Understand the project
Practical overview
When it fits
Use Darts when a Python-based project needs an experimentation toolkit that covers forecasting and anomaly detection through one library, especially when the work involves univariate or multivariate series, external covariates, or probabilistic predictions.
Who it is for
- Python data scientists
- Machine-learning engineers building time-series applications
- Teams analyzing operational, business, financial or sensor data
- Developers experimenting with forecasting and anomaly-detection models
Documented uses
- Convert tabular observations into time-series objects
- Train forecasting models and predict future values
- Compare models through backtesting and shared fit/predict workflows
- Use external data and multiple time series in forecasting
- Generate probabilistic forecasts
Before you choose
Requirements & limits
The repository is public, open source, and licensed under Apache-2.0. The recommended setup uses Python 3.11 or newer and pip; Docker is identified as a supported packaging direction.
- Python 3.11 or newer is the documented recommended environment.
- The shared fit() and predict() pattern can simplify comparisons across forecasting models.
- Support for univariate and multivariate series affects how the library can be applied to a dataset.
- Probabilistic forecasting, external data, backtesting and model combinations are available workflow considerations.
- Anomaly workflows can combine scorers such as KMeansScorer with detectors such as QuantileDetector.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Evidence and transparency
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 2 October 2026
- Original website title
- unit8co/darts · Official homepage title · retained in its source language