NAG Library Routine Document

s17dlf (hankel_complex)


s17dlf returns a sequence of values for the Hankel functions Hν+n 1 z or Hν+n 2 z for complex z, non-negative ν and n=0,1,,N-1, with an option for exponential scaling.


Fortran Interface
Subroutine s17dlf ( m, fnu, z, n, scal, cy, nz, ifail)
Integer, Intent (In):: m, n
Integer, Intent (Inout):: ifail
Integer, Intent (Out):: nz
Real (Kind=nag_wp), Intent (In):: fnu
Complex (Kind=nag_wp), Intent (In):: z
Complex (Kind=nag_wp), Intent (Out):: cy(n)
Character (1), Intent (In):: scal
C Header Interface
#include <nagmk26.h>
void  s17dlf_ (const Integer *m, const double *fnu, const Complex *z, const Integer *n, const char *scal, Complex cy[], Integer *nz, Integer *ifail, const Charlen length_scal)


s17dlf evaluates a sequence of values for the Hankel function Hν 1 z or Hν 2 z, where z is complex, -π<argzπ, and ν is the real, non-negative order. The N-member sequence is generated for orders ν, ν+1,,ν+N-1. Optionally, the sequence is scaled by the factor e-iz if the function is Hν 1 z or by the factor eiz if the function is Hν 2 z.
Note:  although the routine may not be called with ν less than zero, for negative orders the formulae H-ν 1 z=eνπiHν 1 z, and H-ν 2 z=e-νπiHν 2 z may be used.
The routine is derived from the routine CBESH in Amos (1986). It is based on the relation
Hν m z=1pe-pνKνze-p,  
where p= iπ2  if m=1 and p=- iπ2  if m=2, and the Bessel function Kνz is computed in the right half-plane only. Continuation of Kνz to the left half-plane is computed in terms of the Bessel function Iνz. These functions are evaluated using a variety of different techniques, depending on the region under consideration.
When N is greater than 1, extra values of Hν m z are computed using recurrence relations.
For very large z or ν+N-1, argument reduction will cause total loss of accuracy, and so no computation is performed. For slightly smaller z or ν+N-1, the computation is performed but results are accurate to less than half of machine precision. If z is very small, near the machine underflow threshold, or ν+N-1 is too large, there is a risk of overflow and so no computation is performed. In all the above cases, a warning is given by the routine.


NIST Digital Library of Mathematical Functions
Amos D E (1986) Algorithm 644: A portable package for Bessel functions of a complex argument and non-negative order ACM Trans. Math. Software 12 265–273


1:     m – IntegerInput
On entry: the kind of functions required.
The functions are Hν 1 z.
The functions are Hν 2 z.
Constraint: m=1 or 2.
2:     fnu – Real (Kind=nag_wp)Input
On entry: ν, the order of the first member of the sequence of functions.
Constraint: fnu0.0.
3:     z – Complex (Kind=nag_wp)Input
On entry: the argument z of the functions.
Constraint: z0.0,0.0.
4:     n – IntegerInput
On entry: N, the number of members required in the sequence Hν m z,Hν+1 m z,,Hν+N-1 m z.
Constraint: n1.
5:     scal – Character(1)Input
On entry: the scaling option.
The results are returned unscaled.
The results are returned scaled by the factor e-iz when m=1, or by the factor eiz when m=2.
Constraint: scal='U' or 'S'.
6:     cyn – Complex (Kind=nag_wp) arrayOutput
On exit: the N required function values: cyi contains H ν+i-1 m z , for i=1,2,,N.
7:     nz – IntegerOutput
On exit: the number of components of cy that are set to zero due to underflow. If nz>0, then if Imz>0.0 and m=1, or Imz<0.0 and m=2, elements cy1,cy2,,cynz are set to zero. In the complementary half-planes, nz simply states the number of underflows, and not which elements they are.
8:     ifail – IntegerInput/Output
On entry: ifail must be set to 0, -1 or 1. If you are unfamiliar with this argument you should refer to Section 3.4 in How to Use the NAG Library and its Documentation 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 argument, 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).

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:
On entry, fnu=value.
Constraint: fnu0.0.
On entry, m has illegal value: m=value.
On entry, n=value.
Constraint: n1.
On entry, scal has an illegal value: scal=value.
On entry, z=0.0,0.0.
No computation because z=value<value.
No computation because fnu+n-1=value is too large.
Results lack precision, fnu+n-1=value>value.
Results lack precision because z=value>value.
No computation because z=value>value.
No computation because fnu+n-1=value>value.
No computation – algorithm termination condition not met.
An unexpected error has been triggered by this routine. Please contact NAG.
See Section 3.9 in How to Use the NAG Library and its Documentation for further information.
Your licence key may have expired or may not have been installed correctly.
See Section 3.8 in How to Use the NAG Library and its Documentation for further information.
Dynamic memory allocation failed.
See Section 3.7 in How to Use the NAG Library and its Documentation for further information.


All constants in s17dlf are given to approximately 18 digits of precision. Calling the number of digits of precision in the floating-point arithmetic being used t, then clearly the maximum number of correct digits in the results obtained is limited by p=mint,18. Because of errors in argument reduction when computing elementary functions inside s17dlf, the actual number of correct digits is limited, in general, by p-s, where s max1, log10z , log10ν  represents the number of digits lost due to the argument reduction. Thus the larger the values of z and ν, the less the precision in the result. If s17dlf is called with n>1, then computation of function values via recurrence may lead to some further small loss of accuracy.
If function values which should nominally be identical are computed by calls to s17dlf with different base values of ν and different n, the computed values may not agree exactly. Empirical tests with modest values of ν and z have shown that the discrepancy is limited to the least significant 3 – 4 digits of precision.

Parallelism and Performance

s17dlf is not threaded in any implementation.

Further Comments

The time taken for a call of s17dlf is approximately proportional to the value of n, plus a constant. In general it is much cheaper to call s17dlf with n greater than 1, rather than to make N separate calls to s17dlf.
Paradoxically, for some values of z and ν, it is cheaper to call s17dlf with a larger value of n than is required, and then discard the extra function values returned. However, it is not possible to state the precise circumstances in which this is likely to occur. It is due to the fact that the base value used to start recurrence may be calculated in different regions for different n, and the costs in each region may differ greatly.


This example prints a caption and then proceeds to read sets of data from the input data stream. The first datum is a value for the kind of function, m, the second is a value for the order fnu, the third is a complex value for the argument, z, and the fourth is a character value to set the argument scal. The program calls the routine with n=2 to evaluate the function for orders fnu and fnu+1, and it prints the results. The process is repeated until the end of the input data stream is encountered.

Program Text

Program Text (s17dlfe.f90)

Program Data

Program Data (s17dlfe.d)

Program Results

Program Results (s17dlfe.r)