Support promolecules with single atom

[msricher]

Fixes #134

[PaulWAyers]

Thanks @msricher . I'd like to get @marco-2023 and/or @gabrielasd 's feedback too, but it looks good to me.

[PaulWAyers]

Don't we need to do the same trick for charges/mults in lines 582-592?

[msricher]

Yep you're right. I didn't think about both of those also being single values.

[PaulWAyers]

Looks good to me. @marco-2023 and/or @gabrielasd

[gabrielasd]

Changes look good to me too.

@msricher could you add a test case for this fix to test_promolecule.py
It could be a slight change of this test case in the module:

AtomDB/atomdb/test/test_promolecule.py

Lines 83 to 88 in abf1008

	pytest.param(
	{
	"atnums": [4],
	"charges": [1.2],
	"mults": [1.2],
	"coords": np.asarray(

[gabrielasd]

I don't think this is related to the changes Michelle introduced, but the CI checks are failing when building the website.

It mentions an incompatibility between Scipy library and python version. This doesn't make sense to me since the rest of the checks for different python envs are passing.
It could be that the Scipy version listed here differs from the one in the pyproject.toml, but I am not sure.

@msricher or @marco-2023 could you look into this?

[FanwangM]

I am not familiar with the technical part, but the failing test is fixed now.
@PaulWAyers @msricher @marco-2023 @gabrielasd

[PaulWAyers]

We used to limit the top version of Scipy because we were using its old-fashioned generalized inverse. We don't use that anymore (and probably not at all in this package) so this is a good update on the dependencies.