v0.2.0 - 2026-09-15

Major refactor, add lots of new functions, and add documentation. Can now run time-series cross-validation with ARIMA, Prophet, Naive, ensemble and ETS models, and then calculate forecast errors and generate various plots. ARIMA and Prophet both support multiple exogenous regressors.

Added

• Add functions for creating ensemble models (ensemble.py), calculating forecast errors (errors.py), running ETS models (ets.py), naive models (naive.py), more plots (plots.py), and dividing data for validation (splits.py).
• Add great-docs documentation site (including GitHub actions to render and deploy docs and lint alt text, plus two pages in user_guide/).
• Add all-contributors.

Changed

• Major refactor! To improve usability. Changes throughout existing codebase (e.g., arima.py, prophet.py, forecast.py, processing.py).
• Improved README (e.g., illustration, further descriptions).
• Dependencies no longer pinned.