authored by Johannes Bracher; note that the package re-uses code from the surveillance package, some of which is itself adopted from base R
This repository contains a development version of hhh4addon, an add-on package to the surveillance package. hhh4addon adds functionality to surveillance:hhh4 which implements endemic-epidemic modelling for infectious disease count time series. Specifically it extends the model class towards distributed lags and adds functions for the analytical calculation of probabilistic path forecasts and cyclostationary moments. Note that it is currently only partly compatible with the hhh4contacts package, another add-on package for surveillance::hhh4 by Sebastian Meyer.
The exported functions are by now largely documented, including short examples. To get an overview over the functionality of the package please consider the vignette vignette(hhh4addon).
The theory behind longterm predictions and cyclostationary moments can be found in Held, Meyer and Bracher (2017): Probabilistic forecasting in infectious disease epidemiology: the 13th Armitage lecture and the technical report Bracher and Held (2017): Periodically stationary multivariate non-Gaussian autoregressive models. An R file to reproduce the results in this report can be found in example_analyses/Bracher_Held_2017.R . A second example analysis can be found in example_analyses/noro_rota_berlin.R
The analysis in the manuscript Bracher/Held: Endemic-epidemic models with discrete-time serial interval distributions for infectious disease prediction are moved to a different repository: https://github.com/jbracher/dengue_noro_rota.
To install the package directly from github and explore its functionality type:
library(devtools)
install_github("jbracher/hhh4addon", build_vignettes = TRUE)
library(hhh4addon)
vignette("hhh4addon")
You can find my personal website here and contact me via [email protected]