Athan

> JAX supporting 1-dimensional arrays. @jakevdp Would be good to update the docstring then for `bincount`, as currently it suggests that N-dimensional support is present. It is also not clear...

> I'd vote for keyword-only, since it's a very descriptive name and there's no real reason to use positional-only as far as I can tell. @rgommers I am fine making...

For anyone hoping to work on this issue, note that, IMO, the only appropriate way to provide a robust implementation is by implementing the Unicode sentence segmentation algorithm.

Looks like the tests still need to be updated.

For those interested in this project, you can find out more about the current state of things at the following places in the code base: - https://github.com/stdlib-js/stdlib/blob/develop/tools/make/lib/docs/typedoc.mk - https://github.com/stdlib-js/stdlib/tree/develop/etc/typedoc

For those interested in this project, please note that we are NOT interested in developing a test runner in the style of mocha, jest, ava, chai, or anything other than...

@mdhaber Do you have a sense as to how common array `fill_value` support is across the ecosystem? (e.g., NumPy, PyTorch, JAX, Dask, MLX, ndonnx, CuPy)

I've opened another PR which should supersede this one (see https://github.com/data-apis/array-api/pull/896) and includes the 3-way merge edits (including backport) which was included here.

Looks like NumPy's behavior does stem from NumPy 2.0: https://numpy.org/doc/1.26/reference/generated/numpy.broadcast_arrays.html. Previously, it returned a list.

IMO, the approach in this PR is preferred when compared to https://github.com/data-apis/array-api/pull/801. The main issue which I raised there is that changes should actually be constrained to normative content. At...