Open AccessArticle

Local Convergence of an Efﬁcient High Convergence Order Method Using Hypothesis Only on the First Derivative

Ioannis K. Argyros

¹,

Ramandeep Behl

^2,* and

S.S. Motsa

Department of Mathematics Sciences, Cameron University, Lawton, OK 73505, USA

School of Mathematics, Statistics and Computer Science, University of KwaZulu-Natal, Private Bag X01, Scottsville 3209, Pietermaritzburg, South Africa

Author to whom correspondence should be addressed.

Algorithms 2015, 8(4), 1076-1087; https://doi.org/10.3390/a8041076

Submission received: 25 September 2015 / Accepted: 11 November 2015 / Published: 20 November 2015

(This article belongs to the Special Issue Numerical Algorithms for Solving Nonlinear Equations and Systems)

Download Versions Notes

Abstract

We present a local convergence analysis of an eighth order three step method in order to approximate a locally unique solution of nonlinear equation in a Banach space setting. In an earlier study by Sharma and Arora (2015), the order of convergence was shown using Taylor series expansions and hypotheses up to the fourth order derivative or even higher of the function involved which restrict the applicability of the proposed scheme. However, only first order derivative appears in the proposed scheme. In order to overcome this problem, we proposed the hypotheses up to only the first order derivative. In this way, we not only expand the applicability of the methods but also propose convergence domain. Finally, where earlier studies cannot be applied, a variety of concrete numerical examples are proposed to obtain the solutions of nonlinear equations. Our study does not exhibit this type of problem/restriction.

Keywords:

Newton-like method; local convergence; efficiency index; optimum method

MSC Classification:

65G99; 65H10; 47J25; 47J05

1. Introduction

Numerical analysis is a wide-ranging discipline having close connections with mathematics, computer science, engineering and the applied sciences. One of the most basic and earliest problem of numerical analysis concerns with finding efficiently and accurately the approximate locally unique solution

x^{*}

of the equation of the form

F (x) = 0,

(1)

where F is a Fréchet differentiable operator defined on a convex subset D of X with values in Y, where X and Y are the Banach spaces.

Analytical methods for solving such equations are almost non-existent for obtaining the exact numerical values of the required roots. Therefore, it is only possible to obtain approximate solutions and one has to be satisfied with approximate solutions up to any specified degree of accuracy, by relying on numerical methods which are based on iterative procedures. Therefore, researchers worldwide resort to an iterative method and they have proposed a plethora of iterative methods [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16]. While, using these iterative methods researchers face the problems of slow convergence, non-convergence, divergence, inefficiency or failure (for detail please see Traub [15] and Petkovic et al. [13]).

The convergence analysis of iterative methods is usually divided into two categories: semi-local and local convergence analysis. The semi-local convergence matter is, based on the information around an initial point, to give criteria ensuring the convergence of iteration procedures. On the other hand, the local convergence is based on the information around a solution, to find estimates of the radii of convergence balls. A very important problem in the study of iterative procedures is the convergence domain. Therefore, it is very important to propose the radius of convergence of the iterative methods.

We study the local convergence analysis of three step method defined for each

n = 0, 1, 2, \dots

\begin{matrix} y_{n} & = x_{n} - F^{'} {(x_{n})}^{- 1} F (x_{n}), \\ z_{n} & = ϕ_{4} (x_{n}, y_{n}), \\ x_{n + 1} & = ϕ_{8} (x_{n}, y_{n}, z_{n}) \\ = z_{n} - {[z_{n}, x_{n}; F]}^{- 1} [z_{n}, y_{n}; F] {(2 [z_{n}, y_{n}; F] - [z_{n}, x_{n}; F])}^{- 1} F (z_{n}), \end{matrix}

(2)

where

x_{0} \in D

is an initial point,

[\cdot, \cdot; F] : D^{2} \to L (X)

ϕ_{4}

is any two-point optimal fourth-order scheme. The eighth order of convergence of Scheme (2) was shown in [1] when

X = Y = R

and

[x, y; F] = \frac{F (x) - F (y)}{x - y}

for

x \neq y

and

[x, x; F] = F^{'} (x)

. That is when

[\cdot, \cdot; F]

is a divided difference of first order of operator F [5,6]. The local convergence was shown using Taylor series expansions and hypotheses reaching up to the fifth order derivative. The hypotheses on the derivatives of F and H limit the applicability of Scheme (2). As a motivational example, define function F on

X = Y = R

D = [- \frac{1}{π}, \frac{2}{π}]

F (x) = \{\begin{matrix} x^{3} log (π^{2} x^{2}) + x^{5} sin (\frac{1}{x}), & x \neq 0, \\ 0, & x = 0 . \end{matrix}

Then, we have that

F^{'} (x) = 2 x^{2} - x^{3} cos (\frac{1}{x}) + 3 x^{2} log (π^{2} x^{2}) + 5 x^{4} sin (\frac{1}{x}),

F^{''} (x) = - 8 x^{2} cos (\frac{1}{x}) + 2 x (5 + 3 log (π^{2} x^{2})) + x (20 x^{2} - 1) sin (\frac{1}{x})

and

F^{'''} (x) = \frac{1}{x} [(1 - 36 x^{2}) cos (\frac{1}{x}) + x (22 + 6 log (π^{2} x^{2}) + (60 x^{2} - 9) sin (\frac{1}{x}))] .

