Chapter # 2 Solution of Algebraic and Transcendental Equations

Numerical Analysis
Chapter # 2
Solution of Algebraic and Transcendental Equations
Notes:
(1)
Linear Equations: An equation in which the highest power of variable is One

is called linear equation. e.g.
x+2=0
2x + 10 = 0 etc
(2)
Non-Linear Equation: Equation involving higher powers of a variable is called

Non-Linear Equation.e.g.
x3 + 4x2 + 10x + 3 = 0
x4 + 9x2 + 20 = 0
(3) Transcendental Equation: Equation involving functions of variable

(e.g. Trigonometric, Exponential and logarithmic functions etc) is called
Transcendental Equation.e.g.
Sinx = 6x + 2
ex lnx = 20
e-x Sinx = 0
(4) There are two methods of solving Non-Linear and Transcendental Equations.
Numerical Method.
Analytical Method
Types of Numerical Method
(i) Newton Raphson Method
(ii) Secant Method
(iii) False Position Method
(iv) Bisection Method
Prepared by: Inayat Ullah Lecturer in Mathematics Edwardes College Peshawar
Numerical Analysis
(v) Iterative Process

The process of finding Successive approximation to a quantity is called an iterative
process.
Type No # 1
Newton Raphson method
KEY
.f(x) =
Find f (x) i.e.
(Given)
d
f (x ) and put x = xn in f(x) and f (x) .
dx
Interval or initial point x0 will be given if not, put x = 0, 1, 2, 3, 4, .in f(x).

If f(x) changes its sign for any two close values of x, the interval will be that one
and the small value of the interval will be x0.
Finally put n = 0, 1, 2, 3, 4, in the formula:
x n 1 x n
f ( xn )
to get x1, x2, x3,
f ( x n )
If the two values of x are approximately equal then we stop the iteration, and the
last value of x will be a positive root of f(x).
Q.21/Page # 47
Apply any numerical method to calculate the square root of 3 up to six decimal places.
Ans:
Square root of 3 =
Let x =
Numerical Analysis
Squaring both sides
X2 = 3 x2 3 = 0
Let f(x) = x2 3 f(xn) = xn2 - 3
d
d
f (x ) =
( x2 3) f (x ) = 2x f ( x n ) = 2 xn
dx
dx
Using Newton Raphson Method
f ( xn )
f ( x n )
x n 1 x n
x n 1 x n
xn 3 2 xn xn 3 xn 3
----------> (1)
2 xn
2 xn
2xn
To find x0
As f(x) = x2 3
For x = 0, f(0) = 0 3 = -3 (-ve)
For x = 1, f(1) = 1 3 = -2 (-ve)
For x = 2, f(2) = 22 3 = 4 3 = 1 (+ve)
Since f(x) changes its sign at x = 1 and x = 2
The root lies in the interval (1, 2) x0 = 1
2
For n = 0, equation (1) x1
x0 3 1 3 4
2
2 x0
2
2
2
x 3 22 3 4 3 7
1.75
For n = 1, equation (1) x 2 1
2 x1
2( 2)
4
4
Numerical Analysis
2
x 3 (1.75) 2 3 6.0625
1.732149
For n = 2, equation (1) x3 2
2x2
2(1.75)
3 .5
2
For n = 3, equation (1) x 4
x 3 3 (1.732149) 2 3
1.7320508
2 x3
2 x3
We stop the iteration because the value of x3 and x4 are approximately equal. Hence
the root is x = 1.7320508.
(ANSWER)
Q.16/page # 47
Find the +ve real root of xex = 1, using Newton Raphson method.
Use x0=1.
Ans:
Given that: xex = 1 ==> xex 1 = 0
Let f(x) = xex 1 ==> f(xn) = xnexn 1
d
( xe x 1) = xex + ex.1 0
dx
f (x) =
f (x) = xex + ex
f ( x n ) =exn ( xn + 1)
Using Newton Raphson Method
x n 1
f ( xn )
x n e xn 1
xn
=> x n 1 x n xn
f ( x n )
e ( x n 1)
x n 1
x n e xn ( x n 1) ( x n e xn 1)
e xn ( x n 1)
Numerical Analysis
2
x n 1
x n e xn x n e xn x n e xn 1
e xn ( x n 1)
x n 1
x n e xn 1
e xn ( x n 1)
---------> (1)
For n = 0, equation (1) =>

2
x0 e x0 1
x1 x0
e ( x0 1)
x1
12.e1 1 2.7183 1 3.7183
0.6839
e1 (1 1) 2.7183 2 5.4365
x1 = 0.6839
x1 e x1 1 0.6839 e 0.6839 1 0.4677 1.9816 1 1.9267

x2 x1
0.5776
1.9816 1.6839
3.3368
e ( x1 1) e 0.6839 (0.6839 1)
2
x2 e x2 1 0.5776 e 0.5776 1 0.3336 1.7816 1 1.5944

x 3 x2
0.5672
1.7816 1.5776
2.8109
e ( x2 1) e 0.5776 (0.5776 1)
2
After 3 iterations the values of x2 and x3 coincide. Hence the root is x = 0.5672.
(ANSWER)
TYPE NUMBER 2
SECANT METHOD
Key
The formula for Secant Method is:
Numerical Analysis
x n 1 f ( x n ) x n f ( x n 1 )
f ( x n ) f ( x n 1 )
x n 1
f(x) will be given.

For f(xn) and f(xn-1) put x = xn and x = xn-1 in the given function f(x).
Interval (x0, x1) will be given or individually the values of x0 and x1 will be given,
if not then 1st we find the values of x0 and x1 by putting x = 0, 1, 2, 3, 4,
.. to get the change in sign for f(x).
Finally put n = 1, 2, 3, 4, .. in the formula. The root will be that for
which two values of x coincide.
(ii) To solve xx = 10 by Secant Method.
Let f(x) = xx 10 f(xn) = x n
xn
10 f(xn-1) = xn1
xn 1
10
Using Secant Method

x n 1
x n 1 f ( x n ) x n f ( x n 1 )
--------> (1)
f ( x n ) f ( x n 1 )
as root lies in the interval (2, 3) x0 = 2 and x1 = 3

equation (1)
x
x n 1
x n 1 ( x n n 10) x n ( x n 1
x
( x n n 10) ( x n 1
x n 1
x n 1 ( x n n 10) x n ( x n 1
xn
x n x n 1
x n 1
10)
10)
x n 1
x n n 10 x n 1
x n 1
x n 1
10)
10
10)
x n 1
x n 1
x n 1 ( x n n 10) x n ( x n 1
x n 1 .x n n 10 x n 1 x n .x n 1
x
x n n x n 1
x n 1
x n 1
10 x n
(1)
For n = 1, equation (1)

x
x2
x 0 .x1 1 10 x 0 x1 .x0 0 10 x1
x
x1 1 x 0
x0
2( 3 )3 10( 2 ) 3( 2 ) 2 10( 3 ) 54 20 12 30 52
2.2609
27 4
23
33 2 2
x3
x3
x1 .x 2
x2
10 x1 x 2 .x1 1 10 x 2
x2
x2
x1
x1
3( 2.2609 ) 2.2609 10( 3 ) 2.2609( 3 )3 10( 2.2609 )

( 2.2609 ) 2.2609 3 3
18.9721 30 61.0443 22.6090 49.4632
2.3923
6.3240 27
20.6760
Numerical Analysis

x 2 .x 3
x4
x4
x3
10 x 2 x 3 .x 2
x3
x3
x2
x2
10 x 3
x2
2.2609( 2.3923 ) 2.3923 10( 2.2609 ) 2.3923( 2.2609 ) 2.2609 10( 2.3923 )
( 2.3923 ) 2.3923 ( 2.2609 ) 2.2609
182189 22.6090 15.1290 23.9230 4.4039
2.4546
8.0582 6.3240
1.7342
x5
x5
x 3 .x 4
x4
10 x 3 x 4 .x 3
x4
x4
x3
x3
10 x 4
x3
2.3923( 2.4546 ) 2.4546 2.4546( 2.3923 ) 2.3923 10( 2.3923 ) 10( 2.4546 )
( 2.4546 ) 2.4546 ( 2.3923 ) 2.3923
21.6802 19.7798 23.9230 24.5460 2.5234
2.5126
9.0625 8.0582
1.0043
We stop the iteration because the values of x4 and x5 are approximately equal. Hence
the root is x = 2.5126
(ANSWER)
TYPE NO. 3
METHOD OF FALSE POSITION
OR
REGULA FALSI METHOD
KEY
(1)
Formula for False position is
xi x L
f ( x L )[ x R x L ]
f ( xL ) f ( xR )
Where
xL = Lower value of the interval

xR = Upper value of the interval
(2)
The function f(x) will be given.
(3)
The interval will be given, if not then put x = 0, 1, 2, 3,.in f(x).

If f(x) changes its sign for any two consecutive values of x, the interval will be
that one.
(4)
We put all these values in the formula (1) to get xi.
(5)
Find f(xi), if f(xi) is nearly equal to ZERO stop the iteration process otherwise
proceed by reducing the interval in the following ways:
Numerical Analysis
(6)
F(xL)
F(xR)
F(xi)
Replace xL or xR by xi
Positive
Negative
Positive
Replace xL by xi
Negative
Positive
Positive
Replace xR by xi
Positive
Negative
Negative
Replace xR by xi
Negative
Positive
Negative
Replace xL by xi
We stop the iteration process when the value of f(xi) is nearly equal to zero and
the root will be xi.
Q.14/Page # 46
To solve the question f(x) = sinx 0.5x by using the False position Method.
From the previous question the root lies in the interval (1,2) xL = 1, xR= 2
Now f(xL) = f(1) = sin1 0.5(1) = 0.8414 0.5 = 0.3414
And f(xR) = f(2) = sin2 0.5(2) = 0.9093 1 = -0.0907
Using the formula for Regula Falsi Method
f ( x L )[ x R x L ]
xi x L
f ( x L ) f ( x R ) --------->(1)
xi 1
0.3414[ 2 1]
0.3414
1
1 0.7900 1.79
0.3414 0.0907
0.4321
now f(xi) = f(1.79) = sin(1.79) -0.5(1.79) = 0.9760-0.895=0.081

(Not nearly equal to zero. So we proceed)
We have the following situation
.f(xL)
= 0.3414
(+ve)
.f(xR) = - 0.0907 (-ve)

.f(xi)
= 0.081
(+ve)
xi = xL = 1.79, xR = 2
f(xi) = 0.081 = f(xL) [i.e. we replace xL by xi and f(xL) by f(xi) in (1)]
equation (1)
xi 1.79
0.081[ 2 1.79]
0.0170
1.79
1.79 0.0990 1.8890
0.081 0.0907
0.1717
Numerical Analysis
Now f(xi) = f(1.8890) = sin(1.8890) 0.5(1.8890) = 0.0053
Since the value of f(xi) is nearly equal to Zero the root is x = 1.8890 Answer
TYPE NO. 4
BISECTION METHOD
OR
BOLZANO METHOD
KEY
The value of f(x) will be given.
Find those values of x for which f(x) changes its sign. The root will lie in that
interval.
Find the mid point of the interval and find f(x) for that value of x.
Reduce the interval for which f(x) changes its sign [Already discussed in Type #3]
and go on till the value of f(x) nearly equal to zero. This will be your root.
Q.11/Page # 46
Find a positive real root of f(x) = sinx x/2, with x1 = /2 and x2 = using Bisection
Method. Do six iterations.
Ans: Given function is: f(x) = sinx x/2 -------->(1)
1st we find the interval for which f(x) changes its sign.
Since x1 = /2 and x2 =
Therefore
f(x1) = f(/2) = sin/2 -/2/2 = 1 /4 = 1 0.785 = 0.215 (+ve)
And
f(x2) = f() = sin - /2 = 0 1.571 = -1.571(-ve)
As f(x) changes its sign for x1 = /2 and x2 =

the interval is (/2, ) or (1.571, 3.142)
Now we find the mid point of x1 and x2
Mid point = x =
x1 x 2 1.571 3.142
2.356
2
2
Put this value of x in (1)

.f(x) = f(2.356) = sin(2.356) 2.356/2 = -0.471
(not nearly equal to zero, so we proceed)
Numerical Analysis
For the next step we have:
.f(x1) is +ve
f(x2) is ve
f(x) is ve
x1 = 1.571 and x2= x = 2.356

[we choose x as x2 because f(x2) and f(x) have same sign]
And f(x1) = 0.215 (+ve), f(x2) = f(x) = - 0.471 (-ve)
Now mid point = x =
x1 x 2 1.571 2.356
1.963
2
2
Put this value of x in equation (1)

