R/plot.gg_minimal_vimp.R

In ggRandomForests: Visually Exploring Random Forests

####**********************************************************************
####**********************************************************************
####
####  ----------------------------------------------------------------
####  Written by:
####  ----------------------------------------------------------------
####    John Ehrlinger, Ph.D.
####
####    email:  john.ehrlinger@gmail.com
####    URL:    https://github.com/ehrlinger/ggRandomForests
####  ----------------------------------------------------------------
####
####**********************************************************************
####**********************************************************************
#'
#' Plot a \code{\link{gg_minimal_vimp}} object for comparing the Minimal
#' Depth and VIMP variable rankings.
#'
#' @param x \code{\link{gg_minimal_depth}} object created from a
#' \code{\link[randomForestSRC]{var.select}}
#' object
#' @param nvar should the figure be restricted to a subset of the points.
#' @param lbls a vector of alternative variable names.
#' @param ... optional arguments (not used)
#'
#' @return \code{ggplot} object
#'
#'
#' @importFrom ggplot2 ggplot aes_string geom_point labs geom_abline coord_flip
#'  scale_x_discrete
#'
#' @seealso \code{\link{gg_minimal_vimp}}
#' \code{\link[randomForestSRC]{var.select}}
#'
#' @examples
#' \dontrun{
#' ## Examples from RFSRC package...
#' ## ------------------------------------------------------------
#' ## classification example
#' ## ------------------------------------------------------------
#' ## -------- iris data
#' ## You can build a randomForest
#' rfsrc_iris <- rfsrc(Species ~ ., data = iris)
#' varsel_iris <- var.select(rfsrc_iris)
#'
#' # Get a data.frame containing minimaldepth measures
#' gg_dta<- gg_minimal_vimp(varsel_iris)
#'
#' # Plot the gg_minimal_depth object
#' plot(gg_dta)
#'
#' ## ------------------------------------------------------------
#' ## Regression example
#' ## ------------------------------------------------------------
#' ## -------- air quality data
#' rfsrc_airq <- rfsrc(Ozone ~ ., data = airquality, na.action = "na.impute")
#' varsel_airq <- var.select(rfsrc_airq)
#'
#' # Get a data.frame containing error rates
#' gg_dta<- gg_minimal_vimp(varsel_airq)
#'
#' # Plot the gg_minimal_vimp object
#' plot(gg_dta)
#'
#' ## -------- Boston data
#' data(Boston, package="MASS")
#' rfsrc_boston <- randomForestSRC::rfsrc(medv~., Boston)
#' 
#' varsel_boston <- var.select(rfsrc_boston)
#' 
#' # Get a data.frame containing error rates
#' gg_dta<- gg_minimal_vimp(varsel_boston)
#'
#' # Plot the gg_minimal_vimp object
#' plot(gg_dta)
#'
#' ## -------- mtcars data
#' rfsrc_mtcars <- rfsrc(mpg ~ ., data = mtcars)
#' varsel_mtcars <- var.select(rfsrc_mtcars)
#'
#' # Get a data.frame containing error rates
#' gg_dta<- gg_minimal_vimp(varsel_mtcars)
#'
#' # Plot the gg_minimal_vimp object
#' plot(gg_dta)
#'
#' ## ------------------------------------------------------------
#' ## Survival example
#' ## ------------------------------------------------------------
#' ## -------- veteran data
#' ## randomized trial of two treatment regimens for lung cancer
#' data(veteran, package = "randomForestSRC")
#' rfsrc_veteran <- rfsrc(Surv(time, status) ~ ., data = veteran, ntree = 100)
#' varsel_veteran <- var.select(rfsrc_veteran)
#' 
#' gg_dta <- gg_minimal_vimp(varsel_veteran)
#' plot(gg_dta)
#'
#' ## -------- pbc data
#' # We need to create this dataset
#' data(pbc, package = "randomForestSRC",) 
#' # For whatever reason, the age variable is in days... makes no sense to me
#' for (ind in seq_len(dim(pbc)[2])) {
#'  if (!is.factor(pbc[, ind])) {
#'    if (length(unique(pbc[which(!is.na(pbc[, ind])), ind])) <= 2) {
#'      if (sum(range(pbc[, ind], na.rm = TRUE) == c(0, 1)) == 2) {
#'        pbc[, ind] <- as.logical(pbc[, ind])
#'      }
#'    }
#'  } else {
#'    if (length(unique(pbc[which(!is.na(pbc[, ind])), ind])) <= 2) {
#'      if (sum(sort(unique(pbc[, ind])) == c(0, 1)) == 2) {
#'        pbc[, ind] <- as.logical(pbc[, ind])
#'      }
#'      if (sum(sort(unique(pbc[, ind])) == c(FALSE, TRUE)) == 2) {
#'        pbc[, ind] <- as.logical(pbc[, ind])
#'      }
#'    }
#'  }
#'  if (!is.logical(pbc[, ind]) &
#'      length(unique(pbc[which(!is.na(pbc[, ind])), ind])) <= 5) {
#'    pbc[, ind] <- factor(pbc[, ind])
#'  }
#' }
#' #Convert age to years
#' pbc$age <- pbc$age / 364.24
#'
#' pbc$years <- pbc$days / 364.24
#' pbc <- pbc[, -which(colnames(pbc) == "days")]
#' pbc$treatment <- as.numeric(pbc$treatment)
#' pbc$treatment[which(pbc$treatment == 1)] <- "DPCA"
#' pbc$treatment[which(pbc$treatment == 2)] <- "placebo"
#' pbc$treatment <- factor(pbc$treatment)
#' dta_train <- pbc[-which(is.na(pbc$treatment)), ]
#' # Create a test set from the remaining patients
#' pbc_test <- pbc[which(is.na(pbc$treatment)), ]
#'
#' #========
#' # build the forest:
#' rfsrc_pbc <- randomForestSRC::rfsrc(
#'   Surv(years, status) ~ .,
#'  dta_train,
#'  nsplit = 10,
#'  na.action = "na.impute",
#'  forest = TRUE,
#'  importance = TRUE,
#'  save.memory = TRUE
#' )
#' 
#' varsel_pbc <- var.select(rfsrc_pbc)
#'
#' gg_dta <- gg_minimal_vimp(varsel_pbc)
#' plot(gg_dta)
#' }
#'
#' @export
plot.gg_minimal_vimp <- function(x, nvar, lbls, ...) {
  gg_dta <- x
  
