install numbagg from conda-forge

[keewis]

It seems there is a numbagg package on conda-forge now.

Not sure what to do about the min-all-deps CI, but given that the most recent version of numbagg happened more than 12 months ago (more than 18 months, even) maybe we can just bump it to the version on conda-forge?

[keewis]

that looks like it works, besides the fact that conda reports numbagg=0.2.1 but

print(f"numbagg={numbagg.__version__}")

prints numbagg=0.2.0 (cc @ocefpaf)

[ocefpaf]

that looks like it works, besides the fact that conda reports numbagg=0.2.1 but

It is not a conda problem but a bug upstream. The v0.2.1 tag hasv0.2.0 hardcoded in it.

See https://github.com/numbagg/numbagg/blob/v0.2.1/setup.py#L24

[keewis]

ah, that makes sense, thanks. I must have missed that line while checking the package metadata. This has even been fixed since then, but not released. @max-sixty, would it be possible to issue a new release?

[max-sixty]

This has even been fixed since then, but not released. @max-sixty, would it be possible to issue a new release?

This should now be releasing! https://github.com/numbagg/numbagg/releases/tag/v0.2.2

[keewis]

do we need to worry about this:

Traceback (most recent call last):
  File "/home/runner/micromamba-root/envs/xarray-tests/lib/python3.10/doctest.py", line 1350, in __run
    exec(compile(example.source, filename, "single",
  File "<doctest xarray.core.rolling_exp.RollingExp.sum[1]>", line 1, in <module>
  File "/home/runner/work/xarray/xarray/xarray/core/rolling_exp.py", line 162, in sum
    return self.obj.reduce(
  File "/home/runner/work/xarray/xarray/xarray/core/dataarray.py", line 3667, in reduce
    var = self.variable.reduce(func, dim, axis, keep_attrs, keepdims, **kwargs)
  File "/home/runner/work/xarray/xarray/xarray/core/variable.py", line 1955, in reduce
    data = func(self.data, axis=axis, **kwargs)
  File "/home/runner/work/xarray/xarray/xarray/core/rolling_exp.py", line 41, in move_exp_nansum
    raise ValueError("`rolling_exp(...).sum() requires numbagg>=0.2.1.")
ValueError: `rolling_exp(...).sum() requires numbagg>=0.2.1.

not being caught by CI?

Edit: to answer my own question: no, because most of the the rolling_exp tests are explicitly skipped if the numbagg version is too old