mlr-org · mllg · Nov 13, 2022 · Oct 1, 2022 · Oct 1, 2022 · Oct 20, 2022
diff --git a/.github/workflows/update-tic.yml b/.github/workflows/update-tic.yml
@@ -2,7 +2,7 @@ on:
   workflow_dispatch:
   schedule:
     # * is a special character in YAML so you have to quote this string
-    - cron: "0 4 * * *"
+    - cron: "0 4 * * 0" # every week
 
 name: Update tic
 
@@ -23,27 +23,26 @@ jobs:
       GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
 
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v3
         with:
           persist-credentials: false
 
-      - uses: r-lib/actions/setup-r@master
+      - uses: r-lib/actions/setup-r@v2
         with:
           r-version: ${{ matrix.config.r }}
           Ncpus: 4
 
       - name: "[Stage] Dependencies"
         run: |
-            sudo apt install libcurl4-openssl-dev libsodium-dev libharfbuzz-dev libfribidi-dev libgit2-dev
-            Rscript -e "if (!requireNamespace('remotes')) install.packages('remotes', type = 'source')"
-            Rscript -e "remotes::install_github('ropensci/tic', dependencies = TRUE)"
+          Rscript -e "install.packages('pak', repos = 'https://r-lib.github.io/p/pak/stable')"
+          Rscript -e "if (grepl('Ubuntu', Sys.info()[['version']])) {options(repos = c(CRAN = sprintf('https://packagemanager.rstudio.com/all/__linux__/%s/latest', system('lsb_release -cs', intern = TRUE))))}; pak::pkg_install('ropensci/tic', dependencies = TRUE)"
 
       - name: "[Stage] Update YAMLs"
         run: |
             Rscript -e "tic::update_yml()"
 
       - name: "[Stage] Create Pull Request"
-        uses: peter-evans/create-pull-request@v3
+        uses: peter-evans/create-pull-request@v4
         with:
           author: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
           token: ${{ secrets.TIC_UPDATE }}

diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
@@ -2,7 +2,7 @@
 # R specific hooks: https://github.com/lorenzwalthert/precommit
 repos:
 -   repo: https://github.com/lorenzwalthert/precommit
-    rev: v0.3.2
+    rev: v0.3.2.9003
     hooks:
     -   id: style-files
         args: [--style_pkg=styler.mlr, --style_fun=mlr_style]

diff --git a/NEWS.md b/NEWS.md
@@ -5,21 +5,6 @@
 ## Chore
 
 - do not use deprecated function from mlr3misc
-
-## Uncategorized
-
-- Improve PipeOpFilter docu (@66853113+pre-commit-ci[bot], #135)
-
-- [github.com/lorenzwalthert/precommit: v0.3.0 → v0.3.2](https://github.com/lorenzwalthert/precommit/compare/v0.3.0...v0.3.2) (@66853113+pre-commit-ci[bot], #129)
-
-- [github.com/pre-commit/pre-commit-hooks: v4.2.0 → v4.3.0](https://github.com/pre-commit/pre-commit-hooks/compare/v4.2.0...v4.3.0) (@66853113+pre-commit-ci[bot], #129)
-
-- [pre-commit.ci] pre-commit autoupdate (@66853113+pre-commit-ci[bot], #128)
-
-- [github.com/lorenzwalthert/precommit: v0.2.2.9015 → v0.3.0](https://github.com/lorenzwalthert/precommit/compare/v0.2.2.9015...v0.3.0) (@66853113+pre-commit-ci[bot], #128)
-
-- [pre-commit.ci] auto fixes from pre-commit.com hooks (@66853113+pre-commit-ci[bot], #128)
-
 - Use featureless learner in defaults (#124)
 
 

diff --git a/R/Filter.R b/R/Filter.R
@@ -27,13 +27,13 @@ Filter = R6Class("Filter",
     #'   Can be used in tables, plot and text output instead of the ID.
     label = NA_character_,
 
-    #' @field task_type (`character(1)`)\cr
-    #'   Task type, e.g. `"classif"` or `"regr"`.
-    #'   Can be set to `NA` to allow all task types.
+    #' @field task_types (`character()`)\cr
+    #'   Set of supported task types, e.g. `"classif"` or `"regr"`.
+    #'   Can be set to the scalar value `NA` to allow any task type.
     #'
     #'   For a complete list of possible task types (depending on the loaded packages),
     #'   see [`mlr_reflections$task_types$type`][mlr_reflections].
-    task_type = NULL,
+    task_types = NULL,
 
     #' @field task_properties (`character()`)\cr
     #'   [mlr3::Task]task properties.
@@ -67,9 +67,9 @@ Filter = R6Class("Filter",
     #' @description Create a Filter object.
     #' @param id (`character(1)`)\cr
     #'   Identifier for the filter.
-    #' @param task_type (`character()`)\cr
+    #' @param task_types (`character()`)\cr
     #'   Types of the task the filter can operator on. E.g., `"classif"` or
-    #'   `"regr"`. Can be set to `NA` to allow all task types.
+    #'   `"regr"`. Can be set to scalar `NA` to allow any task type.
     #' @param param_set ([paradox::ParamSet])\cr
     #'   Set of hyperparameters.
     #' @param feature_types (`character()`)\cr
@@ -90,18 +90,18 @@ Filter = R6Class("Filter",
     #'   String in the format `[pkg]::[topic]` pointing to a manual page for
     #'   this object. The referenced help package can be opened via method
     #'   `$help()`.
-    initialize = function(id, task_type, task_properties = character(),
+    initialize = function(id, task_types, task_properties = character(),
       param_set = ps(), feature_types = character(),
       packages = character(), label = NA_character_, man = NA_character_) {
 
       self$id = assert_string(id)
       self$label = assert_string(label, na.ok = TRUE)
-      if (!test_scalar_na(task_type)) {
+      if (!test_scalar_na(task_types)) {
         # we allow any task type here, otherwise we are not able to construct
         # the filter without loading additional packages like mlr3proba
-        assert_character(task_type, any.missing = FALSE)
+        assert_character(task_types, any.missing = FALSE)
       }
-      self$task_type = task_type
+      self$task_types = task_types
       self$task_properties = assert_subset(
         task_properties,
         unlist(mlr_reflections$task_properties, use.names = FALSE))
@@ -124,7 +124,7 @@ Filter = R6Class("Filter",
     #' Printer for Filter class
     print = function() {
       catn(format(self), if (is.na(self$label)) "" else paste0(": ", self$label))
-      catn(str_indent("Task Types:", self$task_type))
+      catn(str_indent("Task Types:", self$task_types))
       catn(str_indent("Task Properties:", self$task_properties))
       catn(str_indent("Packages:", self$packages))
       catn(str_indent("Feature types:", self$feature_types))
@@ -162,6 +162,11 @@ Filter = R6Class("Filter",
       task = assert_task(as_task(task),
         feature_types = self$feature_types,
         task_properties = self$task_properties)
+
+      if (!is_scalar_na(self$task_types) && task$task_type %nin% self$task_types) {
+        stopf("Filter '%s' does not support the type '%s' of task '%s'",
+          self$id, task$task_type, task$id)
+      }
       fn = task$feature_names
 
       if (task$nrow == 0L) {
@@ -176,6 +181,11 @@ Filter = R6Class("Filter",
           nfeat = min(nfeat, length(fn))
         }
 
+        if (any(task$missings() > 0L)) {
+          stopf("Cannot apply filter '%s' on task '%s', missing values detected",
+            self$id, task$id)
+        }
+
         # calculate filter values using the dedicated filter
         require_namespaces(self$packages)
         scores = private$.calculate(task, nfeat)

diff --git a/R/FilterAUC.R b/R/FilterAUC.R
@@ -18,7 +18,7 @@
 #' @template seealso_filter
 #' @export
 #' @examples
-#' task = mlr3::tsk("pima")
+#' task = mlr3::tsk("sonar")
 #' filter = flt("auc")
 #' filter$calculate(task)
 #' head(as.data.table(filter), 3)
@@ -43,7 +43,7 @@ FilterAUC = R6Class("FilterAUC",
     initialize = function() {
       super$initialize(
         id = "auc",
-        task_type = "classif",
+        task_types = "classif",
         task_properties = "twoclass",
         feature_types = c("integer", "numeric"),
         packages = "mlr3measures",

diff --git a/R/FilterAnova.R b/R/FilterAnova.R
@@ -49,7 +49,7 @@ FilterAnova = R6Class("FilterAnova",
         id = "anova",
         packages = "stats",
         feature_types = c("integer", "numeric"),
-        task_type = "classif",
+        task_types = "classif",
         label = "ANOVA F-Test",
         man = "mlr3filters::mlr_filters_anova"
       )

diff --git a/R/FilterCMIM.R b/R/FilterCMIM.R
@@ -51,7 +51,7 @@ FilterCMIM = R6Class("FilterCMIM",
 
