R/PCC_method.r

In andregustavom/mlquantify: Algorithms for Class Distribution Estimation

#' Probabilistic Classify and Count
#'
#' It quantifies events based on testing scores, applying the Probabilistic Classify
#' and Count (PCC) method.
#' @param test a numeric \code{vector} containing the score estimated for the positive class from
#' each test set instance. (NOTE: It requires calibrated scores. See \link[CORElearn]{calibrate}
#' from \pkg{CORElearn}).
#' @return A numeric vector containing the class distribution estimated from the test set.
#' @usage PCC(test)
#' @references Bella, A., Ferri, C., Hernández-Orallo, J., & Ramírez-Quintana,
#' M. J. (2010). Quantification via probability estimators. In IEEE International
#' Conference on Data Mining (pp. 737–742). Sidney.<doi.org/10.1109/ICDM.2010.75>.
#' @export
#' @examples
#' library(randomForest)
#' library(caret)
#' cv <- createFolds(aeAegypti$class, 3)
#' tr <- aeAegypti[cv$Fold1,]
#' validation <- aeAegypti[cv$Fold2,]
#' ts <- aeAegypti[cv$Fold3,]
#'
#' # -- Getting a sample from ts with 80 positive and 20 negative instances --
#' ts_sample <- rbind(ts[sample(which(ts$class==1),80),],
#'                    ts[sample(which(ts$class==2),20),])
#' scorer <- randomForest(class~., data=tr, ntree=500)
#' scores <- cbind(predict(scorer, validation, type = c("prob")), validation$class)
#' test.scores <- predict(scorer, ts_sample, type = c("prob"))[,1]
#'
#' # -- PCC requires calibrated scores. Be aware of doing this before using PCC --
#' # -- You can make it using calibrate function from the CORElearn package --
#' # if(requireNamespace("CORElearn")){
#' #   cal_tr <- CORElearn::calibrate(as.factor(scores[,3]), scores[,1], class1=1,
#' #   method="isoReg",assumeProbabilities=TRUE)
#' #   test.scores <- CORElearn::applyCalibration(test.scores, cal_tr)
#' # }
#' PCC(test=test.scores)
PCC <- function(test){

  result <- mean(test)
  if(result < 0 ) result <- 0
  if(result > 1 ) result <- 1
  result <- c(result, 1 - result)
  names(result) <- c("+", "-")
  return(result)
}