Add more similarity measurements, fixes #124

[FanwangM]

This PR addresses #124. The SimilarityIndex class was moved to similarity module to make things consistent, which is mainly for calculating similarity matrix. The smilarity.py in method module was renamed to sim.py because we have two similarity.py files, which can be very confusing. Actually I would perfer to use nsim.py, but to keep things consistent with other files name, I chose sim.py.

[codecov]

Codecov Report

Attention: Patch coverage is 97.62846% with 6 lines in your changes missing coverage. Please review.

Project coverage is 96.96%. Comparing base (0de2c9b) to head (d5cc58d).
Report is 37 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #188      +/-   ##
==========================================
- Coverage   97.57%   96.96%   -0.61%     
==========================================
  Files           9        9              
  Lines         907      924      +17     
==========================================
+ Hits          885      896      +11     
- Misses         22       28       +6     

Files	Coverage Δ
selector/methods/__init__.py	100.00% <100.00%> (ø)
selector/methods/sim.py	96.05% <96.05%> (ø)
selector/similarity.py	98.61% <98.29%> (-1.39%)	⬇️

... and 1 file with indirect coverage changes

[FanwangM]

I am going to close this PR as this has been addressed by

• New average similarity method #189
• Rename simlarity.py to sim.py #219