Add AdaBoost, IBk, LMT, OneR, PART (#6)

DESCRIPTION

@@ -1,12 +1,14 @@
 Package: mlr3learners.rweka
-Title: M5Rules, JRip and J48 Learner for mlr3
+Title: AdaBoostM1, IBk ,JRip, J48, LMT, M5Rules, OneR and PART
+    Learner for mlr3
 Version: 0.1.0.9000
 Authors@R:
     person(given = "Henri",
            family = "Funk",
            role = "cre",
            email = "[email protected]")
-Description: JRip and J48 Learner for mlr3.
+Description: AdaBoostM1, IBk ,JRip, J48, LMT, M5Rules, OneR and
+    PART Learner for mlr3.
 License: LGPL-3
 Depends:
     R (>= 3.1.0)

NAMESPACE

@@ -1,7 +1,13 @@
 # Generated by roxygen2: do not edit by hand
 
+export(LearnerClassifAdaBoostM1)
+export(LearnerClassifIBk)
 export(LearnerClassifJ48)
 export(LearnerClassifJRip)
+export(LearnerClassifLMT)
+export(LearnerClassifOneR)
+export(LearnerClassifPART)
+export(LearnerRegrIBk)
 export(LearnerRegrM5Rules)
 import(data.table)
 import(mlr3misc)

R/LearnerClassifAdaBoostM1.R

@@ -0,0 +1,104 @@
+#' @title Classification AdaBoostM1 Learner
+#'
+#' @name mlr_learners_classif.AdaBoostM1
+#'
+#' @description
+#' A [mlr3::LearnerClassif] implementing classification AdaBoostM1 from package \CRANpkg{RWeka}.
+# Calls [RWeka::AdaBoostM1()].
+#'
+#' @section Custom mlr3 defaults:
+#' - `output_debug_info`:
+#'   - original id: output-debug-info
+#'
+#' - `do_not_check_capabilities`:
+#'   - original id: do-not-check-capabilities
+#'
+#' - `num_decimal_places`:
+#'   - original id: num-decimal-places
+#'
+#' - `batch_size`:
+#'   - original id: batch-size
+#'
+#' - Reason for change: This learner contains changed ids of the following control arguments
+#' since their ids contain irregular pattern
+#'
+#' @templateVar id classif.AdaBoostM1
+#' @template section_dictionary_learner
+#'
+#' @references
+#'  Freund Y, Schapire, R (1993).
+#' Experiments with a New Boosting Algorithm
+#' \url{http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.51.6252&rep=rep1&type=pdf}
+#'
+#' @export
+LearnerClassifAdaBoostM1 = R6Class("LearnerClassifAdaBoostM1",
+  inherit = LearnerClassif,
+  public = list(
+    #' @description
+    #' Creates a new instance of this [R6][R6::R6Class] class.
+    initialize = function() {
+      ps = ParamSet$new(
+        params = list(
+          ParamUty$new(id = "subset", tags = c("train", "pars")),
+          ParamUty$new(id = "na.action", tags = c("train", "pars")),
+          ParamInt$new(
+            id = "P", default = 100L, lower = 90L, upper = 100L,
+            tags = c("train", "control")),
+          ParamLgl$new(id = "Q", default = FALSE, tags = c("train", "control")),
+          ParamInt$new(id = "S", default = 1L, lower = 1L, tags = c("train", "control")),
+          ParamInt$new(id = "I", default = 10L, lower = 1L, tags = c("train", "control")),
+          ParamUty$new(
+            id = "W", default = "weka.classifiers.trees.DecisionStump",
+            tags = c("train", "control")),
+          ParamLgl$new(id = "output_debug_info", default = FALSE, tags = c("train", "control")),
+          ParamLgl$new(
+            id = "do_not_check_capabilities", default = FALSE,
+            tags = c("train", "control")),
+          ParamInt$new(
+            id = "num_decimal_places", default = 2L, lower = 1L,
+            tags = c("train", "control")),
+          ParamInt$new(id = "batch_size", default = 100L, lower = 1L, tags = c("train", "control")),
+          ParamUty$new(id = "options", default = NULL, tags = c("train", "pars"))
+        )
+      )
+
+      super$initialize(
+        id = "classif.AdaBoostM1",
+        packages = "RWeka",
+        feature_types = c("numeric", "factor", "ordered"),
+        predict_types = c("response", "prob"),
+        param_set = ps,
+        properties = c("twoclass", "multiclass"),
+        man = "mlr3learners.rweka::mlr_learners_classif.AdaBoostM1"
+      )
+    }
+  ),
+
+  private = list(
+    .train = function(task) {
+      ctrl = self$param_set$get_values(tags = "control")
+      if (length(ctrl) > 0L) {
+        names(ctrl) = gsub("_", replacement = "-", x = names(ctrl))
+        ctrl = mlr3misc::invoke(RWeka::Weka_control, ctrl)
+      }
+
+      pars = self$param_set$get_values(tags = "pars")
+      f = task$formula()
+      data = task$data()
+      mlr3misc::invoke(RWeka::AdaBoostM1, formula = f, data = data, control = ctrl, .args = pars)
+    },
+
+    .predict = function(task) {
+      response = NULL
+      prob = NULL
+      newdata = task$data(cols = task$feature_names)
+
+      if (self$predict_type == "response") {
+        response = mlr3misc::invoke(predict, self$model, newdata = newdata, type = "class")
+      } else {
+        prob = mlr3misc::invoke(predict, self$model, newdata = newdata, type = "prob")
+      }
+      PredictionClassif$new(task = task, response = response, prob = prob)
+    }
+  )
+)

