R/logscale.R

In fractalRegression: Performs Fractal Analysis and Fractal Regression

#' logscale
#' 
#' Create logarithmically spaced scales
#' 
#' @param scale_min an integer indicating the minimum scale to be resovled
#' @param scale_max an integer indicating the maximum scale to be resolved
#' @param scale_ratio a double indicating the ratio by which scale successive 
#' scales. For example, scale_ratio = 2 would create a scales increasing by 
#' a power of 2.
#' 
#' @return A vector of of logarithmically spaced scales. 
#' 
#' @examples
#' scales <- logscale(scale_min = 16, scale_max = 1024, scale_ratio = 2)
#' 
#' @export
logscale = function(scale_min, scale_max, scale_ratio){
  number_of_scales =  ceiling(log(scale_max/scale_min)/log(scale_ratio));
  scales = rep(NA, number_of_scales)
  scales[1] = scale_min;
  for (i in 2:(number_of_scales+1)){
    scales[i] = scales[i-1]*scale_ratio
  }
  scales <- unique(floor(scales))
  if (max(scales) == scale_max){
  return(scales) 
  } else {
  return(scales[scales<=scale_max])
  }
  
}