MT 1117: Linear Algebra for ICT

Instructor: A.V. Mathias

Department of Mathematics & Statistics
University of Dodoma

Lecture 3

December 2, 2021.

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 1 / 26
 Matrices

1 Concept of Determinant
Determinant by row expansion
Determinant by column expansion

2 Properties of determinants

3 Determinant by row reduction

4 Matrix Inverse
Concept of Inverse
Inverse by determinant
Inverse by row reduction
Further theorems and notes

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 2 / 26
 Determinant by row expansion

Definition
Determinant of n × n square matrix A =P [aij ] denoted by det(A) or |A|
is a scalar quantity defined by det(A) = nk=1 aik (−1)k+i Mik (A)

Mij (A) is the determinant of (n − 1) × (n − 1) matrix obtained

from A by deleting the ith row and jth column of A
The value Mij (A) is referred to as the (i, j)th minor of matrix A
 
2 1 0
Consider the matrix 1 1 −1
0 1 2
Let us determine M21 (A)

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 3 / 26
 Determinant by row expansion . . .
Erase the second row and the first column of matrix A

1 0
1 2
 
 1 0 
It implies that M21 (A) =   = 1(2) − 0(1) = 2
1 2 
Therefore, M21 (A) = 2

Example
 
 a11 a12 
  = a11 M11 (A) − a12 M12 (A) = a11 a22 − a12 a21
 a21 a22 

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 4 / 26
 Determinant by column expansion

Definition
The determinant of n × n square matrix A = [aij ] is defined by
det(A) = nk=1 akj (−1)k+j Mkj (A)
P

 
b11 b12 b13
Consider the matrix B given by B = b21 b22 b23 
b31 b32 b33
Expansion can be done along any column of matrix B
By expanding along the first column we obtain
det(B) = |B| = b11 M11 (B) − b21 M21 (B) + b31 M31 (B)
A matrix with a zero determinant is referred to as a singular
matrix other wise it is non-singular matrix

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 5 / 26
 Properties of determinant
D1
If A is an upper triangular matrix, then the determinant of A is the
product of all the diagonal elements of A.
 
4 4 1 1
0 −1 2 3
Consider the matrix A =  0 0 2 3


0 0 0 2
The determinant of matrix A is given by
|A| = 4 × −1 × 2 × 2 = −16
By using this property, we find that |In | = 1, that is
 1 0 ... 0 
 
 0 1 ... 0 
|In | = 
  = 1 × 1 × ... × 1 = 1
 . . . . . 0 

 0 0 ... 1 
Instructor: A.V. Mathias Department of Mathematics & MT 1117:University
Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 6 / 26
 Properties of determinants cont . . .
D2
If matrix B is obtained from matrix A by multiplying one row of A by a
scalar λ then det(B) = λdet(A).
 
5 0 10
Consider the matrix A = 5 5 5 
0 0 2
It follows
 that   
 5 0 10   1 0 2 
   
|A| =  5 5 5  = 5 ×  5 5 5 
 0 0 2   0 0 2 
 
 1 0 2 
 
= 5 × 5 ×  1 1 1 
 0 0 2 

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 7 / 26
 Properties of determinants cont . . .
 
 1 0 2 
 
= 5 × 5 × 2 ×  1 1 1 

 0 0 1 
      
 1 1   0 2   0 2 
= 50 × 1   − 1 + 0  
0 1   0 1   1 1 
= 50 × (1 − 0 + 0)
= 50 × 1
= 50
Therefore,
 
 5 0 10 
 
|A| =  5 5 5  = 50
 0 0 2 

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 8 / 26
 Properties of determinants cont . . .
D3
If matrix B is obtained from matrix A by interchanging two rows of A
then det(B) = −det(A).

Example
Use D3 to show that if the matrix has a repeated row then its
determinant must be zero.

Let B be a matrix obtained by interchanging two identical rows of

matrix A.
It implies that matrix A = matrix B
But by D3, det(B) = −det(A)
Since A = B, then det(A) = −det(A)
Therefore 2det(A) = 0, implying det(A) = 0
Instructor: A.V. Mathias Department of Mathematics & MT 1117:University
Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 9 / 26
 Properties of determinants cont . . .
D4
If matrix B is obtained from matrix A by adding a multiple of one row
of A to another row of A then det(B) = det(A).
 
