BUG: Fixed timedelta numeric operations #18892
Conversation
TomAugspurger
added
Numeric Operations
| @@ -888,6 +888,19 @@ def test_nanstd_roundoff(self): | |||
| result = data.std(ddof=ddof) | |||
| assert result == 0.0 | |||
|
|
|||
| def test_nanvar_timedelta(self): | |||
There was a problem hiding this comment.
I know this should be done up in TestnanopsDataFrame but I cannot figure out what's going on in those tests. test_nanstd and test_nanvar both have allow_tdelta=True, but were not catching the bug in #18880
|
Doesn't seem like this is going to work without causing overflows everywhere... |
| @@ -338,7 +338,8 @@ Reshaping | |||
| Numeric | |||
| ^^^^^^^ | |||
|
|
|||
| - | |||
| - Fixed ``std`` and ``var`` computations for timedelta arrays not returning results in timedelta units (:issue:`18880`) | |||
There was a problem hiding this comment.
this is a fix for 0.22? or for 0.23? (IOW do you need to have this for nansum)?
There was a problem hiding this comment.
This was was specific to nanstd, so I think I can work around it. The skipna stuff affected all the ops I think. I'll see how hard that is to work around.
There was a problem hiding this comment.
yeah rather leave this alone for now (IOW do it in 0.23).
There was a problem hiding this comment.
rebase on master. can you move to 0.23 (docs were renamed), prob easiest to just check this file from master and past in new one
| @@ -338,7 +338,8 @@ Reshaping | |||
| Numeric | |||
| ^^^^^^^ | |||
|
|
|||
| - | |||
| - Fixed ``std`` and ``var`` computations for timedelta arrays not returning results in timedelta units (:issue:`18880`) | |||
There was a problem hiding this comment.
rebase on master. can you move to 0.23 (docs were renamed), prob easiest to just check this file from master and past in new one
|
can you move the release note and update |
|
can you rebase |
|
needs a rebase |
Closes #18880