issue_comments: 889483293

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/5647#issuecomment-889483293 https://api.github.com/repos/pydata/xarray/issues/5647 889483293 IC_kwDOAMm_X841BHAd 4160723 2021-07-29T21:49:02Z 2021-07-29T21:49:02Z MEMBER

We should raise an error in this cases (and/or suggest setting a MultiIndex). It should also be OK if not every index implements alignment, in which case they should raise an error if coordinates do not match exactly.

Agreed.

I don't think we should try to support alignment with multi-dimensional (non-orthogonal) inside align(). Instead we can just require the coordinates corresponding to such indexes to match exactly (or within a tolerance).

I could see some examples (e.g., union or intersection of staggered grids) where it could still be useful to have a (meta-)index that implements alignment, though.

Actually, while working on #5636 I was thinking more about how to perform alignment based on xarray.Index objects given that the concept of an xarray.Index should be clearly distinct from the concept of a Variable.

We could maybe add an xarray.Index.sizes property similarly to Variable.sizes.

Alternatively, we could allow xarray.Index.union() and xarray.Index.intersection() returning both the joined index and the new Variable objects created from the latter index.

I'd lean towards the 2nd option, which is consistent with the class method constructors added in #5636. Not sure the 1st option is a good idea and it doesn't solve all the issues mentioned in my comment above.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   955936490