Rust implementation of spectral clustering

Spectral clustering on a pre-calculated similarity matrix.

Usage

rs_spectral_clustering_sim(
  similarities,
  k_neighbours,
  n_clusters,
  max_iters,
  seed
)

Arguments

similarities

Numerical matrix representing the similarities. Needs to be symmetric!

k_neighbours

Integer. Number of neighbours to consider in the kNN graph generation

n_clusters

Integer. Number of clusters to identify

max_iters

Integer. Number of iterations for k-means clustering

seed

Integer. Seed for reproducibility

Value

A vector with the membership of the samples