R/learner_CoxBoost_surv_coxboost.R

In mlr-org/mlr3extralearners: Extra Learners For mlr3

#' @title Survival Cox Model with Likelihood Based Boosting Learner
#' @author RaphaelS1
#' @name mlr_learners_surv.coxboost
#'
#' @description
#' Fit a Survival Cox model with a likelihood based boosting algorithm.
#' Calls [CoxBoost::CoxBoost()] from package 'CoxBoost'.
#'
#' @template learner
#' @templateVar id surv.coxboost
#'
#' @template install_coxboost
#'
#' @details
#' Use [LearnerSurvCoxboost] and [LearnerSurvCVCoxboost] for Cox boosting without and with internal
#' cross-validation of boosting step number, respectively. Tuning using the internal optimizer in
#' [LearnerSurvCVCoxboost] may be more efficient when tuning `stepno` only. However, for tuning
#' multiple hyperparameters, \CRANpkg{mlr3tuning} and [LearnerSurvCoxboost] will likely give better
#' results.
#'
#' Three prediction types are returned for this learner, using the internal
#' `predict.CoxBoost()` function:
#' 1. `lp`: a vector of linear predictors (relative risk scores), one per
#' observation.
#' 2. `crank`: same as `lp`.
#' 3. `distr`: a 2d survival matrix, with observations as rows and time points
#' as columns. The internal transformation uses the Breslow estimator to compose
#' the survival distributions from the `lp` predictions.
#'
#' @references
#' `r format_bib("binder2009boosting")`
#'
#' @template seealso_learner
#' @template example
#' @export
LearnerSurvCoxboost = R6Class("LearnerSurvCoxboost",
  inherit = mlr3proba::LearnerSurv,

  public = list(
    #' @description
    #' Creates a new instance of this [R6][R6::R6Class] class.
    initialize = function() {
      ps = ps(
        unpen.index = p_uty(tags = "train"),
        standardize = p_lgl(default = TRUE, tags = "train"),
        stepno = p_int(default = 100, lower = 0, tags = "train"),
        penalty = p_dbl(tags = "train"),
        criterion = p_fct(default = "pscore", levels = c("pscore", "score", "hpscore", "hscore"), tags = "train"),
        stepsize.factor = p_dbl(default = 1, tags = "train"),
        sf.scheme = p_fct(default = "sigmoid", levels = c("sigmoid", "linear"), tags = "train"),
        pendistmat = p_uty(tags = "train"),
        connected.index = p_uty(tags = "train"),
        x.is.01 = p_lgl(default = FALSE, tags = "train"),
        return.score = p_lgl(default = TRUE, tags = "train"),
        trace = p_lgl(default = FALSE, tags = "train"),
        at.step = p_uty(tags = "predict")
      )

      super$initialize(
        id = "surv.coxboost",
        packages = c("mlr3extralearners", "CoxBoost", "pracma"),
        feature_types = c("integer", "numeric"),
        predict_types = c("distr", "crank", "lp"),
        param_set = ps,
        properties = c("weights", "selected_features"),
        man = "mlr3extralearners::mlr_learners_surv.coxboost",
        label = "Likelihood-based Boosting"
      )
    },

    #' @description
    #' Returns the set of selected features which have non-zero coefficients.
    #' Calls the internal `coef.CoxBoost()` function.
    #'
    #' @param at_step (`integer(1)`)\cr
    #' Which boosting step to get the coefficients for. If no step is given
    #' (default), the final boosting step is used.
    #'
    #' @return (`character()`) vector of feature names.
    selected_features = function(at_step = NULL) {
      if (is.null(self$model)) {
        stopf("No model stored")
      }

      coefs = invoke(stats::coef, self$model, at.step = at_step)
      coefs = coefs[coefs != 0]
      names(coefs)
    }
  ),

  private = list(
    .train = function(task) {
      pars = self$param_set$get_values(tags = "train")

      if ("weights" %in% task$properties) {
        pars$weights = as.numeric(task$weights$weight)
      }

      data = task$data()
      tn = task$target_names
      time = data[[tn[1L]]]
      status = data[[tn[2L]]]
      data = as.matrix(data[, !tn, with = FALSE])

      with_package("CoxBoost", {
        invoke(
          CoxBoost::CoxBoost,
          time = time,
          status = status,
          x = data,
          .args = pars
        )
      })
    },

    .predict = function(task) {

      pars = self$param_set$get_values(tags = "predict")

      # get newdata and ensure same ordering in train and predict
      newdata = as.matrix(ordered_features(task, self))

      lp = as.numeric(invoke(predict,
        self$model,
        newdata = newdata,
        .args = pars,
        type = "lp"))

      surv = invoke(predict,
        self$model,
        newdata = newdata,
        .args = pars,
        type = "risk",
        times = sort(unique(self$model$time)))

      mlr3proba::.surv_return(times = sort(unique(self$model$time)),
        surv = surv,
        lp = lp)
    }
  )
)

.extralrns_dict$add("surv.coxboost", LearnerSurvCoxboost)