.f(x) = f(1.963) = sin(1.963) 1.963/2 = -0.058
(not nearly equal to zero, so we proceed)
For the next step, we have:
.f(x1) = 0.215 (+ve), f(x2) = - 0.471 (-ve), f(x) = -0.058 (-ve)
x1 is the same and x2 = x = 1.963
So, mid point = x =
x1 x 2 1.571 1.963
1.767
2
2
Put in equation (1)

.f(x) = f(1.767) = sin(1.767) 1.767/2 = 0.097
(Not nearly equal to zero, so we proceed)
.f(x1) = 0.215 (+ve), f(x2) = - 0.058 (-ve), f(x) = 0.097 (+ve)
x1 = x = 1.767 and x2 = 1.963 (which is the same)
So, Mid point = x =
x1 x2 1.767 1.963
1.865
2
2
Put in equation (1)

.f(x) = f(1.865) = sin(1.865) 1.865/2 = 0.024
(Not nearly equal to zero, so we proceed)
.f(x1) = 0.097 (+ve), f(x2) = - 0.058 (-ve), f(x) = 0.024 (+ve)
x1 = x = 1.865 (Changed)
So , Mid point = x =
and
x2 = 1.963 (Not changed)
x1 x2 1.867 1.963
1.889
2
2
10
Numerical Analysis
11
Put this value of x in equation (1)

.f(1.889) = sin(1.889) 1.889/2 = 0.004
This value of f(x) is nearly equal to Zero. Hence the root is x = 1.889(ANSWER)
CHAPTER # 3
LINEAR EQUATIONS
Consider the system of equations.
2x1 3 x2 + 7x3 = 10
8x1 + 2 x2 + 5x3 = 12
4x1 3x2 9 x3 = 16
All the above equations involve three unknowns i.e. x1, x2, x3 with maximum power
ONE. So this system of equations is called system of linear equations. We can write
it in matrix form as:
2
8
3
2
3
i.e.
7
5
9
x1
10
x 2 12
x3
16
Ax = c
Where A = 8
4
10
c = 12
16
3
2
3
7
5
9
and is called coefficient matrix
which is called column matrix or column vector.

Augmented Matrix is:
2
Ac = 8
4
3
2
3
10
5 12
9 16
The system of linear equations can be solved by two methods.
Direct Method.
Iterative Method.
Types of Direct Method
Crammers Rule
Gauss Elimination Method.
Gauss Jordan Method.
Method of Inversion.
Choleskis Factorization Method.
Numerical Analysis
12
TYPE NO. 1
CRAMMERS RULE
KEY
Write the system of equations in Matrix form i.e. Ax = c

Find |A|. If |A| 0, solution exists and if |A| = 0, then solution does not exit.
Find |A1|, |A2|, |A3|, where |A1|, |A2| and |A3| are the determinants obtained by
replacing the 1st , 2nd and 3rd column of A by c respectively.
The solution is: x1 =
A1
A
, x2 =
A2
A
, x3 =
A3
A
The solution set will be written as: a = (x1, x2, x3) T

Where T stands for Transpose
Q.3/Page # 84
Solve the system of equations by using Crammers Rule
2.16x1 + 3.75x2 + 0.98x3 = 8.68
4.12x1 + 1.67x2 + 7.21x3 = 0.25
3.14x1 + 4.02x2 + 2.16x3 = 9.16
Ans:
The system can be written in matrix form as:
2.16 3.75 0.98 x1
8.68
4.12 1.67 7.21 x 0.25
3.14 4.02 2.16 x 3

9.16
i.e.
Ax = c
1st we find |A|

|A| =
2.16
4.12
3.14
3.75
1.67
4.02
0.98
7.21
2.16
Expanding by R1
|A| =
|A| =
|A| =
|A| =
2.16
1.67
7.21
4.02
2.16
3.75
4.12
7.21
3.14
2.16
0.98
4.12
1.67
3.14
4.02
2.16[ (1.67)(2.16) (4.02)(7.21)]

- 3.75[(4.21)(2.16) (7.21)(3.14)]
+ 0.98[(4.12)(4.02) (3.14)(1.67)]
2.16(3.60 28.98) 3.75(8.90 22.64) + 0.98(16.56 5.24)
- 54.82 + 51.52 + 11.09 = 7.79 0
Solution exists.
Numerical Analysis
Now | A 1 |
8.68
0.25
9.16
3.75
1.67
4.02
13
0.98
7.21
2.16
Expanding by R 1
8.68
1.67
7.21
4.02
2.16
3.75
0.25
7.21
9.16
2.16
0.98
0.25
1.67
9.16
4.02
8.68(3.60 - 28.98) - 3.75(0.54 - 66.04) 0.98(1.005 - 15.29)

220.298 245.62 13.999 11 .32
Also | A 2 |
2.16
8.68
0.98
4.12
3.14
0.25
9.16
7.21
2.16
Expanding by R 1
2.16(0.54 - 66.04) - 8.68(8.89 - 22.64) 0.98(37.74 - 0.785)
- 141.48 119.35 36.21 14.08
2.16
And | A 3 | 4.12
3.75
1.67
8.68
0.25
3.14
4.02
9.16
Expanding by R 1
2.16(15.29 - 1.005) - 3.75(37.74 - 0.785) 8.68(16.56 - 5.24)

30.85 - 138.58 98.257 - 9.4724
| A1 | 11.32
So x 1
1.44
|A|
7.79
| A 2 | 14.08
x2
1.798
|A|
7.79
| A3 | 9.4724
x3
1.209
| A|
7.79
Hence the solution is : x ( x1 , x 2 , x3 ) T (1.44,1.798,1.209) T
Some important Hints

(i) Elementary Row Operation (ERO)
The operations over a matrix A are called elementary row operations if:
Any two rows of A are interchanged i.e. Ri j
Any row of A is multiplied by a non-zero element. i.e. KRi
Any row of A is added to a scalar multiple of another row of A. i.e. Ri KRj
(ii) Echelon Form of a Matrix

A matrix is said to be echelon matrix if:
The first non-zero element of each row is non zero [as a special case 1]
The number of zeros on L.H.S of 1 (one) increases row by row.
Zero row is the last one
Example
Numerical Analysis
1
0
A = 0
1
0
0
5
1
0
4
6
3
(iii) Reduced Echelon form

