dataDepPrior.preCgm(object, nPriorObs, algorithm = "da")
"preCgm"
.
"em"
for the EM algorithm, or
"da"
for data augmentation.
The prior probabilities are computed assuming all variables are independent. For each variable, the marginal probability is computed using the observed data for that variable.
Schafer, J. L. (1997), Analysis of Incomplete Multivariate Data, Chapman & Hall, London.
dataDepPrior.preCgm(object = preCgm(stlouis3), nPriorObs = 5, algorithm = "da")