You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@JeroenVerstraelen do we understand why those keys are there in the first place?
Is this the issue that evoland user reported? Do we need to get this through review sufficiently fast?
@jdries It looks like the multiPolygonBuffered variable causes this.
In the screenshot you can see that the layout is constructed using the original multipolygons (passed to datacube_seq as 'polygons'). However, the requiredSpatialKeysForFeatures are constructed using multiPolygonBuffered. This buffered polygon goes outside of the layoutdefinition's extent so it creates both negative and too large spatial keys.
Description
In
org.openeo.geotrellissentinelhub.PyramidFactory
SpatialKeys that go outside of the layout definition bounds are not filtered out:Should become:
Minimal PG:
https://gist.github.com/JeroenVerstraelen/0971bffc8936b604dd67d3194a3b276e
Here, the layout definition is 2x1 but SpatialKey(0,1) and SpatialKey(1,1) are included in the requiredKeys. This causes an error in partitioners later down the line that is difficult to understand for users.
The text was updated successfully, but these errors were encountered: