Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.

1, March 2021

Fractional Newton-Raphson Method

A. Torres-Hernandez *,a,? and F. Brambila-Paz †,b,?
a Department of Physics, Faculty of Science - UNAM, Mexico
b Department of Mathematics, Faculty of Science - UNAM, Mexico

Abstract
The Newton-Raphson (N-R) method is useful to find the roots of a polynomial of degree n, with n ∈ N. How-
ever, this method is limited since it diverges for the case in which polynomials only have complex roots if a real
initial condition is taken. In the present work, we explain an iterative method that is created using the fractional
calculus, which we will call the Fractional Newton-Raphson (F N-R) Method, which has the ability to enter the
space of complex numbers given a real initial condition, which allows us to find both the real and complex roots
of a polynomial unlike the classical Newton-Raphson method.
Keywords: Newton-Raphson Method, Fractional Calculus, Fractional Derivative.

1. Newton-Raphson Method
Let Φ : Rn → Rn be a function. It is possible to build a sequence {xi }∞
i=0 by defining the following iterative method

xi+1 := Φ(xi ), (1)

if it fulfills that xi → ξ ∈ Rn and if the function Φ is continuous around ξ, we obtain that
 
ξ = lim xi+1 = lim Φ(xi ) = Φ lim xi = Φ(ξ), (2)
i→∞ i→∞ i→∞
the above result is the reason by which the method (1) is known as the fixed point method. Moreover, the
function Φ is called an iteration function. To understand the nature of the convergence of the iteration function
Φ, the following definition is necessary [1]:
Definition 1.1. Let Φ : Rn → Rn be an iteration function. The method (1) for determining ξ ∈ Rn is called (locally)
convergent, if there exists δ > 0 such that for every initial value
n o
x0 ∈ B(ξ; δ) := y ∈ Rn : y − ξ < δ ,
it holds that

lim kxi − ξk → 0 ⇒ lim xi = ξ. (3)

i→∞ i→∞
For the one-dimensional case, the N-R method is one of the most used method to find the roots ξ of a function
f : Ω ⊂ R → R, that is, {ξ ∈ Ω : f (ξ) = 0}, due to its easy implementation and rapid convergence, the N-R method
is expressed in terms of an iteration function Φ : R → R, as follows [1]:

 −1
xi+1 := Φ(xi ) = xi − f (1) (xi ) f (xi ), i = 0, 1, 2, · · · . (4)
The N-R method is based on creating a sequence {xi }∞
by means of the intersection of the tangent line of
i=0
the function f (x) at the xi point with the x axis, if the initial condition x0 is close enough to the root ξ then the
sequence {xi }∞
i=0 should be convergent to the root ξ [2].
* E-mail: [Link]@[Link]; ORCID: 0000-0001-6496-9505
† E-mail: fernandobrambila@[Link]; ORCID: 0000-0001-7896-6460
?Corresponding authors

DOI : 10.5121/mathsj.2021.8101 1
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

Figure 1: Illustration of the Newton-Raphson method.

For the following results in this section, it is necessary to mention that although the absolute value is used,
these are also valid for the case of n dimensions [2–5], in that case, it is necessary to substitute the absolute value
for some norm, that is, | · | → k · k . Before continuing it is necessary to consider the following definition [1]:
Definition 1.2. Let Φ : Ω ⊂ R → R be an iteration function with a fixed point ξ ∈ Ω. Then the method (1) is called
(locally) convergent of (at least) order p (p ≥ 1), if there are exists δ > 0 and C a non-negative constant, with C < 1 if
p = 1, such that for any initial value x0 ∈ B(ξ; δ) it holds that

|xi+1 − ξ| ≤ C |xi − ξ|p , i = 0, 1, 2, · · · , (5)

where C is called convergence factor.

The order of convergence is usually related to the speed at which the sequence generated by (1) converges. For
the particular case p = 1 it is said that the method (1) has an order of convergence (at least) linear, and for the
case p = 2 it is said that the method (1) has an order of convergence (at least) quadratic. The following theorem,
allows characterizing the order of convergence of an iteration function Φ with its derivatives [1, 2] :
Theorem 1.3. Let Φ : Ω ⊂ R → R be an iteration function with a fixed point ξ ∈ Ω. Assuming that Φ is p-times
differentiable in ξ for some p ∈ N, and moreover


(k)
 Φ (ξ) = 0, ∀k < p,

 if p ≥ 2
, (6)
Φ (1) (ξ) < 1,


 if p = 1

then Φ is (locally) convergent of (at least) order p.

