[GRASS-user] Problem running i.segment.stats

[Moritz Lennert]

On 18/10/16 17:44, James Duffy wrote:
> On 18 October 2016 at 16:37, Moritz Lennert
> <mlennert at club.worldonline.be <mailto:mlennert at club.worldonline.be>> wrote:
>
>     On 18/10/16 17:28, James Duffy wrote:
>
>         On 18 October 2016 at 16:17, Moritz Lennert
>         <mlennert at club.worldonline.be
>         <mailto:mlennert at club.worldonline.be>
>         <mailto:mlennert at club.worldonline.be
>         <mailto:mlennert at club.worldonline.be>>> wrote:
>
>             On 18/10/16 14:27, James Duffy wrote:
>
>                 Hello,
>
>                 I'm using GRASS 7.0.4 on the OSGeoLive pre-built
>         operating system. I
>                 have installed the addon 'i.segment.stats' to use with
>         my output
>                 from
>                 'i.segment.uspo'.
>
>                 When selecting the 'rasters' input for this tool, I
>         choose the
>                 original
>                 layers that were used to create the segmented map in the
>         first
>                 place.
>                 Upon execution of 'i.segment.stats', i get the following
>         error:
>
>                 Element type should be "cell" and not "raster.
>
>
>             The above message is just a warning and should not keep the
>         module
>             from working. I just fixed this (r69700), so try reinstalling
>             i.segment.stats.
>
>
>         Ok thank you, have reinstalled and the message no longer appears.
>
>
>
>                 The tool does not continue and outputs a blank .csv file.
>
>                 Any help solving this would be much appreciated.
>
>
>             Could you send us the command line used (you can get that by
>         pushing
>             the 'Copy' button in the GUI of the module or through the
>         command
>             history in the GUI 'command console')
>
>
>         i.segment.stats --overwrite --verbose
>         map=angle_seg_uspo_angle_region_1_rank1 at angle1
>         rasters=angle_ortho.1 at angle1
>         raster_statistics=min,max,mean,stddev,sum
>         csvfile=/home/jd/Wales_GRASS/isegstats_gp_test separator=comma
>
>         With this command, the output .csv is still empty apart from the
>         headers.
>
>
>     Are you sure the region is set to
>     angle_seg_uspo_angle_region_1_rank1 ? I.e.:
>
>     g.region rast=angle_seg_uspo_angle_region_1_rank1
>
>
> This fixed it. Sorry, I assumed that the region was already set to the
> rasters that were used to create the segmented map (which are identical
> in extent to the segmented map itself). Will make a note in my script to
> always set the region to the input for i.segment.stats.

It all depends on how you used i.segment.uspo and what the region 
setting before running it was...

Glad you found the issue.

Moritz