A matrix is called in reduced Echelon form if:
It is in Echelon form.
If 1 lies in any column, the remaining Entries in that column will be Zero.
Example:
1
A= 0
0
0
1
0
0 1
0 2
1 3
Type No. 2
Gauss Elimination Method
KEY
Write the system of equations in matrix form.
Write the matrix form in tabular form.
Change the co-efficient matrix to Echelon form.
Find the values of unknowns. i.e. x1, x2, x3,..
T
Solution is: x ( x1 , x2 , x3 )
Q.2/Page # 83
Solve by Gauss Elimination Method
2x1 x2 + 3x3 x4 = 7
x1 x2 + 4x3 2x4 = 5
3x1 + 2x2 + x3 + 4x4 = 31
4x1 3x2 + 3x3 3x4 = -5
Ans:
Writing the system of equations in matrix form
2 1 3
1 1 4
1 x1 7
2 x2 5
3 2 1 4 x3 31

4 3 3 3 x4 5
Ax = c
Using Gauss Elimination Method, We write the system in tabular form as:
14
Numerical Analysis
R20
15
Rows
X1
X2
X3
X4
Row
Explanation
R1
R2
R3
R4
2
1
3
4
-1
-1
2
-3
3
4
1
3
-1
-2
4
-3
7
5
31
-5
Eq(1)
Eq(2)
Eq(3)
Eq(4)
R5
R6
R7
R8
1
0
0
0
-1/2
-1/2
7/2
-1
3/2
5/2
-7/2
-3
-1/2
-3/2
11/2
-1
7/2
3/2
41/2
-19
R1
R2 R5
R3 3R5
R4 4R5
R9
R10
R11
R12
1
0
0
0
-1/2
1
0
0
3/2
-5
-14
-8
-1/2
3
5
2
7/2
-3
-31
-22
-2R6
R7 7/2R10
R8 + R10
R13
R14
R15
R16
1
0
0
0
-1/2
1
0
0
3/2
-5
1
0
-1/2
3
-5/14
-12/14
7/2
-3
31/14
-60/14
-1/14 R11
R12 + 8R15
R17
R18
R19
R20
1
0
0
0
-1/2
1
0
0
3/2
-5
1
0
-1/2
3
-5/14
1
7/2
-3
31/14
60/12
-14/12 R16
x4 = 60/12 = 5
x4 = 5
5
31
5
31
x4
x 3 (5)
14
14
14
14
31 25 31 25 56
x3
4
14 14
14
14
R18 x 2 5 x3 3 x 4 3 x 2 5(4) 3(5) 3 x 2 20 15 3 x 2 3 5 x 2 2
R19 x3
1
3
1
7
1
3
1
7
5 7
x 2 x3 x 4 x1 ( 2) ( 4) (5) x1 1 6
2
2
2
2
2
2
2
2
2 2
7 5
7 5 2 12 2
x1 1 6 x1
1
2 2
2
2
R17 x1
Numerical Analysis
16
T
T
Hence x ( x1 , x 2 , x3 , x 4 ) (1,2,4,5)
Q.1(ii)/Page # 83
X1 + x2 + x3 + x4 = 4
2x1 + 3x2 + 7x3 x4 = 11
3x1 2x2 + 5x3 3x4 = 3
4x1 5x2 2x3 3x4 = -6
Ans:
Matrix form is
1
2
1
3
2
5
1
1
3
1
7
5
2
x1
x 2

x3
x 4
4
11
3
Ax = c
Using Gauss Jordon Method in tabular form:
Row
Explanation
Eq(1)
Eq(2)
Eq(3)
Eq(4)
Rows
X1
X2
X3
X4
R1
R2
R3
R4
1
2
3
4
1
3
-2
-5
1
7
5
-2
1
-1
-3
-3
4
11
3
-6
R5
R6
R7
R8
1
0
0
0
1
1
-5
-9
1
5
2
-6
1
-3
-6
-7
4
3
-9
-22
R9
R10
R11
R12
0
1
0
0
1
0
0
0
5
-4
27
39
-3
4
-21
-34
3
1
6
5
R6
R5 R9
R7 + 5R9
R8 + 9R9
R13
R14
R15
R16
0
0
1
0
0
1
0
0
1
0
0
0
-7/9
8/9
8/9
-11/3
2/9
17/9
17/9
-11/3
1/27 R11
R9 5R13
R10 + 4R13
R12 39R13
R17
R18
R19
R20
0
0
0
1
0
0
1
0
0
1
0
0
0
0
0
1
1
1
1
1
-3/11 R16
R13+7/9R17
R14 8/9R17
R158/9 R17
R21
R22
R23
R24
1
0
0
0
0
1
0
0
0
0
1
0
0
0
0
1
1
1
1
1
R20
R19
R18
R17
R1
R2 2R5
R3 3R5
R4 4R5
Numerical Analysis
R24
R23
R22
R21
Hence
17
x4 = 1
x3 = 1
x2 = 1
x1 = 1
T
T
the solution is: x ( x1 , x2 , x3 , x4 ) (1,1,1,1) Answer
Type No. 4
Method of Inversion
Key
Write the system of equation in matrix form.

Re-write it as x = A-1c
Find A-1 by Gauss Jordan Method.
Tabular form of this method contains co-efficient Matrix and the identity matrix.
Change the co-efficient Matrix to identity Matrix by ERO and get the values of
unknowns.
Identity Matrix is:
1
I= 0
0
0
1
0
0
1
etc.
Q.4/Page # 84
Solve the system of equation by Inversion of the matrix.
4.65x1 + 1.98x2 4.62x3 = 11.25
0.79x1 2.54x2 + 3.16x3 = -4.91
4.02x1 8.17x2 + 2.93x3 = -6.90
Ans:
Matrix form of the system is
4.65
0.79
4.02
1.98
2.54
8.17
4.62 x1
11.25
3.16 x 2 4.91
6.90
2.93 x3
Ax = c
x = A-1c -(1)
We 1st find A-1 by Gauss Jordan Method
Numerical Analysis
18
Rows
X1
X2
X3
C1
C2
C3
Row
Explanation
R1
R2
R3
4.65
0.79
4.02
1.98
-2.54
-8.17
-4.62
3.16
2.93
1
0
0
0
1
0
0
0
1
Eq(1)
Eq(2)
Eq(3)
R4
R5
R6
1
0
0
0.42
-2.88
-9.87
0.99
2.38
-1.04
0.25
-0.11
-1
0
1
0
0
0
1
1/4.65 R1
R2 0.79R4
R3 4.02R4
R7
R8
R9
0
1
0
1
0
0
-0.82
1.34
-9.20
0.07
0.22
-0.32
-0.34
0.14
-3.43
0
0
1
-1/2.88 R5
R4 0.42R7
R6 + 9.87R7
R10
R11
R12
0
0
1
0
1
0
1
0
0
0.34
0.36
-0.24
0.36
-0.04
-0.34
-0.11
-0.08
0.14
-1/9.20 R9
R7 + 0.82R10
R8 1.34R10
R13
R14
R15
1
0
0
0
1
0
0
0
1
-0.24
0.36
0.34
-0.34
-0.04
0.36
0.14
-0.08
-0.11
R12
R11
R10
Hence A
-1
0.24
= 0.36
0.34
0.34
0.04
0.36
0.14
0.08
0.11
Equation (1)
x1
x 2
x3
x1
x 2
x3
0.24
0.36
0.34
0.34
0.04
0.36
0.14
0.08
0.11
11.25
4.91
6.90
2.74 1.70 0.99

5.40
4.01 0.24 0.61 4.38

6.43
3.90 1.78 0.74
T
T
Hence the solution is: x ( x1 , x 2 , x3 ) ( 5.40,4.38,6.43)
NOTES
(i)
Square Matrix
Number of Rows = Number of Columns
(ii)
(iii)
Rectangular Matrix
Number of Rows Number of Columns
Diagonal Matrix
Numerical Analysis
If in a Matrix, aij =0 for i j and atleast one aii 0
For example
2
A= 0
0
(iv)
0
3
6
0
where 2, 6, 3 are called diagonal elements.
Upper Triangular Matrix

If aij = o, for i j and there exists at least one non-zero entry for i j .
For example
3
U= 0
0
(v)
2
7
0
5
9
6
Lower Triangular Matrix

If aij = 0, for i j and there exists at least one non-zero entry for i j
For example
4
L = 9
1
0
2
6
0
7
(vi)
Singular Matrix
If |A| = 0 A is singular.
(vii) Non-singular Matrix
If |A| 0 A is non-singular.
(viii) How to express a Matrix A in the product of L and U.
i.e.
A = LU
where L is the unit lower triangular and U is the upper triangular. i.e.
a11 a12 a13
If A = a21 a22 a23 , then

a31 a32 a33
a11 a12 a13
1 0 0 u11 u12 u13
a
a22 a23 = l21 1 0 . 0 u 22 u 23
21
a31 a32 a33

l31 l32 1 0
0 u33
TYPE NO. 5
CHOLESKIS METHOD
KEY
Given a system of equation: Ax = c
Find L and U such that: LU = A
Where L is the lower triangular matrix with diagonal elements 1 and U is upper
triangular Matrix.
19
Numerical Analysis
20
Let y = (y1, y2, y3,., yn)T

Find y by using:
Ly = c
Find x = (x1, x2, x3, ..xn)T by using Ux = y which is the solution
Q.17/Page # 88
1
Decompose the Matrix A = 2

2
7
12
8
16
in to the product of LU. Where L is the unit
lower triangular and U is upper triangular. Hence solve the system of equations.
.x1 + 2x2 + 3x3 = 2
2x1 + 8x2 + 7x3 = 7
2x1 + 16x2 + 12x3 = 10
Ans:
1st we express the matrix in the LU form. i.e. A = LU
1
2
2
8
16
1
3
7 = l
21
l
12
31
0
1
l32
u11 = 1
l21.u11 + 0 + 0 = 2
l21(1) = 2
l21= 2
l31.u11 + 0 = 2
l31(1) = 2
l31 = 2
hence L = 2
2
0
1
3
0
0
1
u
u
11 12
. 0 u 22
1 0
0

u13
u 23
u 33
u12 + 0 + 0 = 2
u12 = 2
l21.u12 + u22 = 8
(2)(2) + u22 = 8
u22 = 8 4 = 4
u22 = 4
l31.u12 + l32.u22 + 0 = 16
2.2 + l32 (4) = 16
4l32 = 16 4 = 12
l32 = 3
1
and U = 0
0
2
4
0
u13 + 0 + 0 = 3
u13 = 3
l21.u13 + u23 = 7
(2)(3) + u23 = 7
u23 = 7 6 = 1
l31.u13 + l32.u23 + u33 = 12
2.3 + 3.1 + u33 = 12
u33 = 12 6 3 = 3
u33 = 3
3
1
3
Now we find the solution set of the system

To find y such that:
L
y = c
Numerical Analysis
1 0 0 y1
2
2 1 0 y 7
2

2 3 1 y
10
3
y 00
2
1

2 y1 y 2 0 7
2 y 3y y
10
1
2
3
y1 = 2
2y1 + y2 = 7
(2)(2) + y2 = 7
y2 = 7 4 = 3
y2 = 3
R3
2y1 + 3y2 + y3 = 10
(2)(2) + (3)(3) + y3 = 10
y3 = 10 4 9 = - 3
y3 = -3
Therefore y = (y1, y2, y3)T = (2, 3, -3)T
Now we find x= (x1, x2, x3)T such that:
U x = y
R1
R2
x1
2
3
x
x
3
0 0
3
x1 2 x 2 3x3
2
0 4 x 2 x3 3
0 0 3x3
3
1
0
2
4
3
1
3
3x3 = -3
x3 = -1
R2
4x2 + x3 = 3
4x2 + (-1) = 3
4x2 = 3 + 1 = 4
x2 = 1
R3
x1 + 2x2 + 3x3 = 2
x1 + 2.1 + 3. -1 = 2
x1 + 2 3 = 3
x1 = 4
Hence the solution is: x = (x1, x2, x3)T = (4, 1, -1)T
R3
Diagonally Dominant Matrix

A square Matrix A whose elements are aij is said to be diagonally dominant if:
n
aii aij
, i = 1,2, n
i1
i j
Example
21
Numerical Analysis
a
11
Let A = a 21
a
31
a12
a 22
a 32
22
a13 5
1
2

a 23 3 12
3
a33 1 4 20
In the above matrix the Diagonal elements are 5, -12, -20. the absolute of these value are 5, 12, 20. These
values are greater than the absolute sum of the remaining elements two elements in that row. For example 5
is greater than |1| + |-2| = 1 + 2 = 3. Similar case for the other two rows.
TYPE NO. 2
ITERATIVE METHODS
There are two types of Iterative Methods
Jacobi Method
Gauss Seidel Method.
Type # 1
Jacobis Method
KEY
Write down the system in Diagonally Dominant form
Write down the Diagonally Dominant form in Iterative form.
Find 1st Iteration by Putting x0 = (0, 0, 0, 0,)T
We continue this process until we get the required accuracy.

