Starting Values in Conditional Multivariate Gaussian Models

DESCRIPTION:

Generates starting values for conditional multivariate gaussian models. This routine is not normally called directly by users.

USAGE:

cgm.start(object) 

REQUIRED ARGUMENTS:

object
an object of class "preCgm".

VALUE:

a list with three components:
mu

a matrix of zeros with object$q rows and object$ncells columns.

sigma
an object$q by object$q identity matrix, and
pi
a vector of length object$ncells of ones.

DETAILS:

This routine is not normally called directly by users.

SEE ALSO:

, , .

EXAMPLES:

language.pre <- preCgm(language) 
cgm.start(object = language.pre)