PROF. MORE G.R. LINEAR ALGEBRA S.N.D.

COE & RC, YEOL A

UNIT 5 : LINEAR ALGEBRA – MATRICES, SYSTEM OF LINEAR EQUATIONS

 RANK: A Number 𝑟 > 0 is said to be rank of matrix A if

(i) There exists at least one minor of order r is non-zero.
(ii) All minor of order 𝑟 + 1 is zero.
OR
The rank of matrix A is the maximum order of its non-vanishing minor and
it is denoted as 𝑹𝒂𝒏𝒌(𝑨) = 𝝆(𝑨) = 𝒓.

 Note:
i. If matrix A has non-zero minor of order r then 𝐑𝐚𝐧𝐤(𝐀) = 𝛒(𝐀) ≥ 𝐫.
ii. If matrix A has all minor of order 𝑟 + 1 is zero then 𝐑𝐚𝐧𝐤(𝐀) = 𝛒(𝐀) ≤ 𝐫.
iii. If A is 𝑚 × 𝑛 matrix, where 𝑚 ≠ 𝑛 then 𝐑𝐚𝐧𝐤(𝐀) = 𝛒(𝐀) ≤ 𝐦𝐢𝐧(𝐦, 𝐧).
 NORMAL FORM:
By performing elementary transformations we can reduce any matrix to one of the
following forms called as a normal form of that matrix
[Ir ], [Ir 0] , [Ir ], [Ir 0]
0 0 0
Where, Ir is identity matrix of order r & r is rank of matrix

 Note:
i. If matrix B is obtained from A by subjecting it to any elementary row or column
transformations then we write A ~ B [A is equivalent to B].
ii. Elementary transformations of a matrix do not alter the rank and order of matrix.
iii. Rank of A and B are same.

 PROCEDURE TO CONVERT GIVEN MATRIX IN NORMAL FORM

If A is matrix of order 𝑚 × 𝑛
STEP I: Make 𝑎11 = 1 by using elementary transformation
STEP II: Make all entries which are below 𝑎11 as a zero using row transformation
STEP III: Make all entries which are besides 𝑎11 as a zero using column transformation
STEP IV: Make 𝑎22 = 1 by using elementary transformation
STEP V: Make all entries which are below 𝑎22 as a zero using row transformation
STEP VI: Make all entries which are besides 𝑎22 as a zero using column transformation
STEP VII: Continue the process until we get the following Normal form of a matrix

I Ir 0
[ Ir ] , [Ir 0] , [ r ], [ ]
0 0 0

The study of mathematics begins in minuteness but ends in magnificence - Charles caleb colton
PROF. MORE G.R. LINEAR ALGEBRA S.N.D. COE & RC, YEOL A

 ECHELON OR CANONICAL FORM OF A MATRIX:

Let, a matrix A be of order 𝑚 ∗ 𝑛 then the canonical or echelon form of A is row
equivalent matrix of rank r with the following property
1) One or more element in each of the first r-rows are non-zero or unity ‘1’ & elements
in the remaining row are zeros.
2) In the first r-rows the first non-zero element in each row appears in a column to the
right of the first non-zero element of preceding row.
 Remark:
1 2 3 4 5
0 1 2 3 4
i. Matrix A must be reduced to the Echelon form such as 
0 0 1 2 3
 
0 0 0 0 0
(by performing only row transformations)
ii. For finding the rank of a matrix, Echelon form is very useful as it shows the maximum
order of non-vanishing minor and thus the rank is immediately obtained.
iii. To determine the rank of matrix Am x n reduce it to echelon form and if
m = total no. of rows in A and 𝑘 = total no. of rows which contain all zero elements.
Then 𝜌(𝐴) = 𝑚 − 𝑘. Also if k = 0 then 𝜌(𝐴) = 𝑚.
OR
𝑅𝑎𝑛𝑘 (𝐴) = 𝜌(𝐴) = No. of non − zero rows in row echelon form of matrix A
 Note: To convert a matrix into a “Echelon Form” or “Canonical Form” we use only row
transformation.
 PROCEDURE TO CONVERT GIVEN MATRIX IN ECHELOM FORM
If A is matrix of order 𝑚 × 𝑛
STEP I: Make 𝑎11 = 1 by using row transformation
STEP II: Make all entries which are below 𝑎11 as a zero using row transformation
STEP III: Make 𝑎22 = 1 by using row transformation
STEP IV: Make all entries which are below 𝑎22 as a zero using row transformation
STEP V: Make 𝑎33 = 1 by using row transformation
STEP VI: Make all entries which are below 𝑎33 as a zero using row transformation
STEP VII: Continue the process until we get the Echelon form of a matrix like
′ ′ ′
1 𝑎12 𝑎13 … … … . 𝑎1𝑛
′ ′
0 1 𝑎23 … … … . 𝑎2𝑛
′
𝐴 = 0 0 1 … … …. 𝑎3𝑛
. . . .
. . . .
[ 0 0 0………. 1 ]

The study of mathematics begins in minuteness but ends in magnificence - Charles caleb colton
PROF. MORE G.R. LINEAR ALGEBRA S.N.D. COE & RC, YEOL A

 FIND 𝑨−𝟏 ( IF NON-SINGULAR MATRIX P & Q ARE GIVEN) :

A−1 = QP
 SYSTEM OF LINEAR EQUATIONS:
Consider, System of ‘𝑚’ linear equations in ‘n’ unknowns X1 , X 2 , … … . . X n
𝑎11 𝑥1 + 𝑎12 𝑥2 + 𝑎13 𝑥3 + ⋯ … … + 𝑎1𝑛 𝑥𝑛 = 𝑏1
𝑎21 𝑥1 + 𝑎22 𝑥2 + 𝑎23 𝑥3 + ⋯ … … + 𝑎2𝑛 𝑥𝑛 = 𝑏2
𝑎31 𝑥1 + 𝑎32 𝑥2 + 𝑎33 𝑥3 + ⋯ … … + 𝑎3𝑛 𝑥𝑛 = 𝑏3
.
.
𝑎𝑚1 𝑥1 + 𝑎𝑚2 𝑥2 + 𝑎𝑚3 𝑥3 + ⋯ … … + 𝑎𝑚𝑛 𝑥𝑛 = 𝑏𝑚
The above system of linear equation can be written in matrix form
𝑎11 𝑎12 𝑎13 … … … . 𝑎1𝑛 𝑥 𝑏1
1
𝑎21 𝑎22 𝑎23 … … … . 𝑎2𝑛 𝑥 𝑏2
2
𝑎31 𝑎32 𝑎33 … … … . 𝑎3𝑛 𝑥3
. . . . . = 𝑏.3
. . . . . .
[𝑎𝑚1 𝑎𝑚2 𝑎𝑚3 … … … . 𝑎𝑚𝑛 ] [ 𝑥𝑛 ] [𝑏𝑚 ]

𝑖. 𝑒. AX = B

Where,
𝑎11 𝑎12 𝑎13 … … … . 𝑎1𝑛 𝑥1 𝑏1
𝑎21 𝑎22 𝑎23 … … … . 𝑎2𝑛 𝑥2 𝑏2
𝑎31 𝑎32 𝑎33 … … … . 𝑎3𝑛 𝑥
A = . . . . , X = .3 , B = 𝑏.3
. . . . . .
[𝑎𝑚1 𝑎𝑚2 𝑎𝑚3 … … … . 𝑎𝑚𝑛 ] [𝑥𝑛 ] [𝑏𝑚 ]

