[Qgis-user] How to determine the inner surface on a sphere or ellipsoid

[Nicolas Cadieux]

Hi Nyall,

Just out of curiosity, would a very large polygon slow down a spatial query or would just a complex one? My testing with using shapefiles to reclassify LiDAR data did show that things could be sped up by splitting large polygons with a smaller grid but I had attributed this to the complexity of the polygons and not the size.  Was this a wrong assumptions?

Nicolas Cadieux
Ça va bien aller!

> Le 16 juin 2020 à 18:21, Nyall Dawson <nyall.dawson at gmail.com> a écrit :
> 
> On Wed, 17 Jun 2020 at 02:10, Richard Duivenvoorde <rdmailings at duif.net> wrote:
>> 
>>> On 6/16/20 4:17 PM, Tudorache, Marian wrote:
>>> Hi everyone,
>>> 
>>> I have a list of polygons given by a list of points.
>>> The polygons are properly drawn on QGIS canvas by creating the geometries, the each geometry is used to create a Qgsfeature which are saved on a shapefile.
>>> The problem appears in detecting if a point on the Earth is inside or outside the polygon.
>>> Using pyqgis QgsGeometry intersects function returns proper value.
>>> However when I export the polygon to intermaphics from Kongsberg Geospatial (former Gallium) sometime the point is inside other time is outside the polygon.
>>> I tried to switch the order of the points is QGIS ,but the intersects function always give me the same result regardless of the order of the points.
>>> In intermaphics the intersection between a polygon and a point varies with the order of the points which define a polygon.
>>> I talked to people from gallium and they confirmed the order of the points is important.
>>> In one direction a inner area of the polygon is the small surface and if I switch the order the inner area is the outside and it wraps the Earth on the opposite side.
>>> 
>>> Does QGIS or pyqgis has a similar mechanism to determine which is the inner part of the polygon on a sphere or ellipsoid?
>> 
>> Hi Marian,
>> 
>> are'nt we here talking about the so called Right Hand Rule?
> 
> That's just a convention -- it doesn't change what the boundary
> actually represents.
> 
> If you want to do analysis based on points which fall outside a
> digitized polygon, you should use a "disjoint" relationship. The
> alternative is to do what Nicolas suggested and make a polygon which
> covers the globe minus a small hole, but you'll get terrible
> performance with any analysis using that approach...!
> 
> Nyall
> _______________________________________________
> Qgis-user mailing list
> Qgis-user at lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user