R/LearnerClustAgnes.R

In mlr-org/mlr3cluster: Cluster Extension for 'mlr3'

#' @title Agglomerative Hierarchical Clustering Learner
#'
#' @name mlr_learners_clust.agnes
#'
#' @description
#' A [LearnerClust] for agglomerative hierarchical clustering implemented in [cluster::agnes()].
#' The predict method uses [stats::cutree()] which cuts the tree resulting from
#' hierarchical clustering into specified number of groups (see parameter `k`).
#' The default number for `k` is 2.
#'
#' @templateVar id clust.agnes
#' @template learner
#'
#' @references
#' `r format_bib("kaufman2009finding")`
#'
#' @export
#' @template seealso_learner
#' @template example
LearnerClustAgnes = R6Class("LearnerClustAgnes",
  inherit = LearnerClust,
  public = list(
    #' @description
    #' Creates a new instance of this [R6][R6::R6Class] class.
    initialize = function() {
      param_set = ps(
        metric = p_fct(default = "euclidean", levels = c("euclidean", "manhattan"), tags = "train"),
        stand = p_lgl(default = FALSE, tags = "train"),
        method = p_fct(
          default = "average",
          levels = c("average", "single", "complete", "ward", "weighted", "flexible", "gaverage"),
          tags = "train"
        ),
        trace.lev = p_int(0L, default = 0L, tags = "train"),
        k = p_int(1L, default = 2L, tags = "predict"),
        par.method = p_uty(tags = "train", custom_check = crate(function(x) {
          if (!(test_numeric(x) || test_list(x))) {
            return("`par.method` needs to be a numeric vector")
          }
          if (length(x) %in% c(1L, 3L, 4L)) {
            TRUE
          } else {
            "`par.method` needs be of length 1, 3, or 4"
          }
        }))
      )

      # param deps
      param_set$add_dep("par.method", "method", CondAnyOf$new(c("flexible", "gaverage")))

      param_set$set_values(k = 2L)

      super$initialize(
        id = "clust.agnes",
        feature_types = c("logical", "integer", "numeric"),
        predict_types = "partition",
        param_set = param_set,
        properties = c("hierarchical", "exclusive", "complete"),
        packages = "cluster",
        man = "mlr3cluster::mlr_learners_clust.agnes",
        label = "Agglomerative Hierarchical Clustering"
      )
    }
  ),
  private = list(
    .train = function(task) {
      pv = self$param_set$get_values()
      m = invoke(cluster::agnes,
        x = task$data(),
        diss = FALSE,
        .args = remove_named(pv, "k")
      )
      if (self$save_assignments) {
        self$assignments = stats::cutree(m, pv$k)
      }

      return(m)
    },

    .predict = function(task) {
      pv = self$param_set$get_values(tags = "predict")
      if (pv$k > task$nrow) {
        stopf("`k` needs to be between 1 and %i", task$nrow)
      }

      warn_prediction_useless(self$id)

      PredictionClust$new(task = task, partition = self$assignments)
    }
  )
)

#' @include aaa.R
learners[["clust.agnes"]] = LearnerClustAgnes