𝑅2 (−1) 1 2 3 ⋮ 14
~ [0 5 7 ⋮ 17]
𝑅3 (−1) 0 1 5 ⋮ 31

1 2 3 ⋮ 14
𝑅32 (−5)
[0 1 5 ⋮ 17 ]
~
0 0 −18 ⋮ −54
1 2 3 : 14
𝑅3 (−1/18) [0 1 5 : 17]
0 0 1 : 3

which is in Echelon form.

Clearly 𝜌[𝐀 ⋮ 𝐁] = 𝜌(𝐀) = 3 = Number of unknowns.
The given equations are consistent and have a unique solution. Hence equation (1) can be
written as follows:

1 2 3 𝑥 14
[0 1 5 ] [𝑦 ] = [ 17] (1)
0 0 1 𝑧 3

∴ 𝑥 + 2𝑦 + 3𝑧 = 14 (2)
 𝑦 + 5𝑧 = 17 (3)

𝑧=3 (4)

Solving we get, 𝑥 = 1, 𝑦 = 2, 𝑧 = 3.
Problem 19. Test the consistency and solve

5𝑥 + 3𝑦 + 7𝑧 = 4,3𝑥 + 26𝑦 + 2𝑧 = 9,7𝑥 + 2𝑦 + 10𝑧 = 5

Solution. The matrix equation 𝐀𝐗 = 𝐁 is given by

5 3 7 𝑥 4
[3 26 2] [𝑦] = [9] (1)
7 2 10 𝑧 5
5 3 7 ⋮ 4 1 49 −3 ⋮ 14
∴ [𝐀 ⋮ 𝐁] = [3 26 2 ⋮ 9] ∼ [3 26 2 ⋮ 9] [ by 𝑅1 → −𝑅1 + 2𝑅2 ]
7 2 10 ⋮ 5 7 2 10 ⋮ 5
1 49 −3 ⋮ 14
∼ [3 26 2 ⋮ 9] [ by 𝑅3 → 𝑅3 − 2𝑅2 ]
7 −50 6 ⋮ −13

1 49 −3 ⋮ 14
∼ [0 −121 11 ⋮ −33] [ by 𝑅2 → 𝑅2 − 3𝑅1 , 𝑅3 → 𝑅3 − 𝑅1 ]
0 −99 9 ⋮ −27
1 49 −3 ⋮ 14 1 1
∼ [0 11 −1 ⋮ −3] [ by 𝑅2 → (− ) 𝑅2 , 𝑅3 → ( ) 𝑅3 ]
11 9
0 −11 1 ⋮ −3
1 49 −3 ⋮ 14
∼ [0 11 −1 ⋮ 3] [by 𝑅3 → 𝑅3 + 𝑅2 ]
0 0 0 ⋮ 0

Clearly 𝜌(𝐀 ⋮ 𝐁) = 𝜌(𝐀) = 2 < 3 i.e., number of unknowns.

Therefore, the given system is consistent and has infinite number of solutions.
Thus (1) reduces to

𝑥 + 49𝑦 − 3𝑧 = 14 (2)

11𝑦 − 𝑧 = 3 (3)

3 𝑘
Let z = k, then from (3), 𝑦= +
11 11

Putting values of 𝑦 and 𝑧 in (2), we get

3+𝑘 7 − 16𝑘
𝑥 = 14 + 3𝑘 − 49 ( )=
11 11

Hence the required infinite number of solutions are :

7 − 16𝑘 3+𝑘
𝑥= ,𝑦 = ,𝑧 = 𝑘
11 11

where 𝑘 has infinite number of values.

Problem 20. Prove that the following equations are consistent and solve
 2𝑥 + 4𝑦 − 𝑧 = 9,3𝑥 − 𝑦 + 5𝑧 = 5,8𝑥 + 2𝑦 + 9𝑧 = 19

Solution. The matrix equation 𝐴𝑋 = 𝐵 is given by

2 4 −1 𝑥 9
[3 −1 5] [𝑦] = [ 5] (1)
8 2 9 𝑧 19
2 4 −1 ⋮ 9
∴ [𝐀 ⋮ 𝐁] = [3 −1 5 ⋮ 5]
8 2 −1 ⋮ 19
𝑅12 1 4 −1 ⋮ 9
~ [2 −5 6 ⋮ −4]
𝑅31 (−4) 0 −14 13 ⋮ −17

1 −5 6 ⋮ −4
𝑅12
[2 0 −1 ⋮ 9]
∼
0 −14 13 ⋮ −17
1 −5 6 ⋮ −4
𝑅21 (−2)
[0 14 −13 ⋮ 17]
∼
0 −14 13 ⋮ −17
1 −5 6 ⋮ −4
𝑅32 (1)
[0 14 −13 ⋮ 17]
∼
0 0 0 ⋮ 0
1 −5 6 ⋮ −4
𝑅2 (1/14)
[0 1 −13/14 ⋮ 17/14]
∼
0 0 0 ⋮ 0

which is in Echelon form.