One can easily find that the function

F^{'''} (x)

is unbounded on

D

at the point

x = 0

. Hence, the results in [1], cannot apply to show the convergence of Scheme (2) or its special cases requiring hypotheses on the fifth derivative of function F or higher. Notice that, in particular, there is a plethora of iterative methods for approximating solutions of nonlinear equations [1,2,3,4,5,6,7,8,10,11,12,13,14,15,16]. These results show that initial guess should be close to the required root for the convergence of the corresponding methods. However, how close an initial guess would be required for the convergence of the corresponding method? These local results give no information on the radius of the ball convergence for the corresponding method. The same technique can be applied to other methods.

In the present study we expand the applicability of Scheme (2) using only hypotheses on the first order derivative of function F. We also propose the computable radii of convergence and error bounds based on the Lipschitz constants. We further present the range of initial guess

x^{*}

that tells us how close the initial guess would be required for granted convergence of the Scheme (2). This problem was not addressed in [1]. The advantages of our approach are similar to the ones already mentioned for Scheme (2).

The rest of the paper is organized as follows: in Section 2, we present the local convergence analysis of Scheme (2). Section 3 is devoted to the numerical examples which demonstrate our theoretical results. Finally, the conclusion is given in the Section 4.

2. Local Convergence: One Dimensional Case

In this section, we define some scalar functions and parameters to study the local convergence of Scheme (2).

Let

K_{0} > 0, K_{1} > 0, K > 0, L_{0} > 0, L > 0, M \geq 1

λ \geq 1

, be given constants. Let us also assume

g_{2} : [0, \frac{1}{L_{0}}) \to R

, be nondecreasing and continuous function. Further, define function

h_{2} : [0, \frac{1}{L_{0}}) \to R

and

h_{2} (t) = g_{2} (t) t^{λ - 1} - 1

Suppose that

\begin{matrix} g_{2} (t) t^{λ - 1} < 1, for each [0, \frac{1}{L_{0}}), \\ h_{2} (t) \to a positive number or + \infty, \\ as t \to l < \frac{1}{L_{0}} for some l > 0 . \end{matrix}

(3)

Then, we have

h_{2} (0) = - 1 < 0

. By Equation (3) and the intermediate value theorem, function

h_{2}

has zeros in the interval

(0, l)

. Further, let

r_{2}

be the smallest such zero. Moreover, define functions

g_{1}, p

and

h_{p}

in the interval

[0, \frac{1}{L_{0}})

\begin{matrix} g_{1} (t) = \frac{L t}{2 (1 - L_{0} t)}, \\ p (t) = (K_{0} g_{2} (t) t^{λ - 1} + K_{1}) t, \\ h_{p} (t) = p (t) - 1 \\ and parameter r_{1} by \\ r_{1} = \frac{2}{2 L_{0} + L} . \end{matrix}

We have

g_{1} (r_{1}) = 1

and for each

t \in [0, r_{1}) : 0 \leq g_{1} (t) < 1

. We also get

h_{p} (0) = - 1

and

h_{p} (t) \to + \infty

t \to \frac{1^{-}}{L_{0}}

. Denote by

r_{p}

the smallest zero of function

h_{p}

on the interval

(0, \frac{1}{L_{0}})

. Furthermore, define functions q and

h_{q}

on the interval

[0, \frac{1}{L_{0}})

q (t) = p (t) + 2 [K_{0} g_{2} (t) t^{λ} + K_{1} g_{1} (t) t]

and

h_{q} (t) = q (t) - 1

Using

h_{q} (0) = - 1 < 0

and Equation (3), we deduce that function

h_{q}

has a smallest zero denoted by

r_{q}

Finally define functions

g_{3}

and

h_{3}

on the interval

[0, min {r_{p}, r_{q}})

\begin{matrix} g_{3} (t) = (1 + \frac{K M}{(1 - p (t)) (1 - q (t))}) g_{2} (t) t^{λ} \\ and \\ h_{3} = g_{3} (t) - 1 . \end{matrix}

Then, we get

h_{3} (0) = - 1

and

h_{3} (t) \to + \infty

t \to min {r_{p}, r_{q}}

. Denote by

r_{3}

the smallest zero of function

h_{3}

on the interval

(0, min {r_{p}, r_{q}})

. Define

r = min {r_{1}, r_{2}, r_{3}} .

(4)

Then, we have that

0 < r \leq r_{1} < \frac{1}{L_{0}}

(5)

and for each

t \in [0, r)

0 \leq g_{1} (t) < 1,

(6)

0 \leq p (t) < 1,

(7)

0 \leq q (t) < 1,

(8)

0 \leq g_{2} (t) < 1,

(9)

and

0 \leq g_{3} (t) < 1 .

