m_rank should be a 1-indexed array

lewisacidic / scikit-chem

A high level cheminformatics package for the Scientific Python stack, built on RDKit.

http://scikit-chem.readthedocs.io/en/latest/index.html

Other

63 stars 13 forks source link

m_rank should be a 1-indexed array #78

Closed jRicciL closed 4 years ago

jRicciL commented 4 years ago

I think m_rank, which represents r_i in the Truchon and Bayly paper, should be a 1-indexed array. Otherwise, the BEDROC value will reach values above 1. Changing to a 1-indexed array, I managed to get similar results to the CalcBEDROC() function from the rdkit.ML.Scoring.Scoring module.

lewisacidic commented 4 years ago

Hi @jRicciL thanks for the PR, and sorry for missing it. There is a rewrite of scikit-chem coming soon, and I believe the BEDROC implementation was redone from scratch (also RDKit now implements it!). I'll merge this, but once a few more issues are ironed out, this code will be replaced - just a heads up. Thanks again!