Justus Magin

looks like there's a upcoming change to `scipy.sparse` that causes `flox` to break. Not sure what exactly, but this should be fixed within `flox`. Edit: looks like it's the same...

`zarr` might be fixed by #9673, and the failing `gappy` tests will have to be fixed in `flox` (the issue is a change in upstream-dev `scipy.sparse`). I think since the...

we've got a couple new errors related to `zarr` (first seen two days ago): ``` xarray/tests/test_backends.py::TestZarrDictStore::test_manual_chunk[2]: ValueError: ndarray is not C-contiguous xarray/tests/test_backends.py::TestZarrDictStore::test_warning_on_bad_chunks[2]: Failed: DID NOT WARN. No warnings of type...

The problem with dropping the index is that it will require dequantifying the indexed coordinates, otherwise recreating the index would drop the units entirely. The only ways I can see...

> Notice that `...` is the only way, I think, to get data from a scalar. you can use `arr.item()` if you know that you have a scalar / 0d...

since I just saw this: you might be able to use ```python if isinstance(self.da.data, dask_array_type): data = self.da.data.map_blocks(cp.asarray) return self.da.copy(data=data) return self.da.copy(data=cp.asarray(self.da.data)) ```

I just saw that this is also mentioned in the [review comments](https://github.com/xarray-contrib/xarray-tutorial/pull/291#discussion_r1697683957) of the PR that added this section. Feel free to close this PR if you'd like to address...

all I'd really need is some way to decide whether a directory named ".parquet" is a "kerchunk reference file". If we can rely on `.zmetadata` to exist, that in combination...

the reason `itertools.groupby` needs sorted iterables is that it combines groups locally, not globally. So `list(map(list, itertools.groupby([1, 1, 2, 1, 3], key=lambda x: x)))` would result in `[(1, [1, 1]),...

This also appears with `open_mfdataset`, but appears to only happen with a threading scheduler. In any case, the reason is that `netcdf4` is not thread-safe anymore since `netcdf4=1.6.1` (I think)....