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
We've noticed overlaps is returning the incorrect answer when one of our spherical polygons has an edge that lies on one of the poles. Minimum reproducing example:
Based on my understanding of overlaps it should produce the same output as intersects for these cases. A colleague has also reported strange behaviour from intersection when dealing with this "bigTriangle", which I could try and look into as well if that's useful---I wouldn't be surprised if other functions have issues as well and would be happy to test them if requested.
For reference, "bigTriangle" is one face of an icosahedron. A similar issues happens with one of the faces at the south pole, but interestingly, the mirror of "bigTriangle" that also has an edge on the pole doesn't exhibit these issues.
I believe the edge crossing the pole is the issue because if you slightly change "bigTriangle" (for example, making the first vertex = {31.71747441146, 58.282525588538995}) it produces the expected output.
The text was updated successfully, but these errors were encountered:
A colleague suggested we change the longitude domain from [-pi, pi] to [0, 2pi]. Doing this seems to fix the issue (at least in this example). Is this the expected domain for spherical_equatorial?
Doing this changed which triangle exhibits the problem; the issue now occurs with the face of the icosahedron sharing the pole edge with the one in the example. I'll get an updated minimum reproducing example soon.
We've noticed overlaps is returning the incorrect answer when one of our spherical polygons has an edge that lies on one of the poles. Minimum reproducing example:
And the output I get when using Boost 1.83
Based on my understanding of overlaps it should produce the same output as intersects for these cases. A colleague has also reported strange behaviour from
intersection
when dealing with this "bigTriangle", which I could try and look into as well if that's useful---I wouldn't be surprised if other functions have issues as well and would be happy to test them if requested.For reference, "bigTriangle" is one face of an icosahedron. A similar issues happens with one of the faces at the south pole, but interestingly, the mirror of "bigTriangle" that also has an edge on the pole doesn't exhibit these issues.
I believe the edge crossing the pole is the issue because if you slightly change "bigTriangle" (for example, making the first vertex =
{31.71747441146, 58.282525588538995}
) it produces the expected output.The text was updated successfully, but these errors were encountered: