Hey,
small update on this. The NEP draft has not changed much, but you can
now try the full power of the proposed new indexing types [1]:
* arr.oindex[...] # orthogonal/outer indexing
* arr.vindex[...] # vectorized (like fancy, but different ;))
* arr.lindex[...] # legacy/fancy indexing
Hi all,
at scipy discussing with Nathaniel and others, we thought that maybe we
can push for orthogonal type indexing into numpy. Now with the new
version out and some other discussions done, I thought it is time to
pick it up :).
The basic ideas are twofold. First make indexing easier and less
c
