daLoglin.compute(object, nmarg, prior, start, control = daLoglin.control(), saturated.model)
"preLoglin"
object containing the table to be analyzed.
c(1,2,0,2,3,0,1,3)
would indicate fitting
the (1,2), (2,3), and (1,3) margins in a three-way table, i.e., the
model of no three-way interaction.
"priorLoglin"
, or an array of
hyperparameters.
"ml"
(maximum likelihood) or
"noninformative"
.
String matching is used,
so the characters
"m"
or
"n"
are sufficient. The values
of the hyperparameters changes with the algorithm (see
for details). E.g.
"noninformative"
means a common value of 1 for
EM, and a common value of 0.5 for DA.
"priorLoglin"
object is created by routine
priorLoglin
.
This routine allows you to easily specify data dependent priors.
start
for the order to use in specifying a vector of
hyperparameters. If a single numeric value is input, its value is
replicated for all cells in the table.
"noninformative"
.
NA
) when a vector of
hyperparameters is input as argument
prior
.
mdLoglin
are the cell probabilities. Thus,
start
is a vector with
length equal to the total length of the table containing a probability
estimate for each cell in the table. Starting values for cells that
are structural zeros in the table should be zero. The default
starting values are all equal to one divided by the number of cells in
the table. Suppose that the table is defined by the variables
X1
,
X2
, and
X3
.
Then the cells in the table are ordered such
that the index for variable
X1
varies fastest, the index for
variable
X2
varies next fastest, etc.
"missmodel"
is returned; see
for details.
daLoglin.compute
creates the data
set
.Random.seed
if it does not already exist,
otherwise its value is updated.
See the help file for for additional details. This function is not normally called directly by users.