[Freegis-list] gstat

[Markus Neteler]

Hi Markus,
hi all,

On Thu, Feb 08, 2001 at 07:47:36AM +0100, Markus Böhm wrote:
> Freegis-list, Felipe,
> 
> "Felipe G. Nievinski" wrote:
> 
> > http://www.gstat.org/
> > Do you know it?
> >
> > (sic)
> > "Gstat is a computer program for geostatistical modelling, prediction and
> > simulation. Gstat works at least on the following platforms: Linux, hp-ux
> > (9.x, 10.x), aix (3.2, 4.1), OSF/Alpha, SunOS, SGI, win32 and MS-DOS (dpmi).
> > Gstat is Open Source software, distributed under the terms of the GNU
> > General Public Licence (meaning it's freely available in binary and source
> > code form). "
> >
> > Although it is not a GIS, it can be classified as a 'GeoInformatics'
> > software, and therefore I think it can be in FreeGIS.org software list.
> >
> > Best regards,
> > Felipe Nievinski.
> >
> 
> nice hint, I think.
> I used gstat for my master thesis in combination with other GIS operations.
> Particularly I had very few success with the geostatistical modules of GRASS
> GIS 5x (BTW (to Markus Neteler?): will these modules get further developement?)
> so I had to try different programs to do the geostatistical interpolations.
The GRASS 5beta11 s.* modules are somewhat better than in past. Generally
we'll continue to work on the "R statistical language" interface (Roger
Bivand) - no need to re-invent the wheel in GRASS.
But...
 
> gstat did the job very well and the file exchange with GRASS was not really a
> problem because gstat supports a lot of Grid map formats (for example idrisi,
> ArcInfo/View, ER-Mapper, GMT grid maps, Surfer). As you can see there is
> missing the GRASS raster format which would make the combined use of these
> programs even more comfortable. Perhaps it is possible to develop an interface
> :-)
It's already there :-) I have been to Italy last week, the Politecnico di
Milano (Prof. Maria Brovelli) has been developing a small gstat interface to
GRASS. I'll try to get that. General agreement is there.

> In my opinion the documentation of gestat is quite good as well.
> 
> So I can say that it would be a further good thing for the FreeGIS-CD
I agree.

Cheers

 Markus Neteler