ergm.ego() currently calls ergm() with force.main=TRUE, even when the model is dyad-independent and sufficient statistic is achievable and achieved by SAN. This is because it requires the Hessian returned by ergm for its error calculations. If ergm returns the Hessian even for MPLE, it cold just use the MPLE.
ergm.ego()
currently callsergm()
withforce.main=TRUE
, even when the model is dyad-independent and sufficient statistic is achievable and achieved by SAN. This is because it requires the Hessian returned byergm
for its error calculations. Ifergm
returns the Hessian even for MPLE, it cold just use the MPLE.