Matthew Barber
Matthew Barber
Looking at the diff again, this seems like it was an underlying issue that's only come up because this PR handles string labels correctly for categoricals? Mind I'm pretty ignorant...
> Yeah, feel free to xfail that test for now, on all platform Done!
I'll submit a PR for this
Ah so you fixed the issue I was alluding to in #2122 ```diff - buffer = _VaexBuffer(self._col.values) + buffer = _VaexBuffer(indices.to_numpy()) ``` Before a test like the following would fail...
Wrote a regression test https://github.com/vaexio/vaex/pull/2135
> Is there somewhere to track this work? This now lives in https://github.com/data-apis/array-api-compat, which has a compat layer for both NumPy and CuPy, and PyTorch soon too.
> @honno Is this something that we can move forward? I think we'd really want to get in https://github.com/data-apis/array-api/pull/589 first (which LGTM but we'd want someone else to give the...
Proposal: add APIs for getting and setting elements via a list of indices (i.e., `take`, `put`, etc)
I think if we were to introduce something like `xp.put(x, indices, value)` to the spec we can seemingly all agree on 1. **Only specifying a single array as the `indices`...
> Has this been resolved or does more work remain? Haven't touched this yet (specifically writing a note in the README on tags and what they do/don't mean), need to...
The issue @jakevdp had should be resolved by https://github.com/data-apis/array-api-tests/pull/213 now. > we should figure out if we actually want to keep the tags on this repo. @asmeurer If this is...