I tried to link a PG table to GRASS for point sampling using v.external (in GRASS 7.0.6).
With that linked layer (which is not in public schema and has column “point_id” as primary key) I get a lot of:
WARNING: No record for category XXXX in table
<my_schema.my_point_layer>
when I use v.what.rast on that layer.
The results suggest that not the proper IDs were used when values were loaded to attribute table. Because values in the attribute table of the points do not match with values in the raster at point locations. Furthermore, some points, which are supposed to have values in the attribute table, have none…
Did I overlook something at import or should I open a ticket?
When I try to connect to a remote server in GRASS 7.2.1svn with v.external, I get "database not found" error.
In GRASS 7.0 the database is found with exactly the same command.
However, linking the table fails in 7.0 because the ID column is in mixed cases and column names are not quoted in the SQL generated by v.external.
BTW. Is it necessary to sort the ID column? Does it make sense if UUIDs are used as primary keys?
Cheers
Stefan
-----Original Message-----
From: Martin Landa [mailto:landa.martin@gmail.com]
Sent: mandag 9. januar 2017 18.17
To: Blumentrath, Stefan <Stefan.Blumentrath@nina.no>
Cc: GRASS developers list (grass-dev@lists.osgeo.org) <grass-dev@lists.osgeo.org>
Subject: Re: [GRASS-dev] v.external: ID confusion
Hi Stefan,
2017-01-09 16:42 GMT+01:00 Blumentrath, Stefan <Stefan.Blumentrath@nina.no>:
Did I overlook something at import or should I open a ticket?
yes, please. Also provide all command and sample data to reproduce this issue. Thanks, Ma
2017-01-11 12:54 GMT+01:00 Blumentrath, Stefan <Stefan.Blumentrath@nina.no>:
When I try to connect to a remote server in GRASS 7.2.1svn with v.external, I get "database not found" error.
new issue? I cannot reproduce it. Can you provide any details?
However, linking the table fails in 7.0 because the ID column is in mixed cases and column names are not quoted in the SQL generated by v.external.
BTW. Is it necessary to sort the ID column? Does it make sense if UUIDs are used as primary keys?