  # Test that object is the correct class object
  if (!inherits(gg_dta, "gg_minimal_vimp")) {
    gg_dta <- gg_minimal_vimp(x, ...)
  }
  
  if (missing(nvar))
    nvar <- nrow(gg_dta)
  if (nvar > nrow(gg_dta))
    nvar <- nrow(gg_dta)
  if (length(unique(gg_dta$col)) > 1) {
    gg_dta$col <- factor(gg_dta$col)
  }
  gg_dta$names <- factor(gg_dta$names, 
                         levels = gg_dta$names[order(as.numeric(gg_dta$depth))])
  
  gg_dta <- gg_dta[1:nvar, ]
  
  # If we only have one class for coloring, just paint them black.
  if (length(unique(gg_dta$col)) > 1) {
    gg_plt <-
      ggplot(gg_dta, aes_string(x = "names", y = "vimp", col = "col")) +
      labs(x = "Minimal Depth (Rank Order)", y = "VIMP Rank", color = "VIMP")
  } else {
    gg_plt <- ggplot(gg_dta, aes_string(x = "names", y = "vimp")) +
      labs(x = "Minimal Depth (Rank Order)", y = "VIMP Rank")
  }
  if (!missing(lbls)) {
    if (length(lbls) >= length(gg_dta$names)) {
      st_lbls <- lbls[as.character(gg_dta$names)]
      names(st_lbls) <- as.character(gg_dta$names)
      st_lbls[which(is.na(st_lbls))] <-
        names(st_lbls[which(is.na(st_lbls))])
      
      gg_plt <- gg_plt +
        scale_x_discrete(labels = st_lbls)
    }
  }
  
  gg_plt <- gg_plt + geom_point() +
    geom_abline(
      intercept = 0,
      slope = 1,
      col = "red",
      size = .5,
      linetype = 2
    )
  
  # Draw a line between + and - vimp values.
  if (length(unique(gg_dta$col)) > 1) {
    gg_plt <- gg_plt +
      geom_hline(
        yintercept = sum(gg_dta$col == "+") + .5,
        col = "red",
        size = .5,
        linetype = 2
      )
  }
  
  if (nrow(gg_dta) > attributes(gg_dta)$modelsize) {
    gg_plt <- gg_plt +
      geom_vline(
        xintercept = attributes(gg_dta)$modelsize + .5,
        col = "red",
        size = .5,
        linetype = 2
      )
  }
  
  gg_plt + coord_flip()
  
}