-
-
Notifications
You must be signed in to change notification settings - Fork 1.2k
Closed
Labels
ReleasePlanning and tracking progress of releasesPlanning and tracking progress of releases
Description
What is your issue?
This issue tracks the release for allowing non-nanosecond datetimes/timedeltas (adapting to pandas) behaviour. We aimed for a release in January, so there quite some references to that release in the codebase/docs.
Must have:
- Enable passing a
CFTimedeltaCoder
todecode_timedelta
#9966 - Fix infer_freq, check for subdtype "datetime64"/"timedelta64" #9999
- Revert "Use flox for grouped first, last (#9986)" #10001
- Use
freq="D"
instead offreq="d"
inpd.timedelta_range
#10004 - Doctest failing with numpy 2.1.3 #10005
- remove outdated type-ignores #10006
- cast
numpy
scalars to arrays inNamedArray.from_array
#10008 - tbc
Nice to have:
- Preserve order of variables in combine_by_coords #9070
- Add FAQ answer about API stability & backwards compatibility #9855
- tbc
I'll start preparing release tomorrow. @pydata/xarray please add items as needed. Also raise concerns about the release date, if there are any.
Metadata
Metadata
Assignees
Labels
ReleasePlanning and tracking progress of releasesPlanning and tracking progress of releases