On Fri, Dec 6, 2013 at 1:17 PM, Blumentrath, Stefan
stefan.blumentr...@nina.no wrote:
Hi
Maybe it is not of importance, because no one else would use such extreme
settings, but in case someone considers it as a bug:
I was testing a bit with v.outlier in GRASS 7 on dense LiDAR data (~4 points
per m2).
When I use a very low Tykhonov regularization parameter (lambda_i) of
0.001 and rather short spline steps (1.5) at 0.5m resolution
I get an error-message:
“Decomposition failed at row 16499 and col 0”
I found similar errors with very small lambda values, i guess it's
some floating point precision limit. It seems that lambda should not
be much smaller than 0.0001. For v.outlier, I would leave lambda at
the default and rather use shorter spline steps, if you want to
fine-tune the results.
Markus M
___
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev
