Raster brick from raster masks by sectors (R)

Question

The goal is to divide a raster file into quadrants in order to make some computation in each. I'd like to split the raster and create a raster brick with those parts. I don't need to get perfect clips of the raster so the cell with data in the border of the clip could be also present in the contiguous raster brick.

I'm thinking about creating a radial grid and then extract the raster by masking it with each of the polygons within. I found these interesting questions about the topic.

• How to split circles in 12 sections in PostGIS?

• How to split circles in 8 quadrants

• How to split a circle in equal parts in QGIS?

How can be done this in R ?

Here you have the layer (dem) to tryout.

• Note 1: number of quadrants must be a parameter.
• Note 2: raster brick must have same extension as the original raster.
• Note 3: process must be done in R (a function would be perfect)

Update 10/06/2020: I hope this schema clarifies the purpose: A Function that given a raster and a number of sectors it creates a raster brick (with number of rasters = number of sectors) where each raster keeps just the values within the correspondent sector.

Answer 1

EDIT: Changed the function so it can now generate number of slices other than multiples of 4.

I've come up with a solution using the sf package and rotating points around the raster. Here's what a slice looks like.

It works perfectly, you can test it out yourself:

#' Sector Masking
#' @description Slices Raster object into sections
#'
#' @param raster Raster* object
#' @param segments Number of segments to be sliced
#'
#' @return RasterBrick containing slices in bands
#'
#' @import raster
#' @import sf
#' @import magrittr
#' 
sector_masking <- function(raster, segments){
  library(raster)
  library(sf)
  library(magrittr)
  
  # Get properties of raster
  dem <- brick(raster)
  ext <- extent(dem)
  crs <- as.character(crs(dem))
  
  # Calculate mid and max point of the raster
  half_x <- (ext@xmax - ext@xmin) / 2
  half_y <- (ext@ymax - ext@ymin) / 2
  middle <- st_point(c(ext@xmin+half_x, ext@ymin+half_y))
  max_ext <- st_point(c(ext@xmax, ext@ymax))
  
  # Rotates a starting point around the middle of the raster
  # These points are used to generate triangles which are masking the raster
  slices <- list()
  starting_point <- max_ext
  rot <- function(a) matrix(c(cos(a), sin(a), -sin(a), cos(a)), 2, 2)
  rotation <- segments/2
  for(i in 1:segments){
    next_point <- (starting_point - middle) * rot(pi/rotation) + middle
    slice_matrix <- rbind(middle,
                          starting_point,
                          next_point,
                          middle)
    slice_polygon <- slice_matrix %>% 
      list() %>% 
      st_polygon() %>% 
      st_sfc(crs = crs) %>% 
      st_sf
    slice_raster <- rasterize(slice_polygon, dem, mask = T)
    slices[[i]] <- slice_raster
    starting_point <- next_point
  }
  return(stack(slices))
}