- Fix dependency problems
- Bump version up for JOSS correction
- Bump version up for JOSS
- Test multiple python versions
- Installation problems: remove pandoc from setup.py
- PyPi RST format description
- Hoaglin's Approximation for p-value based metric
- repo moved to the organization
github.com/satzbeleg
- Default settings changed for
transition_simulation
- regularize zero-elements in
mle_btl_sparse
- Scoring by Orme (2009) added to docs
- Bradley-Terry-Luce (BTL) Model added
- Platt scaling added from scikit-learn
- sparse averaging for ratio-based and p-value-based ranking
- derive scores by solving an eigenvalue problem
- Code refactoring
- Main wrapper functions: bws.sampling, bws.counting, bws.ranking
- Identify pairs by logical inference between two BWS sets
- Scaling of just two frequencies based on simple averaging
- Use p-values of an one-sided Chi-Sequared test as scaled value
- Count pairs from a BWS set
- Converter to scipy sparse matrix
- Initial Release