Comparative Study of Interpolation

Formula

Research Project
Submitted to the Department of Mathematic in Partial Fulfillment of the
Requirements for the Degree of BSc. In Mathematic

By
Eman Abdulwahid Mousa

Supervised by
Dr. Ivan Subhi Latif

April 2024
Certification of the supervisors

I certify that this work was prepared under my supervision at the Department of

Mathematics/College of Education/Salahaddin University-Erbil in partial

fulfillment of the requirements for the degree of Bachelor of philosophy of Science

in Mathematics.

Signature:
Supervisor: Dr. Ivan Subhi Latif

Scientific grade: Assist. Professor

Date: /4/ 2023

In view of the available recommendations, I forward this work for debate by the

examining committee.

Signature:

Name: Dr. Rashad Rashid Haji

Scientific grade: Assist. Professor

Chairman of the Mathematics Department

Date: /4/ 2023

II
 Acknowledgment

I would like to thank Allah for giving me the power to complete this work. I
would like to present my profound thanks to supervisor and lecturer

Dr. Ivan S.Latif for his kind and valuable suggestions that assisted me to
accomplish this work.

I would also like to extend my gratitude to the head of the Mathematic

Department Assist. Prof. Dr. Rashad Rashid Haji, and thank my family for all the
love and support they have provided

Eman Abdulwahid Mousa

2024

III
 Dedication To

- My father and mother

- My dear supervisor

- My brothers and sisters

- All who want to read it

Eman Abdulwahid Mousa

2024

IV
 Contents

Certification of the supervisors:................................. Error! Bookmark not defined.

Acknowledgment ....................................................................................................... III
Dedication To ............................................................................................................ IV
Contents ...................................................................................................................... V
Abstract ...................................................................................................................... VI
Chapter one .................................................................................................................. 1
1. Introduction ........................................................................................................ 1
Chapter Two ................................................................................................................ 2
2- Some basic definitions: ....................................................................................... 2
2.1 - Interpolation: ............................................................................................... 2
2.2 - Premise of interpolation: ............................................................................. 2
3- Study of different central difference Interpolation formulas: ............................. 4
Chapter Three ............................................................................................................ 12
3. Numerical discussion: ........................................................................................ 12
Conclusion ................................................................................................................. 23
Reference ................................................................................................................... 24
Kurdish Abtract.......................................................................................................... 26

V
 Abstract

When taking care of issues utilizing numerical strategies it is generally important to

set up a model and to record conditions communicating the limitations and actual
laws that apply. These conditions should now be tackled and a decision introduces
itself. One route is to continue utilizing regular strategies for math, getting an
answer as a recipe, or set of formulae. Another strategy is to communicate the
conditions so that they might be addressed computationally, i.e. by utilizing
strategies for mathematical examination or Numerical analysis. In this paper we are
aimed to discuss interpolation, various methods to solve central difference
interpolation, their generalizations. Applications of interpolation are also discussed
and one can easily understand the concepts of the paper.

VI
 Chapter one

1. Introduction

At times, determining the value of a dependent variable for a given independent

variable poses a challenge when their explicit relationship is unknown. For
instance, consider estimating India's population for the year 2000, where no precise
mathematical expression exists for such predictions. However, employing
interpolation allows for approximate estimation of the population for any given
year. Numerous scholars have explored various interpolation techniques, including
central difference interpolation formulas.

Akima (1970) introduced a novel mathematical method involving fitting a smooth curve
to a set of numbers in a plane, focusing on local steepness and polynomial
representation. Other researchers like Atkinson (1989), Carl and Boor (1980) discussed
Newton's divided difference formula for unevenly spaced points, while Burden and
Faires (2001) and Suli and Mayers (2003) delved into the Lagrange formula for
interpolation. Abdulla et al. (2004) proposed an efficient formula for central difference
interpolation, demonstrating its effectiveness compared to existing methods.

Further studies explored various aspects of interpolation, including finite difference

methods, radial point interpolation, and void-filling interpolation. Muthumalai
(2008) investigated new iterative interpolation methods, while Singh and Bhadauria
(2009) proposed finite difference formulae for unequal sub-intervals. Interpolation
techniques have found applications in diverse fields such as lidar technology for
vegetation assessment.

More recent research has focused on refining interpolation methods, comparing different
central difference formulas and their applications. Scholars like Roseline et al. (2019)
have developed and analyzed central difference interpolation formulas, aiming to
provide clarity and utility for students and researchers alike. These studies contribute to
a better understanding of interpolation techniques and their practical implications,
offering valuable insights for further exploration in academic settings.

1
 Chapter Two

2- Some basic definitions:

In this section, some basic terms regarding interpolation have been discussed:

2.1 - Interpolation: (Muthumalai, , 2014)“Interpolation is the art of

reading between the lines of the table”, says Thiele.

