R/fmeasure.R

In EGAD: Extending guilt by association by degree

#' Fmeasure of precision-recall
#'
#' The function calculates fmeasure for a given beta of a precision-recall curve 
#' 
#' @param recall numeric array
#' @param precis numeric array
#' @param beta numeric value, default is 1 
#'
#' @return fmeasure Numeric value 
#' 
#' @keywords precision-recall fmeasure
#' 
#' @examples 
#' labels <- c(rep(0,10))
#' labels[c(1,3,5)] <- 1 
#' scores <- 10:1
#' prc <- get_prc(scores, labels)
#' fm <- fmeasure(prc[,1], prc[,2])
#'
#' @export
#'
fmeasure <- function(recall, precis, beta = 1) {
    return(((beta^2 + 1) * precis * recall)/(beta^2 * precis + recall))
}