[GRASS-dev] [grass-code I][335] Wrong output from v.voronoi and v.delaunay

code I item #335, was opened at 2007-03-16 12:48
Status: Open
Priority: 3
Submitted By: Māris Nartišs (marisn)
Assigned to: Nobody (None)
Summary: Wrong output from v.voronoi and v.delaunay
Issue type: module bug
Issue status: None
GRASS version: CVS HEAD
GRASS component: vector
Operating system: Linux
Operating system version: Gentoo ~x86 (32 bit)
GRASS CVS checkout date, if applies (YYMMDD): 070316

Initial Comment:
v.voronoi and v.delaunay seems not to work correctly with large(?) (46647 points) datasets.

Steps to reproduce problem in Spearfish6:
r.contour input=elevation.dem@PERMANENT output=contours minlevel=1000 maxlevel=2000 step=100 cut=10
v.to.points -v input=contours type=line output=cont_point llayer=1
v.delaunay input=cont_point output=cont_delaunay
v.voronoi input=cont_point output=cont_voronoi

Bottom (south) part of output looks OK, but at some point output turns into garbage - overlaping areas without any logic.

Attached gis.m screen dumps showing output overview and some place in more detail.

----------------------------------------------------------------------

You can respond by visiting:
http://wald.intevation.org/tracker/?func=detail&atid=204&aid=335&group_id=21

http://wald.intevation.org/tracker/?func=detail&atid=204&aid=335&group_id=21

----------------------------------------------------------------------
code I item #335

..

Summary: Wrong output from v.voronoi and v.delaunay
Issue type: module bug

..

Initial Comment:
v.voronoi and v.delaunay seems not to work correctly with large(?)
(46647 points) datasets.

Steps to reproduce problem in Spearfish6:
r.contour input=elevation.dem@PERMANENT output=contours minlevel=1000
  maxlevel=2000 step=100 cut=10
v.to.points -v input=contours type=line output=cont_point llayer=1
v.delaunay input=cont_point output=cont_delaunay
v.voronoi input=cont_point output=cont_voronoi

Bottom (south) part of output looks OK, but at some point output turns
into garbage - overlaping areas without any logic.

Unrelated to the validity of the bug, but what's the reason for using
v.voronoi with the output of "v.to.points type=line"? Is it a "real
world" problem or a test case? (help with setting the bug's priority)

are you trying to make a TIN?

#spearfish
g.region vect=roads
v.random out=rand_150k n=150000
v.voronoi in=rand_150k out=rand_150k_Vor
..
Number of nodes : 450002
Number of primitives: 600001
Number of points : 0
Number of lines : 0
Number of boundaries: 450001
Number of centroids : 150000
Number of areas : 150000
Number of isles : 1

works beautifully for me.

(what's the 1 island? outer box?)

Hamish