[GRASS-stats] wrong values imported to R

[Roger Bivand]

On Thu, 16 Oct 2008, Jarek Jasiewicz wrote:

> upps...
> I've just updated and....

to what? Please include sessionInfo() output and exact GRASS version, and 
exact GDAL version (what you see when you load rgdal) and whether this is 
the old (deprecated) or the new plugin. There are lots of things that 
could go wrong, and this looks like a plugin built against a different 
version of GRASS.

This problem (r.out.bin always using 16-bit integers for CELL rasters) was 
discussed in a thread on R-sig-geo recently:

https://stat.ethz.ch/pipermail/r-sig-geo/2008-September/004184.html

and resolved off-list by implementing the plugin. The CVS rgdal on 
sourceforge and spgrass6 on sourceforge also have a useGDAL= argument to 
use a GTiff for the intermediate file, so r.out.gdal from GRASS >= 6.3.

See if gdalinfo on your GRASS raster also crashes - that would suggest a 
plugin that is not synchronised with your GRASS and/or GDAL.

Hope this helps,

Roger

>
>>  r=readRAST6("sca_mroga",plugin=TRUE)
>
> *** caught segfault ***
> address (nil), cause 'unknown'
>
> Traceback:
> 1: .Call("RGDAL_OpenDataset", as.character(filename), TRUE, PACKAGE = 
> "rgdal")
> 2: .local(.Object, ...)
> 3: initialize(value, ...)
> 4: initialize(value, ...)
> 5: new("GDALReadOnlyDataset", filename)
> 6: GDAL.open(fname)
> 7: readGDAL(fname, silent = ignore.stderr)
> 8: readRAST6("sca_mroga", plugin = TRUE)
>
> Possible actions:
> 1: abort (with core dump, if enabled)
> 2: normal R exit
> 3: exit R without saving workspace
> 4: exit R saving workspace
>
> and I have very big problem....
> ...
> Markus Neteler pisze:
>>  On Thu, Oct 16, 2008 at 9:02 PM, Jarek Jasiewicz <jarekj at amu.edu.pl>
>>  wrote:
>> 
>> >  Hi:
>> > 
>> >  data from Grass: (integer)
>> > 
>> >  Range of data:    min = 30  max = 31589370
>> >  (there are more or less 14000 categories ranges as above)
>> > 
>> >  imported to R:
>> > 
>> >  summary(r at data$sca_mroga)
>> >    Min.  1st Qu.   Median     Mean  3rd Qu.     Max.
>> >  -32770.0     60.0    150.0    532.1    330.0  32760.0
>> > 
>> >  I think the problems lies in 16bit proxy .bin file during export, and we
>> >  have overflow
>> > 
>>
>>  Jarek,
>>
>>  if your GRASS-R interface isn't too old, please try
>>  to add plugin=TRUE as parameter.
>>  It should enable the new method of reading in data.
>>
>>  Please report if it (not) works.
>>
>>  Best,
>>  Markus
>> 
> _______________________________________________
> grass-stats mailing list
> grass-stats at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-stats
>
>

-- 
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