R/LearnerClassifIBk.R

@@ -0,0 +1,104 @@
+#' @title Classification IBk Learner
+#'
+#' @name mlr_learners_classif.IBk
+#'
+#' @description
+#' A [mlr3::LearnerClassif] implementing classification IBk from package \CRANpkg{RWeka}.
+#' Calls [RWeka::IBk()].
+#'
+#' @section Custom mlr3 defaults:
+#' - `output_debug_info`:
+#'   - original id: output-debug-info
+#'
+#' - `do_not_check_capabilities`:
+#'   - original id: do-not-check-capabilities
+#'
+#' - `num_decimal_places`:
+#'   - original id: num-decimal-places
+#'
+#' - `batch_size`:
+#'   - original id: batch-size
+#'
+#' - Reason for change: This learner contains changed ids of the following control arguments
+#' since their ids contain irregular pattern
+#'
+#' @templateVar id classif.IBk
+#' @template section_dictionary_learner
+#'
+#' @references
+#' Aha D, Kibbler D, Alber M (1991).
+#' Instance-based learning algorithms
+#' \url{https://link.springer.com/content/pdf/10.1007/BF00153759.pdf}
+#'
+#' @export
+LearnerClassifIBk = R6Class("LearnerClassifIBk",
+  inherit = LearnerClassif,
+  public = list(
+    #' @description
+    #' Creates a new instance of this [R6][R6::R6Class] class.
+    initialize = function() {
+      ps = ParamSet$new(
+        params = list(
+          ParamUty$new(id = "subset", tags = c("train", "pars")),
+          ParamUty$new(id = "na.action", tags = c("train", "pars")),
+          ParamLgl$new(id = "I", default = FALSE, tags = c("train", "control")),
+          ParamLgl$new(id = "F", default = FALSE, tags = c("train", "control")),
+          ParamInt$new(id = "K", default = 1L, lower = 1L, tags = c("train", "control")),
+          ParamLgl$new(id = "E", default = FALSE, tags = c("train", "control")),
+          ParamInt$new(id = "W", default = 0L, lower = 0L, tags = c("train", "control")),
+          ParamLgl$new(id = "X", default = FALSE, tags = c("train", "control")),
+          ParamUty$new(
+            id = "A", default = "weka.core.neighboursearch.LinearNNSearch",
+            tags = c("train", "control")),
+          ParamLgl$new(id = "output_debug_info", default = FALSE, tags = c("train", "control")),
+          ParamLgl$new(
+            id = "do_not_check_capabilities", default = FALSE,
+            tags = c("train", "control")),
+          ParamInt$new(
+            id = "num_decimal_places", default = 2L, lower = 1L,
+            tags = c("train", "control")),
+          ParamInt$new(id = "batch_size", default = 100L, lower = 1L, tags = c("train", "control")),
+          ParamUty$new(id = "options", default = NULL, tags = c("train", "pars"))
+        )
+      )
+
+      super$initialize(
+        id = "classif.IBk",
+        packages = "RWeka",
+        feature_types = c("numeric", "factor", "ordered"),
+        predict_types = c("response", "prob"),
+        param_set = ps,
+        properties = c("twoclass", "multiclass"),
+        man = "mlr3learners.rweka::mlr_learners_classif.IBk"
+      )
+    }
+  ),
+
+  private = list(
+    .train = function(task) {
+      ctrl = self$param_set$get_values(tags = "control")
+      if (length(ctrl) > 0L) {
+        names(ctrl) = gsub("_", replacement = "-", x = names(ctrl))
+        ctrl = mlr3misc::invoke(RWeka::Weka_control, ctrl)
+      }
+
+      pars = self$param_set$get_values(tags = "pars")
+      f = task$formula()
+      data = task$data()
+      mlr3misc::invoke(RWeka::IBk, formula = f, data = data, control = ctrl, .args = pars)
+    },
+
+    .predict = function(task) {
+      response = NULL
+      prob = NULL
+      newdata = task$data(cols = task$feature_names)
+
+      if (self$predict_type == "response") {
+        response = mlr3misc::invoke(predict, self$model, newdata = newdata, type = "class")
+      } else {
+        prob = mlr3misc::invoke(predict, self$model, newdata = newdata, type = "prob")
+      }
+      PredictionClassif$new(task = task, response = response, prob = prob)
+    }
+  )
+)

