Package ranky
Compute rankings in Python.
import ranky as rk
All functions are available from the top-level rk namespace. They are
organized in the following modules:
ranky.ranking: ranking systems (Borda count, Copeland, Kemeny-Young, …).ranky.metric: scoring metrics, rank distances and correlations, agreement measures.ranky.duel: pairwise comparison functions and statistical tests.ranky.visualization: plots of rankings and preference matrices.ranky.generator: generation of synthetic judges.ranky.utilities: reading Codalab leaderboards.
A preference matrix has one row per candidate and one column per judge. Most functions accept np.ndarray or pd.DataFrame; with a DataFrame, the names of the candidates and judges are kept in the results.
Made by Adrien PAVAO, 2020.
Sub-modules
ranky.duelranky.generatorranky.metricranky.rankingranky.utilitiesranky.visualization