Convenience: compute HKS descriptors from an adjacency matrix — compute_hks_from_adjacency • manifoldalign

Convenience: compute HKS descriptors from an adjacency matrix

Usage

compute_hks_from_adjacency(
  A,
  k_embed = 30L,
  q = 16L,
  use_normalized_laplacian = TRUE,
  store_basis = FALSE,
  ...
)

Arguments

A: Square adjacency/affinity matrix (dense or sparse).
k_embed: Number of eigenpairs used for the descriptor basis.
q: Number of HKS time steps.
use_normalized_laplacian: Logical; use the normalized Laplacian (default TRUE).
store_basis: Logical; if TRUE, return `list(descriptors=, basis=)`.
...: Passed to [compute_hks_descriptors()] (e.g., time grid settings).

Value

A numeric matrix (n x q), or a list when `store_basis=TRUE`.