R/binary_bbrier.R

In mlr3measures: Performance Measures for 'mlr3'

#' @title Binary Brier Score
#'
#' @details
#' The Binary Brier Score is defined as \deqn{
#'    \frac{1}{n} \sum_{i=1}^n w_i (I_i - p_i)^2,
#' }{
#'    weighted.mean(((t == positive) - p)^2, w),
#' }
#' where \eqn{w_i} are the sample weights,
#'  and \eqn{I_{i}} is 1 if observation \eqn{x_i} belongs to the positive class, and 0 otherwise.
#'
#' Note that this (more common) definition of the Brier score is equivalent to the
#' original definition of the multi-class Brier score (see [mbrier()]) divided by 2.
#'
#' @templateVar mid bbrier
#' @template binary_template
#'
#' @references
#' \url{https://en.wikipedia.org/wiki/Brier_score}
#'
#' `r format_bib("brier_1950")`
#'
#' @inheritParams binary_params
#' @template binary_example
#' @export
bbrier = function(truth, prob, positive, sample_weights = NULL, ...) {
  assert_binary(truth, prob = prob, positive = positive)
  wmean(.se(truth == positive, prob), sample_weights)
}

#' @include measures.R
add_measure(bbrier, "Binary Brier Score", "binary", 0, 1, TRUE, obs_loss = "se_binary")


#' @title Binary Squared Error
#'
#' @description
#' Observation-wise loss function for [bbrier()].
#'
#' @export
#' @keywords internal
se_binary = function(truth, prob, positive, ...) {
  assert_binary(truth, prob = prob, positive = positive)
  ((truth == positive) - prob)^2
}