Clarify Laziness in Iris Functions #5066
Conversation
for more information, see https://pre-commit.ci
There was a problem hiding this comment.
Here are objects where I would expect to see a note, but didn't find one:
iris.analysis.MEDIANiris.analysis.PROPORTIONiris.analysis.WPERCENTILEiris.analysis.cartography.projectiris.analysis.cartography.rotate_grid_vectorsiris.analysis.geometry.geometry_area_weightsiris.analysis.maths.apply_ufunciris.cube.Cube.interpolate
You've covered everything in iris.cube.Cube that mentions coords, probably a few more places than you needed to, but we should have consistency with cell measures and ancillary variables, which have analogous behaviour with coords. Same goes for aux_factories even though they are a bit different.
The note isn't applicable for iris.palette.is_brewer, since it doesn't touch any array objects.
There was a problem hiding this comment.
Great! @ESadek-MO and I have agreed that we'll continue with #3292 after v3.4, but this is already a big improvement for user clarity 😁
I've spot checked some of the notes and confirmed that they are correct.
Happy to merge once this comes out of Draft.
🚀 Pull Request
Addresses #3292.
Description
Add comments in docstrings for each iris function clarifying if it can maintain laziness or not.
https://esadek-mo-irisdocs.readthedocs.io/en/lazydata/