Ⅱ-Matrices

Throught this chapter 𝐾 is 𝑅 or ℂ ,𝑚 and 𝑛 are integers in ℕ∗ .

Ⅱ-1-Matrix of linear map

Let 𝐸, 𝐹 be 𝐾vector spases of dimensions 𝑛 and 𝑚 respectivly and let 𝐵 =
{𝑒1 , 𝑒2 , … , 𝑒𝑛 } be basis over 𝐸 and 𝐵′ = {𝑒1′ , 𝑒2′ , … , 𝑒𝑚
′
} is basis over 𝐹.

𝑇 is the linear map from 𝐸 to 𝐹, where the vectors 𝑇(𝑒1 ), 𝑇(𝑒2 ), … , 𝑇(𝑒𝑛 ) are vectors
in 𝐹 and 𝐵′ is basis over 𝐹 (𝑓𝑜𝑟 𝑗 = 1, … , 𝑛)

𝑇(𝑒𝑗 ) = 𝜆1𝑗 𝑒1′ + 𝜆2𝑗 𝑒2′ + ⋯ + 𝜆𝑚𝑗 𝑒𝑚

′

the table

𝑇(𝑒1 ) 𝑇(𝑒2 ) … 𝑇(𝑒𝑛 )

… 𝑒1′
𝜆11 𝜆12 𝜆1𝑛
𝑒2′
𝜆21 𝜆22 … 𝜆2𝑛
⋮
⋮ ⋮ ⋮ ⋮
𝑒′
( 𝜆𝑚1 𝜆𝑚2 … 𝜆𝑚𝑛 ) 𝑚

Is called matrix of linear map 𝑇 related to the basis 𝐵 and 𝐵′ is noted mostly
𝑀𝑎𝑡𝐵𝐵′ (𝑇)

Remarks :

1. If 𝐸 = 𝐹, then 𝑀𝑎𝑡𝐵 (𝑇).

2. The matrix 𝑀𝑎𝑡𝐵𝐵′ (𝑇) depends on the basis 𝐵 vectors number and the basis 𝐵′ .

Example :

Let 𝑇 be the linear map of 𝑅3 to 𝑅2 defined as follows

𝑇: 𝑅3 → 𝑅2
(𝑥, 𝑦, 𝑧) → (2𝑥 − 𝑦, 𝑥 + 𝑦 + 𝑧)

Let the usual basis of 𝑅3 ∶ 𝐵 = {(1,0,0), (0,1,0), (0,0,1)} and 𝐵′ = {(1.0), (0,1)} the usual
basis of 𝑅2 . Then

𝑇((1,0,0)) = (2,1) = 2 × (1.0) + (0,1).

𝑇((0,1,0)) = (−1,1) = −1 × (1.0) + (0,1).

𝑇((0,0,1)) = (0,1) = 0 × (1.0) + (0,1).

The matrix of 𝑇
 2 −1 0
𝑀𝑎𝑡𝐵𝐵′ (𝑇) = ( )
1 1 1
2. let 𝑔 be a linear map over the 𝑅3 [𝑋] into 𝑅3 [𝑋] . where 𝑔 : 𝑝 → 𝑋 2 𝑝′′ + 𝑝(1) with
the useal basis of 𝑅3 [𝑋] is given 𝐵 = {1, 𝑋, 𝑋 2 , 𝑋 3 }

𝑔(1) = 1, 𝑔(𝑋) = 1, 𝑔(𝑋 2 ) = 2𝑋 2 + 1 and 𝑔(𝑋 3 ) = 6𝑋 3 + 1.

then

1 1 0 1
𝑀𝑎𝑡𝐵 (𝑔) = (0 0 0 0)
0 0 2 0
0 0 0 6
Ⅱ-2-Matrix
Definition : we call a matrix of type (𝒎, 𝒏) (or dimensions × 𝒏 ) over 𝑲 field any numbers of
the form
𝑎11 𝑎12 … 𝑎1𝑛
𝑎21 𝑎22 … 𝑎2𝑛
𝐴=( ⋮ ⋮ ⋮ )
𝑎𝑚1 𝑎𝑚2 … 𝑎𝑚𝑛

𝐴 = (𝑎𝑖𝑗 ) ( where 1 ≤ 𝑖 ≤ 𝑚 , 1 ≤ 𝑗 ≤ 𝑛) is a called a rectangular matrix where 𝒎 the

rows number and 𝒏 the columns number. The numbers 𝑎11 , 𝑎12 , … . 𝑎𝑚𝑛 are knowen as
elements of the matrix 𝐴, In any matrix the elment 𝑎𝑖𝑗 the first suffix 𝒊 indecates that the
element stands in the 𝑖 𝑡ℎ row and the second suffix 𝒋 indecates that the element stands in
the 𝑗𝑡ℎ column.