Proof. Let Φ : R → R be an iteration function, and using the Taylor series expansion of Φ, we obtain two cases:

i) Case p ≥ 2 :

p
X Φ (k) (ξ)
Φ(xi ) = Φ(ξ) + (xi − ξ)k + o ((xi − ξ)p ) ,
k!
k=1

then

p
X Φ (k) (ξ)
|Φ(xi ) − Φ(ξ)| ≤ |xi − ξ|k + o (|xi − ξ|p ) ,
k!
k=1

assuming that ξ is a fixed point of Φ and that Φ (k) (ξ) = 0 ∀k < p is fulfilled, the previous expression implies
that

2
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

|Φ(xi ) − Φ(ξ)| |xi+1 − ξ| Φ (p) (ξ) o (|xi − ξ|p )

= ≤ + ,
|xi − ξ|p |xi − ξ|p p! |xi − ξ|p

therefore

|xi+1 − ξ| Φ (p) (ξ)

lim p ≤ ,
i→∞ |xi − ξ| p!

as a consequence, if the sequence {xi }∞

i=0 generated by (1) converges to ξ, there exists a value k > 0 such that

Φ (p) (ξ)
|xi+1 − ξ| ≤ |xi − ξ|p , ∀i ≥ k,
p!

then Φ is (locally) convergent of (at least) order p.

ii) Case p = 1 :

Φ(xi ) = Φ(ξ) + Φ (1) (ξ)(xi − ξ) + o ((xi − ξ)) ,

then

|Φ(xi ) − Φ(ξ)| ≤ Φ (1) (ξ) |xi − ξ| + o (|xi − ξ|) ,

assuming that ξ is a fixed point of Φ, the previous expression implies that

|Φ(xi ) − Φ(ξ)| |xi+1 − ξ| o (|xi − ξ|)

= ≤ Φ (1) (ξ) + ,
|xi − ξ| |xi − ξ| |xi − ξ|

therefore

|xi+1 − ξ|
lim ≤ Φ (1) (ξ) ,
i→∞ |xi − ξ|

as a consequence, if the sequence {xi }∞

i=0 generated by (1) converges to ξ, there exists a value k > 0 such that

|xi+1 − ξ| ≤ Φ (1) (ξ) |xi − ξ| , ∀i ≥ k,

then considering m ≥ 1

2 m
|xi+m − ξ| ≤ Φ (1) (ξ) |xi+m−1 − ξ| ≤ Φ (1) (ξ) |xi+m−2 − ξ| ≤ · · · ≤ Φ (1) (ξ) |xi − ξ| ,

and assuming that Φ (1) (ξ) < 1 is fulfilled

m
lim |xi+m − ξ| ≤ lim Φ (1) (ξ) |xi − ξ| → 0,
m→∞ m→∞

then Φ is (locally) convergent of order (at least) linear.

3
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

The N-R method is characterized by having an order of convergence at least quadratic for the case where
f (1) (ξ) , 0, but if to the previous case it is added that f (2) (ξ) = 0, then the N-R method presents an order of
convergence at least cubic. On other hand, for the case where the function f has a root ξ with a certain algebraic
multiplicity m ≥ 2, that is,

f (x) = (x − ξ)m g(x), g(ξ) , 0,

the N-R method presents an order of convergence at least linear [1]. The aforementioned may be formalized
by the following proposition:
Proposition 1.4. Let f : Ω ⊂ R → R be a function with a zero ξ ∈ Ω. Then the iteration function Φ of the N-R method,
given by (4), fulfills the following condition:

Φ (p) (ξ)
|xi+1 − ξ| ≤ |xi − ξ|p , (7)
p!
where

if f (x) = (x − ξ)m g(x)




 1,
if f (1) (ξ) , 0, and f (x) , (x − ξ)m g(x)

p= 2, , (8)



 3,

if f (1) (ξ) , 0, f (2) (ξ) = 0, and f (x) , (x − ξ)m g(x)

with g(ξ) , 0 and m ≥ 2.

Proof. Considering that the form of the function f is not explicitly determined, it is possible to consider two
possibilities:

i) Assuming the function may be written as f (x) = (x − ξ)m g(x) with g(ξ) , 0 and m ≥ 2, then

h i
f (1) (x) = (x − ξ)m−1 (x − ξ)g (1) (x) + mg(x) ,

as a consequence, the iteration function of N-R method takes the following form

Φ(x) = x − (x − ξ)h(x)g(x),

with

h i−1
h(x) = (x − ξ)g (1) (x) + mg(x) ,

then

h i
Φ (1) (x) = 1 − h(x) (x − ξ)g (1) (x) + g(x) − (x − ξ)h(1) (x)g(x),

where

h i−2 h i
h(1) (x) = − (x − ξ)g (1) (x) + mg(x) (1 + m) g (1) (x) + (x − ξ)g (2) (x) ,

therefore

1
lim Φ (1) (x) = |1 − h(ξ)g(ξ)| = 1 − < 1, (9)
x→ξ m
and from the Theorem 1.3, the N-R method has an order of convergence at least linear, that is, the N-R
method fulfills the equation (7) with p = 1.

