R/IRT.expectedCounts.R

In alexanderrobitzsch/CDM: Cognitive Diagnosis Modeling

## File Name: IRT.expectedCounts.R
## File Version: 0.21

###########################################################
# extracts expected counts
IRT.expectedCounts <- function(object, ...)
{
    UseMethod("IRT.expectedCounts")
}
###########################################################

###########################################################
# object of class gdm
IRT.expectedCounts.gdm <- function( object, ... )
{
    ll <- aperm( object$n.ik, c(2,3,1,4) )
    attr(ll,"theta") <- object$theta.k
    attr(ll,"prob.theta") <- object$pi.k
    attr(ll,"G") <- object$G
    return(ll)
}
###########################################################

###########################################################
# object of class din
IRT.expectedCounts.din <- function( object, ... )
{
    Ilj <- object$I.lj
    D1 <- dim(Ilj)
    ll <- array( 0, dim=c( D1[1], 2, D1[2], 1) )
    ll[,2,,1] <- object$R.lj
    ll[,1,,1] <- object$I.lj - object$R.lj
    attr(ll,"theta") <- object$attribute.patt.splitted
    attr(ll,"prob.theta") <- object$attribute.patt$class.prob
    attr(ll,"G") <- 1
    attr(ll,"dimnames")[[1]] <- colnames(object$dat)
    return(ll)
}
###########################################################

###########################################################
# object of class gdina
IRT.expectedCounts.gdina <- function( object, ... )
{
    G <- object$G
    Ilj <- object$control$I.lj
    D1 <- dim(Ilj)
    ll <- array( 0, dim=c( D1[1], 2, D1[2], G ) )
    if (G==1){
        ll[,2,,1] <- object$control$R.lj
        ll[,1,,1] <- Ilj - object$control$R.lj
    }
    if (G>1){
        ll[,2,,] <- object$control$R.lj.gg
        ll[,1,,] <- object$control$I.lj.gg - object$control$R.lj.gg
    }
    attr(ll,"theta") <- object$attribute.patt.splitted
    attr(ll,"prob.theta") <- object$attribute.patt[, 1:object$G ]
    attr(ll,"G") <- object$G
    attr(ll,"dimnames")[[1]] <- colnames(object$dat)
    return(ll)
}
############################################################

###########################################################
# object of class slca
IRT.expectedCounts.slca <- function( object, ... )
{
    ll <- aperm( object$n.ik, c(2,3,1,4) )
    res <- list( "delta"=object$delta,
                "delta.designmatrix"=object$delta.designmatrix )
    attr(ll,"skillspace") <- res
    attr(ll,"prob.theta") <- object$pi.k
    attr(ll,"G") <- object$G
    return(ll)
}
############################################################

###########################################################
# object of class mcdina
IRT.expectedCounts.mcdina <- function( object, ... )
{
    ll <- object$n.ik
    attr(ll,"theta") <- object$attribute.patt.splitted
    attr(ll,"prob.theta") <- object$attribute.patt
    attr(ll,"G") <- object$G
    return(ll)
}
############################################################

###########################################################
# object of class reglca
IRT.expectedCounts.reglca <- function( object, ... )
{
    ll <- aperm( object$n.ik, c(2,3,1,4) )
    attr(ll,"theta") <- NA
    attr(ll,"prob.theta") <- object$class_probs
    attr(ll,"G") <- object$G
    return(ll)
}
############################################################