R/colSums-method.R

In pik-piam/magclass: Data Class and Tools for Handling Spatial-Temporal Data

#' ~~ Methods for Function colSums and colMeans ~~
#'
#' ~~ Methods for function \code{colSums} and \code{colMeans} ~~
#'
#'
#' @name colSums-methods
#' @aliases colSums-methods colSums,ANY-method colSums,magpie-method
#' colMeans-methods colMeans,ANY-method colMeans,magpie-method
#' @param x object on which calculation should be performed
#' @param na.rm logical. Should missing values (including NaN) be omitted from the calculations?
#' @param dims integer: Which dimensions are regarded as "rows" or "columns" to sum over. For row*,
#' the sum or mean is over dimensions dims+1, ...; for col* it is over dimensions 1:dims.
#' @param ... further arguments passed to other colSums/colMeans methods
#' @docType methods
#' @section Methods: \describe{
#'
#' \item{list("signature(x = \"ANY\")")}{ normal colSums and colMeans method }
#'
#' \item{list("signature(x = \"magpie\")")}{ classical method prepared to
#' handle MAgPIE objects } }
#' @keywords methods ~~ other possible keyword(s) ~~
#' @importFrom methods new
#' @exportMethod colSums
#'
setMethod("colSums",
  signature(x = "magpie"),
  function(x, na.rm = FALSE, dims = 1, ...) { #nolint
    xGlo <- colSums(as.array(x), na.rm = na.rm, ...)
    out <- new("magpie", array(xGlo, dim = c(1, dim(xGlo)), dimnames = c("GLO", dimnames(xGlo))))
    return(out)
  }
)