Clearly 𝜌[𝐀 ⋮ 𝐁] = 𝜌(𝐀) = 2
∴ The given equations are consistent. Since the rank of 𝐴 is less than the number of unknowns therefore the
given equations have infinite many solutions. The equations (1) become :

1 −5 6 𝑥 −4
[0 1 −13/14] [𝑦] = [17/14]
0 0 0 𝑧 0
𝑥 − 5𝑦 + 6𝑧 = −4, 𝑦 − (13/14)𝑧 = 17/14

Taking 𝑧 = 𝑘 (arbitrary value) then

𝑥 = (−19𝑘 + 29)/14, 𝑦 = (13𝑘 + 17)/14, 𝑧 = 𝑘

Ans.
Problem 21. Prove that the following equations are consistent and solve

𝑥1 − 𝑥2 + 𝑥3 = 2,3𝑥1 − 𝑥2 + 2𝑥3 = −6,3𝑥1 + 𝑥2 + 𝑥3 = −18

Solution. The matrix equation 𝐀𝐗 = 𝐁 is given by

1 −1 1 𝑥1 2
[3 −1 2] [𝑥2 ] = [ −6] (1)
3 1 1 𝑥3 −18
 1 −1 1 ⋮ 2
∴ [𝐀 ⋮ 𝐁] = [3 −1 2 ⋮ −6]
3 1 1 ⋮ −18
𝑅21 (−3) 1 −1 1 ⋮
~ [0 2 −1 −12]
𝑅31 (−3) 0 4 −2 ⋮ −24
1 −1 1 ⋮ 2
𝑅32 (−2)
[0 2 −1 ⋮ −12]
~
0 0 0 ⋮ 0

1 1 −1 1 ⋮ 2
𝑅2 ( ) [0 1 −1/2 ⋮ −6]
2
0 0 0 ⋮ 0

which is in Echelon form.

Clearly 𝜌[𝐀 ⋮ 𝐁] = 𝜌[𝐀] = 2

The given equations are consistent. Since the rank of 𝐴, is less than the number of unknowns, therefore the
given equations will have infinite many solutions. The equation (1) becomes

1 −1 1 𝑥1 2
[0 1 −1/2] [𝑥2 ] = [6]
0 0 0 𝑥3 0
1
∴ 𝑥1 − 𝑥2 + 𝑥3 = 2, 𝑥2 − 𝑥3 = −6
2
1 1
⇒ 𝑥1 = − 𝑘 − 4, 𝑥2 = 𝑘 − 6, 𝑥3 = 𝑘( say ).
2 2

Ans.

Problem 22. Show that the following equations are consistent:

5𝑥 + 3𝑦 + 14𝑧 = 4, 𝑦 + 2𝑧 = 1, 𝑥 − 𝑦 + 2𝑧 = 0,2𝑥 + 𝑦 + 6𝑧 = 2

Solution. The matrix equations 𝐀𝐗 = 𝐁 is given by

5 3 14 𝑥 4
0 1 2 𝑦 1
[ ][ ] = [ ]
1 −1 2 𝑧 0
2 1 6 2
5 3 14 ⋮ 4 1 −1 2 ⋮ 0
0 1 2 ⋮ 1 𝑅13 0 1 2 ⋮ 1
∴ [𝐀 ⋮ 𝐁] = [ ] [ ]
1 −1 2 ⋮ 0 ~ 5 3 14 ⋮ 4
2 1 6 ⋮ 2 2 1 6 ⋮ 2

𝑅31 (−5) 1 −1 2 ⋮ 0
~ 0 1 2 ⋮ 1
[ ]
𝑅41 (−2) 0 8 4 ⋮ 4
0 3 2 ⋮ 2
1 −1 2 ⋮ 0
𝑅32 (−8) 0 1 2 ⋮ 1
~ [ ]
0 0 −12 ⋮ −4
𝑅42 (−3)
0 0 −4 ⋮ −1
 1 −1 2 ⋮ 0
𝑅34 (−3) 0 1 2 ⋮ 1
~ [ ]
0 0 0 ⋮ −1
0 0 −4 ⋮ −1
1 1 −1 2 ⋮ 0
𝑅3 (− )0 1 2 ⋮ 1
4 ]
∼0 0 1 ⋮ 1/4
𝑅4 (−1)0 0 0 ⋮ 1

which is in Echelon form.

∴ 𝜌[𝐴 ⋮ 𝐵] = 4

Now deleting the last column from the Echelon form of [A 𝐁 ], we have
1 −1 2
0 1 2
Echelon form of 𝐴 = [ ]
0 0 1
0 0 0

∵ 𝜌(𝐴) = 3

∴ 𝜌(𝐀) ≠ 𝜌[𝐀 ⋮ 𝐁].

Hence the given equations are inconsistent.

Problem 23. Solve 2𝑥 − 𝑦 + 3𝑧 = 3, 𝑥 + 2𝑦 − 𝑧 − 5𝑤 = 4, 𝑥 + 3𝑦 − 2𝑧 − 7𝑤 = 5.

2 −1 3 0 ⋮ 3
Solution. [𝐀 ⋮ 𝐁] = [1 2 −1 −5 ⋮ 4]
1 3 −2 −7 ⋮ 5
1 2 −1 −5 ⋮ 4
𝑅12
[2 −1 3 0 ⋮ 3]
~
1 3 −2 −7 ⋮ 5
𝑅21 (−2) 1 2 −1 −5 ⋮ 4
~ [0 −5 5 10 ⋮ −5]
𝑅31 (−1) 0 1 −1 −2 ⋮ 1

1 1 2 −1 −5 ⋮ 4
𝑅2 (− ) [0 1 −1 −2 ⋮ 1]
5
~ 0 1 −1 −2 ⋮ 1
1 2 −1 −5 ⋮ 4
𝑅32 (−1) [0 1 −1 −2 ⋮ 1]
0 0 0 0 ⋮ 0

which is in Echelon form.

