Skip to contents

k-means initialisation

Source: R/init.R
init_kmeans.Rd

Runs stats::kmeans() on the supplied samples and uses the resulting cluster centres and within-cluster covariances to seed an EM-style fitter.

Usage

init_kmeans(samples, N = 2L, ridge_eps = 1e-06, nstart = 10L)

Arguments

samples

An n by p numeric matrix of samples from (or close to) the target.

N

Number of components.

ridge_eps

Ridge added to each cluster covariance for numerical stability when a cluster has fewer than two points.

nstart

stats::kmeans nstart argument.

Value

A gmm of N components in dimension ncol(samples).

See also

Other init: init_moment_seed(), init_random(), init_warm_start(), multi_start_best_of()

Examples

x <- matrix(stats::rnorm(200), ncol = 2)
init_kmeans(x, N = 3L)
#> <init_kmeans>: K = 3 components in p = 2 dimensions
#>   [1] w = 0.2600, |mu| = 1.4362, tr(Sigma) = 1.0920
#>   [2] w = 0.4600, |mu| = 0.7757, tr(Sigma) = 0.7767
#>   [3] w = 0.2800, |mu| = 1.0508, tr(Sigma) = 0.7932