Generate groups based on balanced clustering — par_group

For balancing computational loads, the function uses the anticlust package to cluster the input points. The number of clusters is determined by the num_cluster argument. Each cluster will have equal number of points. Grids will be generated based on the cluster extents. At the lower level, the function uses terra::distance() function to calculate the Euclidean distance between points.

Usage

par_group_balanced(points_in = NULL, n_clusters = NULL)

Arguments

points_in: sf or SpatVector object. Target points of computation.
n_clusters: integer(1). The number of clusters.

Value

SpatVector object with a field "CGRIDID".

Note

This function is only for two-dimensional points. The results will be inexhaustive grids.

Author

Insang Song

Examples

library(terra)
library(anticlust)
data(ncpoints, package = "chopin")
ncp <- terra::vect(
  ncpoints, geom = c("X", "Y"),
  keepgeom = FALSE, crs = "EPSG:5070"
)
# 2,304 points / 12 = 192 points per cluster
ncpbal <- par_group_balanced(ncp, 12)
ncpbal
#>  class       : SpatVector 
#>  geometry    : points 
#>  dimensions  : 2304, 1  (geometries, attributes)
#>  extent      : 1055215, 1830766, 1356003, 1686298  (xmin, xmax, ymin, ymax)
#>  coord. ref. : NAD83 / Conus Albers (EPSG:5070) 
#>  names       : CGRIDID
#>  type        :   <num>
#>  values      :       1
#>                      2
#>                      1