The Statnet / EpiModel Family Tree

network · networkDynamic · sna · ergm · tergm · tsna · ndtv · EpiModel · EpiModelHIV · EpiModelCOVID

statnet.org  ·  CRAN: statnet

Outline for EpiModel in SISMID Modules

  • Basic Epidemics on Networks
    • Modeling epidemics + networks = modeling epidemics over networks
    • Core assumption: no feedback of epidemiology on networks
      • One important implication: closed populations
      • Still feedback: network structure ⟹ epidemiology, and incidence ⟹ prevalence
    • Built-in epidemiology types (SI, SIR, SIS)
      • Working with nodal attributes, with heterogeneity in network structure and epidemiological parameters
  • Open Population Models
    • Feedback: epidemiology ⟹ network structure
      • Vital dynamics, “sero-sorting” (edge formation based on changing nodal attributes)
    • Simple vaccine intervention
    • Built-in epidemiology types (SI, SIR, SIS), then getting started with extensions

EpiModel: Built-in and Extended

  • EpiModel is designed for both built-in (“toy models”) and user-defined extensions (“research models”).
  • This course focuses on built-in network models. Extensions are more complicated, and are the focus of NME-II.

Closed Population

EpiModel Workflow for Built-In Models

  1. Construct the (empty) network data structure
  2. Parameterize the TERGM (formation and dissolution formulas and target statistics)
  3. Fit the TERGM, and diagnose the model fit
  4. Parameterize the epidemic model
  5. Simulate the epidemic
  6. Analyze the simulation data

EpiModel Workflow for Built-In Models

  1. Construct the (empty) network data structure: network_initialize, set_vertex_attribute
  2. Parameterize the TERGM: ~, dissolution_coefs
  3. Fit the TERGM, and diagnose the model fit: netest, netdx
  4. Parameterize the epidemic model: param.net, init.net, control.net
  5. Simulate the epidemic: netsim
  6. Analyze the model data: print, plot, summary, as.data.frame, …