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/3322#issuecomment-1101533629,https://api.github.com/repos/pydata/xarray/issues/3322,1101533629,IC_kwDOAMm_X85BqBG9,2448579,2022-04-18T16:14:03Z,2022-04-18T16:14:03Z,MEMBER,Closing in favour of xarray-einstats,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,495799492
https://github.com/pydata/xarray/issues/3322#issuecomment-822124811,https://api.github.com/repos/pydata/xarray/issues/3322,822124811,MDEyOklzc3VlQ29tbWVudDgyMjEyNDgxMQ==,2448579,2021-04-19T02:28:41Z,2021-04-19T02:28:41Z,MEMBER,">  ""xarray linalg library""? 

Thanks @OriolAbril . This sounds like a nice xarray-contrib repo if it's general enough. Note that we do have `xarray.dot` https://xarray.pydata.org/en/stable/generated/xarray.dot.html","{""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,495799492
https://github.com/pydata/xarray/issues/3322#issuecomment-533765137,https://api.github.com/repos/pydata/xarray/issues/3322,533765137,MDEyOklzc3VlQ29tbWVudDUzMzc2NTEzNw==,1217238,2019-09-21T03:59:26Z,2019-09-21T03:59:33Z,MEMBER,"I think the way to write something like an LU decomposition would be to use xarray's `apply_ufunc`, which in theory could wrap `scipy.linalg.lu_factor` in one line.

I don't know if this could/should live in xarray proper. We tend to have pretty high standards, which slows development.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,495799492
https://github.com/pydata/xarray/issues/3322#issuecomment-533248639,https://api.github.com/repos/pydata/xarray/issues/3322,533248639,MDEyOklzc3VlQ29tbWVudDUzMzI0ODYzOQ==,5635139,2019-09-19T18:16:04Z,2019-09-19T18:16:04Z,MEMBER,Somewhat related: https://github.com/pydata/xarray/pull/2766,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,495799492
https://github.com/pydata/xarray/issues/3322#issuecomment-533228830,https://api.github.com/repos/pydata/xarray/issues/3322,533228830,MDEyOklzc3VlQ29tbWVudDUzMzIyODgzMA==,6213168,2019-09-19T17:23:10Z,2019-09-19T17:23:10Z,MEMBER,"Hi @weipeng1999 ,
could you link the reference implementation in numpy/scipy?

I think this would be niche-ish. I would personally try to keep xarray free of functionality that only a tiny fraction of the users actually use - particularly when such functionality can be implemented with a trivial wrapper by the users themselves.
e.g. at the moment we have exactly one scipy function being wrapped, and that's linear interpolation which is useful to a lot of people.

I think this falls into a more general discussion on how niche a function must be in order to be excluded from the library - @shoyer what's your opinion?

Regardless, I would like to point you to <https://xarray-extras.readthedocs.io> which is a module that I created exactly for this kind of cases (PRs are welcome).","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,495799492