It also refers to the addition or filling up of middle terms of the series. Hence,
Interpolation is thetechnique of approximating the function value for any
undetermined independent variable, whereas Extrapolation is the technique of
computing the function value outside the stated range.

2.2 - Premise of interpolation:

For interpolation to work, certain premises must be fulfilled:

1. During the period under examination, there are no significant changes in the values.

2. Values should increase and fall consistently. For example, if the given data
is on the number of fatalities in various years in a particular town, and some of
the observations are for years when the town was affected by war or epidemic,
interpolation approaches are ineffective.

3. In the methodology of finite differences, when a given set of values is expressed

in polynomial form, it is observed that if values of the function are given in a

clear and verified manner then one can easily compute ‘𝑌’ in accordance with ‘
𝑋 '.

And if the value of the function is not known, then one will introduce some simpler
function. For e.g. ∅(𝑥) such that both the functions have the same set of values. This
process is what someone called Interpolation.

Then after that, there comes a major term that has its importance in interpolation i.e.
Finite differences.

2
Finite difference function (:[singh A. , 2009)The math of finite differences gives
out the changes that occur in the value of the dependent variable by the finite
changes in the independent variable.

So, there are have three types of differences in the math of finite differences:

Let us consider a function x =g(y) and tabulate it for the uniformly spaced values

y=y0,y0+h,y0+2h… y0+nh

given. x=x0,x1,x2…xn

To find the values of f(y) and f’(y), for some middle values of x, the given three
differences are helpful:

Forward difference: The given set of values such as

x0-x1,x2-x1,x3-x2… xn-xn-1

When expressed by ∆x0, ∆x1, ∆x2,…,∆xn−1

respectively are known as first forward differences and the symbol∆ is known as the
forward difference operator.

Backward difference: The given set of values such as

x1 − x0, x2 − x1, x3 −x2 , … … … … … . xn – xn−1

when expressed by ∇x1, ∇x2, ∇x3 … ∇xn ,respectively, are known as first backward
differences and the symbol ∇is known as the backward difference operator.

Central difference: The relation that defines the central difference operator ‘d’ is
given as :

x1 − x0 =𝛿𝑋1 , x2 − x1 = 𝛿𝑋3 , …, xn – xn−1 = 𝛿𝑋𝑛−1

2 2 2

3
3- Study of different central difference Interpolation formulas
(kumar, 2018)

It is easy to calculate the value of y corresponding to any value of x if the value of

the function g(x) is known. On the other hand, if one is unfamiliar with the value of
g(x), then it is quite difficult to find the solution of g(x) by using a tabulated
collection of numbers (xi, yi).

In such circumstances, simply substitute a normal function ∅(x) for g(x) which
shows the same values as that of g(x) at the given tabulated collection of numbers.
As a result, one can find any other value from ∅(x) which is called as an
interpolating or smoothing function

The methodology of finite differences helps in understanding the concept of

Interpolation. By using forward and backward interpolation differences of a function
the two fundamental interpolation formulae can be obtained. These formulas are
frequently used in engineering and scientific research. Before discussing central
difference interpolation formulas let’s take a small look at the forward and backward
interpolation formulas.

Newton's forward interpolation formula:

Newton’s forward interpolation formula: let us consider a function x = g(y), take the
values 𝑥0 ,𝑥1 ,𝑥2 ,𝑥3 ,𝑥4 ...............𝑥n in accordance with the values 𝑦0 ,𝑦1 ,𝑦2 ,𝑦3
........𝑦n of y. Let these values of y to be stated in a spaced manner such that 𝑦j =𝑦0
+𝑗h(j =0

2........). assuming x(y) to be an 𝑛th degree polynomial in y such that x (𝑦0 ) = 𝑥0 ,

𝑥(𝑦n ) = 𝑥1 , .......... 𝑥(𝑦n ) = 𝑥n , we can write,

𝑝(𝑝−1) 𝑝(𝑝−1)(𝑝−2)
𝑥p =𝑥0+𝑝∆𝑥 + ∆2𝑥0 + ∆3𝑥0+……………..(1)
2! 3!

4
This method is beneficial for determining values of x at the beginning of a tabulated
set of values and extrapolating values of x a bit backward (i.e. to the left) of 𝑥0.

One more result is there: The linear interpolation is given by the first two terms of
this formula, while the parabolic interpolation is given by the first three terms, and
so on.

