This is the mail archive of the
gsl-discuss@sources.redhat.com
mailing list for the GSL project.
Re: truncated normal distribution
- To: dupas at pge dot cnrs-gif dot fr
- Subject: Re: truncated normal distribution
- From: Brian Gough <bjg at network-theory dot co dot uk>
- Date: Tue, 20 Feb 2001 10:39:14 +0000 (GMT)
- Cc: gsl-discuss at sources dot redhat dot com, Olivier dot Langella at pge dot cnrs-gif dot fr
- References: <4.2.0.58.20010219142125.00c0be70@mailhost.pge.cnrs-gif.fr>
- Reply-To: gsl-discuss at sources dot redhat dot com
dupas@pge.cnrs-gif.fr writes:
> Is there tables of the normal distribution available, calculating
> the upper value or the mean of the tail knowing the surface of the
> tail ? (and reciprocally) ?
> What is the best method to perform numerical integration of normal
> distribution ?
> How to obtain the upper value knowing the tail ?
Hi,
The specfunc directory includes the functions erf(x) and erfc(x),
which compute the area of the tail. See the header file gsl_sf_erf.h
for details. I'm not sure if there are any functions for finding the
upper value given the tail -- you could find the value by using one of
the root-finding algorithms to solve for the inverse.
See the integration chapter in the manual for details of numerical
integration. The routines gsl_integration_qag... will be the most useful
for working with probability distributions.
regards
Brian Gough