arundo/adtk

Health v2 · maintenance only

Adoption not part of the score

1211 stars · 163 forks observed · 2026-08-28

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

ADTK is a Python package for unsupervised and rule-based anomaly detection in time series. It provides detectors, transformers, and aggregators with unified APIs plus pipeline classes to combine them into detection models.

Use cases

• detect anomalies in time series data
• unsupervised anomaly detection in sensor data
• rule-based spike and level-shift detection
• build anomaly detection pipelines in Python
• visualize time series anomaly events
• monitor metrics for outliers

When to choose

• you need unsupervised or rule-based time series anomaly detection without labeled training data
• you want composable detectors, transformers, and aggregators with a unified API
• you need to visualize and process time series anomaly events

When to avoid

• you need supervised anomaly detection with labeled training data
• you need deep learning-based anomaly detection models
• you need streaming/real-time detection at scale rather than batch analysis

Facets

library · maturity maintenance

machine-learning data-science data-visualization time-series data-science analytics python anomaly-detection time-series unsupervised-learning detectors transformers

Member repositories

For agents

markdown · JSON · MCP: product_card(name="arundo/adtk")