Q.16/ Page # 87
Solve the system of equations
x1 + 0.1x2 = 1.0
0.1x1 + x2 + 0.1x3 = 2
0.1x2 + x3 = 3
Using (i) Jacobis Method.
(ii) Gauss Seidel Method.
Ans:
(i) 1st we will use the Jacobis Method
x1 + 0.1x2 = 1.0
0.1x1 + x2 + 0.1x3 = 2
0.1x2 + x3 = 3
The system is already in diagonally dominant form. So it can be written as:
. x1 = 1 0.1x2
. x2 = 2 0.1x2 0.1x3 -----------------> (A)
. x3 = 3 0.1x2
1st Iteration
Put x1 = x2 = x3 = 0 in (A)
x1 = 1 , x2 = 2 and x3 = 3
2nd Iteration
Put x1 = 1, x2 = 2 and x3 = 3 in (A)
Numerical Analysis
x1 = 1 (0.1)2 = 1 0.2 = 0.80

x2 = 2 (0.1) 1 (0.1) 3 = 2 0.1 0.3 = 1.6
x3 = 3 (0.1)2 = 3 0.2 = 2.8
3rd Iteration
Put x1 = 0.80, x2 = 1.6 and x3 = 2.8 in (A)
x1 = 1 (0.1) (1.60) = 0.84

.
x2 = 2 (0.1)(0.80) (0.1)(2.8) = 1.64
x3 = 3 (0.1)(1.6) = 2.836
th
4 Iteration
Put x1 = 0.84, x2 = 1.64 and x3 = 2.836 in (A)
x1 = 1 (0.1) (1.64) = 0.836

x2 = 2 (0.1)(0.84) (0.1)(2.846) = 1.6324
x3 = 3 (0.1)(1.64) = 2.836
This converges to the exact solution. Hence the solution is:
x1 = 0.836, x2 = 1.6324 and x3 = 2.836
(ii) Now we will use Gauss Seidel Method to solve the system.
Since
. x1 = 1 0.1x2-----------------------> (1)
. x2 = 2 0.1x2 0.1x3-------------> (2)
. x3 = 3 0.1x2 ----------------------> (3)
1st Iteration
Put x2 = 0 in (1) x1 = 1
Put x1 = 1 and x3 = 0 in (2)
x2 = 2 (0.1)(1) 0 = 1.99
Put x2 = 1.99 in (3)
x3 = 3 (0.1)(1.99) = 2.801
nd
2 Iteration
Put x2 = 1.99 in (1)
x1 = 1 (0.1)(1.99) = 0.801
Put x1 = 0.801 and x3 = 2.801 in (2)
x2 = 2 (0.1)(0.801) (0.1)(2.801) = 1.640

Put x2 = 1.640 in (3)
x3 = 3 (0.1)(1.640) = 2.836
3rd Iteration
Put x2 = 1.640 in (1)
x1 = 1 (0.1)(1.640) = 0.836
Put x1 = 0.836 and x3 = 2.836 in (2)
x2 = 2 (0.1)(0.836) (0.1)(2.836) = 1.632

Put x2 = 1.632 in (3)
x3 = 3 (0.1)(1.632) = 2.836
This converges to the real solution. So the solution is:
x1 = 0.836, x2 = 1.632 and x3 = 2.836
Q.10/Page # 86
Solve the system of equations.
6x1 3x2 + x3 = 11
23
Numerical Analysis
24
2x1 + x2 8x3 = -15

x1 7x2 + x3 = 10
Using Gauss Seidel method with initial vector x0 = (0, 0, 0)T
Ans:
1st we will write the system in Diagonally Dominant form by interchanging 2nd and 3rd equations.i.e.
6x1 3x2 + x3 = 11
x1 7x2 + x3 = 10
-----------> (A)
2x1 + x2 8x3 = -15
Writing the system (A) as:
x1 = 1/6(11 + 3x2 x3)----------> (1)
x2 = 1/7(-10 + x1 + x3)----------> (2)
x3 = 1/8(15 + 2x1 + x2)---------> (3)
1st Iteration
Put x2 = x3 = 0 in (1)
x1 = 1/6(11+0 0) = 1.833
Put x1 = 1.833 and x3 = 0 in (2)
x2 = 1/7(- 10 + 1.833 + 0) = - 1.166

Put x1 = 1.833 and x2 = - 1.166 in (3)
x3 = 1/8(15 + 2(1.833) 1.166) = 2.125

nd
2 Iteration
Equation (1) x1 = 1/6[11 + 3(- 1.166) 2.125] = 0.885
Equation (2) x2 = 1/7[- 10 + 0.885 + 2.125] = - 0.989
Equation (3) x3 = 1/8[15 + 2(0.885) 0.989] = 1.972
3rd Iteration
Equation (1) x1 = 1/6[11 + 3(- 0.989) 1.972] = 1.001
Equation (2) x2 = 1/7[- 10 + 1.001 + 1.972] = - 1.003
Equation (3) x3 = 1/8[15 + 2(1.001) 1.003] = 2.000
Which converges to the real solution. i.e x1 = 1, x2 = - 1 and x3 = 2
hence the solution is x = (x1, x2, x3)T = (1, -1, 2)T
Q.9/Page # 86
The pair of equations
.x1 + 2x2 = 3.0
.3x1 + x2 = 4.0
Can be re-arranged to give:
x1 = 3 2x2
x2 = 4 3x1
Apply Jacobis Method to this arrangement, beginning with x = (1.01, 1.01)T and notice the system diverges.
Now apply Gauss Seidel method to system and comment which method diverges more rapidly.
Ans: Given that:
x1 = 3 2x2
.x2 = 4 3x1
st
1 we use Jacobis Method
1st Iteration
put x1 = 1.01 and x2 = 1.01 in the above equations.
.x1 = 3 2(1.01) = 0.98
Numerical Analysis
.x2 = 4 3(1.01) = 0.97
2nd Iteration
x1 = 3 2(0.97) = 1.06
x2 = 4 3(0.98) = 1.06
3rd Iteration
x1 = 3 2(1.06) = 0.88
x2 = 4 3(1.06) = 0.82
4th Iteration
x1 = 3 2(0.82) = 1.36
x2 = 4 3(0.88) = 1.36
5th Iteration
x1 = 3 2(1.36) = 0.28
x2 = 4 3(1.36) = - 0.08
6th Iteration
x1 = 3 2(- 0.08) = 3.16
x2 = 4 3(0.28) = 3.16
after 6th Iteration x1 = 3.16 and x2 = 3.16. i.e. x = (3.16, 3.16)T, which is diverging.
Now using Gauss Seidel Method
x1 = 3 2x2---------->(1)
x2 = 4 3x1---------->(2)
1st Iteration
put x2 = 1.01 in (1)
x1 = 3 2(1.01) = 0.98 put in (2) x2 = 4 3(0.98) = 1.06

