Mathematical modeling of infectious disease dynamics in R.
EpiModel simulates epidemics with three classes of models, from deterministic compartmental models to stochastic network models in which partnerships form, persist, and dissolve according to rules estimated with temporal exponential-family random graph models (TERGMs) from the Statnet suite of R packages. Built-in SI, SIR, and SIS models run out of the box, and an extension API supports custom disease states, demography, and interventions for new research questions.
| Model class | Simulate with | Contact structure |
|---|---|---|
| Deterministic compartmental models | dcm() |
Population-level ODEs with random mixing |
| Stochastic individual contact models | icm() |
Individuals with random mixing |
| Stochastic network models |
netest(), netdx(), netsim()
|
Dynamic partnerships estimated from network data |
All three share one interface: set inputs with param.*(), init.*(), and control.*(), then simulate.
EpiModel is the core package of a broader platform for infectious disease modeling. Visit epimodel.org for the extension packages, the learning pathway, and the team.
The current release version can be found on CRAN and installed with:
install.packages("EpiModel", dependencies = TRUE)
To install the development version, use the remotes package:
if (!require("remotes")) install.packages("remotes")
remotes::install_github("EpiModel/EpiModel")
A network model in three steps: estimate a dynamic network from target statistics, check that simulations from it reproduce those targets, then simulate an epidemic over it. This runs in under a minute on a laptop.
library(EpiModel)
# 1. Estimate: 500 people, 175 partnerships, 110 with 2+ partners, none with 4+
nw <- network_initialize(n = 500)
est <- netest(nw,
formation = ~edges + concurrent + degrange(from = 4),
target.stats = c(175, 110, 0),
coef.diss = dissolution_coefs(~offset(edges), duration = 50) # mean duration
)
# 2. Diagnose: simulated networks should match the targets
dx <- netdx(est, nsims = 5, nsteps = 500)
plot(dx)
# 3. Simulate an SIS epidemic over the dynamic network
param <- param.net(inf.prob = 0.4, act.rate = 2, rec.rate = 0.05)
init <- init.net(i.num = 10)
control <- control.net(type = "SIS", nsims = 5, nsteps = 500)
sim <- netsim(est, param, init, control)
plot(sim)
The Intro to EpiModel vignette walks through each step and the analysis of the output.
 Package Documentation Function reference, vignettes, and getting started guides |
 Training Materials Network Modeling for Epidemics (NME), our open course at SISMID |
 EpiModel Gallery Worked examples and templates for extending network models |
The main methods paper in the Journal of Statistical Software describes the modeling framework and software design: doi:10.18637/jss.v084.i08. The learning pathway on epimodel.org orders these resources from beginning to advanced.
Users are encouraged to use GitHub Issues to ask questions (both technical coding questions and conceptual modeling questions), report bugs, and request new features.
If using EpiModel for teaching or research, please include a citation to our main methods paper:
Jenness SM, Goodreau SM and Morris M. EpiModel: An R Package for Mathematical Modeling of Infectious Disease over Networks. Journal of Statistical Software. 2018; 84(8): 1-47. doi: 10.18637/jss.v084.i08
Please also send us an email if you have used EpiModel in your work so we can add it to our records. EpiModel and its extension packages have been used in 130+ published studies; see the full list.
EpiModel is developed by Samuel Jenness, Steven Goodreau, Martina Morris, and Adrien Le Guillou. Additional contributors are listed on the contributors page.
Development has been supported by grants from the National Institutes of Health (NIH) and the Centers for Disease Control and Prevention (CDC). Details are on the project page.
These materials are distributed under the GPL-3 license, with the copyright and attribution requirements listed in the LICENSE.