Package index • FiberMargin

Minimal interfaces for regular masks and irregular spatial fields.

clean_categorical_mask(): Clean a noisy categorical mask
refine_spatial_labels(): Refine noisy categorical labels on a spatial domain

Mask Evaluation

Reproducible mask corruptions and damage-aware evaluation.

corrupt_categorical_mask(): Generate a corrupted categorical mask
evaluate_mask_cleaning(): Evaluate categorical mask cleaning

Simulation

Reproducible planar, gradient, junction, and volumetric generators.

simulate_spatial_clusters(): Simulate labelled spatial clusters
simulate_spatial_domains(): Simulate spatial tissue domains with challenging geometry
simulate_gradient_regions(): Simulate layered gradient-mixture tissue regions
simulate_complex_spatial_domains(): Simulate held-out complex spatial domain geometries
simulate_volumetric_domains(): Simulate complex three-dimensional tissue domains

Evaluation

Common benchmark construction, metrics, and timed method comparison.

spatial_benchmark(): Construct a spatial refinement benchmark
evaluate_spatial_refinement(): Evaluate a spatial label refinement
benchmark_spatial_refiners(): Benchmark spatial label refinement methods

Benchmark Data

Licensed real-data benchmark access and provenance.

available_spatial_benchmarks(): List real spatial benchmark datasets
load_spatial_benchmark(): Load a bundled real spatial benchmark
dlpfc_benchmark: Human DLPFC Visium refinement benchmark
colorectal_benchmark: Human CRC Visium HD refinement benchmark