[GRASS-dev] Re: R + grass7

GRASS GIS GRASS GIS developer
1

fyi a question regarding an “translation” (flags,parameter) table between grass6x and grass7.

Helmut

-------- Original-Nachricht --------
Datum: Wed, 29 Aug 2012 14:35:37 +0200 (CEST)
Von: Roger Bivand
An: Helmut Kudrnovsky
Betreff: Re: R + grass7

On Wed, 29 Aug 2012, Helmut Kudrnovsky wrote:

Dear Roger,

I’m trying to improve the GRASS GIS-R-coupling on windows
(http://trac.osgeo.org/grass/ticket/1149).

May I ask you if spgrass6
(http://cran.at.r-project.org/web/packages/spgrass6/index.html) is GRASS
GIS 7-aware.

Yes, the description on:

http://cran.r-project.org/web/packages/spgrass6/index.html

uses “6+” to signal that subsequent versions are intended to work. Since
the package only uses system(), it will work as long as
–interface-description returns an XML file. Plenty of examples and
scripts stop working because GRASS 6 and 7 modules have different flags
and/or parameter names, but in principle only for that reason. It’s hard
for me to check - it would be useful to have a translation table for flag
and parameter names by module where changes have occurred.

Best wishes,

Roger

–
Roger Bivand
Department of Economics, NHH Norwegian School of Economics,
Helleveien 30, N-5045 Bergen, Norway.

2

On Wed, Aug 29, 2012 at 2:50 PM, Helmut Kudrnovsky <hellik@web.de> wrote:

fyi a question regarding an "translation" (flags,parameter) table between
grass6x and grass7.

We are maintaining a list of changes here:
http://trac.osgeo.org/grass/wiki/Grass7/NewFeatures#Renamedoptions

The idea to be least invasive while gaining maximal consistency.
I hope that only a few scripts/examples are failing after migrating
from GRASS 6 to 7.

Markus

3

Hi,

You can find a list of changes in option settings and names here: http://trac.osgeo.org/grass/wiki/Grass7/NewFeatures#Renamedoptions

Cheers,

Paulo

On 08/29/2012 02:50 PM, Helmut Kudrnovsky wrote:

fyi a question regarding an "translation" (flags,parameter) table between grass6x and grass7.

Helmut

-------- Original-Nachricht --------
Datum: Wed, 29 Aug 2012 14:35:37 +0200 (CEST)
Von: Roger Bivand
An: Helmut Kudrnovsky
Betreff: Re: R + grass7

On Wed, 29 Aug 2012, Helmut Kudrnovsky wrote:

> Dear Roger,
>
> I'm trying to improve the GRASS GIS-R-coupling on windows
> (http://trac.osgeo.org/grass/ticket/1149). <https://service.gmx.net/de/cgi/derefer?TYPE=3&DEST=http%3A%2F%2Ftrac.osgeo.org%2Fgrass%2Fticket%2F1149).&gt;
>
> May I ask you if spgrass6
> (http://cran.at.r-project.org/web/packages/spgrass6/index.html) <https://service.gmx.net/de/cgi/derefer?TYPE=3&DEST=http%3A%2F%2Fcran.at.r-project.org%2Fweb%2Fpackages%2Fspgrass6%2Findex.html)&gt; is GRASS
> GIS 7-aware.

Yes, the description on:

http://cran.r-project.org/web/packages/spgrass6/index.html <https://service.gmx.net/de/cgi/derefer?TYPE=3&DEST=http%3A%2F%2Fcran.r-project.org%2Fweb%2Fpackages%2Fspgrass6%2Findex.html&gt;

uses "6+" to signal that subsequent versions are intended to work. Since
the package only uses system(), it will work as long as
--interface-description returns an XML file. Plenty of examples and
scripts stop working because GRASS 6 and 7 modules have different flags
and/or parameter names, but in principle only for that reason. It's hard
for me to check - it would be useful to have a translation table for flag
and parameter names by module where changes have occurred.

Best wishes,

Roger

--
Roger Bivand
Department of Economics, NHH Norwegian School of Economics,
Helleveien 30, N-5045 Bergen, Norway.

_______________________________________________
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user

4

On Wed, 29 Aug 2012, Paulo van Breugel wrote:

Hi,

You can find a list of changes in option settings and names here: http://trac.osgeo.org/grass/wiki/Grass7/NewFeatures#Renamedoptions

Thanks, this may be helpful, but is misses small things like the name being the same but the representation changing. For example, in some v.* like v.info, the layer parameter in 6* is integer, but 7 wants character.

http://grass.osgeo.org/grass64/manuals/html64_user/v.info.html

http://grass.osgeo.org/grass70/manuals/html70_user/v.info.html

r.info is also changed, with 6* -t having to be re-created for 7 -g then find the "datatype" in the output.

http://grass.osgeo.org/grass64/manuals/html64_user/r.info.html

http://grass.osgeo.org/grass70/manuals/html70_user/r.info.html

Those broke the R code, so spgrass6 has to branch on GRASS 6/7 version in these cases; others I've seen break example scripts. For script maintainers, it may be unavoidable that changes have to be made, I agree, but it can be puzzling.

Roger

Cheers,

Paulo

On 08/29/2012 02:50 PM, Helmut Kudrnovsky wrote:

fyi a question regarding an "translation" (flags,parameter) table between grass6x and grass7.

Helmut

-------- Original-Nachricht --------
Datum: Wed, 29 Aug 2012 14:35:37 +0200 (CEST)
Von: Roger Bivand
An: Helmut Kudrnovsky
Betreff: Re: R + grass7

On Wed, 29 Aug 2012, Helmut Kudrnovsky wrote:

> Dear Roger,
>
> I'm trying to improve the GRASS GIS-R-coupling on windows
> (http://trac.osgeo.org/grass/ticket/1149). <https://service.gmx.net/de/cgi/derefer?TYPE=3&DEST=http%3A%2F%2Ftrac.osgeo.org%2Fgrass%2Fticket%2F1149).&gt;
>
> May I ask you if spgrass6
> (http://cran.at.r-project.org/web/packages/spgrass6/index.html) <https://service.gmx.net/de/cgi/derefer?TYPE=3&DEST=http%3A%2F%2Fcran.at.r-project.org%2Fweb%2Fpackages%2Fspgrass6%2Findex.html)&gt; is GRASS
> GIS 7-aware.

Yes, the description on:

http://cran.r-project.org/web/packages/spgrass6/index.html <https://service.gmx.net/de/cgi/derefer?TYPE=3&DEST=http%3A%2F%2Fcran.r-project.org%2Fweb%2Fpackages%2Fspgrass6%2Findex.html&gt;

uses "6+" to signal that subsequent versions are intended to work. Since
the package only uses system(), it will work as long as
--interface-description returns an XML file. Plenty of examples and
scripts stop working because GRASS 6 and 7 modules have different flags
and/or parameter names, but in principle only for that reason. It's hard
for me to check - it would be useful to have a translation table for flag
and parameter names by module where changes have occurred.

Best wishes,

Roger

--
Roger Bivand
Department of Economics, NHH Norwegian School of Economics,
Helleveien 30, N-5045 Bergen, Norway.

_______________________________________________
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user

--
Roger Bivand
Department of Economics, NHH Norwegian School of Economics,
Helleveien 30, N-5045 Bergen, Norway.
voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: Roger.Bivand@nhh.no