Skip to contents

Parallel Assessment of Clustering Stability Using Bootstrap Resampling

Source: R/auto_tuning.R
assess_clustering_stability_parallel.Rd

Parallel version of assess_clustering_stability that processes bootstrap samples concurrently for significant performance improvements. Uses foreach/doParallel for cross-platform compatibility.

Usage

assess_clustering_stability_parallel(
  mx,
  small.cell.reduction,
  cut.off = 1,
  n_bootstrap = 50,
  sample_fraction = 0.8,
  segment.levels = 3,
  n_cores = NULL,
  parallel = TRUE,
  seed = NULL,
  verbose = FALSE
)

Arguments

mx

Mobility matrix for analysis.

small.cell.reduction

Small cell reduction parameter to test.

cut.off

Edge weight threshold for network construction. Default is 1.

n_bootstrap

Number of bootstrap samples to generate. Default is 50.

sample_fraction

Fraction of data to use in each bootstrap sample. Default is 0.8.

segment.levels

Number of hierarchical levels for clustering. Default is 3.

n_cores

Number of CPU cores to use. NULL (default) uses all available cores - 1.

parallel

Logical indicating whether to use parallel processing. Default is TRUE.

seed

Random seed for reproducibility. Default is NULL.

verbose

Logical indicating whether to show progress messages. Default is FALSE.

Value

Numeric stability score between 0 and 1, where higher values indicate more stable clustering results.

Details

This function parallelizes the bootstrap sampling process: - Distributes bootstrap samples across available cores - Each core independently runs moneca_fast on its assigned samples - Results are combined to compute overall stability - Falls back to sequential processing if parallel packages unavailable

Performance improvements: - With 4 cores: ~3.5x faster than sequential - With 8 cores: ~6-7x faster than sequential - Scales linearly with number of cores up to n_bootstrap

See also

assess_clustering_stability, auto_tune_small_cell_reduction

Examples

if (FALSE) { # \dontrun{
# Generate sample data
mobility_data <- generate_mobility_data(n_classes = 6, seed = 123)

# Parallel stability assessment (auto-detect cores, default n_bootstrap = 50)
stability_parallel <- assess_clustering_stability_parallel(
  mobility_data,
  small.cell.reduction = 5,
  parallel = TRUE
)

# Sequential for comparison
stability_seq <- assess_clustering_stability_parallel(
  mobility_data,
  small.cell.reduction = 5,
  parallel = FALSE
)

# For more thorough assessment, increase n_bootstrap
stability_thorough <- assess_clustering_stability_parallel(
  mobility_data,
  small.cell.reduction = 5,
  n_bootstrap = 100,
  parallel = TRUE
)
} # }