CUB (version 0.1)

cubeforsim: Simulation routine for CUBE models without covariates

Description

Fit CUBE models without covariates to given ordinal data. It is useful for simulation experiments since it performs the same steps as the CUBE function, but with no printed output.

Usage

cubeforsim(m, ordinal, starting, maxiter = 500, toler = 1e-06, expinform=FALSE)

Arguments

m

Number of ordinal categories

ordinal

Vector of ordinal responses

starting

Vector of initial parameters estimates to start the optimization algorithm

maxiter

Maximum number of iterations allowed for running the optimization algorithm (defaul: maxiter = 500)

toler

Fixed error tolerance for final estimates (default: toler = 1e-6)

expinform

Logical: if TRUE, the function returns the expected variance covariance matrix (default is FALSE)

Value

An object of the class "CUBE", with null output for $BIC since the routine is only for simulation purposes

See Also

CUBE, loglikCUBE

Examples

data(relgoods)
m<-10
ordinal<-na.omit(relgoods[,37])
starting<-rep(0.1,3) 
simul<-cubeforsim(m,ordinal,starting)
param<-simul$estimates    # Estimated parameters vector (pai,csi,phi)
####################
data(univer)
m<-7
ordinal<-univer[,11]
starting<-rep(0.1,3)
simul<-cubeforsim(m,ordinal,starting)
param<-simul$estimates   # Estimated parameters vector (pai,csi,phi)