Augurs is a versatile time series analysis library for Rust with JavaScript bindings. It offers tools for clustering, outlier detection, forecasting, and changepoint detection. The library’s documentation provides detailed information, and users are encouraged to show support by starring it on GitHub. Notably, augurs allows for parallelized clustering with DBSCAN to identify series behaving similarly. Outlier detection with DBSCAN detects series that deviate from the norm. Forecasting options include the MSTL algorithm for multiple seasonalities and the Prophet algorithm for trend, seasonality, and holidays. Changepoint detection using a Bayesian Normal Gamma approach showcases changepoints in the data plot.
https://demo.augu.rs/