handle all nan case for loess smoothing

[juliettelavoie]

Pull Request Checklist:

• This PR addresses an already opened issue (for bug fixes / features)
  • This PR fixes #xyz
• Tests for the changes have been added (for bug fixes / features)
  • (If applicable) Documentation has been added / updated (for bug fixes / features)
• CHANGES.rst has been updated (with summary of main changes)
  • Link to issue (:issue:number) and pull request (:pull:number) has been added

What kind of change does this PR introduce?

• Trying to fix https://stackoverflowteams.com/c/ouranos/questions/103
• When the axis is only nan and skipna=True, don't even try to calculate a loess (because it fails), just return nan.

Does this PR introduce a breaking change?

before the all nan case would fail.

Other information:

[aulemahal]

Looks good!

[juliettelavoie]

Can I ignore the License compliance issues and merge (it doesn't look blocked) ? Maintainer access is needed to see the details of the 2 fossa issues..

[aulemahal]

Yes you can! I also don't have access to the report, but the issues reported by @tlogan2000 are more confusing than critical...

[coveralls]

coverage: 90.19%. remained the same
when pulling db819b3 on fix-loess-nan
into c6c10cb on main.