Use cumsum from flox

[Illviljan]

• Closes cumsum drops index coordinates #6528
• Tests added
• User visible changes (including notable bug fixes) are documented in whats-new.rst

The non-flox version reduces chunksizes significantly:

x = xr.DataArray([1, 1, 1, 1, 1], name="x").chunk()
grp_idx = xr.DataArray([-1, 0, 0, -1, 1])
with xr.set_options(use_flox=False):
    print(x.groupby(grp_idx).cumsum())
<xarray.DataArray 'x' (dim_0: 5)> Size: 40B
dask.array<getitem, shape=(5,), dtype=int64, chunksize=(2,), chunktype=numpy.ndarray>
Dimensions without coordinates: dim_0

With flox the chunksize is retained:

x = xr.DataArray([1, 1, 1, 1, 1], name="x").chunk()
grp_idx = xr.DataArray([-1, 0, 0, -1, 1])
with xr.set_options(use_flox=True):
    print(x.groupby(grp_idx).cumsum())
<xarray.DataArray 'x' (dim_0: 5)> Size: 40B
dask.array<_finalize_scan, shape=(5,), dtype=int64, chunksize=(5,), chunktype=numpy.ndarray>
Dimensions without coordinates: dim_0

Other changes:

• Changes DataArray.cumsum/Dataset.cumsum/DataTree.cumsum/DataArray.groupby.cumsum/Dataset.groupby.cumsum etc.
• Coordinates are now retained

Notes
groupby_scan was added in: https://github.com/xarray-contrib/flox/releases/tag/v0.9.9
cumsum was added in: https://github.com/xarray-contrib/flox/releases/tag/v0.10.5

[dcherian]

why is this needed

cc @TomNicholas

[Illviljan]

Because of #6528 and #10987 (comment). Since the problem of Datasets dropping the coords is solved in .cumsum in this PR, DataTree must also use ds.cumsum instead of ds.reduce("cumsum") in order to avoid dropping the coordinates.

[dcherian]

let's update the code comment to reflect your comment above

[Illviljan]

@dcherian, this is ready for a another review now. It was only changes in tests since the last time.

[dcherian]

Are you able to address the extra testing requested in #10987 (comment)?

If you're too busy, we can just merge. This is a good improvement.