∴ 𝜌[𝐀 ⋮ 𝐁] = 𝜌[𝐀] = 2 < 𝑛 [𝑛 = 4 the number of unknowns ]

∴ The equations are consistent and have infinite many solutions. The matrix equation 𝐀𝐗 = 𝐁 can be written as
follows :
𝑥
1 2 −1 −5 𝑦 4
[0 1 −1 −2] [ ] = [1]
𝑧
0 0 0 0 𝑤 0
∴ 𝑥 + 2𝑦 − 𝑧 − 5𝑤 = 4, 𝑦 − 𝑧 − 2𝑤 = 1.
Taking 𝑧 = 𝜆1 , 𝑤 = 𝜆2 (arbitrary values)
∴ 𝑥 = −𝜆1 + 𝜆2 + 2, 𝑦 = 𝜆1 + 2𝜆2 + 1, 𝑧 = 𝜆1 , 𝑤 = 𝜆2 .
Ans.
Problem 24. Find that for what values of 𝜆, 𝜇 the equations

𝑥 + 𝑦 + 𝑧 = 6, 𝑥 + 2𝑦 + 3𝑧 = 10, 𝑥 + 2𝑦 + 𝜆𝑧 = 𝜇

have (i) no solution (ii) 𝑎 unique solution (iii) infinite many solutions.
Solution. The matrix equation 𝐀𝐗 = 𝐁 is given by

1 1 1 𝑥 6
𝑦 10
[1 2 3] [ ] = [ ] (1)
1 2 𝜆 𝑧 𝜇
1 1 1 ⋮ 6
[𝐀 ⋮ 𝐁] = [1 2 3 ⋮ 10]
1 2 𝜆 ⋮ 𝜇

𝑅21 (−1) 1 1 1 ⋮ 6
~ [0 1 2 ⋮ 4]
𝑅32 (−1) 1 0 𝜆−3 ⋮ 𝜇 − 10

Now consider the following cases :

Case I. If 𝜆 ≠ 3, then 𝜌(𝐴) = 𝜌[𝐴 ⋮ 𝐵] = 3 = Number of unknowns.
Hence in this case the equations are consistent and will have a unique solution.
Case II. If 𝜆 = 3, 𝜇 = 10, then
Clearly 𝜌(𝐴) = 𝜌[𝐴 ⋮ 𝐵] = 2 < 3 (number of unknowns).

Hence in this case the equations are consistent and will have infinite many solutions.
Case III. If 𝜆 = 3, 𝜇 ≠ 10, then

𝜌[𝐀 ⋮ 𝐁] = 3, 𝜌(𝐴) = 2

∴ 𝜌(𝐀) ≠ 𝜌[𝐀 ⋮ 𝐁]

Hence in this case the equations are inconsistent and have no solution.
Problem 25. For what values of 𝜆, the equations 𝑥 + 𝑦 + 𝑧 = 1, 𝑥 + 2𝑦 + 4𝑧 = 𝜆. 𝑥 + 4𝑦 + 10𝑧 = 𝜆2 have a
solution and solve completely in each case.
Solution. The matrix equation 𝐀𝐗 = 𝐁 is given by

1 1 1 𝑥 1
[1 2 4 ] [𝑦] = [ 𝜆 ] (1)
1 4 10 𝑧 𝜆2
1 1 1 ⋮ 1
[𝐴 ⋮ 𝐵] = [1 2 4 ⋮ 𝜆]
1 4 10 ⋮ 𝜆2
𝑅21 (−1) 1 1 1 ⋮ 1
~ [0 1 3 ⋮ 𝜆−1]
𝑅31 (−3) 0 3 9 ⋮ 𝜆2 − 2

1 1 1 ⋮ 1
𝑅32 (−3)
[0 1 3 ⋮ 𝜆 − 1 ].
~
0 0 0 ⋮ 𝜆2 − 3𝜆 + 2

Deleting the last column of matrix [A:B], clearly 𝜌(𝑨) = 2 Therefore the equations are consistent if the rank of
the augmented matrix [𝑨: 𝑩] is also 2 .
∴ 𝜆2 − 3𝜆 + 2 = 0 ⇒ 𝜆 = 1,2
Thus the equations have infinite many solutions when 𝜆 = 1 or 𝜆 = 2
Case I. When 𝜆 = 1, then equation (1) is given by

1 1 1 𝑥 1
[0 1 3] [𝑦] = [0]
0 0 0 𝑧 0

⇒ 𝑥 + 𝑦 + 𝑧 = 1, 𝑦 + 3𝑧 = 0
⇒ If 𝑧 = 𝑘, then 𝑥 = 2𝑘 + 1, 𝑦 = −3𝑘.

Problem 26. Prove that the following system of equations is inconsistent

𝑥 − 2𝑦 + 𝑧 − 𝑤 + 1 = 0,3𝑥 − 2𝑧 + 3𝑤 + 4 = 0,5𝑥 − 4𝑦 + 𝑤 + 3 = 0

Solution. Here the augmented matrix is

1 −2 1 −1 ⋮ −1
[𝐀 ⋮ 𝐁] = [3 0 −2 3 −4]
5 −4 0 1 ⋮ −3
𝑅21 (−3) 1 −2 1 −1 −1
∼ [0 6 −5 6 −1]
𝑅31 (−5) 0 6 −5 6 2
1 −2 1 −1 ⋮ −1
𝑅32 (−1)
[0 6 −5 6 ⋮ −1]
~
0 0 0 0 ⋮ 3
1
𝑅2 ( ) 1 −2 1 −1 ⋮ −1
6
~ [0 1 −5/6 6 ⋮ −1/6]
1 0 0 0 0 ⋮ 1
𝑅3 ( )
3

Clearly 𝜌[𝐀 ⋮ 𝐁] = 3, 𝜌(𝐴) = 2

∴ 𝜌(𝐀) < 𝜌[𝐀 ⋮ 𝐁].

Hence the equations are inconsistent.

9. Some Theorems on Characteristic Roots and Vectors :

(I) There exists one and only one characteristic root corresponding to a characteristic vector 𝐗 of a square
matrix A.

Proof. If possible let for a given characteristic vector 𝐗 of a square matrix 𝐀 there exists two distinct
characteristic roots 𝜆1 and 𝜆2 .

Hence for characteristic roots 𝜆1 and 𝜆2 , we have

𝐀𝐗 = 𝜆1 𝐗, and 𝐀𝐗 = 𝜆2 𝐗
∴ 𝜆1 𝐗 = 𝜆2 𝐗 ⇒ (𝜆1 − 𝜆2 )𝐗 = 0 ⇒ 𝜆1 − 𝜆2 = 0 [∵ X ≠ 0]
∴ 𝜆1 = 𝜆2 .

Hence there can not exist more than one characteristic root corresponding to a characteristic vector 𝐗 of a
square matrix 𝐀
(II) 𝜆 is a characteristic root of a square matrix 𝐀, if and only if there exists a non-zero vector 𝐗 such that 𝐀𝐗 =
𝜆𝐗

Proof. Let 𝐀 be a 𝑛 × 𝑛 square matrix. If 𝜆 is a characteristic root of the matrix 𝐀, then 𝜆 will satisfy the
characteristic equation of 𝐀 Hence |𝐀 − 𝜆𝐈| = 0 i.e., 𝐀 − 𝜆𝐈 is a singular matrix. Therefore, there exist a non
-zero solution of the equation

(A − 𝜆I)X = 0 or 𝐀𝐗 = 𝜆𝐗

i.e, there exist a non-zero vector 𝐗 such that 𝐀𝐗 = 𝜆𝐗.

Conversely. Suppose there exists a non-zero vector 𝐗, such that 𝐀𝐗 = 𝜆𝐗 or (𝐀 − 𝜆𝐈)𝐗 = 𝟎.

Since the solution of the equation (𝐴 − 𝜆𝐼)𝑋 = 0 is non-zero, therefore the coefficient matrix 𝐀 − I𝜆 is singular
[Link] |𝐀 − 𝜆I| = 0. It concludes that 𝜆 is a characteristic root of the matrix 𝐀.

Note. The above theorem can be stated as. 'If 𝜆 is a latent root of a square matrix 𝐀; then the equation 𝐀𝐗 = 𝜆𝐗
has a non-trivial solution.'
(III) If 𝜆 is a characteristic root of a square matrix 𝐀, then there exists more than one characteristic vector
corresponding to 𝜆.

Proof. Let 𝑘 ≠ 0 be a scalar. If 𝐗 is a characteristic vector corresponding to 𝜆, then we shall prove that 𝑘𝐗 is also
a characteristic vector corresponding to 𝜆.

Since 𝐗 is a characteristic vector of A corresponding to 𝜆, hence we have

𝐀𝐗 = 𝜆𝐗, then 𝑘(𝐀𝐗) = 𝑘(𝜆𝐗) [∵ 𝑘 ≠ 0]

𝐀(𝑘𝐗) = 𝜆(𝑘𝐗)

It shows that 𝑘𝐗 is also a characteristic vector of 𝐀 corresponding to the characteristic root 𝜆.

(IV) The characteristic (eigen) roots of a diagonal matrix are given by its diagonal elements.

Proof. Let Abe a 𝑛 × 𝑛 diagonal matrix. Hence the characteristic equation of 𝐀 is |𝐀 − 𝜆𝐈| = 0
𝑎11 − 𝜆 0 0…………0
0 𝑎22 − 𝜆 0 … … … … 0
i.e., | |=0
…………………… …………
0 0 0𝑎𝑛𝑛 − 𝜆

⇒ (𝑎11 − 𝜆)(𝑎22 − 𝜆) … (𝑎𝑛𝑛 − 𝜆) = 0

⇒ 𝜆 = 𝑎11 , 𝜆 = 𝑎22 , … , 𝜆 = 𝑎𝑛𝑛 .

Hence the characteristic roots (or eigen values) are 𝑎11 , 𝑎22 , … , 𝑎𝑛𝑛 which are the diagonal elements of A.
(𝐕) The product of eigen values of a square matrix A of order 𝑛 is equal to the determinant of the matrix.

Proof. Let 𝐀 = [𝑎𝑖𝑗 ] be a square matrix of order 𝑛 and let 𝜆1 , 𝜆2 , … , 𝜆𝑛 be its eigen values. If 𝐹(𝜆) is characteristic
polynomial, then

𝑎11 − 𝜆 𝑎12 … … … … … 𝑎1𝑛

𝐹(𝜆) = |𝐴 − 𝜆𝐼| = || 𝑎21 𝑎22 − 𝜆 … … … 𝑎2𝑛 ||
……… ……………………
an1 𝑎𝑛2 … 𝑎𝑛𝑛 − 𝜆

= (−1)𝑛 (𝜆𝑛 + 𝑏1 𝜆𝑛−1 + ⋯ + 𝑏𝑛−1 𝜆 + 𝑏𝑛 ) (1)

= (−1)𝑛 (𝜆 − 𝜆1 )(𝜆 − 𝜆2 ) … (𝜆 − 𝜆𝑛 ) (2)

Putting 𝜆 = 0, we have

𝐹(0) = |𝐀| = 𝜆1 𝜆2 … 𝜆𝑛 = (−1)𝑛 𝑏𝑛

If shows that the product of eigen values of 𝐀 is equal to the determinant |𝐀|.
(VI) The matrices 𝐀 and 𝐏 −1 𝐀𝐏 both have the same characteristic roots where 𝐏 is an invertible matrix of the
same order.

