completeGauss(data, subset, prior = <<see below>>, start = <<see below>>, control = emGauss.control())
data
is a data frame,
this expression may use variables in the data frame.
"priorGauss"
giving the hyperparameters
of the prior distribution. Routine
priorGauss
is used to create
the class
"priorGauss"
object. Alternatively, the character
strings "ml" (for no prior, i.e., maximum likelihood estimation),
"noninformative" (for a noninformative prior), or "ridge" (for the
default ridge prior) may be used. Pattern matching means that only
the first character in the string is required. See
for
details.
completeGauss
, but is included to conform with other
missing data functions.
completeGauss
, but is included to conform with other
missing data functions.
"missmodel"
is returned; see
for details. In the class
"missmodel"
object returned by
completeGauss
, the
paramIter
component
contains one or more rows of parameter estimates, and the
algorithm
element contains an object of class
"em"
.
The
completeGauss
function computes
Bayes estimates of the parameters in a multivariate normal model.
Schafer, J. L. (1997), Analysis of Incomplete Multivariate Data, Chapman & Hall, London.
completeGauss(data = na.omit(cholesterol))