28 CHAPTER 2: Solution of Algebraic and Transcendental Equations

2.3 METHOD OF FALSE POSITION

This is the oldest method for finding the real root of a nonlinear equation
f (x) = 0 and closely resembles the bisection method. In this method, also
known as regula–falsi or the method of chords, we choose two points a and
b such that f (a) and f (b) are of opposite signs. Hence, a root must lie in
between these points. Now, the equation of the chord joining the two points
[a, f (a)] and [b, f (b)] is given by
y − f (a) f (b) − f (a )
= . (2.6)
x−a b−a
The method consists in replacing the part of the curve between the points
[a, f (a)] and [b, f (b)] by means of the chord joining these points, and taking
the point of intersection of the chord with the x-axis as an approximation to
the root. The point of intersection in the present case is obtained by putting
y = 0 in Eq. (2.6). Thus, we obtain
f (a) af (b) − bf (a)
x1 = a − (b − a ) = , (2.7)
f (b) − f (a) f (b) − f ( a)
which is the first approximation to the root of f (x) = 0. If now f (x1) and
f (a) are of opposite signs, then the root lies between a and x1, and we
replace b by x1 in Eq. (2.7), and obtain the next approximation. Otherwise, we
replace a by x1 and generate the next approximation. The procedure is repeated
till the root is obtained to the desired accuracy. Figure 2.2 gives
a graphical representation of the method. The error criterion Eq. (2.5) can be
used in this case also.
Y
y = f (x)
* B [b, f (b)]
O x1 Y
x2 X

*
A [a, f (a)]
Figure 2.2 Method of false position.

Example 2.6 Find a real root of the equation:

f (x) = x3 – 2x – 5 = 0.

We find f (2) = –1 and f (3) = 16. Hence a = 2, b = 3, and a root lies

between 2 and 3. Equation (2.7) gives
 SECTION 2.3: Method of False Position 29

2(16) − 3(−1) 35
x1 = = = 2.058823529.
16 − (−1) 17

Now, f (x1) = –0.390799917 and hence the root lies between 2.058823529 and
3.0. Using formula (2.7), we obtain

2.058823529(16) − 3(−0.390799917)
x2 = = 2.08126366.
16.390799917

Since f (x2) = –0.147204057, it follows that the root lies between 2.08126366
and 3.0. Hence, we have

2.08126366(16) − 3(−0.147204057)
x3 = = 2.089639211.
16.147204057
Proceeding in this way, we obtain successively:
x4 = 2.092739575, x5 = 2.09388371,
x6 = 2.094305452, x7 = 2.094460846,…
The correct value is 2.0945…, so that x7 is correct to five significant figures.

Example 2.7 Given that the equation x2.2 = 69 has a root between 5 and 8.
Use the method of regula–falsi to determine it.
Let f (x) = x2.2 – 69. We find

f (5) = –34.50675846 and f (8) = 28.00586026.

Hence
5(28.00586026) − 8(−34.50675846)
x1 = = 6.655990062.
28.00586026 + 34.50675846

Now, f (x1) = –4.275625415 and therefore, the root lies between 6.655990062
and 8.0. We obtain
x2 = 6.83400179, x3 = 6.850669653.
The correct root is 6.8523651…, so that x3 is correct to three significant
figures.

Example 2.8 The equation 2x = log10 x + 7 has a root between 3 and 4. Find
this root, correct to three decimal places, by regula–falsi method.
Let
f (x) = 2x – log10 x – 7, a = 3 and b = 4.
Then we find
f (3) = –1.4771 and f (4) = 0.3979.
30 CHAPTER 2: Solution of Algebraic and Transcendental Equations

Hence
af (b) − bf (a)
x1 =
f (b) − f (a)
3(0.3979) − 4(−1.4771)
=
0.3979 + 1.4771
7.1021
= = 3.7878.
1.8750
Therefore, the root lies between 3 and 3.7878. Now, we take a = 3 and b = 3.7878
Then,
f (b) = 2(3.7878) – log10 3.7878 – 7 = –0.002787
Hence,
3( 0.002787)  3.7878(–1.4771)
x2 
– 0.002787  1.4771
 3.7893,
and
f ( x2 ) = 2(3.7893) − log10 (3.7893) − 7
= 0.000041,
which shows that x = 3.789 is the root correct to three decimal places.

Example 2.9 Find a root of the equation 4e–x sin x – 1 = 0 by regular–falsi

method given that the root lies between 0 and 0.5.
Let
f (x) = 4e–x sin x – 1, a = 0, b = 0.5.
We have
f (a) = –1 and f (b) = 4e–0.5 sin 0.5 – 1 = 0.163145
Therefore,
0(0.163145)  0.5( 1)
x1 
1.163145
0.5
  0.4298690
1.163145
Now, we take
a = 0 and b = 0.4298690
Then
f (x) = 0.08454
Therefore,
0(0.08454)  0.42987( 1)
x2 
1.08454
 0.39636
Now,
a = 0, b = 0.39636 and f (b) = 0.038919
 SECTION 2.4: Iteration Method 31

Hence
0(0.038919)  0.39636( 1)
x3 
1.038919
 0.381512,
and
f (x3) = 0.016934
Taking a = 0 and b = 0.381512, we obtain
0(0.016934) − 0.381512(–1)
x4 =
1.016934
= 0.375159,
and
f (x4) = 0.0071873
Proceeding as above, we obtain
x5 = 0.37248, x6 = 0.37136,
x7 = 0.37089, x8 = 0.370697
It follows that the required root is 0.371, correct to three decimal places.

2.4 ITERATION METHOD

We have so far discussed root-finding methods which require an interval in
which the root lies. We now describe methods which require one or more
approximate values to start the solution and these values need not necessarily
bracket the root. The first is the iteration method which requires one starting
value of x.
To describe this method for finding a root of the equation
f (x) = 0, (2.1)
we rewrite this equation in the form
x = f (x) (2.8)
There are many ways of doing this. For example, the equation
x3 + x2 – 2 = 0
can be expressed in different forms
2
x= , x = 2 − x3 , x = (2 − x 2 )1/3 , etc.
1+ x
Now, let x0 be an approximate root of Eq. (2.8). Then, substituting in
Eq. (2.8), we get the first approximation as
x1 = f (x0)
Successive substitutions give the approximations
x2 = f (x1), x3 = f (x2), … , xn = f (xn–1).
68 CHAPTER 2: Solution of Algebraic and Transcendental Equations

EXERCISES

2.1 Explain the bisection method for finding a real root of the equation
f (x) = 0 and write an algorithm for its implementation with a test for
relative accuracy of the approximation.
Obtain a root, correct to three decimal places, of each of the following
equations using the bisection method (Problems 2.2–2.5):
2.2 x3 – 4x – 9 = 0
2.3 x3 + x2 – 1 = 0
2.4 5x log10 x – 6 = 0
2.5 x2 + x – cos x = 0
2.6 Give the sequence of steps in the regula–falsi method for determining
a real root of the equation f (x) = 0.
Use the method of false position to find a real root, correct to three decimal
places, of the following equations (Problems 2.7–2.10):
2.7 x3 + x2 + x + 7 = 0
2.8 x3 – x – 4 = 0
2.9 x = 3e–x
2.10 x tan x + 1 = 0
2.11 Find the real root, which lies between 2 and 3, of the equation
x log10 x – 1.2 = 0
using the methods of bisection and false–position to a tolerance of
0.5%.
2.12 Explain briefly the method of iteration to compute a real root of the
equation f (x) = 0, stating the condition of convergence of the sequence
of approximations. Give a graphical representation of the method.
Use the method of iteration to find, correct to four significant figures, a real
root of each of the following equations (Problems 2.13–2.16):
2.13 ex = 3x

1
2.14 x =
( x + 1)2
2.15 1 + x2 = x3

1
2.16 x – sin x =
2
2.17 Establish an iteration formula to find the reciprocal of a positive number
N by Newton–Raphson method. Hence find the reciprocal of 154 to
four significant figures.
 EXERCISES 69

