Merge branch 'main' into uni_cox

bblodfon · web-flow · commit d90569484de8 · 2023-11-02T12:16:45.000+01:00
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -58,6 +58,7 @@ Collate:
     'FilterCorrelation.R'
     'FilterDISR.R'
     'FilterFindCorrelation.R'
+    'FilterLearner.R'
     'FilterImportance.R'
     'FilterInformationGain.R'
     'FilterJMI.R'
diff --git a/NEWS.md b/NEWS.md
@@ -2,6 +2,8 @@
 
 - Fixed issue with `FilterPerformance` where the arg `measure` wasn't passed on
 - Added `FilterUnivariateCox` (thanks to @bblodfon)
+- Parameter value `na.rm` is properly initialized to `TRUE` (thanks to @bblodfon)
+- Bugfix: property `missings` is now set correctly for `FilterFindCorrelation`
 
 # mlr3filters 0.7.1
 
diff --git a/R/Filter.R b/R/Filter.R
@@ -212,6 +212,22 @@ Filter = R6Class("Filter",
       assert_ro_binding(rhs)
       get_properties = get0(".get_properties", private)
       if (is.null(get_properties)) character() else get_properties()
+    },
+
+    #' @field hash (`character(1)`)\cr
+    #' Hash (unique identifier) for this object.
+    hash = function(rhs) {
+      assert_ro_binding(rhs)
+      calculate_hash(class(self), self$id, self$param_set$values, mget(private$.extra_hash, envir = self))
+    },
+
+    #' @field phash (`character(1)`)\cr
+    #' Hash (unique identifier) for this partial object, excluding some components
+    #' which are varied systematically during tuning (parameter values) or feature
+    #' selection (feature names).
+    phash = function(rhs) {
+      assert_ro_binding(rhs)
+      calculate_hash(class(self), self$id, mget(private$.extra_hash, envir = self))
     }
   )
 )
diff --git a/R/FilterAUC.R b/R/FilterAUC.R
@@ -15,6 +15,7 @@
 #' `r format_bib("bommert_2020")`
 #'
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterAnova.R b/R/FilterAnova.R
@@ -15,6 +15,7 @@
 #' `r format_bib("bommert_2020")`
 #'
 #' @family Filter
+#' @include Filter.R
 #' @importFrom stats aov
 #' @template seealso_filter
 #' @export
diff --git a/R/FilterCMIM.R b/R/FilterCMIM.R
@@ -16,6 +16,7 @@
 #'
 #' @template details_praznik
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterCarScore.R b/R/FilterCarScore.R
@@ -11,6 +11,7 @@
 #' Argument `verbose` defaults to `FALSE`.
 #'
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterCarSurvScore.R b/R/FilterCarSurvScore.R
@@ -10,6 +10,7 @@
 #' `r format_bib("bommert_2021")`
 #'
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 FilterCarSurvScore = R6Class("FilterCarSurvScore",
diff --git a/R/FilterCorrelation.R b/R/FilterCorrelation.R
@@ -20,6 +20,7 @@
 #' `r format_bib("bommert_2020")`
 #'
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterDISR.R b/R/FilterDISR.R
@@ -16,6 +16,7 @@
 #'
 #' @template details_praznik
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterFindCorrelation.R b/R/FilterFindCorrelation.R
@@ -16,6 +16,7 @@
 #' that are excluded with `FilterFindCorrelation` at score 0.1 (= 1 - 0.9).
 #'
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
@@ -93,6 +94,14 @@ FilterFindCorrelation = R6Class("FilterFindCorrelation",
       # The following has the correct names and values, BUT we need scores in
       # reverse order. Shift by 1 to get positive values.
       1 - apply(cm, 2, max)
+    },
+    .get_properties = function() {
+      use = self$param_set$values$use %??% "everything"
+      if (use %in% c("complete.obs", "pairwise.complete.obs")) {
+        "missings"
+      } else {
+        character(0)
+      }
     }
   )
 )
diff --git a/R/FilterImportance.R b/R/FilterImportance.R
@@ -8,6 +8,7 @@
 #' and extracts the importance values to use as filter scores.
 #'
 #' @family Filter
+#' @include FilterLearner.R
 #' @template seealso_filter
 #' @export
 #' @examples
@@ -34,7 +35,7 @@
 #'   graph$train(task)
 #' }
 FilterImportance = R6Class("FilterImportance",
-  inherit = Filter,
+  inherit = FilterLearner,
 
