Skip to contents

AMPH Forecast Suite

Suite of tools and helper functions to build and run an infectious disease forecasting pipeline and hub.

Overview

The AMPH Forecast Suite is an R package that provides a comprehensive toolkit for building infectious disease forecasting pipelines and hubs. It streamlines the process of working with multiple forecasting packages by:

  • Installing and managing dependencies for key forecasting packages
  • Setting up appropriate directory structure
  • Pulling and manipulating data to use with forecasting models
  • Providing helper functions to translate data between different package formats
  • Running forecasts and nowcast using multiple packages
  • Ensembling, visualizing, and evaluating forecasts
  • Producing outputs compatible with forecast hubs and evaluation tools

For more information, see the package documentation and function help pages.

Installation

You can install the development version from GitHub:

# Install remotes if you haven't already
install.packages("remotes")

# Install AMPH Forecast Suite
remotes::install_github("ACCIDDA/AMPH_Forecast_Suite")

Install Packages and Build Directory Structure

The package includes a function to install all required packages and set up a directory structure for your forecasting project.

library(AMPHForecastSuite)

# Install all forecasting packages
install_forecast_packages()

# Or install specific categories
install_forecast_packages(
  install_hubverse = TRUE,
  install_forecasting = TRUE,
  install_epi = TRUE,
  install_data = TRUE,
  install_nowcast = TRUE
)

Packages for Infectious Disease Forecasting

The package helps you install and work with:

Hubverse Packages

  • hubData: Data management for forecast hubs
  • hubUtils: Utility functions for hubs
  • hubValidations: Validation tools for hub submissions
  • hubVis: Visualization tools for hub data
  • hubAdmin: Administrative tools for hubs

Forecasting Packages

  • fable: Forecasting models for tidy time series
  • fabletools: Tools for working with fable models
  • feasts: Feature extraction and statistics for time series
  • epiforecast: Tools for infectious disease forecasting
  • tsibble: Tidy temporal data frames

Epidemiological Packages

  • EpiEstim: Estimate time-varying reproduction numbers
  • EpiNow2: Estimate Rt and forecast with delays
  • epipredict: Epidemiological forecasting tools

Nowcasting packages

  • epinowcast: Flexible Hierarchical Nowcasting
  • baselinenowcast: Baseline nowcasting methods for handling delays in epidemiological data

Data Access

  • epidatr: Access epidemiological data from various sources

Additional Resources

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the GPL-3 License - see the LICENSE file for details.

Citation

If you use this package in your research, please cite it as:

ACCIDDA (2025). AMPH Forecast Suite: Tools for Infectious Disease Forecasting.
R package version 0.1.0. https://github.com/ACCIDDA/AMPH_Forecast_Suite

Support

For questions and issues, please open an issue on GitHub.