R/tileEquivalence.R

In bogdanoancea/deduplication: R package for computing the duplicity probility for each mobile device.

#' @title Transforms the tiles indices from the notation used by the simulation 
#' software to the one used by the \pkg{raster} package.
#'
#' @description In order to perform the population estimations, the area of 
#' interest is overlapped with a rectangular grid of tiles. Each tile is a 
#' rectangle with predefined dimensions. This function is a utility function 
#' which transform the tiles indexes from the numbering system used by the 
#' simulation software to the one used by the \pkg{raster} package. 
#' The simulation software uses a notation where the tile with index 0 is the 
#' bottom left tile while the \pkg{raster} package uses another way to number 
#' the tiles, tiles being numbered starting with 1 for the upper left tile.
#'
#' @param nrow Number of rows in the grid overlapping the area of interest.
#'
#' @param ncol Number of columns in the grid overlapping the area of interest.
#'
#' @return Returns a data.frame object with two columns: on the first column are 
#' the tile indexes according to the \pkg{raster} package numbering and on the 
#' second column are the equivalent tile indexes according to the simulation
#' software numbering.
#'
#'
#' @export
tileEquivalence <- function(nrows, ncols){
  
  if (nrows <= 0 )
    stop("nrows should be a positive number")
  if (ncols <= 0 )
    stop("ncols should be a positive number")
  
  nCells <- nrows * ncols
  
  tileID_raster <- as.vector(matrix(1:nCells, ncol = ncols, byrow = TRUE))
  tileID_simulator <- as.vector(matrix(0:(nCells - 1), ncol = ncols, byrow = TRUE)[nrows:1, ])
  
  tileCorresp <- cbind(rasterCellID = tileID_raster, tileID = tileID_simulator)
  tileCorresp <- tileCorresp[order(tileCorresp[, 'rasterCellID']), ]
  colnames(tileCorresp)<-c('rasterCell', 'tile')
  return(tileCorresp)
}