PERF/COMPAT: define platform int to np.intp by chris-b1 · Pull Request #13972 · pandas-dev/pandas (original) (raw)

This currently fails on 32-bit linux (before your PR). I recall had to do something on windows to get this to pass. Something wrong somewhere. Your PR fixes this! yeah!

======================================================================
FAIL: test_alignment_non_pandas (pandas.tests.frame.test_operators.TestDataFrameOperators)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/jreback/pandas/pandas/tests/frame/test_operators.py", line 1210, in test_alignment_non_pandas
    Series([1, 2, 3], index=df.index))
  File "/home/jreback/pandas/pandas/util/testing.py", line 1154, in assert_series_equal
    assert_attr_equal('dtype', left, right)
  File "/home/jreback/pandas/pandas/util/testing.py", line 878, in assert_attr_equal
    left_attr, right_attr)
  File "/home/jreback/pandas/pandas/util/testing.py", line 1018, in raise_assert_detail
    raise AssertionError(msg)
AssertionError: Attributes are different

Attribute "dtype" are different
[left]:  int32
[right]: int64

So as-is this passes on linux32 (and fixes the above).

lmk when good to go.