Simulate spatial tissue domains with challenging geometry — simulate_spatial

Generates irregular 2D tissue sections or 3D tissue volumes with known domain labels and corrupted initial cluster assignments.

Usage

simulate_spatial_domains(
  n = 50000L,
  pattern = c("jagged_stripes", "wavy_layers", "rings", "spiral", "branching", "lobes",
    "islands", "disconnected", "thin_layers", "intermixed", "layers3d"),
  k = 5L,
  noise = 0.2,
  dimensions = if (identical(pattern[1L], "layers3d")) 3L else 2L,
  samples = 1L,
  noise_type = c("random", "boundary", "patch", "region"),
  feature_scale = 1,
  density_profile = c("uniform", "moderate", "strong", "extreme", "hotspot"),
  seed = 1L
)

Arguments

n: Number of observations.
pattern: Domain geometry. See Details.
k: Number of tissue domains.
noise: Fraction of initial labels to corrupt. Each true class present in a specimen retains one correct exemplar; a rate that makes this impossible is rejected.
dimensions: Either 2 or 3. Pattern `layers3d` requires 3.
samples: Number of independent slides or samples.
noise_type: One of `"random"`, `"boundary"`, `"patch"`, or `"region"`.
feature_scale: Relative width of islands and thin layers. Values below one create more difficult sub-neighborhood structures.
density_profile: Relative observation concentration across tissue regions. Use `"uniform"`, `"moderate"`, `"strong"`, `"extreme"`, `"hotspot"`, or a positive numeric vector of length `k`. Character profiles permute region weights reproducibly to avoid tying density to a particular class identifier.
seed: Random seed.

Value

A `spatial_refinement_benchmark` with coordinates, noisy labels, truth, samples, geometry metadata, and boundary and sparse-region indicators.

Details

Available geometries include jagged and wavy layers, concentric rings, spiral arms, branching sectors, lobes, rare islands, disconnected domains, thin layers, interleaved microdomains, and curved 3D layers.