matrixpls method for generic function
fitted computes the model implied
covariance matrix by combining
formative as a
simultaneous equations system. The error terms are constrained to be uncorrelated and
covariances between exogenous variables are fixed at their sample values. Defining a
composite as dependent variable in both inner and formative creates an impossible model
and results in an error.
matrixpls estimation result object produced by the
All other arguments are ignored.
a matrix containing the model implied covariances.