poly_overlap.R

poly_df <- read.csv(
  url("https://gist.github.com/will-r-chase/ac1c1b3ba96a761a7d8d70af860bf663/raw/97c68537ad35254721ed55b12b696af6e424cd4d/overlapping_polys.csv")
)

# geos approach
poly_geos <- geos::geos_make_polygon(
  poly_df$x, poly_df$y, 
  feature_id = poly_df$id
)

# to use sf, you could use st_as_sfc(poly_geos) and use st_intersects()
poly_sf <- sf::st_as_sf(poly_geos)

# set up loop variables
poly_geos_iter <- poly_geos

# st_intersects will give you the same object
poly_geos_intersections <- geos::geos_intersects_matrix(
  poly_geos_iter, 
  poly_geos_iter
)
max_iter <- 1000

for (i in seq_len(max_iter)) {
  n_intersections <- vapply(poly_geos_intersections, length, integer(1))
  if (all(n_intersections == 1)) {
    break;
  }
  
  poly_geos_iter <- poly_geos_iter[-which.max(n_intersections)]
  poly_geos_intersections <- geos::geos_intersects_matrix(
    poly_geos_iter, 
    poly_geos_iter
  )
}

plot(poly_geos)
plot(poly_geos_iter, border = "red", add = T)

poly_df <- read.csv(
  url("https://gist.github.com/will-r-chase/ac1c1b3ba96a761a7d8d70af860bf663/raw/97c68537ad35254721ed55b12b696af6e424cd4d/overlapping_polys.csv")
)

# geos approach
poly_geos <- geos::geos_make_polygon(
  poly_df$x, poly_df$y, 
  feature_id = poly_df$id
)

# to use sf, you could use st_as_sfc(poly_geos) and use st_intersects()
poly_sf <- sf::st_as_sf(poly_geos)

# set up loop variables
poly_geos_iter <- poly_geos

# st_intersects will give you the same object
poly_geos_intersections <- geos::geos_intersects_matrix(
  poly_geos_iter, 
  poly_geos_iter
)
max_iter <- 1000

for (i in seq_len(max_iter)) {
  n_intersections <- vapply(poly_geos_intersections, length, integer(1))
  if (all(n_intersections == 1)) {
    break;
  }
  
  poly_geos_iter <- poly_geos_iter[-which.max(n_intersections)]
  poly_geos_intersections <- geos::geos_intersects_matrix(
    poly_geos_iter, 
    poly_geos_iter
  )
}

plot(poly_geos)
plot(poly_geos_iter, border = "red", add = T)

^{Created on 2020-08-01 by the reprex package (v0.3.0)}