A Text Book for S.Y.B.Sc./S.Y.B.A.

Mathematics (2013 Pattern),

Savitribai Phule Pune University, Pune.

T-222(B): NUMERICAL METHODS AND ITS APPLICATIONS

Panel of Authors
P.G.Jadhav(Convenor)
Hon.Balasaheb Jadhav Arts, Commerce and
Science College, Ale, Tal.: Junnar, Dist: Pune.
(Email:pgjmaths@gmail.com)
B. B.Divate
G.E.Societys
HPT Arts and RYK Science College, Nashik.
(Email:rykmathsbbd@gmail.com)
B. S. Darekar
Arts, Commerce and Science College,
Narayangaon,Dist: Pune.
(Email:bsdarekar@redimail.com, sbdarekar@gmail.com)

Editors
Dr. P. M. Avhad

Dr. S. A. Katre

Conceptualized by Board of Studies(BOS) in Mathematics, Savitribai Phule Pune

University, Pune.

Preface
We are glad to present this book to S.Y.B.Sc. Mathematics students of Pune University. This book is strictly written according to the revised syllabus of second
semester of S.Y.B.Sc. Mathematics, implemented by University of Pune, since June
2014. The book deals with errors, solution of algebraic and transcendental equations, Interpolation, Least Square Curve tting procedure, Numerical dierentiation
and integration and numerical solution of rst order ODE.
This book is based on our teaching experience of last 16 17 years. Simple and
lucid language is used to explain the fundamental concept in each chapter. Book
contains large number of examples with all types. We request the students to solve
these problems on their own.
We are thankful to Prof. P.M.Avhad (chairman) and all other BOS members in
Mathematics, University of Pune for their continuous co-operation, encouragement,
and motivation to write this book. We are also thankful to University authorities
and press for publishing this book.
Suggestions and comments for further development of this book are welcome and
will be acknowledge.
In case of queries/suggestions, send an email to: pgjmaths@gmail.com

-Authors

Acknowledgment
We sincerely thank the following University authorities (Savitribai Phule Pune University, Pune) for their constant motivation and valuable guidance in the preparation
of this book.

Dr. W. N. Gade, Hon. Vice Chancellor, Savitribai Phule Pune University,

Pune.
Dr. V. B. Gaikwad, Director BCUD, Savitribai Phule Pune University, Pune.
Dr. K. C. Mohite, Dean, Faculty of Science, Savitribai Phule Pune University,
Pune.
Dr. B. N. Waphare, Professor, Department of Mathematics, Savitribai Phule
Pune University, Pune.
Dr. M. M. Shikare, Professor, Department of Mathematics, Savitribai Phule
Pune University, Pune.
Dr. V. S. Kharat, Professor, Department of Mathematics, Savitribai Phule
Pune University, Pune.
Dr. V. V. Joshi, Professor, Department of Mathematics, Savitribai Phule Pune
University, Pune.
Mr. Dattatraya Kute, Senate Member, Savitribai Phule Pune University; Manager, Savitribai Phule Pune University Press.
All the sta of Savitribai Phule Pune University press.

Syllabus: PAPER-II:MT-222(B): NUMERICAL METHODS AND ITS

APPLICATIONS
1. Errors: [4 ]
1.1 Errors and Their Computations
1.2 Rounding o numbers to n signicant digits, to n decimal places.
1.3 Absolute, relative and percentage errors.
1.4 A general error formula.
2. Solution of Algebraic and Transcendental Equations: [10]
2.1 Bisection method.
2.2 The method of False position.
2.3 The iteration method, Aitkens 2 process
2.4 Newton- Raphson Method.
3. Interpolation: [16]
3.1 Finite Dierence Operators and their relations.
3.2 Detection of Errors using dierence table.
3.3 Dierences of a polynomial
3.4 Newtons Interpolation Formulae (Forward and Backward )
3.5 Lagranges Interpolation Formula
3.6 Divided dierences and Newtons General Interpolation formula.
4. Least Squares Curve Fitting Procedures [4]
4.1 Fitting a Straight Line
4.2 Nonlinear curve tting: Power function y = axc ,polynomials of degree 2 and
3,Exponential function y = cedx
5. Numerical Dierentiation and Integration: [8 ]
5.1Numerical Dierentiation
5.2Numerical Integration,General quadrature formula.
5.3 Trapezoidal rule.
5.4 Simpsonss 13 rule.
5.5 Simpsonss 13 rule.

ii

6. Numerical solution of rst order ordinary dierential equations: [6]

6.1 Taylor Series method
6.2 Eulers method.
6.3 Modied Eulers methods.
6.4Runge - Kutta Methods 2nd and4th order.
Text Books : Prepared by the BOS Mathematics, University of Pune.
Recommended Book:
1. S.S. Sastry; Introductory Methods of Numerical Analysis, 3rdedition, Prentice
Hall of India.
Sections:1.3,1.4,2.1,2.2,2.3,2.4,2.5,3.3,3.4,3.5,3.6,3.9.1,3.10(3.10.1only), 4.2.1,4.2.2,5.2(excluding5.2.1,5.2.2),5.4.1,5.4.2,5.4.3,7.2,7.4,7.4.1,7.4.2,7.5
Reference Book:
1. K.E. Atkinson; An Introduction to Numerical Analysis, Wiley Publications.
2. H.C.Saxena; Finite dierences and Numerical Analysis, S.Chand
and Company

