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/1471#issuecomment-431051341,https://api.github.com/repos/pydata/xarray/issues/1471,431051341,MDEyOklzc3VlQ29tbWVudDQzMTA1MTM0MQ==,1217238,2018-10-18T15:21:24Z,2018-10-18T15:21:24Z,MEMBER,"I'm marking #1408 as a bug so we won't forget about it. Hopefully it should
be fixed automatically as part of the ""explicit indexes"" refactor.

On Thu, Oct 18, 2018 at 2:48 AM Ondrej Grover <notifications@github.com>
wrote:

> I indeed often resort to using a pandas.MultiIndex, but especially the
> dropping of the selected coordinate value (#1408
> <https://github.com/pydata/xarray/issues/1408>) makes it quite
> inconvenient.
>
> —
> You are receiving this because you commented.
> Reply to this email directly, view it on GitHub
> <https://github.com/pydata/xarray/issues/1471#issuecomment-430946620>, or mute
> the thread
> <https://github.com/notifications/unsubscribe-auth/ABKS1pnDztKWoTaWEjzPpP6orveOMNWRks5umE6BgaJpZM4ORDdd>
> .
>
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,241290234
https://github.com/pydata/xarray/issues/1471#issuecomment-430358013,https://api.github.com/repos/pydata/xarray/issues/1471,430358013,MDEyOklzc3VlQ29tbWVudDQzMDM1ODAxMw==,1217238,2018-10-16T19:00:16Z,2018-10-16T19:00:34Z,MEMBER,"You can use a `pandas.MultiIndex` with xarray. The interface/abstraction could be improved and has some rough edges (e.g., see especially https://github.com/pydata/xarray/issues/1603), but I think this is the preferred way to support these use cases. It does already work for indexing.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,241290234
https://github.com/pydata/xarray/issues/1471#issuecomment-313719395,https://api.github.com/repos/pydata/xarray/issues/1471,313719395,MDEyOklzc3VlQ29tbWVudDMxMzcxOTM5NQ==,1217238,2017-07-07T15:48:05Z,2017-07-07T15:48:05Z,MEMBER,"I'm afraid this isn't possible, by design. Every variable in a Dataset sharing the same coordinate system is enforced as part of the xarray data model. This makes data analysis and comparison with a Dataset quite straightforward, since everything is already on the same grid.

For cases where you need different coordinate values and/or dimension sizes, your options are to either rename dimensions for different variables or use multiple Dataset/DataArray objects (Python has nice built-in data structures).

In theory, we could add something like an ""UnalignedDataset"" that supports most of the Dataset methods without requiring alignment but I'm not sure it's worth the trouble.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,241290234