nag_kelvin_bei (s19abc) (PDF version)
s Chapter Contents
s Chapter Introduction
NAG Library Manual

NAG Library Function Document

nag_kelvin_bei (s19abc)

+ Contents

    1  Purpose
    7  Accuracy

1  Purpose

nag_kelvin_bei (s19abc) returns a value for the Kelvin function beix.

2  Specification

#include <nag.h>
#include <nags.h>
double  nag_kelvin_bei (double x, NagError *fail)

3  Description

nag_kelvin_bei (s19abc) evaluates an approximation to the Kelvin function beix.
Note:  bei-x=beix, so the approximation need only consider x0.0.
The function is based on several Chebyshev expansions:
For 0x5,
beix = x24 r=0 ar Tr t ,   with ​ t=2 x5 4 - 1 ;
For x>5,
beix=ex/22πx 1+1xat sinα-1xbtcosα
+ex/22π x 1+1xct cosβ-1xdtsinβ
where α= x2- π8 , β= x2+ π8 ,
and at, bt, ct, and dt are expansions in the variable t= 10x-1.
When x is sufficiently close to zero, the result is computed as beix= x24 . If this result would underflow, the result returned is beix=0.0.
For large x, there is a danger of the result being totally inaccurate, as the error amplification factor grows in an essentially exponential manner; therefore the function must fail.

4  References

Abramowitz M and Stegun I A (1972) Handbook of Mathematical Functions (3rd Edition) Dover Publications

5  Arguments

1:     xdoubleInput
On entry: the argument x of the function.
2:     failNagError *Input/Output
The NAG error argument (see Section 3.6 in the Essential Introduction).

6  Error Indicators and Warnings

NE_INTERNAL_ERROR
An internal error has occurred in this function. Check the function call and any array sizes. If the call is correct then please contact NAG for assistance.
NE_REAL_ARG_GT
On entry, x=value.
Constraint: xvalue.
x is too large for an accurate result to be returned and the function returns zero.

7  Accuracy

Since the function is oscillatory, the absolute error rather than the relative error is important. Let E be the absolute error in the function, and δ be the relative error in the argument. If δ is somewhat larger than the machine precision, then we have:
E x2 - ber1x+ bei1x δ
(provided E is within machine bounds).
For small x the error amplification is insignificant and thus the absolute error is effectively bounded by the machine precision.
For medium and large x, the error behaviour is oscillatory and its amplitude grows like x2π ex/2. Therefore it is impossible to calculate the functions with any accuracy when xex/2> 2πδ . Note that this value of x is much smaller than the minimum value of x for which the function overflows.

8  Parallelism and Performance

Not applicable.

9  Further Comments

None.

10  Example

This example reads values of the argument x from a file, evaluates the function at each value of x and prints the results.

10.1  Program Text

Program Text (s19abce.c)

10.2  Program Data

Program Data (s19abce.d)

10.3  Program Results

Program Results (s19abce.r)


nag_kelvin_bei (s19abc) (PDF version)
s Chapter Contents
s Chapter Introduction
NAG Library Manual

© The Numerical Algorithms Group Ltd, Oxford, UK. 2014