[GRASS-dev] [GRASS-user] v.neighors error:region do not overlap

On Wed, Aug 5, 2015 at 10:25 AM, patrick s. <patrick_gis@gmx.net> wrote:

UPDATE:

The error of v.neighbors does not appear when importing points only for a
region: "v.in.ogr -r"

So definitely a problem of v.neighbors.

Looking at v.neighbors source code [1], the message seems wrong, it should
be something like "Some input vector points are outside of computational
region". But I don't understand why this should prevent v.neighbors from
running. Perhaps this test could be removed?

Anna

[1]
https://trac.osgeo.org/grass/browser/grass/trunk/vector/v.neighbors/main.c#L95

Patrick

On 01.08.2015 23:56, Markus Neteler wrote:

On Wed, Jul 29, 2015 at 4:40 PM, patrick s. <patrick_gis@gmx.net> wrote:

Hi,

I need to run v.neighbors on multiple datasets for the identic region
(extend and resolution). One dataset has observations outside of this
region, and causes an error: "ERROR: Input vector and computational
region
do not overlap." However, viewing the extend in v.info and r.info (see
below) and overlaying the layers in the GUI, I see overlaps for 90% of my
points. Is this expected behavior or do I miss something? (GRASS70 from
Repos, OS:Ubuntu 15.04)

#Code
r.in.gdal in="./00_REFERENCE/reference_map.tif" out=region_ref --o
g.region rast=region_ref

Just to be sure, please post

g.region raster=region_ref -p

I generated a location with epsg:2056 (is that the right one for
CH_1903+ ?) and there is indeed map overlap.
But just to be sure that the computational region is right, too.

Markus

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

On Wed, Aug 5, 2015 at 4:51 PM, Anna Petrášová <kratochanna@gmail.com> wrote:

On Wed, Aug 5, 2015 at 10:25 AM, patrick s. <patrick_gis@gmx.net> wrote:

UPDATE:

The error of v.neighbors does not appear when importing points only for a
region: "v.in.ogr -r"

So definitely a problem of v.neighbors.

Looking at v.neighbors source code [1], the message seems wrong, it should
be something like "Some input vector points are outside of computational
region". But I don't understand why this should prevent v.neighbors from
running. Perhaps this test could be removed?

The test got added in
https://trac.osgeo.org/grass/changeset/62746

The motivation I don't know. How about changing it from error msg to a
warning for now?

Markus

Anna

[1]
https://trac.osgeo.org/grass/browser/grass/trunk/vector/v.neighbors/main.c#L95

On Thu, Aug 6, 2015 at 2:08 PM, Markus Neteler <neteler@osgeo.org> wrote:

On Wed, Aug 5, 2015 at 4:51 PM, Anna Petrášová <kratochanna@gmail.com> wrote:

...

Looking at v.neighbors source code [1], the message seems wrong, it should
be something like "Some input vector points are outside of computational
region". But I don't understand why this should prevent v.neighbors from
running. Perhaps this test could be removed?

The test got added in
https://trac.osgeo.org/grass/changeset/62746

The motivation I don't know. How about changing it from error msg to a
warning for now?

... done so in 7.1.svn and 7.0.svn.
The module now continues.

Markus