Justus Magin

would that just delegate to the underlying index, or also wrap it (probably the former, but I wanted to make sure)? In any case, I wonder if we should just...

> Isn't it addressed by `pint.Quantity.to_base_unit`? No, because `to_base_unit` will *convert* the ~index~ index's data to the base units (i.e. combinations of the base units `m`, `s`, `kg`, ...). Instead,...

@benbovy, something I just noticed: where do we need `PintIndex.from_variables`? As far as I understand it, this is used for `set_xindex`, but since `PintIndex` is a meta index (or rather,...

we also need to figure out how to handle `set_xindex` on top of a quantified dataset (so a `PintIndex` would be overwritten – if that makes sense?)

> (or an object of the default index type used by Xarray) is there a way to access this default index type? Otherwise it's probably fine to keep hard-coding `PandasIndex`...

now that I have more time to work on this, I think the only things left to fix right now is: - [ ] figure out what to do with...

we're already using some part of the private API of `xarray` (`Coordinates._construct_direct`), so I would be fine to use `Coordinates` either way. However, from my experiments it appears just passing...

`Coordinates._construct_direct` works, but `Coordinates(coords=coords, indexes={})` still tries to create default indexes. Not sure why. Edit: It appears `as_variable` converts to an index variable.

a couple more notes: - the `diff` called by `assert_identical` appears to "work" just by changing `a.indexes` to `a.xindexes` (it doesn't raise, at least). However, `Dataset.equals` and `Dataset.identical` don't compare...

looks like @TomNicholas' efforts to allow creating `xarray` objects from `Coordinates` objects without indexes fixed a lot of the issues I had here, as well. One of the few things...