This is the mail archive of the
gsl-discuss@sources.redhat.com
mailing list for the GSL project.
Re: GSL LU decomposition
- From: Gerard Jungman <jungman at lanl dot gov>
- To: emmanuel dot rouat at st dot com, gsl-discuss at sources dot redhat dot com
- Date: Wed, 05 Dec 2001 14:43:38 -0700
- Subject: Re: GSL LU decomposition
- Organization: LANL T-3
- References: <3C0E50CA.6E8BA733@st.com>
Hi. You should always send messages like this to the
gsl mailing list, gsl-discuss@sources.redhat.com;
other people are interested and can often help.
Emmanuel Rouat wrote:
>
> is known as 'sparse matrix technique' .In the spice simulator,
> these routines were coded by Kundert.
>
> So my question is: do your routines use this technique, or
> are they more straightforward than this?
GSL has no support for sparse matrix methods.
> Also, in a pretty large circuit, the matrices can easily
> reach sizes of 10000x10000 or more... Is the GSL library
> efficient with matrices of that size?
If they are sparse, then you should definitely use a
sparse matrix package instead.
--
G. Jungman