Chooses initial cluster locations using Kmeans++

Source: R/fit_assign.R

Chooses initial cluster locations using Kmeans++

init_centers(X, k)

Arguments

X

array Data points of dimension (n,d)
k

int The number of desired clusters

Value

array Initial coordinates of clusters

Examples

X = rbind(c(0,0), c(1,1))
init_centers(X, 2)

#>      [,1] [,2]
#> [1,]    1    1
#> [2,]    0    0