Skip to contents

Generate Toroidal Shift Permutation Indices

Source: R/C_resampling_function.R
generate_toroidal_permutations.Rd

Shifts spatial coordinates in a toroidal (wrap-around) manner, perfectly preserving spatial autocorrelation structure. This is useful for permutation testing when you want to break cross-type coordination while preserving within-type spatial patterns.

Usage

generate_toroidal_permutations(location_data, n_permu = 100, seed = NULL)

Arguments

location_data

Data frame with x, y, cell_ID columns

n_permu

Number of permutations to generate

seed

Optional random seed for reproducibility

Value

Matrix of permutation indices (n_cells x n_permu). Each column contains a permutation of row indices that can be used to reorder cells.

Details

The toroidal shift works by:

  1. Applying a random shift to all coordinates (wrapping at boundaries)

  2. Matching cells based on their new positions to original positions

This preserves ALL spatial autocorrelation within each cell type because the relative positions of cells are unchanged - only their absolute positions are shifted.

Examples

if (FALSE) { # \dontrun{
# Create example location data
loc_data <- data.frame(
  x = runif(100, 0, 10),
  y = runif(100, 0, 10),
  cell_ID = paste0("cell_", 1:100)
)

# Generate 100 toroidal permutations
perm_matrix <- generate_toroidal_permutations(loc_data, n_permu = 100)

# Apply first permutation
permuted_cells <- loc_data$cell_ID[perm_matrix[, 1]]
} # }