Proof. The two matrices 𝐀 and 𝐏 −1 𝐀𝐏 will have the same characteristic roots if their characteristic equations
are same.

Let 𝐁 = 𝐏 −1 𝐀𝐏

∴ 𝐁 − 𝜆𝐈 = 𝐏 −1 𝐀𝐏 − 𝜆𝐈 = 𝐏 −1 𝐀𝐏 − 𝜆(𝐏 −1 𝐏) [∵ 𝐈 = 𝐏 −1 𝐏]

= 𝐏 −1 𝐀𝐏 − 𝐏 −1 𝜆𝐏 = 𝐏 −1 (𝐀 − 𝜆𝐈)𝐏

∴ |𝐁 − 𝜆𝐈|=∣ 𝐏 −1 (𝐀 − 𝜆𝐈)𝐏|= |𝐏 −1 ||𝐀 − 𝜆𝐈||𝐏 ∣

𝐀 − 𝜆𝐈||𝐏 −1 ||𝐏| = |𝐀 − 𝜆𝐈||𝐏 −1 𝐏| = |𝐀 − 𝜆𝐈||𝐈| = |𝐀 − 𝜆𝐈|

Hence |𝐀 − 𝜆𝐈| = 0 and |𝐁 − 𝜆𝐈| = 0 are same.

(VII) The characteristic roots of an idempotent matrix are either zero or unity.
Proof. Let 𝐀 be an indempotent matrix.
∴ 𝐀𝟐 = 𝐀. Let 𝜆 be a characteristic root of 𝐀 and the corresponding vector be 𝐗. Hence 𝐗 ≠ 0 and

𝐀𝐗 = 𝜆𝐗

𝐀(𝐀𝐗) = 𝐀(𝜆𝐗) = 𝜆(𝐀𝐗)

(𝐀𝐀)𝐗 = 𝜆(𝜆𝐗) [∵ from , 𝐀𝐗 = 𝜆𝑋]

𝐀2 𝐗 = 𝜆2 𝐗 or 𝐀𝐗 = 𝜆2 𝐗 [∵ 𝐀𝟐 = 𝑨]

𝜆𝐗 = 𝜆2 𝐗 [∵ from , 𝐀𝐗 = 𝜆𝑋]

(𝜆2 − 𝜆)𝐗 = 0 or 𝜆2 − 𝜆 = 0 [∵ X ≠ 0]

𝜆(𝜆 − 1) = 0 or 𝜆 = 0,1.

(VIII) A square matrix A and its transpose matrix 𝐀′ have the same characteristic roots. 1

Proof. Let 𝜆 be a characteristic root of 𝐀. The two matrices 𝐀 and 𝐀′ will have the same characteristic roots if
their characteristic equations namely |𝐀 − 𝜆𝐈| = 0 and |"𝐀′ − 𝜆𝐈| = 0 are same.

Let 𝐀 = [𝑎𝑖𝑗 ]𝑛×𝑛

𝑎11 − 𝜆 𝑎12 ⋯ 𝑎𝑛1

𝑎 𝑎22 − 𝜆 ⋯ 𝑎𝑛2
|𝑨 − 𝜆𝐈| = 0 ⇒ | 21 | = 0. (1)
⋯ ⋯ ⋯ ⋯
𝑎1𝑛 𝑎2𝑛 ⋯ 𝑎𝑛𝑛 − 𝜆

𝑎11 − 𝜆 𝑎21 ⋯ 𝑎𝑛1

𝑎 𝑎22 − 𝜆 ⋯ 𝑎𝑛2
Again |𝐀′ − 𝜆𝐈| = 0 ⇒ | 12 |=0
⋯ ⋯ ⋯ ⋯
𝑎1𝑛 𝑎2𝑛 ⋯ 𝑎𝑛𝑛 − 𝜆

Interchange of rows and columns do not alter the value of the determinant, therefore
 𝑎11 − 𝜆 𝑎12 … 𝑎1𝑛
𝑎 𝑎22 − 𝜆 … 𝑎2𝑛
|𝐀′ − 𝜆I| = | 21 |= (2)
… … … …
𝑎1𝑛 𝑎2𝑛 … 𝑎𝑛𝑛 − 𝜆

Hence from (1) and (2), we have

|𝐴 − 𝜆𝐼| = 0 and |𝐴′ − 𝜆𝐼| = 0 are same.

(IX) The characteristic roots of a Hermitian matrix are real.

