Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

post 1.11 cleanup tasks #258

Open
2 of 3 tasks
ev-br opened this issue Feb 27, 2025 · 4 comments
Open
2 of 3 tasks

post 1.11 cleanup tasks #258

ev-br opened this issue Feb 27, 2025 · 4 comments

Comments

@ev-br
Copy link
Member

ev-br commented Feb 27, 2025

  • Sort out xfails: added quite a few for 1.11 / spec 2024.12; some of these are actual 2024.12 issues, and some are probably just errors in the test suite
  • drop python 3.9 / numpy 1.21 -- CI: drop python 3.9, numpy 1.21 #266
  • see if conda-forge updates automagically or if it needs a nudge
@crusaderky
Copy link
Contributor

I'm very nervous about these new lines in dask-xfails.txt:

array_api_tests/test_array_object.py::test_setitem
array_api_tests/test_array_object.py::test_getitem_arrays_and_ints_1[1]
array_api_tests/test_array_object.py::test_getitem_arrays_and_ints_1[None]
array_api_tests/test_array_object.py::test_getitem_arrays_and_ints_2[1]
array_api_tests/test_array_object.py::test_getitem_arrays_and_ints_2[None]

as to my understanding they suppress many checks on 2023.12 functionality due to some new edge cases introduced by 2024.12?

@ev-br
Copy link
Member Author

ev-br commented Feb 27, 2025

Getitem tests are new, data-apis/array-api-tests#343
Basically, dask's fancy indexing is limited in many respects.

Setitem test, I've no idea TBH. There were no changes in either -compat or -tests, so is most likely a fluke.

@crusaderky
Copy link
Contributor

setitem fails in dask on an edge case:

>>> import dask.array as da
>>> a = da.zeros((0, 2))
>>> b = da.ones((0, 2))
>>> a[:, :] = b
ValueError: shape mismatch: value array of shape (0, 2) could not be broadcast to indexing result of shape (0, 2)

Obviously the edge case itself is unimportant but the XFAIL neuters the whole test for setitem.

@crusaderky
Copy link
Contributor

XREF dask/dask#11800

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants