Hi Matt, No, we don't suggest they should be reversed.
If you look at the union, you could also make the minimal bounding box bigger and it will still include the union, thus we dedine the union to be the smallest of all bounding boxes that contains that blue figure. Likewise, for the intersection, we would use the maximum bounding box that still fully contained inside that blue area. Idan. On Wed, 19 Oct 2022, 23:00 Matt.Wilkie, <matt.wil...@yukon.ca> wrote: > *The definitions are correct. For the union you want a rectangle (or > polygon) that's larger than the input, indeed. But you want the smallest of > those rectangles, because there's an infinity of them. The larger ones > would cover an arbitrary area of the plane (or the globe, if you want).* > > > > *Same for the intersection. You want the largest polygon equal to or > contained in the inputs. Otherwise a smaller answer could just as well be a > single point (or empty), even in nondegenerate cases.* > > > > Please verify I’m parsing these words the way you intend. To me it sounds > like you’re saying the labels for these figures should be reversed? (for > gdal) > > > > > > > > -Matt > _______________________________________________ > gdal-dev mailing list > gdal-dev@lists.osgeo.org > https://lists.osgeo.org/mailman/listinfo/gdal-dev >
_______________________________________________ gdal-dev mailing list gdal-dev@lists.osgeo.org https://lists.osgeo.org/mailman/listinfo/gdal-dev
