[GRASS-user] [GRASS-dev] execGRASS("r.diversity"): [Error 3] Cannot find path specified \\.grass7\\r.li

Markus Neteler neteler at osgeo.org
Thu May 22 01:48:02 PDT 2014


Hi Sandra,

(adding Luca Delucchi, author, in cc)

back to the full problem:

On Wed, May 21, 2014 at 12:08 PM, Sandra MacFadyen <sandramf at live.co.za> wrote:
...
> I am new to GRASS and have a problem executing the r.diversity function via
> R 3.1.0 and GRASS 7beta2:
>
> When running the code:
>
> # Calculate spectral diversity per band with windows size 5x5
> execGRASS("r.diversity",flags="overwrite", parameters=list(input="name.1",
> prefix="name_divers.1", alpha=0.5, size=5))

Could you run the command directly in GRASS (rather than R) for a test?

On Thu, May 22, 2014 at 9:31 AM, Sandra MacFadyen <sandramf at live.co.za> wrote:
...
> I have the correct update of r.diversity.py in the correct directory but now I am getting an error, " Illegal areas disposition"
>
>> execGRASS("r.diversity",flags="overwrite", parameters=list(input="s1993_tm5_rad.1", prefix="s1993_tm5_rad_divers.1", alpha=0.5, size=5))
> ERROR: Illegal areas disposition
> ERROR: Illegal areas disposition
> ERROR: Illegal areas disposition
> ERROR: Illegal areas disposition
> Done.
>
>
> As replied to Glynn earlier. I created a configuration file using r.li.setup in this directory as follows:

The command r.diversity does not require the creation of a config
file. This is done automatically for the indices calculated by the
command (on the contrary to the r.li.* commands which do need such a
config file).

> "conf_diversity_5.0"
> SAMPLINGFRAME 0|0|1|1
> SAMPLEAREA -1|-1|0.010526315789473684|0.009487666034155597
> MOVINGWINDOW
>
> This file disappears after the error occurs above.

You just named it by chance as the internal file name of the command.
@author: Perhaps a PID or something random (urandom?) might be added
to avoid this non-issue :)

If you want to control everything, probably you need to pick the
r.li.* command of interest which accepts configuration files.

Best
Markus


More information about the grass-user mailing list