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.

Sign up for GitHub

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

ENH: correct Dask capabilities #290

Merged
merged 1 commit into from
Mar 26, 2025
Merged

ENH: correct Dask capabilities #290

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 16 additions & 6 deletions array_api_compat/dask/array/_info.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,11 +68,22 @@ def capabilities(self):
The resulting dictionary has the following keys:

- **"boolean indexing"**: boolean indicating whether an array library
supports boolean indexing. Always ``False`` for Dask.
supports boolean indexing.

Dask support boolean indexing as long as both the index
and the indexed arrays have known shapes.
Note however that the output .shape and .size properties
will contain a non-compliant math.nan instead of None.

- **"data-dependent shapes"**: boolean indicating whether an array
library supports data-dependent output shapes. Always ``False`` for
Dask.
library supports data-dependent output shapes.

Dask implements unique_values et.al.
Note however that the output .shape and .size properties
will contain a non-compliant math.nan instead of None.

- **"max dimensions"**: integer indicating the maximum number of
dimensions supported by the array library.

See
https://data-apis.org/array-api/latest/API_specification/generated/array_api.info.capabilities.html
Expand All @@ -99,9 +110,8 @@ def capabilities(self):

"""
return {
"boolean indexing": False,
"data-dependent shapes": False,
# 'max rank' will be part of the 2024.12 standard
"boolean indexing": True,
"data-dependent shapes": True,
"max dimensions": 64,
}

Expand Down
8 changes: 5 additions & 3 deletions dask-xfails.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -28,12 +28,14 @@ array_api_tests/test_has_names.py::test_has_names[array_method-to_device]
array_api_tests/test_has_names.py::test_has_names[array_attribute-device]
array_api_tests/test_has_names.py::test_has_names[array_attribute-mT]

# Fails because shape is NaN since we don't materialize it yet
# Data-dependent output shape
# These tests fail as array-api-tests doesn't cope with unknown shapes
# Also, output shape is (math.nan, ) instead of (None, )
# Also, da.unique() doesn't accept equals_nan which causes non-compliant
# output when there are NaNs in the input.
array_api_tests/test_searching_functions.py::test_nonzero
array_api_tests/test_set_functions.py::test_unique_all
array_api_tests/test_set_functions.py::test_unique_counts

# Different error but same cause as above, we're just trying to do ndindex on nan shape
array_api_tests/test_set_functions.py::test_unique_inverse
array_api_tests/test_set_functions.py::test_unique_values

Expand Down
Loading