R/LearnerClassifJ48.R

@@ -6,12 +6,21 @@
 #' A [mlr3::LearnerClassif] implementing classification J48 from package \CRANpkg{RWeka}.
 #' Calls [RWeka::J48()].
 #'
-#' This learner contains changed ids of the following control agruments
-#' since their ids contain irregular pattern:
-#' * mlr3learner: output_debug_info RWeka: output-debug-info
-#' * mlr3learner: do_not_check_capabilities RWeka: do-not-check-capabilities
-#' * mlr3learner: num_decimal_places RWeka: num-decimal-places
-#' * mlr3learner: batch_size RWeka: batch-size
+#' @section Custom mlr3 defaults:
+#' - `output_debug_info`:
+#'   - original id: output-debug-info
+#'
+#' - `do_not_check_capabilities`:
+#'   - original id: do-not-check-capabilities
+#'
+#' - `num_decimal_places`:
+#'   - original id: num-decimal-places
+#'
+#' - `batch_size`:
+#'   - original id: batch-size
+#'
+#' - Reason for change: This learner contains changed ids of the following control arguments
+#' since their ids contain irregular pattern
 #'
 #' @templateVar id classif.J48
 #' @template section_dictionary_learner
@@ -34,7 +43,8 @@ LearnerClassifJ48 = R6Class("LearnerClassifJ48",
           ParamUty$new(id = "na.action", tags = c("train", "pars")),
           ParamLgl$new(id = "U", default = FALSE, tags = c("train", "control")),
           ParamLgl$new(id = "O", default = FALSE, tags = c("train", "control")),
-          ParamDbl$new(id = "C", default = 0.25, lower = .Machine$double.eps,
+          ParamDbl$new(
+            id = "C", default = 0.25, lower = .Machine$double.eps,
             upper = 1 - .Machine$double.eps, tags = c("train", "control")),
           ParamInt$new(id = "M", default = 2L, lower = 1L, tags = c("train", "control")),
           ParamLgl$new(id = "R", default = FALSE, tags = c("train", "control")),
@@ -45,13 +55,15 @@ LearnerClassifJ48 = R6Class("LearnerClassifJ48",
           ParamLgl$new(id = "A", default = FALSE, tags = c("train", "control")),
           ParamLgl$new(id = "J", default = FALSE, tags = c("train", "control")),
           ParamInt$new(id = "Q", default = 1L, lower = 1L, tags = c("train", "control")),
-          ParamLgl$new(id = "doNotMakeSplitPointActualValue", default = FALSE,
-            tags = c("train", "control")),
-          ParamLgl$new(id = "output_debug_info", default = FALSE,
+          ParamLgl$new(
+            id = "doNotMakeSplitPointActualValue", default = FALSE,
             tags = c("train", "control")),
-          ParamLgl$new(id = "do_not_check_capabilities", default = FALSE,
+          ParamLgl$new(id = "output_debug_info", default = FALSE, tags = c("train", "control")),
+          ParamLgl$new(
+            id = "do_not_check_capabilities", default = FALSE,
             tags = c("train", "control")),
-          ParamInt$new(id = "num_decimal_places", default = 2L, lower = 1L,
+          ParamInt$new(
+            id = "num_decimal_places", default = 2L, lower = 1L,
             tags = c("train", "control")),
           ParamInt$new(id = "batch_size", default = 100L, lower = 1L, tags = c("train", "control")),
           ParamUty$new(id = "options", default = NULL, tags = c("train", "pars"))

