ergm.ego: Inference for Exponential-Family Random Graph Models based on...

View source: R/ergm.ego.R

ergm.egoR Documentation

Inference for Exponential-Family Random Graph Models based on Egocentrically Sampled Data


A wrapper around the ergm to fit an ERGM to an egor.


  popsize = 1,
  offset.coef = NULL,
  constraints = ~.,
  control = control.ergm.ego(),
  na.action =,
  na.rm = FALSE, = TRUE



An formula object, of the form e ~ <model terms>, where e is a egor object. See ergm for details and examples.

For a list of currently implemented egocentric terms for the RHS, see ergm.ego-terms.


The size |N| of the finite population network from which the egocentric sample was taken; only affects the shift in the coefficients of the terms modeling the overall propensity to have ties. Setting it to 1 (the default) essentially uses the -\log |N'| offset on the edges term. Passing 0 disables network size adjustment and uses the egocentric sample size; passing I(N) uses the specified size N (though can be overridden by the ppop control.ergm.ego() option) and disables network size adjustment.


A vector of coefficients for the offset terms.


A one-sided formula formula giving the sample space constraints. See ergm for details and examples.


Additional arguments passed to ergm.


A control.ergm.ego control list.


How to handle missing actor attributes in egos or alters, when the terms need them for models that scale.


How to handle missing actor attributes in egos or alters, when the terms need them for models that do not scale.

Whether to actually call ergm


An object of class ergm.ego inheriting from ergm, with the following additional or overridden elements:


Variance-covariance matrix of the estimate of the sufficient statistics


Estimate of the sufficient statistics


The egor object passed


Population network size used


Pseudopopulation size used, see control.ergm.ego


The coefficients, along with the network size adjustment netsize.adj coefficient.


Pseudo-MLE estimate of the variance-covariance matrix of the parameter estimates under repeated egocentric sampling


The variance-covariance matrix of parameter estimates under the ERGM superpopulation process (without incorporating sampling).


Estimated Jacobian of the expectation of the sufficient statistics with respect to the model parameters


Pavel N. Krivitsky


Pavel N. Krivitsky and Martina Morris (2017). "Inference for social network models from egocentrically sampled data, with application to understanding persistent racial disparities in HIV prevalence in the US." Annals of Applied Statistics, 11(1): 427–455. doi: 10.1214/16-AOAS1010

Pavel N. Krivitsky, Martina Morris, and Michał Bojanowski (2019). "Inference for Exponential-Family Random Graph Models from Egocentrically-Sampled Data with Alter–Alter Relations." NIASRA Working Paper 08-19.

Pavel N. Krivitsky, Michał Bojanowski, and Martina Morris (2020). "Impact of survey design on estimation of exponential-family random graph models from egocentrically-sampled data." Social Networks, to appear. doi: 10.1016/j.socnet.2020.10.001

Pavel N. Krivitsky, Mark S. Handcock, and Martina Morris (2011). "Adjusting for Network Size and Composition Effects in Exponential-Family Random Graph Models." Statistical Methodology, 8(4): 319–339. doi: 10.1016/j.stamet.2011.01.005

See Also



fmh.ego <- as.egor(faux.mesa.high)


egofit <- ergm.ego(fmh.ego~edges+degree(0:3)+nodefactor("Race")+nodematch("Race")

# Run convergence diagnostics

# Estimates and standard errors

statnet/ergm.ego documentation built on June 13, 2022, 5:20 p.m.