borg_trainControl: BORG-Guarded trainControl

In BORG: Bounded Outcome Risk Guard for Model Evaluation

BORG-Guarded trainControl

Description

A guarded version of caret::trainControl() that validates CV settings against data dependencies.

Usage

borg_trainControl(
  data,
  method = "cv",
  number = 10,
  coords = NULL,
  time = NULL,
  groups = NULL,
  target = NULL,
  allow_override = FALSE,
  ...
)

Arguments

data	A data frame. Required for dependency checking.
method	Character. Resampling method.
number	Integer. Number of folds or iterations.
coords	Character vector. Coordinate columns for spatial check.
time	Character. Time column for temporal check.
groups	Character. Group column for clustered check.
target	Character. Target variable.
allow_override	Logical. Allow random CV despite dependencies.
...	Additional arguments passed to caret::trainControl().

Value

A trainControl object, potentially modified for blocked CV.

Examples

if (requireNamespace("caret", quietly = TRUE)) {
  spatial_data <- data.frame(
    lon = runif(50), lat = runif(50), response = rnorm(50)
  )
  ctrl <- borg_trainControl(
    data = spatial_data,
    method = "cv",
    number = 5,
    coords = c("lon", "lat")
  )
}

BORG documentation built on March 20, 2026, 5:09 p.m.