See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/236203809

A New Efficient Optimal Eighth-Order Iterative Method for Solving Nonlinear

Equations

Article · April 2013

Source: arXiv

CITATIONS READS
4 123

2 authors:

Jai Prakash Jaiswal Neha Choubey

Guru Ghasidas Vishwavidyalaya VIT Bhopal University
80 PUBLICATIONS 382 CITATIONS 20 PUBLICATIONS 85 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Neha Choubey on 07 November 2015.

The user has requested enhancement of the downloaded file.

 A NEW EFFICIENT OPTIMAL EIGHTH-ORDER
ITERATIVE METHOD FOR SOLVING NONLINEAR
EQUATIONS
arXiv:1304.4702v1 [math.NA] 17 Apr 2013

J. P. JAISWAL AND NEHA CHOUBEY

Abstract. We established a new eighth-order iterative method, con-

sisting of three steps, for solving nonlinear equations. Per iteration the
method requires four evaluations (three function evaluations and one
evaluation of the first derivative). Convergence analysis shows that this
method is eighth-order convergent which is also substantiated through
the numerical works. Computational results ascertain that our method
is efficient and demonstrate almost better performance as compared to
the other well known eighth-order methods.

Mathematics Subject Classification (2000). 65H05, 65H10,

41A25.

Keywords and Phrases. Nonlinear equations, Order of conver-

gence, Efficiency index, Simple root, Hermite polynomial.

1. Introduction
This paper concerns the numerical solution of non-linear equations
of the general form f (x) = 0. Such equations appear in real world
situations frequently while there is no closed form solution for them.
That is why the numerical solution of these types of equations draw
much attention to itself day by day. One of the common problems
encountered in science and engineering problems is that given a single
variable function f (x), find the values of x for which f (x) = 0. The
root of such nonlinear equations may be real or complex. There are
two general types of methods available to find the roots of algebraic
and transcendental equations. First, direct methods, which are not al-
ways applicable to find the roots, and second, iterative methods based
on the concept of successive approximations. In this case, the general
procedure is to start with one or more initial approximation(s) to the
root and attain a sequence of iterates, which in the limit converges
to the true solution. Multipoint iterative methods for solving nonlin-
ear equations are of great practical importance since they overcome
1
2 J. P. JAISWAL AND NEHA CHOUBEY

theoretical limits of one-point methods concerning the convergence or-

der and computational efficiency. Here, we focus on the simple root
of nonlinear scalar equations by iterative method. Let the function
f : D ⊆ ℜ −→ ℜ be a sufficiently differentiable function and α ∈ D be
a simple root of f (x) = 0. The famous Newton’s method [4] of order
two which can be defined as xn+1 = xn − f (xn )/f ′ (xn ) is one of the
oldest and the most applicable method in the literature. Traub [11]
proposed the concept of efficiency index as a measure for comparing
methods. This index is prescribed by p1/n , where p is the order of
convergence and n is the whole number of evaluations per iteration.
Kung and Traub [9] then presented a hypothesis on the optimality of
roots by giving 2n−1 as the optimal order. This means that the Newton
iteration by two evaluations per iterations is optimal with 1.414 as the
efficiency index. By taking into account the optimality concept many
authors have tried to build iterative methods of optimal higher order
of convergence. Anyway, these schemes are divided into two main cat-
egories. First the derivative free and second, the methods in which
(first, second, ...) derivative evaluation is used per cycle. In this paper
we present a new eighth-order iterative method to find a simple root α
of the nonlinear equation f (x) = 0. We will compare our new method
with well known existing eighth-order methods, namely proposed in
[20], [12], [15], [14], [22], [8], [7], [3], [21], .
This paper is organized as follows: In section 2, we describe the
eighth-order method and prove that the method obtained preserves
their convergence order. This new method agree with the Kung and
Traub conjecture for n = 4. In section 3, we will briefly state the well
known established methods in order to compare the effectiveness of the
new method. Finally, in section 4, the new method is compared in the
performance with some well known eighth-order methods. Numerical
results indicate that the our proposed method give better performance.
Consequently, we have found that the new eighth-order method is con-
sistent, stable and convergent.

2. Development of the method and analysis of

convergence
In this section, we will define a new eighth-order method. In order
to establish the order of convergence of this new method, we state fol-
lowing definitions:
 A NEW EFFICIENT OPTIMAL EIGHTH-ORDER ITERATIVE METHOD. . . 3

Definition 2.1. Let f(x) be a real function with a simple root α and let
xn be a sequence of real numbers that converge towards α. The order
of convergence m is given by
xn+1 − α
lim = ζ 6= 0, (2.1)
n→∞ (xn − α)m

where ζ is the asymptotic error constant and m ∈ R+ .

