[GRASS-dev] Re: [GRASS GIS] #465: r.proj.seg thins along null areas and raster bounds for bilinear and cubic methods

[GRASS GIS]

#465: r.proj.seg thins along null areas and raster bounds for bilinear and cubic
methods
--------------------------+-------------------------------------------------
  Reporter:  kyngchaos    |       Owner:  grass-dev at lists.osgeo.org
      Type:  enhancement  |      Status:  new                      
  Priority:  major        |   Milestone:                           
 Component:  Raster       |     Version:  svn-develbranch6         
Resolution:               |    Keywords:                           
  Platform:  All          |         Cpu:  All                      
--------------------------+-------------------------------------------------
Comment (by kyngchaos):

 Replying to [comment:11 glynn]:
 > There is no need to modify the existing functions; just check whether
 the calculated value is null.

 Checking if the output is null doesn't tell us if it's null because the
 nearest input cell is null or because it interpolated to null from nulls
 in the matrix.  If the nearest input cell is null, there is no point to
 try simpler interpolations (this mainly affects trying bilinear after
 cubic, for saving any computaion time).

 I can try some speed tests Monday to see if this is worthwhile.

 > Also, it would have been better to attach the changes as a single patch,
 i.e. "svn add" the new files then use "svn diff" to create a patch
 containing all of the changes.

 I could apply them to svn if they're OK.

-- 
Ticket URL: <http://trac.osgeo.org/grass/ticket/465#comment:12>
GRASS GIS <http://grass.osgeo.org>