issues Search Results · repo:scikit-hep/awkward language:Python
Filter by
1k results
(50 ms)1k results
inscikit-hep/awkward (press backspace or delete to remove)Version of Awkward Array
master branch
Description and code to reproduce
tests-cuda/test_2922a_new_cuda_kernels.py::test_2071_unflatten_non_packed_indexed_counts
tests-cuda/test_2922a_new_cuda_kernels.py::test_2071_unflatten_non_packed_counts_indexed_layout ...
bug
ianna
- 6
- Opened yesterday
- #3431
I m pretty sure _all_ of our nplike functions will now need `| VirtualArray` in the input signature. Can you update the nplikes, and the array modules, etc?
Originally posted by @agoose77 in ...
ianna
- Opened yesterday
- #3429
Perhaps this is a good candidate for https://docs.python.org/3/library/functools.html#functools.cached_property
Originally posted by @agoose77 in https://github.com/scikit-hep/awkward/pull/3364#discussion_r2002974502 ...
ikrommyd
- 1
- Opened yesterday
- #3428
Hi awkward. Thanks for this great codebase, and I m sorry if I m asking a question which has been answered elsewhere. I
looked around and couldn t find anything.
Do you understand why array_equal gives ...
bug
ianna
- Opened yesterday
- #3426
Do we assume that the result of each generator invocation is unique (i.e. doesn t need to be copied)? That would be nice, and means that this is correct.
Originally posted by @agoose77 in https://github.com/scikit-hep/awkward/pull/3364#discussion_r1985119884 ...
pfackeldey
- 2
- Opened 2 days ago
- #3422
Why this restriction? We do allow NumPy → JAX, so we should allow Virtual[Numpy
→ JAX, no?
Originally posted by @agoose77 in https://github.com/scikit-hep/awkward/pull/3364#discussion_r1985084963 ...
pfackeldey
- 2
- Opened 2 days ago
- #3421
This is much better. In future, I d like to update `_backends` so that we don t have to hard-code the nplike-to-backend relationship.
Originally posted by @agoose77 in https://github.com/scikit-hep/awkward/pull/3364#discussion_r1985081845 ...
ianna
- Opened 2 days ago
- #3420
In https://github.com/scikit-hep/awkward/pull/3364, we make .raw()/._raw() calls materialize. Some more thinking can be
put into whether this should be the case. Perhaps we need a materializing and a non ...
ikrommyd
- 1
- Opened 13 days ago
- #3414
It seems like we some-where rely on array.ctypes internally. We should try and abstract that to inside the nplike /
kernels.
cleanup
agoose77
- Opened 13 days ago
- #3413
I don t think this is a good idea if we can avoid it — a VirtualArray _is _not_ the same as e.g. a Jax array or CuPy array. We also use this to memoize `type(array)→nplike` which would fail for ...
ianna
- Opened 14 days ago
- #3412

Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip!
Restrict your search to the title by using the in:title qualifier.
Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip!
Restrict your search to the title by using the in:title qualifier.