BUG: fix+test quantile with empty DataFrame, closes #23925 #27436
Conversation
pandas/tests/frame/test_quantile.py
Outdated
| tm.assert_series_equal(result, expected) | ||
|
|
||
| result = df.quantile([0.5]) | ||
| expected = pd.DataFrame([], index=[], columns=[0.5]) |
There was a problem hiding this comment.
I don't quite understand this expected. Should index and columns be flipped (and in the implementation too)?
With DataFrame.quantlie([0.5]) the
- index is the
q([0.5]) - columns are the (numeric) columns
There was a problem hiding this comment.
I think you're right, yah. will update
There was a problem hiding this comment.
what do we do in the scalar-q case then?
There was a problem hiding this comment.
I think
In [11]: pd.Series(name=0.5)
Out[11]: Series([], Name: 0.5, dtype: float64)
right? So that it's equivalent to
In [18]: pd.DataFrame({"A": [1]}).quantile().drop("A")
Out[18]: Series([], Name: 0.5, dtype: float64)
gfyoung
added
Algos
| # GH#23925 _get_numeric_data may drop all columns | ||
| df = pd.DataFrame(pd.date_range("1/1/18", periods=5)) | ||
| result = df.quantile(0.5) | ||
| expected = pd.Series([], index=[], name=0.5) |
There was a problem hiding this comment.
can you add a name to the index of the frame which I think should propagate here?
There was a problem hiding this comment.
good point. i think it should be df.columns.name that propagates.
It looks like DataFrame_get_numeric_data isn't retaining the columns names, will open a separate issue
There was a problem hiding this comment.
also I pushed the 1.0 whatsnew if you rebase
doc/source/whatsnew/v1.0.0.rst
Outdated
| @@ -193,6 +193,7 @@ ExtensionArray | |||
| Other | |||
| ^^^^^ | |||
|
|
|||
| - Bug in :meth:`DataFrame.quantile` with zero-column :class:`DataFrame` incorrectly raising (:issue:`23925`) | |||
There was a problem hiding this comment.
move this to numeric & remove the Other section so its not used anymore.
| @@ -467,3 +467,17 @@ def test_quantile_empty(self): | |||
|
|
|||
| # FIXME (gives NaNs instead of NaT in 0.18.1 or 0.19.0) | |||
| # res = df.quantile(0.5, numeric_only=False) | |||
|
|
|||
| def test_quantile_empty_no_columns(self): | |||
There was a problem hiding this comment.
might be worth testing this with timedelta & period data as well (just to confirm that they work), but can do followup.
|
thanks @jbrockmendel |
black pandasgit diff upstream/master -u -- "*.py" | flake8 --diff