Summary
EpiNow2 estimates the time-varying reproduction number, growth rate, and doubling time using a range of open-source tools (Abbott et al.), and current best practices (Gostic et al.). It aims to help users avoid some of the limitations of naive implementations in a framework that is informed by community feedback and is actively supported.
Forecasting is also supported for the time-varying reproduction number, infections, and reported cases using the same generative process approach as used for estimation.
More details
EpiNow2 estimates the time-varying reproduction number on cases by date of infection (using a similar approach to that implemented in {EpiEstim}). True infections, treated as latent and unobserved, are estimated and then mapped to observed data (for example cases by date of report) via one or more delay distributions (in the examples in the package documentation these are an incubation period and a reporting delay) and a reporting model that can include weekly periodicity.
Uncertainty is propagated from all inputs into the final parameter estimates, helping to mitigate spurious findings. This is handled internally. The time-varying reproduction estimates and the uncertain generation time also give time-varying estimates of the rate of growth.
Models provided
EpiNow2 provides three models:
estimate_infections(): Reconstruct cases by date of infection from reported cases.estimate_secondary(): Estimate the relationship between primary and secondary observations, for example, deaths (secondary) based on hospital admissions (primary), or bed occupancy (secondary) based on hospital admissions (primary).estimate_truncation(): Estimate a truncation distribution from multiple snapshots of the same data source over time. For more flexibility, check out the{epinowcast}package.
The default model in estimate_infections() uses a non-stationary Gaussian process to estimate the time-varying reproduction number and infer infections. Other options, which generally reduce runtimes at the cost of the granularity of estimates or real-time performance, include:
- A stationary Gaussian process (faster to estimate but currently gives reduced performance for real time estimates).
- User specified breakpoints.
- A fixed reproduction number.
- A piecewise constant, combining a fixed reproduction number with breakpoints.
- A random walk, combining a fixed reproduction number with regularly spaced breakpoints (i.e weekly).
- A deconvolution/back-calculation method for inferring infections, followed with calculating the time-varying reproduction number.
- Adjustment for the remaining susceptible population beyond the forecast horizon.
By default, all these models are fit with MCMC sampling using the rstan R package as the backend. Users can, however, switch to use approximate algorithms like variational inference, the pathfinder algorithm, or Laplace approximation especially for quick prototyping. The latter two methods are provided through the cmdstanr R package, so users will have to install that separately.
The documentation for estimate_infections provides examples of the implementation of the different options available.
EpiNow2 is designed to be used via a single function call to two functions:
epinow(): Estimate Rt and cases by date of infection and forecast these infections into the future.regional_epinow(): Efficiently runepinow()across multiple regions in an efficient manner.
These two functions call estimate_infections(), which works to reconstruct cases by date of infection from reported cases.
For more details on using each function corresponding function documentation.
Installation
Install the released version of the package:
install.packages("EpiNow2")Install the development version of the package with:
install.packages("EpiNow2", repos = c("https://epiforecasts.r-universe.dev", getOption("repos")))Alternatively, install the development version of the package with pak as follows (few users should need to do this):
# check whether {pak} is installed
if (!require("pak")) {
install.packages("pak")
}
pak::pkg_install("epiforecasts/EpiNow2")If using pak fails, try:
# check whether {remotes} is installed
if (!require("remotes")) {
install.packages("remotes")
}
remotes::install_github("epiforecasts/EpiNow2")To build EpiNow2 from source, users will need to configure their C toolchain. This is because EpiNow2 implements the underlying models in Stan (a statistical modelling programming language), which is built on C++.
Each operating system has a different set up procedure. Windows users need to install an appropriate version of RTools. Mac users can follow these steps, and Linux users can use this guide. For simple deployment/development a prebuilt docker image is also available (see documentation here).
Resources
Getting Started
The Getting Started vignette (see vignette("EpiNow2")) is your quickest entry point to the package. It provides a quick run through of the two main functions in the package and how to set up them up. It also discusses how to summarise and visualise the results after running the models.
More broadly, users can also learn the details of estimating delay distributions, nowcasting, and forecasting in a structured way through the free and open short-course, “Nowcasting and forecasting infectious disease dynamics”, developed by some authors of this package.
Package website
The package has two websites: one for the stable release version on CRAN, and another for the version in development. These two provide various resources for learning about the package, including the function reference, details about each model (model definition), workflows for each model (usage), and case studies or literature of applications of the package. However, the development website may contain experimental features and information not yet available in the stable release.
End-to-end workflows
The workflow vignette (see vignette("estimate_infections_workflow")) provides guidance on the end-to-end process of estimating reproduction numbers and performing short-term forecasts for a disease spreading in a
Model definitions
In different vignettes we provide the mathematical definition of each model. For example, the model definition vignette for estimate_infections() can be found in vignette("estimate_infections").
Example implementations
A simple example of using the package to estimate a national Rt for Covid-19 can be found here.
Contributing
We welcome all contributions. If you have identified an issue with the package, you can file an issue here. We also welcome additions and extensions to the underlying model either in the form of options or improvements. If you wish to contribute in any form, please follow the package contributing guide.
Contributors
All contributions to this project are gratefully acknowledged using the allcontributors package following the all-contributors specification. Contributions of any kind are welcome!
Code
seabbs, sbfnk, jamesmbaazam, joeHickson, hsbadr, pitmonticone, actions-user, ellisp, jdmunday, pearsonca, JAllen42, kaitejohnson, adamkucharski, andrjohns, Bisaloo, LloydChapman, medewitt, nikosbosse, sophiemeakin
Issue Authors
raulfernandezn, pcarbo, johnaponte, sophie-schiller, munozedg, kathsherratt, yungwai, kgostic, fkrauer, philturk, krageth, tony352, username-rp, HAKGH, AndrewRiceMGW, brynhayder, RichardMN, andrybicio, rhamoonga, furqan915, MFZaini1984, fabsig, affans, GauriSaran, davidvilanova, jrcpulliam, dajmcdon, joshwlambert, avallecam, athowes, lorenzwalthert, nlinton, martinamcm, adrian-lison, zsusswein