R/print.pec.R

In pec: Prediction Error Curves for Risk Prediction Models in Survival Analysis

#' Printing a `pec' (prediction error curve) object.
#' 
#' Print the important arguments of call and the prediction error values at
#' selected time points.
#' 
#' @aliases print.pec summary.pec
#' @param x Object of class \code{pec}
#' @param times Time points at which to show the values of the prediction error
#' curve(s)
#' @param what What estimate of the prediction error curve to show. Should be a
#' string matching an element of x. The default is determined by
#' splitMethod.
#' @param digits Number of decimals used in tables.
#' @param print Set to FALSE to suppress printing.
#' @param \dots Not used
#' @return The first argument in the invisible cloak.
#' @author Thomas A. Gerds \email{tag@@biostat.ku.dk}
#' @seealso \code{\link{pec}}
#' @keywords survival
#' @export
print.pec <- function(x,
                      times,
                      digits=3,
                      what=NULL,
                      ...){
    cat("\nPrediction error curves\n\n")
    # {{{ echo models
    cat("Prediction models:\n\n")
    printModels <- sapply(x$models,function(m){
        if (inherits(x = m,what = "character")||inherits(x = m,what = "call"))
            m
        else
            if (inherits(try(m$call,silent=TRUE),what="try-error"))
                "unknown formula"
        else
            m$call
    })
    print(printModels,quote=FALSE)
    # }}}
    # {{{ echo response
    print(x$response)
    # }}}
    # {{{ echo cens model
    if (!is.null(x$cens.model)){
        if (x$cens.model!="none")
            cat("\nIPCW:",x$cens.model,"model")
        else cat("\nno censoring")}
    # }}}
    # {{{ discover what to print
    if (is.null(what))
        if (x$splitMethod$internal.name=="noPlan")
            what="AppErr"
        else
            what <- paste(x$splitMethod$internal.name,"Err",sep="")
    # }}}
    # {{{ echo estimation splitMethod
    if (!is.null(x$splitMethod)) {
        cat("\n")
        print(x$splitMethod)}
    if (missing(times)){
        ## times <- x$maxtime ## times <- quantile(x$time,.9)
        times <- x$minmaxtime
        ##     naPos <- sapply(x[[what]],function(pec){
        ##       length(pec)-sum(is.na(pec))-1
        ##     })
        ##     times <- min(x$time[naPos],times)
    }
    # }}}
    # {{{ cumulative prediction errors
    ## cat("\n",paste(rep("_",options()$width),collapse=""),"\n")
    tnames <- paste("time=",round(times,1),sep="")
    tnames[times<1] <- paste("time=",signif(times[times<1],2),sep="")
    if (x$exact==TRUE){
        cat("\nCumulative prediction error, aka Integrated Brier score  (IBS)\n aka Cumulative rank probability score\n\nRange of integration:",x$start,"and",tnames,":\n\n")
        out <- crps(object=x,times=times,start=x$start,what=what)
        if (is.matrix(out))
            print(out,digits=digits,quote=FALSE)
        else{
            print.listof(out,digits=digits,quote=FALSE)
        }
    }
    else{
        sx <- summary(x,times=times,what=what,digits=digits,print=FALSE)
        if (length(sx)==1)
            print(round(sx[[1]],digits=digits))
        else{
            names(sx) <- c("AppErr"="Apparent prediction error","Boot632plusErr"="Boostrap .632+","Boot632Err"="Bootstrap .632","BootCvErr"="Bootstrap cross-validation","loocvErr"="Leave one out cross validation","crossval"="cross-validation","NoInfErr"="No information error")[names(sx)]
            cat("\nPrediction error at",tnames,":\n\n",sep="")
            print.listof(sx,digits=digits,quote=FALSE)
        }
    }
    # }}}
    # {{{ warn about failed computations
    failed <- sapply(x$failed,length)>0
    if (sum(failed)>0){
        cat("\nWarning: In some bootstrap samples the model failed:\n")
        print(sapply(x$failed[failed],table))
    }
    # }}}
    # {{{ echo test results
    if (!is.null(x$multiSplitTest)){
        cat("\n",paste(rep("_",options()$width/2),collapse=""),"\n")
        print(x$multiSplitTest)
    }
    # }}}
    invisible(x)
}