(10)

U (γ, s)

and

\bar{U} (γ, s)

stand, respectively for the open and closed balls in X with center

γ \in X

and radius

s > 0

Next, we present the local convergence analysis of Scheme (2) using the preceding notations.

Theorem 1.

Let us consider

F : D \subset X \to Y

be a Fréchet differentiable operator. Let us also assume

[\cdot, \cdot; F] : D^{2} \to L (X)

be a divided difference of order one. Suppose that there exist

x^{*} \in D, L_{0} > 0, λ \geq 1

such that Equation (3) holds and for each

x \in D

F (x^{*}) = 0, F^{'} {(x^{*})}^{- 1} \in L (Y, X),

(11)

∥ z (x) - x^{*} ∥ \leq g_{2} (∥ x - x^{*} ∥) ∥ x - x^{*} ∥^{λ}

(12)

and

∥F^{'} {(x^{*})}^{- 1} (F^{'} (x) - F^{'} (x^{*}))∥ \leq L_{0} ∥ x - x^{*} ∥ .

(13)

Moreover, suppose that there exist

K_{0} > 0, K_{1} > 0, K > 0, L > 0

and

M \geq 1

such that for each

x, y \in U (x^{*}, \frac{1}{L_{0}}) \cap D

∥F^{'} {(x^{*})}^{- 1} ([x, y; F] - F^{'} (x^{*}))∥ \leq K_{0} ∥ x - x^{*} ∥ + K_{1} ∥ y - x^{*} ∥,

(14)

∥F^{'} {(x^{*})}^{- 1} [x, y; F]∥ \leq K,

(15)

∥F^{'} {(x^{*})}^{- 1} (F^{'} (x) - F^{'} (y))∥ \leq L ∥ x - y ∥,

(16)

∥F^{'} {(x^{*})}^{- 1} F^{'} (x)∥ \leq M

(17)

and

\bar{U} (x^{*}, r) \subseteq D,

(18)

where the radius of convergence r is defined by Equation (4) and

z (x) = ϕ_{4} (x, x - F^{'} {(x)}^{- 1} F (x))

. Then, the sequence

{x_{n}}

generated by Scheme (2) for

x_{0} \in U (x^{*}, r) - {x^{*}}

is well defined, remains in

U (x^{*}, r)

for each

n = 0, 1, 2, \dots

and converges to

x^{*}

. Moreover, the following estimates hold

∥ y_{n} - x^{*} ∥ \leq g_{1} (∥ x_{n} - x^{*} ∥) ∥ x_{n} - x^{*} ∥ < ∥ x_{n} - x^{*} ∥ < r,

(19)

∥ z_{n} - x^{*} ∥ \leq g_{2} (∥ x_{n} - x^{*} ∥) ∥ x_{n} - x^{*} ∥ < ∥ x_{n} - x^{*} ∥

(20)

and

∥ x_{n + 1} - x^{*} ∥ \leq g_{3} (∥ x_{n} - x^{*} ∥) ∥ x_{n} - x^{*} ∥ < ∥ x_{n} - x^{*} ∥,

(21)

where the

“ g ”

functions are defined by previously. Furthermore, for

T \in [r, \frac{2}{L_{0}})

, the limit point

x^{*}

is the only solution of equation

F (x) = 0

\bar{U} (x^{*}, r) \cap D .

Proof.

We shall show estimates Equations (19)–(21) hold with the help of mathematical induction. By hypotheses

x_{0} \in U (x^{*}, r) - {x^{*}}

, Equations (5) and (13), we get that

∥F^{'} {(x^{*})}^{- 1} (F^{'} (x_{0}) - F^{'} (x^{*}))∥ \leq L_{0} ∥ x - x^{*} ∥ < L_{0} r < 1 .

(22)

It follows from Equation (22) and the Banach Lemma on invertible operators [5,14] that

F^{'} {(x_{0})}^{- 1} \in L (Y, X), y_{0}

is well defined and

∥F^{'} {(x_{0})}^{- 1} F^{'} (x^{*})∥ \leq \frac{1}{1 - L_{0} ∥ x_{0} - x^{*} ∥} .

(23)

Using the first sub step of Scheme (2) for

n = 0

, Equations (4), (6), (11) and (23), we get in turn

\begin{matrix} ∥ y_{0} - x^{*} ∥ & = ∥x_{0} - x^{*} - F {(x_{0})}^{- 1} F (x_{0})∥ \\ \leq ∥F^{'} {(x_{0})}^{- 1} F^{'} (x^{*})∥ ∥\int_{0}^{1} F^{'} {(x_{0})}^{- 1} (F^{'} (x^{*} + θ (x_{0} - x^{*})) - F^{'} (x_{0})) (x_{0} - x^{*}) d θ∥ \\ \leq \frac{L ∥ x_{0} - x^{*} ∥^{2}}{1 - L ∥ x_{0} - x^{*} ∥} = g_{1} ∥ ∥ x_{0} - x^{*} ∥) ∥ x_{0} - x^{*} ∥ \\ < ∥ x_{0} - x^{*} ∥ < r, \end{matrix}

