BUG: fix reducing numpy ufuncs for Series/Index #23314
Conversation
|
Hello @jorisvandenbossche! Thanks for submitting the PR.
|
| @@ -833,6 +833,15 @@ def test_ufunc_coercions(self, holder): | |||
| exp = tm.box_expected(exp, box) | |||
| tm.assert_equal(result, exp) | |||
|
|
|||
| @pytest.mark.parametrize('holder', [pd.Int64Index, pd.UInt64Index, | |||
There was a problem hiding this comment.
The tests above also use holder, so I wanted to be consistent, but can change them all
There was a problem hiding this comment.
yes, we have settled on box a while back
| def test_ufunc_reduce(self, holder): | ||
| idx = holder([1, 2, 3, 4, 5], name='x') | ||
|
|
||
| result = np.add.reduce(idx) |
There was a problem hiding this comment.
use paramterize over some more ops here
There was a problem hiding this comment.
We would only be testing that those work on the underlying numpy array (which they do :)), so I don't think it is that important
There was a problem hiding this comment.
well can't hurt to have a few more then
There was a problem hiding this comment.
What might be more useful is with other result types, like np.logical_and.reduce
jreback
added
Bug
Numeric Operations
Codecov Report
@@ Coverage Diff @@
## master #23314 +/- ##
==========================================
+ Coverage 92.22% 92.22% +<.01%
==========================================
Files 169 169
Lines 51258 51263 +5
==========================================
+ Hits 47274 47279 +5
Misses 3984 3984
Continue to review full report at Codecov.
|
|
needs a rebase |
|
@jorisvandenbossche can you merge master and update |
|
can you merge master (and move notes to 0.25) / or close if going to handle in |
closes #23312