TanaT Documentation#

TanaT is an extensible Python library for temporal sequence analysis with a primary focus on patient care pathways.

The objective of this library is to gather a collection of tools related to the analysis of timed sequences (or trajectories). It has been strongly inspired by the TraMineR library for the analysis of state sequences in R and libraries dedicated to time series analysis such as aeon or tslearn.

The originality of TanaT is to support multi-sequence trajectories that can combine three types of temporal data: states, intervals, and events. This allows tracking subjects (patients, users, customers, etc.) over time across multiple dimensions, providing a richer and more complete view of their temporal evolution. The library provides a comprehensive toolkit for multidimensional temporal pattern discovery and analysis.

On this page