[GRASS-dev] [GRASS-user] r.series weights to zero

GRASS GRASS developer
1

On Fri, Jun 8, 2018 at 12:30 PM, Frank DAVID <frank.david@geophom.fr> wrote:

Cordialement,
Frank David

-------- Message d’origine --------
De : Markus Metz
Date :07/06/2018 17:30 (GMT+01:00)
À : Frank David
Cc : grass-user , GRASS developers list
Objet : Re: [GRASS-user] r.series weights to zero

On Tue, Jun 5, 2018 at 2:51 PM, Frank David <frank.david@geophom.fr> wrote:

Hi all Grass users,

It’s seems that r.series weights does not accept zero as weight… Is it possible ?

Yes, weights must be positive.

and so, why ?

A zero weight means that all values in this map are ignored. Zero weights can also cause the sum of weights to be zero, leading to division by zero and finally to +inf / -inf as a result

Thank you Markus for your reply.
I would think it would have been interesting to use zero as weight on some raster in order to use (or not) them depending on the context. For my purpose i will use a value next to zero.

Agreed. The underlying functions test if the sum of weights is zero, thus it is ok if at least one weight is larger than zero, otherwise the result is NULL. Zero weights are now allowed in r.series, trunk r72789.

Markus M

Thanks again
Frank

Markus M

I’ve got the following error message when running r.series with one weight set to 0

“/usr/local/grass-7.4.0svn/etc/python/grass/script/core.py”,
line 332, in handle_errors
returncode=returncode)
grass.exceptions.CalledModuleError: Module run None

Thank you for your comment.

Frank

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

2

Cordialement,
Frank David

-------- Message d’origine --------
De : Markus Metz
Date :07/06/2018 17:30 (GMT+01:00)
À : Frank David
Cc : grass-user , GRASS developers list
Objet : Re: [GRASS-user] r.series weights to zero

Hi all Grass users,

It’s seems that r.series weights does not accept zero as weight… Is it possible ?

Yes, weights must be positive.

and so, why ?

A zero weight means that all values in this map are ignored. Zero weights can also cause the sum of weights to be zero, leading to division by zero and finally to +inf / -inf as a result

Thank you Markus for your reply.
I would think it would have been interesting to use zero as weight on some raster in order to use (or not) them depending on the context. For my purpose i will use a value next to zero.

Agreed. The underlying functions test if the sum of weights is zero, thus it is ok if at least one weight is larger than zero, otherwise the result is NULL. Zero weights are now allowed in r.series, trunk r72789.

Great !. Thank you Markus
Frank

···

–

Geophom
327 rue de Vieille Cour 44521 OUDON
Tel +33(0)2 85 52 02 59 - Port +33(0)6 04 47 91 06
www.geophom.fr