Ouranosinc / xclim / 7048914432

90%

main: 92%

push

github

Ensure inputs of cosine of solar zenith angle are chunked when needed (#1542)

<!--Please ensure the PR fulfills the following requirements! -->
<!-- If this is your first PR, make sure to add your details to the
AUTHORS.rst! -->
### Pull Request Checklist:
- [x] This PR addresses an already opened issue (for bug fixes /
features)
    - This PR fixes #1536
- [x] Tests for the changes have been added (for bug fixes / features)
- [ ] (If applicable) Documentation has been added / updated (for bug
fixes / features)
- [x] CHANGES.rst has been updated (with summary of main changes)
- [x] Link to issue (:issue:`number`) and pull request (:pull:`number`)
has been added

### What kind of change does this PR introduce?

* New function `xc.core.utils._chunk_like` to chunk a list of inputs
according to one chunk dictionary. It also circumvents
https://github.com/pydata/xarray/issues/6204 by recreating DataArrays
that where obtained from dimension coordinates.
* Generalization of `uses_dask` so it can accept a list of objects.
* Usage of `_chunk_like` to ensure the inputs of
`cosine_of_solar_zenith_angle` are chunked when needed, in
`mean_radiant_temperature` and `potential_evapotranspiration`.

The effect of this is simply that the `cosine_of_solar_zenith_angle`
will be performed on blocks of the same size as in the original data,
even though its inputs (the dimension coordinate) did not carry that
information. Before this PR, the calculation was done as a single block,
of the same size as the full array.

### Does this PR introduce a breaking change?
No.

### Other information:
Dask might warn something like `PerformanceWarning: Increasing number of
chunks by factor of NN`. Where NN should be the number of chunks along
the `lat` dimension, if present. That exactly what we want, so it's ok.

17 of 19 new or added lines in 3 files covered. (89.47%)

8545 of 9449 relevant lines covered (90.43%)

4.37 hits per line