GPU-aware k-means clustering

GPU-aware k-means clustering

Usage

cuda_kmeans(
  x,
  centers,
  iter.max = 100L,
  tolerance = 1e-06,
  seed = NULL,
  batch_size = 256L,
  device = c("auto", "cuda", "cpu")
)

Arguments

x

Numeric matrix with observations in rows.

centers

Number of clusters or a matrix of initial centres.

iter.max

Maximum Lloyd iterations.

tolerance

Convergence tolerance for centre movement.

seed

Optional random seed used for initial centres.

batch_size

Maximum number of observations whose centre-distance block is materialized at once. The native backend keeps observations, centres, assignments, and updates on the GPU while bounding temporary distance storage to approximately batch_size * n_centers values.

device

Device used for the numerical clustering stages.

Value

A cuda_kmeans list containing integer cluster assignments, final centers, per-cluster withinss, tot.withinss, the number of iteration count in iter, a logical converged flag, and the actual distance device. Observation and feature names are retained when supplied.

Details

The native CUDA backend uploads the observations and initial centres once, then keeps distance calculation, deterministic assignment, accumulation, and centre updates on the device. Only the small convergence movement summary is inspected between iterations; final assignments, centres, and within-cluster sums are transferred to R. Compatibility backends without a resident k-means operation retain the established distance-on-backend and update-on-CPU implementation.

Examples

set.seed(1)
x <- rbind(matrix(rnorm(40), 20, 2), matrix(rnorm(40, 4), 20, 2))
cuda_kmeans(x, centers = 2, seed = 1, device = "cpu")
#> <cuda_kmeans clusters=2 iterations=2 converged=TRUE distance_device=cpu compute=cpu backend=base>
#>           [,1]         [,2]
#> [1,] 0.1905239 -0.006471519
#> [2,] 4.1387968  4.101736906