R/sastraj_to_lctm.R

In hlennon/LCTMtools: Latent Class Trajectory Models: Tools for checking adequacy

#' Converts a SAS proc traj model to the format of R's hlme class
#'
#' @param oe contains model parameter estimates and maximised likelihood, AIC, BIC values
#' @param of contains posterior probabilities
#' @param op contains predictors
#' @param os containts fixed effect  and class membershop parameter estimates
#' @return  A format to feed into the LCTMtoolkit() R function
#' @examples
#' \dontrun{sastraj_to_lctm(oe, of, op, os)
#' os$PI/100
#' exp(c(M$best[1:4], 0))/(sum(exp(c(M$best[1:4], 0))))
#' }
#' @export

sastraj_to_lctm <- function(oe, of, op, os) {
    n <- nrow(of)
    K <- nrow(os)
    mod <- list(NULL)
    mod$pprob <- as.data.frame(of[, c("ID", "GROUP", paste0("GRP", 1:K, "PRB",
        sep = ""))])
    colnames(mod$pprob) <- c("ID", "class", paste0("prob", 1:K, sep = ""))
    mod$call <- "SAS"
    mod$os   <- os
    mod$oe   <- oe
    mod$of   <- of
    mod$op   <- op
    # model$par <- os$PI
    model <- mod
    return(model)
}