2.18 Explain Newton–Raphson method to compute a real root of the equation

f (x) = 0 and find the condition of convergence. Hence, find a non-zero
root of the equation x2 + 4sin x = 0.
2.19 Using Newton–Raphson method, derive a formula for finding the kth
root of a positive number N and hence compute the value of (25)1/4.
Use the Newton–Raphson method to obtain a root, correct to three decimal
places, of each of the following equations (Problems 2.20–2.25):
2.20 xsin2 – 4 = 0
2.21 ex = 4x
2.22 x3 – 5x + 3 = 0
2.23 xex = cos x
1 + cos x
2.24 x =
3
2.25 cot x = –x
2.26 Describe a computational procedure to implement Newton–Raphson
method for computing the square root of a positive number to an accuracy
e. Write a flow-chart for the same.
2.27 Compute, to four decimal places, the root between 1 and 2 of the
equation
x3 – 2x2 + 3x – 5 = 0
by (a) Method of False Position and (b) Newton–Raphson method.
Using Ramanujan’s method, find the smallest root of each of the following
equations (Problems 2.28–2.30):
2.28 x3 – 6x2 + 11x – 6 = 0
2.29 x + x3 – 1 = 0
2.30 sin x + x –1 = 0
2.31 Use the secant method to determine the root, between 5 and 8, of
the equation x2.2 = 69. Compare your result with that obtained in
Example 2.7.
2.32 Determine the real root of the equation x = e–x, using the secant method.
Compare your result with that obtained in Example 2.26.
2.33 Point out the difference between the regula–falsi and the secant methods
for finding a real root of f (x) = 0. Apply both the methods to find a real
root of the equation x3 – 2x – 5 = 0 to an accuracy of 4 decimal places.
2.34 Describe briefly Muller’s method and use it to find (a) the root, between
2 and 3, of the equation x3 – 2x – 5 = 0 and (b) the root, between 0 and
1, of the equation x = e–x cos x.
 EXERCISES 71

2.45 To find the smallest root of the equation f (x) = x3 – x – 1 = 0 by the

iteration method, f (x) = 0 should be rewritten as
(a) x = x3 – 1 (b) x = (x + 1)1/3
1 x +1
(c) x = 2 (d) x =
x –1 x2
Find the correct choice in the above.
2.46 Newton–Raphson formula converges if

f ′( x) f ′′( x) f ( x) f ′′( x)
(a) <1 (b) <1
[ f ( x) ]
2
[ f ′( x)]2

f ( x) f ′( x)
(c) <1 (d) None of these.
[ f ′′( x)]2
2.47 Which one of the following is not correct?
(a) Newton–Raphson method has quadratic convergence.
(b) The bisection method converges slowly.
(c) To solve f (x) = 0 by iteration method, the given equation is written in
the form x = f (x) where | f ¢(x)| < 1 in an interval containing the root.
(d) The method of regula–falsi converges faster than the secant method.
2.48 Which one of the following is correct?
(a) The bisection method has quadratic convergence.
(b) The iteration method is a self-correction method.
(c) The equation x3 – 2x – 5 = 0 has two positive roots.
(d) The equation x = cos x can be solved by Graeffe’s method.

Answers to Exercises

2.2 2.706 2.3 0.755

2.4 2.741 2.5 0.550
2.7 –2.105 2.8 1.796
2.9 1.0499 2.10 2.798
2.11 Successive values are (By bisection method)
2.5, 2.75, 2.625, 2.6875, 2.7187, 2.7344, 2.7422.
By false position method
2.7210, 2.7402, 2.7407, 2.7406, 2.7406
2.13 0.6190 2.14 0.4656
2.15 1.466 2.16 1.497
2.17 0.006494 2.18 –1.9338
2.19 2.23607 2.20 4.5932
72 CHAPTER 2: Solution of Algebraic and Transcendental Equations

