theochem / Selector

Python library of algorithms for selecting diverse subsets of data for machine-learning.
https://selector.qcdevs.org
GNU General Public License v3.0
22 stars 22 forks source link

Exclude neighbours of the ref_index in DISE algorithm #223

Closed FarnazH closed 3 months ago

FarnazH commented 5 months ago

In DISE algorthim, the ref_index was not included as a selected sample, and its neighbors were not excluded. I believe this was not a desired feature, so this PR fixed that. Feel Free to let me know what you think.

FarnazH commented 5 months ago

@FanwangM and @marco-2023, the tests currently fail. If you agree with this change, I will update the tests. Just to let you know the OptiSim algorithm adds ref_index as its first selection; see https://github.com/theochem/Selector/blob/main/selector/methods/distance.py#L293

FanwangM commented 5 months ago

It looks good to me. Thanks for fixing this.