Feature/rdf normalization

[vyasr]

Description

Enable normalization of RDF to 1 in the large R limit. Also fix a bug in Jupyter plotting code.

Motivation and Context

Resolves: #396. Also fixes an unrelated bug.

How Has This Been Tested?

Screenshots

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds or improves functionality)
• Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

• I have read the CONTRIBUTING document.
• My code follows the code style of this project.
• I have updated the documentation (if relevant).
• I have added tests that cover my changes (if relevant).
• All new and existing tests passed.
• I have updated the credits.
• I have updated the Changelog.

[vyasr]

@joaander see #396 for the relevant discussion, I wanted to get your input before finalizing this.

[mphoward]

Thanks for fixing! This looks like it should work to me.

[codecov]

Codecov Report

Merging #529 into master will increase coverage by 0.05%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #529      +/-   ##
==========================================
+ Coverage   90.33%   90.39%   +0.05%     
==========================================
  Files          16       16              
  Lines        2049     2050       +1     
  Branches       27       27              
==========================================
+ Hits         1851     1853       +2     
+ Misses        185      184       -1     
  Partials       13       13

Impacted Files	Coverage Δ
freud/density.pyx	96.85% <100%> (+0.65%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update c9ea2f7...a7c25f4. Read the comment docs.

[joaander]

Looks good!

Levine, Stone, and Kohlmeyer have a clean way of expressing g(r) normalization with arbitrary particle selections if you would like to generalize this or use their notation in the documentation.

[vyasr]

Thanks for the paper! I think we could implement that generally here, it would just require maintaining a counter inside the loop in compute to determine the total number of pairs that we computed for and then normalizing for that case. However, it seems like a pretty esoteric use-case, so I don't think it's a priority and I'm happy to punt on it for now. @bdice thoughts? If you agree with me, then go ahead and merge this PR when you're happy and we can reconsider that at a future date.

[mphoward]

Josh makes a good point. It actually probably isn't too hard to treat this normalization exactly here because freud does not have such a strong sense of "selections" / "groups" or "tags" as other programs. The number of pairs in the normalization is N*(N-1) iff query_points is neighbors AND exclude_ii is True. Otherwise, the normalization is for N*N pairs, some of which may be excluded. The normalize flag would activate this pair-counting logic (and could include a link to that paper); otherwise, it would fall back on the old logic and always use the N*N pairs.

[bdice]

Looks good!