EpiModel package provides functions for
building, solving, and plotting mathematical models of
infectious disease. The goals of the package are to provide
basic tools for modeling in multiple frameworks for
pedagogical purposes, and to support users to develop and
expand these tools using the package's utility functions
for their own research.EpiModel currently provides functionality for three
classes of epidemic models:
EpiModelallows for easy specification of
sensitivity models to compare multiple runs of the same
model with different parameter values.Statnetsuite of software in R, our
network models simulate partnership formation and
dissolution stochastically according to a user-defined
statistical model, as well as disease spread on that
network.EpiModel supports three infectious disease types to
be run across all of the three classes:
The core functions for parameterizing and solving the three model classes are:
epiDCMfor
deterministic compartmental epidemic models.epiICMfor individual contact epidemic
models.epiNet.estfor estimating the
statistical models underlying partnership formation and
dissolution used in stochastic network epidemic models.
This function is a wrapper around theergmandstergmfunctions in theergmandtergmpackages, respectively, with additional diagnostic tables
and plots useful for epidemic modeling.epiNet.simNetfor simulating networks given a
model fit withepiNet.est. These network simulations
are used for network epidemic models in which there is no
dependence between the network structure and the disease
process (thus, the network structure may be fully simulated
ahead of the disease simulation).epiNet.simTransfor stochastic network
epidemic models, with a given network model fit or set of
network simulations fromepiNet.estorepiNet.simNet, respectively. For models
involving dependence of disease trajectories on the network
structure (e.g., disease causes death, which dissolves
partnerships), it is not necessary to pre-simulate the
networks since each disease simulation re-simulates the
network at each time step. A help page providing an
overview of the internals ofepiNet.simTransthat may be useful for adapting and expanding the software
for novel research is available atepiNetModules.