This function computes deviance between the full model and the null (intercept-only) model.
It uses built-in R functions, namely glm; slow but no C++ dependencies.
Make sure that LFr1 and LFr0 do not have intercept terms.
dev.R(dat, LFr1, LFr0 = NULL, p = FALSE)a matrix with m rows and n columns.
alternative logistic factors (an output from lfa or lfa.corpcor)
null logistic factors (an output from lfa or lfa.corpcor)
estimate p-values (by default, "FALSE")
When p=FALSE (by default), dev.R returns a vector of m deviances.
When p=TRUE, a list consisting of
the m deviances
the m p-values based on a chisq distribution