2.21 0.3574
2.22 With x0 = 0.5, x1 = 0.6470, x2 = 0.65656, …
2.23 With x0 = 0.5, x1 = 0.5180, x2 = 0.5180, …
2.24 x0 = 1.047, x1 = 0.6224, x2 = 0.6071, x3 = 0.6071.
2.25 2.798
2.27 1.8437; 1.8438
b1 b2
2.28 = 0.54546, = 0.77647,
b2 b3
b3
b4
= 0.88696,
b4
b5
 0.94237, |
Smallest root is 1.
2.29 Convergents are:
1.0, 0.5, 0.66666, 0.75, 0.66666, 0.66666, 0.69231, 0.68421, …
2.30 Convergents are:
0.5, 0.5, 0.51064, 0.51087, 0.51097, …
Required root is 0.5110.
2.31 Root = 6.85236.
2.32 Fifth iteration value = 0.567143
2.33 Regula–falsi: 2.0945 in 7 iterations.
Secant method: 2.0946.
2.34 (a) 2.09462409 (b) 0.51752
2.35 Third approximation to the roots: 0.9168, 1.0897, 2.0019
2.36 Third approximation: 3.014443, 1.991425, 0.999494.
2.37 x2 – 0.0165x + 0.9394 (Exact factor is x2 + 1).
2.38 x2 – 1.9485x + 1.942982 (Exact factor is x2 – 2x + 2).
2.39 Q1 » 3.13145, Q2 » 1.898, Q3 » 0.9706.
2.42 x1 = 0.8261, y1 = 0.5636.
2.43 x0 = 3, y0 = 2, x1 = 3.1667, y1 = 2.5
2.44 x = 3.584, y = –1.848
2.45 (b)
2.46 (b)
2.47 (d)
2.48 (a)
52 NUMERICAL ANALYSIS

2.7 REGULA-FALSI METHOD

Consider the equation f(x) = 0 and let a, b be two values of x such that f(a) and f(b) are of opposite
signs. Also let a < b. The graph of y = f(x)will meet the x-axis at the same point between a and
b. The equation of the chord joining the two points [a, f(a)] and [b, f(b)] is
y− f a bg
=
f b − f a bg bg (17)
x−a b−a
In the small interval (a, b) the graph of the function can be considered as a straight line. So
that x-coordinate of the point of intersection of the chord joining [a, f(a)] and [b, f(b)] with the x-
axis will give an approximate value of the root. So putting y = 0 in (17) we get

−
bg
f a
=
bg bg
f b − f a
x − x1 b−a

b g bb − ag ,
f a
or x=a−
bg bg
f a − f a

af bbg − bc f ba gh
f bbg − f ba g
or x= = x (say).
0

If f(a) and f(x0) are of apposite signs then the root lies between a and x0 otherwise it lies
between x0 and b.
If the root lies between a and x0 then the next approximation

b g bg
af x0 − x0 f a
f b x g − f ba g
x1 =
0

x f bbg − bf b x g
f bbg − f b x g
otherwise x1 = 0 0
.
0

The above method is applied repeatedly till the desired accuracy is obtained.
The Geometrical interpretation of the method is as follows In Fig. 2.4, the curve y = f(x)
between A(x = a) and B(x = b) cuts OX at Q. The chord AB cuts OX at P. It is clear that x = OQ
is the actual value of the root whereas x = OP = x0 is the first approximation to the root f(x0) and
f(a) are of opposite signs. So we apply the false position method to the interval (a, x0) and get OP1
the next approximation to the root. The procedure is continued till the root is obtained to the desired
degree of accuracy. The points of intersection of the successive chords with x-axis, namely P1,
P2, … tend to coincide with Q the point where the curve y = f(x) cuts the x-axis and so we get
successive approximate values of the root of the equation.
SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS 53

y = f(x)
A [a, f(a)]

a P2 P Px0 Db
x1 X
O C Q
[b, f(b)]

B
Fig. 2.4

Example 2.25 Find an approximate value of the root of the equation x3 + x – 1 = 0 near x = 1, by the method of
Falsi using the formula twice.
Solution Here f(x) = x3 + x – 1
f(0.5) = – 0.375, f(x) = 1
Hence the root lies between 0.5 and 1.
We take a = 0.5, b = 1

b g b g b gb g b g
af b − bf a 0.5 1 − 1 −0.375
x0 =
bg bg
f b − f a
=
b g
1 − −0.375
= 0.64.

Now f(0.64) = – 0.0979 < 0

∴ The root lies between 0.64 and 1,
applying the formula again we get

b0.64gb1g − b1gb−0.0979g = 0.7379 = 0.672.

1 − b −0.0979g
∴ x1 =
10979
.

Example 2.26 Find the real root of the equation x log10 x – 1.2 = 0 correct to five decimal places by Regula–Falsi
method using the formula four times.
Solution Here f(x) = x log10 x – 1.2
f(2) = –0.6, f(3) = 0.23.
Thus the root lies between a and 3 and it is nearer to 3.
54 NUMERICAL ANALYSIS

We take a = 2, b = 3.
Let x0 denote the first approximation

bg bg b g b g
af b − bf a 2 0.23 − 3 −0.6 2.26
∴ x0 =
bg bg
f b − f a b g
=
0.23 − −0.6
=
0.83
= 2.72.

Now f(2.7) = – 0.04

The root lies between 2.7 and 3.
We note that f(2.8) = 0.05

b2.7gb0.05g − 2.8b−0.04g = 0.247 = 2.74,

0.05 − b −0.04g
∴ x1 =
0.09
since f(2.74) = –0.0006
∴ Root lies between 2.74 and 2.8.
But f(2.75) = 0.0081

b2.84gb0.0081g − b2.75gb−0.0006g = 0.023844 = 2.7407,

0.0081 − b −0.0009g
∴ x2 =
0.0087

since f(2.7407) = 0.000045, the root is < 2.7407,

but f(2.7407) = –0.000039

b2.7406gb0.000045g − b2.7407gb− 0.000039g = 0.0002301

0.000045 – b −0.000039g
∴ x3 =
0.000084

= 2.7392
is the required value.
Example 2.27 Solve the equation x tan x = –1, by Regula falsi method starting with 2.5 and 3.0 as the initial
approximations to the root.
Solution. We have f(x) = x tan x + 1
f(a) = f(2.5) = 2.5 tan (2.5) + 1 = – 0.8675
f(b) = f(3) = 3 tan 3 + 1 = 0.5724
By regula falsi method, the first approximation is given by
af (b) − bf ( a )
x1 = f (b ) − f ( a )

3 f (2.5) − 2.5 f ( 3) 3 ( −0.8675 ) − 2.5 ( 0.5724 )

= =
f ( 2.5) − f ( 3) −0.8675 − 0.5724
= 2.8012
Now, f(x1) = f(2.8012) = 2.8012 tan (2.8012) + 1 = 0.00787
f(2.5) f(2.8012) < 0
therefore, the root lies between 2.5 and 2.8012.
The second approximation to the root is given by
2.8012 f ( 2.5) − 2.5 f ( 2.8012 )
x2 = f ( 2.5) − f ( 2.8012 )
SOLUTION OF ALGEBRAIC AND TRANSCENDENTAL EQUATIONS 55

( 2.8012 ) ( −0.8675 ) − ( 2.5) ( 0.00787 )

=
−0.8675 − 0. 00787
= 2.7984
⇒ f(x2) = f(2.7984) = 2.7984 tan (2.7984) + 1
= 0.000039
⇒ f(2.5) f(2.7984) < 0
The root lies between 2.5 and 2.79 84
∴ The third approximation to the root is given by

2.7984 f ( 2.5) − 2.5 f ( 2. 7984 )

x3 =
f ( 2.5) − f ( 2.7984 )
( 2. 7984 ) ( −0.8675) − ( 2.5) ( 0. 000039 )
= −0.8675 − 0. 000039
= 2.7982
The required root is 2.798.

