Tom Nicholas

Looks like these > https://data-apis.org/array-api-tests/. use these > [experimental array api strategies](https://hypothesis.readthedocs.io/en/latest/numpy.html#array-api) > Would that be useful here? I think they are complementary. In **theory** if xarray supports the array...

I've also accidentally done this before and would also prefer a guard against my own trigger finger :sweat_smile:

My understanding is that if `__array_function__` is working correctly you should never need to call `cupy.round` on your dataarray. Instead you should always be able to call `np.round` and trust...

Hi @zoj613 , thanks for the bug report. This should be a very simple fix, and your MVCE would make a great test for the behaviour. Would you be interested...

This sounds like an interesting project - I'm also about to be able to work on xarray much more directly (thanks @rabernat ). Should I add this as another xarray...

Flagging another possible use case, this time in Magnetic Confinement Fusion: representing the IMAS data model. IMAS is currently closed-source (being part of the ITER project), but there is a...

I also added my chunking strategy from https://github.com/HypothesisWorks/hypothesis/issues/3433

@Zac-HD would you be able to give your hypothesis-expert opinions on the general design of these strategies? Especially the whole "strategies that accept multiple other strategies" thing.

> I'll aim for a proper review tonight! Amazing! Thank you! > Quick remarks: strategies accepting strategies is fine, though [our API style guide](https://github.com/HypothesisWorks/hypothesis/blob/master/guides/api-style.rst) suggests accepting strategies xor values Oh...

Thanks for the feedback @Zac-HD, that's extremely helpful already! Sounds like we need to decide on the general API approach, then I can go back and fix all the internals.