Removing the behavior of i.maxlik which add the output to input group

[grass-svn2git]

Reported by spareeth on 20 Mar 2018 15:51 UTC The final result from i.maxlik is added to input group by default. which in my opinion is strange. I dont know if there is any reason behind this behaviour.

In the mapset nc_spm_08_grass7/landsat;

i.group group=lsat7_2000_NW subgroup=lsat7_2000_NW input=lsat7_2000_10,lsat7_2000_20,lsat7_2000_30,lsat7_2000_40,lsat7_2000_50
i.cluster group=lsat7_2000_NW subgroup=lsat7_2000_NW signaturefile=sig_cluster_lsat2000 classes=10 reportfile=rep_clust_lsat2000.txt
i.maxlik group=lsat7_2000_NW subgroup=lsat7_2000_NW  signaturefile=sig_cluster_lsat2000 output=lsat7_2000_cluster_10classes reject=lsat7_2002_cluster_reject
i.group lsat7_2000_NW -l
group <lsat7_2000_NW> references the following raster maps
-------------
<lsat7_2000_10@landsat>                   
<lsat7_2000_20@landsat>                   
<lsat7_2000_30@landsat>                   
<lsat7_2000_40@landsat>                   
<lsat7_2000_50@landsat>                   
<lsat7_2000_cluster_10classes@landsat>    
<lsat7_2002_cluster_reject@landsat>  

Would be better to remove this behaviour in future release.

GRASS GIS version and provenance

svn-trunk

Migrated-From: https://trac.osgeo.org/grass/ticket/3525

[grass-svn2git]

Comment by @veroandreo on 1 Nov 2018 12:44 UTC +1!

It's a quite annoying to have to remove the classification and reject maps from the group in each new run, better leave optional to the user to add it if he/she wants it as an additional input layer.

[grass-svn2git]

Comment by @landam on 1 Nov 2018 12:48 UTC +0 Key point is that output maps are not added to subgroup (which is used by most of imagery modules for reading list of input maps).

[grass-svn2git]

Comment by @landam on 1 Nov 2018 12:48 UTC So you don't need to remove output maps from group in each new run...

[grass-svn2git]

Modified by @landam on 1 Nov 2018 12:49 UTC

[grass-svn2git]

Comment by @veroandreo on 1 Nov 2018 13:32 UTC Replying to [comment:3 martinl]:

So you don't need to remove output maps from group in each new run... Good! Thanks for clarification, Martin! I still prefer not to have classification results in the group, though ;)

[grass-svn2git]

Comment by @metzm on 1 Nov 2018 21:00 UTC Replying to [comment:5 veroandreo]:

Replying to [comment:3 martinl]:

So you don't need to remove output maps from group in each new run... Good! Thanks for clarification, Martin! I still prefer not to have classification results in the group, though ;)

Assuming that imagery modules accept a group/subgroup as input that is not in the current mapset, the output can not be added to the input group if it is not in the current mapset.

The metadata of the output (datasource, description, comments) should in any case contain information about how this output has been generated: fully qualified input group/subgroup name, fully qualified input signature file name if appropriate, and other parameters used by i.maxlik.

[grass-svn2git]

Comment by marisn on 2 Nov 2018 14:11 UTC Replying to [comment:6 mmetz]:

Replying to [comment:5 veroandreo]: Assuming that imagery modules accept a group/subgroup as input that is not in the current mapset, the output can not be added to the input group if it is not in the current mapset.

This is a good argument. Still at the moment only library and i.group has been changed to allow reading group contents from other mapsets. See https://trac.osgeo.org/grass/changeset/73632 and https://trac.osgeo.org/grass/changeset/73633

This issue should be viewed together with signature files as those also are placed into the group/subgroup.

[grass-svn2git]

Comment by @mlennert on 5 Nov 2018 10:39 UTC See also https://trac.osgeo.org/grass/ticket/2483.