The Trivariate Moment Decomposition model involves the computation of
principal polynomial components, which are derived for each of the three categorical
variables, and of the polynomial core array which is akin to the generalised correlations
between these components. The function step.g3ordered
computes the core array.
step.g3ordered(param)
The core matrix whose the general element can be interpreted as a generalized singular value.
The weighted three-way table and the matrices of the row, column and tube components derived via the Trivariate Moment Decomposition algorithm.
Rosaria Lombardo, Eric J Beh, Ida Camminatiello.
Beh EJ and Lombardo R (2014) Correspondence Analysis, Theory, Practice and New Strategies. John Wiley & Sons.