4
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

ii) Assuming that f (x) , (x − ξ)m g(x) with g(ξ) , 0 and m ≥ 2, the first derivative of the iteration function of
N-R method takes the following form

 −2
Φ (1) (x) = f (1) (x) f (x)f (2) (x),

and if it fulfills that f (1) (ξ) , 0, then

lim Φ (1) (x) = 0, (10)

x→ξ

and from the Theorem 1.3, the N-R method has an order of convergence at least quadratic, that is, the N-R
method fulfills the equation (7) with p = 2. On other hand, the second derivative of the iteration function of
N-R method takes the following form

 −1  −2  −3  2 

Φ (2) (x) = f (1) (x) f (2) (x) + f (x) f (1) (x) f (3) (x) − 2 f (1) (x) f (2) (x) ,

and if it fulfills that f (1) (ξ) , 0 and f (2) (ξ) = 0, then

lim Φ (1) (x) = lim Φ (2) (x) = 0, (11)

x→ξ x→ξ

and from the Theorem 1.3, the N-R method has an order of convergence at least cubic, that is, the N-R
method fulfills the equation (7) with p = 3.

The previous proposition, illustrates two important points that are worth mentioning when using the N-R
method to find the zeros of a function f :

i) When it is not evident, unless it is explicitly specified that the function f has no roots of algebraic multiplic-
ity m ≥ 2, technically there exists the possibility that the N-R method has an order of convergence at least
linear, that is, the N-R method may fulfill the equation (7) with p ≥ 1.
ii) Due that the N-R method is a local iterative method, even if it proves that for a root ξ ∈ Ω the method has
an order of convergence at least linear, this does not rule out that for the same function f it may present
a higher order of convergence over the same region Ω. As an example of the above, we may consider the
following function

f (x) = (x − η)(x − ξ)m g(x), g(η) , g(ξ) , 0,

with η, ξ ∈ Ω, η − ξ < , and m ≥ 2.

The previous points are important, because when the N-R method is implemented in a function f , the zeros
of the function are assumed to be unknown, and their algebraic multiplicities m ≥ 2, in case they exist, are also
unknown. With the above in mind, the following corollary is obtained, which is derived from the Theorem 1.3:
Corollary 1.5. Let Φ : R → R be an iteration function. If Φ defines a sequence {xi }∞
i=0 such that xi → ξ, and if the
following condition is fulfilled

lim Φ (1) (x) , 0, (12)

x→ξ

then Φ has an order of convergence (at least) linear in B(ξ; δ).

5
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

2. Fractional Calculus
The fractional calculus is a mathematical analysis branch whose applications have been increasing since the end
of the XX century and beginnings of the XXI century [6–8], the fractional calculus arises around 1695 due to
Leibniz’s notation for the derivatives of integer order

dn
f (n) (x) := f (x), n ∈ N,
dxn
thanks to this notation L’Hopital could ask in a letter to Leibniz about the interpretation of taking n = 1/2 in a
derivative, since at that moment Leibniz could not give a physical or geometrical interpretation to this question, he
simply answered L’Hopital in a letter, “. . . is an apparent paradox of which, one day, useful consequences will be
drawn” [9]. The name of fractional calculus comes from a historical question since in this branch of mathematical
analysis it is studied the derivatives and integrals of a certain order α, with α ∈ R or C.
Currently, the fractional calculus does not have a unified definition of what is considered a fractional deriva-
tive, because one of the conditions required to consider an expression as a fractional derivative is to recover the
results of conventional calculus when the order α → n, with n ∈ N [10], among the most common definitions
of fractional derivatives are the Riemann-Liouville (R-L) fractional derivative and the Caputo fractional deriva-
tive [11], the latter is usually the most studied since the Caputo fractional derivative allows us a physical inter-
pretation to problems with initial conditions, this derivative fulfills the property of the classical calculus that the
derivative of a constant is null regardless of the order α of the derivative, however this does not occur with the
R-L fractional derivative, and this characteristic can be used to solve nonlinear systems [4, 5, 12].
Unlike the Caputo fractional derivative, the R-L fractional derivative does not allow for a physical interpreta-
tion to the problems with initial condition because its use induces fractional initial conditions, however the fact
that this derivative does not cancel the constants for α, with α < N, allows to obtain a “spectrum” of the behavior
of the constants for different orders of the derivative, which is not possible with conventional calculus.

2.1. Introduction to the Riemann-Liouville Fractional Derivative

One of the key pieces in the study of fractional calculus is the iterated integral, which is defined as follows [11]
Definition 2.1. Let L1loc (a, b), the space of locally integrable functions in the interval (a, b). If f is a function such that
f ∈ L1loc (a, ∞), then the n-th iterated integral of the function f is given by

Z x
n

n−1
 1
a Ix f (x) = a Ix a Ix f (x) = (x − t)n−1 f (t)dt, (13)
(n − 1)! a

where

Z x
a Ix f (x) := f (t)dt.
a

Considerate that (n − 1)! = Γ (n) , a generalization of (13) may be obtained for an arbitrary order α > 0

Z x
α 1
a Ix f (x) = (x − t)α−1 f (t)dt, (14)
Γ (α) a

similarly, if f ∈ L1loc (−∞, b), we may define

Z b
α 1
x Ib f (x) = (t − x)α−1 f (t)dt, (15)
Γ (α) x

the equations (14) and (15) correspond to the definitions of right and left Riemann-Liouville fractional inte-
gral, respectively. The fractional integrals fulfill the semigroup property, which is given in the following propo-
sition [11]:

6
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

Proposition 2.2. Let f be a function. If f ∈ L1loc (a, ∞), then the fractional integrals of f fulfill that

α β α+β
a Ix a Ix f (x) = a Ix f (x), α, β > 0. (16)
From the previous result, and considering that the operator d/dx is the inverse operator to the left of the
operator a Ix , any integral α-th of a function f ∈ L1loc (a, ∞) may be written as

α dn n α dn
a Ix f (x) = ( a I x a I x f (x)) = ( I n+α f (x)) . (17)
dxn dxn a x
Considering (14) and (17), we can build the Riemann-Liouville fractional derivative a Dxα , as follows [11, 13]:

 −α
 n a Ix f (x), if α < 0

α

a Dx f (x) :=  d , (18)
 ( I n−α f (x)) , if α ≥ 0
dxn a x


where n = dαe, then applying the operator (18) to the function xµ , with α ∈ R \ Z and µ > −1, we obtain the
following result

α µ Γ (µ + 1) µ−α
0 Dx x = x . (19)
Γ (µ − α + 1)

2.2. Introduction to the Caputo Fractional Derivative

Michele Caputo (1969) published a book and introduced a new definition of fractional derivative, he created this
definition with the objective of modeling anomalous diffusion phenomena. The definition of Caputo had already
been discovered independently by Gerasimov (1948). This fractional derivative is of the utmost importance since it
allows us to give a physical interpretation of the initial value problems, moreover to being used to model fractional
time. In some texts, it is known as the fractional derivative of Gerasimov-Caputo.
Let f be a function, such that f is n-times differentiable with f (n) ∈ L1loc (a, b), then the (right) Caputo fractional
derivative is defined as [13]

Zx
dn
!
1
C α
a Dx f (x) :=a Ixn−α f (x) = (x − t)n−α−1 f (n) (t)dt, (20)
dxn Γ (n − α) a
where n = dαe. It should be mentioned that the Caputo fractional derivative behaves as the inverse operator to
the left of the Riemann-Liouville fractional integral, that is,

C α α
a Dx (a Ix f (x)) = f (x).
On the other hand, the relation between the fractional derivatives of Caputo and Riemann-Liouville is given
by the following expression [13]

 n−1 (k)

C α
 X f (a) 
a Dx f (x) = a Dxα f (x) − k
(x − a)  ,
 
k!
k=0

then, if f (k) (a) = 0 ∀k < n, we obtain

C α
a Dx f (x) = a Dxα f (x),
considering the previous particular case, it is possible to unify the definitions of R-L fractional integral and
Caputo fractional derivative as follows

 −α

 a Ix f (x), ! if α < 0
C D α f (x) := n

a x

n−α d . (21)
 a Ix f (x) , if α ≥ 0


dxn


7
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

3. Fractional Newton-Raphson Method

Let Pn (R) be the space of polynomials of degree less than or equal to n ∈ N with real coefficients. The N-R method
is useful for finding the roots of a function f ∈ Pn (R). However, this method is limited because it cannot find
roots ξ ∈ C \ R, if the sequence {xi }∞
i=0 generated by (4) has an initial condition x0 ∈ R. To solve this problem and
develop a method that has the ability to find roots, both real and complex, of a polynomial if the initial condition
x0 is real, we propose a new method, which consists of the Newton-Raphson method with the implementation of
the fractional derivatives. Before continuing, it is necessary to define the following notation

dα
f (α) (x) := f (x), (22)
dxα
where the operator d α /dxα denotes any fractional derivative applied on the variable x, that fulfills the follow-
ing condition of continuity respect to the order of the derivative

lim f (α) (x) = f (1) (x). (23)

α→1

Considering a function Φ : (R \ Z) × C → C. Then, using as a basis the idea of the N-R method (4), and con-
sidering any fractional derivative that fulfills the condition (23), we can define the Fractional Newton-Raphson
method as follows (for the case in n dimensions consult the reference [5]):

 −1
xi+1 := Φ (α, xi ) = xi − f (α) (xi ) f (xi ), i = 0, 1, 2, · · · . (24)

For the above expression to make sense, due to the part of the integral operator that fractional derivatives
usually have, and that the F N-R method can be used in a wide variety of functions [5], we consider in the expres-
sion (24) that the fractional derivative is obtained for a real variable x, and if the result allows it, this variable is
subsequently substituted by a complex variable xi , that is,

f (α) (xi ) := f (α) (x) , x ∈ R, xi ∈ C. (25)

x−→xi

It should be mentioned that in general, in the F N-R method Φ (1) (α, ξ) , 0 if f (ξ) = 0, and from the Corollary
1.5, the Proposition 1.4 and the condition (23), any sequence {xi }∞ i=0 generated by the iterative method (24) has
an order of convergence at least linear, that is, the F N-R method may fulfill the equation (7) with p ≥ 1, which
becomes more evident when considering α ∈ [1 − , 1 + ] \ {1}.
To understand why the F N-R method, if f ∈ Pn (R), has the ability to enter the complex space using a real
initial condition unlike the classical N-R method, it is enough to observe the R-L fractional derivative (19), with
α = 1/2, of the constant function f0 (x) = x0 and the identity function f1 (x) = x1 :

1/2 Γ (1) −1/2 1/2 Γ (2) 1/2

0 Dx f0 (x) = x , 0 Dx f1 (x) = x .
Γ (1/2) Γ (3/2)

Figure 2: The R-L fractional derivatives of f0 (x) and f1 (x), with α ∈ [0, 1].

8
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

For polynomials of degree n ≥ 1, in the F N-R method the initial condition x0 must be taken different to zero,
as a consequence of the R-L fractional derivative of order α, with α < Z, of the constants are proportional to the
function x−α . When using the F N-R method, with the R-L fractional derivative, on a function f ∈ Pn (R), presents
among its behaviors, the following particular cases depending on the initial condition x0 :
i) If we take an initial condition x0 > 0, the sequence {xi }∞
i=0 may be divided into three parts, this occurs because
it may exists a value M ∈ N for which {xi }M−1
i=0 ⊂ R >0 with {xM } ⊂ R<0 , in consequence {xi }i≥M+1 ⊂ C.
ii) On the other hand, if we take an initial condition x0 < 0, the sequence {xi }∞ i=0 may be divided into two parts,
{x0 } ⊂ R<0 and {xi }i≥1 ⊂ C.

3.1. Advantages of the Fractional Newton-Raphson Method

One of the main advantages of the F N-R method is that the initial condition x0 can be left fixed, and so vary
the order α of the derivative to obtain both real and complex roots of a polynomial. Due that the order α of the
derivative is varied, different values of α can throw the same root but with a different number of iteration, so to
optimize the method, it is possible to implement a filter in which once we have obtained the roots, only those
whose orders of the derivatives have generated a smaller number of iterations are extracted.

Figure 3: Illustration of some lines generated by the F N-R method.

a) α = −0.77 b) α = −0.32 c) α = 0.19 d) α = 1.87

Figure 4: llustrations of some trajectories generated by the F N-R method for the same initial condition x0 but
with different values of α.

Another advantage is a consequence that the method provides complex roots, so once a root is obtained it is
enough to obtain its conjugate complex to obtain another root, in essence, it could be considered that they extract
two roots with the same order of the derivative and the same number of iterations. The method does not guarantee
that all roots of the polynomial are found by leaving an initial condition fixed and by varying the orders α of the
derivative, as in the classical N-R method, finding the roots will depend on giving an appropriate initial condition.

9
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

3.2. Results of the Fractional Newton-Raphson Method

The following examples are solved using the R-L fractional derivative (19). Instructions for implementing the F N-
R method, along with information to provide values α ∈ [0.7, 1.3] \ {1} are found in the reference [5]. For rounding
reasons, for the examples the following function is defined


 Re (x) ,

 if |Im (x)| ≤ 10−m
Rndm (x) :=  . (26)

 x, if |Im (x)| > 10−m

Combining the function (26) with the method (24), the following iterative method is defined

xi+1 := Rnd5 (Φ(α, xi )) , i = 0, 1, 2 · · · . (27)

Example 3.1. Let f be a function, with

f (x) = − 64.23x14 − 72.74x13 − 61.66x12 + 32.26x11

+ 32.3x10 − 41.37x9 + 20.18x8 + 4.32x7
− 5.67x6 + 17.41x5 − 78.6x4 − 48.27x3
− 19.31x2 + 77.92x − 45.03.

Then the initial condition x0 = 0.68 is chosen to use the iterative method given by (27). Consequently, we obtain the
results of the Table 1.

α xn kxn − xn−1 k2 kf (xn )k2 n

1 0.97399 0.89785306 − 0.29205148i 1.65245E − 06 9.65908E − 05 17
2 0.97455 0.53497472 − 0.82703363i 6.75426E − 07 7.71022E − 05 19
3 0.97631 −0.07482893 − 1.0188353i 2.60768E − 07 3.30325E − 05 17
4 0.98539 −0.6673652 − 1.16572645i 1.06301E − 07 3.96105E − 05 26
5 0.99275 −0.67766753 − 0.66659064i 1.34536E − 07 6.10746E − 06 18
6 1.00575 −0.07482895 + 1.01883532i 3.83275E − 07 1.21870E − 05 22
7 1.00623 −0.6673652 + 1.16572646i 1.66433E − 07 5.94548E − 05 32
8 1.00635 −0.67766754 + 0.66659064i 1.50333E − 07 2.08269E − 06 19
9 1.00643 0.53497473 + 0.82703358i 1.70294E − 07 9.71477E − 06 21
10 1.03515 −1.09479584 − 0.25179059i 1.44222E − 07 5.21878E − 05 26
11 1.15163 −1.09479581 + 0.25179059i 6.08276E − 08 9.98791E − 05 25
12 1.15715 0.51558361 − 0.33422342i 1.03121E − 06 5.06052E − 05 15
13 1.16239 0.51558364 + 0.33422325i 1.37117E − 06 6.92931E − 05 14
14 1.16731 0.89785308 + 0.29205152i 1.70880E − 07 9.00285E − 05 18

Table 1: Results obtained using the iterative method (27).

Example 3.2. Let f be a function, with

f (x) = − 96.98x15 − 96.82x14 − 3.87x13 + 25.78x12

+ 90.68x11 + 48.05x10 + 50.54x9 − 5.16x8
+ 47.01x7 + 90.23x6 + 87.09x5 + 53.09x4
+ 15.38x3 + 97.98x2 − 61.98x + 14.69.

Then the initial condition x0 = 0.15 is chosen to use the iterative method given by (27). Consequently, we obtain the
results of the Table 2.

10
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

α xn kxn − xn−1 k2 kf (xn )k2 n

1 0.88451 0.7739975 − 0.54762173i 1.17047E − 07 5.48998E − 05 28
2 0.90499 −0.82526288 + 0.64969528i 1.14018E − 07 7.66665E − 05 24
3 0.90731 0.03271742 + 1.02608471i 1.14018E − 07 4.71095E − 05 19
4 0.90863 −0.48361539 + 0.928383i 1.16619E − 07 6.49343E − 05 28
5 0.90923 0.03271738 − 1.02608473i 1.30384E − 07 5.48149E − 05 22
6 0.93627 0.28667103 + 0.19437684i 4.73466E − 06 3.81068E − 05 12
7 0.94059 0.30392964 + 0.77330882i 9.96092E − 07 6.75516E − 05 16
8 0.94155 0.77399751 + 0.54762167i 1.14018E − 07 3.18386E − 05 14
9 0.95179 0.2866711 − 0.19437464i 1.46720E − 05 9.07085E − 05 9
10 0.95499 −1.16959779 + 0.06354745i 1.01980E − 07 1.49833E − 05 15
11 0.99283 1.16397068 1.90000E − 07 6.26264E − 05 6
12 1.04799 −0.48361536 − 0.92838301i 2.15407E − 07 7.67702E − 05 31
13 1.05463 0.30392985 − 0.77330891i 7.60000E − 07 5.64786E − 05 18
14 1.06431 −1.16959776 − 0.06354748i 2.00998E − 07 6.59751E − 05 14
15 1.06455 −0.82526289 − 0.64969531i 4.12311E − 08 2.72233E − 05 21

Table 2: Results obtained using the iterative method (27).

Example 3.3. Let f be a function, with

f (x) = − 57.62x16 − 56.69x15 − 37.39x14 − 19.91x13 + 35.8312

− 72.4711 + 44.41x10 + 43.53x9 + 59.93x8
− 42.9x7 − 54.24x6 + 72.12x5 − 22.92x4
+ 56.39x3 + 15.8x2 + 60.05x + 55.31.

Then the initial condition x0 = 0.83 is chosen to use the iterative method given by (27). Consequently, we obtain the
results of the Table 3.

α xn kxn − xn−1 k2 kf (xn )k2 n

1 0.81691 0.8812118 − 0.42696217i 1.08167E − 07 8.16395E − 05 49
2 0.83851 1.03423973 6.00000E − 08 7.69988E − 05 56
3 0.97383 −1.0013396 6.37436E − 05 6.64131E − 05 7
4 0.99055 −0.35983764 + 1.18135267i 6.70820E − 08 2.53547E − 05 21
5 0.99059 −0.70050491 + 0.78577099i 1.70294E − 07 9.13799E − 06 17
6 0.99219 −0.70050494 − 0.78577099i 1.18229E − 06 5.28258E − 05 7
7 0.99283 0.36452491 − 0.83287828i 3.63610E − 06 4.30167E − 05 17
8 0.99347 −0.28661378 − 0.8084062i 1.38226E − 05 9.04752E − 05 8
9 0.99427 −0.35983765 − 1.18135267i 1.26491E − 07 4.09162E − 05 16
10 0.99539 −1.3699527 2.30000E − 07 7.02720E − 05 14
11 1.12775 −0.62435238 1.25000E − 06 6.46233E − 05 4
12 1.16423 0.58999229 − 0.86699687i 7.07107E − 08 7.38972E − 05 25
13 1.16595 0.36452487 + 0.83287805i 3.06105E − 07 9.42729E − 05 15
14 1.16607 0.58999222 + 0.86699689i 5.00000E − 08 5.09054E − 05 18
15 1.16647 0.88121183 + 0.42696223i 4.12311E − 08 5.37070E − 05 39
16 1.20923 −0.28661363 + 0.8084063i 1.94165E − 07 5.02799E − 05 16

Table 3: Results obtained using the iterative method (27).

11
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

Example 3.4. Let f be a function, with

x
f (x) = sin(x) − ,
50
and assuming that

 50 
(α)
X (−1)k 2k+1 x 
f (x) ≈ 0 Dxα  x −  .

Γ (2k + 2) 50
k=0

Then the initial condition x0 = 1.27 is chosen to use the iterative method given by (27). Consequently, we obtain the
results of the Table 4.

α xn kxn − xn−1 k2 kf (xn )k2 n

1 0.94227 3.07995452 8.41000E − 06 2.07258E − 08 6
2 0.96327 −33.81479691 6.34000E − 06 2.09752E − 07 12
3 0.96352 −32.11337988 6.26000E − 06 2.12854E − 07 11
4 0.96378 −27.68747328 6.85000E − 06 1.66654E − 07 11
5 0.96433 −25.67192859 8.22000E − 06 2.42883E − 07 10
6 0.96472 −21.54564423 9.87000E − 06 1.68227E − 07 10
7 0.96531 −19.24464801 9.97000E − 06 2.24310E − 07 10
8 0.96643 −15.39497861 9.28000E − 06 9.12247E − 08 9
9 0.97064 −9.23893132 9.97000E − 06 3.50056E − 08 8
10 0.97829 −6.41177493 8.59000E − 06 3.92845E − 08 6
11 1.05026 −3.07995462 8.44000E − 06 8.10843E − 08 2
12 1.08752 0.0000102 6.95000E − 06 9.99600E − 06 14
13 1.16342 −12.82578588 7.44000E − 06 2.91050E − 07 8
14 1.23044 38.58047692 8.29000E − 06 1.61756E − 06 8
15 1.23297 33.81479547 4.70000E − 06 8.79795E − 07 6
16 1.23412 32.11337776 7.20000E − 06 1.36969E − 06 9
17 1.23728 27.6874726 5.66000E − 06 7.46478E − 07 5
18 1.24368 25.67192621 7.27000E − 06 1.75186E − 06 7
19 1.24500 21.54564384 6.34000E − 06 5.27961E − 07 5
20 1.24978 19.24464986 9.27000E − 06 1.89479E − 06 6
21 1.25961 15.394979 9.08000E − 06 2.87629E − 07 3
22 1.26676 6.41177455 2.66000E − 06 3.29978E − 07 6
23 1.26837 12.82578672 2.84000E − 06 5.04044E − 07 5
24 1.29773 9.23893132 1.04000E − 06 3.50056E − 08 3

Table 4: Results obtained using the iterative method (27).

