R/RLearner_cluster_SimpleKMeans.R

#' @export
makeRLearner.cluster.SimpleKMeans = function() {
  makeRLearnerCluster(
    cl = "cluster.SimpleKMeans",
    package = "RWeka",
    par.set = makeParamSet(
      makeUntypedLearnerParam(id = "A", default = "weka.core.EuclideanDistance"),
      makeLogicalLearnerParam(id = "C", default = FALSE),
      makeLogicalLearnerParam(id = "fast", default = FALSE),
      makeIntegerLearnerParam(id = "I", default = 100L, lower = 1L),
      makeIntegerLearnerParam(id = "init", default = 0L, lower = 0L, upper = 3L),
      makeLogicalLearnerParam(id = "M", default = FALSE),
      makeIntegerLearnerParam(id = "max-candidates", default = 100L, lower = 1L),
      makeIntegerLearnerParam(id = "min-density", default = 2L, lower = 1L),
      makeIntegerLearnerParam(id = "N", default = 2L, lower = 1L),
      makeIntegerLearnerParam(id = "num-slots", default = 1L, lower = 1L),
      makeLogicalLearnerParam(id = "O", default = FALSE),
      makeIntegerLearnerParam(id = "periodic-pruning", default = 10000L, lower = 1L),
      makeIntegerLearnerParam(id = "S", default = 10L, lower = 0L),
      makeNumericLearnerParam(id = "t2", default = -1),
      makeNumericLearnerParam(id = "t1", default = -1.5),
      makeLogicalLearnerParam(id = "V", default = FALSE, tunable = FALSE),
      makeLogicalLearnerParam(id = "output-debug-info", default = FALSE, tunable = FALSE)
    ),
    properties = "numerics",
    name = "K-Means Clustering",
    short.name = "simplekmeans",
    callees = c("SimpleKMeans", "Weka_control")
  )
}

#' @export
trainLearner.cluster.SimpleKMeans = function(.learner, .task, .subset, .weights = NULL,  ...) {
  ctrl = RWeka::Weka_control(...)
  RWeka::SimpleKMeans(getTaskData(.task, .subset), control = ctrl)
}

#' @export
predictLearner.cluster.SimpleKMeans = function(.learner, .model, .newdata, ...) {
  # SimpleKMeans returns cluster indices (i.e. starting from 0, which some tools don't like
  as.integer(predict(.model$learner.model, .newdata, ...)) + 1L
}
guillermozbta/mir documentation built on May 11, 2019, 6:27 p.m.