   public = list(
 
@@ -61,6 +62,7 @@ FilterImportance = R6Class("FilterImportance",
     }
   ),
 
+
   private = list(
     .calculate = function(task, nfeat) {
       learner = self$learner$clone(deep = TRUE)
diff --git a/R/FilterInformationGain.R b/R/FilterInformationGain.R
@@ -13,6 +13,7 @@
 #'   `TRUE` for regression tasks.
 #'
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterJMI.R b/R/FilterJMI.R
@@ -17,6 +17,7 @@
 #'
 #' @template details_praznik
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterJMIM.R b/R/FilterJMIM.R
@@ -16,6 +16,7 @@
 #'
 #' @template details_praznik
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterKruskalTest.R b/R/FilterKruskalTest.R
@@ -23,6 +23,7 @@
 #' `r format_bib("bommert_2020")`
 #'
 #' @family Filter
+#' @include Filter.R
 #' @importFrom stats kruskal.test
 #' @template seealso_filter
 #' @export
diff --git a/R/FilterLearner.R b/R/FilterLearner.R
@@ -0,0 +1,20 @@
+#' @include Filter.R
+FilterLearner = R6Class("FilterLearner", inherit = Filter,
+  active = list(
+    #' @field hash (`character(1)`)\cr
+    #' Hash (unique identifier) for this object.
+    hash = function(rhs) {
+      assert_ro_binding(rhs)
+      calculate_hash(class(self), self$id, self$param_set$values, self$learner$hash)
+    },
+
+    #' @field phash (`character(1)`)\cr
+    #' Hash (unique identifier) for this partial object, excluding some components
+    #' which are varied systematically during tuning (parameter values) or feature
+    #' selection (feature names).
+    phash = function(rhs) {
+      assert_ro_binding(rhs)
+      calculate_hash(class(self), self$id, self$learner$hash)
+    }
+  )
+)
diff --git a/R/FilterMIM.R b/R/FilterMIM.R
@@ -16,6 +16,7 @@
 #'
 #' @template details_praznik
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterMRMR.R b/R/FilterMRMR.R
@@ -16,6 +16,7 @@
 #'
 #' @template details_praznik
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterNJMIM.R b/R/FilterNJMIM.R
@@ -16,6 +16,7 @@
 #'
 #' @template details_praznik
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterPerformance.R b/R/FilterPerformance.R
@@ -9,6 +9,7 @@
 #'   to be minimized.
 #'
 #' @family Filter
+#' @include FilterLearner.R
 #' @template seealso_filter
 #' @export
 #' @examples
@@ -33,7 +34,7 @@
 #'   graph$train(task)
 #' }
 FilterPerformance = R6Class("FilterPerformance",
-  inherit = Filter,
+  inherit = FilterLearner,
 
   public = list(
 
@@ -57,7 +58,7 @@ FilterPerformance = R6Class("FilterPerformance",
       self$learner = learner = assert_learner(as_learner(learner, clone = TRUE))
       self$resampling = assert_resampling(as_resampling(resampling))
       self$measure = assert_measure(as_measure(measure,
-        task_type = learner$task_type, clone = TRUE), learner = learner)
+        task_type = learner$task_type), learner = learner)
       packages = unique(c(self$learner$packages, self$measure$packages))
 
       super$initialize(
diff --git a/R/FilterPermutation.R b/R/FilterPermutation.R
@@ -19,6 +19,7 @@
 #' }
 #'
 #' @family Filter
+#' @include FilterLearner.R
 #' @template seealso_filter
 #' @export
 #' @examples
@@ -45,7 +46,7 @@
 #'   graph$train(task)
 #' }
 FilterPermutation = R6Class("FilterPermutation",
-  inherit = Filter,
+  inherit = FilterLearner,
   public = list(
 
