Justus Magin

you can only interpolate along a single dimension with that method: `u.interpolate_na(dim="level", method="linear")` should work. If you absolutely need multi-dim missing value interpolation on a geoscience dataset, you might want...

right, that looks like a documentation bug, and we might want to consider making `dim` a required parameter.

we had some more discussion about this in #8051, although these are probably more about arguments explicitly passed on by `xarray` instead of new kwargs for the backend.

I can't reproduce this in my environment, I get the expected masked values (thanks for the code snippet, though). Can you please add the output of `xr.show_versions()` to the original...

looking at the docstring of the chunk manager base class, it appears this is not used for all reductions: https://github.com/pydata/xarray/blob/d26144db962d9d620c7f7c4e7398c2f6b3bf6d47/xarray/namedarray/parallelcompat.py#L386-L390 In other words, this seems to be expected behavior? I...

looks like `iris` depends on `netcdf4`, so we need to merge #9136 to see if anything fails. Edit: the most recent commit doesn't seem to cause any additional failing tests,...

@trexfeathers, looks like there was nothing to do here.

not yet, no. This just makes sure that the formatting of the coordinates representation does not call `.indexes`.

This should be a duplicate of #91 (unless I'm misunderstanding something?). Can you try to reproduce with `xarray.core.datatree.DataTree`? I believe this should have been fixed in pydata/xarray#8930.

be aware that currently this also fails on `main` – a dependency stopped pinning to `numpy