(24)

which shows Equation (18) for

n = 0

and

y_{0} \in U (x^{*}, r)

. Then, from Equations (3) and (12), we see that Equation (20) follows. Hence,

z_{0} \in U (x^{*}, r)

. Next, we shall show that

{[z_{0}, x_{0}; F]}^{- 1} \in L (Y, X)

and

{(2 [z_{0}, y_{0}; F] - [z_{0}, x_{0}; F])}^{- 1} \in L (Y, X)

Using Equations (4), (5), (7), (13), (14) and (24), we get in turn that

\begin{matrix} ∥F^{'} {(x^{*})}^{- 1} ([z_{0}, x_{0}, F] - F^{'} (x^{*}))∥ \leq K_{0} ∥ z_{0} - x^{*} ∥ + K_{1} ∥ x_{0} - x^{*} ∥ \\ \leq K_{0} g_{2} (∥ x_{0} - x^{*} ∥) ∥ x_{0} - x^{*} ∥^{λ} + K_{1} ∥ x_{0} - x^{*} ∥, \\ = p (∥ x_{0} - x^{*} ∥) < p (r) < 1 . \end{matrix}

(25)

It follows from Equation (25) that

∥{[z_{0}, x_{0}; F]}^{- 1} F^{'} (x^{*})∥ \leq \frac{1}{1 - p (∥ x_{0} - x^{*} ∥)} .

(26)

Similarly, but using Equation (8) instead of Equation (7), we obtain in turn that

\begin{matrix} ∥F^{'} {(x^{*})}^{- 1} [2 ([z_{0}, y_{0}; F] - F^{'} (x^{*})) - ([z_{0}, x_{0}; F] - F^{'} (x^{*}))]∥ \\ \leq 2 ∥F^{'} {(x^{*})}^{- 1} ([z_{0}, y_{0}; F] - F^{'} (x^{*}))∥ + ∥F^{'} {(x^{*})}^{- 1} ([z_{0}, x_{0}; F] - F^{'} (x^{*}))∥, \\ \leq 2 (K_{0} ∥ z_{0} - x^{*} ∥ + K_{1} ∥ y_{0} - x^{*} ∥) + p (∥ x_{0} - x^{*} ∥), \\ \leq 2 (K_{0} g_{2} (∥ x_{0} - x^{*} ∥) ∥ x_{0} - x^{*} ∥^{λ} + K_{1} g_{1} (∥ x_{0} - x^{*} ∥) ∥ x_{0} - x^{*} ∥) + p (∥ x_{0} - x^{*} ∥), \\ = q (∥ x_{0} - x^{*} ∥) < q (r) < 1 . \end{matrix}

(27)

That is

∥2 {([z_{0}, y_{0}; F] - [z_{0}, x_{0}; F])}^{- 1} F^{'} (x^{*})∥ \leq \frac{1}{1 - q (∥ x_{0} - x^{*} ∥)} .

(28)

Hence,

x_{1}

is well defined by the third sub step of Scheme (2) for

n = 0

. We can write by Equation (11)

F (x_{0}) = F (x_{0}) - F (x^{*}) = \int_{0}^{1} F^{'} (x^{*} + θ (x_{0} - x^{*})) (x_{0} - x^{*}) d θ .

(29)

Notice that

∥ x^{*} + θ (x_{0} - x^{*}) - x^{*} ∥ = θ ∥ x_{0} - x^{*} ∥ < r

. Hence, we have that

x^{*} + θ (x_{0} - x^{*}) \in U (x^{*}, r)

. Then, by Equations (17) and (29) we get that

∥F^{'} {(x^{*})}^{- 1} F (x_{0})∥ = ∥\int_{0}^{1} F^{'} {(x^{*})}^{- 1} F^{'} (x^{*} + θ (x_{0} - x^{*})) (x_{0} - x^{*}) d θ∥ \leq M ∥ x_{0} - x^{*} ∥ .

(30)

We also have that by replacing

x_{0}

z_{0}

in Equation (30) that

∥F^{'} {(x^{*})}^{- 1} F (z_{0})∥ \leq M ∥ z_{0} - x^{*} ∥,

(31)

since

z_{0} \in U (x^{*}, r)

Then, using the last substep of Scheme (2) for

n = 0

, Equations (4), (10), (15), (20) (for

n = 0

), (26), (28), and (31) that

\begin{matrix} ∥ x_{1} - x^{*} ∥ & \leq ∥ z_{0} - x^{*} ∥ + ∥{[z_{0}, x_{0}; F]}^{- 1} F^{'} (x^{*})∥ ∥F^{'} {(x^{*})}^{- 1} [z_{0}, x_{0}; F]∥ \\ \times ∥{([z_{0}, y_{0}; F] - [z_{0}, x_{0}; F])}^{- 1} F^{'} (x^{*})∥ ∥F^{'} {(x^{*})}^{- 1} F (z_{0})∥, \\ \leq ∥ z_{0} - x^{*} ∥ + \frac{K M ∥ z_{0} - x^{*} ∥}{(1 - p (∥ x_{0} - x^{*} ∥)) (1 - q (∥ x_{0} - x^{*} ∥))}, \\ \leq (1 + \frac{K M}{(1 - p (∥ x_{0} - x^{*} ∥)) (1 - q (∥ x_{0} - x^{*} ∥))}) ∥ z_{0} - x^{*} ∥, \\ \leq g_{3} (∥ x_{0} - x^{*} ∥) ∥ x_{0} - x^{*} ∥ < ∥ x_{0} - x^{*} ∥ < r, \end{matrix}

(32)

which shows Equation (21) for

n = 0

and

x_{1} \in U (x^{*}, r)

. By simply replacing

x_{0}

y_{0}, z_{0}

x_{m}

y_{m}, z_{m}

in the preceding estimates we arrive at Equations (19)–(21). Then, from the estimates

∥ x_{m + 1} - x^{*} ∥ < ∥ x_{m} - x^{*} ∥ < r,

we conclude that

lim_{m \to \infty} x_{k} = x^{*}

and

x_{m + 1} \in U (x^{*}, r)

. Finally, to show the uniqueness part, let

y^{*} \in \bar{U} (x^{*}, T)

be such that

F (y^{*}) = 0

. Set

Q = \int_{0}^{1} F^{'} (x^{*} + θ (y^{*} - x^{*})) d θ

. Then, using Equation (14), we get that

\begin{matrix} ∥F^{'} {(x^{*})}^{- 1} (Q - F^{'} (x^{*}))∥ & \leq L_{0} \int_{0}^{1} θ ∥ x^{*} - y^{*} ∥ d θ = \frac{L_{0}}{2} T < 1 . \end{matrix}

(33)

Hence,

Q^{- 1} \in L (Y, X)

. Then, in view of the identity

F (y^{*}) - F (x^{*}) = Q (y^{*} - x^{*})

, we conclude that

x^{*} = y^{*}

☐

Remark 2.2

(a): In view of Equation (11) and the estimate

$\begin{matrix} ∥F^{'} {(x^{*})}^{- 1} [x, x^{*}; F]∥ & = & ∥F^{'} {(x^{*})}^{- 1} ([x, x^{*}; F] - F^{'} (x^{*}) - F^{'} (x^{*})) + I∥, \\ \leq & 1 + ∥F^{'} {(x^{*})}^{- 1} ([x, x^{*}; F] - F^{'} (x^{*}))∥, \\ \leq & 1 + L_{0} ∥ x_{0} - x^{*} ∥, \end{matrix}$

condition Equation (13) can be dropped and M can be replaced by

$M = M (t) = 1 + L_{0} t,$

or $M = 2,$ since $t \in [0, \frac{1}{L_{0}}) .$
(b): The results obtained here can be used for operators F satisfying the autonomous differential equation [5,6] of the form

$F^{'} (x) = P (F (x)),$

where P is a known continuous operator. Since $F^{'} (x^{*}) = P (F (x^{*})) = P (0),$ we can apply the results without actually knowing the solution $x^{*} .$ Let as an example $F (x) = e^{x} + 2 .$ Then, we can choose $P (x) = x - 2$ .
(c): The radius $r_{1}$ was shown in [5,6] to be the convergence radius for Newton’s method under conditions Equations (11) and (12). It follows from Equation (4) and the definition of $r_{1}$ that the convergence radius r of the Scheme (2) cannot be larger than the convergence radius $r_{1}$ of the second order Newton’s method. As already noted, $r_{1}$ is at least the size of the convergence ball given by Rheinboldt [14]

$\begin{matrix} r_{R} = \frac{2}{3 L} . \end{matrix}$

In particular, for $L_{0} < L$ , we have that

$r_{R} < r_{1}$

and

$\frac{r_{R}}{r_{1}} \to \frac{1}{3} as \frac{L_{0}}{L} \to 0 .$

That is our convergence ball $r_{1}$ is at most three times larger than Rheinboldt’s. The same value for $r_{R}$ given by Traub [15].
(d): We shall show that how to define function $g_{2}$ and l appearing in condition Equation (3) for the method

$\begin{matrix} y_{n} & = x_{n} - F^{'} {(x_{n})}^{- 1} F (x_{n}), \\ z_{n} & = ϕ_{4} (x_{n}, y_{n}) : = y_{n} - {[y_{n}, x_{n}; F]}^{- 1} F^{'} (x_{n}) {[y_{n}, x_{n}; F]}^{- 1} F^{'} (y_{n}), \\ x_{n + 1} & = ϕ_{8} (x_{n}, y_{n}, z_{n}) . \end{matrix}$

(34)

Clearly method (34) is a special case of Scheme (2). If $X = Y = R,$ then Method (34) reduces to Kung-Traub method [15]. We shall follow the proof of Theorem 1 but first we need to show that ${[y_{n}, x_{n}; F]}^{- 1} \in L (Y, X)$ . We get that

