Test Clustering Method on Synthetic Data

Runs a clustering method on synthetic ground truth data and reports accuracy metrics.

Usage

test_clustering_method(
  method,
  noise_sd = 0,
  n_clusters = 27,
  seed = 42,
  verbose = TRUE,
  ...
)

Arguments

method

Clustering method name (e.g., "supervoxels", "snic", "flash3d")

noise_sd

Noise level for synthetic data. Default 0 (perfect separability).

n_clusters

Number of ground truth clusters (must be perfect cube). Default 27.

seed

Random seed for reproducibility.

verbose

Print results. Default TRUE.

...

Additional arguments passed to cluster4d()

Value

A list with:

metrics

Output from clustering_accuracy()

result

Full clustering result object

data

Synthetic data used

time

Elapsed time in seconds

Examples

if (FALSE) { # \dontrun{
# Test FLASH-3D with 27 cubic clusters
test_clustering_method("flash3d")

# Test with 8 clusters
test_clustering_method("slic", n_clusters = 8)

# Test with noise
test_clustering_method("snic", noise_sd = 0.5)
} # }