Opened 10 years ago
Last modified 5 years ago
#2670 new defect
v.voronoi skeleton: nonsense result with many adjacent polygons
| Reported by: | mlennert | Owned by: | |
|---|---|---|---|
| Priority: | normal | Milestone: | 7.8.3 |
| Component: | Default | Version: | svn-trunk |
| Keywords: | Cc: | ||
| CPU: | Unspecified | Platform: | Unspecified |
Description
v.voronoi -s census_wake2000 out=census_skel
The output is nonsense (see attached image).
I can get more sensible results (see second image) with:
for cat in $(v.category census_wake2000 op=print); do v.extract census_wake2000 cat=$cat out=census$cat; g.region vect=census$cat; v.voronoi -s census$cat out=skel$cat; done; v.patch $(g.list vect pat="skel*" sep=,) out=census_skel_patched --o
How difficult would it be to integrate such a loop into v.voronoi to ensure more useful centerlines ?
Attachments (5)
Change History (20)
by , 10 years ago
| Attachment: | v_voronoi_allpolystogether.png added |
|---|
comment:2 by , 9 years ago
I can second that.
I tried v.voronoi -s on both a set of isolated lake and valley polygons. There using a loop helped, but results were not very promising never the less, esp. for relatively complex polygons with many vertices. See e.g. how the centerline is routed around islands...
So I tried to generalize the polygons in advance. That improved the results in areas where vertex density was high but led to unuseful results where vertex density was low (within one polygon)...
by , 9 years ago
| Attachment: | lake_without_loop.png added |
|---|
Lake skeleton produced by v.voronoi -s on the entire set of (a little more than 1350) lakes...
by , 9 years ago
| Attachment: | lake_in_loop_non_generalized.png added |
|---|
Lake skeleton produced by v.voronoi on a single, relatively complex lake...
by , 9 years ago
| Attachment: | lake_in_loop_generalized.png added |
|---|
Lake skeleton produced by v.voronoi -s on a single lake which was generalized with douglas_reduction in advance...
comment:6 by , 9 years ago
| Milestone: | 7.0.4 → 7.0.5 |
|---|
comment:7 by , 8 years ago
| Milestone: | 7.0.5 → 7.0.6 |
|---|
comment:8 by , 7 years ago
| Milestone: | 7.0.6 → 7.0.7 |
|---|
comment:9 by , 6 years ago
With GRASS 7.6.0 I am getting
v.voronoi -s census_wake2000 out=census_skel
Reading features...
ERROR: Found 0 vertices in <census_wake2000@PERMANENT>, but at least 2 are
needed
?
follow-up: 11 comment:10 by , 6 years ago
| Milestone: | 7.0.7 → 7.8.0 |
|---|
Moving this on to milestone 7.8, but not sure if this is really fixable.
MarkusM, should this be a no fix ?
comment:11 by , 6 years ago
Replying to martinl:
With GRASS 7.6.0 I am getting
v.voronoi -s census_wake2000 out=census_skel Reading features... ERROR: Found 0 vertices in <census_wake2000@PERMANENT>, but at least 2 are needed?
From the manual: "The bounds of the output map are limited by the current region (see g.region)."
Then you get no error but nonsense results.
Replying to mlennert:
Moving this on to milestone 7.8, but not sure if this is really fixable.
MarkusM, should this be a no fix ?
Not sure, v.voronoi generally suffers from numerical instability. Previously, I could improve it a bit, but it's not perfect, but I have one more idea (maybe in the next few weeks I find time to try this idea: scale coordinates to region center).
Furtheron, something seems to be conceptually wrong with the skeletons for adjacent areas, that should be fixable.
comment:15 by , 5 years ago
| Milestone: | → 7.8.3 |
|---|
all polygons in one go