![]() Requirement can be relaxed - any changes here require corresponding We do not specify a future-compatible path by which this Objects that do not satisfy this cannot override any Numpy ![]() Implemented in a specific way and be compatible with Numpy’s ndarray The _array_ufunc_ as described below requires that anyĬorresponding Python binary operations ( _mul_ et al.) should be Right behaviour, hence the change in name.) An implementation was made, but had not quite the (Note that in earlier iterations, the override was called Specifically addresses how binary operators and ufuncs should interact. The mechanism here follows more closely the way Python enablesĬlasses to override _mul_ and other binary operations. Retro-fit NumPy with multi-methods, which would solve the same This covers some of the same ground as Travis Oliphant’s proposal to On discovery of _array_ufunc_ the ufunc will hand off the Here we propose adding a mechanism to override ufuncs based on the ufuncĬhecking each of it’s arguments for a _array_ufunc_ method. Ndarray using _array_prepare_ and _array_wrap_, and there is little to no support for arbitrary NumPy’s universal functions (ufuncs) currently have some limitedįunctionality for operating on user defined subclasses of
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |