MATRICES, DETERMINANT,

SYSTEM OF LINEAR EQUATION

Prepared by: Dr. ARNAB KUMAR DE
Assistant Professor, W.B.E.S.
H.O.D., Basic Science and Humanities Department
Author of “Multi-Objective Stochastic Programming in Fuzzy Environments”
Publisher: IGI Global, USA
 MATRICES
A rectangular array of elements (real or complex) into rows
and columns is called a matrix of order . An is expressed

in the form

If , then the matrix is said to be a square matrix of order .

In an matrix, if , then the matrix is said to be a row matrix
or row vector and if , then the matrix is said to be a column
matrix or column vector.
 TYPES OF MATRICES
Null matrix
If each element of an is zero, then the matrix is called a null matrix.
Diagonal matrix
A square matrix is called a diagonal matrix if all the non-diagonal elements are zero.
Scalar matrix
A diagonal matrix is said to be a scalar matrix if all the diagonal elements be the same
scalar.
Unit matrix or Identity matrix
A scalar matrix whose diagonal elements are all one, is called an unit matrix.
Triangular matrix
A square matrix is said to be an upper triangular matrix if .
×
A square matrix is said to be a lower triangular matrix if .
×
A square matrix is said to be a triangular matrix if it is either upper triangular or lower
triangular.
 ALGEBRAIC OPERATIONS ON MATRICES
1. Scalar multiplication
Let 𝐴 = (𝑎 ) × be a matrix of order 𝑚 × 𝑛 and 𝑐 be any scalar. Then the scalar multiplication of 𝐴 by 𝑐 is
defined as 𝑐𝐴 = (𝑐𝑎 ) × .
2. Addition of matrices
Two matrices 𝐴 and 𝐵 are said to be conformable for addition if they have the same order. Let 𝐴 = (𝑎 ) × and
𝐵 = (𝑏 ) × be two matrices of order 𝑚 × 𝑛. Then their sum 𝐴 + 𝐵 is defined as 𝐴 + 𝐵 = (𝑐 ) × where
𝑐 = 𝑎 + 𝑏 , 𝑖 = 1,2, … , 𝑚; 𝑗 = 1,2, … , 𝑛.
Properties
i. Matrix addition is commutative, i.e., 𝐴 + 𝐵 = 𝐵 + 𝐴.
ii. Matrix addition is associative, i.e., 𝐴 + 𝐵 + 𝐶 = 𝐴 + 𝐵 + 𝐶.
3. Matrix Multiplication
Two matrices 𝐴 and 𝐵 are said to be conformable for multiplication if the number of columns of 𝐴 be equal to the
number of rows of 𝐵. Let 𝐴 = (𝑎 ) × and 𝐵 = (𝑏 ) × be two matrices of order 𝑚 × 𝑛 and 𝑛 × 𝑝. Then their
product 𝐴𝐵 is a matrix order 𝑚 × 𝑝 and is defined as 𝐴𝐵 = (𝑐 ) × where 𝑐 = ∑ 𝑎 𝑏 , 𝑖 = 1,2, … , 𝑚; 𝑗 =
1,2, … , 𝑝.
Properties
i. Matrix multiplication is non-commutative in general, i.e., 𝐴𝐵 ≠ 𝐵𝐴.
ii. Matrix multiplication is associative, i.e., 𝐴 𝐵𝐶 = (𝐴𝐵)𝐶.
iii) Matrix multiplication is distributive with respect to matrix addition, i.e., 𝐴 𝐵 + 𝐶 = 𝐴𝐵 + 𝐴𝐶.
 INVERSE OF A MATRIX
A non-zero square matrix of order is said to be invertible if there exists a
non-zero matrix of the same order such that .
Properties
1. An invertible matrix has unique inverse.
2. A non-zero square matrix is invertible iff it is non-singular.
( )
3.
4. If and be invertible matrices of the same order, then is invertible
and .
5. If be an invertible matrix, then is invertible and .
 ELEMENTARY OPERATIONS
An elementary operation on a matrix is an operation of the following
three types.
1. Interchange of two rows or columns of .
2. Multiplication of a row or column by a non-zero scalar.
3. Addition of a scalar multiple of one row or column to another row or
column.
 ROW-REDUCED MATRIX
An matrix is called row-reduced if
i) the first non-zero element in each non-zero row is one (called the
leading )
ii) each column containing the leading of some row has all its other
elements zero.
 ROW-REDUCED ECHELON MATRIX
An matrix is called a row-reduced echelon matrix if
i) is row-reduced.
ii) there is an integer such that the first rows of are
non-zero rows and the remaining rows are all zero rows.
iii) if the leading of the non-zero row occurs at the th
column of , then .
 RANK OF A MATRIX
Let be a non-zero matrix of order . The rank of is defined to be the
greatest positive integer such that has at least one non-zero minor of
order and every minor of order is zero.
The rank of a zero matrix is zero. For a non-zero matrix ,
.
Properties
1. .
2. If a row-reduced echelon matrix of has non-zero rows, then
.
3. The rank of a matrix remains invariant under an elementary row
operation.
 LINEAR DEPENDENCE AND LINEAR
INDEPENDENCE
A finite set of vectors is said to be linearly dependent if
there exists scalars not all zero, such that

The set is said to be linearly independent if the equality

is satisfied only when,
.
 DETERMINANT
Let 𝑆 be the set of all square matrices of order 𝑛 with real entries.

A scalar function 𝑓: 𝑆 → ℝ that assign to each square matrix 𝐴 = (𝑎 ) × , a real number 𝑓 𝐴 = det(𝐴), is called a determinant function. A

𝑎 𝑎 ⋯ 𝑎
𝑎 𝑎 ⋯ 𝑎
determinant is denoted by 𝑓 𝐴 = det 𝐴 = ⋯ ⋯ ⋯ ⋯ .
𝑎 𝑎 ⋯ 𝑎
Properties
1. For any square matrix 𝐴, det 𝐴 = det 𝐴 .
2. If any two rows or columns of a determinant be interchanged, then its numerical value is same but the sign is reversed.
3. If any two rows or columns of a determinant be identical, then the determinant value is zero.
4. If each element of any row or column is the sum of two elements, then the determinant can be expressed as the sum of two determinants.
5. If each element of any row or column be multiplied by a scalar, then the determinant be multiplied by the same scalar.
6. If a scalar multiple of one row or column be added to another row or column, then the value of the determinant remains unaltered.
7. If one row or column of a determinant be expressed as a linear combination of the remaining rows or columns, then the determinant value
is zero.
8. If all the elements of a determinant are polynomials in 𝑥 and if two rows or columns of the determinant become identical when 𝑥 = 𝑎, then
𝑥 − 𝑎 is the factor of the determinant.
 SYSTEM OF LINEAR EQUATION
Consider a system of 𝑚 linear equations in 𝑛 unknowns as
𝑎 𝑥 + 𝑎 𝑥 + ⋯+ 𝑎 𝑥 = 𝑏
𝑎 𝑥 + 𝑎 𝑥 + ⋯+ 𝑎 𝑥 = 𝑏
⋯ ⋯ ⋯
𝑎 𝑥 +𝑎 𝑥 +⋯+𝑎 𝑥 =𝑏
Where 𝑎 ′𝑠 and 𝑏 ′𝑠 are real numbers.
If 𝑏 = 0 ∀ 𝑖, then the system is said to be a homogeneous system, otherwise the system is called non-homogeneous.
If the above system has a solution, then the system is said to be consistent, otherwise the system is called inconsistent.
Matrix representation
𝑎 𝑎 ⋯ 𝑎 𝑏 𝑥
𝑎 𝑎 ⋯ 𝑎 𝑏 𝑥
If = ⋯ ⋯ ⋯ ⋯ ,𝐵= ⋯ and 𝑋 = ⋯ , then the above system of linear equations can be expressed
𝑎 𝑎 ⋯ 𝑎 𝑏 𝑥 ×
× ×
in matrix form as 𝐴𝑋 = 𝐵.
 CRAMER’S RULE
Let us consider a system of 𝑛 linear equations in 𝑛 unknowns
𝑎 𝑥 + 𝑎 𝑥 + ⋯+ 𝑎 𝑥 = 𝑏
𝑎 𝑥 + 𝑎 𝑥 + ⋯+ 𝑎 𝑥 = 𝑏
⋯ ⋯ ⋯
𝑎 𝑥 + 𝑎 𝑥 + ⋯+ 𝑎 𝑥 =𝑏

Where 𝑎 ′𝑠 and 𝑏 ′𝑠 are real numbers.

𝑎 𝑎 ⋯ 𝑎 𝑏 𝑎 ⋯ 𝑎 𝑎 𝑏 ⋯ 𝑎
𝑎 𝑎 ⋯ 𝑎 𝑏 𝑎 ⋯ 𝑎 𝑎 𝑏 ⋯ 𝑎
Let ∆= ⋯ ⋯ ⋯ ⋯ , then ∆ is called the coefficient determinant. For Cramer’s rule ∆≠ 0. In Cramer’s rule, let ∆ = ⋯ ⋯ ⋯ ⋯ ,∆ = ⋯ ,
⋯ ⋯ ⋯
𝑎 𝑎 ⋯ 𝑎 𝑏 𝑎 ⋯ 𝑎 𝑎 𝑏 ⋯ 𝑎
⋯, ⋯,

𝑎 𝑎 ⋯ 𝑏 𝑏
𝑎 𝑎 ⋯ 𝑏
∆ = ⋯ ⋯ i.e., ∆ is obtained from ∆ by replacing the 𝑖 − 𝑡ℎ column of ∆ by the right hand side vectors 𝑏⋯
⋯ ⋯
𝑎 𝑎 ⋯ 𝑏 𝑏

∆ ∆ ∆
Thus the solution of the system in Cramer’s rule is obtained as, 𝑥 = ∆
,𝑥 = ∆
,…,𝑥 = ∆
.
 GAUSS ELIMINATION METHOD
Let us consider a system of linear equations in unknowns

Where and are real numbers.

Then the coefficient matrix is and the augmented

matrix is

.
In Gauss elimination method the augmented matrix is modified as follows
𝑚𝑢𝑙𝑡𝑖𝑝𝑙𝑖𝑒𝑟 𝑎 𝑎 ⋯ 𝑎 𝑏
𝑎 ⁄𝑎 𝑎 𝑎 ⋯ 𝑎 𝑏
𝑎 ⁄𝑎 𝑎 𝑎 ⋯ 𝑎 𝑏 , provided 𝑎 ≠ 0.
… ⋯ ⋯ ⋯ ⋯ ⋯
𝑎 ⁄𝑎 𝑎 𝑎 ⋯ 𝑎 𝑏
Then, after first elimination

𝑚𝑢𝑙𝑡𝑖𝑝𝑙𝑖𝑒𝑟 𝑎 𝑎 ⋯ 𝑎 𝑏
( ) ( ) ( )
𝑎 ⋯ 𝑎 𝑏
( )
𝑎
( )
𝑎
( )
𝑎
( ) ⋯ 𝑎( ) 𝑏
( ) , provided 𝑎 ≠ 0.
… ⋯ ⋯ ⋯ ⋯
( ) ( ) ⋯
𝑎 𝑎 𝑎
( ) ( ) ( )
𝑎 𝑏
The second elimination gives,

𝑎 𝑎 𝑎 ⋯ 𝑎 𝑏
( ) ( ) ( ) ( )
𝑎 𝑎
⋯ 𝑎 𝑏
( )
𝑎 ⋯ 𝑎( ) 𝑏
( )
⋯ ⋯ ⋯
⋯ ⋯
( ) ( ) ( )
𝑎 𝑎 𝑏
Repeating the process for times, an upper triangular matrix is
obtained as,

From the last system, by back substitution process we get the required
solution of the system of linear equations. This process of elimination is
known as Gauss Elimination method.
 GAUSS JORDAN METHOD
Let us consider a system of linear equations in unknowns

Where and are real numbers.

Then the coefficient matrix is and the augmented

matrix is

.
In Gauss Jordan method the augmented matrix is modified as follows

, provided .

Then, after first elimination

( ) ( ) ( ) ( )
( )
( ) ( ) ( ) ( ) ( ) , provided .

( ) ( ) ( ) ( ) ( )
The second elimination gives,

( )
𝑎 𝑎 ⋯ 𝑎( )
𝑏
( )

( ) ( ) ( ) ( )
𝑎 𝑎
⋯ 𝑎 𝑏
( )
𝑎 ⋯ 𝑎( ) 𝑏
( )
⋯ ⋯ ⋯
⋯ ⋯
( ) ( ) ( )
𝑎 𝑎 𝑏

Repeating the process for (𝑛 − 1) times, an upper triangular matrix is obtained as,

( )
𝑎 ⋯ 𝑏
( ) ( )
𝑎 ⋯ 𝑏
( ) ⋯ ( )
𝑎 𝑏
⋯ ⋯ ⋯
⋯ ( ) ( )
𝑎 𝑏
From the last system, the 1 , 2 , … , 𝑛 variables can be obtained from the 1 , 2 ,…,𝑛 equation directly. This process of
elimination is known as Gauss Jordan method.