# unit8co/darts

A python library for user-friendly forecasting and anomaly detection on time series.

Repository: https://github.com/unit8co/darts
Canonical: https://ross.abutalabs.com/products/darts
Homepage: https://unit8co.github.io/darts/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: python, time-series, forecasting, machine-learning, deep-learning, anomaly-detection, data-science
Last push: 2026-08-21T08:30:55+00:00

## Health v2 (maintenance only)
Score: 92/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 98, release rhythm 81, longevity 100
- inputs: {"age_days": 2911, "days_push": 12, "days_rel": 44, "gap_med": 38, "n_releases_24m": 20}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 9505, forks 1030 (observed 2026-08-28T04:10:31.536454+00:00)

## What it is
Darts is a Python library for user-friendly time series forecasting and anomaly detection. It provides a unified scikit-learn-like fit/predict API across models ranging from classical methods like ARIMA to deep neural networks, with support for probabilistic forecasting, backtesting, and multivariate series.

## Use cases
- forecast future values of a time series in python
- detect anomalies in time series data
- train forecasting models on multiple time series
- compare ARIMA vs neural network forecasting models
- produce probabilistic forecasts with confidence intervals
- backtest a forecasting model on historical data
- use covariates like holidays or weather in time series forecasting

## When to choose
- you want one consistent API for many forecasting models, from ARIMA to deep learning
- you need anomaly detection built on top of forecasting or filtering models
- you want probabilistic forecasting and model backtesting out of the box
- you work with univariate or multivariate time series in pandas

## When to avoid
- you need lightweight, dependency-free statistical forecasting only
- your use case is general tabular ML regression rather than time series
- you require real-time streaming anomaly detection at very low latency

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, data-science, nlp
- domain: time-series, machine-learning, data-science, deep-learning
- platform: python
- tags: forecasting, anomaly-detection, time-series, probabilistic-forecasting, arima, deep-learning, scikit-learn-style

## Member repositories
- unit8co/darts (main) score 92

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:31.536454+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:22:01.103318+00:00, confidence not recorded.
  - readme: https://github.com/unit8co/darts (fetched 2026-08-28T04:10:31.536454+00:00, sha 2a42866c3cb2)
  - homepage: https://unit8co.github.io/darts/ (fetched 2026-08-29T08:21:52.179355+00:00, sha a932ffc2b185)
  - registry_pypi: https://pypi.org/pypi/darts/json (fetched 2026-08-29T08:21:52.196036+00:00, sha 40e988f8e8c9)
- Data as of 2026-08-30T08:39:29.467469+00:00.