Tom White

`take_along_axis` is difficult to implement efficiently for chunked arrays (see discussion and ideas in https://github.com/dask/dask/issues/3663), so I think it's worth declaring support for 2024.12 even without it.

The Xarray change that enables this has been merged (https://github.com/pydata/xarray/pull/9530). I've re-created the branch from #468 here: https://github.com/cubed-dev/cubed/tree/np-array-function. However, when I run the [cubed-xarray](https://github.com/cubed-dev/cubed-xarray) tests I get some failures: ```...

Fixed in #670

Here is a possible plan of attack. This is not in any particular order (since it would depend on what the demand is), but I've tried to tease out the...

> Yes to all of these! Also an example showing that xarray groupby with flox works. We already have https://flox.readthedocs.io/en/latest/user-stories/climatology-hourly-cubed.html, but we could add something to the Cubed docs too...

As noted in https://github.com/cubed-dev/cubed-xarray/issues/26, in the case of `index`, the special case of slices that don't cross chunk boundaries it would be easy to implement, and useful since it's so...

> the special case of slices that don't cross chunk boundaries it would be easy to implement, and useful since it's so common Fixed in #586

#701 is related

The reason for the failures is because Zarr Python v3 makes more memory copies than v2 in some cases, see https://github.com/tomwhite/memray-array

I'm going to merge this now. Note that this runs the mem utilization tests only on Zarr v2 for the time being. Most work on Zarr v3, but there is...