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

net/publication/344558001

Study of Numerical Accuracy in Different Spline Interpolation Techniques

Article in Global Journal of Pure and Applied Mathematics · October 2020

CITATIONS READS

6 1,471

2 authors:

Najmuddin Ahmad Farah Deeba

Integral University Integral University
50 PUBLICATIONS 149 CITATIONS 7 PUBLICATIONS 19 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Najmuddin Ahmad on 09 October 2020.

The user has requested enhancement of the downloaded file.

Global Journal of Pure and Applied Mathematics.
ISSN 0973-1768 Volume 16, Number 5 (2020), pp. 687-693
c Research India Publications
http://www.ripublication.com/gjpam.htm

Study of Numerical Accuracy in Different Spline

Interpolation Techniques

Najmuddin Ahmad, Khan Farah Deeba

Department of Mathematics, Integral University, Kursi Road, Lucknow (India).
Corresponding author’s E-mail ID: najmuddinahmad33@gmail.com

Abstract

In this paper, we introduced the accuracy of results using different spline

functions. Some examples are also discussed. This work presents a
numerical accuracy of results by different spline functions. Results are
analyzed by comparing the actual sampled values with the values obtained
by linear, quadratic and cubic spline interpolation.
Keywords: Interpolation, Linear Spline, Quadratic Spline, Cubic Spline,
Trigonometric function.

INTRODUCTION:

Numerical analysis is the area of mathematics and computer sciences that

creates analyzes and implements algorithms for solving numerically the problems
of continuous mathematics.[1] In the mathematical field of numerical analysis,
interpolation is a method of constructing new data points within the range of a discrete
set of known data points. It is often required to interpolate, i.e., estimate the value of that
function for an intermediate value of the independent variable. We have n + 1 distinct
nodal point x0 , x1 , . . . , xn and we want to determine an interpolating polynomial.
The interpolating polynomial is linear in each subinterval (xi−1 , xi ) and it agrees
with the function f(x) at the n+1 nodal points. The subintervals or the line segment
are called finite elements in one space dimension and the nodal points are called
knots. [2] A low-order polynomial approximation in each subinterval provides a better
688 Najmuddin Ahmad, Khan Farah Deeba

approximation to the tabulated function than fitting a single high-order polynomial to

the entire interval. Cubic spline interpolation is a special case for spline interpolation
that is used very often to avoid the problem of Runge’s phenomenon. This method
gives an interpolating polynomial that is smoother and has smaller error than some
other interpolating polynomials such as Lagrange polynomial and Newton polynomial.

ANALYSIS AND IMPLEMENTATION:

LINEAR SPLINE INTERPOLATION:

Let the given data points be

(xi , yi ) i = 0, 1, 2, ...n

where,
a = x0 < x1 < x2 < ... < xn = b
and let,
hi = (xi − xi−1 ) i = 1, 2, 3, ..., n
Further let si (x) be the spline of degree one defined in the interval [xi−1 , xi ]. Obviously,
si (x) represents a straight line joining the points (xi−1 , yi−1 ) and (xi , yi ). Hence, we
write
si (x) = yi−1 + mi (x − xi−1 )
Where,
(yi − yi−1 )
mi = .
(xi − xi−1 )

QUADRATIC SPLINE INTERPOLATION:

Let the given data points be

(xi , yi ), i = 0, 1, 2, ..., n

Where
a = x0 < x1 < x2 < ... < xn = b
and let
hi = xi − x( i − 1), ı = 1, 2, ..., n
0
Let si (x) and si (x) be continuous in [x0 , xn ] and let

si xi = yi , i = 0, 1, 2, ..., n.
Study of Numerical Accuracy in Different Spline Interpolation Techniques 689

Since si (x) is a quadratic in the interval [xi−1 , xi ].

1 (xi − x)2 (x − xi−1 )2 hi

si (x) = [− mi−1 + mi ] + yi−1 + mi−1 .
hi 2 2 2

CUBIC SPLINE INTERPOLATION:

Let the given data points be

(xi , yi ), i = 0, 1, 2, ..., n

Where
a = x0 < x1 < x2 < ... < xn = b
and let
hi = xi − xi−1 , i = 1, 2, ..., n
Let si (x) be the cubic spline defined in the interval [xi−1 , xi ]. The conditions for the
natural cubic spline are,

1. si (x) is atmost a cubic in each subinterval [xi−1 , xi ], i = 1, 2, 3, ..., n

2. si xi = yi , i = 0, 1, 2, ..., n
0 00
3. si (x) and si (x) and si (x) are continuous in [x0 , xn ] and
00 00
4. si (x0 ) = si (xn ) = 0.

1 (xi − x)3 (x − xi−1 )3

si (x) = [ Mi−1 + Mi
hi 6 6
(h2i ) (h2 )
+(yi−1 − Mi−1 )(xi − x) + (yi − i Mi )(x − xi−1 )]
6 6

NUMERICAL EXAMPLES:

EXAMPLE:1

x 1 2 4 8
f(x) 3 7 21 73

Table 1: Example:1
690 Najmuddin Ahmad, Khan Farah Deeba

In the interval [1, 2], we have

(x − 2)
L1 (x) = (3) + (x − 1)(7) = 4x − 1
(−1)

In the interval [2, 4], we have

(x − 4) (x − 2)
L2 (x) = (7) + (21) = 7x − 7
(−2) 2

In the interval [4, 8], we have

(x − 8) ((x − 4)
L3 (x) = (21) + (73) = 13x − 31
(−4) 4

Hence the linear interpolating polynomials are:

L1 (x) = 4x–1. 1 ≤ x ≤ 2

L2 (x) = 7x–7. 2 ≤ x ≤ 4
L3 (x) = 13x–31. 4 ≤ x ≤ 8

EXAMPLE 2:

x 1 2 3
f(x) -8 -1 18

Table 2: Example:2

L1 (x) = 7x–15. Ininterval[1, 2]

L2 (x) = 19x–39. Ininterval[2, 3]

EXAMPLE:3

x 1 2 3
f(x) -1 -8 18

Table 3: Example:3

Q(x) = 12x2 –41x + 33

.
Study of Numerical Accuracy in Different Spline Interpolation Techniques 691

EXAMPLE:4

x -3 -2 -1 1 3 6 7
f(x) 369 222 171 165 207 990 1779

Table 4: Example:4

Q1 (x) = 48x2 + 93x+

Q2 (x) = 6x2 –3x + 162.

Q3 (x) = 132x2 –927x + 1800.

EXMPLE:5

x 0 π/2 π
f(x) 0 1 0

Table 5: Example:5

In the interval [0, π/2], the natural cubic spline is given by

c1 (x) = 2/π[−(2x3 )/π 2 + 3x/2]

y(π/6) = c1 (π/6) = 2/π(−π/108 + π/4) = 0.4815

In the interval [0, π/4], the natural cubic spline is given by

c1 (x) = 4/π(−0.1240x3 + 0.7836x)

y(π/6) = c1 (π/6) = 0.4998

COMPARISON OF RESULTS:

From above examples we see that when we half the interval then cubic spline has
produced a better approximation.The cubic spline values together with the exact values
are given in the following table.
y = sin x
692 Najmuddin Ahmad, Khan Farah Deeba

X(in degree ) Cubic spline values Exact values

5 0.087155743 0.087155530
15 0.258819045 0.258818415
25 0.422618262 0.422617233
35 0.573576436 0.573575040
45 0.707106781 0.707105059

Table 6: Example:6

CONCLUSION:

In this paper, accordingly to the analysis the performance of function is presented.

Experimental result shows that the cubic spline has produced a better approximation
when the interval is halved. We finally consider values of y = sin x in intervals of
100 from x = 0 to π and then interpolate for x = 50, 150, 250, 350 and 450, using the
natural cubic spline.

ACKNOWLEDGEMENT :

Manuscript communication number (MCN) : IU/R&D/2020 − M CN 00845office of

research and development integral university , Lucknow.

REFERENCES

[1] Najmuddin, A., Singh, V.P., Journal of Mathematical and Computational Science,
6(5), 730, 2016.
[2] Jain, M.K., Iyenger, S.R.K., Numerical methods for Scientific and Engineering
Computation, New Age International Publishers, 2015
[3] Sastry, S.S., Introductory Methods of Numerical Analysis, PHI Learning Private
Limited, 2012.
[4] Sastry, S.S., Introductory Methods of Numerical Analysis, PHI Learning Private
Limited, 2012.
[5] Jain, M.K., Iyenger, S.R.K., Numerical methods for Scientific and Engineering
Computation, New Age International Publishers, 2015
[6] Chapra, S.C., Applied Numerical Methods with MATLAB, Tata McGraw Hill
Education Private Limited, 2010
[7] Sastry, S.S., Introductory Methods of Numerical Analysis, PHI Learning Private
Limited, 2012.
 Study of Numerical Accuracy in Different Spline Interpolation Techniques 693

[8] Agrawal, U., Computer based Numerical and Statistical Techniques, Dhanpat Rai,
New Delhi, 2015.
[9] Shao, Y., Matric Approach: the relationship among Lagrange, Monomial and
Newton’s interpolation, Report - University of St. Thomas.
[10] Hussien, K.A., International Journal of Computer Science and Network Security,
11(3), 255, 2011.

View publication stats