R/LearnerClassifJRip.R

@@ -6,12 +6,21 @@
 #' A [mlr3::LearnerClassif] implementing classification JRip from package \CRANpkg{RWeka}.
 #' Calls [RWeka::JRip()].
 #'
-#' This learner contains changed ids of the following control agruments
-#' since their ids contain irregular pattern:
-#' * mlr3learner: output_debug_info RWeka: output-debug-info
-#' * mlr3learner: do_not_check_capabilities RWeka: do-not-check-capabilities
-#' * mlr3learner: num_decimal_places RWeka: num-decimal-places
-#' * mlr3learner: batch_size RWeka: batch-size
+#' @section Custom mlr3 defaults:
+#' - `output_debug_info`:
+#'   - original id: output-debug-info
+#'
+#' - `do_not_check_capabilities`:
+#'   - original id: do-not-check-capabilities
+#'
+#' - `num_decimal_places`:
+#'   - original id: num-decimal-places
+#'
+#' - `batch_size`:
+#'   - original id: batch-size
+#'
+#' - Reason for change: This learner contains changed ids of the following control arguments
+#' since their ids contain irregular pattern
 #'
 #' @templateVar id classif.Jrip
 #' @template section_dictionary_learner
@@ -41,9 +50,11 @@ LearnerClassifJRip = R6Class("LearnerClassifJRip",
           ParamLgl$new(id = "E", default = FALSE, tags = c("train", "control")),
           ParamLgl$new(id = "P", default = FALSE, tags = c("train", "control")),
           ParamLgl$new(id = "output_debug_info", default = FALSE, tags = c("train", "control")),
-          ParamLgl$new(id = "do_not_check_capabilities", default = FALSE,
+          ParamLgl$new(
+            id = "do_not_check_capabilities", default = FALSE,
             tags = c("train", "control")),
-          ParamInt$new(id = "num_decimal_places", default = 2L, lower = 1L,
+          ParamInt$new(
+            id = "num_decimal_places", default = 2L, lower = 1L,
             tags = c("train", "control")),
           ParamInt$new(id = "batch_size", default = 100L, lower = 1L, tags = c("train", "control")),
           ParamUty$new(id = "options", default = NULL, tags = c("train", "pars"))