A is Coefficient matrix
X is Matrix of unknowns (or variables
B is Matrix of constant
 Remark: The system AX = B has a solution i.e. the set of value 𝑥1 , 𝑥2 , 𝑥3 , … … … , 𝑥𝑛 satisfies
simultaneously all m linear equations.
 AUGMENTED MATRIX [𝑨 ⋮ 𝑩]:
If 𝐴𝑋 = 𝐵 is a system of m equations in n unknowns then matrix written as [𝐴 ⋮ 𝐵] is
called as augmented matrix.
𝑎11 𝑎12 𝑎13 … … … . 𝑎1𝑛 ⋮ 𝑏1
𝑎21 𝑎22 𝑎23 … … … . 𝑎2𝑛 ⋮ 𝑏2
𝑎 𝑎 𝑎 … … … . 𝑎3𝑛 ⋮ 𝑏3
Hence, [𝐴 ⋮ 𝐵] = 31 32 33
. . . . ⋮ .
. . . . ⋮ .
[𝑎𝑚1 𝑎𝑚2 𝑎𝑚3 … … … . 𝑎𝑚𝑛 ⋮ 𝑏𝑚 ]
The study of mathematics begins in minuteness but ends in magnificence - Charles caleb colton
PROF. MORE G.R. LINEAR ALGEBRA S.N.D. COE & RC, YEOL A

 CONSISTENT: A system of equation AX = B has a solution then system is consistent.

 INCONSISTENT: A system of equation AX = B has no solution then system is Inconsistent.
 NON-HOMOGENOUS SYSTEM: A system of equation is of the form AX = B,
If B is a non-zero matrix, then the system is said to be non-homogenous system of
equation.
 HOMOGENOUS SYSTEM: A system of equation is of the form AX = B,
If B is a zero (null) matrix, then the system is said to be homogenous system of equation.

 Note: Homogenous system is always Consistent.

(𝒙𝟏 = 𝒙𝟐 = 𝒙𝟑 = ⋯ … … = 𝒙𝒏 = 𝟎)
 PROCEDURE TO SOLVE NON-HOMOGENOUS SYSTEM OF EQUATION
(AX = B):
Consider, a system of ‘m’ equations in ‘n’ unknowns given by AX = B.
Let, 𝑚 = total number of equations, 𝑛 = total number of unknowns.
STEP I: Consider, given system of equations
𝑎11 𝑥1 + 𝑎12 𝑥2 + 𝑎13 𝑥3 + ⋯ … … + 𝑎1𝑛 𝑥𝑛 = 𝑏1
𝑎21 𝑥1 + 𝑎22 𝑥2 + 𝑎23 𝑥3 + ⋯ … … + 𝑎2𝑛 𝑥𝑛 = 𝑏2
𝑎31 𝑥1 + 𝑎32 𝑥2 + 𝑎33 𝑥3 + ⋯ … … + 𝑎3𝑛 𝑥𝑛 = 𝑏3
.
.
𝑎𝑚1 𝑥1 + 𝑎𝑚2 𝑥2 + 𝑎𝑚3 𝑥3 + ⋯ … … + 𝑎𝑚𝑛 𝑥𝑛 = 𝑏𝑚
STEP II: Write given system of linear equations in matrix form as
𝐀𝐗 = 𝐁
Where,
A is coefficient matrix, X is matrix of unknown, B is matrix of constant
STEP III: Write augmented matrix [A ⋮ B]

STEP IV: Reduce augmented matrix into echelon form

STEP V: Write 𝑅𝑎𝑛𝑘 (A) & 𝑅𝑎𝑛𝑘 (A ⋮ B)

STEP VII: Observe that, whether 𝑅𝑎𝑛𝑘 (A) = 𝑅𝑎𝑛𝑘 (A ⋮ B) = 𝑟 or 𝑅𝑎𝑛𝑘 (A) ≠ 𝑅𝑎𝑛𝑘 (A ⋮ B)
STEP VIII: If 𝑅𝑎𝑛𝑘 (A) ≠ 𝑅𝑎𝑛𝑘 (A ⋮ B) implies that system has no solution → process stop.
If 𝑅𝑎𝑛𝑘 (A) = 𝑅𝑎𝑛𝑘 (A ⋮ B) = 𝑟 then system has either unique solution or infinite
Solution.

