R/param_parallel_reg.R

Defines functions parallel_reg

Documented in parallel_reg

#' Parallel regression assumption (logical)
#'
#' Whether the parallel regression assumption is imposed in an ordinal
#' regression model.
#'
#' @param values A vector of possible values (`TRUE` or `FALSE`).
#'
#' @details This parameter is used by ordinal regression models specified by
#'   `parsnip::ordinal_reg()`, for example with the `"clm"`, `"vglm"`, or
#'   `"ordinalNet"` engines. It controls whether the regression coefficients are
#'   shared across all ordinal thresholds (e.g. proportional odds in
#'   logit-linked models) or allowed to vary (generalized ordered logit). `TRUE`
#'   makes all effects parallel, `FALSE` makes all effects nominal or
#'   category-specific.
#' @examples
#' values_parallel_reg
#' parallel_reg()
#' @export
parallel_reg <- function(values = values_parallel_reg) {
  new_qual_param(
    type = "logical",
    values = values,
    label = c(parallel_reg = "Parallel Regression"),
    finalize = NULL
  )
}

#' @rdname parallel_reg
#' @export
values_parallel_reg <- c(TRUE, FALSE)

Try the dials package in your browser

Any scripts or data that you put into this service are public.

dials documentation built on Nov. 11, 2026, 5:06 p.m.