stability_evaluation: Clustering stability evaluation

In vittoriofortino84/COPS: Clustering algorithms for Omics-driven Patient Stratification

Clustering stability evaluation

Description

Performs stability analysis on cross-validated clusterings.

Usage

stability_evaluation(
  clusters,
  by = c("datname", "drname", "run", "k", "m"),
  parallel = 1,
  reference_fold = NULL,
  ...
)

stability_evaluation_pac(
  clusters,
  by = c("datname", "drname", "k", "m"),
  parallel = 1,
  reference_fold = NULL,
  ...
)

Arguments

clusters	clustering data.frame such as returned by subsample_clustering_evaluation
by	vector of column names to keep
parallel	number of threads
reference_fold	fold number that corresponds to reference which other folds are compared against, inferred from input by default
...	extra arguments are ignored

Details

Default settings work with subsample_clustering_evaluation output 'clusters'.

Value

Returns a data.frame where each row corresponds to clustering stability with respect to kept column variables

Functions

• stability_evaluation_pac(): Stability evaluation with Proportion of Ambiguously Clustered pairs (PAC)

vittoriofortino84/COPS documentation built on Jan. 28, 2025, 3:16 p.m.