Skip to contents

Leave-Disc-Out Cross-Validation

Source: R/borg_disc_cv.R
borg_disc_cv.Rd

Spatial cross-validation with circular exclusion buffers around test points. For each fold, all training points within a specified radius of any test point are removed, creating a spatial gap that eliminates autocorrelation leakage. More principled than rectangular or k-means blocking for point data.

Usage

borg_disc_cv(
  data,
  coords,
  target = NULL,
  radius = NULL,
  v = 5,
  min_train = 0.5,
  seed = 42,
  output = c("list", "rsample"),
  verbose = FALSE
)

Arguments

data

A data frame with coordinate and predictor columns.

coords

Character vector of length 2. Coordinate column names.

target

Character. Target variable name (for diagnostics).

radius

Numeric. Exclusion buffer radius in coordinate units. If NULL, automatically set to the autocorrelation range estimated by borg_diagnose().

v

Integer. Number of folds. Default: 5.

min_train

Numeric. Minimum fraction of data that must remain in training after exclusion. Folds that violate this are dropped. Default: 0.5.

seed

Integer. Random seed. Default: 42.

output

Character. Output format: "list" (default), "rsample".

verbose

Logical. Print diagnostic messages. Default: FALSE.

Value

A list with class "borg_disc_cv" containing:

folds

List of train/test index vectors

radius

Exclusion radius used

n_excluded

Number of training points excluded per fold (in the buffer zone)

effective_training

Fraction of data available for training per fold (after exclusion)

strategy

"leave_disc_out"

params

List of parameters used

Compatible with other BORG functions that accept fold lists.

Details

How it works

  1. Partition data into v spatial folds using k-means on coordinates (same as borg_cv).

  2. For each fold, identify the test points.

  3. Remove from training all points within radius of any test point. These excluded points are neither train nor test — they form a buffer zone.

  4. If the remaining training set is too small (< min_train * n), the fold is dropped with a warning.

Choosing the radius

The radius should match the autocorrelation range of the spatial process. borg_diagnose() estimates this via variogram analysis. Setting radius = NULL uses this estimate automatically.

Examples

set.seed(42)
d <- data.frame(x = runif(200, 0, 100), y = runif(200, 0, 100))
d$z <- sin(d$x / 10) + rnorm(200, sd = 0.5)
cv <- borg_disc_cv(d, coords = c("x", "y"), target = "z", radius = 15)
cv