Welcome to pairwise-ranking’s documentation!

pairwise-ranking is a Python library for producing ranking information from pairwise data. Particularly, it implements Bayesian inference on extensions of the popular Bradley-Terry model popularly known as the basis for Elo scores.

Check out the Usage section for further information, including on Installation.

Note

This project is under active development.

Contents

• Usage
  • Installation
  • Loading data
    • read_match_list()
    • read_match_list_from_match_list()
    • read_match_list_from_gml()
    • read_match_list_from_adj_matrix()
  • Inference
    • scores()
    • ranks()
    • probability()
  • Sampling
    • samples()
    • draw_depth_and_luck_posterior()
• Models
• Data sets
  • Sports
  • Human hierarchies
  • Animal hierarchies