CODEWITHSUNDEEP
algorithmrinitializationk-means
CodeKingPlusPlus
CodeKingPlusPlus

Reputation: 16081

R kmeans initialization

In the R programming environment, I am currently using the standard implementation of the kmeans algorithm (type: help(kmeans)). It appears that I cannot initialize the starting centroids. I specify the kmeans algorithm to give me 4 clusters and I would like to pass the vector coordinates of the starting centroids.

  1. Is there an implementation of kmeans to allow me to pass initial centroid coordinates?

Upvotes: 3

Views: 3960

Answers (1)

Dason
Dason

Reputation: 61913

Yes. The implementation you mention allows you to specify starting positions. You pass them in through the centers parameter

> dat <- data.frame(x = rnorm(99, mean = c(-5, 0 , 5)), y = rnorm(99, mean = c(-5, 0, 5)))
> plot(dat)
> start <- matrix(c(-5, 0, 5, -5, 0, 5), 3, 2)
> kmeans(dat, start)
K-means clustering with 3 clusters of sizes 33, 33, 33

Cluster means:
           x           y
1 -5.0222798 -5.06545689
2 -0.1297747 -0.02890204
3  4.8006581  5.00315151

Clustering vector:
 [1] 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2
[51] 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3

Within cluster sum of squares by cluster:
[1] 58.05137 73.81878 52.45732
 (between_SS / total_SS =  94.7 %)

Available components:

[1] "cluster"      "centers"      "totss"        "withinss"     "tot.withinss" "betweenss"   
[7] "size"

Upvotes: 6

Related Questions