search

MobleyLab / chemper

Repository for Chemical Perception Sampling Tools
MIT License
19 stars 10 forks source link

Add functions for clustering fragments based on data #51

Closed bannanc closed 5 years ago

bannanc commented 5 years ago

@vtlim This pull request I think adds everything you need to finish your pipeline.

This PR adds clustering using scikit-learn's BayesianGaussianMixture as a default, but would support any clustering function that can call .fit and .predict.

It builds on PRs #45, #48, and #50 so those should be merged in that order first and then I will deal with it if conflicts arise.

I will add an example notebook here eventually, but it will require, I think being able to interact externally with OpenEye or RDKit so I need to find a way to skip those notebooks while testing if a tool isn't installed.

bannanc commented 5 years ago

These tests will fail because I still haven't fixed the tests that I intentionally broke in PR #50, but that should just be two tests in test_graphs_output.py otherwise everything else should still be working.

codecov[bot] commented 5 years ago

Codecov Report

Merging #51 into master will decrease coverage by <.01%. The diff coverage is 90.41%.