Notation

 The set of matrices over 𝐾 of type (𝒎, 𝒏) denoted by 𝑴𝒎,𝒏 (𝑲), if 𝑚 = 𝑛 the set
denoted by 𝑴𝒏 (𝑲).
 If 𝑚 = 𝑛 the matrix 𝐴 = [𝑎𝑖𝑗 ] is a called square matrix
𝒂𝟏𝟏
𝒂
 If 𝑛 = 1 then 𝐴 is called a column matrix (column vector) : 𝑨 = ( 𝟐𝟏 )
⋮
𝒂𝒎𝟏
 If 𝑚 = 1 then 𝐴 is called a row matrix (row vector) : 𝑨 = (𝒂𝟏𝟏 𝒂𝟏𝟐 … 𝒂𝟏𝒏 )
 The Zero matrix of the type (𝒎, 𝒏) is the matrix where all 𝑎𝑖𝑗 elements are equal to
0 it can be denoted by 𝟎𝒎×𝒏
𝟎 𝟎 𝟎
Example : 𝟎𝟑×𝟑 = (𝟎 𝟎 𝟎).
𝟎 𝟎 𝟎
Ⅱ-3-Operation on Matrices

Ⅱ-3-1. Equality of two matrices :

Two matrices 𝐴 and 𝐵 are said to be equal if

1. They are of same order (type)

2. Their corresponding elements are equal.
That is if 𝐴 = (𝑎𝑖𝑗 )𝑚×𝑛 and 𝐵 = (𝑏𝑖𝑗 )𝑚×𝑛 then 𝑎𝑖𝑗 = 𝑏𝑖𝑗 for all 𝑖and 𝑗.

Ⅱ-3-2. Scalar multiplication of matrix:

Let 𝜆 be a scalar then scalar product of matrix 𝐴 = (𝑎𝑖𝑗 )𝑚×𝑛 given denoted by 𝜆𝐴

𝜆𝑎11 𝜆𝑎12 … 𝜆𝑎1𝑛

𝜆𝐴 = ( ⋮ ⋮ ⋱ ⋮ )
𝜆𝑎𝑚1 𝜆𝑎𝑚2 … 𝜆𝑎𝑚𝑛

Ⅱ-3-3. Addition of two matrices :

Let 𝐴 = (𝑎𝑖𝑗 )𝑚×𝑛 and 𝐵 = (𝑏𝑖𝑗 )𝑚×𝑛 are two matrices with same order then the sum 𝐴 + 𝐵
is given by 𝐴 + 𝐵 = (𝑎𝑖𝑗 )𝑚×𝑛 + (𝑏𝑖𝑗 )𝑚×𝑛 = (𝑎𝑖𝑗 + 𝑏𝑖𝑗 )𝑚×𝑛

Ⅱ-3-4. Multiplication of two matrices :

Two matrices are said to be confirmable for product if 𝐴𝐵 if number of columns in 𝐴 is equal
to the number of rows in matrix 𝐵. Let 𝐶 = 𝐴𝐵 be the matrix product of the order 𝑚 × 𝑟
where

𝐴 = (𝑎𝑖𝑗 )𝑚×𝑛 and 𝐵 = (𝑏𝑖𝑗 )𝑛×𝑟

𝑐𝑖𝑗 = ∑𝑛𝑘=1 𝑎𝑖𝑘 𝑏𝑘𝑗 = 𝑎𝑖1 𝑏1𝑗 + 𝑎𝑖2 𝑏2𝑗 + … … … + 𝑎𝑖𝑛 𝑏𝑛𝑗

Theorem

Let 𝐸, 𝐹 𝑎𝑛𝑑 𝐺 be vector-spaces. 𝐵1 , 𝐵2 , 𝐵3 are basis of these vector spaces

Let 𝑇 ∈ 𝑳(𝑬, 𝑭) and 𝑆 ∈ 𝑳(𝑭, 𝑮)

𝑀𝑎𝑡𝐵1 𝐵3 (𝑆°𝑇) = 𝑀𝑎𝑡𝐵2𝐵3 (𝑆)𝑀𝑎𝑡𝐵1𝐵2 (𝑇)

Ⅱ-3-5. Transpose:
The transpose of matrix 𝐴 = (𝑎𝑖𝑗 )𝑚×𝑛 writen 𝐴′ (𝐴′ 𝑜𝑟 𝐴𝑡 ) is the matrix obtained by writing
the rows of 𝐴 in order as columns. That

𝐴 = (𝑎𝑗𝑖 )𝑛×𝑚
Example (Ⅱ-3) :
Let 𝐴, 𝐵 :

1 3 −4 1 0 3 1
𝐴 = (0 −5 2 ) ,𝐵 = (0 2 1 4)
2 1 3 0 −3 −1 6

Calculate each case if it’s possible 𝐴 + 𝐵, 𝐴𝐵 , 𝐵𝐴, 𝐵′ , 𝐴′ , 𝐵 + 𝐴𝐵.

𝐴 + 𝐵 is not possible because 𝐴, 𝐵 are not of the same order .

1 3 −4 1 0 3 1
𝐴𝐵 = (0 −5 2 ) × (0 2 1 4)
2 1 3 0 −3 −1 6
𝐴𝐵
1 × 1 + 0 × 3 + 0 × −4 0 × 1 + 3 × 2 + (−3) × −4 1 × 3 + 3 × 1 + (−1) × −4 1 × 1 + 3 × 4 + 6 × −4
= (0 × 1 + 0 × −5 + 0 × 2 0 × 0 + (−5) × 2 + 2 × −3 0 × 3 + (−5) × 1 + (−1) × 2 0 × 0 + 4 × −5 + 6 × 2)
1×2+0×1+0×3 0 × 2 + 2 × 1 + 3 × −3 2 × 3 + 1 × 1 + (−1) × 3 2 × 1 + 1 × 4 + 6 × 3

1 18 10 −11
𝐴𝐵 = (0 −16 −7 −8 )
2 −7 4 24

𝐵𝐴 it’s impossible because 𝐴 𝑎𝑛𝑑 𝐵 are not confirmable.

1 0 0 1 0 2
𝐵 = (0
′ 2 −3), 𝐴′ = ( 3 −5 1)
3 1 −1
−4 2 3
1 4 6
1 0 3 1 1 18 10 −11
𝐵 + 𝐴𝐵 = (0 2 1 4 ) + (0 −16 −7 −8 )
0 −3 −1 6 2 −7 4 24
1+1 0 + 18 10 + 3 1 + (−11)
= (0 + 0 −16 + 2 1 + (−7) −8 + 4 )
2 + 0 −3 + (−7) 4 + (−1) 24 + 6

2 18 13 −10
𝐵 + 𝐴𝐵 = (0 −14 −6 −4 )
2 −10 3 30
Properties of some oprations
 (𝐴′ )′ = 𝐴
 (𝜆𝐴)′ = 𝜆𝐴′
 (𝐴 + 𝐵)′ = 𝐴′ + 𝐵′
 (𝐴𝐵)′ = 𝐵′ 𝐴′
 Matrix addition is commutative 𝐴 + 𝐵 = 𝐵 + 𝐴
 Matrix addition is associative 𝐴 + (𝐵 + 𝐶 ) = (𝐴 + 𝐵) + 𝐶
 The matrix product is distributative with respect to addition that :
 𝐴(𝐵 + 𝐶 ) = 𝐴𝐵 + 𝐴𝐶

(𝐵 + 𝐶 )𝐴 = 𝐵𝐴 + 𝐶𝐴

 The matrix product is associative

A(BC) = (AB)C = ABC

 𝜆(𝐴𝐵) = (𝜆𝐴)𝐵 = 𝐴(𝜆𝐵)

 If 𝐴 𝑎𝑛𝑑 𝐵 are matrices of the same type and 𝐴𝐵 = 𝐵𝐴 we say 𝐴 𝑎𝑛𝑑 𝐵 commut.
In this case the can define (𝐴 + 𝐵)𝑃 the binomial formula for ∀𝑃 ∈ ℕ∗
𝑃
𝑃
(𝐴 + 𝐵 )𝑃 = ∑ ( ) 𝐴𝑃−𝑘 𝐵𝑘
𝑘
𝑘=1

Ⅱ-4-Square Mrtices
As we defined privously the square matrix is the matrix of type 𝑛 × 𝑛

𝑎11 𝑎12 … 𝑎1𝑛

𝑎 𝑎22 … 𝑎2𝑛
𝐴 = ( 21 )
⋮ ⋮ ⋱ ⋮
𝑎𝑛1 𝑎𝑛2 … 𝑎𝑛𝑛

Ⅱ-4-1. Integral power of matrices:

Let 𝐴 be square matrix of the order 𝑛, and 𝑟 be positive integer then we define

𝐴𝑟 = 𝐴 × 𝐴× … … × 𝐴 (𝑟 𝑡𝑖𝑚𝑒 𝑚𝑢𝑙𝑡𝑖𝑝𝑙𝑖𝑐𝑎𝑡𝑖𝑜𝑛)

𝐴2 = 𝐴 × 𝐴

𝐴3 = 𝐴 × 𝐴 × 𝐴

 The power of matrix are commutative that is

𝐴𝑟 𝐴𝑠 = 𝐴𝑟+𝑠 = 𝐴𝑠 𝐴𝑟

Ⅱ-4-2. Special Types of square matrices :

Definition :(Diagonal matrix)

A square matrix is 𝐴 = (𝑎𝑖𝑗 )𝑛×𝑛 is called a diagonal matrix if each of its non-diagonal
element is zero (the diagonal element the elements of the indices 𝑖 = 𝑗 ). Example : 𝐴 =
1 0 0
(0 −9 0)
0 0 3

Definition :(Identity matrix)

A diagonal matrix whose diagonal elements are equal to 1 denoted by 𝑰𝑛

 1 0 0
Example : 𝑰𝟑 = (0 1 0)
0 0 1

Definition :(Upper Triangular matrix)

A square matrix said to be a Upper tringular matrix if 𝑎𝑖𝑗 = 0 for 𝑖 > 𝑗.

5 7 4
Example : 𝐴 = (0 2 0)
0 0 3
Definition :(Lower Triangular matrix)

A square matrix said to be a Lower tringular matrix if 𝑎𝑖𝑗 = 0 for 𝑖 < 𝑗.

−1 0 0
Example : 𝐴 = ( 4 2 0)
5 8 3

Definition :(Symmetric matrix)

A square matrix is 𝐴 = (𝑎𝑖𝑗 ) is said to be a symmetric if 𝐴 = 𝐴′

𝑛×𝑛

8 −2 7
Example : 𝐴 = (−2 −9 3)
7 3 5

𝐴 is a symmetric matrix.

Definition :(Skew- Symmetric matrix)

A square matrix is 𝐴 = (𝑎𝑖𝑗 )𝑛×𝑛 is said to be a skew- symmetric if the elements 𝑎𝑖𝑗 of the
Uper tringular part are equal in lower tringular part −𝑎𝑖𝑗 . Example :

8 −2 7
(
𝐵= 2 −9 3)
−7 −3 5

𝐵 is skew- symmetric matrix.

Trace of matrix
The trace of square matrix 𝐴 is the sum of the diagonal entries of 𝐴 denoted 𝑻𝒓(𝐴).

Example

trace of 𝐵 from the previous example is 4 ∶ 𝑻𝒓(𝐵) = 𝟖 + (−𝟗) + 𝟓 = 𝟒.

Some definitions
 A matrix 𝐴 such that 𝐴2 = 𝐴 is called Idenpotant.
  If 𝐴 is a matrix and 𝑟 is the least positive integer such that 𝐴𝑟+1 = 𝐴, Then 𝐴 is called
periodic of period 𝑟.
 If 𝐴 matrix for which 𝐴𝑟 = 𝟎 the zero matrix. 𝐴 is called nilpotent.If 𝑟 is the least
positive integer for which this is true. 𝐴 is said to be nilpotent of order .

Ⅱ-4-3. Inverse of matrix and Elementary row Operations

Ⅱ-4-3.a. Determinant
Definition : Let 𝐴 = (𝑎𝑖𝑗 ) be square matrix of order , then the number |𝐴| called
𝑛×𝑛
determinant of the matrix 𝐴.

1. Determinant of 𝟐 × 𝟐 :
𝑎 𝑏 𝑎 𝑏|
Let 𝐴 = ( ) then |𝐴| = | = 𝑎𝑑 − 𝑐𝑏
𝑐 𝑑 𝑐 𝑑
2. Determinant of 𝟑 × 𝟑:
𝑎11 𝑎12 𝑎13
Let 𝐵 = (𝑎21 𝑎22 𝑎23 )
𝑎31 𝑎32 𝑎33
𝑎11 𝑎12 𝑎13
𝑎22 𝑎23 𝑎21 𝑎23 𝑎21 𝑎22
Then 𝐵 = |𝑎21 𝑎22 𝑎23 | = 𝑎11 |𝑎
| | | − 𝑎12 | | + 𝑎13 |𝑎31 𝑎32 |
32 𝑎33 𝑎31 𝑎33
𝑎31 𝑎32 𝑎33
|𝐵| = 𝑎11 (𝑎22 𝑎33 − 𝑎23 𝑎32 ) − 𝑎12 (𝑎21 𝑎33 − 𝑎23 𝑎31 ) + 𝑎13 (𝑎21 𝑎32 − 𝑎22 𝑎31 )

Properties of Determinant
 det(𝐴𝐵) = det(𝐴) det(𝐵)
 det(𝐴′ ) = det(𝐴)
 det(𝜆𝐴) = 𝜆𝑛 det(𝐴

Singular and non-singular matrix

Definition Let 𝐴 be matrix of order 𝑛. 𝐴 is called a singular matrix is |𝐴| = 0 and non-
singular otherwise.

Ⅱ-4-3.b. Minor and cofactors

Definition Let 𝐴 = (𝑎𝑖𝑗 )𝑛×𝑛 be square matrix of order 𝑛. Then 𝑀𝑖𝑗 denoted a sub matrix
of 𝐴 with order (𝑛 − 1) × (𝑛 − 1) obtaind by deleting its 𝑖 𝑡ℎ row and 𝑗𝑡ℎ column. The
determinant |𝑀𝑖𝑗 | is called the minor of the element 𝑎𝑖𝑗 of 𝐴.

The cofactor of 𝑎𝑖𝑗 denoted by 𝐴𝑖𝑗 and is equal to (−1)𝑖+𝑗 |𝑀𝑖𝑗 |

Theorem 1 (The laplace Expansion theorem)

Let 𝐴 = (𝑎𝑖𝑗 )𝑛×𝑛 be square matrix of order 𝑛. det(𝐴)is computed by
det(𝐴) = 𝑎𝑖1 𝐶𝑖1 + 𝑎𝑖2 𝐶𝑖2 + ⋯ + 𝑎𝑖𝑛 𝐶𝑖𝑛
or
det(𝐴) = 𝑎1𝑗 𝐶1𝑗 + 𝑎2𝑗 𝐶2𝑗 + ⋯ + 𝑎𝑛𝑗 𝐶𝑛𝑗
Where 𝐶𝑖𝑗= (−1)𝑖+𝑗 |𝑀𝑖𝑗 |
Example
1. compute the determinat of 𝐴 and the cofactor matrrix of 𝐴

1 0 1
𝐴=( 2 −2 4)
−3 −1 −3
With use of the first row

det(𝐴) = 1 |
−2 4 | − 0 | 2 4 | + (−3) | 2 −2|
−1 −3 −3 −3 −3 −1

det(𝐴) = 1((−2 × −3) − (4 × −1)) − 0((2 × −3) − (4 × −3)) + 1((2 × −1) − (−2 × −3))
det(𝐴) = 2 ≠ 0

Or using the first column

det(𝐴) = 1 |−2 4 |− 2| 0 1 | + (−3) | 0 1|

−1 −3 −1 −3 −2 4
det(𝐴) = 2 ≠ 0

𝐴 is non-singular matrix

−2 4 2 4 2 −2
| | −| | | |
−1 −3 3 −3 3 −1 10 −6 −8
0 1 1 1 1 0
𝑐𝑜𝑓𝑎𝑐𝑡𝑜𝑟 (𝐴) = −| | | | −| | = (−1 0 1)
−1 −3 3 −3 3 −1
0 1 1 1 1 0 2 −2 −2
|
( −2 4 | − | | | |
2 4 2 −2 )

2. Let 𝐴 = (𝑎 𝑏 ) the cofactor matrix of 𝐴

𝑐 𝑑
𝑑 −𝑏
𝑐𝑜𝑓𝑎𝑐𝑡𝑜𝑟(𝐴) = ( )
−𝑐 𝑎

Ⅱ-4-3.c. Adjoin Matrix

Definition The transpose of the matrix cofactors of the element 𝑎𝑖𝑗 of 𝐴 denoted by 𝑎𝑑𝑗 𝐴 is
called adjoin of matrix 𝐴.

Example
find the adjoin matrix of the above Example

10 −1 2
( ) ′
𝐴𝑑𝑗 𝐴 = 𝑐𝑜𝑓𝑎𝑐𝑡𝑜𝑟 𝐴 = (−6 0 −2)
−8 1 −2
Ⅱ-4-3.1. Inverse of matrix
Definition If 𝐴 and 𝐵 are two matrices such that 𝐴𝐵 = 𝐵𝐴 = 𝑰 then each is said to be
inverse of the other.

The inverse of 𝐴 is denoted by 𝐴−1

𝐴−1 𝐴 = 𝐴𝐴−1 = 𝑰

Theorem 2 (Existance of the inverse) :

The necessary and sufficient condition for square matrix 𝐴 to have an inverse is that
|𝐴 | ≠ 0
(that 𝐴 is non-singular).

