stumpy-dev/stumpy

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

4158 stars · 371 forks observed · 2026-09-30

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

STUMPY is a Python library that efficiently computes the matrix profile for time series, enabling fast nearest-neighbor subsequence searches. It supports time series data mining tasks such as motif discovery, anomaly (discord) detection, and semantic segmentation, with scalable GPU and distributed (Dask) backends.

Use cases

• find repeated patterns (motifs) in a long time series
• detect anomalies or discords in sensor data
• compute the matrix profile for time series
• segment a time series into semantic regions
• mine time series data at scale with Dask or GPUs
• find nearest-neighbor subsequences in streaming data

When to choose

• you need efficient motif or anomaly discovery in univariate or multivariate time series
• you want a well-tested, NumFOCUS-affiliated library with 100% test coverage
• you need scalability via Numba, GPU, or Dask parallelization

When to avoid

• you need general-purpose forecasting or classical time series models like ARIMA
• you work outside Python
• you need real-time streaming analytics with sub-millisecond latency guarantees

Facets

library · maturity stable

data-science math data-science time-series analytics python cross-platform matrix-profile time-series-analysis motif-discovery anomaly-detection numba dask

Member repositories

For agents

markdown · JSON · MCP: product_card(name="stumpy-dev/stumpy")