nd
2 Iteration
equation (1) x1 = 3 2(1.06) = 0.88
equation (2) x2 = 4 3(0.88) = 1.36
3rd Iteration
equation (1) x1 = 3 2(1.36) = 0.28
equation (2) x2 = 4 3(0.28) = 3.16
4th Iteration
equation (1) x1 = 3 2(3.16) = - 3.32
equation (2) x2 = 4 3(- 3.32) = 13.96
5th Iteration
equation (1) x1 = 3 2(13.96) = - 24.92
equation (2) x2 = 4 3(- 24.92) = 78.76
6th Iteration
equation (1) x1 = 3 2(78.76) = - 154.52
equation (2) x2 = 4 3(- 154.52) = 467.56
After 6th Iteration x1 = - 154.52 and x2 = 467.56. i.e. x = (- 154.52, 467.56)T. which is diverging.
Comment: After six iterations we see that Gauss Seidel Method diverges rapidly than the Jacobis Method.
CHAPTER#4
Numerical Integration
1. Trapezoidal Rule or Trapezium Rule.
25
Numerical Analysis
26
1
1
f ( x0 ) f ( x1 ) f ( x 2 ) .............. f ( x n )
f ( x ) dx h
2
2
Where
ba
n
2. Simpsons Rule
f ( x) dx
h
f ( x0 ) 4 f ( x1 ) 2 f ( x 2 ) 4 f ( x3 ) 2 f ( x4 ) 4 f ( x5 ) ................. f ( xn )
3
ba
n
Where
Error = Approximate value True value

EXERCISE ON PAGE 112
Q #1
:
P 112
f ( x ) dx
Here
(x
f ( x) x
h
1
) dx
x
1
, a 2, b 3, n 4
x
b a 32 1
n
4
4
f ( x) x
x0 2
1
x
1
2.5
2
f ( x0 ) f (2) 2
f ( x1 ) f ( 2.25) 2.25
x1 2
1
2.25
4
x2 2
2
2.5
4
f ( x 2 ) f (2.5) 2.5
x3 2
3
2.75
4
f ( x3 ) f ( 2.75) 2.75
x4 2
4
3
4
f ( x 4 ) f (3) 3
1
2.694
2.25
1
2.9
2.5
1
3.114
2.75
1
3.333
3
Numerical Analysis
Apply Trapezoidal Rule
1
1
f ( x0 ) f ( x1 ) f ( x 2 ) f ( x3 ) f ( x 4 )
2
2
f ( x) dx h
1
1 2 .5
3.333
)dx
2.694 2.9 3.114
x
4 2
2
(x
1
1.25 8.708 1.667
4
1
11.625 2.9063
4
For true value
(x
1
) dx
x
x2
2
xdx
1
dx
x
ln x
3
2
1 2
(3 2 2 ) (ln 3 ln 2)
2
1
(9 4) (1.099 0.693)
2
2.5 0.406 2.906
Error = Approximate value True value = 2.9063 2.906 0.0003
(ii)
f ( x )dx
1
dx
e4x
f ( x)
Here
1
e 4 x , a 0, b 2, n 4
4x
e
ba 20 2 1

n
4
4 2
f ( x ) e 4 x
x0 0
f ( x0 ) f (0) e 0 1
x1 0
1
0 .5
2
f ( x1 ) f (0.5) e 4( 0.5) 0.135
27
Numerical Analysis
x2 0
2
1
2
f ( x 2 ) f (1) e 4 0.018
x3 0
3
1.5
2
f ( x3 ) f (1.5) e 4(1.5) 0.002
x4 0
4
2
2
f ( x 4 ) f (2) e 8 0.0003
Trapezoidal rule
1
1
f ( x 0 ) f ( x1 ) f ( x 2 ) f ( x3 ) f ( x 4 )
2
2
f ( x)dx h
1 1
0.0003
0.135 0.018 0.002
2 2
2
1
0.5 0.1550 0.0002
2
1
0.6550 0.3275
2
For true value
e 4 x dx
e 4 x
4
1 4 x
e
4
2
0
1 8
(e e 0 )
4
1
(0.0003 1)
4
1
(0.9997) 0.2499
4
Error = Approximate value True value