1 4 1 1
1 −1 2 3
Consider a matrix A = 0 0 2 3


0 0 0 2
Suppose that matrix B is obtained by applying the operation
R2 → R2 + (−R1 )  
1 4 1 1
0 −5 1 2
It implies that B = 
0 0 2 3


0 0 0 2
Therefore, det(B) = 1 × −5 × 2 × 2 = −20 = det(A)
Instructor: A.V. Mathias Department of Mathematics & MT 1117:University
Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 10 / 26
 Properties of determinants cont . . .
D5
If two matrices A and B have the same size then
det(AB) = det(A)det(B).

It should be noted that det(A + B) 6= det(A) + det(B)

D6
For all square matrices A, det(AT ) = det(A).
   
4 0 0 0 4 4 1 1
4 1 0 0
, then AT = 0 1 2 0
 
Let A = 1 2 −2 0 0 0 −2 1
1 0 1 2 0 0 0 2
T
By D6 and D1; det(A) = det(A ) = 4 × 1 × −2 × 2 = −16

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 11 / 26
 Determinant by row reduction
This involves evaluating the determinant of a given matrix by
considering properties of determinant.
The matrix is reduced to r.e.f keeping the determinant symbol and
then the diagonal entries are multiplied to obtain the determinant.
 
1 2 1
Consider the matrix A = 2 3 4
1 2 2
Now,  
 1 2 1 
 
|A| =  2 3 4 
 1 2 2 
Apply the operation R2 → R2 + (−2R1 )

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 12 / 26
 Determinant by row reduction
This operation does not change the determinant, (Refer to D4)
So
 we have,  
 1 2 1   1 2 1 
   
 2 3 4  =  0 −1 2 
   
 1 2 2   1 2 2 
Apply the operation R3 → R3 + (−R1 ).
This operation also does not change the determinant (Refer to D4)
   
 1 2 1   1 2 1 
   
 0 −1 2  =  0 −1 2 
   
 1 2 2   0 0 1 
We then apply D1,
Therefore, det(A) = 1 × −1 × 1 = −1.
Note: You can use any operation of your choice at any stage
Instructor: A.V. Mathias Department of Mathematics & MT 1117:University
Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 13 / 26
 Matrix Inverse

Definition
The n × n matrix A is invertible if there exist n × n matrix B such that
AB = BA = In .

In this particular case, matrix B is called the inverse of matrix A

and is denoted by A−1 .
If there is no such B, then we say that A is singular (has no
inverse).
D7
Matrix A is invertible if and only if det(A) 6= 0
 
1 1
Consider a matrix A =
 2 3
3 −1
Therefore, A−1 =
−2 1
Instructor: A.V. Mathias Department of Mathematics & MT 1117:University
Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 14 / 26
 Inverse by determinant
Adjoint, minor and cofactor matrices

The matrix of minors of n × n matrix A = [aij ] is the matrix

M(A) = [Mij (A)] of the same size.
The matrix of cofactors of matrix A denoted by Acof is the matrix
Acof = [Aij ] of the same size.
The adjoint matrix of matrix A is the transpose of matrix of
cofactors of matrix A.
It implies that adj(A) = ATcof
 
1 2 0
Consider the matrix A = 0 0 −1
0 2 1
The determinant, |A| = 2
Instructor: A.V. Mathias Department of Mathematics & MT 1117:University
Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 15 / 26
 Inverse by determinant cont . . .
The matrix of minors M(A) is given by
 
M11 (A) M12 (A) M13 (A)
M(A) = M21 (A) M22 (A) M23 (A)
M31 (A) M32 (A) M33 (A)
It follows that,
     
 0 −1   0 −1   0 0 
     
 2 1   0 1   0 2 
     
 2 0   1 0   1 2 
M(A) =  
      
 2 1 
 
 0 1   0
  2 

 2 0   1 0   1 2 
     
 0 −1   0 −1   0 0 
 
2 0 0
Therefore, M(A) =  2 1 2
−2 −1 0
Instructor: A.V. Mathias Department of Mathematics & MT 1117:University
Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 16 / 26
 Inverse by determinant cont . . .
To get the matrixof cofactors
 Acof of matrix A overlay M(A) with
+ − +
the check board − + −
+ − +
Therefore,
 
2 0 0
Acof = −2 1 −2
−2 1 0
The adjoint matrix adj(A) is the transpose of cofactor matrix Acof .
adj(A) = ATcof
Thus,  
2 −2 −2
adj(A) = 0 1 1
0 −2 0
Instructor: A.V. Mathias Department of Mathematics & MT 1117:University
Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 17 / 26
 Inverse by determinant cont . . .
By using matrix A, we can now compute A adj(A),
  
1 2 0 2 −2 −2
A adj(A) = 0 0 −1 0 1 1
0 2 1 0 −2 0
 
2 0 0
= 0 2 0

0 0 2
 
1 0 0
= 2 0 1 0 = 2I3
0 0 1
But 2 = det(A),
This implies that A adj(A) = det(A)I3 .

