tergm_MCMC_sample is a low-level internal function not intended to
be called directly by end users. It collects a sample of networks and
returns the statistics of each sample, along with a toggle matrix of the
changes needed from the original network to each in the sample.
tergm_MCMC_slave is an even
lower-level function that actually calls the C code.
tergm_MCMC_sample(
nw,
model,
model.mon = NULL,
proposal,
control,
theta,
verbose = FALSE,
...,
eta = ergm.eta(theta, model$etamap)
)tergm_MCMC_slave(state, eta, control, verbose)
returns the MCMC sample as a list containing:
statsmatrix.gen: the matrix of sampled statistics for model,
relative to the initial network
statsmatrix.mon: the matrix of sampled statistics for model.mon,
relative to the initial network
newnetwork: ergm_state with the final network from the
sampling process
changed: a matrix of changes, where the first column is
the timestamp of the change, the second and third columns are the tail and head
(respectively) of the changed dyad, and the fourth column is the edge state to which
the dyad was changed; this is only returned if control$changes is TRUE
maxchanges: the maxchanges value from the control list
a network object
the model, as returned by ergm_model
the optional monitoring model, as returned by ergm_model
the proposal, as returned by ergm_proposal()
the list of control parameters
the vector of curved parameters
A logical or an integer to control the amount of
progress and diagnostic information to be printed. FALSE/0
produces minimal output, with higher values producing more
detail. Note that very high values (5+) may significantly slow
down processing.
the vector of natural parameters
This function is normally called inside simulate.tergm() functions
to prepare inputs for the C sampling code and return its results
simulate.tergm()