Geolet is a Python library that offers an interpretable transformation and classification approach for trajectory data. Geolet first partitions trajectories into multiple segments, yielding candidate subtrajectories. Secondly, it normalizes and filters them to produce a set of prototypical subtrajectories. Then, Geolet transforms the dataset using a custom Shapelet Transform adapted to Trajectory data. Finally, any interpretable classification model, such as random forests or SVM, can be used to classify the transformed data. The package offers multiple distance methods, partitioning strategies, and subtrajectories selection filters.