The study of mathematics begins in minuteness but ends in magnificence - Charles caleb colton
PROF. MORE G.R. LINEAR ALGEBRA S.N.D. COE & RC, YEOL A

1. Unique solution: If 𝑅𝑎𝑛𝑘 (A) = 𝑅𝑎𝑛𝑘 (A ⋮ B) = 𝑟 = 𝑛(𝑛𝑜. 𝑜𝑓 𝑢𝑛𝑘𝑛𝑜𝑤𝑛𝑠) implies that

system has unique solution.
(Find unique solution by writing echelon form of matrix [𝐴 ⋮ 𝐵] into system of equations
and find the values of 𝑥1 , 𝑥2 , 𝑥3 , … . , 𝑥𝑛 )
2. Infinite solution: If 𝑅𝑎𝑛𝑘 (A) = 𝑅𝑎𝑛𝑘 (A ⋮ B) = 𝑟 < 𝑛(𝑛𝑜. 𝑜𝑓 𝑢𝑛𝑘𝑛𝑜𝑤𝑛𝑠) implies that
system has infinite solution.
(Find infinite solution by writing echelon form of matrix [𝐴 ⋮ 𝐵] into system of equations
and take (𝑛 − 𝑟) unknowns as some parameter (𝑡, 𝑠, 𝑝, 𝑞, … )and find solution)

CHART
NON-HOMOGENOUS SYSTEM OF EQUATIONS
Write given system of linear equations in matrix form as
AX = B (Non-homogenous system of equations)
Write augmented matrix [A ⋮ B]
Reduce augmented matrix [A ⋮ B] into row echelon form
and write 𝑅𝑎𝑛𝑘 (A), 𝑅𝑎𝑛𝑘 (A ⋮ B)

If 𝑅𝑎𝑛𝑘 (A) = 𝑅𝑎𝑛𝑘 (A ⋮ B) = 𝑟 If 𝑅𝑎𝑛𝑘 (A) ≠ 𝑅𝑎𝑛𝑘 (A ⋮ B)

then System is consistent then System is Inconsistent

No Solution
𝑟 = 𝑛 (𝑛𝑜. 𝑜𝑓 𝑢𝑛𝑘𝑛𝑜𝑤𝑛𝑠) 𝑟 < 𝑛 (𝑛𝑜. 𝑜𝑓 𝑢𝑛𝑘𝑛𝑜𝑤𝑛𝑠)
Unique Solution Infinite Solution

 Note: For non-homogenous system AX = B

In special case, No. of equation = 3
No. of unknown = 3
1. |A| ≠ 0 implies that system is consistent and has unique solution
2. |A| = 0 implies that system is either inconsistent or having
infinite number of solution

 PROCEDURE TO SOLVE HOMOGENOUS SYSTEM OF EQUATION

(AX = B, B is null matrix):
Consider, a system of ‘m’ equations in ‘n’ unknowns given by
AX = B, B is null matrix
Let, 𝑚 = total number of equations, 𝑛 = total number of unknowns.

The study of mathematics begins in minuteness but ends in magnificence - Charles caleb colton
PROF. MORE G.R. LINEAR ALGEBRA S.N.D. COE & RC, YEOL A

CHART
HOMOGENOUS SYSTEM OF EQUATIONS
Write given system of linear equations in matrix form as
AX = B (Homogenous system of equations), B is null matrix

Write augmented matrix [A ⋮ B]

Convert augmented matrix [A ⋮ B] into row echelon form

and write 𝑅𝑎𝑛𝑘 (A), 𝑅𝑎𝑛𝑘 (A ⋮ B)

