nparams() currently not implemented. The number of parameters should be divided by r, where r is the number of classes, since each local model has r-1 dummy classes. This should be updated locally (for each model) as it may effect BIC and AIC scores. See test-multinet.r
nparams()
currently not implemented. The number of parameters should be divided by r, where r is the number of classes, since each local model has r-1 dummy classes. This should be updated locally (for each model) as it may effect BIC and AIC scores. See test-multinet.r