facebook/prophet

Tool for producing high quality forecasts for time series data that has multiple seasonality with linear or non-linear growth.

What it solves

Prophet provides an automatic procedure for forecasting time series data. It is specifically designed to handle data with strong seasonal effects, missing values, trend shifts, and outliers, making it easier to generate reliable forecasts without requiring deep expertise in time series modeling.

How it works

It uses an additive model where non-linear trends are combined with yearly, weekly, and daily seasonality, as well as holiday effects. The system is built on a Stan backend (using cmdstan or rstan) to fit these components to historical data.

Who it’s for

Data scientists and analysts who need to perform time series forecasting at scale, particularly those working with data that exhibits strong seasonality and has several seasons of historical records.

Highlights

  • Robust to missing data and outliers.
  • Handles multiple seasonalities (daily, weekly, yearly) and holiday effects.
  • Available as both Python and R packages.
  • Supports extra regressors and custom performance metrics.

Related

  • Project
    WenjieDu/PyPOTS

    A Python toolbox for machine learning on partially-observed time series, providing unified APIs for imputation, forecasting, and anomaly detection on data with missing values.

  • Project
    facebookresearch/Kats

    A comprehensive toolkit for time series analysis that provides a unified framework for forecasting, anomaly detection, and feature extraction.

  • Project
    Nixtla/nixtla

    A foundation model for time series forecasting and anomaly detection that provides zero-shot inference and fine-tuning capabilities across multiple domains.

  • Project
    wwrechard/pydlm

    A flexible Bayesian dynamic linear model library for Python that enables time series decomposition and forecasting through additive components.

  • Project
    Nixtla/statsforecast

    A high-performance library for univariate time series forecasting that provides fast, scalable implementations of statistical and econometric models.