Proof. Let 𝐀 be a Hermitian matrix and 𝜆 be its one characteristic root. Let 𝐗 be the eigen vector corresponding
to 𝜆, then

𝐀𝐗 = 𝜆𝐗 (1)

Pre-multiplying (1) by 𝐗 ⊖ (Conjugate transpose of 𝐗 ), we have

𝐗 Θ 𝐀𝐗 = 𝐗 Θ 𝜆𝐗 or 𝐗 Θ 𝐀𝐗 = 𝜆𝐗 Θ 𝐗 (2)

Taking transpose conjugate of both sides of (2), we have

𝚯 Θ Θ
(𝑿𝚯 𝑨𝑿) = (𝜆𝑿Θ 𝑿)Θ ⇒ 𝑿Θ 𝑨Θ (𝑿Θ ) = 𝜆‾𝑿Θ (𝑿Θ )

𝑿Θ 𝑨𝑿 = 𝜆‾𝑿Θ 𝑿 (3)

[∵ 𝐗 Θ 𝐗 = 0]

∴ From (2) and (3)

𝜆𝐗 Θ 𝐗 = 𝜆‾𝐗 Θ 𝐗 ⇒ (𝜆 − 𝜆‾)𝐗 Θ 𝐗 = 𝟎

⇒ 𝜆 − 𝜆‾ = 0

⇒ 𝜆 = 𝜆‾ [∴ 𝐗 Θ 𝐗 ≠ 0]

It shows that 𝜆 is real.

(X) The characteristic roots of a real symmetric matrix are all real.

Proof. Let 𝐀 be a real symmetric matrix, then we know that 𝐀 = 𝐀 and 𝐀′ = 𝐀; thus A⊖ = A i.e., A is a
Hermitian matrix. Hence proceed as in theorem IX above.
(XI) The characteristic roots of a skew Hermitian matrix are either purely imaginary or zero.

Proof. Let 𝐀 be a skew Hermitian matrix, then 𝑖𝐀 is Hermitian matrix. Let 𝜆 be a characteristic root of 𝐀 and 𝐗
the eigen vector corresponding to 𝜆, then

𝐀𝐗 = 𝜆𝐗 ⇒ 𝑖(𝐀𝐗) = 𝑖(𝜆𝐗) ⇒ (𝑖 A)𝐗 = (𝑖𝜆)𝐗

It concludes that 𝑖𝜆 is a characteristic root of 𝑖𝐀, but 𝑖𝐴 is Hermitian matrix and so 𝑖𝜆 is real [See theorem IX
above]. Hence 𝜆 is either zero or purely imaginary.
(XII) The characteristic roots of a real skew symmetric matrix are either all zeroes or purely imaginary.

Proof. We know that if all elements of a skew-Hermitian matrix are real, then it is called a real skew-symmetric
matrix. Hence for its proof, proceed as in theorem XI above.
(XIII) The characteristic roots of a unitary matrix are of unit modulus.
Proof. Let 𝐴 be a unitary matrix, then

𝐀𝚯 𝐀 = 𝐈 (1)

Let 𝜆 be a characteristic root of 𝐀 and 𝐗 be the eigen vector corresponding to 𝜆, then

𝐀𝐗 = 𝜆𝐗 (2)

Taking conjugate transpose of both sides of (2), we have

(𝐀𝐗)Θ = (𝜆𝐗)Θ ⇒ 𝐗 Θ 𝐀Θ = 𝜆‾𝐗 Θ (3)

Multiplying corresponding sides of (2) and (3), we have

(𝐗 Θ 𝐀Θ )(𝐀𝐗) = (𝜆‾𝐗 Θ )(𝜆𝐗)

𝐗 Θ (𝐀Θ 𝐀)𝐗 = 𝜆𝜆‾𝐗 Θ 𝐗 or 𝐗 Θ 𝐈 = 𝜆𝜆‾𝐗 Θ 𝐗 [∵ 𝐀ɵ 𝑨 = 𝑰]

𝐗 Θ 𝐗 = 𝜆𝜆‾𝐗 Θ 𝐗 ⇒ (1 − 𝜆𝜆‾)𝐗 Θ 𝐗 = 0 (4)

∴ From (4), 1 − 𝜆𝜆‾ = 0 [∵ 𝐗 ɵ 𝑿 ≠ 0]

𝜆𝜆‾ = 1, or |𝜆|2 = 0, or |𝜆| = 1

It shows that the characteristic roots of unitary matrix are of unit modulus.
(XIV) The characteristic roots of an orthogonal matrix are of unit modulus.

Proof. Let A be an orthogonal matrix, then

𝐀′ 𝐀 = I (1)

now proceed as in above theorem XIII.

10. THE CAYLEY-HAMILTON THEOREM

Statement. Every square matrix satisfies its characteristic equation.

Or

If the characteristic polynomial of a 𝐧 × 𝐧 square matrix 𝐴 is

|𝐀 − 𝜆𝐈| = (−1)𝑛 [𝜆𝑛 + 𝑎1 𝜆𝑛−1 + 𝑎2 𝜆𝑛−2 + ⋯ + 𝑎𝑛 ]

then matrix equation 𝑿𝒏 + 𝒂𝟏 𝑿𝒏−𝟏 + 𝒂𝟐 𝑿𝒏−𝟐 + ⋯ + 𝒂𝒏 𝑰 = 𝟎.

is satisfied by 𝑿 = 𝑨
Proof. Let 𝐀 = [𝛼𝑖𝑗 ]𝑛×𝑛 , hence

𝑎11 − 𝜆 𝑎12 … … … … … 𝑎1𝑛

𝐴 − 𝜆𝐼 = || 𝑎21 𝑎22 − 𝜆 … … … 𝑎2𝑛 ||
……… ……………………
an1 𝑎𝑛2 … 𝑎𝑛𝑛 − 𝜆

∴ Let

|𝐀 − 𝜆I| = (−1)𝑛 [𝜆𝑛 + 𝑎1 𝜆𝑛−1 + 𝑎2 𝜆𝑛−2 + ⋯ + 𝑎𝑛 ] (1)

Clearly the elements of the matrix 𝐀 − 𝜆𝐈 are at the most of first degree in 𝜆, therefore, the co-factors of |𝐀 − 𝜆𝐈|
will be at most of degree (𝑛 − 1) in 𝜆 i.e.; the elements of adj. (𝐀 − 𝜆𝐈) will be at most of degree (𝑛 − 1) in 𝜆.
Hence adj. (A - 𝜆𝐈 ) can be expressed as a matrix polynomial in 𝜆 as follows:

adj. (𝐀 − 𝜆𝐈) = 𝐵0 𝜆𝑛−1 + 𝐵1 𝜆𝑛−2 + ⋯ + 𝐵𝑛−2 𝜆 + 𝐵𝑛−1 (2)

where every 𝐵0 , 𝐵1 , … , 𝐵𝑛−1 , are square matrices of order 𝑛 × 𝑛 and their elements are functions of 𝑎𝑖𝑗 's.