$\begin{matrix} ∥F^{'} {(x^{*})}^{- 1} ([y_{n}, x_{n}; F] - F^{'} (x^{*}))∥ & \leq K_{0} ∥ y_{n} - x^{*} ∥ + K_{1} ∥ x_{n} - x^{*} ∥, \\ \leq (K_{0} g_{1} (∥ x_{n} - x^{*} ∥) + K_{1}) ∥ x_{n} - x^{*} ∥, \\ = p_{0} (∥ x_{n} - x^{*} ∥) . \end{matrix}$

(35)

As in the case of function p, function $h_{p_{0}} = p_{0} (t) - 1$ , where $p_{0} (t) = (K_{0} g_{1} (t) + K_{1}) t$ has a smallest zero denoted by $r_{p_{0}}$ in the interval $(0, \frac{1}{L_{0}})$ . Set $l = r_{p_{0}}$ . Then, we have from the last sub step of Method (34) that

$\begin{matrix} ∥ z_{n} - x^{*} ∥ & \leq ∥ y_{n} - x^{*} ∥ + ∥{[y_{n}, x_{n}; F]}^{- 1} F^{'} (x^{*})∥ ∥F^{'} {(x^{*})}^{- 1} F^{'} (x_{n})∥, \\ ∥{[y_{n}, x_{n}; F]}^{- 1} F^{'} (x_{n})∥ ∥F^{'} {(x^{*})}^{- 1} F (y_{n})∥, \\ \leq ∥ y_{n} - x^{*} ∥ + \frac{M^{2}}{{(1 - p_{0} (∥ x_{n} - x^{*} ∥))}^{2}} ∥ y_{n} - x^{*} ∥, \\ (1 + \frac{M^{2}}{{(1 - p_{0} (∥ x_{n} - x^{*} ∥))}^{2}}) g_{1} (∥ x_{n} - x^{*} ∥) ∥ x_{n} - x^{*} ∥, \\ (1 + \frac{M^{2}}{{(1 - p_{0} (∥ x_{n} - x^{*} ∥))}^{2}}) \frac{L ∥ x_{n} - x^{*} ∥^{2}}{1 - L_{0} ∥ x_{n} - x^{*} ∥} . \end{matrix}$

(36)

It follows from Equation (36) that $λ = 2$ and $g_{2} (t) = \frac{L}{1 - L_{0} t} (1 + \frac{M^{2}}{{(1 - p_{0} (t))}^{2}})$ . Then, the convergence radius is given by

$r = min {r_{1}, r_{2}, r_{p_{0}}, r_{3}} .$

(37)

3. Numerical Example and Applications

In this section, we shall check the effectiveness and validity of our theoretical results which we have proposed in Section 2 on the scheme proposed by Sharma and Arora [1]. For this purpose, we shall choose a variety of nonlinear equations which are mentioned in the following examples including motivational example. At this point, we chose the following eighth order methods proposed by Sharma and Arora [1]

\{\begin{matrix} y_{n} & = x_{n} - F^{'} {(x_{n})}^{- 1} F (x_{n}), \\ z_{n} & = y_{n} - {(2 [y_{n}, x_{n}; F] - F^{'} (x_{n}))}^{- 1} F (y_{n}), \\ x_{n + 1} & = ϕ_{8} (x_{n}, y_{n}, z_{n}), \end{matrix}

(38)

\{\begin{matrix} y_{n} & = x_{n} - F^{'} {(x_{n})}^{- 1} F (x_{n}), \\ z_{n} & = y_{n} - {({[y_{n}, x_{n}; F]}^{2})}^{- 1} F^{'} (x_{n}) F (y_{n}), \\ x_{n + 1} & = ϕ_{8} (x_{n}, y_{n}, z_{n}) \end{matrix}

(39)

and

\{\begin{matrix} y_{n} & = x_{n} - F^{'} {(x_{n})}^{- 1} F (x_{n}), \\ z_{n} & = y_{n} - (2 {[y_{n}, x_{n}; F]}^{- 1} - F^{'} {(x_{n})}^{- 1}) F (y_{n}), \\ x_{n + 1} & = ϕ_{8} (x_{n}, y_{n}, z_{n}), \end{matrix}

(40)

denoted by

M_{1}

M_{2}

and

M_{3}

, respectively.

The initial guesses

x_{0}

are selected with in the range of convergence domain which gives guarantee for convergence of the iterative methods. Due to the pages limit, all the values of parameters are done for only 5 significant digits and displayed in the Table 1, Table 2 and Table 3 and examples Equations (1)–(3), although 100 significant digits are available. The considered test examples with corresponding initial guess, radius of convergence and necessary number of iterations (n) for getting the desired accuracy are displayed in Table 1, Table 2 and Table 3.

