BUG: agg with axis=1 #19605
BUG: agg with axis=1 #19605
Conversation
…aggregate functions. Provides the user with a correct Exception for this issue.
Codecov Report
@@ Coverage Diff @@
## master #19605 +/- ##
==========================================
- Coverage 91.62% 91.62% -0.01%
==========================================
Files 150 150
Lines 48790 48804 +14
==========================================
+ Hits 44703 44715 +12
- Misses 4087 4089 +2
Continue to review full report at Codecov.
|
|
Hello @spencerogden! Thanks for updating the PR. Cheers ! There are no PEP8 issues in this Pull Request. 🍻 Comment last updated on February 09, 2018 at 20:51 Hours UTC |
| @@ -1012,3 +1012,18 @@ def test_non_callable_aggregates(self): | |||
| expected = df.size | |||
|
|
|||
| assert result == expected | |||
|
|
|||
| def test_frame_row_and_column_aggregates(self): | |||
| df = DataFrame({'B': [4, 5, 6, 7, 8], | |||
There was a problem hiding this comment.
add the issue number as a comment.
| df = DataFrame({'B': [4, 5, 6, 7, 8], | ||
| 'C': [7, 8, 9, 10, 11]}, | ||
| index=[1, 2, 3, 4, 5]) | ||
|
|
There was a problem hiding this comment.
rather than just adding a new test. ideally would like to systematically test all routines.
you can create a fixture
@pytest.fixture(params=[0, 1])
def axis(request):
return request.param
| assert_series_equal(df.agg('min', axis=1), | ||
| Series({1: 4, | ||
| 2: 5, | ||
| 3: 6, |
There was a problem hiding this comment.
also add a test with mixed dtypes (note that they will generally be excluded on the agg so this works ok)
| result, how = self._aggregate(func, axis=0, *args, **kwargs) | ||
| result, how = self._aggregate(func, | ||
| _axis=0, | ||
| *args, **kwargs) |
There was a problem hiding this comment.
though would prefer to pass in axis and have it used down in the stack
There was a problem hiding this comment.
as .apply does generally work with axis=1, it is prob just not passed thru
|
closing in favor of #21224 |
This is a potentially inefficient fix to the this issue which requires a double transpose. I am unsure how inefficient this is, and if this is an acceptable solution. It solves the issue, but perhaps it would be preferable to implement this further down the stack in
_aggregate_multiple_funcs.An additional small fix is to pass
_axisdown the stack.axiswas passed toself._aggregateprevious, but had not effect since it was not referenced.git diff upstream/master -u -- "*.py" | flake8 --diff