tslearn

Why we included this project

tslearn fills a gap that generic scikit-learn leaves open: it gives you time series specific algorithms with the same familiar API. You can fit a KNN classifier or a TimeSeriesKMeans model just like any other scikit-learn estimator, and the underlying machinery handles dynamic time warping, barycenters, variable-length series, and a distance metrics module that most general-purpose libraries skip. Preprocessing utilities such as scaling and resampling, plus a synthetic data generator, make it easy to prototype before moving to heavier forecasting stacks. Teams already using scikit-learn will appreciate how directly it drops into existing pipelines.