Assign points to a fixed number of clusters

Splits the points in the array into k clusters by nearest centre. One label per row, in the order you gave the points.

Usage

ga_kmeans(geometry, k, seed = NULL)

Arguments

geometry: a GeoArrow point array
k: how many clusters to produce
seed: a seed for reproducible starts, or NULL to vary each run

Value

an integer array of cluster labels, the same length as geometry

Details

k-means always produces exactly k clusters and no noise, so it labels every point. It favours round, similarly sized clusters; use ga_dbscan() when the shapes are irregular or the count is unknown.

ga_kmeans() clusters over the whole array, not within each row, so k is a single value rather than one per row. The algorithm starts from a random seed, so ga_kmeans() can return different results on each run unless you give seed. It leaves a row that is not a single point, or is null, out of the clustering and returns null for it.

References

KMeans

Examples

g <- ga_point(
  c(0, 0.1, 0.2, 5, 5.1, 5.2),
  c(0, 0.1, 0.2, 5, 5.1, 5.2)
)

as.vector(ga_kmeans(g, k = 2, seed = 1))
#> [1] 1 1 1 2 2 2