Skip to contents

Summary

This article lists open source implementations and interesting applications of dynamic time warping I have found in the literature, patents, and other resources. It is supposed to be a living document, so it will be updated from time to time with new findings.

Open Source Implementations

This section lists several open source implementations of DTW in R and Python, in case you want to play with DTW.

R

  • dtw, by Toni Giorgino: most complete DTW suite in R, with a companion paper here. It also has a Python implementation linked below.
  • dtwclust, by Alexis Sarda-Espinosa: focused on applying DTW for time series clustering. It has a very comprehensive technical vignette here.
  • IncDTW, by Maximilian Leodolter: efficient implementation with a companion paper, focused on fast DTW computations for real-time applications.

Python

  • dtw-python: Python implementation of dtw.
  • dtaidistance: fast C implementation of DTW, with a great documentation page here.
  • tslearn, by Romain Tavenard: focused on time series clustering via DTW for machine learning modeling.
  • fastdtw, by Stan Salvador and Philip Chan: python implementation of FastDTW, a fast approximation to DTW described in this paper.

Other Languages

The Wikipedia page on Dynamic Time Warping lists implementations in other languages in the section Open Source Software.