Skip to content

TST: Multiindex slicing with NaNs, unexpected results for #25154 #39356

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

Merged
merged 7 commits into from
Feb 5, 2021
Merged

TST: Multiindex slicing with NaNs, unexpected results for #25154 #39356

Changes from 2 commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
a9a4ff4
TST: Added test case for Multiindex slicing with NaNs issue #25154
theodorju Jan 21, 2021
c0cb975
Merge remote-tracking branch 'upstream/master' into pandas-25154
theodorju Jan 21, 2021
da9a5f0
Removed unecessary definitions.
theodorju Jan 23, 2021
df14980
Addition of test cases.
theodorju Feb 4, 2021
071256c
Modifications in test_frame_getitem_nan_cols_multiindex to use tm.ass…
theodorju Feb 4, 2021
548ca8a
Reverting changes in test_series_getitem_indexing_errors
theodorju Feb 4, 2021
ada8989
Added new line in test_series_getitem_multiindex
theodorju Feb 5, 2021
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
20 changes: 20 additions & 0 deletions pandas/tests/indexing/multiindex/test_getitem.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -229,6 +229,26 @@ def test_frame_getitem_nan_multiindex(nulls_fixture):
tm.assert_frame_equal(result, expected)


def test_frame_getitem_nan_cols_multiindex(nulls_fixture):
Copy link
Contributor

Choose a reason for hiding this comment

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

can you add all of the cases in the original post (working an non-working), there are 5 cases i think. pls parameterize

Copy link
Contributor Author

@theodorju theodorju Feb 4, 2021 •
edited
Loading

Choose a reason for hiding this comment

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

Thank you for the suggestion. The additional cases were added and parameterized.

Keyword argument check_column_type was needed when slicing with (["b"], [np.nan]), in test case 5, the reason is explained below:

Test Case 5:
Slicing out (["b"], [np.nan]):

When asserting the types of the columns of the actual and expected result for the second level of the multiindex, on _check_types of assert_index_equal on asserters.py the left argument that is based on the sliced multiindex (argument left in _check_types) is:

  • left: Index(['bar', 'foo'], dtype='object')

while the right argument that is based on the expected result is:

  • right: Index([], dtype='object')

Based on that the left has "string" as inferred type, while the right one is empty which causes the test to fail, even though the resulting dataframes are identical.

I think this is because when slicing a dataframe with a multi-index the resulting levels of columns are the initial ones present in the original dataframe and are not updated.

In order to avoid that comparison I passed check_column_type=False as keyword argument.

# Slicing MultiIndex including levels with nan values, for more information
# see GH#25154
data = [[1, 2, 3], [4, 5, 6]]
index = ["First", nulls_fixture]
Copy link
Member

Choose a reason for hiding this comment

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

Define the index only once if it is the same. You could remove the other definitions and write the data directly into the object.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Thanks for the comments.
Suggested changes were implemented.

columns = MultiIndex.from_tuples([("a", "foo"), ("b", "foo"), ("b", nulls_fixture)])
df = DataFrame(data=data, columns=columns, index=index, dtype="int64")

# Slicing out 'b', ['foo', nan]
cols = (["b"], ["foo", nulls_fixture])
result = df.loc[:, cols]
expected_columns = MultiIndex.from_tuples([("b", "foo"), ("b", nulls_fixture)])
expected_index = ["First", nulls_fixture]
expected = DataFrame(
[[2, 3], [5, 6]], columns=expected_columns, index=expected_index, dtype="int64"
)

tm.assert_frame_equal(result, expected)


# ----------------------------------------------------------------------------
# test indexing of DataFrame with multi-level Index with duplicates
# ----------------------------------------------------------------------------
Expand Down