Theorem 3 (uniqueness of the inverse) :

Inverse of a matrix if it exixts is unique

Properties of the inverse

 (𝐴−1 )−1 = 𝐴
1
 (𝜆𝐴)−1 = 𝐴−1 , 𝜆 ∈ 𝑅 ∗
𝜆
 (𝐴 −1 )′
= (𝐴′ )−1
 (𝐴𝐵)−1 = 𝐴−1 𝐵−1

Definition ( linear Group )

We call a linear Group of the order 𝑛 over 𝐾, denoted 𝑮𝑳𝒏 (𝑲), the group of all
invertable elements of 𝑴𝑛 (𝐾).

Theorem 4
1
If 𝐴 is non-singular matrix then 𝐴−1 = det(𝐴) 𝐴𝑑𝑗𝐴

Example
The inverse of the previous 𝐴 is

1
5 − 1
1 1 10 −1 2 2
𝐴−1 = 𝐴𝑑𝑗 𝐴 = (−6 0 −2) = −3 0 −1
det 𝐴 2 1
−8 1 −2
−4 1
( 2 )
 1 3
Let 𝐵 = ( )
2 4
det(𝐵) = (1)(4) − (3)(2) = −2 ≠ 0

𝐵 is non-singular, then

3
1 4 −2
−3 2 )
𝐵−1 = ( )=(
−2 −2 1 1
1 −
2

Theorem 5
Let 𝐸, 𝐹 be 𝐾vector spases of the same dimension 𝑛 with respectivly the basis 𝐵 𝑎𝑛𝑑 𝐵′
Let 𝑇 ∈ 𝑳(𝑬, 𝑭), the linear map 𝑇 is linear is bijective if and only if 𝑴𝒂𝒕𝑩𝑩′ (𝑻) is non-
singular.

.Thereom 6(Similar Matrix)

Let , 𝐵 ∈ 𝑴𝑛 (𝐾), 𝐴 and 𝐵 are said to be a similar matrices if and only if ∃𝑃 ∈ 𝑮𝑳𝒏 (𝑲), 𝐵 =
𝑃𝐴𝑃−1

Ⅱ-4-3.2.Elementary row Operations

Elementary transformations (operations)

Some operations on matrices called as elementary transformations. There are six types of
elementry transformations three of them are row transformation and other three of them
are column transformations there are as follows

1- Interchange of any two rows or columns

(for 𝑖 𝑡ℎ row and 𝑗𝑡ℎ denoted by 𝑅𝑖 ↔ 𝑅𝑗 )
2- Multiplication of elements of any row or column by non-zero number 𝜆.
(for 𝑖 𝑡ℎ row multiplied by number 𝑅𝑖 ↔ 𝜆𝑅𝑖 )
3- Multiplication to elements of any row or column by a scalar 𝜆 and addition of it to
the corespending elements of any other row or column
(for 𝑗𝑡ℎ row multiplied by 𝜆 + to the corresponding elements of 𝑖 𝑡ℎ row denoted by
𝑅𝑖 ↔ 𝑅𝑖 + 𝜆 𝑅𝑗 )

Definition (Equivalent matrix)

A matrix 𝐵 is said to be equivalent to amatrix 𝐴 if 𝐵 can be obtainded from 𝐴 by for forming

finitely many successive elementary transformations on matrix 𝐴.
Denoted 𝐴~𝐵

Example on how to compute the inverse using elementary transformations

The folloing method can be used with any non singulat square matrix

0 11
Let the matrix 𝐴 = ( )
−2 3

First we write the matrix in the form (𝐴|𝑰)

0 11 1 0
( | )
−2 3 0 1

We reduce(𝐴|𝑰) to (𝑰|𝑨−𝟏 ) by forming finitely many successive elemnetary transformations

−2 3 0 1
𝑅1 ↔ 𝑅2 ( | )
0 11 1 0
1 1
𝑅1 → − 𝑅1 3 0 −
2 (1 − 2| 2)
1 1
𝑅2 → 𝑅 0 1
11 0
11 2
3 1
3 1 0 22 − 2
𝑅1 → 𝑅1 + 𝑅2 ( | )
2 0 1 1 0
11
3 1
−
𝑨−𝟏 = (22 2)
1
0
11

this method is of finding the invese is called Gauss Jordan Eliminatin.

Ⅱ-4-4.Rank of matrix
The rank of matrix 𝐴 is the dimension of the subvector space spanned by its columns.

To find the rank of matrix, we transform the matrix into its echelon form using the
elementry transformations.Then ,determine the rank by the non-zero rows.