=0.3275-0.2499=0.0776
Q #2
:
P 112
Here
f ( x ) dx
( x 4 x 2 1)dx
f ( x) x 4 x 2 1, a 1, b 2, n 2
b a 2 1 1
n
2
2
f ( x) x 4 x 2 1
f ( x 0 ) f (1) 1 1 1 1
x0 1
x1 1
1
1.5
2
f ( x1 ) f (1.5) (1.5) 4 (1.5) 2 1 3.8125
28
Numerical Analysis
x2 1
2
2
2
f ( x 2 ) f (2) (2) 4 (2) 2 1 13
Simpsons Rule
h
f ( x0 ) 4 f ( x1 ) f ( x 2 )
a
3
3
1
4
2
1 ( x x 1)dx 2 3 1 4(3.8125) 13
f ( x) dx
1
(1 15.2500 13)
6
1
29.2500 4.8760
6
(ii)
f ( x ) dx e x sin xdx
0
Here
f ( x) e x sin x, a 0, b 2, n 2
ba 20 2
1
n
2
2
f ( x) e x sin x
x0 0
f ( x0 ) f (0) e 0 sin 0 0
x1 0 1 1
f ( x1 ) f (1) 2.2874
x 2 0 1(2) 2
f ( x 2 ) f (2) 6.7188
Simpsons Rule
h
f ( x0 ) 4 f ( x1 ) f ( x 2 )
3
2
1
x
0 e sin xdx 3 0 4 2.2874 6.7188
f ( x) dx
1
15.8668
3
5.2886
Q #3
:
P 112
Here
f ( x ) dx
ydx
a 0, b 1, h 0.1
29
Numerical Analysis
x
f ( x) y
x 0 0.2
f ( x0 ) 1
x1 0.1
f ( x1 ) 0.99
x 2 0.2
f ( x 2 ) 0.961
x3 0.3
f ( x 3 ) 0.9141
x 4 0.4
f ( x 4 ) 0.852
x5 0.5
f ( x 5 ) 0.779
x 6 0.6
f ( x 6 ) 0.697
x 7 0.7
f ( x 7 ) 0.613
x8 0.8
f ( x8 ) 0.527
x9 0.9
f ( x 9 ) 0.445
x10 1
f ( x10 ) 0.369
Trapezoidal Rule
f
(
x
)
f
(
x
)
f
(
x
)
f
(
x
)
f
(
x
)
f
(
x
)
f
(
x
)
f
(
x
)
f
(
x
)
f
(
x
)
2 0 1 2 3 4 5 6 7 8 9
b
a f (x)dx h 1
f ( x10 )
2
1
0.359
1
ydx
0
.
1
0
.
99
0
.
96
0
.
914
0
.
852
0
.
779
0
.
697
0
.
613
0
.
527
0
.
445
0 2
2
0.1 6.7780 0.6678
Q #4
:
P 112
Here
f ( x)dx
f ( x)
h
x
x0 1
dx
1 x4
1
, a 1, b 2, n 6
1 x4
b a 2 1 1
n
6
6
f ( x)
1
1 x4
f ( x0 ) 0.5
30
Numerical Analysis
x1 1
1
1.2
6
f ( x1 ) 0.325
x2 1
2
1.3
6
f ( x 2 ) 0.259
x3 1
3
1.5
6
f ( x3 ) 0.165
x4 1
4
1.7
6
f ( x 4 ) 0.107
x5 1
5
1.8
6
f ( x5 ) 0.087
x6 1
6
2
6
f ( x 6 ) 0.059
Trapezoidal Rule
1
1
f ( x0 ) f ( x1 ) f ( x 2 ) f ( x3 ) f ( x 4 ) f ( x5 ) f ( x 6 )
2
2
f ( x)dx h
1
1 0. 5
0.059
dx
0.325 0.259 0.165 0.107 0.087
4
6
2
2
1 x
1
1.223 0.205
6
= 0.4088
Answer
31

Chapter # 2 Solution of Algebraic and Transcendental Equations

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter # 2 Solution of Algebraic and Transcendental Equations

Uploaded by

Copyright:

Available Formats

Numerical Analysis

Linear Equations: An equation in which the highest power of variable is One

Non-Linear Equation: Equation involving higher powers of a variable is called

(3) Transcendental Equation: Equation involving functions of variable

(ii) Secant Method

(iii) False Position Method

(iv) Bisection Method

Prepared by: Inayat Ullah Lecturer in Mathematics Edwardes College Peshawar

(v) Iterative Process

Interval or initial point x0 will be given if not, put x = 0, 1, 2, 3, 4, .in f(x).

Prepared by: Inayat Ullah Lecturer in Mathematics Edwardes College Peshawar

Using Newton Raphson Method

For n = 0, equation (1) x1

Prepared by: Inayat Ullah Lecturer in Mathematics Edwardes College Peshawar

For n = 3, equation (1) x 4

Using Newton Raphson Method

Prepared by: Inayat Ullah Lecturer in Mathematics Edwardes College Peshawar

For n = 0, equation (1) =>

12.e1 1 2.7183 1 3.7183

For n = 1, equation (1) =>

x1 e x1 1 0.6839 e 0.6839 1 0.4677 1.9816 1 1.9267

For n = 2, equation (1) =>

x2 e x2 1 0.5776 e 0.5776 1 0.3336 1.7816 1 1.5944

Prepared by: Inayat Ullah Lecturer in Mathematics Edwardes College Peshawar

f(x) will be given.

Using Secant Method

as root lies in the interval (2, 3) x0 = 2 and x1 = 3

For n = 1, equation (1)

For n = 2, equation (1)

3( 2.2609 ) 2.2609 10( 3 ) 2.2609( 3 )3 10( 2.2609 )

18.9721 30 61.0443 22.6090 49.4632

For n = 3, equation (1)

182189 22.6090 15.1290 23.9230 4.4039

For n = 4, equation (1)

21.6802 19.7798 23.9230 24.5460 2.5234

Formula for False position is

xL = Lower value of the interval

The function f(x) will be given.

The interval will be given, if not then put x = 0, 1, 2, 3,.in f(x).

We put all these values in the formula (1) to get xi.

Prepared by: Inayat Ullah Lecturer in Mathematics Edwardes College Peshawar

now f(xi) = f(1.79) = sin(1.79) -0.5(1.79) = 0.9760-0.895=0.081

.f(xR) = - 0.0907 (-ve)

Prepared by: Inayat Ullah Lecturer in Mathematics Edwardes College Peshawar

The value of f(x) will be given.

f(x1) = f(/2) = sin/2 -/2/2 = 1 /4 = 1 0.785 = 0.215 (+ve)

f(x2) = f() = sin - /2 = 0 1.571 = -1.571(-ve)

As f(x) changes its sign for x1 = /2 and x2 =

Put this value of x in (1)

x1 = 1.571 and x2= x = 2.356

Put this value of x in equation (1)

Put in equation (1)

Put in equation (1)

x2 = 1.963 (Not changed)

Prepared by: Inayat Ullah Lecturer in Mathematics Edwardes College Peshawar

Put this value of x in equation (1)

and is called coefficient matrix

which is called column matrix or column vector.

The system of linear equations can be solved by two methods.

Prepared by: Inayat Ullah Lecturer in Mathematics Edwardes College Peshawar

Write the system of equations in Matrix form i.e. Ax = c

The solution set will be written as: a = (x1, x2, x3) T

3.14 4.02 2.16 x 3