And

f ( x0 ) 1  2 f ( x0 )
pn ( x)  f ( x0 )  ( x  x0 )  ( x  x0 )( x  x1 )  ...
h 2! h2
1  n f ( x0 )
 ( x  x0 )( x  x1 )...( x  xn1 ) .
n! h n

x  x0
Form (5.25) let s  ,
h

then, we have x  xi  x  x0  ( xi  x0 )  sh  ih  (s  i )h , i=0,1, …, n.

With this, the Newton forward formula (5.25) becomes

s( s  1) 2 s( s  1)( s  2)...(s  n  1) n
pn ( x)  f ( x0 )  sf ( x0 )   f ( x0 )  ...   f ( x0 ),
2! n!

and the corresponding error (5.8) becomes

s( s  1)( s  2)...( s  n) n1 ( n1)

En ( x)  f ( x)  pn ( x)  h f (c), where a<c<b.
(n  1)!

Newton Backward Difference Interpolation Formula

let us consider a function x= g(y), take the values

𝑥0, 𝑥1, 𝑥2 ... ... ... ... ...in accordance with the values𝑦0, 𝑦0 + h,

𝑦0 + 2h.......... of y. Consider

5
the problem of determining the values of g(y) for y = 𝑦 + 𝑝h, where p may be any
real

number. Then we have:

𝑝(𝑝+1) 2 𝑝(𝑝+1)(𝑝+2) 3
𝑥p =𝑥n +𝑝∇𝑥n + ∇ 𝑥n+ ∇ 𝑥n+ ⋯ … … … (2)
2! 3!

It is termed Newton's backward interpolation formula as the given equations consist

of 𝑥𝑛 and the backward differences of 𝑥n.

This method is beneficial for determining values of x at the end of the tabulated set
of values and for extrapolating values of x a little ahead (to the right) of 𝑥n.

x  xn
Following similar steps as in subsection 5.4.1, with s  .
h
We can obtain the following Newton backward difference interpolation formula:
s ( s  1) 2 s( s  1)( s  2)...( s  n  1) n
pn ( x)  f ( xn )  s f ( xn )   f ( xn ) ...   f ( xn ) ,
2! n!

and the corresponding error:

s( s  1)( s  2)...( s  n) n1 ( n1)

En ( x)  f ( x)  pn ( x)  h f (c). where a<c<b.
(n  1)!

Central difference interpolation: (Uddin, 2019)

So, in Newton's backward and forward difference interpolation formula, it is learned

that they can be used for interpolation around the beginning and end of a table of
data. Now let’s introduce the central difference interpolation formula that is best
applicable for determining the values near the middle of the table. In the central
difference interpolation, there are four methods, that are required to formulate the
central difference:

Gauss Forward interpolation formula:

6
If y takes the values y0 − 2ℎ, y0 − ℎ, y0, y0 + ℎ, y0 + 2ℎ and corresponding values of
x = g(y) are x −2, x−1, x0, x1, x2, then the formula for Gauss forward interpolation is:

𝑝(𝑝−1) (𝑝+1)𝑝(𝑝−1) (𝑝+1)𝑝(𝑝−1)(𝑝−2)

𝑥𝑝 = 𝑥0 + 𝑝∆𝑥0 + ∆2 𝑥−1 + ∆3 𝑥−1 + ∆4 𝑥𝑥−2 +
2! 3! 4!

⋯ … … … …. (3)

As seen below, this formula uses odd differences slightly below the central line,and
even differences are taken on the central line:

𝐶𝑒𝑛𝑡𝑟𝑎𝑙 𝑙𝑖𝑛𝑒 𝑥
0 ∆⬚2 − 1
∆ 𝑥

∆𝑥0∆3𝑥−1 ∆5 𝑥−2

This method is beneficial to estimate values of x for p (0<p<1), calculated in a

forward direction from the origin.

When ‘p’ is between 0 and , this formula is applied.

Gauss backward interpolation:

Gauss backward interpolation formula is derived by modifying Newton’s forward

interpolation formula. It's beneficial for interpolating x values for a negative ‘p’
value that's between -1 and 0 or in terms of central difference notation the values lie
between -1/2 and 0.

The gauss backward formula is written as:

𝑝(𝑝+1) (𝑝+1)𝑝(𝑝−1)
𝑥𝑝 =𝑥0 +p∆𝑥−1 + ∆2 𝑥−1 + ∆3 𝑥−2
2! 3!
(𝑝+2)(𝑝+1)𝑝(𝑝−1)
+ ∆4 𝑥−2 … … … …. (4)
4!

As illustrated below, this approach has odd differences above the central line , and
even differences are taken on the central line:
∆𝑥−1 ∆3 𝑥−2 ∆5 𝑥−3

7
 2
Central line: 𝑥0 ∆2 𝑥−1 ∆4 𝑥−2

This is the rule formulation of gauss backward interpolation.

The Gauss forward and backward interpolation formulas aren't very useful. These,
on the other hand, might be considered as the first phase in attaining the crucial
formula for the next sections:

Stirling’s central difference interpolation formula:

After taking the arithmetic mean of Gauss forward and backward interpolation we
will obtain Stirling's Interpolation formula. This formula is used for an odd number
of equally spaced values.

So, aftertaking the mean of Gauss forward and backward interpolation we get
Stirling's formula as

∆𝑥0+ ∆𝑥−1 𝑝2
𝑥 = 𝑥0 + 𝑝 [ ] + ∆2 𝑥−1 ± 1 +
2 2!
𝑝(𝑝2 −1) ∆𝑥−1+∆𝑥−2 𝑝2 (𝑝2 −1)
[ ]+ ∆4 𝑥−2+⋯……………(5)
3! 2 4!

The means of odd differences just above and below the central line and even
differences onthe central line are used in this formula, as illustrated below:

∆𝑥−1 ∆3 𝑥−2

Central line: 𝑥0 ∆2𝑥−1 ∆4 𝑥−2

∆𝑥0 ∆3𝑥−1

By applying this format one can make up Stirling's formula an easy way to remind
the terms in the formula.

Stirling’s interpolation formula gives the best approximate result when ‘p’ lies

8
 −1 1
between and .
4 4

9
Bessel's Interpolation Formula

To derive BIF, since  y 0   y i for i=1, 2, … from (5.28), we get:

i i

s( s  1) 2 s( s  1)( s  2) 3 s( s  1)( s  2)...( s  n  1) n

pn ( x)  y0  s y 1   y1   y 3  ...   yn .
2
2! 3! 2
n ! 2

Using average operator, it yields:

1 1 1 1 1 1 1
y0  y0  y0  y1  y1  ( y1  y0 )  ( y1  y0 )   y 1   y 1 .
2 2 2 2 2 2 2
2 2
1 1 1 1
 2 y1   2 y1   2 y1   2 y0   2 y0
2 2 2 2

1 1 1
 ( 2 y1   2 y0 )  ( 2 y1   2 y0 )   2 y 1   3 y 1 .
2 2 2
2 2

 3 y 3   3 y 3   3 y 1   3 y 1   3 y 1  ( 3 y 3   3 y 1 )   3 y 1   4 y1
2 2 2 2 2 2 2 2

1 1 1 1
=  3 y 1   4 y1   4 y1   4 y0   4 y0
2
2 2 2 2

1 1
  3 y 1  ( 4 y1   4 y0 )  ( 4 y1   4 y0 )
2
2 2

1
=  3 y 1   4 y 1   5 y 1 .
2 2
2 2

and so on.

Substitute these values in (5.37) and add the coefficients, we get:

s( s  1) 2 s( s  1)( s  0.5) 3 s( s 2  1)( s  2) 4

pn ( x)   y 1  ( s  0.5) y 1   y 1   y1   y1  .
2 2
2! 2
3! 2
4! 2

10
The formula (5.38) is known as Bessel's interpolation formula.

1- Newton’s forward interpolation formula :

𝑝(𝑝 − 1) 2 𝑝(𝑝 − 1)(𝑝 − 2) 3

𝑥𝑝 = 𝑥0 + 𝑝∆𝑥0 + ∆ 𝑥0 + ∆ 𝑥0 + ⋯
2! 3!

2- Newton’s forward interpolation formula :

𝑝(𝑝 + 1) 2 𝑝(𝑝 + 1)(𝑝 + 2) 3
𝑥𝑝 = 𝑥𝑛 + 𝑝∇𝑥𝑛 + ∇ 𝑥𝑛 + ∇ 𝑥𝑛 + ⋯
2! 3!
3- Gauss forward interpolation formula :

𝑝(𝑝 − 1) 2 𝑝(𝑝 + 1)(𝑝 − 1) 3

𝑥𝑝 = 𝑥0 + 𝑝∆𝑥0 + ∆ 𝑥−1 + ∆ 𝑥−1
2! 3!
𝑝(𝑝 + 1)(𝑝 − 1)(𝑝 − 2) 4
+ ∆ 𝑥−1 + ⋯
4!

4- Gauss backward interpolation :

𝑝(𝑝 − 1) 2 𝑝(𝑝 + 1)(𝑝 − 1) 3
𝑥𝑝 = 𝑥0 + 𝑝∆𝑥0 + ∆ 𝑥−1 + ∆ 𝑥−1
2! 3!
𝑝(𝑝 + 1)(𝑝 − 1)(𝑝 − 2) 4
+ ∆ 𝑥−1 + ⋯
4!