If 𝑅𝑎𝑛𝑘 (A) = 𝑅𝑎𝑛𝑘(A ⋮ B) = 𝑟 = 𝑛 If 𝑅𝑎𝑛𝑘 (A) = 𝑅𝑎𝑛𝑘 (A ⋮ B) = 𝑟 < 𝑛

then System has trivial solution then System has infinite solution
(𝑖. 𝑒. 𝑥1 = 𝑥2 = 𝑥3 = ⋯ . = 𝑥𝑛 = 0)
 Note: For homogenous system AX = B, B 𝑖𝑠 null matrix
In special case, No. of equation = 3
No. of unknown = 3
1. |A| ≠ 0 implies that system has trivial solution 𝑥 = 𝑦 = 𝑧 = 0
2. |A| = 0 implies that system has infinite number of solution.
 LINEARLY INDEPENDENT & DEPENDENT :
Let, X1 , X 2 , X 3 , … … X n are vector’s
Consider the equation,
C1 X1 + C2 X 2 + ⋯ … … + Cn X n = 0
Where, 𝐶1 , 𝐶2 , 𝐶3 , … … 𝐶𝑛 are constant.
(i) If C1 = C2 = C3 = … … = Cn = 0 then the vectors X1 , X 2 , X 3 , … … X n are said to
be Linearly Independent.
(ii) If C1 , C2 , C3 , … … Cn not all zero then the vectors are said to be Linearly Dependent.

CHART
LINEARLY DEPENDENT/ LINEARLY INDEPENDENT
Let, X1 , X 2 , X 3 , … … X n are vector’s

Consider the equation, C1 X1 + C2 X 2 + ⋯ … … + Cn X n = 0

C1 , C2 , C3 , … … Cn not all zero C1 = C2 = C3 = ⋯ . = Cn = 0

Linearly Dependent Linearly Independent

Find relation between X1 , X2 , X3 , … … Xn . Stop

The study of mathematics begins in minuteness but ends in magnificence - Charles caleb colton
PROF. MORE G.R. LINEAR ALGEBRA S.N.D. COE & RC, YEOL A

 LINEAR TRANSFORMATION:
The equations of linear transformation from (𝑥1 , 𝑥2 ) variables to (𝑦1 , 𝑦2 ) variables is
given by 𝑦1 = 𝑎11 𝑥1 + 𝑎12 𝑥2 , 𝑦2 = 𝑎21 𝑥1 + 𝑎22 𝑥2
𝑦1 𝑎11 𝑎12 𝑥1
[𝑦 ] = [𝑎 𝑎22 ] [𝑥2 ]
2 21
𝑖. 𝑒. Y = AX
gives linear transformation from (𝑥1 , 𝑥2 ) variables to (𝑦1 , 𝑦2 ) variables.
i.e. the transformation of the vector X to the vector Y.

 Remark: In Linear transformation Y = AX,

(i) If A is singular (i.e. |𝐀| = 𝟎) then the transformation is Singular.
(ii) If A is Non-singular (i.e. . |𝐀| ≠ 𝟎) then the transformation is Non- Singular.
(iii) If A is Orthogonal (i.e. . |𝐀| = ±𝟏 𝒐𝒓 𝐀𝐀𝐓 = 𝐀𝐓 𝐀 = 𝐈) then the transformation is
Orthogonal.

 Composite Transformation:
If transformation from (𝑥1 , 𝑥2 , 𝑥3 , … … , 𝑥𝑛 ) to (𝑦1 , 𝑦2 , 𝑦3 , … … , 𝑦𝑛 ) is given by Y = AX
& another transformation (𝑦1 , 𝑦2 , 𝑦3 , … … , 𝑦𝑛 ) to (𝑧1 , 𝑧2 , 𝑧3 , … … , 𝑧3 ) is given by Z = BY
then the transformation from (𝑥1 , 𝑥2 , 𝑥3 , … … , 𝑥𝑛 ) to (𝑧1 , 𝑧2 , 𝑧3 , … … , 𝑧3 ) is given by
Z = BY
Z = B(AX)
Z = (BA)X
Called as ‘Composite Transformation’.

 Inverse Transformation: for a non-singular transformation Y = AX, Since ′A′ is non-

singular (𝑖. 𝑒. |A| ≠ 0). A−1 exists carries the vector Y back into the vector X as
X = A−1 Y

 Orthogonal Matrix: A square matrix ′A′ is said to be orthogonal if

AAT = AT A = I

 Note:
1. If A is orthogonal then A−1 = AT .
2. Matrix A is orthogonal iff |A| = ±1

The study of mathematics begins in minuteness but ends in magnificence - Charles caleb colton