Skip to content

Investigate impact of rejecting sparse inputs in get_namespace #29470

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

Closed
wants to merge 3 commits into from
Closed

Investigate impact of rejecting sparse inputs in get_namespace #29470

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6138a92
FIX make check_array accept sparse inputs when array api dispatch is …
ogrisel Jul 11, 2024
39760c8
Add a changelog entry
ogrisel Jul 11, 2024
cf21705
Investigate impact of rejecting sparse inputs in get_namespace
ogrisel Jul 11, 2024
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
4 changes: 4 additions & 0 deletions doc/whats_new/v1.6.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -56,6 +56,10 @@ See :ref:`array_api` for more details.
compatible inputs when their base estimators do. :pr:`27096` by :user:`Tim
Head <betatim>` and :user:`Olivier Grisel <ogrisel>`.

- |Fix| func:`validation.check_array` now accepts scipy sparse inputs without error
even when array API dispatch is enabled.
:pr:`29469` by :user:`Olivier Grisel <ogrisel>`.

Metadata Routing
----------------

Expand Down
8 changes: 8 additions & 0 deletions sklearn/utils/_array_api.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@
from functools import wraps

import numpy
import scipy.sparse as sp
import scipy.special as special

from .._config import get_config
Expand Down Expand Up @@ -528,6 +529,13 @@ def get_namespace(*arrays, remove_none=True, remove_types=(str,), xp=None):
True if the arrays are containers that implement the Array API spec.
Always False when array_api_dispatch=False.
"""
if any(sp.issparse(a) for a in arrays):
# Consistently reject scipy sparse arrays, whether or not array_api_dispatch
# is enabled.
raise ValueError(
"Scipy sparse arrays or matrices are not supported in get_namespace."
)

array_api_dispatch = get_config()["array_api_dispatch"]
if not array_api_dispatch:
if xp is not None:
Expand Down
10 changes: 10 additions & 0 deletions sklearn/utils/tests/test_validation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2018,6 +2018,16 @@ def test_check_array_array_api_has_non_finite(array_namespace):
check_array(X_inf)


@skip_if_array_api_compat_not_configured
def test_check_array_on_sparse_inputs_with_array_api_enabled():
X_sp = sp.csr_array([[0, 1, 0], [1, 0, 1]])
with config_context(array_api_dispatch=True):
assert sp.issparse(check_array(X_sp, accept_sparse=True))

with pytest.raises(TypeError):
check_array(X_sp)


@pytest.mark.parametrize(
"extension_dtype, regular_dtype",
[
Expand Down
9 changes: 6 additions & 3 deletions sklearn/utils/validation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -838,7 +838,10 @@ def check_array(
"https://numpy.org/doc/stable/reference/generated/numpy.matrix.html"
)

xp, is_array_api_compliant = get_namespace(array)
if sp.issparse(array):
xp, is_array_api_compliant = None, False
else:
xp, is_array_api_compliant = get_namespace(array)

# store reference to original array to check if copy is needed when
# function returns
Expand Down Expand Up @@ -931,7 +934,7 @@ def is_sparse(dtype):
)
)

if dtype is not None and _is_numpy_namespace(xp):
if dtype is not None and xp is not None and _is_numpy_namespace(xp):
# convert to dtype object to conform to Array API to be use `xp.isdtype` later
dtype = np.dtype(dtype)

Expand Down Expand Up @@ -1667,7 +1670,6 @@ def check_non_negative(X, whom):
whom : str
Who passed X to this function.
"""
xp, _ = get_namespace(X)
# avoid X.min() on sparse matrix since it also sorts the indices
if sp.issparse(X):
if X.format in ["lil", "dok"]:
Expand All @@ -1677,6 +1679,7 @@ def check_non_negative(X, whom):
else:
X_min = X.data.min()
else:
xp, _ = get_namespace(X)
X_min = xp.min(X)

if X_min < 0:
Expand Down
Loading