In addition, we also want to verify the theoretical order of convergence of Methods (38)–(40). Therefore, we calculate the computational order of convergence (COC) [9] approximated by using the following formulas

ρ = \frac{l n \frac{∥ x_{n + 2} - x^{*} ∥}{∥ x_{n + 1} - x^{*} ∥}}{l n \frac{∥ x_{n + 1} - x^{*} ∥}{∥ x_{n} - x^{*} ∥}}, for each = 0, 1, 2, \dots

(41)

or the approximate computational order of convergence (ACOC) [9]

ρ^{*} = \frac{l n \frac{∥ x_{n + 2} - x_{n + 1} ∥}{∥ x_{n + 1} - x_{n} ∥}}{l n \frac{∥ x_{n + 1} - x_{n} ∥}{∥ x_{n} - x_{n - 1} ∥}}, for each = 1, 2, \dots

(42)

During the current numerical experiments with programming language Mathematica (Version 9), all computations have been done with multiple precision arithmetic, which minimize round-off errors. We use

ϵ = 10^{- 200}

as a tolerance error. The following stopping criteria are used for computer programs:

(i) | x_{n + 1} - x_{n} | < ϵ

and

(i i) | f (x_{n + 1}) | < ϵ

Further, we use

λ = 2

and function

g_{2}

as defined above Equation (37) in all the examples.

Example 1.

Let

S = R, D = [- 1, 1], x^{*} = 0

and define function F on D by

F (x) = sin x .

(43)

Then, we get

L_{0} = L = M = K = 1

and

K_{0} = K_{1} = \frac{L_{0}}{2}

. We obtain different radius of convergence, COC (ρ) and n in the following Table 1.

Table 1. Different values of parameters which satisfy Theorem 1.

**Table 1.** Different values of parameters which satisfy Theorem 1.
Cases	$r_{R}$	$r_{1}$	$r_{2}$	$r_{p_{0}}$	$r_{3}$	r	$x_{0}$	n	ρ
$M_{1}$	0.66667	0.66667	0.28658	0.27229	0.76393	0.27229	0.25	4	9.0000
$M_{2}$	0.66667	0.66667	0.28658	0.27229	0.76393	0.27229	0.25	4	9.0000
$M_{3}$	0.66667	0.66667	0.28658	0.27229	0.76393	0.27229	0.25	4	9.0000

Example 2

Let

X = Y = C [0, 1]

, the space of continuous functions defined on

[0, 1]

be and equipped with the max norm. Let

D = \bar{U} (0, 1)

. Define function F on

D

F (φ) (x) = φ (x) - 5 \int_{0}^{1} x τ φ {(τ)}^{3} d τ,

(44)

we have that

F^{'} (φ (ξ)) (x) = ξ (x) - 15 \int_{0}^{1} x τ φ {(τ)}^{2} ξ (τ) d τ, f o r e a c h ξ \in D .

(45)

Then, for

x^{*} = 0,

we obtain that

L_{0} = 7.5, L = 15, M = K = 2

and

K_{0} = K_{1} = \frac{L_{0}}{2}

. We obtain different radius of convergence in the following Table 2.

Table 2. Different values of parameters which satisfy Theorem 1.

**Table 2.** Different values of parameters which satisfy Theorem 1.
$r_{R}$	$r_{1}$	$r_{2}$	$r_{p_{0}}$	$r_{3}$	r
0.044444	0.066667	0.011303	0.022046	0.088889	0.011303

Example 3

Returning back to the motivation example at the introduction on this paper, we have

L = L_{0} = \frac{2}{2 π + 1} (80 + 16 π + (11 + 12 log 2) π^{2})

M = K = 2, K_{0} = K_{1} = \frac{L_{0}}{2}

and our required zero is

x^{*} = \frac{1}{π}

. We obtain different radius of convergence, COC (ρ) and n in the following Table 3.

Table 3. Different values of parameters which satisfy Theorem 1.

**Table 3.** Different values of parameters which satisfy Theorem 1.
Cases	$r_{R}$	$r_{1}$	$r_{2}$	$r_{p_{0}}$	$r_{3}$	r	$x_{0}$	n	ρ
$M_{1}$	0.0075648	0.0075648	0.0016852	0.0094361	0.0086685	0.0016852	0.310	6	8.0000
$M_{2}$	0.0075648	0.0075648	0.0016852	0.0094361	0.0086685	0.0016852	0.310	6	8.0000
$M_{3}$	0.0075648	0.0075648	0.0016852	0.0094361	0.0086685	0.0016852	0.310	6	8.0000

4. Conclusions

Most of the time, researchers mentioned that the initial guess should be close to the required root for the granted convergence of their proposed schemes for solving nonlinear equations. However, how close an initial guess would be required to grantee the convergence of the proposed method? We propose the computable radius of convergence and error bound by using Lipschitz conditions in this paper. Further, we also reduce the hypotheses from fourth order derivative of the involved function to only first order derivative. It is worth noticing that Scheme (2) is not changing if we use the conditions of Theorem 1 instead of the stronger conditions proposed by Sharma and Arora (2015). Moreover, to obtain the error bounds in practice and order of convergence, we can use the computational order of convergence which is defined in numerical Section 3. Therefore,we obtain in practice the order of convergence in a way that avoids the bounds involving estimates higher than the first Fréchet derivative. Finally, on account of the results obtained in Section 3, it can be concluded that the proposed study not only expands the applicability but also gives the computable radius of convergence and error bound of the scheme given by Sharma and Arora (2015) for obtaining simple roots of nonlinear equations.

