You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In short, the broader ecosystem is working to support non-nanosecond datetime objects (see: pydata/xarray#7493) in Pandas 2+ and Xarray's release last month. This topic was brought to my attention by the recent upstream dev issues w/ Xarray (#681).
As I was digging through our climatology code today, I noticed we have some things hard coded that will break (in some cases) with these ecosystem changes.
We'll likely want to pin some dependencies and/or make changes and add tests to support this.
It looks like we already have versions of these packages with support for non-nanosecond datetimes installing in our test environments. Our tests aren't failing, but I suspect that's because we're not testing for this.
_infer_calendar_name and the test suite are probably where changes are needed.
Also worth noting that there's some overlap here with #508.
In short, the broader ecosystem is working to support non-nanosecond datetime objects (see: pydata/xarray#7493) in Pandas 2+ and Xarray's release last month. This topic was brought to my attention by the recent upstream dev issues w/ Xarray (#681).
As I was digging through our climatology code today, I noticed we have some things hard coded that will break (in some cases) with these ecosystem changes.
We'll likely want to pin some dependencies and/or make changes and add tests to support this.
Related:
The text was updated successfully, but these errors were encountered: