[GRASS-user] v.cellstats

[Hamish]

Carlos "Guâno" Grohmann wrote:

> > > are there iny news on v.cellstats?
> >
> > try r.in.xyz.
> >
> > http://grass.ibiblio.org/grass61/manuals/html61_user/r.in.xyz.html
>
> Thanks Hamish. It seems to be what I need, just that.. it doesen't
> have a 'mode' option in the stats... that's what I really need

No mode, sorry. r.univar, r.in.xyz, etc still wait for someone to help
write advanced statistics in an efficient way. From the TODO section of
r.in.xyz help page:

TODO

* Support for advanced statistics (in parallel with r.univar).
Especially useful for dealing with outliers would be median and 5-10%
trimmed means.

The equivalent module from GRASS 5 (s.cellstats) contains code for
additional statistical options, claiming to use only 16 bytes per cell:
skewness, kurtosis, mean of squares, mean of absolute values, first
quartile, median, third quartile.



Note that "mode" isn't on that list, however if you want one of those
you can use GRASS 5.4 to create a raster which is fully compatible with
GRASS 6. You can have multiple versions of GRASS installed without
problems- you might try GRASS 5.4 from CVS to get around problems
reading 3D region settings from GRASS 6 which don't exist in GRASS 5.

s.in.ascii
s.cellstats
s.to.rast


out of curiosity, what is the nature of the dataset? LIDAR? Swath sonar?



Hamish