[MRG] ENH: Geometric-SMOTE implementation

[joaopfonseca]

Reference Issue

Fixes #881

What does this implement/fix? Explain your changes.

Implements the Geometric SMOTE, as described in Geometric SMOTE a geometrically enhanced drop-in replacement for SMOTE by Douzas and Bacao.

[lgtm-com]

This pull request introduces 1 alert when merging fa3ffe5 into f1abf75 - view on LGTM.com

new alerts:

• 1 for Unused import

[codecov]

Codecov Report

Attention: Patch coverage is 98.90110% with 2 lines in your changes missing coverage. Please review.

Project coverage is 97.26%. Comparing base (f1abf75) to head (c9945c0).
Report is 132 commits behind head on master.

Files with missing lines	Patch %	Lines
imblearn/over_sampling/_smote/geometric.py	97.84%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #882      +/-   ##
==========================================
+ Coverage   97.21%   97.26%   +0.04%     
==========================================
  Files          96       98       +2     
  Lines        6253     6435     +182     
  Branches      712      733      +21     
==========================================
+ Hits         6079     6259     +180     
- Misses        103      104       +1     
- Partials       71       72       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[joaopfonseca]

@glemaitre, @chkoar I believe this is ready for revision. Some of the tests are not passing but from my understanding they are not related to the modifications in this PR.

If I may add, I just noticed imbalanced-learn's SMOTE-based over-samplers are only implementing modifications to the data selection mechanism. Geometric-SMOTE is a modification to the data generation mechanism. I hope you find this implementation useful and appropriate.