Hierarchical Clustering with Optional PCA Dimensionality Reduction

This function performs hierarchical clustering on the input data matrix, optionally reducing dimensionality with PCA first. Supports different distance and linkage methods.

Usage

cluster_k_hc(
  data,
  k,
  pca = TRUE,
  n_pcs = 10,
  dist_method = "euclidean",
  hc_method = "complete"
)

Arguments

data

A numeric matrix or data frame (features x samples). Rows are features (e.g., genes), columns are samples.

k

Integer. The number of clusters to cut the hierarchical tree into. Must be a positive integer.

pca

Logical. If TRUE (default), perform PCA before clustering.

n_pcs

Integer. Number of principal components to use if `pca = TRUE`. Default is 10. Must be a positive integer.

dist_method

Distance method to use: "euclidean" (default), "pearson", "spearman", etc. Should be one of the supported methods.

hc_method

Linkage method for hierarchical clustering. Default is "complete". Should be one of the supported methods.

Value

A named integer vector with cluster assignments for each sample.