University of Palestine

Gaza Strip

Civil Engineering College

Numerical analysis
CIVL 3309
Formerly CVL 3308

Dr. Suhail Lubbad

 So far we have dealt with a one single equation in one variable

f(x)=0

We were looking for the value(s) of x that satisfies the above equation:
Those values were called the roots
In this part we will be dealing with a set of equations in
n-unknowns or variables

f 1 . x1, x 2, x3, ⋯, xn /=0

f 2 . x1, x 2, x 3, ⋯, x n /=0
f 3 . x1, x 2, x3, ⋯, x n /=0
⋮
f n .x 1, x2, x 3, ⋯, xn /=0

The goal is to find the values of

x1, x 2, x 3, ⋯, x n that satisfy the set of equation

simultaneously that is, at the same time

 Herein this part we deal with
a set of LINEAR algebraic equations ≡ A system of linear algebraic equations

The general form of such systems is:

a 11 x 1 ,a 12 x 2,a 13 x 3 ,⋯,a 1n x n=b1

a 21 x 1,a 22 x 2 ,a 23 x 3 ,⋯,a 2n x n =b2
a 31 x 1,a 32 x 2,a 33 x 3,⋯,a 3n x n =b3
⋮
a n1 x 1 ,a n2 x 2,a n3 x 3,⋯,a nn x n =bn

a’s are constant coefficients, b’s are constants

n is the number of equations
 What is a linear algebraic equation?

In 2 variables: y= ax + b
A straight line (thus the name)

In 3 variables: a1x + a2y + a3z = b

A plane

In n variables: a1x1+ a2x2 + a3x3 … anxn = b

Hard to view or imagine (but it is a plane in nD space)

In all the above, LINEARITY means

Each variable appears to the first power only
“no other exponents, including negatives or fractions, are allowed”
 For the cases of 2 or 3 equations, one can sort things out by simple
calculations and eliminations technique
As you have done in Schools

When the number of equations is 4 or more, calculation are long, time

consuming and boring

Computers must be utilized

Our Goal in Numerical analysis
Matrix notation to represent and manipulate linear algebraic equations.

HOW?

What is a matrix?
Properties of matrices
Special Matrices
And
What are the defined mathematical operations on matrices?
(refreshing students with Linear algebra)

Our today's lecture

A matrix is an object of ordered entries in a rectangular array

0
Column

[ ]
a11 a12 a13 ⋯ a 1m

A = [ A ]= ⋮
a21 a 22 a23 ⋯ a 2m -
row
⋮
an−1,1 an−1, 2 an−1, 3 ⋯ a n−1, m
an , 1 an2 an3 ⋯ a nm

A ij= the Element in i th row and j th column =aij

The size of A is number of rows BY number of columns
SIZE (A) ≡ n BY m or n×m
 Row matrix: Column matrix:
1 row but many columns many rows but 1 column

[]
C11
C21
B= [ B ] = [ b11 b12 ⋯ b1m ] [ ]
C= C =
⋯
Size(B) =1×m
Cn1
Size(C) =n×1

Square matrices

Size =n×n , that is number of rows equals number of columns

The smallest 1×1 matrix [r] , where r is a number

 Special SQUARE Matrices

A symmetric matrix is one where aij = aji for all i’s and j’s
Notice symmetry around the main diagonal

[ ]
5 1 2
M= [ M ] = 1 3 7
2 7 8

[ ]
5 1 2
M= [ M ] = 1 3 7
2 7 8
An upper triangular matrix: A lower triangular matrix:
Elements below the main Elements above the main
diagonal are zero diagonal are zero

[ ] [ ]
a 11 a 12 a13 a14 a 11 0 0 0
0 a 22 a23 a 24 a 21 a 22 0 0
M= [ M ] = M= [ M ] =
0 0 a33 a34 a 31 a 32 a33 0
0 0 0 a 44 a 41 a 42 a 43 a 44
 A diagonal matrix square matrix
Elements off the main diagonal are equal to zero

[ ]
a11 0 0 0
0 a 22 0 0
M= [ M ] =
0 0 a33 0
0 0 0 a 44

The identity matrix is a diagonal matrix

whose elements on the main diagonal are 1s

[ ]
1 0 0 0
0 1 0 0
M= [ M ] =
0 0 1 0
0 0 0 1
 A banded matrix
has all elements equal to zero, with the exception of a band centered on
the main diagonal

[ ]
a 11 a 12 0 0
a 21 a 22 a 23 0
M= [ M ] =
0 a 32 a33 a34
0 0 a 43 a 44

tridiagonal matrix
Matrix Operations

Addition add corresponding terms in each matrix

Commutative
Associative

Subtraction subtract corresponding terms in each matrix

Commutative
Associative

Multiplication of a matrix by a scaler (c)

Multiply each element by c
Matrix Multiplication

C=[A][B]

k=n
C ij= ∑ a ik bkj
k =1
Multiplication is NOT commutative
[ A ][ B]≠[ B][A ]

BUT associative,
.[ A][ B]/[C]=[A ].[B][C]/

And distributive
[ A ].[ B],[C ]/=[ A ][ B],[A ][C ]
.[ A ],[B ]/[C]=[ A ][C],[ B][C ]
 If a matrix [A] is square and nonsingular,
there is another matrix [A]−1, called the inverse of [A]
Such that
−1 −1
[A][A ] =[A] [A ]=[I]
Transpose of a matrix:
ROWS become Columns

[ ]
5 1 9
F= [ F ] = 6 3 7
2 4 8

[ ]
5 6 2
Transpose of F ≡ F =
T
1 3 4
9 7 8
tr [A]:
the trace of a matrix is the sum of the elements on its principal
diagonal.

n
tr [A]=∑a ii
i=1

=a 11,a 22 ,a 33,...,a nn
 Augmented matrix
addition of a column (or columns) to the original matrix.

Example: [M] a matrix of coefficients:

[ ]
5 1 2
M= [ M ] = 1 3 7
2 7 8

AUGMENTATION

[ ∣] [ ∣ ]
5 1 2 3 5 1 2 1 0 0
M1=[ M1 ] = 1 3 7 3 M2=[ M 2 ] = 1 3 7 0 1 0
2 7 8 3 2 7 8 0 0 1
Representing a system of linear algebraic equations by MATRICES
a 11 x 1,a12 x 2 ,a13 x 3 ,⋯,a1n xn =b1
a 21 x 1,a 22 x2 ,a 23 x 3,⋯,a 2n x n=b2

a m1 x 1,am2 x 2 ,am3 x 3,⋯,amn x n =b m

[A][X]=[B]
where A is the matrix of cofactors in equations,

[ ] [] []
B is the column of constants, a11 a12 a13 ⋯ a 1n b1 x1
X is the column of unknowns A= a21 a22 a 23 ⋯ a 2n b2 x2
B= X=
⋮ ⋮ ⋮ ⋮
am1 am2 am3 ⋯ a mn bm xm
Which can be solved using the inverse as follows

−1 −1
[A] [A ][X]=[A] [B]
Giving
−1
[X]=[ A] [B]
A linear algebraic equations can be represented by MATRICES to be
solved by ELIMINATION,
where A is the matrix of coefficients, and B is the matrix of constants

a11 x 1,a 12 x 2 ,a 13 x3 ,⋯,a1n x n=b1

a 21 x 1,a 22 x 2,a 23 x3 ,⋯,a 2n x n =b 2

am1 x 1,a m2 x 2,a m3 x3 ,⋯,amn xn =bm

0
[ A∣B ]
0

[ ]
a 11 a12 a13 ⋯ a1n b1
a 21 a 22 a 23 ⋯ a 2n b 2

a m1 a m2 am3 ⋯ a mn bm
 AFTER THEORETICAL BACKGROUND

WE WILL EMPLOY COMPUTERS

TO WRITE CODES

MUCH IS ALREADY DONE IN MATLAB FOR YOU

YOU NEED TO LEARN IT AND USE IT

PARTS YOU MUST HAVE ALREADY DONE