S19AQF (PDF version)
S Chapter Contents
S Chapter Introduction
NAG Library Manual

NAG Library Routine Document

S19AQF

Note:  before using this routine, please read the Users' Note for your implementation to check the interpretation of bold italicised terms and other implementation-dependent details.

 Contents

    1  Purpose
    7  Accuracy

1  Purpose

S19AQF returns an array of values for the Kelvin function kerx.

2  Specification

SUBROUTINE S19AQF ( N, X, F, IVALID, IFAIL)
INTEGER  N, IVALID(N), IFAIL
REAL (KIND=nag_wp)  X(N), F(N)

3  Description

S19AQF evaluates an approximation to the Kelvin function kerxi for an array of arguments xi, for i=1,2,,n.
Note:  for x<0 the function is undefined and at x=0 it is infinite so we need only consider x>0.
The routine is based on several Chebyshev expansions:
For 0<x1,
kerx=-ftlogx+π16x2gt+yt  
where ft, gt and yt are expansions in the variable t=2x4-1.
For 1<x3,
kerx=exp-1116x qt  
where qt is an expansion in the variable t=x-2.
For x>3,
kerx=π 2x e-x/2 1+1xct cosβ-1xdtsinβ  
where β= x2+ π8 , and ct and dt are expansions in the variable t= 6x-1.
When x is sufficiently close to zero, the result is computed as
kerx=-γ-logx2+π-38x2 x216  
and when x is even closer to zero, simply as kerx=-γ-log x2 .
For large x, kerx is asymptotically given by π 2x e-x/2 and this becomes so small that it cannot be computed without underflow and the routine fails.

4  References

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

5  Parameters

1:     N – INTEGERInput
On entry: n, the number of points.
Constraint: N0.
2:     XN – REAL (KIND=nag_wp) arrayInput
On entry: the argument xi of the function, for i=1,2,,N.
Constraint: Xi>0.0, for i=1,2,,N.
3:     FN – REAL (KIND=nag_wp) arrayOutput
On exit: kerxi, the function values.
4:     IVALIDN – INTEGER arrayOutput
On exit: IVALIDi contains the error code for xi, for i=1,2,,N.
IVALIDi=0
No error.
IVALIDi=1
xi is too large, the result underflows. Fi contains zero. The threshold value is the same as for IFAIL=1 in S19ACF, as defined in the Users' Note for your implementation.
IVALIDi=2
xi0.0, the function is undefined. Fi contains 0.0.
5:     IFAIL – INTEGERInput/Output
On entry: IFAIL must be set to 0, -1​ or ​1. If you are unfamiliar with this parameter you should refer to Section 3.3 in the Essential Introduction for details.
For environments where it might be inappropriate to halt program execution when an error is detected, the value -1​ or ​1 is recommended. If the output of error messages is undesirable, then the value 1 is recommended. Otherwise, if you are not familiar with this parameter, the recommended value is 0. When the value -1​ or ​1 is used it is essential to test the value of IFAIL on exit.
On exit: IFAIL=0 unless the routine detects an error or a warning has been flagged (see Section 6).

6  Error Indicators and Warnings

If on entry IFAIL=0 or -1, explanatory error messages are output on the current error message unit (as defined by X04AAF).
Errors or warnings detected by the routine:
IFAIL=1
On entry, at least one value of X was invalid.
Check IVALID for more information.
IFAIL=2
On entry, N=value.
Constraint: N0.
IFAIL=-99
An unexpected error has been triggered by this routine. Please contact NAG.
See Section 3.8 in the Essential Introduction for further information.
IFAIL=-399
Your licence key may have expired or may not have been installed correctly.
See Section 3.7 in the Essential Introduction for further information.
IFAIL=-999
Dynamic memory allocation failed.
See Section 3.6 in the Essential Introduction for further information.

7  Accuracy

Let E be the absolute error in the result, ε be the relative error in the result and δ be the relative error in the argument. If δ is somewhat larger than the machine precision, then we have:
E x2 ker1x+ kei1x δ,  
ε x2 ker1x + kei1x kerx δ.  
For very small x, the relative error amplification factor is approximately given by 1logx , which implies a strong attenuation of relative error. However, ε in general cannot be less than the machine precision.
For small x, errors are damped by the function and hence are limited by the machine precision.
For medium and large x, the error behaviour, like the function itself, is oscillatory, and hence only the absolute accuracy for the function can be maintained. For this range of x, the amplitude of the absolute error decays like πx2e-x/2 which implies a strong attenuation of error. Eventually, kerx, which asymptotically behaves like π2x e-x/2, becomes so small that it cannot be calculated without causing underflow, and the routine returns zero. Note that for large x the errors are dominated by those of the standard function exp.

8  Parallelism and Performance

Not applicable.

9  Further Comments

Underflow may occur for a few values of x close to the zeros of kerx, below the limit which causes a failure with IFAIL=1.

10  Example

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

10.1  Program Text

Program Text (s19aqfe.f90)

10.2  Program Data

Program Data (s19aqfe.d)

10.3  Program Results

Program Results (s19aqfe.r)


S19AQF (PDF version)
S Chapter Contents
S Chapter Introduction
NAG Library Manual

© The Numerical Algorithms Group Ltd, Oxford, UK. 2015