Acknowledgments

We would like to express our gratitude to the anonymous reviewers for their help with the publication of this paper.

Author Contributions

The contributions of all of the authors have been similar. All of them have worked together to develop the present manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

References

Sharma, J.R.; Arora, H. A new family of optimal eighth order methods with dynamics for nonlinear equations. Appl. Math. Comput. submitted..
Amat, S.; Busquier, S.; Plaza, S. Dynamics of the King and Jarratt iterations. Aequ. Math. 2005, 69, 212–223. [Google Scholar] [CrossRef]
Amat, S.; Busquier, S.; Plaza, S. Chaotic dynamics of a third-order Newton-type method. J. Math. Anal. Appl. 2010, 366, 24–32. [Google Scholar] [CrossRef]
Amat, S.; Hernández, M.A.; Romero, N. A modified Chebyshev’s iterative method with at least sixth order of convergence. Appl. Math. Comput. 2008, 206, 164–174. [Google Scholar] [CrossRef]
Argyros, I.K. Convergence and Application of Newton-Type Iterations; Springer: New York, NY, USA, 2008. [Google Scholar]
Argyros, I.K.; Hilout, S. Numerical Methods in Nonlinear Analysis; World Scientific Publishing Company: Hackensack, NJ, USA, 2013. [Google Scholar]
Behl, R.; Motsa, S.S. Geometric construction of eighth-order optimal families of Ostrowski’s method. Sci. World J. 2015, 2015. [Google Scholar] [CrossRef] [PubMed]
Ezquerro, J.A.; Hernández, M.A. New iterations of R-order four with reduced computational cost. BIT Numer. Math. 2009, 49, 325–342. [Google Scholar] [CrossRef]
Grau-Sánchez, M.; Noguera, M.; Gutiérrez, J.M. On some computational orders of convergence. Appl. Math. Lett. 2010, 23, 472–478. [Google Scholar] [CrossRef]
Kanwar, V.; Behl, R.; Sharma, K.K. Simply constructed family of a Ostrowski’s method with optimal order of convergence. Comput. Math. Appl. 2011, 62, 4021–4027. [Google Scholar] [CrossRef]
Magreñán, Á.A. Different anomalies in a Jarratt family of iterative root–finding methods. Appl. Math. Comput. 2014, 233, 29–38. [Google Scholar] [CrossRef]
Magreñán, Á.A. A new tool to study real dynamics: The convergence plane. Appl. Math. Comput. 2014, 248, 215–224. [Google Scholar] [CrossRef]
Petkovic, M.S.; Neta, B.; Petkovic, L.; Džunič, J. Multipoint Methods for Solving Nonlinear Equations; Academic press: Amsterdam, NY, USA, 2013. [Google Scholar]
Rheinboldt, W.C. An adaptive continuation process for solving systems of nonlinear equations. Pol. Acad. Sci. Banach Cent. Publ. 1978, 3, 129–142. [Google Scholar]
Traub, J.F. Iterative Methods for the Solution of Equations; Prentice-Hall Series in Automatic Computation: Englewood Cliffs, NJ, USA, 1964. [Google Scholar]
Weerakoon, S.; Fernando, T.G.I. A variant of Newton’s method with accelerated third order convergence. Appl. Math. Lett. 2000, 13, 87–93. [Google Scholar] [CrossRef]

© 2015 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/4.0/).

Share and Cite

MDPI and ACS Style

Argyros, I.K.; Behl, R.; Motsa, S.S. Local Convergence of an Efﬁcient High Convergence Order Method Using Hypothesis Only on the First Derivative. Algorithms 2015, 8, 1076-1087. https://doi.org/10.3390/a8041076

AMA Style

Argyros IK, Behl R, Motsa SS. Local Convergence of an Efﬁcient High Convergence Order Method Using Hypothesis Only on the First Derivative. Algorithms. 2015; 8(4):1076-1087. https://doi.org/10.3390/a8041076

Chicago/Turabian Style

Argyros, Ioannis K., Ramandeep Behl, and S.S. Motsa. 2015. "Local Convergence of an Efﬁcient High Convergence Order Method Using Hypothesis Only on the First Derivative" Algorithms 8, no. 4: 1076-1087. https://doi.org/10.3390/a8041076

APA Style

Argyros, I. K., Behl, R., & Motsa, S. S. (2015). Local Convergence of an Efﬁcient High Convergence Order Method Using Hypothesis Only on the First Derivative. Algorithms, 8(4), 1076-1087. https://doi.org/10.3390/a8041076

Article Menu

Local Convergence of an Efﬁcient High Convergence Order Method Using Hypothesis Only on the First Derivative

Abstract

1. Introduction

2. Local Convergence: One Dimensional Case

3. Numerical Example and Applications

4. Conclusions

Acknowledgments

Author Contributions

Conflicts of Interest

References

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI