Given the new trends surrounding Ensemble Feature Selection and the needs of comparing discrepancies between different feature subsets generated by function or data perturbation, this repository aims to implement a correct and reliable Python computation of the Kuncheva Index measure, proposed by Ludmila I. Kuncheva. A stability index for feature selection. In Artificial Intelligence and Applications (AIAP’07), pages 390–395, 2007.

In order to use get_kuncheva_index method, one must provide a python list containing all subsets which also must be python lists. Along with the subsets containing the selected features, it is required by the n parameter the length of the original set. If complete rankings are given for each subset, i.e., it's not a subset but an ordered ranking containing all features, a threshold value must be provided and the length of the rankings shall be used as the n value, hence dismissing its input as a parameter.