Skip to content

BUG/TST: Precision on downcast #5363

Closed
Closed
@TomAugspurger

Description

@TomAugspurger

See #5174

May not be just Macs. From that thread, yarikoptic's OS was: OS: Linux 3.9-1-amd64 #1 SMP Debian 3.9.8-1 x86_64.

A sample failure is from the generic.interpolate suite (modified to downcast, which is the default):

    def test_interp_quad(self):
        sq = Series([1, 4, np.nan, 16], index=[1, 2, 3, 4])
        result = sq.interpolate(method='quadratic')
        expected = Series([1., 4, 9, 16], index=[1, 2, 3, 4])
        assert_series_equal(result, expected)

On (some?) macs and apparently some linux builds this will fail because it the result is not downcast to an int. See also stepping through here and the PR to patch this failure here.

I'm going to see if I can get my failing scipy/numpy tests to pass cleanly. Hopefully that will fix these.

Metadata

Metadata

Assignees

No one assigned

    Labels

    Dtype ConversionsUnexpected or buggy dtype conversionsNumeric OperationsArithmetic, Comparison, and Logical operations

    Type

    No type

    Projects

    No projects

    Milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions