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

BUG/CoW: is_range_indexer can't handle very large arrays #53672

Merged
merged 2 commits into from
Jun 15, 2023
Merged

BUG/CoW: is_range_indexer can't handle very large arrays #53672

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
47c8373
BUG: is_range_indexer can't handle very large arrays
lithomas1 Jun 14, 2023
486b5d8
fix test on 32-bit
lithomas1 Jun 14, 2023
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
2 changes: 1 addition & 1 deletion doc/source/whatsnew/v2.1.0.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -396,7 +396,7 @@ Indexing
^^^^^^^^
- Bug in :meth:`DataFrame.__setitem__` losing dtype when setting a :class:`DataFrame` into duplicated columns (:issue:`53143`)
- Bug in :meth:`DataFrame.__setitem__` with a boolean mask and :meth:`DataFrame.putmask` with mixed non-numeric dtypes and a value other than ``NaN`` incorrectly raising ``TypeError`` (:issue:`53291`)
-
- Bug in indexing methods (e.g. :meth:`DataFrame.__getitem__`) where taking the entire :class:`DataFrame`/:class:`Series` would raise an ``OverflowError`` when Copy on Write was enabled and the length of the array was over the maximum size a 32-bit integer can hold (:issue:`53616`)

Missing
^^^^^^^
Expand Down
2 changes: 1 addition & 1 deletion pandas/_libs/lib.pyx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -668,7 +668,7 @@ ctypedef fused int6432_t:

@cython.wraparound(False)
@cython.boundscheck(False)
def is_range_indexer(ndarray[int6432_t, ndim=1] left, int n) -> bool:
def is_range_indexer(ndarray[int6432_t, ndim=1] left, Py_ssize_t n) -> bool:
"""
Perform an element by element comparison on 1-d integer arrays, meant for indexer
comparisons
Expand Down
13 changes: 13 additions & 0 deletions pandas/tests/libs/test_lib.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,7 @@
lib,
writers as libwriters,
)
from pandas.compat import IS64

from pandas import Index
import pandas._testing as tm
Expand Down Expand Up @@ -248,6 +249,18 @@ def test_is_range_indexer(self, dtype):
left = np.arange(0, 100, dtype=dtype)
assert lib.is_range_indexer(left, 100)

@pytest.mark.skipif(
not IS64,
reason="2**31 is too big for Py_ssize_t on 32-bit. "
"It doesn't matter though since you cannot create an array that long on 32-bit",
)
@pytest.mark.parametrize("dtype", ["int64", "int32"])
def test_is_range_indexer_big_n(self, dtype):
# GH53616
left = np.arange(0, 100, dtype=dtype)

assert not lib.is_range_indexer(left, 2**31)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
assert not lib.is_range_indexer(left, 2**31)
assert not lib.is_range_indexer(left, np.iinfo(np.int32).max)

To fix the failing 32 bit build

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for the catch. I can't change the size here unfortunately (need it to be 2**31 == np.iinfo(np.int32).max + 1 to trigger the OverflowError previously).

Gonna skip on 32-bit instead, since the issue can't happen there.
(Max array len should be less than int32 max on 32-bit platforms).


@pytest.mark.parametrize("dtype", ["int64", "int32"])
def test_is_range_indexer_not_equal(self, dtype):
# GH#50592
Expand Down