automatically chunk in groupby binary ops

[dcherian]

What happened?

From https://discourse.pangeo.io/t/xarray-unable-to-allocate-memory-how-to-size-up-problem/3233/4

Consider

# ds is dataset with big dask arrays
mean = ds.groupby("time.day").mean()
mean.to_netcdf()
mean = xr.open_dataset(...)

ds.groupby("time.day") - mean

In GroupBy._binary_op

xarray/xarray/core/groupby.py

Line 616 in 39caafa

expanded = other.sel({name: group})

we will eagerly construct other that is of the same size as ds.

What did you expect to happen?

I think the only solution is to automatically chunk if ds has dask arrays, and other (or mean) isn't backed by dask arrays. A chunk size of 1 seems sensible.

Minimal Complete Verifiable Example

No response

MVCE confirmation

• Minimal example — the example is as focused as reasonably possible to demonstrate the underlying issue in xarray.
• Complete example — the example is self-contained, including all data and the text of any traceback.
• Verifiable example — the example copy & pastes into an IPython prompt or Binder notebook, returning the result.
• New issue — a search of GitHub Issues suggests this is not a duplicate.

Relevant log output

No response

Anything else we need to know?

No response

Environment