Next: calling lapack routine from C++
From: ying.linying on 12 Jul 2005 05:01
Actually it is one standard fourier transform integration for kx and
ky. Inner integral (kx)is infinit and outer limit of integral (ky)is
function of kx. I am not familiar with such programm, so is there
anybody having experience on such kind of programm? Thanks a lot.

From: Andrew Smith on 12 Jul 2005 18:09
I used IMSL succesfully for a two dimensional integration but I am not sure
what you mean by the function is infinite. This is an extract from the
online manual:

TWODQ/DTWODQ (Single/Double precision)
Compute a two-dimensional iterated integral.

Usage

CALL TWODQ (F, A, B, G, H, ERRABS, ERRREL, IRULE, RESULT,

ERREST)

Arguments

F - User-supplied FUNCTION to be integrated. The form is F(X, Y), where

X - First argument of F. (Input)

Y - Second argument of F. (Input)

F - The function value. (Output)

F must be declared EXTERNAL in the calling program.

A - Lower limit of outer integral. (Input)

B - Upper limit of outer integral. (Input)

G - User-supplied FUNCTION to evaluate the lower limits of the inner
integral.

The form is G(X), where

X - Only argument of G. (Input)

G - The function value. (Output)

G must be declared EXTERNAL in the calling program.

H - User-supplied FUNCTION to evaluate the upper limits of the inner
integral.

The form is H(X), where

X - Only argument of H. (Input)

H - The function value. (Output)

H must be declared EXTERNAL in the calling program.

ERRABS - Absolute accuracy desired. (Input)

ERRREL - Relative accuracy desired. (Input)

IRULE - Choice of quadrature rule. (Input)

The Gauss-Kronrod rule is used with the following points:

IRULE Points

1 7-15

2 10-21

3 15-31

4 20-41

5 25-51

6 30-61

If the function has a peak singularity, use IRULE = 1. If the function is
oscillatory,

use IRULE = 6.

RESULT - Estimate of the integral from A to B of F. (Output)


<ying.linying(a)gmail.com> wrote in message
news:1121158874.730115.218840(a)f14g2000cwb.googlegroups.com...
> Actually it is one standard fourier transform integration for kx and
> ky. Inner integral (kx)is infinit and outer limit of integral (ky)is
> function of kx. I am not familiar with such programm, so is there
> anybody having experience on such kind of programm? Thanks a lot.
>


From: Madhusudan Singh on 12 Jul 2005 18:12
ying.linying(a)gmail.com wrote:

> Actually it is one standard fourier transform integration for kx and
> ky. Inner integral (kx)is infinit and outer limit of integral (ky)is
> function of kx. I am not familiar with such programm, so is there
> anybody having experience on such kind of programm? Thanks a lot.

You might wish to consult quadpack.
From: grace on 13 Jul 2005 00:05
Thanks. I also have this one, but there is exact upper and lower limit
for internal integral. For my case, the internal upper limit and lower
limit are both infinite number, not a function or a finite number.

From: Andres Kuusk on 13 Jul 2005 02:52
grace wrote:

> For my case, the internal upper limit and lower
> limit are both infinite number, not a function or a finite number.

Look at Section 4.4 (Improper Integrals) in Numerical Recipes by Press et
al., Cambridge University Press. Subroutines midsql() and midsqu() may
help you.


Andres Kuusk
Tartu Observatory, Estonia
 | 
Pages: 1
Next: calling lapack routine from C++