Merge pull request #779 from mrapp-ke/ordinal-feature-vector

Neue Klasse OrdinalFeatureVector

cpp/subprojects/common/include/mlrl/common/input/feature_vector_nominal.hpp

@@ -18,10 +18,15 @@ class NominalFeatureVector : public AbstractFeatureVector {
 
         uint32* indptr_;
 
-        const uint32 numValues_;
-
         const int32 majorityValue_;
 
+    protected:
+
+        /**
+         * The number of distinct values of the nominal feature, excluding the majority value.
+         */
+        const uint32 numValues_;
+
     public:
 
         /**

cpp/subprojects/common/include/mlrl/common/input/feature_vector_ordinal.hpp

@@ -0,0 +1,61 @@
+/*
+ * @author Michael Rapp ([email protected])
+ */
+#pragma once
+
+#include "mlrl/common/input/feature_vector_nominal.hpp"
+
+/**
+ * A feature vector that stores the indices of the examples that are associated with each value, except for the majority
+ * value, i.e., the most frequent value, of an ordinal feature.
+ */
+class OrdinalFeatureVector : public NominalFeatureVector {
+    private:
+
+        uint32* order_;
+
+    public:
+
+        /**
+         * @param numValues     The number of distinct values of the ordinal feature, excluding the majority value
+         * @param numElements   The number of elements in the vector, i.e., the number of examples not associated with
+         *                      the majority value
+         * @param majorityValue The majority value, i.e., the most frequent value, of the ordinal feature
+         */
+        OrdinalFeatureVector(uint32 numValues, uint32 numElements, int32 majorityValue);
+
+        ~OrdinalFeatureVector() override;
+
+        /**
+         * Returns an `index_iterator` to the beginning of the ordered indices of the values of the ordinal feature.
+         *
+         * @param index The index of the value
+         * @return      An `index_iterator` to the beginning
+         */
+        index_iterator order_begin(uint32 index);
+
+        /**
+         * Returns an `index_iterator` to the end of the ordered indices of the values of the ordinal feature.
+         *
+         * @param index The index of the value
+         * @return      An `index_iterator` to the end
+         */
+        index_iterator order_end(uint32 index);
+
+        /**
+         * Returns an `index_const_iterator` to the beginning of the ordered indices of the values of the ordinal
+         * feature.
+         *
+         * @param index The index of the value
+         * @return      An `index_const_iterator` to the beginning
+         */
+        index_const_iterator order_cbegin(uint32 index) const;
+
+        /**
+         * Returns an `index_const_iterator` to the end of the ordered indices of the values of the ordinal feature.
+         *
+         * @param index The index of the value
+         * @return      An `index_const_iterator` to the end
+         */
+        index_const_iterator order_cend(uint32 index) const;
+};

cpp/subprojects/common/meson.build

@@ -45,6 +45,7 @@ source_files = [
     'src/mlrl/common/input/feature_vector_equal.cpp',
     'src/mlrl/common/input/feature_vector_nominal.cpp',
     'src/mlrl/common/input/feature_vector_numerical.cpp',
+    'src/mlrl/common/input/feature_vector_ordinal.cpp',
     'src/mlrl/common/input/label_matrix_c_contiguous.cpp',
     'src/mlrl/common/input/label_matrix_csc.cpp',
     'src/mlrl/common/input/label_matrix_csr.cpp',

cpp/subprojects/common/src/mlrl/common/input/feature_vector_nominal.cpp

@@ -2,7 +2,7 @@
 
 NominalFeatureVector::NominalFeatureVector(uint32 numValues, uint32 numElements, int32 majorityValue)
     : values_(new int32[numValues]), indices_(new uint32[numElements]), indptr_(new uint32[numValues + 1]),
-      numValues_(numValues), majorityValue_(majorityValue) {
+      majorityValue_(majorityValue), numValues_(numValues) {
     indptr_[numValues] = numElements;
 }
 

cpp/subprojects/common/src/mlrl/common/input/feature_vector_ordinal.cpp

@@ -0,0 +1,24 @@
+#include "mlrl/common/input/feature_vector_ordinal.hpp"
+
+OrdinalFeatureVector::OrdinalFeatureVector(uint32 numValues, uint32 numElements, int32 majorityValue)
+    : NominalFeatureVector(numValues, numElements, majorityValue), order_(new uint32[numValues]) {}
+
+OrdinalFeatureVector::~OrdinalFeatureVector() {
+    delete[] order_;
+}
+
+OrdinalFeatureVector::index_iterator OrdinalFeatureVector::order_begin(uint32 index) {
+    return order_;
+}
+
+OrdinalFeatureVector::index_iterator OrdinalFeatureVector::order_end(uint32 index) {
+    return &order_[numValues_];
+}
+
+OrdinalFeatureVector::index_const_iterator OrdinalFeatureVector::order_cbegin(uint32 index) const {
+    return order_;
+}
+
+OrdinalFeatureVector::index_const_iterator OrdinalFeatureVector::order_cend(uint32 index) const {
+    return &order_[numValues_];
+}