Compute the Model Matrix

DESCRIPTION:

Compute the model matrix for a hierarchical log-linear model. This is an internal routine and is not normally called directly by users.

USAGE:

Loglin.get.x(data, margins) 

REQUIRED ARGUMENTS:

data
a data frame or matrix containing the raw data. When a data frame is input, the table is specified by the levels of the factor variables. When the input object is a matrix, it is assumed that the levels of each variable (column) form a sequence of integers from one to the maximum value of the variable.

OPTIONAL ARGUMENTS:

margins
a formula or a list of vectors containing the marginal totals to be fit. A margin is described by the factors not summed over. Thus list(1:2, 3:4) would indicate fitting the 1,2 margin (summing over variables 3 and 4) and the 3,4 margin in a four-way table. This same model can be specified using the names of the variables (e.g., list(c("V1", "V2"), c("V3", "V4"))), or using formula notation, as in margins = ~V1:V2 + V3:V4. When formula notation is used, the argument frequency can be included as the dependent variable (as in margins = frequency~V1:V2 + V3:V4). If margins is not specified, a saturated models (a single interaction term containing all table variables) is fit. When a matrix is input, every column in the matrix is used to define the table. When a data frame is input, the table is defined by the factor variables in the data frame.

VALUE:

a list containing:
x
the model matrix.
levs
the number of levels in each column in x.
iy
the column number of the frequency variable argument in argument data.
nmarg
a vector containing a listing of the margins to be fitted.
saturated.model
TRUE if the model is saturated, and FALSE otherwise
slevs
a list containing the names associated with each of the levels in each column of the design matrix.

DETAILS:

This is an internal routine and is not usually called directly by users.

SEE ALSO:

.

EXAMPLES:

Loglin.get.x(data = crime)