Theorem

the rank of matrix 𝐴 of 𝑴𝒏 (𝑲) is equal 𝒏 if and only if 𝐴 is non-singular matrix

Example
1 1 0 0 1 0
𝑟𝑎𝑛𝑘 (−2 0 2) = 𝑟𝑎𝑛𝑘 (−2 0 2) 𝐶1 → 𝐶1 − 𝐶2
1 2 3 −1 2 3
 0 1 0
= 𝑟𝑎𝑛𝑘 (−2 0 0) 𝐶3 → 𝐶3 + 𝐶1
−1 2 2
0 1 0 1
= 𝑟𝑎𝑛𝑘 (−2 0 0) 𝑅3 → 𝑅3 − 𝑅2
2
0 2 2
0 1 0
= 𝑟𝑎𝑛𝑘 (−2 0 0) 𝑅3 → 𝑅3 − 2𝑅1
0 0 2
1 0 0
= 𝑟𝑎𝑛𝑘 (0 −2 0) 𝐶1 ↔ 𝐶2
0 2 2

=3
Ⅲ-System of Linear Equations

Ⅲ-1. Solution of the linear system 𝑨𝑿 = 𝒃

We now study how to find the solution of the system of 𝑚 linear equations in 𝑛 unknowns

Consider the system of equations in unknowns 𝑥1 , 𝑥2 , … … … , 𝑥𝑛 as

𝑎11 𝑥1 + 𝑎12 𝑥2 + … +𝑎1𝑛 𝑥𝑛 = 𝑏1

𝑎 𝑥 + 𝑎22 𝑥2 + … +𝑎2𝑛 𝑥𝑛 = 𝑏2
(𝑆) { 21 1
⋮ ⋮ ⋮ ⋮
𝑎𝑚1 𝑥1 + 𝑎𝑚2 𝑥2+ … +𝑎𝑚𝑛 𝑥𝑛 = 𝑏𝑚

Is called system of linear equations with 𝑛 unknowns 𝑥1 , 𝑥2 , … … … , 𝑥𝑛 .

If the constants 𝑏1 , 𝑏2 , … . 𝑏𝑚 are all zero the system is said to be homogeneous type 𝑏𝑗 ∈ 𝐾

The system can be put in the matrix form as 𝐴𝑋 = 𝑏

𝑎11 ⋯ 𝑎1𝑛 𝑥1 𝑏1
Where 𝐴 = ( ⋮ ⋱ ⋮ ) , 𝑋 = ( ⋮ ) ,𝑏 = ( ⋮ )
𝑎𝑚1 ⋯ 𝑎𝑚𝑛 𝑥𝑛 𝑏𝑚

The matrix 𝐴 = (𝑎𝑖𝑗 )𝑛×𝑛 is called coefficient matrix , the column matrix 𝑋 is called the
matrix of the unknowns and 𝑏 is called as column matrix of constants .

Ⅲ-2. Methods of solving system of linear equations

Solving the system (𝑆) means to find the set of solution ( the unknowns column matrix
values)

Given any system of linear equations there are threepossible outcoms:

 The system has a unique solution
 The system has Infinitely many solutions
 The system has no solutions.

Ⅲ-2.1 –Method of Inversesion

Consider the matrix equation 𝐴𝑋 = 𝑏 where det 𝐴 ≠ 0 by multiplying 𝐴−1 ,we have

𝐴−1 (𝐴𝑋) = 𝐴−1 𝑏

𝑋 = 𝐴−1 𝑏

Example
Let (𝑆)
 3𝑥 + 3𝑦 − 2𝑧 =5
(𝑆 ) { 2𝑦 + 7𝑧 =0
𝑥+𝑦−𝑧 =3

3 3 −2 5
(
Where 𝐴 = 0 2 7 and 𝑏 = 0). 𝐴 is non singular det 𝐴 = −2 ≠ 0
) (
1 1 −1 3
9 1 25
− −
1 −9 1 25 2 2 2
𝐴−1 = − ( 7 −1 −21) = 7 1 21
2 −
−2 0 6 2 2 2
( 1 0 −3 )
9 1 25
− − 5
2 2 2 −15
𝑋 = 𝐴−1 𝑏 = 7 1 21 × (0) = ( 14 )
− 3 −4
2 2 2
( 1 0 −3 )
The system has unique solution (𝑥, 𝑦, 𝑧) = (−15,14, −4)

Ⅲ-2.2– Cramer’s rule method

Definition A system (𝑆) is said to be Cramer System if the matrix 𝐴 of the system (𝑆) is non-
singular square matrix.

Theorem( Cramer’s rule)

Let (𝑆) be a Cramer System in which 𝐴 is matrix of (𝑆) .The solution fon unknown
det(𝐴𝑗 )
𝑥1 , 𝑥2 , … … … , 𝑥𝑛 is given by : 𝑥𝑖 =
det(𝐴)
𝑡ℎ
Where 𝐴𝑗 is the matrix 𝐴 whose 𝑗 column has been replaced by the constants in 𝑏

Example the same previous system

By Crammer’s rule the solution is

5 3 −2 3 5 −2 3 3 5
|0 2 7| |0 0 7| |0 2 0|
𝑥= 3 1 −1 = −15, 𝑦= 1 3 −1 = 14, 𝑧= 1 1 3 = −4.
−2 −2 −2

Ⅲ-2.3 –Gaussian Elimination method ( using Elamentary row operation)

The Gaussian Elimination method is more general then the previous ones. The matrix 𝐴 of
the system does not has to be non-singular square matrix, it can be of type (𝑚, 𝑛)
(rectangule)

The Gaussain elimination is based on transforming the matrix of system 𝐴 with the vector
𝑏 into (𝐴|𝑏) to some kind of upper tringular matrix form using the row opretions
 𝑎11 ⋯ 𝑎1𝑛 𝑏1
(𝐴 |𝑏 ) = ( ⋮ ⋱ ⋮ |⋮)
𝑎𝑚1 ⋯ 𝑎𝑚𝑛 𝑏𝑛

To

𝑎11 … … 𝑎1𝑛
… 𝑎′ 2𝑛 𝑏1
0 𝑎′ 22 |
(𝐴′|𝑏′) = ⋮ 0 𝑎′𝑟𝑛 𝑏′2
⋮ | ⋮
⋮ 𝑏′𝑛
( 0 0
0 )

Theorem
Any row operation used on matrix of the system with not change the corresponding
solution of the system

Gaussion Elimination steps

𝟏𝒔𝒕 𝒔𝒕𝒆𝒑 starting from the left , find the frst non-zero column. The first pivot is obtain by
taking the bigesst (with absolut value) element in the first column it’s called Gauss pivot
(suppose it’s 𝑎11 as example)

𝟐𝒏𝒅 𝒔𝒕𝒆𝒑 use row operations on 𝑅1 , 𝑅2 … 𝑅𝑚 to make the entiers below the pivot position
𝑎 𝑎
equal zero by 𝑅2 − 𝑎21 𝑅1 , … , 𝑅𝑚 − 𝑎𝑚1 𝑅1
11 11

𝟑𝒓𝒅 𝒔𝒕𝒆𝒑 Ignoring the row containing the pivot position, repeat step 1 and step 2 with the
remaining rows .

Example

1- The previous system solution by Gauss method

3 3 −2 5
(𝐴|𝑏) = (0 2 7 |0) the pivot is 3 is in the first row
1 1 −1 3
3 3 −2 5
1 ( 0 2 7 |0 )
𝑅3 → 𝑅3 − 𝑅1 −1 4
3 0 0 3 3

3 3 −2 5
(0 2 7|0)
𝑅3 → −3𝑅3
0 0 1 −4

The system (𝑆)

3𝑥 + 3𝑦 − 2𝑧 =5
{ 2𝑦 + 7𝑧 =0
𝑧 = −4
We substitute 𝑧 = −4 in 2nd equation 2𝑦 + 7(−4) = 0 we get 𝑦 = 14, then 𝑥 =
−15 from the 1𝑠𝑡 eqaution.

2- Let the system (𝑆)

𝑥 − 𝑦 + 2𝑧 =4
𝑥 +𝑧 =6
{2𝑥 − 3𝑦 + 5𝑧
=4
3𝑥 + 2𝑦 − 𝑧 = 1
The matrix for the Gaussian Elimination
1 −1 2 4
(𝐴|𝑏) = (1 0 4 |6 )
2 −3 5 4
3 2 −1 1

Use the row operation

𝑅2 → 𝑅2 − 𝑅1 1 −1 2 4
𝑅3 → 𝑅3 − 2𝑅1 (0 1 −1| 2 )
0 −1 1 −4
𝑅4 → 𝑅4 − 3𝑅1
0 5 −7 −11

1 −1 2 4
𝑅3 → 𝑅3 + 𝑅2 (0 1 −1| 2 )
0 0 0 −2
0 5 −7 −11

The system (𝑆) has no solution in the 3𝑟𝑑 𝑟𝑜𝑤 0𝑥 + 0𝑦 + 0𝑧 = −2 which is contraduction