API: Always return DataFrame from get_dummies #24284
Conversation
TomAugspurger
added
Reshaping
|
Hello @TomAugspurger! Thanks for submitting the PR.
|
Codecov Report
@@ Coverage Diff @@
## master #24284 +/- ##
==========================================
- Coverage 92.22% 92.22% -0.01%
==========================================
Files 162 162
Lines 51828 51821 -7
==========================================
- Hits 47799 47791 -8
- Misses 4029 4030 +1
Continue to review full report at Codecov.
|
Codecov Report
@@ Coverage Diff @@
## master #24284 +/- ##
==========================================
- Coverage 92.28% 92.28% -0.01%
==========================================
Files 162 162
Lines 51830 51827 -3
==========================================
- Hits 47831 47827 -4
- Misses 3999 4000 +1
Continue to review full report at Codecov.
|
doc/source/whatsnew/v0.24.0.rst
Outdated
| is not dummy encoded. When just ``["B", "C"]`` are passed to ``get_dummies``, | ||
| then all the columns are dummy-encoded, and a :class:`SparseDataFrame` was returned. | ||
|
|
||
| .. ipython:: python |
pandas/core/reshape/reshape.py
Outdated
|
|
||
| # if all NaN | ||
| if not dummy_na and len(levels) == 0: | ||
| return get_empty_Frame(data, sparse) | ||
| return get_empty_Frame(data) |
There was a problem hiding this comment.
did you mean to capitalize? (or was that before)
|
does that specific issues that it closes? |
|
|
||
| .. note:: | ||
|
|
||
| There's no difference in memory usage between a :class:`SparseDataFrame` |
There was a problem hiding this comment.
you might need to update the existing docs slightly and/or change the usage in previous whatsnew notes.
pandas/core/reshape/reshape.py
Outdated
| @@ -865,19 +863,16 @@ def _get_dummies_1d(data, prefix, prefix_sep='_', dummy_na=False, | |||
| if is_object_dtype(dtype): | |||
| raise ValueError("dtype=object is not a valid dtype for get_dummies") | |||
|
|
|||
| def get_empty_Frame(data, sparse): | |||
| def get_empty_Frame(data): | |||
There was a problem hiding this comment.
lowercase this, maybe make a module level function.
|
No specific issue. |
|
thanks ! |
xref #19239
In preparation for hopefully deprecating SparseDataFrame / SparseSeries. Right now, I've just made this an API breaking change.
If we want to do this via deprecation, we'd need a new keyword argument to control the result type. I opted against that, because to get a warning-free
get_dummiespeople would needIMO, that's too big a burden for the common case, but would be curious to hear others' thoughts here.