       super$initialize(
         id = "cmim",
-        task_type = c("classif", "regr"),
+        task_types = c("classif", "regr"),
         param_set = param_set,
         feature_types = c("integer", "numeric", "factor", "ordered"),
         packages = "praznik",

diff --git a/R/FilterCarScore.R b/R/FilterCarScore.R
@@ -53,7 +53,7 @@ FilterCarScore = R6Class("FilterCarScore",
 
       super$initialize(
         id = "carscore",
-        task_type = "regr",
+        task_types = "regr",
         param_set = param_set,
         feature_types = "numeric",
         packages = "care",

diff --git a/R/FilterCarSurvScore.R b/R/FilterCarSurvScore.R
@@ -27,7 +27,7 @@ FilterCarSurvScore = R6Class("FilterCarSurvScore",
         packages = c("carSurv", "mlr3proba"),
         param_set = ps,
         feature_types = c("integer", "numeric"),
-        task_type = "surv",
+        task_types = "surv",
         label = "Correlation-Adjusted coRrelation Survival Score",
         man = "mlr3filters::mlr_filters_carsurvscore"
       )

diff --git a/R/FilterCorrelation.R b/R/FilterCorrelation.R
@@ -51,7 +51,7 @@ FilterCorrelation = R6Class("FilterCorrelation",
 
       super$initialize(
         id = "correlation",
-        task_type = "regr",
+        task_types = "regr",
         param_set = param_set,
         feature_types = c("integer", "numeric"),
         packages = "stats",

diff --git a/R/FilterDISR.R b/R/FilterDISR.R
@@ -51,7 +51,7 @@ FilterDISR = R6Class("FilterDISR",
 
       super$initialize(
         id = "disr",
-        task_type = c("classif", "regr"),
+        task_types = c("classif", "regr"),
         param_set = param_set,
         feature_types = c("integer", "numeric", "factor", "ordered"),
         packages = "praznik",

diff --git a/R/FilterFindCorrelation.R b/R/FilterFindCorrelation.R
@@ -55,7 +55,7 @@ FilterFindCorrelation = R6Class("FilterFindCorrelation",
 
       super$initialize(
         id = "find_correlation",
-        task_type = c("classif", "regr"),
+        task_types = c("classif", "regr"),
         param_set = param_set,
         feature_types = c("integer", "numeric"),
         packages = "stats",

diff --git a/R/FilterImportance.R b/R/FilterImportance.R
@@ -51,7 +51,7 @@ FilterImportance = R6Class("FilterImportance",
 
       super$initialize(
         id = "importance",
-        task_type = learner$task_type,
+        task_types = learner$task_type,
         feature_types = learner$feature_types,
         packages = learner$packages,
         param_set = learner$param_set,

diff --git a/R/FilterInformationGain.R b/R/FilterInformationGain.R
@@ -18,7 +18,7 @@
 #' @examples
 #' if (requireNamespace("FSelectorRcpp")) {
 #'   ## InfoGain (default)
-#'   task = mlr3::tsk("pima")
+#'   task = mlr3::tsk("sonar")
 #'   filter = flt("information_gain")
 #'   filter$calculate(task)
 #'   head(filter$scores, 3)
@@ -61,7 +61,7 @@ FilterInformationGain = R6Class("FilterInformationGain",
 
       super$initialize(
         id = "information_gain",
-        task_type = c("classif", "regr"),
+        task_types = c("classif", "regr"),
         param_set = param_set,
         feature_types = c("integer", "numeric", "factor", "ordered"),
         packages = "FSelectorRcpp",

diff --git a/R/FilterJMI.R b/R/FilterJMI.R
@@ -52,7 +52,7 @@ FilterJMI = R6Class("FilterJMI",
 
       super$initialize(
         id = "jmi",
-        task_type = c("classif", "regr"),
+        task_types = c("classif", "regr"),
         param_set = param_set,
         packages = "praznik",
         feature_types = c("integer", "numeric", "factor", "ordered"),

diff --git a/R/FilterJMIM.R b/R/FilterJMIM.R
@@ -50,7 +50,7 @@ FilterJMIM = R6Class("FilterJMIM",
       param_set$values = list(threads = 1L)
       super$initialize(
         id = "jmim",
-        task_type = c("classif", "regr"),
+        task_types = c("classif", "regr"),
         param_set = param_set,
         packages = "praznik",
         feature_types = c("integer", "numeric", "factor", "ordered"),

diff --git a/R/FilterKruskalTest.R b/R/FilterKruskalTest.R
@@ -50,7 +50,7 @@ FilterKruskalTest = R6Class("FilterKruskalTest",
 
