R/metrics.r

In vcerqueira/tsensembler: Dynamic Ensembles for Time Series Forecasting

#' Computing the squared error
#'
#' Utility function to compute pointwise squared error (SE)
#'
#' @param y A numeric vector representing the actual values.
#' @param y_hat A numeric vector representing the forecasted values.
#'
#' @return squared error of forecasted values.
#'
#' @family error/performance functions
#'
#' @keywords internal
#'
#' @export
se <- function(y, y_hat) {
  stopifnot(length(y) == length(y_hat),
            is.numeric(y),
            is.numeric(y_hat))

  (y - y_hat) ^ 2
}

#' Computing the mean squared error
#'
#' Utility function to compute mean squared error (MSE)
#'
#' @inheritParams se
#'
#' @family error/performance functions
#'
#' @keywords internal
#'
#' @export
mse <- function(y, y_hat) mean(se(y, y_hat), na.rm = TRUE)

#' Computing the root mean squared error
#'
#' Utility function to compute Root Mean Squared Error (RMSE)
#'
#' @inheritParams se
#'
#' @family error functions
#'
#' @keywords internal
#'
#' @export
rmse <- function(y, y_hat) sqrt(mse(y, y_hat))

#' Computing the absolute error
#'
#' Element-wise computation of the absolute error loss function.
#'
#' @inheritParams se
#'
#' @family error/performance functions
#'
#' @keywords internal
#'
#' @export
ae <- function(y, y_hat) {
  stopifnot(length(y) == length(y_hat),
            is.numeric(y),
            is.numeric(y_hat))

  abs(y - y_hat)
}