R/learner_fairml_regr_fairnclm.R

In mlr3fairness: Fairness Auditing and Debiasing for 'mlr3'

#' @title Regression Non-convex Fair Regression Learner
#' @author pfistfl
#' 
#' @details 
#' Fair regression model based on nonconvex optimization from Komiyama et al. (2018).
#' Implemented via package `fairml`.
#' The 'unfairness' parameter is set to 0.05 as a default.
#' 
#' @name mlr_learners_regr.fairnclm
#' @template class_learner
#' @templateVar id regr.fairnclm
#' @templateVar caller nclm
#' @references
#' `r format_bib("komiyama")`
#'
#' @template seealso_learner
#' @template example
#' @export
LearnerRegrFairnclm = R6Class("LearnerRegrFairnclm",
  inherit = LearnerRegr,

  public = list(
    #' @description
    #' Creates a new instance of this [R6][R6::R6Class] class.
    initialize = function() {
      ps = ps(
        lambda = p_dbl(lower = 0, upper = Inf, tags = "train", default = 0),
        save.auxiliary = p_lgl(default = FALSE, tags = "train"),
        covfun = p_uty(tags = "train", default = "stats::cov"),
        unfairness = p_dbl(lower = 0, upper = 1, tags = "train")
      )
      ps$values = list(unfairness = 0.05)

      super$initialize(
        id = "regr.fairnclm",
        packages = "fairml",
        feature_types = c("integer", "numeric", "factor", "ordered"),
        predict_types = c("response"),
        param_set = ps,
        man = "mlr3fairness::mlr_learners_regr.fairnclm"
      )
    }
  ),

  private = list(

    .train = function(task) {
      assert_pta_task(task)
      # get parameters for training
      pars = self$param_set$get_values(tags = "train")

      # set column names to ensure consistency in fit and predict
      self$state$feature_names = task$feature_names
      pta = task$col_roles$pta
      r = as.numeric(task$truth())
      s = get_pta(task, intersect = FALSE)

      pars = remove_named(pars, "intersect")
      p = task$data(cols = setdiff(task$feature_names, pta))
      p = int_to_numeric(p)
      mlr3misc::invoke(fairml::nclm, response = r, predictors = p, sensitive = s, .args = pars)
    },

    .predict = function(task) {
      # get parameters with tag "predict"
      pars = self$param_set$get_values(tags = "predict")

      pta = task$col_roles$pta
      s = get_pta(task, intersect = FALSE)
      p = task$data(cols = setdiff(self$state$feature_names, pta))
      ints = colnames(keep(p, is.integer))
      p = int_to_numeric(p)
      pred = mlr3misc::invoke(predict, self$model, new.predictors = p, new.sensitive = s, .args = pars)
      list(response = pred)
    }
  )
)