5- Stirling’s central difference interpolation formula :

∆𝑥0 + ∆𝑥−1 𝑝2 2 𝑝(𝑝2 − 1) ∆𝑥−1 + ∆𝑥−2
𝑥𝑝 = 𝑥0 + 𝑝 [ ] + ∆ 𝑥−1 + [ ]
2 2! 3! 2
𝑝2 (𝑝2 − 1) 4
+ ∆ 𝑥−2 + ⋯
4!

6- Bessel’s central difference interpolation formula :

1
𝑝(𝑝 − 1) ∆2 𝑥0 + ∆2 𝑥−1 (𝑝 − 2) 𝑝(𝑝 − 1)
𝑥𝑝 = 𝑥0 + 𝑝∆𝑥0 + [ ]+ ∆3 𝑥−1
2! 2 3!

11
 Chapter Three

3. Numerical discussion (1Ndu, 2019)

3.1. Consider a function x = 4𝑦 2 + 𝑦 + 1, and value of x for equidistant spaced

values of y are:

X 1 3 5 7 9 11 13
Y 6 40 106 204 334 496 690

Solution: let us solve this problem with all the four central difference
interpolation formulae:
Here let us find this for y=6, for that, we have 𝑦0

y0=7 and h=2 and p=

y x ∆x ∆2x ∆3x
1 6
34
3 40 32
66 0
5 106 32
98 0
7 204 32
130 0
9 334 32
162 0
11 496 32
194
13 690

12
1. Gauss forward interpolation formula:

(−0.5)(−0.5−1)
(6) = 204 + (−0.5)(130) + (32)......... [ using(3)]
2

= 204 − 65 + (−0.5)(−1.5)(16)
= 204 −65 + 12 = 151
Hence we get x(6) = 151 by gauss forward interpolation formula.
using (3)]
2. Gauss backward interpolation formula:
(−0.5)(−0.5+1)
(6) = 204 + (−0.5)(98) + (3) ... ... ... .... [using (4)]
2

= 204 − 49 + (−0.5)(0.5)(16)
= 204 − 49 − 4
=151
3. Stirling’s formula:
98+130 (−0.5)2
(6) = 204 + (−0.5) [ ]+ (32).............. [using (5)]
2 2!

= 204 + (−0.5)(114) + 0.25(16)

= 204 − 57 + 4
=151
4. Bessel’s formula:
(−0.5)(−0.5−1) 32+32
(6) = 204 + (−0.5)(130) + [ ] ... ... ... [using(6)]
2 2

=204- 65+(0.375)(32)
=151
Hence, by all the methods we get the same result as x (6) =151

13
3.2. The population of a town in a year is: (Das, 2016)

Year 1931 1941 1951 1961 1971

Population(in thousand) 15 20 27 39 52

Sol: let us solve this problem with all the four central difference interpolation
formulae: Here let us find this for the year=1946, for that, we have 𝑥0 = 1951 and
h=2 and

𝑥−𝑥0 1946−1951
p= = =-0.5
ℎ 2

Year Population(in ∆𝑥 ∆2 𝑥 ∆3 𝑥 ∆4 𝑥
thousand)
1931 15
5
1941 20 2
7 3
1951 27 5 -7
12 -4
1961 39 1
13
1971 52

1-Gauss forward interpolation:

(−0.5)(−1.5)
(1946) = 27 + (−0.5)12 + (5) ... ... ... ... . . [using(3)]
2

=27-6+1.875

=22.875 thousand

2- Gauss backward interpolation:

(−0.5)(−0.5 + 1)
(1946) = 27 + (−0.5)7 + (5) ... ... ... . . [using(4)]
2

14
= 27-3.5-0.625

= 22.875

=22875 thousand

3. Stirling’s formula:

7+12 (0.5)2
(1946)=27+(−0.5)[ ]+ (5).............[using(5)]
2 2

= 27-4.75+0.625

= 22.875

= 22875 thousand

4. Bessel’s formula:

(−0.5)(−0.5 − 1) 5 + 1
(1946) = 27 + (−0.5)12 + [ ] ... ... ... ... [using(6)]
2 2

= 27-6+1.875

=22.875

=22875 thousand

Hence, by all the methods we get the same result as x (1946) =22875 thousand

5.Alternatives for interpolation formulas:

So far, we have come to know about the several interpolation formulae

likeNewton’sforward, Newton’s backward, Gauss’s backward and Gauss’s forward,
Stirling’s formula, Bessel’s formula for calculating 𝑥𝑝for equally spaced tabulated
values.Now we will do the comparative study of all of them that which formula or
method is best applicable at which stage:The central difference formula has smaller
coefficients that converge faster than Newton's formula. The coefficients in Stirling's
formula fall more rapidly than those in Bessel's formula after a few terms, while the
15
coefficients in Bessel's formula decline more rapidly than those in Newton's formula
after a few terms.As a result,wherever possible, we will prefer the central difference
formula instead of Newton's formula.The correct interpolation method, on the other
hand, is determined by the position of the interpolated values in the given data. The
following rules will assist you in comprehending all of the strategies.:

1. For calculating the tabulated set of values at the starting of the table, use
Newton’s forward interpolation formula.

2. For calculatingvalue nearby the end of the table,prefer Newton’s backward

interpolation formula.

3. For calculating value close to the center of the table, prefer either Stirling’s
formula or Bessel’s formula.

−1 1
• If someone wants to interpolate a value of 'p' that is between 𝑎𝑛𝑑 , prefer
4 4

Stirling’s formula.

• Use Bessel’s formula, if someone wants to interpolate for a value of 'p' between

1 3
𝑎𝑛𝑑 .
4 4

• By reminding all these rules one can easily make the right choice that in which
problem which method is best applicable.

6. Applications of Interpolation in Real life:

As we know, the method of finding values at unknown points using known values or
sample points is known as interpolation. So, various practical uses of interpolation
as follows:

1. The concept of interpolation can be used to estimate unknown parameters for any
geographic point data, such as rainfall,elevation,noise levels, chemical
concentrations, and so on.

16
2. Using interpolation techniques to zoom digital images: Image processing for low-
resolution digital images is a very challenging problem nowadays. It is because of
errors that occur in quantization and sampling. Zooming in on such images is
extremely difficult. As a result, when zooming in low-resolution images, we use
interpolation functions.

3. Interpolation is often used outside the domain of mathematics to scale images and
transform the sample rate of digital signals.

4. Interpolation is helpful whenever you have to scale things up or down regularly.

For example- We may know how much food costs for a 10person event, 50 person
event, or 100 person event, but we need a precise estimate of how much catering
will cost for 25 people or 75 people. That's what interpolation allows us to achieve.

These are some of the examples of interpolation in real life as well as in other
disciplines.

7. Observation of the Study:Based on the observations and readings, it is

concluded from this paper that Interpolation is mostly used to assist users, who may
be scientists, photographers, engineers, or mathematicians, in determining what
information may exist outside of their collected information. The comparative study
of all the methods helps us to understand that how one can solve various problems
by using the best suitable interpolation method. Or how to make the right choice that
which formula gives the result of the given problem more accurately.

17
EX 1: Determine the best method among N.F.I.F , N.B.I.F and (contral.I.f for
𝜋
estimating (0.1) on the following table , whare f(x)= cos2( 𝑥)
2

Sol 1:
x y 𝛥 ∆2 ∆3 ∆4
Note that , since ∆𝑦𝑖=∇𝑦i+1
- 1 -1 =𝛿𝑦𝑖+1
2
2
Then for above interpolations ,
we have the some table
- 0 1 2 -4
Xm0.1 1

0 1 -1 -2 4 8

1 0 1 2

0⋅1−0
2 1 Xm=0.1 , x0=0 , h=1 ⇒ m= =0.1
1

0.1(0⋅1−1)
I. Forward : f(0.1) ≅ 1+(0.1)(-1) + (2)=0.81
2!
0.1(0⋅1+1)
II. Backward: f(0.1) ≅ 1+(0.1)(1) + (2)=1.21
2!
1 0.1(0.1+1) 0.1(0.1−1)(0.1−0.5)
III. Central : f(0.1) ≅ + (0.1-0.5)(-1)+ (0)+ (4)=0.9194
2 2! 3!

𝜋
Since f(0.1) = cos2 ( (0 ⋅ 1)) ≅0.975528 , then
2

E1= 0.975528 -0.81= 0.,165528 (Forward)

E2= 0.975528 – 1.21 = 0.234472 (Backward)

E3=0.975528 – 0.9194 = 0.056128 (Central)

Thus, the central method is the best to interpolate f(0.1)

EX 2: Determine the best method among N.F.I.F , N.B.I.F and (Contral.i.f for
estimating F(4.5) on the following table , whare f(x)= 3X2+2X+1

18
Sol 2 :
X Y ∆ ∆2

2 17 17

3 34 23 6
4.5
4 57 29 6

5 86 35 6

6 121 41 6

7 162 47 6 4.5−6
XM=4.5 , XO=5 , h=1 ⇒ m= =-0.5
1
8 209
I. Forward: F(4.5) ≅ 86+(-
0.5) (35) + (-0.5-1)3=70.75
II. Backward: F(4.5) ≅86+(-0.5) (29) + (-0.5-1)3=70.75
86−121 (−0⋅5)(−0.5−1) 6+6
III. Central: F(4.5) ≅ ( ) + (-0.5-0.5)+ ( )= 70.75
2 2! 3

since f(4.5) = 3(4.5)2 + 2(4.5)+1=1241

E1=E2=E3 = 1241-70.75= 1170.25.

Thus , All above methods ore the seme to interpolate f(4.5)

19
EX3: Determine the best method among N.F.I.F , N.B.I.F and (Bessel.i.f for
estimating F(6), on the following data , whare f(x)= 4X2+X+1

SOL3:

X Y ∆ ∆2 ∆3

Xm=6 1 6 34

3 40 66 32 0

5 106 98 32 0

7 204 130 32 0

9 334 162 32 0

11 496 194 32

13 690

6−7
XM=6 , XO=7 , h=2 ⇒ m= =-0.5
2

(−0⋅5)(−0⋅5+1)
I. Forward : f(6) ≅ 204+(-0.5)(130)+ (32)=151
2
(−0⋅5)(−0⋅5+1)
II. Backward: f(6) ≅204+(-0.5)(98)+ (32)=151
2
(−0⋅5)(−0⋅5+1) 32+32
III. Central : f(6) ≅204+(-0.5)(130)+ ( ) =151
2 2

Since F(6)=4(62)+(6)+1=0

E1=E2=E3=151-151=0

Thus , All above methods ore the seme to estimate F(6)

And F(6)= 151 is an exact solution.

20
EX4: Determine which one of these methods among newton, backwand forward and
central is the best for estimating f(3.1) where f(x)=x2.on [-3,3]

Sol4:
X Y ∆ ∆2 ∆3

-3 9 -5

-2 4 -3 2 0

-1 1 -1 2 0

0 0 1 2 0
Xm=3.1
1 1 3 2 0

2 4 5 2
3.1−3
XM=3.1 XO=3 , h=1 ⇒ m= =0.1
1
3 9
I. Forward is failing here, because
we can’t determine Y0.
(0.1)(0⋅1+1)
II. Backward: f(3.1) ≅ 9+(0.1)(5) + (2)=9.61
2!

III. Central is failing here, because we can’t determine Y1 to calculate 𝜇𝑦1 .

2

So thet , backward is the best method for (3.1)

21
EX 5: Determine the best method among N.F.I.F , N.B.I.F and (Contral.i.f for
interpolating F(1.45), whure F(x)= [x] is 0 floor function.

Sol 5:
x Y ∆ ∆2

-2.5 -3 1

-1.5 -2 1 0
Xm=1.45
-0.5 -1 1 0

0.5 0 1 0

1.5 1 1 0

2.5 2 1.45−1.5
XM=1.45, XO=1.5 , h=1 ⇒ m= =-0.05
1

I. Forward : f(1.45) ≅ 1+(-0.05)(-1) = 0.95

II. Backward: f(1.45) ≅ 1+(-0.05)(1) =0.95
1+2
III. Central : f(1.45) ≅ ( ) + (-0.05-0.5)(1)=0.95
2

Since f(1.45)= [1.45]=1 , them

E1=E2=E3=1-0.95=0.05

Thus , All the methods ore the seme to interpolate f(1.45).

22
 Conclusion

In conclusion, the field of interpolation offers a rich tapestry of methods and

approaches that researchers have explored over the years. From classical techniques
like Lagrange's and Newton's interpolation to innovative variations such as the
central difference interpolation proposed by Mohiuddin et al. (2019), the spectrum
of tools available for estimating values between data points is diverse and
continually evolving. Studies by various scholars have delved into the nuances of
different interpolation methods, shedding light on their strengths, limitations, and
applications in diverse domains.

The comparative analyses presented in these research efforts have highlighted the
efficiency, accuracy, and versatility of newer interpolation formulas in contrast to
traditional approaches. Singh and Bhandari (2009) introduced finite difference
formulae for unequal sub-intervals, while Batter et al. (2009) explored
interpolation's utility in lidar technology applications. Garner and Go (2013),
Muthumala and Uthra (2014), Srivastav et al. (2015), Das and Chakrabarty (2016)
have contributed unique perspectives on interpolation methods tailored to specific
real-world challenges.The ongoing pursuit of refining interpolation techniques is
evident in the research endeavors documented in the introduction. Researchers have
strived to enhance accuracy, efficiency, and adaptability in interpolation
methodologies to cater to the complexities of modern computational and analytical
tasks. The comparative evaluations undertaken have revealed promising
advancements and novel insights that pave the way for a deeper understanding and
utilization of interpolation in diverse scientific and practical contexts.

23
 Reference

Pal, M., 2007. Numerical analysis for scientists and engineers: theory and C
programs. Alpha Science International, Limited.

Poonia, A.C.R., Comparative Studyof Different Central Difference Interpolation

Formulas.

1. Kumar, R. and Kumar, K., 2018. COMPUTER BASED NUMERICAL AND

STATISTICAL TECHNIQUES. CBS Publishers & Distributors Private Limited.

2. Akima, H., 1970. A new method of interpolation and smooth curve fitting based
on local procedures. Journal of the ACM (JACM), 17(4), pp.589-602.

3. Bater, C.W. and Coops, N.C., 2009. Evaluating error associated with lidar-
derived DEM interpolation. Computers & Geosciences, 35(2), pp.289-300.

4. Mardikis, M.G., Kalivas, D.P. and Kollias, V.J., 2005. Comparison of

interpolation methods for the prediction of reference evapotranspiration—an
application in Greece. Water Resources Management, 19, pp.251-278.

5. Süli, E. and Mayers, D.F., 2003. An introduction to numerical analysis.

Cambridge university press.

6. Abdulla, F., 2014. A New (Proposed) Formula for Interpolation and Comparison
with Existing Formula of Interpolation. Math. Theory Modell, 4(4), pp.33-48.

7. Liu, G.R., Zhang, J., Li, H., Lam, K.Y. and Kee, B.B., 2006. Radial point
interpolation based finite difference method for mechanics
problems. International Journal for Numerical Methods in Engineering, 68(7),
pp.728-754.

8. Reuter, H.I., Nelson, A. and Jarvis, A., 2007. An evaluation of void‐filling

interpolation methods for SRTM data. International Journal of Geographical
Information Science, 21(9), pp.983-1008.

24
9. Singh, A.K. and Bhadauria, B.S., 2009. Finite difference formulae for unequal
sub-intervals using Lagrange’s interpolation formula. Int. J. Math. Anal, 3(17),
p.815.

10.Householder, A.S., 2006. Principles of numerical analysis. Courier Corporation

11.Garnero, G. and Godone, D., 2014. Comparisons between different interpolation

techniques. The International Archives of the Photogrammetry, Remote Sensing
and Spatial Information Sciences, 40, pp.139-144.

12.Poonia, A.C.R., Comparative Studyof Different Central Difference Interpolation

Formulas.

13.Das, B. and Chakrabarty, D., 2016. Lagrange’s interpolation formula:

representation of numerical data by a polynomial curve. International Journal of
Mathematics Trends and Technology-IJMTT, 34.

14.Das, B. and Chakrabarty, D., 2016. Newton’s divided difference interpolation

formula: Representation of numerical data by a polynomial curve. International
Journal of Mathematics Trends and Technology-IJMTT, 35.

15.Roseline 1Ndu, Kingdom 1Nwuju, Wilkox K. (2019) “A Comparative Study of

Interpolation using the concept of Mathematical Norm with a Proposed Model”
International Journal of Scientific and Research Publications, Volume 9, Issue 4,
ISSN 2250-3153.

16.Uddin, M.J., Kowsher, M. and Moheuddin, M.M., 2019. A new method of

central difference interpolation. Applied Mathematics and Sciences: An
International Journal (MathSJ).

25
 ‫پوختە‬

‫کە سنووردارکردن و یاسا ڕاستەقینەکانی کارپێکراو دەگەیەنن‪ .‬ئێستا پێویستە مامەڵە لەگەڵ ئەم مەرجانەدا‬
‫بکرێت و بڕیارێک خۆی بناسێنێت‪ .‬یەکێک لە ڕێگاکان ئەوەیە کە بەردەوام بیت لە بەکارهێنانی ستراتیژییە‬
‫ئاساییەکان بۆ بیرکاری‪ ،‬وەرگرتنی وەاڵم وەک ڕێچکەیەک‪ ،‬یان کۆمەڵێک فۆرمول‪ .‬ستراتیژییەکی دیکە‬
‫بریتییە لە گەیاندنی مەرجەکان بۆ ئەوەی بە شێوەیەکی حیساباتی چارەسەر بکرێن‪ ،‬واتە بە بەکارهێنانی‬
‫ستراتیژییەکان بۆ تاقیکردنەوەی بیرکاری یان شیکاری ژمارەیی‪ .‬لەم توێژینەوەیەدا ئامانجمان باسکردنی‬
‫ئینتەرپۆالسیۆنە‪ ،‬شێوازە جۆراوجۆرەکان بۆ چارەسەرکردنی ئینتەرپۆالسیۆنی جیاوازی ناوەندی‪،‬‬
‫گشتاندنەکانیان‪ .‬هەروەها باس لە بەکارهێنانی ئینتەرپۆالسیۆن دەکرێت و مرۆڤ دەتوانێت بە ئاسانی لە‬
‫چەمکەکانی توێژینەوەکە تێبگات‪.‬‬

‫‪a‬‬