[GRASS-stats] GRASS/R interface

Agustin Lobo Agustin.Lobo at ija.csic.es
Fri Feb 13 06:09:48 EST 2009


Roger,

I also recall not having problems with qgis 0.11, I think this is a
osgeow4 problem.
The sequence (in windows using osgeo4w qgis1.0 and grass) is:

1. Start QGIS 1.0
2. Open project
2. Start GRASS plugin
3. Open mapset
4. Start GRASS tools
5. Start Grass shell
6. Start R from the shell
7. Run spgrass6() : gdal error at loading rgdal (which, actually,
would not be required for
what we are doing). In our previous messages, we concluded that these
errors are osgeo4w issues that we would not be able to solve.

spgrass6() loads rgdal with no problems if R is not launched from the
grass shell, but
then "grass is not running". This is why I asked if the grass info
could be passed to
spgrass6() explicitly by the user as parameters. Another way to
circumvent the problem would be just leaving the rgdal loading up to
the user (i.e., spgrass6(rgdal=F) )

Note that this is not an operational problem for me now, we've just
dropped using
windows for that project. The same sequence works fine in linux. This
has actually been
an advantage, as the student has switched from win to lin. (Many
students don't because the current R default gui in windows is much
better than in linux, but this is another story)

Agus



2009/2/13 Roger Bivand <Roger.Bivand at nhh.no>:
> On Fri, 13 Feb 2009, Agustin Lobo wrote:
>
>> Thanks, where?
>> Also, there is no way the grass environment could be passed
>> to spgrass6 if R had been started from a not-grass terminal?
>> This would circumvent a current problem in the
>> windows osgeo4w QGIS-GRASS-R environment.
>
> I'm still unsure of the steps you are taking. Are you opening a GRASS
> location/mapset - for example from GRASS tools under Plugins ->, then from
> the tools choosing the console, and entering the command to start R there?
> Could you put a sequence of screenshots describing your process on a site
> somewhere - I have no problem with regular QGIS 0.11.0. Starting spgrass6
> outside the GRASS environment will most likely not work well at all, because
> nothing that it needs to run is available (progam PATH, data location,
> etc.).
>
> Roger
>
>> Agus
>>
>> 2009/2/12 Markus Neteler <neteler at osgeo.org>:
>>>
>>> On Wed, Feb 11, 2009 at 8:25 AM, Agustin Lobo <aloboaleu at gmail.com>
>>> wrote:
>>>>
>>>> In relation to current problems with spgrass6() in a
>>>> windows osgeo4w QGIS-GRASS-R environment,
>>>> I'd like to make sure that using spgrass6()
>>>> actually requires having launched R from within a
>>>> GRASS terminal, otherwise the user gets the
>>>> "GRASS not running" error message. Am I wrong? This is an important
>>>> step in practice that is not mentioned in the documents
>>>> that you are refering to.
>>>
>>> Now mentioned!
>>>
>>> best
>>> Markus
>>>
>>
>>
>>
>>
>
> --
> 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
>
>



-- 
Dr. Agustin Lobo
Institut de Ciencies de la Terra "Jaume Almera" (CSIC)
Lluis Sole Sabaris s/n
08028 Barcelona
Spain
Tel. 34 934095410
Fax. 34 934110012
e-mail Agustin.Lobo at ija.csic.es
http://www.ija.csic.es/gt/obster


More information about the grass-stats mailing list