4. Conclusions
The F N-R method is very efficient to find roots of polynomials since it does not present the divergence problems,
like the classical N-R method, for a polynomial with only complex roots when using real initial conditions. How-
ever, the really interesting thing is that this method opens up the possibility of creating new fractional iterative
methods in one dimension [14–17] or in multiple dimensions [5, 18], as well as opens the possibility of creating
new hybrid iterative methods by combining the F N-R method with existing iterative methods [1, 2]. So in this
work it has been given one more application to fractional calculus and has opened the possibility of extending the
capacity of the iterative methods that allow us to find zeros of functions more general than polynomials [4, 5].

12
 Applied Mathematics and Sciences: An International Journal (MathSJ) Vol.8, No.1, March 2021

References
[1] Robert Plato. Concise numerical mathematics. Number 57. American Mathematical Soc., 2003.

[2] Josef Stoer and Roland Bulirsch. Introduction to numerical analysis, volume 12. Springer Science & Business
Media, 2013.
[3] Richard L Burden and J Douglas Faires. Numerical analysis. Thomson Learning,, 2002.
[4] A. Torres-Hernandez, F. Brambila-Paz, P. M. Rodrigo, and E. De-la-Vega. Reduction of a nonlinear system
and its numerical solution using a fractional iterative method. Journal of Mathematics and Statistical Science,
6:285–299, 2020. [Link]
[5] A. Torres-Hernandez, F. Brambila-Paz, and E. De-la-Vega. Fractional newton-raphson method and some
variants for the solution of nonlinear systems. Applied Mathematics and Sciences: An International Journal
(MathSJ), 7:13–27, 2020. DOI: 10.5121/mathsj.2020.7102.

[6] Fernando Brambila. Fractal Analysis: Applications in Physics, Engineering and Technology. IntechOpen, 2017.
[7] Carlos Alberto Torres Martı́nez and Carlos Fuentes. Applications of radial basis function schemes to frac-
tional partial differential equations. Fractal Analysis: Applications in Physics, Engineering and Technology,
pages 4–20, 2017. DOI: 10.5772/67892.

[8] Benito F. Martı́nez-Salgado, Rolando Rosas-Sampayo, Anthony Torres-Hernández, and Carlos Fuentes. Ap-
plication of fractional calculus to oil industry. Fractal Analysis: Applications in Physics, Engineering and Tech-
nology, pages 21–42, 2017. DOI: 10.5772/intechopen.68571.
[9] Kenneth S. Miller and Bertram Ross. An introduction to the fractional calculus and fractional differential equa-
tions. Wiley-Interscience, 1993.

[10] Keith Oldham and Jerome Spanier. The fractional calculus theory and applications of differentiation and integra-
tion to arbitrary order, volume 111. Elsevier, 1974.
[11] Rudolf Hilfer. Applications of fractional calculus in physics. World Scientific, 2000.
[12] A. Torres-Hernandez, F. Brambila-Paz, and J.J. Brambila. A nonlinear system related to investment under
uncertainty solved using the fractional pseudo-newton method. Journal of Mathematical Sciences: Advances
and Application, 63:41–53, 2020. [Link]
[13] A.A. Kilbas, H.M. Srivastava, and J.J. Trujillo. Theory and Applications of Fractional Differential Equations.
Elsevier, 2006.

[14] Krzysztof Gdawiec, Wiesław Kotarski, and Agnieszka Lisowska. Visual analysis of the newton’s method with
fractional order derivatives. Symmetry, 11(9):1143, 2019.
[15] Krzysztof Gdawiec, Wiesław Kotarski, and Agnieszka Lisowska. Newton’s method with fractional derivatives
and various iteration processes via visual analysis. Numerical Algorithms, pages 1–58, 2020.
[16] Ali Akgül, Alicia Cordero, and Juan R Torregrosa. A fractional newton method with 2αth-order of conver-
gence and its stability. Applied Mathematics Letters, 98:344–351, 2019.
[17] Alicia Cordero, Ivan Girona, and Juan R Torregrosa. A variant of chebyshev’s method with 3αth-order of
convergence by using fractional derivatives. Symmetry, 11(8):1017, 2019.
[18] A. Torres-Hernandez, F. Brambila-Paz, P.M. Rodrigo, and E. De-la-Vega. Fractional pseudo-newton method
and its use in the solution of a nonlinear system that allows the construction of a hybrid solar receiver. Applied
Mathematics and Sciences: An International Journal (MathSJ), 7:1–12, 2020. DOI: 10.5121/mathsj.2020.7201.

13