This is the mail archive of the
gsl-discuss@sources.redhat.com
mailing list for the GSL project.
Re: integrating over a multiple step-function
- From: Daniel Rohe <d dot rohe at fkf dot mpg dot de>
- To: Daniel Rohe <d dot rohe at fkf dot mpg dot de>
- Cc: gsl list <gsl-discuss at sources dot redhat dot com>
- Date: Mon, 08 Apr 2002 13:41:19 +0200
- Subject: Re: integrating over a multiple step-function
- References: <3CB16447.203@fkf.mpg.de>
shame on me again! as the manual says: qag with a low key is suitable
for a function with continuities.
I guess I should do a grep on the manual next time; or maybe start a
mailing list with myself as the only participant.
sorry about uselessly sharing my problems with the rest of the world!
daniel
Daniel Rohe wrote:
> here's a somewhat tricky problem to be solved by an integration routine:
>
> the function f is to be integrated say over the intervall (a,b).
> however the kernel
> is folded/multiplied by a theta function turning it into a potentially
> multiple step-function.
> the location of the steps is not known so the routine needs to take
> care of this.
>
> until now I've been using the qagp-routine, but it seems to fail in
> some cases.
> can anyone suggest an alternative, such as maybe the qags-method?
>
> daniel
>