Changed resample filter to resample-unsupervised, and added resample-supervised.

src/clj_ml/filters.clj

@@ -192,12 +192,19 @@
 
 (deffilter remove-useless-attributes)
 
-(defmethod make-filter-options :resample
+(defmethod make-filter-options :resample-unsupervised
   ([kind m]
      (->> (check-option-values m {:seed "-S" :size-percent "-Z"})
           (check-options m {:no-replacement "-no-replacement" :invert "-V"}))))
 
-(deffilter resample)
+(deffilter resample-unsupervised)
+
+(defmethod make-filter-options :resample-supervised
+  ([kind m]
+     (->> (check-option-values m {:seed "-S" :size-percent "-Z" :bias "-B"})
+          (check-options m {:no-replacement "-no-replacement" :invert "-V"}))))
+
+(deffilter resample-supervised)
 
 (defmethod make-filter-options :select-append-attributes
   ([kind m]
@@ -235,7 +242,8 @@
    :remove-percentage weka.filters.unsupervised.instance.RemovePercentage
    :remove-range weka.filters.unsupervised.instance.RemoveRange
    :remove-useless-attributes weka.filters.unsupervised.attribute.RemoveUseless
-   :resample weka.filters.unsupervised.instance.Resample
+   :resample-unsupervised weka.filters.unsupervised.instance.Resample
+   :resample-supervised weka.filters.supervised.instance.Resample
    :select-append-attributes weka.filters.unsupervised.attribute.Copy
    :project-attributes weka.filters.unsupervised.attribute.Remove})
 
@@ -258,7 +266,8 @@
      - :remove-percentage
      - :remove-range
      - :remove-useless-attributes
-     - :resample
+     - :resample-unsupervised
+     - :resample-supervised
      - :select-append-attributes
      - :project-attributes
      - :clj-streamable
@@ -432,7 +441,7 @@
             Note: percentage, not decimal. e.g. 89 not 0.89
             If you pass in a decimal Weka silently sets it to 0.0.
 
-    * :resample
+    * :resample-unsupervised
 
       \"Produces a random subsample of a dataset using either sampling
       with replacement or without replacement. The original dataset
@@ -455,6 +464,38 @@
         - :invert
           Inverts the selection; can only be true if :replacement is false (boolean)
 
+    * :resample-supervised
+
+      \"Produces a random subsample of a dataset using either sampling
+      with replacement or without replacement. The original dataset
+      must fit entirely in memory. The number of instances in the
+      generated dataset may be specified. The dataset must have a
+      nominal class attribute. If not, use the unsupervised
+      version. The filter can be made to maintain the class
+      distribution in the subsample, or to bias the class distribution
+      toward a uniform distribution. When used in batch mode (i.e. in
+      the FilteredClassifier), subsequent batches are NOT resampled.\"
+      -- from Weka JavaDoc.
+
+      Parameters:
+
+        - :seed
+          Random number seed (integer)
+
+        - :size-percent
+          \"The size of the output dataset, as a percentage of
+          the input dataset (default 100)\" (integer)
+
+        - :bias \"Bias factor towards uniform class distribution.0 =
+          distribution in input data -- 1 = uniform
+          distribution. (default 0)\" (0 or 1)
+
+        - :no-replacement
+          Use replacement or not; default is false, i.e., with replacement (boolean)
+
+        - :invert
+          Inverts the selection; can only be true if :replacement is false (boolean)
+
     * :select-append-attributes
 
       Append a copy of the selected columns at the end of the dataset.

test/clj_ml/filters_test.clj

@@ -34,12 +34,20 @@
         filter (make-filter :remove-useless-attributes {:dataset-format ds :max-variance 95})]
     (is (== (.getMaximumVariancePercentageAllowed filter) 95))))
 
-(deftest make-filter-resample
+(deftest make-filter-resample-unsupervised
   (fact
    (let [ds (load-instances :arff "http://repository.seasr.org/Datasets/UCI/arff/iris.arff")
-         options (make-filter-options :resample {:dataset-format ds :seed 10 :size-percent 50 :no-replacement true :invert true})]
+         options (make-filter-options :resample-unsupervised
+                                      {:dataset-format ds :seed 10 :size-percent 50 :no-replacement true :invert true})]
      options => (just ["-S" "10" "-Z" "50" "-V" "-no-replacement"] :in-any-order))))
 
+(deftest make-filter-resample-supervised
+  (fact
+   (let [ds (load-instances :arff "http://repository.seasr.org/Datasets/UCI/arff/iris.arff")
+         options (make-filter-options :resample-supervised
+                                      {:dataset-format ds :seed 10 :size-percent 50 :no-replacement true :invert true :bias 1})]
+     options => (just ["-S" "10" "-Z" "50" "-V" "-no-replacement" "-B" "1"] :in-any-order))))
+
 (deftest make-filter-discretize-sup
   (let [ds (make-dataset :test [:a :b {:c [:g :m]}]
                                      [ [1 2 :g]
@@ -92,12 +100,19 @@
     (is (= weka.filters.unsupervised.attribute.Reorder
            (class f)))))
 
-(deftest make-filter-reorder-attributes
+(deftest make-filter-resample-unsupervised
   (let [ds (load-instances :arff "http://repository.seasr.org/Datasets/UCI/arff/iris.arff")
-        f (make-filter :resample {:dataset-format ds :seed 10 :size-percent 50 :replacement true})]
+        f (make-filter :resample-unsupervised {:dataset-format ds :seed 10 :size-percent 50 :replacement true})]
     (is (= weka.filters.unsupervised.instance.Resample
            (class f)))))
 
+(deftest make-filter-resample-supervised
+  (let [ds (dataset-set-class (load-instances :arff "http://repository.seasr.org/Datasets/UCI/arff/iris.arff")
+                              :class)
+        f (make-filter :resample-supervised {:dataset-format ds :seed 10 :size-percent 50 :replacement true :bias 1})]
+    (is (= weka.filters.supervised.instance.Resample
+           (class f)))))
+
 (deftest make-filter-remove-attributes
   (let [ds (make-dataset :test [:a :b {:c [:g :m]}]
                                      [ [1 2 :g]
@@ -178,9 +193,15 @@
     (is (= (str ds (str (make-dataset :test [{:s nil} :n {:class [:yes :no]}]
                                       [["Hello" 55 :yes] ["World" -100 :no]])))))))
 
-(deftest make-apply-filter-resample
+(deftest make-apply-filter-resample-unsupervised
   (let [ds (load-instances :arff "http://repository.seasr.org/Datasets/UCI/arff/iris.arff")
-        ds2 (make-apply-filter :resample {:seed 10 :size-percent 50 :replacement true} ds)]
+        ds2 (make-apply-filter :resample-unsupervised {:seed 10 :size-percent 50 :replacement true} ds)]
+    (is (= 75 (dataset-count ds2)))))
+
+(deftest make-apply-filter-resample-supervised
+  (let [ds (dataset-set-class (load-instances :arff "http://repository.seasr.org/Datasets/UCI/arff/iris.arff")
+                              :class)
+        ds2 (make-apply-filter :resample-supervised {:seed 10 :size-percent 50 :replacement true :bias 1} ds)]
     (is (= 75 (dataset-count ds2)))))
 
 (deftest make-apply-filters-test