The ergm.degeneracy
function checks a given ergm object for degeneracy by
computing and returning the instability value of the model and the value
of the log-likelihood function at the maximized theta values
ergm.degeneracy(object, control = object$control,
fast = TRUE, test.only = FALSE,
verbose = FALSE)
an ergm
object
the list of control parameters as returned by control.ergm
;
default=control.ergm()
whether the degeneracy check should be "fast", i.e to sample changeobs(?) when there are > 100, rather than use all changeobs; default=TRUE
whether to silence printing of the model instability calculation (T or F); this parameter is ignored if the instability > 1; default=FALSE
whether to print a notification when 'object' is deemed degenerate (T or F); default=FALSE
returns the original ergm object with 2 additional components:
degeneracy.value: the instability of the model
degeneracy.type : a 2-element vector containing
loglikelihood: the value of the log-likelihood function corresponding to 'theta'; if degenerate, this is a vector of Inf
theta : the vector of theta values found through maximixing the log- likelihood; if degenerate, this is 'guess'