     #' @field learner ([mlr3::Learner])\cr
@@ -88,6 +89,26 @@ FilterPermutation = R6Class("FilterPermutation",
     }
   ),
 
+  active = list(
+    #' @field hash (`character(1)`)\cr
+    #' Hash (unique identifier) for this object.
+    hash = function(rhs) {
+      assert_ro_binding(rhs)
+      calculate_hash(class(self), self$id, self$param_set$values, self$learner$hash,
+        self$resampling$hash, self$measure$hash)
+    },
+
+    #' @field phash (`character(1)`)\cr
+    #' Hash (unique identifier) for this partial object, excluding some components
+    #' which are varied systematically during tuning (parameter values) or feature
+    #' selection (feature names).
+    phash = function(rhs) {
+      assert_ro_binding(rhs)
+      calculate_hash(class(self), self$id, self$learner$hash, self$resampling$hash,
+        self$measure$hash)
+    }
+  ),
+
   private = list(
     .calculate = function(task, nfeat) {
       task = task$clone()
diff --git a/R/FilterRelief.R b/R/FilterRelief.R
@@ -13,6 +13,7 @@
 #' If a feature has no non-missing observation, the resulting score will be (close to) 0.
 #'
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
diff --git a/R/FilterSelectedFeatures.R b/R/FilterSelectedFeatures.R
@@ -15,6 +15,7 @@
 #' \CRANpkg{mlr3pipelines}, only the filter criterion `cutoff` makes sense.
 #'
 #' @family Filter
+#' @include Filter.R
 #' @template seealso_filter
 #' @export
 #' @examples
@@ -42,7 +43,7 @@
 #'   graph$train(task)
 #' }
 FilterSelectedFeatures = R6Class("FilterSelectedFeatures",
-  inherit = Filter,
+  inherit = FilterLearner,
 
   public = list(
 
diff --git a/R/FilterVariance.R b/R/FilterVariance.R
@@ -12,6 +12,7 @@
 #' `r format_bib("bommert_2020")`
 #'
 #' @family Filter
+#' @include Filter.R
 #' @importFrom stats var
 #' @template seealso_filter
 #' @export
diff --git a/README.Rmd b/README.Rmd
@@ -96,7 +96,7 @@ tab = as.data.table(mlr_learners)
 tab[sapply(properties, is.element, el = "importance"), key]
 ```
 
-If your learner is not listed here but capable of extracting variable importance from the fitted model, the reason is most likely that it is not yet integrated in the package [mlr3learners](https://github.com/mlr-org/mlr3learners) or the [extra learner organization](https://github.com/mlr3learners).
+If your learner is not listed here but capable of extracting variable importance from the fitted model, the reason is most likely that it is not yet integrated in the package [mlr3learners](https://github.com/mlr-org/mlr3learners) or the [extra learner extension](https://github.com/mlr-org/mlr3extralearners).
 Please open an issue so we can add your package.
 
 Some learners need to have their variable importance measure "activated" during learner creation.
diff --git a/README.md b/README.md
@@ -100,8 +100,8 @@ If your learner is not listed here but capable of extracting variable
 importance from the fitted model, the reason is most likely that it is
 not yet integrated in the package
 [mlr3learners](https://github.com/mlr-org/mlr3learners) or the [extra
-learner organization](https://github.com/mlr3learners). Please open an
-issue so we can add your package.
+learner extension](https://github.com/mlr-org/mlr3extralearners).
+Please open an issue so we can add your package.
 
 Some learners need to have their variable importance measure “activated”
 during learner creation. For example, to use the “impurity” measure of
diff --git a/man/Filter.Rd b/man/Filter.Rd
diff --git a/man/mlr_filters_importance.Rd b/man/mlr_filters_importance.Rd
diff --git a/man/mlr_filters_performance.Rd b/man/mlr_filters_performance.Rd
diff --git a/man/mlr_filters_permutation.Rd b/man/mlr_filters_permutation.Rd
diff --git a/man/mlr_filters_selected_features.Rd b/man/mlr_filters_selected_features.Rd
