Spectral clustering via kernlab

Thin extender over kernlab::specc that performs Ng / Jordan / Weiss spectral clustering on a feature matrix.

Usage

morie_spectral_cluster(x, centers, ...)

Arguments

x

Numeric matrix or data frame of features (rows = observations).

centers

Integer; the number of clusters to extract.

...

Further arguments forwarded to kernlab::specc (e.g. kernel, kpar, nystrom.red, iterations).

Value

A list with $method = "kernlab::specc" and $raw (a specc S4 object containing the cluster assignments, centres and within-cluster sums of squares).

Examples

# \donttest{
  if (requireNamespace("kernlab", quietly = TRUE)) {
    set.seed(1)
    x <- rbind(
      matrix(stats::rnorm(80, mean = -2), ncol = 2),
      matrix(stats::rnorm(80, mean =  2), ncol = 2)
    )
    morie_spectral_cluster(x, centers = 2)
  }
#> $method
#> [1] "kernlab::specc"
#> 
#> $raw
#> Spectral Clustering object of class "specc" 
#> 
#>  Cluster memberships: 
#>  
#> 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 
#>  
#> Gaussian Radial Basis kernel function. 
#>  Hyperparameter : sigma =  4.6282344288711 
#> 
#> Centers:  
#>           [,1]      [,2]
#> [1,]  2.116840  1.689608
#> [2,] -1.907974 -1.879733
#> 
#> Cluster size:  
#> [1] 40 40
#> 
#> Within-cluster sum of squares:  
#> [1] 72.28717 64.30232
#> 
#> 
# }