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/pull/2344#issuecomment-410751334,https://api.github.com/repos/pydata/xarray/issues/2344,410751334,MDEyOklzc3VlQ29tbWVudDQxMDc1MTMzNA==,1217238,2018-08-06T15:37:53Z,2018-08-06T15:37:53Z,MEMBER,"If you really want to save the data / metadata separately for a DataArray, you might as well roll your own version of `to_dict`/`from_dict`. The code for these methods is remarkably simple, all using public APIs: 
https://github.com/pydata/xarray/blob/ded0a684136540962bcc409e6272b1cebb5af30a/xarray/core/dataarray.py#L1662-L1674","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,347712372
https://github.com/pydata/xarray/pull/2344#issuecomment-410564717,https://api.github.com/repos/pydata/xarray/issues/2344,410564717,MDEyOklzc3VlQ29tbWVudDQxMDU2NDcxNw==,1217238,2018-08-06T01:26:01Z,2018-08-06T01:26:10Z,MEMBER,"I'm not sure we want to add this special case.

Logically, if a iterating over a Dataset only includes `data_vars` (as of xarray v0.11), then it would seem unexpected to use `variables` instead when a Dataset is passed as coords into the DataArray constructor.

If you want to copy coordinates `coords` or `variables`, you can do that explicitly, e.g., `DataArray(array, dims=dims, coords=dataset.coords)`.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,347712372