Definition 2.2. Let β be the number of function evaluations of the new

method. The efficiency of the new method is measured by the concept
of efficiency index [19, 10] and defined as
µ1/β , (2.2)
where µ is the order of the method.

2.1. New Eighth-order Method. In this section, we construct our

eight-order method by considering three step cycles. Kung [16] devel-
oped a one-parameter family of fourth-order methods, which is written
as
f (xn )
yn = xn − ′ , (2.3)
f (xn )
 
f (xn ) + βf (yn ) f (yn )
zn = yn − , (2.4)
f (xn ) + (β − 2)f (yn ) f ′ (xn )
where β is a constant. In particular, the special method for β = −1/2
is as follows:
f (xn )
yn = xn − ′ , (2.5)
f (xn )
 
2f (xn ) − f (yn ) f (yn )
zn = yn − . (2.6)
2f (xn ) − 5f (yn ) f ′ (xn )
To achieve an 8th order method, we use the first and second steps of
the three step cycle from the above method and in the third step we
apply the Newtons iteration
f (zn )
xn+1 = zn − ′ . (2.7)
f (zn )
As we can see, this method consists of three evaluation of the func-
tion and two evaluation of the first derivative per iteration, which has
efficiency index 1.5157. To derive a scheme with a higher efficiency in-
dex, we approximate f ′ (zn ) using a Hermite interpolation polynomial
which is equal to the function f (x) in the domain D (D is the interval
in which f has a simple root).
4 J. P. JAISWAL AND NEHA CHOUBEY

To approximate f ′ (zn ), we construct a Hermite interpolation poly-

nomial, H(x), that meets the interpolation conditions
H(xn ) = f (xn ), H(yn ) = f (yn ), H(zn ) = f (zn ) and H ′ (xn ) = f ′ (xn ).
The H(x) could be defined as follows:
H(x) = ω0 (x)f (xn ) + ω1 (x)f (yn ) + ω2 (x)f (zn ) + ω 0 (x)f ′ (xn ), (2.8)
where the interpolation basis functions ω0 (x), ω1 (x), ω2 (x) and ω 0 (x)
are cubic polynomial function that satisfy in the following conditions
(i) ω0 (xn ) = 1, ω0 (yn ) = 0, ω0(zn ) = 0, ω0′ (xn ) = 0,
(ii) ω1 (xn ) = 0, ω1 (yn ) = 1, ω1(zn ) = 0, ω1′ (xn ) = 0,
(iii) ω2 (xn ) = 0, ω2 (yn ) = 0, ω2(zn ) = 1, ω2′ (xn ) = 0,
(iv) ω 0 (xn ) = 1, ω 0 (yn ) = 0, ω 0 (zn ) = 0, ω ′0 (xn ) = 1. (2.9)
From (i), we have ω0 (x) = A(x − yn )(x − zn )(x − B), where A, B are
constants. The conditions ω0 (xn ) = 1 and ω0′ (xn ) = 0 imply
2xn − zn − yn (xn − yn )(xn − zn )
A=− 2 2
, B = xn + .
(xn − yn ) (xn − zn ) 2xn − zn − yn
Thus
 
(x − xn )(x − zn ) (x − xn )(2xn − zn − yn )
ω0 (x) = 1− . (2.10)
(xn − yn )(xn − zn ) (xn − yn )(xn − zn )
From (ii), we have ω1 (x) = C(x − xn )2 (x − zn ) where C is a constant.
The condition ω1 (yn ) = 1 implies
1
C= .
(yn − xn )2 (yn − zn )
Thus
(x − xn )2 (x − zn )
ω1 (x) = . (2.11)
(yn − xn )2 (yn − zn )
Similarly,
(x − xn )2 (x − yn )
ω2 (x) = . (2.12)
(zn − xn )2 (zn − yn )
From (iv), we have ω(x) = D(x − xn )(x − yn )(x − zn ) where D a
constant. The conditions ω ′0 (xn ) = 1 gives
1
D= .
(xn − yn )(xn − zn )
Hence
(x − xn )(x − yn )(x − zn )
ω 0 (x) = . (2.13)
(xn − yn )(xn − zn )
 A NEW EFFICIENT OPTIMAL EIGHTH-ORDER ITERATIVE METHOD. . . 5

With the specific expressions of the interpolation basis functions ω0 (x),

ω1 (x) ω2 (x), ω 0 (x), we have
 
(x − yn )(x − zn ) (x − xn )(2xn − yn − zn )
H(x) = 1− f (xn )
(xn − yn )(xn − zn ) (xn − yn )(xn − zn )
(x − xn )2 (x − zn ) (x − xn )2 (x − yn )
+ f (yn ) + f (zn )
(yn − xn )(yn − zn ) (zn − xn )2 (zn − yn )
(x − xn )(x − yn )(x − zn ) ′
+ f (xn ),
(xn − yn )(xn − zn )
(2.14)
Differentiating the above equation then putting x = zn and Simplifying
we can get
f (xn ) − f (zn ) f (yn ) − f (zn ) f (xn ) − f (yn )
H ′ (zn ) = 2 + −
xn − zn yn − zn xn − yn
 
(xn − zn ) f (yn ) − f (xn ) (xn − zn ) ′
+ − f (xn ) − f ′ (xn ),
(yn − xn ) (yn − xn ) (yn − xn )
(2.15)
and subsequently, we can find
H ′(zn ) = 2f [xn , zn ]+f [yn, zn ]−f [xn , yn ]+(xn −zn )f [yn , xn , xn ]−f ′ (x),
(2.16)
where f [xn , zn ] (similarly f [yn , zn ] and f [xn , yn ]) and f [yn , xn , xn ] are
defined below by taking into consideration the divided differences:
f (xn ) − f (zn )
f [xn , zn ] = ,
xn − zn
f [yn , xn ] − f ′ (xn )
f [yn , xn , xn ] = .
yn − xn
Now replacing f ′ (zn ) in the equation (2.7) by H ′ (zn ) from the equation
(2.16), we find that
xn+1 = 2f [xn , zn ] + f [yn , zn ] − f [xn , yn ] + (xn − zn )f [yn , xn , xn ] − f ′ (x),
(2.17)
Now this method consists of three evaluation of the functions and one
evaluation of the first derivative per iteration, now it has improved
efficiency index 1.6817. Now we prove the convergence of this method
by the following theorem:
Theorem 2.1. Let us consider α as the simple root of the nonlinear
equation f(x)=0 in the domain D and assume that f(x) is sufficiently
smooth in the neighborhood of the root. Then, the iterative scheme
6 J. P. JAISWAL AND NEHA CHOUBEY

defined by (2.5), (2.6) and (2.17) is of local order eight and has the
following error equation
en+1 = (c32 c23 − c32 c3 c4 )e8n + O(e9n ),
f (h) (d)
where en = xn − α and ch = h!
, h=1,2,3....

Proof. We provide the Taylor series expansion of each term involved in

(2.5), (2.6) and (2.17). By Taylor expansion around the simple root in
the nth iteration, we have

f (xn ) = f ′ (α)[en + c2 e2n + c3 e3n + c4 e4n + c55 e5n

+ c6 e6n + c7 e7n + c8 e8n + c9 e9n + c10 e10
n (2.18)
+ c11 e11
n + c12 e12
n + O(e13
n )]
and, we have
f ′ (xn ) = f ′ (α)[1 + 2c2 en + 3c3 e2n + 4c4 e3n + 5c55 e4n
+ 6c6 e5n + 7c7 e6n + 8c8 e7n + 9c9 e8n + 10c10 e9n (2.19)
+ 11c11 e10
n + 12c12 e11
n + O(e12
n )].

Further more it can be easily find

f (xn
= en − c2 e2n + 2c2 e3n − 2c3 e3n + ........ + O(e12
n ). (2.20)
f ′ (xn )
By considering this relation and equation (2.5), we obtain
yn = α + c2 e2n + 2(c3 − c22 )e3n + ...... + O(e12
n ). (2.21)
At this time, we should expand f (yn ) around the root by taking into
consideration (2.21). Accordingly, we have
f (yn ) = f ′ (α)[c2 e2n + 2(−c22 + c3 )e3n + ...... + O(e12 )]. (2.22)
Using (2.21), (2.18), (2.22) and (2.20) in the equation (2.6), we can
obtain
3
zn = α − c2 c3 e42 + ( c42 + 2c22 c3 − 2c4 c2 − 2c23 )e5n + .... + O(e12
n ). (2.23)
2
On the other hand, we have
3
f (zn ) = f ′ (α)[−c2 c3 e4n +( c22 c3 −2c4 c2 −2c23 )e5n +......+O(e12
n )]. (2.24)
2
Now we expand the Taylor series of each existing divided differences in
the denominator of equation (2.17). We obtain
f [xn , zn ] = f ′ (α)[1 + c2 en + c3 e2n + c4 e3n + ..... + O(e12
n )], (2.25)
f [yn , zn ] = f ′ (α)[1 + c2 e2n + (2c2 c3 − 2c32 )e3n + ..... + O(e12
n )], (2.26)
 A NEW EFFICIENT OPTIMAL EIGHTH-ORDER ITERATIVE METHOD. . . 7

f [xn , yn ] = f ′ (α)[1+c2 en +(c22 +c3 )e2n +(−2c32 +3c3 c2 +c4 )e3n +.....+O(e12 n )],
(2.27)
f [yn , xn , xn ] = f ′ (α)[c2 + 2c3 en + (3c4 + c2 c3 )e2n
(2.28)
+ (−2c22 c3 + 2c4 c2 + 2c23 + 4c5 )e3n + .... + O(e12 n )].
By consider the above mentioned relations (2.25)-(2.28) in the equation
(2.17), we can find
3
xn+1 = (c32 c23 − c22 c3 c4 )e8 + [−3c62 c3 + c52 c4 − 4c42 c23 + 8c32 c3 c4
2 (2.29)
2 3 2 2 2 2 9 12
+ 4c2 3c − 2c c
5 2 3c − 2c c
2 4 − 4c c
2 3 4c ]e + ... + O(en ).

This implies that the order of convergence for this method is eight i.e.
optimal order of convergence and its efficiency index is 81/4 ≈ 1.6817,
which is more that 1.4142 of Newton’s method, and 1.5650 of three-
step methods [6, 2], and is equal to 1.6817 of [20], [12], [15], [14], [22],
[8], [7], [3], [21] but numerical performance is better to almost all the
eighth-order methods ( later shown in the Tables 2-8).
3. Well established eighth-order Methods
First we are giving here some well established eighth-order methods:
3.1. Bi et al. Methods, [ [20], 2009]:
Method I
f (xn )
yn = xn − ′ ,
f (xn )
  
2f (xn ) − f (yn ) f (yn )
zn = yn − ,
2f (xn ) − 5f (yn ) f ′ (xn )
  
f (xn ) + (γ + 2)f (zn ) f (zn )
xn+1 = zn − .
f (xn ) + γf (zn ) f [zn , yn ] + f [zn , xn , xn ](zn − yn )
(3.1)
Method II
"  2  3 #  
2f (yn ) f (yn ) f (yn ) f (yn )
zn = yn − 1 + +5 + ,
2f (xn ) f (xn ) f (xn ) f ′ (xn )
  
f (xn ) + (γ + 2)f (zn ) f (zn )
xn+1 = zn − .
f (xn ) + γf (zn ) f [zn , yn ] + f [zn , xn , xn ](zn − yn )
(3.2)
Method II
"  2  3 #−1  
2f (yn ) f (yn ) f (yn ) f (yn )
zn = yn − 1 − − + ,
2f (xn ) f (xn ) f (xn ) f ′ (xn )
8 J. P. JAISWAL AND NEHA CHOUBEY
  
f (xn ) + (γ + 2)f (zn ) f (zn )
xn+1 = zn − .
f (xn ) + γf (zn ) f [zn , yn ] + f [zn , xn , xn ](zn − yn )
(3.3)
Method IV
 −2/3  
2f (xn ) − 3f (yn ) f (yn )
zn = yn − ,
f (xn ) f ′ (xn )
  
f (xn ) + (γ + 2)f (zn ) f (zn )
xn+1 = zn − ,
f (xn ) + γf (zn ) f [zn , yn ] + f [zn , xn , xn ](zn − yn )
(3.4)
where γ ∈ R and denominator is not equal to zero.

3.2. Sharma et al. Methods, [ [12], 2010]:

Method I   
f (xn ) f (yn )
zn = yn − ,
f (xn ) − 2f (yn ) f ′ (xn )
"  2 #  
f (zn ) f (zn ) f [xn , yn ]f (zn )
xn+1 = zn − 1 + +γ . (3.5)
f (xn ) f (xn ) f [yn , zn ]f [xn , zn ]
Method II
  
f (xn ) f (yn )
zn = yn − ,
f (xn ) − 2f (yn ) f ′ (xn )
  
f (xn ) + (γ + 1)f (zn ) f [xn , yn ]f (zn )
xn+1 = zn − . (3.6)
f (xn ) + γf (zn ) f [yn , zn ]f [xn , zn ]
Method III
  
f (xn ) f (yn )
zn = yn − ,
f (xn ) − 2f (yn ) f ′ (xn )
 (1/γ)  
f (zn ) f [xn , yn ]f (zn )
xn+1 = zn − 1 + γ , (3.7)
f (xn ) f [yn , zn ]f [xn , zn ]
where γ ∈ R and denominator is not equal to zero.

3.3. Thukral Method, [ [15], 2010]:

f (xn )2 + f (yn )2
zn = xn − ,
f ′ (xn )(f (xn ) − f (yn ))
" 2 #
1 + µ2n

2 3 f (zn ) f (zn ) f (zn )
xn+1 = zn − − 2(µn ) − 6(µn ) + +4 .
1 − µn f (yn ) f (xn ) f ′ (xn )
(3.8)
 A NEW EFFICIENT OPTIMAL EIGHTH-ORDER ITERATIVE METHOD. . . 9

where µn = (f (yn )/f (xn )) and denominator is not equal to zero.

3.4. Wang et al., [ [21], 2010]:

f (yn )
zn = yn − ,
2f [xn , yn ] − f ′ (xn )
f (zn )
xn+1 = zn − .
2f [xn , zn ] + f [yn , zn ] − 2f [xn , yn ] + (yn − zn )f [yn , xn , xn ]
(3.9)
3.5. Sragolzaei et al. Method, [ [14], 2011]:
 2  
f (yn ) f (yn )
zn = yn − 1 + ,
f (xn ) f ′ (xn )
f (zn )
xn+1 = zn − .
2f [zn , xn ] + f [zn , yn ] − 2f [yn , xn ] + (yn − zn )f [yn , xn , xn ]
(3.10)
3.6. Cordero et al. Method, [ [3], 2011].
 
f (xn ) f (xn ) − f (yn )
zn = xn − ′ ,
f (xn ) f (xn ) − 2f (yn )
 2
f (zn ) f (xn ) − f (yn ) 1 f (zn )
un = zn − ′ + ,
f (xn ) f (xn ) − 2f (yn ) 2 f (yn ) − 2f (zn )
 
f (zn ) un − zn
xn+1 = un − 3 ′ . (3.11)
f (xn ) yn − xn
3.7. Soleymani Methods, [([7], [8]); (2011, 2012)]:
Method I
3 f ′ (yn )2 − f ′ (xn )2 f (xn )
 
zn = xn − 1 − ,
8 f ′ (yn )2 f ′ (xn )
f (zn )
xn+1 = zn − . (3.12)
f ′ (y n ) + 2f [zn , xn , xn ](zn − yn )
Method II
2f (xn )
zn = xn − ,
f ′ (xn ) + f ′ (yn )
f (zn )
kn = zn − ′ ,
f (yn )
f (kn )
xn+1 = kn − f ′ (xn )(3f ′ (yn )−f ′ (xn )) . (3.13)
f ′ (xn )+f ′ (yn )
10 J. P. JAISWAL AND NEHA CHOUBEY

3.8. Kim Method, [ [22], 2012]:

1 + βun + λu2n
 
f (yn )
zn = yn − 2
,
1 + (β − 2)un + µun f ′ (xn )
 
1 + aun + bvn f (zn )
xn+1 = zn − , (3.14)
1 + cun + dvn f (xn ) + f [yn , xn , zn ](zn − xn )
′

where un = f (yn )/f (xn ) , vn = f (zn )/f (xn ) , β, λ, µ, a, b, c, and d

are constant parameters and related by β = (λ − µ − 2/3)/2, a = −2,
c = −3, d = b − 3.
4. Numerical Testing
To demonstrate the performance of the new eighth-order method,
we take seven particular non-linear equations. We will determine the
consistency and stability of results by examining the convergence of
the new iterative method. We will give estimates of the approximate
solution produced be the eighth-order method.
Here we consider, the following test functions to illustrate the accu-
racy of new iterative method. The root of each nonlinear test function is
also listed in from of each page up to fifteen decimal places, when such
roots are non-integers. All the computations reported here we have
done using Mathematica 8, Scientific computations in many branches
of science and technology demand very high precision degree of numer-
ical precision. The test non-linear functions are listed in Table-1. The
Table 1. Test functions and their roots.
Non-linear function Roots
x
f1 (x) = sin(x) − 100 0
1 3 1
f2 (x) = 3x4 − x − 3x +1 1
f3 (x) = exp(sin(x)) − 1 − x5 0
2
x + sin( xπ )
f4 (x) = √ √ 0
x3
f5 (x) = x4 + 8 sin( x2π+2 ) + x4 +1
− 6+ 8
16
-2
f6 (x) = cos(x) − x 0.739085133215160
f7 (x) = exp(x) + cos(x) -1.7461395304080124

results of comparison for the test function are provided in the Table 2-
8. It can be seen that the resulting method from our class are accurate
and efficient in terms of number of accurate decimal places to find the
roots after some iterations.
A NEW EFFICIENT OPTIMAL EIGHTH-ORDER ITERATIVE METHOD. . .11

Table 2. Errors Occurring in the estimates of the root

of function f1 by the method described with initial guess
x0 = 0.7.

Methods |f1 (x1 )| |f1 (x2 )| |f1 (x3 )|

(2.17) 0.695e-5 0.654e-60 0.336e-665
(3.1) with γ = 1 0.791e-3 0.363e-18 0.740e-95
(3.2) with γ = 1 0.205e-3 0.428e-21 0.169e-109
(3.3) with γ = 1 0.628e-3 0.115e-18 0.233e-97
(3.4) with γ = 1 0.887e-2 0.177e-8 0.140e-28
(3.5) with γ = 1 0.689e-4 0.339e-48 0.139e-535
(3.6) with γ = 1 0.821e-4 0.233e-47 0.224e-526
(3.7) with γ = 1 0.754e-4 0.918e-48 0.790e-531
(3.8) 0.129e-2 0.174e-28 0.257e-261
(3.10) 0.275e-4 0.239e-53 0.511e-593
(3.14) with (λ, µ, b) = (0, 0, 4) 0.101e-3 0.414e-46 0.201e-512
(3.13) 0.148e-4 0.116e-65 0.512e-860
(3.12) 0.133e-2 0.601e-10 0.548e-32
(3.11) 0.830e-4 0.409e-39 0.702e-357
(3.9) 0.101e-4 0.414e-58 0.217e-645
12 J. P. JAISWAL AND NEHA CHOUBEY

Table 3. Errors Occurring in the estimates of the root

of function f2 by the method described with initial guess
x0 = 1.2.

Methods |f2 (x1 )| |f2 (x2 )| |f2 (x3 )|

(2.17) 0.111e-3 0.497e-31 0.823e-250
(3.1) with γ = 1 0.372e-3 0.709e-24 0.638e-161
(3.2) with γ = 1 dgt dgt dgt
(3.3) with γ = 1 0.165e+0 .136e-3 0.442e-15
(3.4) with γ = 1 0.768e-2 0.116e-6 0.427e-21
(3.5) with γ = 1 0.317e-2 0.192e-18 0.376e-148
(3.6) with γ = 1 0.547e-2 0.141e-16 0.308e-133
(3.7) with γ = 1 0.420e-2 0.176e-17 0.186e-140
(3.8) 0.213e+1 0.536e-1 0.271e-7
(3.10) 0.109e-1 0.431e-14 0.362e-113
(3.14) with (λ, µ, b) = (0, 0, 4) 0.496e-4 0.101e-33 0.318e-271
(3.13) 0.667e-3 0.119e-24 0.128e-198
(3.12) 0.545e-4 0.419e-9 0.247e-19
(3.11) 0.992e-2 0.683e-15 0.467e-120
(3.9) 0.713e-3 0.409e-24 0.494e-194
A NEW EFFICIENT OPTIMAL EIGHTH-ORDER ITERATIVE METHOD. . .13

Table 4. Errors Occurring in the estimates of the root

of function f3 by the method described with initial guess
x0 = −0.55.

Methods |f3 (x1 )| |f3 (x2 )| |f3 (x3 )|

(2.17) 0.628e-2 0.344e-20 0.168e-184
(3.1) with γ = 1 0.175e-1 0.682e-9 0.549e-146
(3.2) with γ = 1 0.266e+1 0.260e+2 0.527e+3
(3.3) with γ = 1 0.176e+0 0.736e-4 0.871e-17
(3.4) with γ = 1 0.486e-2 0.243e-5 0.614e-12
(3.5) with γ = 1 0.520e-2 .212e-18 0.173e-149
(3.6) with γ = 1 0.883e-2 0.141e-16 0.652e-135
(3.7) with γ = 1 0.684e-2 0.186e-17 0.610e-142
(3.8) 0.271e+1 0.626e+3 0.225e+5
(3.10) 0.107e-1 0.344e-15 0.582e-123
(3.14) with (λ, µ, b) = (0, 0, 4) 0.623e-2 0.337e-18 0.261e-148
(3.13) 0.737e-2 0.568e-17 0.774e-138
(3.12) 0.424e-2 0.198e-5 0.431e-12
(3.11) 0.176e-1 0.284e-14 0.162e-116
(3.9) 0.110e-2 0.141e-24 0.104e-199
14 J. P. JAISWAL AND NEHA CHOUBEY

Table 5. Errors Occurring in the estimates of the root

of function f4 by the method described with initial guess
x0 = 0.1.

Methods |f4 (x1 )| |f4 (x2 )| |f4 (x3 )|

(2.17) 0.467e-14 0.371e-147 0.370e-1478
(3.1) with γ = 1 0.340e-12 0.287e-115 0.629e-1043
(3.2) with γ = 1 0.199e-10 0.174e-97 0.513e-881
(3.3) with γ = 1 0.124e-11 0.128e-109 0.173e-991
(3.4) with γ = 1 0.274e-4 0.568e-15 0.508e-47
(3.5) with γ = 1 0.756e-11 0.106e-91 0.156e-738
(3.6) with γ = 1 0.756e-11 0.106e-91 0.162e-738
(3.7) with γ = 1 0.756e-11 0.106e-91 0.159e-738
(3.8) 0.141e-9 0.322e-80 0.242e-645
(3.10) 0.354e-10 0.130e-85 0.435e-689
(3.14) with (λ, µ, b) = (0, 0, 4) 0.169e-11 0.145e-97 0.442e-786
(3.13) 0.101e-10 0.146e-90 0.283e-729
(3.12) 0.472e-3 0.999e-8 0.447e-17
(3.11) 0.681e-11 0.423e-92 0.950e-742
(3.9) 0.256e-11 0.618e-96 0.706e-773
A NEW EFFICIENT OPTIMAL EIGHTH-ORDER ITERATIVE METHOD. . .15

Table 6. Errors Occurring in the estimates of the root

of function f5 by the method described with initial guess
x0 = −3.

Methods |f5 (x1 )| |f5 (x2 )| |f5 (x3 )|

(2.17) 0.283e-3 0.555e-26 0.108e-207
(3.1) with γ = 1 0.756e-2 0.218e-7 0.129e-29
(3.2) with γ = 1 0.690e-2 0.143e-7 0.243e-30
(3.3) with γ = 1 0.745e-2 0.201e-7 0.107e-29
(3.4) with γ = 1 0.179e-1 0.159e-3 0.101e-7
(3.5) with γ = 1 0.333e-3 0.599e-25 0.659e-199
(3.6) with γ = 1 0.321e-3 0.447e-25 0.633e-200
(3.7) with γ = 1 0.328e-3 0.519e-25 0.209e-199
(3.8) 0.251e-3 0.409e-25 0.199e-199
(3.10) 0.380e-3 0.157e-24 0.138e-195
(3.14) with (λ, µ, b) = (0, 0, 4) 0.432e-3 0.545e-25 0.344e-200
(3.13) 0.705e-3 0.362e-23 0.173e-185
(3.12) 0.622e-2 0.676e-5 0.779e-11
(3.11) 0.122e-2 0.507e-23 0.229e-185
(3.9) 0.307e-3 0.132e-25 0.160e-204
16 J. P. JAISWAL AND NEHA CHOUBEY

Table 7. Errors Occurring in the estimates of the root

of function f6 by the method described with initial guess
x0 = 1.5.

Methods |f6 (x1 )| |f6 (x2 )| |f6 (x3 )|

(2.17) 0.696e-6 0.176e-55 0.300e-452
(3.1) with γ = 1 0.682e-2 0.124e-10 0.132e-45
(3.2) with γ = 1 0.706e-2 0.141e-10 0.227e-45
(3.3) with γ = 1 0.686e-2 0.126e-10 0.143e-45
(3.4) with γ = 1 0.650e-1 0.383e-3 0.129e-7
(3.5) with γ = 1 0.415e-5 0.990e-48 0.103e-388
(3.6) with γ = 1 0.421e-5 0.111e-47 0.261e-388
(3.7) with γ = 1 0.418e-5 0.109e-47 0.164e-388
(3.8) 0.249e-4 0.203e-40 0.385e-329
(3.10) 0.142e-5 0.222e-51 0.797e-418
(3.14) with (λ, µ, b) = (0, 0, 4) 0.228e-5 0.317e-50 0.454e-409
(3.13) 0.234e-5 0.185e-50 0.289e-411
(3.12) 0.129e-1 0.311e-5 0.180e-12
(3.11) 0.893e-5 0.142e-45 0.589e-372
(3.9) 0.870e-6 0.363e-54 0.332e-441
A NEW EFFICIENT OPTIMAL EIGHTH-ORDER ITERATIVE METHOD. . .17

Table 8. Errors Occurring in the estimates of the root

of function f7 by the method described with initial guess
x0 = −2.3.

Methods |f7 (x1 )| |f7 (x2 )| |f7 (x3 )|

(2.17) 0.563e-6 0.167e-54 0.101e-442
(3.1) with γ = 1 0.105e-2 0.464e-14 0.179e-59
(3.2) with γ = 1 0.529e-2 0.301e-11 0.318e-48
(3.3) with γ = 1 0.490e-3 0.222e-15 0.946e-65
(3.4) with γ = 1 0.578e-2 0.554e-6 0.507e-14
(3.5) with γ = 1 0.412e-4 0.749e-39 0.893e-317
(3.6) with γ = 1 0.443e-4 0.135e-38 0.977e-315
(3.7) with γ = 1 0.428e-4 0.101e-38 0.960e-316
(3.8) 0.718e-3 0.478e-28 0.186e-229
(3.10) 0.788e-4 0.774e-37 0.671e-301
(3.14) with (λ, µ, b) = (0, 0, 4) 0.231e-4 0.551e-42 0.580e-343
(3.13) 0.788e-6 0.357e-53 0.647e-432
(3.12) 0.384e-3 0.269e-8 0.132e-18
(3.11) 0.246e-4 0.910e-41 0.314e-332
(3.9) 0.592e-5 0.357e-46 0.630e-376
18 J. P. JAISWAL AND NEHA CHOUBEY

5. Conclusion
In this work, we have developed a new eighth-order convergent method
for solving non-linear equations. Convergence analysis shows that our
new method is eighth-order convergent which is also supported by the
numerical works. This iterative method require evaluation of three
functions and one first derivative during each iterative step. Compu-
tational results demonstrate that the iterative method is efficient and
exhibit better performance as compared with other well known eighth-
order methods.

References
[1] A. Cordero, S. L. Huero, E. Martinez and J. R. Torregrosa: Steffensen type
methods for solving non-linear equations, Applied Mathematics and computa-
tion, vol 198, no. 2, pp. 527-533 (2007).
[2] A. Cordero, J. L. Huero, M. Martinez and J. R. Torregrosa: Efficient three
step iterative methods with sixth order convergence for non linear equations,
Numerical Algorithms 53, pp. 485-495 (2010).
[3] A. Cordero, J. R. Torregrosa and M. P. Vassileva: Three step iterative method
with optimal eighth order convergence, Journal of Computational and Applied
Mathematics, 235, pp. 3189-3194, (2011).
[4] A. Iliev and N. Kyurtrchiev, Nontrivial method in Numerical Analysis selected
topics in Numerical Analysis, LAP Lambert Academic Publishing (2010).
[5] B. I. Yun: A non-iterative method for solving non-linear equations, Applied
Mathematics and Computation, vol. 198, no.2, pp. 691-699 (2008).
[6] C. Chun and Y. Ham: Some sixth-order variation of ostrowski’s root finding
methods, Applied Mathematical Computations, 193, pp. 389-394 (2007).
[7] F. Soleymani: Regarding the accuracy of optimal eighth-order methods, Math-
ematical and Computer Modelling, 53, pp. 1351-1357 (2011).
[8] F. Soleymani: A second derivative free eighth-order non-linear equation solver,
Non-linear Stud. (19) No. 1, pp. 79-86 (2012).
[9] H. T. Kung and J. F. Traub: Optimal order of one-point and multipoint
iteration, J.CAM 21, PP. 643-651 (1974).
[10] J. F. Traub: Iterative methods for solution of equations chelsea Publishing,
New York, NY, USA (1997).
[11] J. F. Traub: Iterative methods for the solution of equation, Prentice-Hall,
Englewood cliffs, New Jerset (1964).
[12] J. R. Sharma and R. Sharma: A new family of modified Ostrowski’s methods
with accelerated eight order convergence, Numerical Algorithms, Vol. 54, no.
4, pp. 445-458 (2010).
[13] M. S. Petkovic: On a general clam of multipoint root-finding methods of
high computational efficiency. SIAM J. Numerical Analysis, 47, pp. 4402-4414
(2010).
[14] P. Sargolzaei and F. Soleymani: Accurate fourteenth order methods for solving
non-linear equations, Numerical Algorithm, vol. 58, pp. 513-527 (2011).
 A NEW EFFICIENT OPTIMAL EIGHTH-ORDER ITERATIVE METHOD. . .19

[15] R. Thukral: A new eight-order iterative method for solving non-linear equa-
tions Applied Mathematics and Computation, vol. 217, no. 1, pp. 222-229
(2010).
[16] R. F. Kung: A family of fourth order methods for nonlinear equations, SIAM
Journal on Numerical Analysis 10, pp. 876-879 (1973).
[17] S. K. Khattri and I. K. Argyrous, Sixth-order derivative free family of iterative
methods, Applied mathematics and Computation, vol. 217, no. 12, pp. 5500-
5507 (2011).
[18] S. Weerakoon and T. G. I. Fernando: A variant of Newtons’s Method with
accelerated third-order convergence, method with accelerated third-order con-
vergence, Applied Mathematics letters, vol. 13, no.8, pp. 87-93 (2000).
[19] W. Gautschi: Numerical Analysis: An Introduction Birkhauser, Barton, Mass,
USA (1997).
[20] W. Bi, H. Ren and Q. Wen: A new family of eight-order iterative methods for
solving non linear equation, Applied Mathematics and Computation, vol. 214,
no. 1, pp. 234-245 (2009).
[21] X. Wang and L. Liu: Modified Ostrowski’s method with eight-order conver-
gence and high efficiency index, Applied Mathematics Letters, 23, pp. 549-554
(2010).
[22] Y. I. Kim: A try parametric family of three step eight order methods for
solving non linear equations, International Journal of Computer Mathematics,
vol. 89, no. 8, pp. 1051-1059 (2012).

Jai Prakash Jaiswal

Department of Mathematics,
Maulana Azad National Institute of Technology,
Bhopal, M.P., India-462051.
Corresponding author:
E-mail: [email protected].

Neha Choubey
Department of Applied Sciences and Humanities,
Oriental Institute of Science and Technology,
Bhopal, M.P., India-462021
E-mail: [email protected].

View publication stats