issue_comments: 696451333

This data as json

html_url issue_url id node_id user created_at updated_at author_association body reactions performed_via_github_app issue
https://github.com/pydata/xarray/issues/4422#issuecomment-696451333 https://api.github.com/repos/pydata/xarray/issues/4422 696451333 MDEyOklzc3VlQ29tbWVudDY5NjQ1MTMzMw== 6628425 2020-09-22T00:29:32Z 2020-09-22T00:32:04Z MEMBER

Also, would this be solved if using use_cftime=True when reading back the file?

To be honest I'm not sure. You could give it a try -- the only thing there is that you'd need to cast the times back to np.datetime64 if that was ultimately the date type you wanted to go with (instead of a cftime date type).

If I understood correctly the code you quote, in that case you are overwritting the units attribute. Maybe the same could be done in this case.

In the test case I think we're merely acknowledging that the remaining part of the test would fail with these units. I agree with you and @dcherian though -- it's super weird that pandas behaves the way it does here. Considering this isn't the first time these units have come up, it might be worth special-casing them in some way.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   701062999