All Versions
14
Latest Version
Avg Release Cycle
115 days
Latest Release
-

Changelog History
Page 2

  • v1.0.1 Changes

    January 02, 2017

    Date: 02/01/17

    ✨ Enhancements

    • ➕ Added the GridSearch feature, by Maher
    • ➕ Added a 'clip' option to the predict() method
    • ➕ Added NMF algorithm
    • ➕ Added entry point for better command line usage.
    • ➕ Added CoClustering algorithm.
    • ➕ Added SlopeOne algorithm.
    • ➕ Added Probabilistic Matrix Factorization as an option SVD
    • Cythonized Baseline Computation

    Other

    • Surprise is now a scikit!
    • 🔄 Changed license to BSD
    • Six is now a dependency

  • v1.0.0 Changes

    November 22, 2016

    Date: 22/11/16

    • 🔄 Changed name from recsys to surprise
    • 👌 Improved printing of accuracy measures.
    • ➕ Added version number.
    • Rewrote the the main.py

  • v0.0.4 Changes

    November 15, 2016

    Date: 15/11/16

    ✨ Enhancements

    • ➕ Added notebooks for comparing and evaluating algorithm performances
    • 👍 Better use of setup.py
    • ➕ Added a min_support parameter to the similarity measures.
    • ➕ Added a min_k parameter to the KNN algorithms.
    • The similarity matrix and baselines are now returned.
    • ✅ You can now train on a whole training set without test set.
    • The estimate method can return a tuple with prediction details.
    • ➕ Added SVD and SVD++ algorithms.
    • ✂ Removed all the x/y vs user/item stuff. That was useless for most algorithms.

    API Changes

    • ✂ Removed the @property decorator for many iterators.
    • It's now up to the algorithms to decide if they can or cannot make a prediction.

  • v0.0.3 Changes

    Date: 25/10/16

    • ➕ Added support for Python 2