Get Design Matrix in Conditional Gaussian Models

DESCRIPTION:

Computes and returns the design matrix in conditional gaussian models.

USAGE:

cgmDesign(object) 

REQUIRED ARGUMENTS:

object
an object of class "missmodel" with a paramIter component that inherits from class "cgm".

VALUE:

the design matrix. This is the matrix used in computing the linear model parameters for the means. If the model is saturated so that there is no linear model, then value NULL is returned.

DETAILS:

The conditional gaussian model allows the multivariate normal means to be predicted as a linear model. This routine returns the design matrix ( X) in the linear model ( mu = X b).

SEE ALSO:

, , , .

EXAMPLES:

stlouis.da3 <- daCgm(stlouis3[,-1], 
                     margins = ~risk:D1 + risk:D2 + D1:D2, 
                     design = ~D1*D2 + G, 
                     optData = data.frame(G = rep(-1:1, 4)), 
                     control = list(save = 101:5100, trace = T)) 
cgmDesign(stlouis.da3)