v0.2.0

Added

• Data drift: Bhattacharyya distance. #123
• Data drift: Hellinger distance. #122
• Concept drift: Incremental Kolmogorov-Smirnov test. #125
• Callbacks: ResetOnDataDrift, PermutationTestOnBatchData, History #124 and WarningSamplesBuffer. #127
• Magic method repr on each object. #131
• Documentation references on each detector, dataset or metric. #136
• Include Test PyPI on publish workflow. #145

Changed

• Rename Unsupervised methods to Data drift methods.
• Rename Supervised methods to Concept drift methods.
• Update documentation API. #133
• Update README.md example and replace tables with diagrams. #140

Removed

• scikit-learn integration dependency. From now on it's possible to use any framework. #121
• Concept drift: MD3-RS and MD3-SVM (will be added in the future).

Fixed

• All documentation examples. #128
• Concept drift: ADWIN epsilon calculation. #141

v0.1.0

Added

• Supervised methods.
• Unsupervised methods.
• Semi-supervised methods.
• Real dataset (Elec2).
• Synthetic generator (SEA).

Changed

Removed

Fixed