This class of objects is returned as the
paramIter
component in the
class
"missmodel"
object returned by the
mdLoglin
,
completeLoglin
,
emLoglin
, and
daLoglin
functions.
The class
"Loglin"
has associated methods:
VarNames.Loglin
,
,
,
,
,
, and
.
The class
"Loglin"
object consists of a matrix. Each row in the
matrix represents one set of parameters estimates, typically produced
using an iterative algorithm like EM or data augmentation. In the
"Loglin"
object, the parameters are the cell probabilities in the log-linear
model. The number of parameters equals the number of cells in the
table specified by the log-linear model. Using the
control
argument
of
Loglin
,
completeLoglin
,
daLoglin
, or
emLoglin
, you can
choose which iterates to save. The iteration number for
each row of saved estimates is given by the row name.