Now we know that if 𝐃 is a square matrix, then

D . Adj. 𝐃 = |𝐃|. 𝐈 (3)

Putting 𝐃 = 𝐀 − 𝜆I in eqn. (3), we get

(𝐀 − 𝜆𝐈)Adj. (𝐀 − 𝜆𝐈) = |𝐀 − 𝜆𝐈|. 𝐈 (4)

Putting values from (1) and (2) in eqn. (4), we get

(𝐀 − 𝜆𝐼)(𝐵0 𝜆𝑛−1 + 𝐵1 𝜆𝑛−2 + ⋯ + 𝐵𝑛−2 𝜆 + 𝐵𝑛−1 )

= (−1)𝑛 (𝜆𝑛 + 𝑎1 𝜆𝑛−2 + 𝑎2 𝜆𝑛−2 + ⋯ + 𝑎𝑛 )𝐈 (𝟓)

Equating the coefficients of like powers of 𝜆 on both sides of equation (5), we have

−𝐈𝐵0 = (−1)𝑛 𝐈

𝐀𝐵0 − 𝐈𝐵1 = (−1)𝑛 𝑎1 𝐈

A𝐵1 − 𝐋𝐵2 = (−1)𝑛 𝑎2 𝐈

……………………………….

……………………………….

𝐀𝐵𝑛−2 − 𝐈𝐵𝑛−1 = (−1)𝑛 𝑎𝑛−1 𝐈

𝐀𝐵𝑛−1 = (−1)𝑛 𝑎𝑛 𝐈

Now pre-multiplying above equations by 𝐀𝐧 , 𝐀𝐧−𝟏 , 𝐀𝐧−𝟐 , … , I respectively and adding, we get

0 = (−1)𝑛 [𝐀𝑛 + 𝑎1 𝐀𝑛−1 + 𝑎2 𝐀𝑛−1 + ⋯ + 𝑎𝑛 𝐈]

𝑨𝑛 + 𝑎1 𝑨𝑛−1 + 𝑎2 𝑨𝑛−2 + ⋯ + 𝑎𝑛 𝐈 = 0 (6)

Equation (6) shows that 𝐗: = 𝐀 satisfies the matrix equation

𝐗 𝑛 + 𝑎1 𝐗 𝑛−1 + 𝑎2 𝐗 𝑛−2 + ⋯ + ⋯ 𝑎𝑛 𝐈 = 0

Cor. To find A−1 . Proved.

If |𝐀| ≠ 0, then the inverse of matrix 𝐀 can be easily evaluated with the help of Cayley-Hamilton theorem. From
above theorem, we have

𝐀𝑛 + 𝑎1 𝐀𝑛−1 + 𝑎2 𝐀𝑛−2 + ⋯ + 𝑎𝑛 𝐈 = 0

If |𝐀| ≠ 0, then pre-multiplying by 𝐀−1 , we get

 𝐀𝑛−1 + 𝑎1 𝐀𝑛−2 + ⋯ + 𝑎𝑛−1 𝐈 + 𝑎𝑛 𝐀−1 = 0
1
⇒ 𝐀−1 = − [ 𝐀𝑛−1 + 𝑎1 𝐀𝑛−2 + ⋯ + 𝑎𝑛−1 𝐈], if 𝑎𝑛 ≠ 0.
𝑎𝑛

SOLVED PROBLEM
0 0 1
Problem 27. Show that Cayley Hamilton theorem is satisfied by the matrix 𝐀 = [ 3 1 0].
−2 1 4
Hence find 𝐀−1 .
0−𝜆 0 1
Solution. |𝐀 − 𝜆𝐈| = [ 3 1−𝜆 0 ]
−2 1 4−𝜆

= −𝜆(1 − 𝜆)(4 − 𝜆) + 1{3 + 2(1 − 𝜆)} = −𝜆3 + 5𝜆2 − 6𝜆 + 5.

Characteristic equation of 𝐀 is given by

|𝐴 − 𝜆𝐈| = −𝜆3 + 5𝜆2 − 6𝜆 + 5 = 0

𝜆3 − 5𝜆2 + 6𝜆 − 5 = 0 (1)

Now we are to show that

𝐀3 − 5𝐀2 + 6𝐀 − 5𝐈 = 0 (2)

0 0 1 0 0 1 −2 1 4
∴ 𝐀2 = [ 3 1 0] [ 3 1 0] = [ 3 1 3]
−2 1 4 −2 1 4 −5 5 14
−2 1 4 0 0 1 −5 5 14
𝐀3 = 𝐀2 ⋅ 𝐀 = [ 3 1 3] [ 3 1 0] = [ −3 4 15]
−5 5 14 −2 1 4 −13 19 51
−5 5 14 −2 1 4 0 0 1 1 0 0
∴ 𝐀3 − 5𝐀2 + 6𝐀 − 5𝐈 = [ −3 4 15 ] − 5 [ 3 1 3 ] + 6 [ 3 1 0 ] − 5 [ 0 1 0]
−13 19 51 −5 5 14 −2 1 4 0 0 1
−5 5 14 10 −5 −20 0 0 6 −5 0 0
= [ −3 4 15] + [−15 −5 −15] + [ 18 6 0] + [ 0 −5 0]
−13 19 51 25 −25 −70 −12 6 24 0 0 −5
0 0 0
= [0 0 0] = 0.
0 0 0

Problem 28. Using Cayley Hamilton theorem express 2𝐀5 − 3𝐀4 + 𝐀2 − 4𝐈 in the form of linear polynomial in
3 1
𝐀, where, 𝐀 = [ ].
−1 2
3−𝜆 1
Solution. |𝐴 − 𝜆𝐼| = | | = (3 − 𝜆)(2 − 𝜆) + 1 = 𝜆2 − 5𝜆 + 7.
−1 2−𝜆

∴ Characteristic equation of 𝐀 is

(𝐴 − 𝜆𝐼) = 0 or 𝜆2 − 5𝜆 + 7 = 0 (1)

Equation (1) is satisfied by matrix A (By Calyley-Hamilton theorem)

 𝐀2 − 5𝐀 + 7𝐈 = 0 or 𝐀2 = 5𝐀 − 7𝐈 (2)

Multiplying both sides of equation (2) successively by 𝐴, 𝐴2 , 𝐴3 , we have

𝐀𝟑 = 5 𝐀2 − 7 𝐀 (3)

𝑨4 = 5𝑨𝟑 − 7𝑨2 (4)

𝑨5 = 5𝑨𝟒 − 7𝑨𝟑 (5)

∴ 2𝑨5 − 3𝑨4 + 𝑨2 − 4𝑰 = 2(5𝑨4 − 7𝑨3 ) − 3𝑨4 + 𝑨2 − 4𝑰 [From (5)]

= 7 𝑨4 − 14 𝑨3 + 𝑨2 − 4𝑰 = 7(5 𝑨3 − 7 𝑨2 ) − 14 𝑨3 + 𝑨2 − 4𝑰 [From (4)]

= 21𝑨3 − 48𝑨2 − 4𝑰 = 21(5𝑨2 − 7𝑨) − 48𝑨2 − 4𝑰 [ From (3)]

= 57 𝑨2 − 147 𝑨 − 4𝑰 = 57(5 𝑨 − 7𝐈) − 147 𝑨 − 4𝑰 [ From (2)]

= 138 𝑨 − 403𝑰

Which is a required linear polynomial in A.

2 −1 1
Problem 29. Find the characteristic equation of the matrix 𝐀 = [−1 2 −1] and verify that it is satisfied
1 −1 2
by 𝐴. Hence find 𝐴−1 .

Or

2 −1 1
Verify Cayley-Hamilton theorem for the matrix 𝐀 = [−1 2 −1]. Hence compute 𝐴−1 .
1 −1 2
2 − 𝜆 −1 1
Solution. |𝐀 − 𝜆𝐈| = | −1 2 − 𝜆 −1 |
1 −1 2 − 𝜆

= (2 − 𝜆){(2 − 𝜆)2 − 1} + 1{(−1) ⋅ (2 − 𝜆) + 1} + 1{1 − (2 − 𝜆)}

= (2 − 𝜆)(3 − 4𝜆 + 𝜆2 ) + (𝜆 − 1) + (𝜆 − 1)
= −𝜆3 + 6𝜆2 − 9𝜆 + 4 = 𝜆3 − 6𝜆2 + 9𝜆 − 4

The characteristic equation of 𝐀 is

|𝐴 − 𝜆𝐼| = 𝜆3 − 6𝜆2 + 9𝜆 − 4 = 0

Now we shall show that

𝐀3 − 6𝐀2 + 9𝐀 − 4𝐈 = 0

2 −1 1 2 −1 1 6 −5 5
𝐴2 = [−1 2 −1] [−1 2 −1] = [−5 6 −5] (1)
1 −1 2 1 −1 2 5 −5 6
2 −1 1 6 −5 5 22 −21 21
𝐀3 = 𝐀 ⋅ 𝐀2 = [−1 2 −1] [−5 6 −5] = [−21 22 −21].
1 −1 2 5 −5 6 21 −21 22

Hence 𝑨𝟑 − 6𝑨𝟐 + 9𝑨 − 4𝑰
 22 −21 21 6 −5 5 2 −1 1 1 0 0
= [−21 22 −21] − 6 [−5 6 −5] + 9 [−1 2 −1] − 4 [0 1 0]
21 −21 22 5 −5 6 1 −1 2 0 0 1
0 0 0
= [0 0 0] = 0.
0 0 0

Now multiplying equation (1) by 𝐴−1 , we get

𝐀2 − 6𝐀 + 9𝐈 − 4𝐀−1 = 0
1
∴ 𝐀−1 = (𝐀2 − 6𝐀 + 9𝐈)
4
6 −5 5 −12 6 −6 9 0 0 3 1 −1
2
Now 𝐀 − 6𝐀 + 9𝐈 = [−5 6 −5] + [ 6 −12 6] + [0 9 0] = [ 1 3 1]
5 −5 6 −6 6 −12 0 0 9 −1 1 3
3 1 −1
1
∴ 𝐀−1 = [ 1 3 1].
4
−1 1 3
Ans.