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/7868#issuecomment-1561584592,https://api.github.com/repos/pydata/xarray/issues/7868,1561584592,IC_kwDOAMm_X85dE-PQ,5821660,2023-05-24T16:50:34Z,2023-05-24T16:50:34Z,MEMBER,"Thanks @ghiggi for your comment.
The problem is we have at least two contradicting user requests here, see #7328 and #7862.
I'm sure there is a solution to accommodate both sides. ","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,1722417436
https://github.com/pydata/xarray/issues/7868#issuecomment-1561358915,https://api.github.com/repos/pydata/xarray/issues/7868,1561358915,IC_kwDOAMm_X85dEHJD,19285200,2023-05-24T15:20:00Z,2023-05-24T15:20:00Z,NONE,"Dask array with dtype `object` can contain whatever python object (i.e. I saw examples of geometry and matplotlib collections within dask arrays with `object` dtype). As a consequence, dask do not try the conversion to i.e. `str` to estimate the array size, since there is no clean way AFAIK to attach an attribute to dtype suggesting that the `object` is actually a `string`.
With your PR, the dtype is not anymore `object` when creating the `dask.array` and this solves the issue I guess.
Did I overlooked something?
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,1722417436
https://github.com/pydata/xarray/issues/7868#issuecomment-1561214028,https://api.github.com/repos/pydata/xarray/issues/7868,1561214028,IC_kwDOAMm_X85dDjxM,5821660,2023-05-24T13:58:16Z,2023-05-24T13:58:16Z,MEMBER,"My main question here is, why is dask not trying to retrieve the object types from dtype.metadata? Or does it and fail for some reason?.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,1722417436
https://github.com/pydata/xarray/issues/7868#issuecomment-1560674198,https://api.github.com/repos/pydata/xarray/issues/7868,1560674198,IC_kwDOAMm_X85dBf-W,5821660,2023-05-24T08:27:11Z,2023-05-24T08:27:11Z,MEMBER,"@ghiggi Glad it works, but we still have to check if that is the correct location for the fix, as it's not CF specific.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,1722417436
https://github.com/pydata/xarray/issues/7868#issuecomment-1560651807,https://api.github.com/repos/pydata/xarray/issues/7868,1560651807,IC_kwDOAMm_X85dBagf,19285200,2023-05-24T08:12:18Z,2023-05-24T08:12:18Z,NONE,"Thanks @kmuehlbauer ! https://github.com/pydata/xarray/pull/7869 solve the issues !
Summarizing:
- With #7869, netCDF4 with `NC_STRING` variable arrays are now read into xarray as `Unicode` dtype (instead of `object`)
- As a consequence `dask` can estimate the array's size and `xr.open_dataset(fpath, chunks=""auto"")` does not raise anymore the `NotImplementedError`.
- `NC_CHAR` variable arrays continue to be read into xarray as fixed-length byte-string dtype. Maybe something more could be done to deserialize also `NC_CHAR` to `Unicode`. However, this might cause some backward incompatibilities and might be better to address this in a separate PR.
Thanks again @kmuehlbauer for having resolved the problem in less than 2 hours :1st_place_medal: ","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,1722417436
https://github.com/pydata/xarray/issues/7868#issuecomment-1559959581,https://api.github.com/repos/pydata/xarray/issues/7868,1559959581,IC_kwDOAMm_X85c-xgd,5821660,2023-05-23T18:42:55Z,2023-05-23T19:01:00Z,MEMBER,@ghiggi Thanks for getting this back into action. I got dragged away from the one string object issue in #7654. I'll split this out and add a PR.,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,1722417436
https://github.com/pydata/xarray/issues/7868#issuecomment-1559973194,https://api.github.com/repos/pydata/xarray/issues/7868,1559973194,IC_kwDOAMm_X85c-01K,5821660,2023-05-23T18:55:46Z,2023-05-23T18:55:46Z,MEMBER,@ghiggi I'd appreciate if you could test your workflows against #7869. Your example and the one over in #7652 are working AFAICT.,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,1722417436