Contents
1

ERRORS
1.1 Introduction: . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2 Errors and Their Computations: . . . . . . . . . . . . . . . . .
1.3 Rounding o errors to n signicant digits,to n decimal places:
1.3.1 Truncation Error: . . . . . . . . . . . . . . . . . . . . .
1.4 Absolute, relative and percentage errors: . . . . . . . . . . . .
1.5 A General Error formula: . . . . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

1
1
1
2
3
3
9

2 SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

12
2.1 Introduction: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2 Mathematical Preliminaries: . . . . . . . . . . . . . . . . . . . . . . . 12
2.3 The Bisection Method: . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.4 The Method of False Position: . . . . . . . . . . . . . . . . . . . . . . 16
2.5 The Iteration Method : . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.5.1 The Aitkens 2 -process : . . . . . . . . . . . . . . . . . . . . 23
2.6 Newton-Raphson Method: . . . . . . . . . . . . . . . . . . . . . . . . 25

iii

Chapter 1
ERRORS
1.1

Introduction:

Now a days we use digital computers to solve mathematical problems.In practical

applications, an engineer also obtain nal results in numerical form with help of
computers. As computers have a nite world length so only xed number of digits
are used for computation. So, when we use any kind of computer or calculator to
solve mathematical problems, we get solutions quickly with some errors. In this
chapter we study various errors and their computations. 

1.2

Errors and Their Computations:

There are two types of numbers viz, exact and approximate numbers. Any real

31
315
number is exact number. e.g. 1, 2, 5, 12, 1000
, 1327
, 7, 10 7, , e.... are exact
numbers. Approximate numbers are those that represent the rational or irrational
numbers to a certain required accuracy.Approximate value of is 3.1416 or its better
approximation is 3.14159265. But we can not write its exact value. Same is true for
any irrational number.
Denition 1.1. Signicant Digits : The digits which are used to express a number are called signicant digits or signicant gures.
The numbers 62.345, 0.32122, 0.10026, 0.99999 contain ve signicant digits each.
23.4, 0.234, 0.0234, 0.000234 have three signicant gures.
The number of signicant digits is uncertain in number which is represented
by whole number like 32, 354; 1, 02, 34, 500 etc. The number 66400 has uncertain
1

CHAPTER 1.

ERRORS

number of signicant digits, whereas the numbers 6.64104 , 6.640104 and 6.6400
104 have three, four and ve signicant gures, respectively. 

1.3

Rounding o errors to n signicant digits,to n decimal

places:

Denition 1.2. Rounding-o process: In numerical computations, we come

across numbers numbers which have large number of digits and it will be necessary
to cut them to a suitable and usable number of gures. This process is called as
rounding-o.
Rules to round-o the number:To round-o a number to n signicant digits,
discards all digits to the right of the nth digit and if this (n + 1)th discarded number
is
1. less than 5,then leave the nth unaltered.
2. greater than 5, then add 1 in nth digit.
3. is 5 then
(a) keep nth digit as it is if (n + 1)th digit is even.
(b) add 1 in nth digit if (n + 1)th digit is odd.
Illustrative Examples
Example 1.1. Round-o following numbers to two, three and four signicant gures.
3.3465827, 15.23538753, 5.375829, 0.00457328, 54.2549757
Solution: The table for round-o(R-o) numbers to 2, 3 and 4 signicant gures
(SF) of given numbers is as follows:
Sr. No. Given number R-o no.to 2-SF R-o no.to 3-SF R-o no.to 4-SF
1
3.3465827
3.35
3.346
3.3466
2
15.23538753
15.24
15.235
15.2354
3
5.375829
5.38
5.376
5.3758
4
0.00457328
0.00
0.004
0.0046
5
54.2549757
54.25
54.255
54.2550

CHAPTER 1.

1.3.1

ERRORS

Truncation Error:

There is another type of error which is known as truncation error which is computed
by using approximate formula obtained by truncation of innite series given real
function. This error is associated with the problem of convergence.
Suppose Taylors series of real function f (x) is
f (x + h) = f (x) + hf (x) +

h2
h3
f (x) + f (x) +
2!
3!

(1.1)

If we neglect second and higher order derivatives in the above expansion (1.3.1)
then the error which arises in computation due to the truncation in formula is a
truncation error.
Remark 1.1. The round-o error can be reduced by computing for more signicant
gures. A useful rule is for each step retain at least one more gure than that given in
the data, round-o after performing last operation. The computer allow a precision
of 7 signicant gures in the range about 1038 to 1039 . Arithmetic carried out with
this precision is called single precision. For more accuracy arithmetic carried out
with 15 signicant gures with range about 10308 to 10308 (by MATLAB) is called
double precision.

1.4

Absolute, relative and percentage errors:

Denition 1.3. If xt is true value and xa is approximate value of a given quantity

then
Error e is dened as
e = xt xa = x

(1.2)

ea = |xt xa |

(1.3)

ea
x
=
xt
xt

(1.4)

Absolute error ea is dened as

Relative error er is dened as

er =

CHAPTER 1.

ERRORS

Percentage error ep is dened as

ep = 100er = 100

x
xt

(1.5)

Let U be number such that

ea = |xt xa | 6 U

(1.6)

then U is an upper limit of the absolute error and is said to measure absolute accuracy. Similarly the quantity
U
U

|xt |
|xa |

(1.7)

measures the relative accuracy.

Result 1: If x is the number rounded to n decimal places then the upper limit
of the absolute error is given by
U=

1
2 10n

(1.8)

Theorem 1.1. Let x1 and x2 be true values of two quantities,x1a and x2a be their
approximate values respectively.If e1 ,e2 and esum are errors in x1 ,x2 and x1 + x2
respectively then
esum = e1 + e2 .

(1.9)

(i.e. Sum of errors of two numbers is error of their sum.)

Proof:
e1 = x1 x1a and e2 = x2 x2a
e1 + e2 = (x1 x1a ) + (x2 x2a )
= (x1 + x2 ) (x1a + x2a )
= esum
Theorem 1.2. Let x1 and x2 be true values of two quantities,x1a and x2a be their
approximate values respectively. If e1 ,e2 and edif f are errors in x1 ,x2 and x1 x2
respectively then
edif f = e1 e2 .

(1.10)

CHAPTER 1.

ERRORS

(i.e. Dierence of errors of two numbers is error of their dierences.)

Proof:
e1 = x1 x1a and e2 = x2 x2a
e1 e2 = (x1 x1a ) (x2 x2a )
= (x1 x2 ) (x1a x2a )
= edif f
Theorem 1.3. Let x1 and x2 be true values of two quantities,x1a and x2a be their
approximate values respectively. If e1 ,e2 and eprod are errors in x1 ,x2 and x1 x2
respectively then
eprod = x1 e2 + x2 e1 (approximately).

(1.11)

Proof:
eprod = (x1 + e1 )(x2 + e2 ) x1 x2
= x1 x2 + x1 e2 + x2 e1 + e1 e2 x1 x2
= x1 e2 + x2 e1 + e1 e2
= x1 e2 + x2 e1 (approximately)
Theorem 1.4. Let x1 and x2 be true values of two quantities,x1a and x2a be their
approximate values respectively. If e1 ,e2 and eq are errors in x1 ,x2 and xx12 respectively
then
(
)
e2
x1 e1

(approximately).
eq =
(1.12)
x2 x1 x2
Proof:
(x1 + e1 ) x1
eq =

(x2 + e2 ) x2
x2 (x1 + e1 ) x1 (x2 + e2 )
=
x2 (x2 + e2 )
x2 e1 x1 e2
)
= (
x22 1 + xe22
x2 e1 x1 e2
e2
(approximately by neglecting
)
2
x2
x2
(
)
x1 e1
e2
=

x2 x1 x2
=

CHAPTER 1.

ERRORS

Result 2: If e1 , e2 , ...., en are errors in n numbers x1 , x2 , ...., xn respectively then

the error in their sum is
e = e1 + e2 + .... + en

(1.13)

Procedure to nd sum of numbers with dierent accuracies :

1. Isolate the number with largest error.
2. Round-o all other numbers retaining in them one digit more than in the isolated number.
3. Add up them.
4. Round-o the sum by discarding the one digit.
Example 1.2. If x = 0.321 is correct to 3 decimal places, then nd upper limit of
its absolute error and its relative accuracy.
Solution:We know that if the number x is rounded to N decimal places, then
upper limit of its absolute error is given by ,
U
U = 12 (10N ) and absolute accuracy is given by |x|
.
The upper limit U is
1
U = (103 ) = 0.0005,
2
and the relative accuracy is ,
U
0.0005
=
= 0.0016
|x|
0.321
Example 1.3. An approximate value of is given by x1 =
x = 3.141592654 Find the relative error.
Solution: We have

22
7

and its true value is

e = x x1 = 0.0030694
0.0030694
er =
= 0.000977
3.141592654
Example 1.4. Three approximate values of the number
0.12. Which of these three is the best approximation ?

1
9

are given by 0.10,0.11and

CHAPTER 1.

ERRORS

Solution: We have



1

 0.10 = 1
9
 90


1

 0.11 = 1
9
 900


1

 0.12 = 8
9
 900

It follows that 0.11 is the best approximation for 19 .

Example 1.5. Find the relative error of the number 11.64 if all of its digit are
correct.
Solution: Here error in the number x = 11.64 is,
1
e = (102 ) = 0.005,
2
Hence relative error is,
e
0.005
=
= 4.30 104 .
|x| |11.64|

Example 1.6. Evaluate the sum S = 3+ 3 11+ 5 to 4 signicant digits. Also,nd

its error and relative error.
er =

Solution: We have,

3
3 = 1.732, 11 = 2.234, 5 = 2.236
correct to 4 signicant digits.Hence,

3
S = 3 + 11 + 5 = 1.732 + 2.234 + 2.236 = 6.602
Now error in each number is 0.0005 hence error in S is,
e = 3 0.0005 = 0.0015.
This total error shows that the sum is correct to 3 signicant digits only. Hence we
take S = 6.60 and then relative error in Sis,
er =

e
0.0015
=
= 2.27 104 = 0.0002.
|x|
|6.60|

CHAPTER 1.

ERRORS

Example 1.7. Find the sum of the numbers:

0.1234, 0.3453, 0.000045, 186.12, 15.42, 0.0265, 304.24, 0.7694, 23.6758
Where in each number all given digits are correct.
Solution: We have three numbers, 186.12, 15.42, 304.24 which have greatest error
which is 0.005. Hence we round-o all the other number to three decimal digits.
These are :
0.123, 0.345, 0.000, 0.027, 0.769, and 23.676
The sum is given by
S = 186.12 + 15.42 + 304.24 + 0.123 + 0.345 + 0.000 + 0.027 + 0.769 + 23.676 = 530.72
To determine the error, we note that the rst three numbers have each an error of
0.005 and remaining have an error of 0.0005 each. Thus error in all the 9 numbers
in the sum is given by
e = 3(0.005) + 6(0.0005)
= 0.015 + 0.003
= 0.018
Thus The sum is given by
S = 530.72 0.018
Example 1.8. Two numbers are given as 4.8 and 53.23, both of which being correct
to the signicant gures given. Find their product.
Solution: Here the number with greatest error is 4.8 Hence we round-o the
second number to three signicant digits, i.e. 53.2 Their product is given by
P = 4.8 53.2 = 255.36 = 2.6 102
We have taken only two signicant digit since one of the given number, 4.8, contained
only two signicant digits.

CHAPTER 1.

1.5

ERRORS

A General Error formula:

In this we derive a general formula for the error committed in using a certain functional relation.Let
u = f (x1 , x2 , ..., xn )

(1.14)

be a function of n variables x1 , x2 , ..., xn , and let the error in each xi be xi . Then

the error u in u is given by
u + u = f (x1 + x1 , x2 x2 , ..., xn xn )

(1.15)

Expanding the right-hand-side by Taylors series we obtain

n

f
xi + term involving(xi )2
u + u = f (x1 , x2 , ..., xn ) +
xi
i=1

(1.16)

Assuming that the errors in xi are small and that (xi )xi 1, so that the squares
and higher powers of xi can be neglected, the above relation yields
n

f
f
f
f
xi =
x1 +
x2 + ..... +
xn
u
x
x
x
x
i
1
2
n
i=1

(1.17)

We observe that this formula has the same form as that for the total dierential of
u. The formula for the relative error follows immediately:
er =

u u x1 u x2
u xn
=
+
+ ..... +
u
x1 u
x2 u
xn u

(1.18)

Example 1.9. If u = 2yzx2 . If the errors in x, y and z is taken as 0.001 and

taking x = y = z = 1 then nd the relative error.
Solution:
u
2yz u 2z u 2y
= 3 ,
= 2,
= 2
x
x
y
x z
x
and
u =

2yz
2z
2y
x
+
y
+
z
x3
x2
x2

CHAPTER 1.

ERRORS

10

In general, the errors x, y and z may be positive or negative, and hence we

take the absolute values of the term on the right side. This gives

 
 

 2yz
  2z
  2y

(u)max  3 x +  2 y  +  2 z 
x
x
x
Now, let x = y = z = 0.001 and x = y = z = 1 then, the relative maximum
error (er )max is given by
(er )max =

0.006
(u)max
=
= 0.003
u
2
Exercise:

1. Round-o the following numbers to 2, 3 and 4 signicant gures

23.4523619826, 0.00154876, 235.69854756, 0.0000011455, 7.2564266

2. Estimate the sum S = 3 + 5 + 7 to four signicant digits and nd its error
and relative errors. (S = 6.614, e = 0.0015, er = 0.0002)

3. Calculate the value of 102 101 (0.04963)

4. Find the sum of the numbers 105.5, 27.25, 6.56, 0.1568, 0.000256, 208.6,
0.0235, 0.538 and 0.0571, where each number is correct to digits given. Estimate the error in the sum. (348.7 0.15)
5. Find the sum of the numbers 0.1532, 15.45, 0.000354, 305.1, 8.12, 143.3,
0.0212, 0.643 and 0.1734, where each number is correct to digits given. Estimate the error in the sum. (472.6 0.015)
6. Find the product of two numbers 56.54 and 12.4 which are both correct to the
signicant digits given. (701)
7. Find the product of two numbers 48.3 and 2.5 which are both correct to the
signicant digits given. (120)
8. Find the quotient q = x/y, where x = 4.536 and y = 1.32, correct to the digits
given. Find also the relative error in the result. (q = 3.44, er = 0.0039)
9. Find the quotient q = x/y, where x = 5.647 and y = 2.52, correct to the digits
given. Find also the relative error in the result. (q = 2.24, er = 0.002)

CHAPTER 1.

ERRORS

11

10. Three approximate values of the number 1/3 are given as 0.30, 0.33 and 0.34
Which of this three is the best approximation. (0.33)
11. Three approximate values of the number 2/3 are given as 0.65, 0.66 and 0.67
Which of this three is the best approximation. (0.67)
12. Three approximate values of the number 1/6 are given as 0.165, 0.166 and 0.167
Which of this three is the best approximation. (0.167)

Chapter 2
SOLUTION OF ALGEBRAIC AND
TRANSCENDENTAL EQUATIONS
2.1

Introduction:

In scientic and engineering studies we come across roots of equations of the form
f (x) = 0.

(2.1)

If f (x) is linear, quadratic, cubic or a biquadratic expression, then algebraic formulae

are available for expressing the roots in terms of the coecients. On the other hand
if f (x) is polynomial of higher degree or an expression involving transcendental
functions then algebraic methods are not available to nd roots of such equations.
In this chapter we study several numerical methods like bisection method, the
method of false position, iteration method, Aitkens 2 process and Newton Raphson method to nd solutions of algebraic and transcendental equation.

2.2

Mathematical Preliminaries:

In this section we state certain denitions and mathematical results which would be
useful in the sequel.
Denition 2.1. Algebraic function: An expression of form
f (x) = a0 + a1 x + a2 x2 + ... + an xn
is called as algebraic function or polynomial function of degree n.

12

(2.2)

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

13

Denition 2.2. Transcendental function: A non algebraic function involving at

least one of exponential, logarithmic, trigonometric, inverse trigonometric function
is called as transcendental function.
Theorem 2.1. Location of roots result: If function f (x) is continuous on closed
and bounded interval [a, b] and if f (a).f (b) < 0 ( i.e.f (a) and f (b) have opposite
signs) then f () = 0 for at least one (a, b).
Theorem 2.2. Mean value Theorem:If function f (x) is,
1. continuous on closed and bounded interval [a, b]
2. dierentiable on (a, b),
then there exist for at least one (a, b) such that
f () =

f (b) f (a)
.
ba

(2.3)

Theorem 2.3. Taylors series:If f (x) possesses all ordered continuous derivatives
in an interval containing a then in that interval f (x) is expressed as series containing
powers of (x a) as follows:
(x a)2
(x a)n (n)
f (a) + +
f (a) +
2!
n!
(This series is called as Taylors series of f (x) about a.)
f (x) = f (a) + (x a)f (a) +

2.3

(2.4)

The Bisection Method:

This method depends on repeated application of the location of root result.

Let f (x) = 0 be equation such that such that f (x) is the continuous real function
on some interval I.We use following steps to nd a approximate root of f (x) = 0.
1. Choose real numbers a and b satisfying a < b in I such that f (a)f (b) < 0.
2. The rth approximate root of the equation is given by,
xr =

(a + b)
,
2

r = 0, 1, 2, .

(2.5)

3. If f (xr ) = 0 then xr is a root/solution of the equation f (x) = 0.Stop the

process.

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

14

4. If f (xr ) = 0 then
(a) If f (a)f (xr ) < 0 then root lies in the interval (a, xr ).Then take b = xr and
go to step 2 above.
(b) If f (xr )f (b) < 0 then root lies in the interval (xr , b).Then take a = xr and
go to step 2 above.
5. Repeat steps 2 , 3 and 4 till we get root correct to required accuracy or till
getting required number of iterations.
Note: In bisection method in each iteration width of interval is reduced by a
factor one half and hence at the end of nth step width of interval containing root
will be |ba|
2n . Now,
|b a|
loge (|b a| )

2n
loge 2

(2.6)

This inequality is useful to nd number of iterations n required to obtain a root

with accuracy .
Example 2.1. Find the root of the equation x2 5x + 3 = 0using bisection method.
Solution:
Since f (4) is negative and f (5) is positive, a root lies between 4 and 5 and therefore
we take x0 = 4.5. Then f (x0 ) = 0.75 which is positive.
Hence the root lies between 4 and 4.5 and we obtain
x1 =

4 + 4.5
= 4.25
2

we nd f (x1 ) = 0.2656 which is negative. We therefore conclude that the root lies
between 4.25 and 4.5.
Proceeding this way, the following table is obtained

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

n.
1
2
3
4
5
6
7
8
9
10
11
12

a
4
4
4.25
4.25
4.25
4.2813
4.2969
4.2969
4.3008
4.3027
4.3042
4.3044

b
5
4.5
4.5
4.375
4.3125
4.3125
4.3125
4.3047
4.3047
4.3047
4.3047
4.3047

x
4.5
4.25
4.375
4.3125
4.2812
4.2969
4.3047
4.3008
4.3027
4.3037
4.3044
4.3046

15

f(x)
0.75
-0.1875
0.2656
0.0352
-0.0771
-0.0212
0.0069
-0.0072
-0.0001
0.0033
0.0060
0.0065

At n = 12, it is seen that the dierence between two successive iterates is 0.0005
which is less than 0.001. Thus x = 4.3046 is the approximate root of the given
equation.
Example 2.2. Find the root of the equation ex = sinx using bisection method.
Solution:
We have f (x) = ex sinx = 0. Since f (0) is positive and f (1) is negative, a root
lies between 1 and 0 and therefore we take x0 = 0.5. Then f (x0 ) = 0.1271 which is
positive.
Hence the root lies between 1 and 0.1271 and we obtain
Proceeding this way, the following table is obtained

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

n
1
2
3
4
5
6
7
8
9
10
11

a
1
1
0.75
0.625
0.625
0.5938
0.5938
0.5938
0.5898
0.5898
0.5889

b
0
0.5
0.5
0.5
0.5625
0.5625
0.5781
0.5859
0.5859
0.5879
0.5879

x
0.5
0.75
0.625
0.5625
0.5937
0.5781
0.5859
0.5898
0.5879
0.5889
0.5884

16

f(x)
0.1271
-0.2093
-0.0498
0.0365
-0.0072
0.0145
0.0036
-0.0018
0.0009
-0.0004
0.0002

At n = 11, it is seen that the absolute dierence between two successive iterates
is 0.0006 which is less than 0.001. Thus x = 0.5884 is the approximate root of the
given equation.
Exercise:
Using Bisection method nd root (correct up to three decimal places) of the following
equations.
1. x3 + x2 + x + 7 = 0

(2.105)

2. x3 x 4 = 0

(1.79604)

3. x3 x2 1 = 0

(1.466)

4. x3 3x 5 = 0

(2.279)

5. x3 5x + 3 = 0

(0.657)

2.4

The Method of False Position:

This is the oldest method for nding the real root of a nonlinear equation y = f (x) =
0. In this method rstly we choose two points a and b such that f (a) and f (b) have

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

17

opposite signs. Hence,a root must lies between these points. Now, the equation of
the chord joining the two points A(a, f (a) and B(b, f (b)) is given by
y f (a) f (b) f (a)
=
xa
ba

(2.7)

Suppose chord represented by above equation intersects X axis at x0 and assumex0

to be root of f (x) = 0. Therefore,y = 0 when x = x0 .Substituting x = x0 and y = 0
in (2.3.1) and simplifying we get,
x0 = a

f (a)
af (b) bf (a)
=
f (b) f (a)
f (b) f (a)

(2.8)

This formula gives rst approximate root of f (x) = 0.

We use following steps to nd a approximate root of f (x) = 0 by False position
(Regula falsi) method:
1. Choose real numbers a and b satisfying a < b in I such that f (a)f (b) < 0.
2. The rth approximate root of the equation is given by,
xr = a

f (a)
af (b) bf (a)
=
,
f (b) f (a)
f (b) f (a)

r = 0, 1, 2, .

(2.9)

3. If f (xr ) = 0 then xr is a root/solution of the equation f (x) = 0.Stop the

process.
4. If f (xr ) = 0 then
(a) If f (a)f (xr ) < 0 then root lies in the interval (a, xr ).Then take b = xr and
go to step 2 above.
(b) If f (xr )f (b) < 0 then root lies in the interval (xr , b).Then take a = xr and
go to step 2 above.
5. Repeat steps 2 , 3 and 4 till we get root correct to required accuracy or till
getting required number of iterations.
Example 2.3. Find the root of the equation x3 x2 1 = 0 using False position
method.

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

18

Solution:
Since f (1) is negative and f (2) is positive, a root lies between 1 and 2 and therefore
we take x0 = 1 and x1 = 2. Now by the formula
f (x0 )
(x1 x0 )
f (x1 ) f (x0 )
f (1)
x2 = 1
(2 1)
3 (1)
x2 = 1.25
x2 = x0

Now f (x2 ) = 0.6094 which is negative so root lies between1.25 and 2 Therefore
now we take x0 = 1.25 and x1 = 2 and using the formula again we get
f (x0 )
(x1 x0 )
f (x1 ) f (x0 )
f (1.25)
x3 = 1.25
(2 1.25)
3 (0.6094)
x3 = 1.3766
x3 = x0

Now f (x3 ) = 0.2863 which is negative so root lies between1.3766 and 2 Therefore
now we take x0 = 1.25 and x1 = 2 and continuing this process we get following
table.
n
1
2
3
4
5
6
7
8
9

x0
x1
1
2
1.25
2
1.3766 2
1.4309 2
1.4524 2
1.4606 2
1.4637 2
1.4649 2
1.4653 2

xn
f (xn )
1.25
-0.6094
1.3766 -0.2863
1.4309 -0.1177
1.4524 -0.0457
1.4606 -0.0173
1.4637 -0.0065
1.4649 -0.00245
1.4653 -0.0009
1.4655 -0.0003

At n = 9, it is seen that the absolute dierence between two successive iterates

is 0.0001 which is less than 0.001. Thus x = 0.4655 is the approximate root of the
given equation.

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

19

Exercise:
1. Using False position method nd root, correct to three decimal places of the
following equations.
(a) x3 x 4 = 0 (1.796)
(b) x3 x 1 = 0 (1.325)
(c) x3 + x 1 = 0 (0.682)
(d) x3 + x2 + x + 7 = 0
(e) x3 x 4 = 0
(f) x3 x2 1 = 0
2. Evaluate

(a) 2 18

(b) 3 13

(c) 2 10

(d) 4 72

2.5

the following by using False position method.

(3.1622)
(4.2426)
(2.3513)
(2.9129)

The Iteration Method :

In previous root nding methods, we use end points of a interval containing a root
to nd an approximate root of the given equation.In iteration method we take value
in a interval containing a root as initial approximate root of the given equation and
improve it to required accuracy.
Let f (x) = 0 be given equation. In iteration method to nd a root of equation
f (x) = 0,we use following steps:
1. Using location of root result nd an interval I containing a root of the equation.
2. Rewrite equation f (x) = 0 as x = (x) where (x) < 1 for each x I.

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

20

3. Take x0 I as an initial approximate root and obtain sequence of approximate

roots by using
xn+1 = (xn ), n = 0, 1, 2,

(2.10)

This sequence converges to the root of the equation.

4. Stop the process of obtaining a sequence when we get approximate root correct
to required accuracy.
Theorem 2.4. Let be a root of the equation f (x) = 0 in an interval I, x0 I be
an approximate root of the equation and x = (x) be an equivalent expression of the
equation f (x) = 0 such that
1. (x) and (x) are continuous on I.
2. | (x) < 1| for all x I.
Then, sequence obtained by relation
xn+1 = (xn ), n = 0, 1, 2,

(2.11)

converges to the root .

Proof: Since is a root of the equation f (x) = 0 x = (x), therefore we have
= ()

(2.12)

x1 = (x0 )

(2.13)

From (2.4.2),

Subtracting (2.4.4) from (2.4.3) we get,

x1 = () (x0 )

(2.14)

Now without loss of generality assume x0 < then by hypothesis (x) satisfy both
conditions of Lagranges mean value theorem on the interval [x0 , ] I. Therefore
there exist 0 (x0 , ) such that
() (x0 )
x0

( x0 ) (0 ) = () (x0 )
(0 ) =

(2.15)
(2.16)

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

21

From (2.4.5) and (2.4.7) we get,

x1 = ( x0 ) (0 ), x0 < 0 <

(2.17)

Similarly we obtain
x2 = ( x1 ) (1 ), x1 < 1 <

(2.18)

x3 = ( x2 ) (2 ), x2 < 2 <
..
.

(2.19)

xn+1 = ( xn ) (n ), xn < n <

(2.20)

Now | (x)| < 1 implies

| (i )| r < 1, i, for some r (0, 1).

(2.21)

Using this relation in equations (2.4.8) to (2.4.12) we get

| x0 | | x1 | | x2 | | xn | | xn+1 |.

(2.22)

This shows that all successive approximations lies in I.

Now multiplying equations (2.4.8) to (2.4.12) we get
xn+1 = ( x0 ) (0 ) (1 ) (2 ) (n )

(2.23)

From (2.4.12) and (2.4.14) we get,

| xn+1 | rn+1 |( x0 )| 0 as n . ( r (0, 1).)

(2.24)

Thus sequence of approximations x0 , x1 , x2 , converges to the root . 

Example 2.4. Find the root of the equation 3x = cosx + 1 using iteration method.
Solution:

3x = cosx + 1
1
(x) = x = (cosx + 1)
3
|sinx| 1
| (x)| =
<
3
3

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

22

Hence the iteration method can be applicable.

We start with x0 = 2 . Therefore x1 = (x0 ) = 13 and f (x1 ) = 0.9451 and
x2 = (x1 ) = 0.6483. Proceeding this way, the following table() is obtained

n
0
1
2
3
4
5

xn
0
0.3333
0.6483
0.5990
0.6086
0.6068

xn+1
0.3333
0.6483
0.5990
0.6086
0.6068
0.6072

f (xn+1 )
-0.9451
0.1479
-0.0288
0.0054
-0.0010
0.0002

At n = 5, it is seen that the dierence between two successive iterates is 0.0003

which is less than 0.001. Thus x = 0.6072 is the approximate root of the given
equation.
1

Example 2.5. Find the root of the equation x = (5 x) 3 using iteration method.
Solution:
1

(x) = x = (5 x) 3
2
1
| (x)| = (5 x) 3 < 1
3
Hence the iteration method can be applicable.
We start with x0 = 1. Therefore x1 = (x0 ) = 1.5874 and f (x1 ) = 0.0819 and
x2 = (x1 ) = 1.5055. Proceeding this way, the following table is obtained

n
0
1
2
3
4

xn
1
1.5874
1.5055
1.5175
1.5158

xn+1 f (xn+1 )
1.5874 0.0819
1.5055 -0.0119
1.5175 0.0017
1.5158 -0.0003
1.5160 0.0001

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

23

At n = 4, it is seen that the dierence between two successive iterates is 0.0004

which is less than 0.001. Thus x = 1.516 is the approximate root of the given equation.

Exercise:
Using Iteration method nd a root, correct to 4 signicant gures of the following
equations.
1. x =

1
(x+1)2

(0.4655)

2. xsinx = 1 (1.068)
3. 5x3 20x + 3 = 0 (0.1514)
4. sin2 x = x2 1 (1.404)
5. ex = cotx (0.5314)
6. ex = 10x (0.0913)
7. x3 = x2 + 1 (1.466)
8. sinx = 10(x 1) (1.088)
2.5.1

The Aitkens 2 -process :

Let xi1 , xi , xi+1 be three successive approximations to the desired root of the
equation x = (x) which is equivalent to f (x) = 0 satisfying conditions of iterative
method.Then from iterative method we have,
xi = k( xi1 ), xi+1 = k( xi )

(2.25)

xi1
xi
=
xi+1
xi

(2.26)

(xi+1 xi )2
= xi+1
xi+1 2xi + xi1

(2.27)

Dividing, we obtain

which gives on simplication

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

24

If we now dene xi and 2 xi by the relations

xi = xi+1 xi and 2 xi = (xi ),
then
2 xi1 = (xi1 )
= (xi xi1 )
= xi xi1
= xi+1 xi (xi xi1 )
= xi+1 2xi + xi1
Using this (2.4.18) is written in simpler form as
= xi+1

(xi )2
2 xi1

(2.28)

This explains the term 2 -process. For computation of 2 process we use following
table
x1
x1 = x2 x1
x2
2 x1 = x2 x1
x2 = x3 x1
x3
Example 2.6. Find the root of the equation 3x = cosx + 1 using iteration method
and Aitkens 2 process.
Solution:
As per the table () given above x1 = 0.3333, x2 = 0.6483 and x3 = 0.599 hence
by formula xi = xi+1 xi we get x1 = 0.315, x2 = 0.0493 and 2 x1 =
x2 x1 = 0.3643
By Aitkens 2 process,
(x2 )2
x4 = x3
2 x1
= 0.6057
Which corresponds to 4 normal iteration in iteration method.

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

2.6

25

Newton-Raphson Method:

Let f (x) = 0 be equation such that f (x) possesses all ordered derivatives in an
interval containing its one root say . Let x0 be an approximate root of f (x) = 0
and x1 = x0 + h be the correct root so that f (x1 ) = 0. Expanding f (x0 + h) by
Taylors series we get,
f (x0 + h) = f (x0 ) + hf (x0 ) +

h2
f (x0 ) + = 0
2!

Neglecting second and higher ordered derivatives we get,

f (x0 ) + hf (x0 ) = 0
This gives,
f (x0 )
f (x0 )
Substituting this value in x1 = x0 + h we get better approximation as
h=

x1 = x0

f (x0 )
f (x0 )

Successive approximations are given by formula,

xn+1 = xn

f (xn )
, n = 0, 1, 2, 3,
f (xn )

(2.29)

This formula is known as Newton-Raphson formula.

Note: This is the fastest method to nd a root of the equation among all methods
discussed in this chapter .
Example 2.7. By Newton-Raphson method nd the roots of the equation x4 x5 =
0 between 1 and 2.

f (x) = x4 x 5
f (x) = 4x3 1
f (1) = 5 and f (2) = 9

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

Therefore root lies between 1 and 2.

We choose x0 = 1.3
We have Newton-Raphsons formula
f (xi )
i = 0, 1, 2, ....
f (xi )
f (x0 )
x1 = x0
f (x0 )
f (1.3)
= 1.3
f (1.3)
(1.3)4 1.3 5
= 1.3
4(1.3) 1
3.4439
= 1.3
7.788
= 1.7422

xi+1 = xi

26

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

f (x1 )
f (x1 )
f (1.7422)
= 1.7422
f (1.7422)
2.4706
= 1.7422
20.1521
= 1.6196
f (x2 )
x3 = x2
f (x2 )
f (1.6196)
= 1.6196
f (1.6196)
4.2611
= 1.6196
15.9035
= 1.3532
f (x3 )
x4 = x3
f (x3 )
f (1.3532)
= 1.3532
f (1.3532)
0.9999
= 1.3532
8.9117
= 1.241
x2 = x1

1.241 is the desired approximate root.

Example 2.8. By Newton-Raphson method evaluate

11.

We have to nd a number x such that

3
x = 11
f (x) = x3 11 = 0
f (x) = 3x2

We know that 3 8 = 2 and 3 27 = 3 so 3 11 lies between 2 and 3.

27

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

we take initial value x0 = 2.1

Thus

f (x0 )
f (x0 )
1.739
= 2.1
13.23
= 2.2314
f (x1 )
x2 = x1
f (x1 )
0.1105
= 2.2314
14.9374
= 2.224
x1 = x0

3
11 2.224

Exercise:
1. Evaluate

(a) 2 18

(b) 3 13

(c) 2 10

(d) 4 72

the following by using Newton-Raphson method.

(3.1622)
(4.2426)
(2.3513)
(2.9129)

2. Using Newton-Raphson method nd root of the following equations.

(a) x2 5x + 3 = 0 (3.3548)
(b) x5 + 5x + 1 = 0 (0.1999)
(c) x3 x 4 = 0 (1.7963)
(d) sinx = (1/2)x (1.896)
(e) x + log x = 2 (1.756)
(f) (1/2)sinx = xe2x

(3.130)

3. Obtain Newton-Raphson formula to nd inverse of a given number.

28

CHAPTER 2. SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS

29

4. Find the root of the equation ex sinx = 0 which lies between 0 and
1 (0.58867).
5. Find the root of the equation 3x = cosx + 1.

(0.6072)

6. Establish the formula

N)
1(
xi+1 = xi +
2
xi

and hence compute the value of 2 correct up to six decimal places.

(1.414214)

7. Find the leasr positive root of the equation tanx = x to an accuracy of 0.0001
by Newton-Raphson method.
(4.4934)