2.8 MULLER’S METHOD

Muller’s method is an iterative method. It requires three starting points. (x0, f(x0)), (x1, f(x1)) and
(x2, f(x2)). A parabola is constructed that passes through these points then the quadratic formula is
used to find a root of the quadratic for the next approximation.
Without loss of generality we assume that x2 is the best approximation to the root and consider
the parabola through the three starting values as shown in Fig. 2.5.
Make the change of variable using the differences
t = x – x2 (18)
using the differences h0 = x0 – x2 and h1 = x1 – x2. (19)
Consider the quadratic polynomial involving t
y = at2 + bt + c (20)
each point is used to obtain an equation involving a, b and c

at t = h0 ; ah02 + bh0 + c = f 0 (21)

at t = h1 ; ah12 + bh1 + c = f 1 (22)

at t = 0; a 02 + b 0 + c = f 2 (23)
from equation (23) we get c = f2. Substituting c = f2 in (21) and (22) and using
e0 = f0 – c, e1 = f1 – c,
we get ah02 + bh0 = f 0 − c = e0 , (24)

ah12 + bh1 = f 1 − c = e1 (25)

 CHAPTER 1

Transcendental and Polynomial

Equations
1.1 INTRODUCTION
We consider the methods for determining the roots of the equation
f (x) = 0 (1.1)
which may be given explicitly as a polynomial of degree n in x or f (x) may be defined
implicitly as a transcendental function. A transcendental equation (1.1) may have no root,
a finite or an infinite number of real and / or complex roots while a polynomial equation (1.1)
has exactly n (real and / or complex) roots. If the function f (x) changes sign in any one of
the intervals [x* – ε, x*], [x*, x* + ε], then x* defines an approximation to the root of f (x)
with accuracy ε. This is known as intermediate value theorem. Hence, if the interval [a, b]
containing x* and ξ where ξ is the exact root of (1.1), is sufficiently small, then
| x* – ξ | ≤ b – a
can be used as a measure of the error.
There are two types of methods that can be used to find the roots of the equation (1.1).
(i) Direct methods : These methods give the exact value of the roots (in the absence of
round off errors) in a finite number of steps. These methods determine all the roots at
the same time.
(ii) Iterative methods : These methods are based on the idea of successive approximations.
Starting with one or more initial approximations to the root, we obtain a sequence of
iterates {xk } which in the limit converges to the root. These methods determine one or
two roots at a time.
Definition 1.1 A sequence of iterates {xk } is said to converge to the root ξ if
lim | xk – ξ | = 0.
k→∞
If xk, xk–1, ... , xk–m+1 are m approximates to a root, then we write an iteration method in
the form
xk + 1 = φ(xk, xk–1, ... , xk–m+1) (1.2)
where we have written the equation (1.1) in the equivalent form
x = φ(x).
The function φ is called the iteration function. For m = 1, we get the one-point iteration
method
xk + 1 = φ(xk ), k = 0, 1, ... (1.3)

1
 2 Numerical Methods : Problems and Solutions

If φ(x) is continuous in the interval [a, b] that contains the root and | φ′(x) | ≤ c < 1 in
this interval, then for any choice of x0 ∈ [a, b], the sequence of iterates {xk } obtained from (1.3)
converges to the root of x = φ(x) or f (x) = 0.
Thus, for any iterative method of the form (1.2) or (1.3), we need the iteration function
φ(x) and one or more initial approximations to the root.
In practical applications, it is not always possible to find ξ exactly. We therefore attempt
to obtain an approximate root xk + 1 such that
| f (xk+1 ) | < ε (1.4)
and / or | xk+1 – xk | < ε (1.5)
where xk and xk + 1 are two consecutive iterates and ε is the prescribed error tolerance.
Definition 1.2 An iterative method is said to be of order p or has the rate of convergence p, if
p is the largest positive real number for which
| εk + 1 | ≤ c | εk |p (1.6)
where εk = xk – ξ is the error in the kth iterate.
The constant c is called the asymptotic error constant. It depends on various order
derivatives of f (x) evaluated at ξ and is independent of k. The relation
εk + 1 = cε kp + O(ε kp + 1)
is called the error equation.
By substituting xi = ξ + εi for all i in any iteration method and simplifying we obtain the
error equation for that method. The value of p thus obtained is called the order of this method.

