Spatial Block Cross-Validation (rsample-compatible)

Creates spatial block CV folds that plug directly into tune::tune_grid() and tune::fit_resamples().

Usage

borg_spatial_cv(
  data,
  coords,
  target = NULL,
  v = 5,
  buffer = NULL,
  repeats = 1L,
  ...
)

Arguments

data

A data frame, sf object, or terra::SpatVector.

coords

Character vector of length 2. Coordinate column names.

target

Character. Target variable for autocorrelation diagnosis.

v

Integer. Number of folds. Default: 5.

buffer

Numeric. Optional spatial buffer distance for exclusion.

repeats

Integer. Number of repeated fold sets. Default: 1.

...

Additional arguments passed to borg_cv().

Value

A borg_rset object (subclass of rset) compatible with tidymodels.

Examples

# \donttest{
if (requireNamespace("rsample", quietly = TRUE)) {
  set.seed(42)
  d <- data.frame(x = runif(100), y = runif(100), z = rnorm(100))
  folds <- borg_spatial_cv(d, coords = c("x", "y"), target = "z")
}
# }