For each of the three model classes in EpiModel (deterministic compartmental models, individual contact models, and network models) the tutorials are organized into basic "built-in" models to guide new users in the features of the model class, and advanced extension models to build out the models to address new research questions.
Before diving into any tutorials, we recommend starting with our main methods paper published in the Journal of Statistical Software. It is available at http://doi.org/10.18637/jss.v084.i08.
In July 2020, we have consolidated all of the tutorial materials for the network model class in our extended materials for the Network Modeling for Epidemics course. This also coincided with the release of EpiModel 2.0, a major update to the software package that improves the functionality of EpiModel but may require some updating of code (see the migration section below).
Basic Built-In Models
Basic DCMs with EpiModel This tutorial
provides some mathematical background for deterministic compartmental
models, with exploration of different model types and parameterizations
Basic ICMs with EpiModel Stochastic
individual contact models (ICMs) are the microsimulation analogs to
DCMs. This tutorial explains the general differences between deterministic
and stochastic modeling, with hands-on basic examples.
Basic Network Models with EpiModel Stochastic network models build in arbitrarily complex contact or partnership relational structures that form and dissolve over time, using the framework of temporal exponential random graph models. See the materials from Days 3 and 4 of the Network Modeling for Epidemics course for these tutorial materials.
Advanced Extension Models
New DCMs with EpiModel Creating new
deterministic compartmental models in EpiModel involves writing new
model functions defining the mathematical transition processes, and then
parameterizing and simulating those models. This tutorial shows examples
of how to write model functions, include new parameters, and run new
New Network Models with EpiModel Writing new network models requires creating modules that are plugged in to the epidemic simulation. This framework allows for new epidemiological processes of arbitrary complexity that interact with the unique dynamic network structures of interest. See the materials from Days 4 and 5 the Network Modeling for Epidemics course for these tutorial materials. In addition, we have a growing library of examples for extension models for EpiModel at our EpiModel Gallery.
EpiModel 2.0 Migration
EpiModel 2.0 incorporates a substantial redesign of many elements of the EpiModel infrastructure and application programming interface (API). We anticipate that there will be some minor backwards incompatibilities with any EpiModel code developed with versions 1.x. We have therefore prepared an EpiModel 2.0 Migration Guide to outline these changes and assist users in transitioning.