EVoC clustering from pre-computed kNN — rs_evoc_from

Wrapper function into the Rust interface for EVoC clustering. This version uses a pre-computed kNN graph, please see new_nearest_neighbour().

Usage

rs_evoc_from_knn(
  embd,
  knn_data,
  n_neighbours,
  evoc_params,
  seed,
  use_high_precision,
  verbose
)

embd: Numerical matrix. The data to cluster. Should be of dimensions samples x features.
knn_data: List. A NearestNeighbours object with k, indices, and dist elements.
n_neighbours: Integer. Number of nearest neighbours for graph construction.
evoc_params: Named list. List that contains all of the key parameters for EVoC clustering.
seed: Integer. Seed for reproducibility.
use_high_precision: Optional logical. Controls fp32 vs fp64 for. If NULL will use sensible default thresholding.
verbose: Integer. If 0L -> silent or 1L for normal verbosity; 2L for detailed verbosity.

A named list with cluster layers, membership strengths, persistence scores, and the kNN graph.