print.moc prints information contained in a fitted moc
object. The attributes parameters of the functions
gmu, gshape, gextra and gmixture will be
used to label the output.
coef.moc returns the coefficients (estimated parameters) of a
fitted moc object.
fitted.moc computes the expected values for each observation
of a moc object using its expected function.
obsfit.moc computes and prints the mean posterior
probabilities and the posterior means of a user specified function of
the expected and observed values, separated with respect
to the specified variable.
# S3 method for moc
print(x, digits = 5, expand = TRUE, transpose = FALSE, …)# S3 method for moc
coef(object, split=FALSE, …)
# S3 method for moc
fitted(object, …)
obsfit.moc(object, along = list(cons = rep(1, object$nsubject)),
FUN = function(x) x)
Objects of class moc.
If split is TRUE, returns a list with elements corresponding to mu, shape, extra and mixture parameters.
Number of digits to be printed.
Expand density, gmu, gshape, gextra, gmixture function body in the print.
Transpose fitted.mean and observed.mean in the print.
Splitting variable.
User defined function to apply to observed and expected values.
Unused.
All these methods return their results invisibly.
obsfit.moc will first compute the posterior probabilities
for all subjects in each mixture using post.moc and
then the weighted posterior mean probabilities
$$\hat{\bar{\tau}}_k = \frac{\sum_i wt_i\,\hat{\tau}_{i,k}}
{\sum_i wt_i}$$
The weighted posterior means of a function \(g()\) of the data
(which are the empirical estimators of the conditional expectation given
mixture group) are computed as
$$\frac{\sum_i wt_i\,\hat{\tau}_{i,k}\,g(y_i)}{\sum_i
wt_i\,\hat{\tau}_{i,k}}$$
where both sums are taken over index of valid data \(y_i\).