BUG: Right result for DatetimeIndex + TimeDelta when timezone is set(… #13935
Conversation
pandas/core/ops.py Outdated
There was a problem hiding this comment.
we should handle is_datetime64_dtype and is_datetime64tz_dtype separately here
There was a problem hiding this comment.
it happens, isn't is?
Current coverage is 85.30% (diff: 100%)@@ master #13935 diff @@ ========================================== Files 139 139 Lines 50157 50161 +4 Methods 0 0 Messages 0 0 Branches 0 0 ========================================== + Hits 42785 42790 +5 + Misses 7372 7371 -1 Partials 0 0
|
pandas/core/ops.py Outdated
There was a problem hiding this comment.
I think this localization can be removed, as adding timedelta is unrelated to the timezone.
| Pls add a test for |
| acutally need to localize if freq < 1D, otherwise can simply add it. E.g. if we are crossing a dst boundary, then need to convert to UTC, operate, then re-convert. So I think we should continue to always convert / operate / re-convert for simplicity. |
| @jreback you're right, but |
conquistador1492 force-pushed the issue_13905 branch from 5150db3 to b17da31 Compare | can you rebase / update? |
| can you rebase / update |
1 similar comment
| can you rebase / update |
| closing, but if you want to update, pls comment |
5 participants
git diff upstream/master | flake8 --diff…#13905)