Period group embedding

[chrisiacovella]

Pull Request Summary

This adds in the ability to do embedding with the atomic group and period along with atomic number, as mentioned in issue #330. Like atomic number, we can specify number of features (i.e, the size of the tensor), as opposed to be a singular value like charge.

Initially I added period and group to the NNPInput, but decided to change this to being set in featurization just based on atomic number. Adding to NNPInput breaks a lot of other things (since it uses slots and is no longer dynamic) and we should really only really need this this if it is requested in featurization (so it's a waste of memory to set this for each batch). For things that don't have a group (specifically the Lanthanoid series) likely need to return a default value of zero. Currently I have Lanthanum set to be group 3, only because it is in TMQM and thus needed a definition (from a chemistry perspective I'm not sure if that makes sense to group it in there, but since Sc or Y in the dataset I'm not sure it matters).

Key changes

Notable points that this PR has either accomplished or will accomplish.

• Adds atomic group and period to featurization (computed if requested)

Associated Issue(s)

• Period/Group embedding #330

Pull Request Checklist

• Issue(s) raised/addressed and linked
• Includes appropriate unit test(s)
• Appropriate docstring(s) added/updated
• Appropriate .rst doc file(s) added/updated
• PR is ready for review

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 79.01%. Comparing base (8120b70) to head (d2a0ff4).

Additional details and impacted files