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 18 / 26
 Inverse by determinant cont . . .

D8
For any n × n square matrix A, A adj(A) = det(A)In

Let A be a square matrix where det(A) 6= 0, applying D8

A adj(A)
det(A)
= In
Multiplying to the left by A−1 ,
A−1 Adet(A)
adj(A)
= A−1 In
Since A−1 A = In and A−1 In = A−1 , then
In adj(A)
det(A)
= A−1 , which implies adj(A)
det(A)
= A−1
Therefore, A−1 = 1
det(A)
adj(A)

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 19 / 26
 Inverse by determinant cont . . .

D9
If A is a square matrix such that det(A) 6= 0 then A−1 = 1
det(A)
adj(A)
 
a b
Consider the matrix A =
c d
   
d c d −c
Using this matrix M(A) = , Acof = and
  b a −b a
d −b
adj(A) =
−c a
So, by D9,
 
−1 1 d −b
A = ad−bc
−c a

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 20 / 26
 Inverse by row reduction
Consider n × n matrix A, to obtain its inverse you need to follow
the following steps:
Form the augmented matrix [A|In ]
Reduce the matrix [A|In ] to its r.r.e.f matrix say [C|B]
If C = In , then A is non singular and A−1 = B
If C 6= In , then C has a row of zeros and A is singular matrix with
no inverse.
Example
 
−1 1 1
Find A given that A =
2 3

Form theaugmented matrix

 [A|I2 ]
1 1 1 0
[A|I2 ] =
2 3 0 1
Instructor: A.V. Mathias Department of Mathematics & MT 1117:University
Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 21 / 26
 Inverse by row reduction cont . . .
Apply the operation R2 → R2 + (−2R1 )
   
1 1 1 0 1 1 1 0
∼
2 3 0 1 0 1 −2 1
Then apply the operation R1 → R1 + (−R2 )
   
1 1 1 0 1 0 3 −1
∼
0 1 −2 1 0 1 −2 1
 
1 0 3 −1
Taking the augmented matrix = [C|B]
0 1 −2 1
Since C = I2 , then B is the inverse of A
Therefore, A is non singular and
 
−1 3 −1
A =
−2 1

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 22 / 26
 Further theorems and notes

Important note:
If A is row-equivalent to In , then it is non-singular.
If A is row-equivalent to a matrix with a row of zeros, then it is
singular.

Theorem
If A is non-singular then A−1 is non-singular and (A−1 )−1 = A

Proof
Given that A is non-singular, it implies that
A A−1 = A−1 A or A−1 A = A A−1
Therefore, A−1 is non-singular and (A−1 )−1 = A 

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 23 / 26
 Further theorems and notes cont . . .

Theorem
If A and B are non-singular matrices, then AB is non-singular and
(AB)−1 = B−1 A−1

Proof
Since both A and B are non-singular, then both A−1 and B−1 exist.
Thus,
AB(B−1 A−1 ) = A(BB−1 )A−1 ; Associative property of matrices
= AIn A−1 ; Definition of matrix inverse
= AA−1 ; Identity property
= In ; Definition of matrix inverse
Thus, AB is non-singular and (AB)−1 = B−1 A−1 

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 24 / 26
 Further theorems and notes cont . . .

Theorem
If A is non-singular, then AT is non-singular and (AT )−1 = (A−1 )T .

Proof
Given that A is non-singular, thus
AA−1 = In = A−1 A
Transposing the whole thing
(AA−1 )T = InT = (A−1 A)T
(A−1 )T AT = In = AT (A−1 )T
Therefore, AT is non-singular and (AT )−1 = (A−1 )T . 

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 25 / 26
 Thank you for your Attention.

Instructor: A.V. Mathias Department of Mathematics & MT 1117:University

Statistics of Dodoma
Linear Algebra for ICT(Lecture 3) December 2, 2021. 26 / 26