R/tam_pv_mcmc_proc_input.R

In TAM: Test Analysis Modules

## File Name: tam_pv_mcmc_proc_input.R
## File Version: 0.15

tam_pv_mcmc_proc_input <- function( tamobj, group, Y )
{

    tam_class <- class(tamobj)
    # available classes
    tam_classes_available <- c( "tam.mml", "tam.mml.3pl")
    is_tam_class <- tam_class %in% tam_classes_available

    #** extract some elements
    person <- tamobj$person
    AXsi <- tamobj$AXsi
    AXsi[ is.na(AXsi ) ] <- -99
    B <- tamobj$B
    guess <- tamobj$guess
    resp <- tamobj$resp
    resp.ind <- tamobj$resp.ind
    pweights <- tamobj$pweights
    pid <- tamobj$pid
    D <- dim(B)[3]
    nstud <- nrow(resp)
    nitems <- nrow(AXsi)
    #** is_tam_class==FALSE
    if ( ! is_tam_class ){
        #.. create person inits
        person <- tam_pv_mcmc_person_inits_from_resp( resp=resp, D=D )
        #.. process resp
        resp.ind <- 1 - is.na(resp)
        resp[ is.na(resp) ] <- 0
    }

    #** more values used in tam.pv.mcmc function
    resp1 <- resp + 1
    resp_ind_bool <- resp.ind==1
    maxK <- ncol(AXsi)
    if ( is.null(guess) ){
        guess <- rep(0,nitems)
    }
    if ( is.null(pweights) ){
        pweights <- rep(1,nstud)
    }
    if ( is.null(pid) ){
        pid <- 1:nstud
    }

    #--- grouping variable
    if (is.null(group)){
        group <- rep(1,nstud)
    }
    groups <- sort(unique(group))
    G <- length(groups)
    group <- match( group, groups )
    group_index <- list()
    N_groups <- rep(0,G)
    for (gg in 1:G){
        group_index[[gg]] <- which( group==gg )
        attr(group_index,"N_groups")[gg] <- v1 <- length(group_index[[gg]])
    }
    c2 <- cumsum( attr(group_index,"N_groups") )
    dfr <- data.frame("start"=c(1, c2[-G] + 1 ), "end"=c2 )
    attr(group_index,"N_groups_cumsum") <- dfr

    if ( is.null(Y) ){
        Y <- matrix(1, nrow=nstud, ncol=1)
    }
    Y <- as.matrix(Y)

    #--- OUTPUT
    res <- list(person=person, is_tam_class=is_tam_class, resp=as.matrix(resp),
                AXsi=AXsi, B=B, guess=guess, resp.ind=as.matrix(resp.ind),
                pweights=pweights, nitems=nitems, nstud=nstud,
                maxK=maxK, D=D, pid=pid, group=group, G=G, groups=groups,
                group_index=group_index, Y=Y, resp_ind_bool=resp_ind_bool)
    return(res)
}