       super$initialize(
         id = "kruskal_test",
-        task_type = "classif",
+        task_types = "classif",
         param_set = param_set,
         packages = "stats",
         feature_types = c("integer", "numeric"),

diff --git a/R/FilterMIM.R b/R/FilterMIM.R
@@ -51,7 +51,7 @@ FilterMIM = R6Class("FilterMIM",
 
       super$initialize(
         id = "mim",
-        task_type = c("classif", "regr"),
+        task_types = c("classif", "regr"),
         param_set = param_set,
         packages = "praznik",
         feature_types = c("integer", "numeric", "factor", "ordered"),

diff --git a/R/FilterMRMR.R b/R/FilterMRMR.R
@@ -51,7 +51,7 @@ FilterMRMR = R6Class("FilterMRMR",
 
       super$initialize(
         id = "mrmr",
-        task_type = c("classif", "regr"),
+        task_types = c("classif", "regr"),
         param_set = param_set,
         packages = "praznik",
         feature_types = c("integer", "numeric", "factor", "ordered"),

diff --git a/R/FilterNJMIM.R b/R/FilterNJMIM.R
@@ -50,7 +50,7 @@ FilterNJMIM = R6Class("FilterNJMIM",
       param_set$values = list(threads = 1L)
       super$initialize(
         id = "njmim",
-        task_type = c("classif", "regr"),
+        task_types = c("classif", "regr"),
         param_set = param_set,
         packages = "praznik",
         feature_types = c("integer", "numeric", "factor", "ordered"),

diff --git a/R/FilterPerformance.R b/R/FilterPerformance.R
@@ -61,7 +61,7 @@ FilterPerformance = R6Class("FilterPerformance",
 
       super$initialize(
         id = "performance",
-        task_type = learner$task_type,
+        task_types = learner$task_type,
         param_set = learner$param_set,
         feature_types = learner$feature_types,
         packages = packages,

diff --git a/R/FilterPermutation.R b/R/FilterPermutation.R
@@ -78,7 +78,7 @@ FilterPermutation = R6Class("FilterPermutation",
 
       super$initialize(
         id = "permutation",
-        task_type = learner$task_type,
+        task_types = learner$task_type,
         feature_types = learner$feature_types,
         packages = packages,
         param_set = param_set,

diff --git a/R/FilterRelief.R b/R/FilterRelief.R
@@ -11,7 +11,7 @@
 #' @examples
 #' if (requireNamespace("FSelectorRcpp")) {
 #'   ## Relief (default)
-#'   task = mlr3::tsk("pima")
+#'   task = mlr3::tsk("sonar")
 #'   filter = flt("relief")
 #'   filter$calculate(task)
 #'   head(filter$scores, 3)
@@ -43,7 +43,7 @@ FilterRelief = R6Class("FilterRelief",
 
       super$initialize(
         id = "relief",
-        task_type = c("classif", "regr"),
+        task_types = c("classif", "regr"),
         param_set = param_set,
         feature_types = c("integer", "numeric", "factor", "ordered"),
         packages = "FSelectorRcpp",

diff --git a/R/FilterSelectedFeatures.R b/R/FilterSelectedFeatures.R
@@ -59,7 +59,7 @@ FilterSelectedFeatures = R6Class("FilterSelectedFeatures",
 
       super$initialize(
         id = "selected_features",
-        task_type = learner$task_type,
+        task_types = learner$task_type,
         feature_types = learner$feature_types,
         packages = learner$packages,
         param_set = learner$param_set,

diff --git a/R/FilterVariance.R b/R/FilterVariance.R
@@ -47,7 +47,7 @@ FilterVariance = R6Class("FilterVariance",
 
       super$initialize(
         id = "variance",
-        task_type = NA_character_,
+        task_types = NA_character_,
         param_set = param_set,
         packages = "stats",
         feature_types = c("integer", "numeric"),

diff --git a/R/mlr_filters.R b/R/mlr_filters.R
@@ -33,7 +33,7 @@ as.data.table.DictionaryFilter = function(x, ..., objects = FALSE) {
   setkeyv(map_dtr(x$keys(), function(key) {
     f = x$get(key)
     insert_named(
-      list(key = key, label = f$label, task_type = list(f$task_type),
+      list(key = key, label = f$label, task_types = list(f$task_types),
         task_properties = list(f$task_properties), params = list(f$param_set$ids()),
         feature_types = list(f$feature_types), packages = list(f$packages)),
       if (objects) list(object = list(f))