wannesm/dtaidistance

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

1248 stars · 191 forks observed · 2026-10-03

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

A Python library for computing time series distances, most notably Dynamic Time Warping (DTW), with a fast C implementation exposed via Cython. It supports N-dimensional series, parallelized distance matrix computation, DTW Barycenter Averaging for clustering, subsequence search, and local concurrences.

Use cases

• compute dtw distance between two time series
• compute pairwise dtw distance matrix for many series fast
• cluster time series with dtw barycenter averaging
• find best matching subsequence in a longer series
• compare multivariate time series with dtw
• speed up dtw computation with pruning

When to choose

• you need fast, C-accelerated DTW in Python with minimal dependencies
• you want DTW-based clustering or subsequence search for time series
• you work with numpy/pandas time series and need efficient distance matrices

When to avoid

• you need many other time series similarity measures beyond DTW variants
• you need a pure non-Python solution or GPU-accelerated DTW
• you need full time series forecasting or classification pipelines rather than distances

Facets

library · maturity stable

math machine-learning data-science time-series machine-learning data-science python cross-platform dtw time-series-distances clustering cython numpy dynamic-time-warping algorithms

Member repositories

For agents

markdown · JSON · MCP: product_card(name="wannesm/dtaidistance")