[STATSGRASS] Re: gmeta6() returns empty string

[Roger Bivand]

On Thu, 1 Mar 2007, Tilmann Steinmetz wrote:

> Hi Roger and all,
> 
> > 
> > Very strange, maybe when Stromboli calms down? 
> > 
> 
> I'm still covered with tephra but starting to see clearer: I tried executing 
> system("g.region -p") in R on a similar machine, running R 2.2.1 on FreeBSD 
> 5.4 and GRASS6.0.2,2 which _does not_ have similar problems- so the shell 
> thing definitely is not a general FreeBSD problem. I am still thankful for 
> every hint that may lead to the wanted rogue environment variable or 
> whatever.

G_gettext is also in GRASS 6.0.2, I just checked.

> 
> > > 
> > > Is there any way I can go from here without running Grass on, 
> > say, Linux?
> > 
> > Yes, of course. Just run them side-by-side, with GRASS using 
> > r.in.gdal and 
> > v.in.ogr to read files written from R by writeGDAL() and writeOGR() 
> > in 
> > rgdal, and r.out.* and v.out.ogr wrt. readOGR() and readGDAL() in 
> > rgdal in 
> 
> oh great! I wasn't aware of that.
>

Then that may be the easiest way to go forward.
 
> > R. This is almost what the interface does anyway, but because the 
> > new 
> > stand-alone r.out.gdal is in 6.3, you may need to choose your 
> > prefered 6.2 
> > raster output format avoiding the older scripted r.out.gdal if you 
> > are not 
> > running the GDAL GRASS plugin (best not to, IMO). All the interface 
> > gives 
> > over and above this is handling the intermediate files as temporary 
> > files, 
> > and adding GRASS categories as factor levels to R factors.
> > 
> > How did you install R, I assume you built from source? Did you 
> > notice any 
> > worrying messages?
> 
> yes - it was installed from the FreeBSD ports collection (i.e. a source 
> install which resolves dependencies, or: is meant to ..) I am now updating - 
> lets see what happens.
> 

OK. Please say if you see anything there. The r.out.gdal in GRASS 6.3 is a 
big improvement over the scripted version in GRASS 6.2 and earlier.

Roger

> Tilmann
> > > 
> > > ----- Originalnachricht -----
> > > Von: Roger Bivand <Roger.Bivand at nhh.no>
> > > Datum: Mittwoch, Februar 28, 2007 14:46
> > > Betreff: Re: WG: Re: WG: Re: gmeta6() returns empty string
> > > An: Tilmann Steinmetz <tilmann.steinmetz at geo.uni-halle.de>
> > > 
> > > > Does this usually correlate with activity on Stromboli?
> > > 
> > > "g.list rast" rather points to rushmore or even better: bugsites :}
> > > 
> > > 
> > 
> > -- 
> > Roger Bivand
> > Economic Geography Section, Department of Economics, Norwegian 
> > School of
> > Economics and Business Administration, Helleveien 30, N-5045 Bergen,
> > Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43
> > e-mail: Roger.Bivand at nhh.no
> > 
> > 
> 

-- 
Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Helleveien 30, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: Roger.Bivand at nhh.no