I have data in a Shapefile which has genuine cases of overlapping geometries. (As documented in v.in.ogr) the common areas appear with multiple categories.
I need to find these common areas and “dedupe” them.
Currently I export the data out via v.out.ogr and in postgis do select distinct on (st_asbinary(geometrycomlum)) from tablename order by priorityattribute
Is there some way to either dedupe it in grass Gis itself, or have v.out.ogr output some common attribute in the two geometries that represent the common area
clip from v.in.ogr man page describing how the overlapping polygons are handled.
Overlapping polygons
When importing overlapping polygons, the overlapping parts will become new areas with multiple categories, one unique category for each original polygon. An original polygon will thus be converted to multiple areas with the same shared category. These multiple areas will therefore also link to the same entry in the attribute table. A single category value may thus refer to multiple non-overlapping areas which together represent the original polygon overlapping with another polygon. The original polygon can be recovered by using v.extract with the desired category value or where statement and the -d flag to dissolve common boundaries
The discussion continued in the thread you linked (see next messages in the thread). I used the -c flag of v.in.ogr to avoid cleaning polygons when importing the shapefile as I needed the simple features to extract zonal stats from raster maps afterwards. Would that be useful to you too in this case?
I have data in a Shapefile which has genuine cases of overlapping geometries. (As documented in v.in.ogr) the common areas appear with multiple categories.
I need to find these common areas and “dedupe” them.
Currently I export the data out via v.out.ogr and in postgis do select distinct on (st_asbinary(geometrycomlum)) from tablename order by priorityattribute
Is there some way to either dedupe it in grass Gis itself, or have v.out.ogr output some common attribute in the two geometries that represent the common area
clip from v.in.ogr man page describing how the overlapping polygons are handled.
Overlapping polygons
When importing overlapping polygons, the overlapping parts will become new areas with multiple categories, one unique category for each original polygon. An original polygon will thus be converted to multiple areas with the same shared category. These multiple areas will therefore also link to the same entry in the attribute table. A single category value may thus refer to multiple non-overlapping areas which together represent the original polygon overlapping with another polygon. The original polygon can be recovered by using v.extract with the desired category value or where statement and the -d flag to dissolve common boundaries
I’m not sure I understand what you mean, but you can see what 2 or n categories an overlapping polygon has by displaying categories and centroids from layer 2 in d.vect. Layer 2 is usually where GRASS will store these topology errors after import.
Hi Veronica,
My understanding (I am not advanced user) that -c will also prevent any other cleaning of polygons imported from shapefile.
There are some small errors in the data that are corrected by v.in.ogr. These are the overlaps well above the “snap” threshold I use to import data.
(I may be completely wrong here and would love to correct my understanding).
I think that the two overlapping polygons are represented in Grass GIS using a single area with the unique centroid (and multiple category ids).
I was hoping that grass has a way of associating and exposing to user a unique “centroid id” for each centroid (So I know that two overlapping parts are from same area).
thanks,
Hemant Bist.
The discussion continued in the thread you linked (see next messages in the thread). I used the -c flag of v.in.ogr to avoid cleaning polygons when importing the shapefile as I needed the simple features to extract zonal stats from raster maps afterwards. Would that be useful to you too in this case?
I have data in a Shapefile which has genuine cases of overlapping geometries. (As documented in v.in.ogr) the common areas appear with multiple categories.
I need to find these common areas and “dedupe” them.
Currently I export the data out via v.out.ogr and in postgis do select distinct on (st_asbinary(geometrycomlum)) from tablename order by priorityattribute
Is there some way to either dedupe it in grass Gis itself, or have v.out.ogr output some common attribute in the two geometries that represent the common area
clip from v.in.ogr man page describing how the overlapping polygons are handled.
Overlapping polygons
When importing overlapping polygons, the overlapping parts will become new areas with multiple categories, one unique category for each original polygon. An original polygon will thus be converted to multiple areas with the same shared category. These multiple areas will therefore also link to the same entry in the attribute table. A single category value may thus refer to multiple non-overlapping areas which together represent the original polygon overlapping with another polygon. The original polygon can be recovered by using v.extract with the desired category value or where statement and the -d flag to dissolve common boundaries
Hi Veronica,
Here is what I meant by the problem I get with using -c. ( Without -c and a snap value v.in.ogr only has duplicates and no errors)
But with -c option there are lot of topological errors as well (since v.in.ogr does not do cleaning without -c)
All the warnings about topological errors/incorrect boundaries,centroid outside areas, duplicate centroid are fixed without -c option
WARNING: Number of centroids exceeds number of areas: 1057 > 627
WARNING: Number of incorrect boundaries: 979
WARNING: Number of centroids outside area: 181
WARNING: Number of duplicate centroids: 331
WARNING: The output contains topological errors:
Number of incorrect boundaries: 979
Number of centroids outside area: 181
Number of duplicate centroids: 331
The input could be cleaned by snapping vertices to each other.
Estimated range of snapping threshold: [1e-08, 10]
Try to import again, snapping with 5: ‘snap=5’
Execution of <v.in.ogr -c input=/tmp/chkgrass/input.shp snap=0.5 output=outtab> finished.
Cleaning up default sqlite database …
I’m not sure I understand what you mean, but you can see what 2 or n categories an overlapping polygon has by displaying categories and centroids from layer 2 in d.vect. Layer 2 is usually where GRASS will store these topology errors after import.
Hi Veronica,
My understanding (I am not advanced user) that -c will also prevent any other cleaning of polygons imported from shapefile.
There are some small errors in the data that are corrected by v.in.ogr. These are the overlaps well above the “snap” threshold I use to import data.
(I may be completely wrong here and would love to correct my understanding).
I think that the two overlapping polygons are represented in Grass GIS using a single area with the unique centroid (and multiple category ids).
I was hoping that grass has a way of associating and exposing to user a unique “centroid id” for each centroid (So I know that two overlapping parts are from same area).
thanks,
Hemant Bist.
The discussion continued in the thread you linked (see next messages in the thread). I used the -c flag of v.in.ogr to avoid cleaning polygons when importing the shapefile as I needed the simple features to extract zonal stats from raster maps afterwards. Would that be useful to you too in this case?
I have data in a Shapefile which has genuine cases of overlapping geometries. (As documented in v.in.ogr) the common areas appear with multiple categories.
I need to find these common areas and “dedupe” them.
Currently I export the data out via v.out.ogr and in postgis do select distinct on (st_asbinary(geometrycomlum)) from tablename order by priorityattribute
Is there some way to either dedupe it in grass Gis itself, or have v.out.ogr output some common attribute in the two geometries that represent the common area
clip from v.in.ogr man page describing how the overlapping polygons are handled.
Overlapping polygons
When importing overlapping polygons, the overlapping parts will become new areas with multiple categories, one unique category for each original polygon. An original polygon will thus be converted to multiple areas with the same shared category. These multiple areas will therefore also link to the same entry in the attribute table. A single category value may thus refer to multiple non-overlapping areas which together represent the original polygon overlapping with another polygon. The original polygon can be recovered by using v.extract with the desired category value or where statement and the -d flag to dissolve common boundaries
