R/get_optimal_res.R

In jmlondon/crawlr: utilities and visualization functions in support of crawl

#' Get optimal land mask raster resolution
#'
#' the \code{crawl} package includes a function \code{fix_path} that will
#' adjust a movement path to move around a restricted area (e.g. land if
#' the animal is a fish or marine mammal). The resolution of this land
#' mask is an important parameter and, often, is related to the
#' spatial scale of the movement. Higher resolution has a significant cost
#' in compute time and memory requirements. So, this function uses a simple
#' rule to determine an optimal resolution for this land mask. The highest
#' resolution is 200 m per side for any region with a max dimension less
#' than or equal to 25 km. Above a max dimension of 25 km, the resolution
#' is determined by dividing the max dimension by 1000 and multiplying by 8.
#' That said, once the max dimension reaches 150km, the resolution is set to
#' a fixed value of 1.2km.
#'
#' @param pts a SpatialPoints, or SpatialPointsDataFrame
#'
#' @return numeric value indicating the square cell dimension in meters
#' @export
#'
get_optimal_res <- function(pts) {
  x_len <- diff(pts@bbox['mu.x',])[[1]]
  y_len <- diff(pts@bbox['mu.y',])[[1]]

  max_len <- max(x_len, y_len)

  res <- ifelse(max_len >= 25000, (max_len/1000) * 8, 200)
  res <- ifelse(max_len >= 150000, 1200, res)
  return(res)
}