Speed up LearnerND

[basnijholt]

(original issue on GitLab)

opened by Jorn Hoofwijk (@Jorn) at 2018-07-08T17:31:01.760Z

To make the code run faster

• cython
• numba

To implement a new algorithm for better performance:

• the _ask function makes a new heap every time it is called, that is wasted time (resolved in gitlab:!88)
• the _ask function recomputes the loss in subtriangulated simplices everytime, should not be needed (resolved in gitlab:!88)

[basnijholt]

originally posted by Joseph Weston (@jbweston) at 2018-07-13T18:20:01.921Z on GitLab

@Jorn do we ever need to call tri.add_point without providing the simplex?

If so I could imagine we could use some spatial tree structure (octree, rtree, mtree) to get fast lookups