1.2 ITERATIVE METHODS FOR SIMPLE ROOTS

A root ξ is called a simple root of f (x) = 0, if f (ξ) = 0 and f ′(ξ) ≠ 0. Then, we can also write
f (x) = (x – ξ) g(x), where g(x) is bounded and g(ξ) ≠ 0.

Bisection Method
If the function f (x) satisfies f (a0) f (b0) < 0, then the equation f (x) = 0 has atleast one real
1
root or an odd number of real roots in the interval (a0, b0). If m1 = 2
(a0 + b0) is the mid point of
this interval, then the root will lie either in the interval (a0, m1) or in the interval (m1, b0)
provided that f (m1) ≠ 0. If f (m1) = 0, then m1 is the required root. Repeating this procedure a
number of times, we obtain the bisection method
1
mk + 1 = ak + (b – ak), k = 0, 1, ... (1.7)
2 k
(ak , mk + 1 ), if f (ak ) f (mk + 1 ) < 0,
where (ak + 1, bk + 1) =
(mk + 1, bk ), if f (mk + 1 ) f (bk ) < 0 .

We take the midpoint of the last interval as an approximation to the root. This method
always converges, if f (x) is continuous in the interval [a, b] which contains the root. If an error
tolerance ε is prescribed, then the approximate number of the iterations required may be
determined from the relation
n ≥ [log(b0 – a0) – log ε] / log 2.
Transcendental and Polynomial Equations 3

Secant Method
In this method, we approximate the graph of the function y = f (x) in the neighbourhood
of the root by a straight line (secant) passing through the points (xk–1, fk–1) and (xk, fk), where
fk = f (xk) and take the point of intersection of this line with the x-axis as the next iterate. We
thus obtain
x k − xk − 1
xk + 1 = x k − fk , k = 1, 2, ...
f k − f k− 1
xk − 1 f k − x k f k − 1
or xk + 1 = , k = 1, 2, ... (1.8)
fk − fk −1
where xk–1 and xk are two consecutive iterates. In this method, we need two initial approxima-
tions x0 and x1. This method is also called the chord method. The order of the method (1.8) is
obtained as
1
p = (1 + 5 ) ≈ 1.62 .
2
If the approximations are chosen such that f (xk–1) f (xk ) < 0 for each k, then the method
is known as Regula-Falsi method and has linear (first order) rate of convergence. Both these
methods require one function evaluation per iteration.

Newton-Raphson method
In this method, we approximate the graph of the function y = f (x) in the neighbourhood
of the root by the tangent to the curve at the point (xk, fk) and take its point of intersection with
the x-axis as the next iterate. We have the Newton-Raphson method as
fk
xk + 1 = x k − , k = 0, 1, ... (1.9)
fk ′
and its order is p = 2. This method requires one function evaluation and one first derivative
evaluation per iteration.
Chebyshev method
Writing f (x) = f (xk + x – xk) and approximating f (x) by a second degree Taylor series
expansion about the point xk, we obtain the method
fk 1 f ″
xk + 1 = x k − − ( xk + 1 − xk ) 2 k
fk ′ 2 fk ′
Replacing xk + 1 – xk on the right hand side by (– fk / fk′ ), we get the Chebyshev method
fk 1 fkFG IJ 2
fk ″
xk + 1 = xk − −
fk ′ 2 fk ′H K fk ′
, k = 0, 1, ... (1.10)

whose order is p = 3. This method requires one function, one first derivative and one second
derivative evaluation per iteration.
Multipoint iteration methods
It is possible to modify the Chebyshev method and obtain third order iterative methods
which do not require the evaluation of the second order derivative. We give below two multipoint
iteration methods.