Cluster Specific Similarities Per Omics Block

Usage

cluster_specific_similarity(
  shared_specific,
  k = NULL,
  method = c("Spectral", "PAM", "Proximity", "Enhanced_Proximity"),
  prox_method_cl = c("PAM", "Spectral"),
  tune_method = "silhouette",
  gap_w = "uniform",
  similarity_type = c("second", "first"),
  ...
)

Arguments

shared_specific: Output from get_shared_specific_weights().
k: Optional specific-clustering k configuration. NULL tunes each block; single integer uses same k for all blocks; named numeric/list allows block-specific k.
method: Clustering backend for each specific block. Choose from "Spectral", "PAM", "Proximity", and "Enhanced_Proximity".
prox_method_cl: Proximity clustering backend used when method is "Proximity" or "Enhanced_Proximity".
tune_method: Tuning criterion for PAM when block-level k is NULL.
gap_w: Weighting scheme for spectral eigengap when block-level k is NULL.
similarity_type: Similarity structure for each specific block.
...: Additional arguments passed to clustering backends.

Value

A list with per-omics specific similarities and clustering outputs.