Neue Klasse NominalFeatureVector

cpp/subprojects/common/include/mlrl/common/data/types.hpp

@@ -6,6 +6,7 @@
 #include <cmath>
 #include <limits>
 
+typedef int int32;
 typedef long int int64;
 typedef unsigned char uint8;
 typedef unsigned int uint32;

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

@@ -0,0 +1,152 @@
+/*
+ * @author Michael Rapp ([email protected])
+ */
+#pragma once
+
+#include "mlrl/common/input/feature_vector_common.hpp"
+
+/**
+ * A feature vector that stores the indices of the examples that are associated with each value, except for the minority
+ * value, i.e., the least frequent value, of a nominal feature.
+ */
+class NominalFeatureVector final : public AbstractFeatureVector {
+    private:
+
+        int32* values_;
+
+        uint32* indices_;
+
+        uint32* indptr_;
+
+        const uint32 numValues_;
+
+        const int32 minorityValue_;
+
+    public:
+
+        /**
+         * @param numValues     The number of distinct values of the nominal feature, excluding the minority value
+         * @param numExamples   The total number of examples
+         * @param minorityValue The minority value, i.e., the least frequent value, of the nominal feature
+         */
+        NominalFeatureVector(uint32 numValues, uint32 numExamples, int32 minorityValue);
+
+        ~NominalFeatureVector() override;
+
+        /**
+         * An iterator that provides access to the values of the nominal feature and allows to modify them.
+         */
+        typedef int32* value_iterator;
+
+        /**
+         * An iterator that provides read-only access to the values of the nominal feature.
+         */
+        typedef const int32* value_const_iterator;
+
+        /**
+         * An iterator that provides access to the indices of the examples that are associated with each value of the
+         * nominal feature and allows to modify them.
+         */
+        typedef uint32* index_iterator;
+
+        /**
+         * An iterator that provides read-only access to the indices of the examples that are associated with each value
+         * of the nominal feature.
+         */
+        typedef const uint32* index_const_iterator;
+
+        /**
+         * An iterator that provides access to the indices that specify the first element in the array of example
+         * indices that corresponds to each value of the nominal feature.
+         */
+        typedef uint32* indptr_iterator;
+
+        /**
+         * Returns a `value_iterator` to the beginning of the values of the nominal feature.
+         *
+         * @return A `value_iterator` to the beginning
+         */
+        value_iterator values_begin();
+
+        /**
+         * Returns a `value_iterator` to the end of the values of the nominal feature.
+         *
+         * @return A `value_iterator` to the end
+         */
+        value_iterator values_end();
+
+        /**
+         * Returns a `value_const_iterator` to the beginning of the values of the nominal feature.
+         *
+         * @return A `value_const_iterator` to the beginning
+         */
+        value_const_iterator values_cbegin() const;
+
+        /**
+         * Returns a `value_const_iterator` to the end of the value of the nominal feature.
+         *
+         * @return A `value_const_iterator` to the end
+         */
+        value_const_iterator values_cend() const;
+
+        /**
+         * Returns an `index_iterator` to the beginning of the indices of the examples that are associated with a
+         * specific value of the nominal feature.
+         *
+         * @param index The index of the value
+         * @return      An `index_iterator` to the beginning
+         */
+        index_iterator indices_begin(uint32 index);
+
+        /**
+         * Returns an `index_iterator` to the end of the indices of the examples that are associated with a specific
+         * value of the nominal feature.
+         *
+         * @param index The index of the value
+         * @return      An `index_iterator` to the end
+         */
+        index_iterator indices_end(uint32 index);
+
+        /**
+         * Returns an `index_const_iterator` to the beginning of the indices of the examples that are associated with a
+         * specific value of the nominal feature.
+         *
+         * @param index The index of the value
+         * @return      An `index_const_iterator` to the beginning
+         */
+        index_const_iterator indices_cbegin(uint32 index) const;
+
+        /**
+         * Returns an `index_const_iterator` to the end of the indices of the examples that are associated with a
+         * specific value of the nominal feature.
+         *
+         * @param index The index of the value
+         * @return      An `index_const_iterator` to the end
+         */
+        index_const_iterator indices_cend(uint32 index) const;
+
+        /**
+         * Returns an `indptr_iterator` to the beginning of the indices that specify the first element in the array of
+         * example indices that corresponds to each value of the nominal feature.
+         *
+         * @return An `indptr_iterator` to the beginning
+         */
+        indptr_iterator indptr_begin();
+
+        /**
+         * Returns an `indptr_iterator` to the end of the indices that specify the first element in the array of example
+         * indices that corresponds to each value of the nominal feature.
+         *
+         * @return An `indptr_iterator` to the end
+         */
+        indptr_iterator indptr_end();
+
+        /**
+         * Returns the minority value, i.e., the least frequent value, of the nominal feature.
+         *
+         * @return The minority value
+         */
+        int32 getMinorityValue() const;
+
+        uint32 getNumElements() const override;
+};

cpp/subprojects/common/meson.build

@@ -42,6 +42,7 @@ source_files = [
     'src/mlrl/common/input/feature_vector.cpp',
     'src/mlrl/common/input/feature_vector_common.cpp',
     '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/label_matrix_c_contiguous.cpp',
     'src/mlrl/common/input/label_matrix_csc.cpp',

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

@@ -0,0 +1,61 @@
+#include "mlrl/common/input/feature_vector_nominal.hpp"
+
+NominalFeatureVector::NominalFeatureVector(uint32 numValues, uint32 numExamples, int32 minorityValue)
+    : values_(new int32[numValues]), indices_(new uint32[numExamples]), indptr_(new uint32[numValues + 1]),
+      numValues_(numValues), minorityValue_(minorityValue) {
+    indptr_[numValues] = numExamples;
+}
+
+NominalFeatureVector::~NominalFeatureVector() {
+    delete[] values_;
+    delete[] indices_;
+    delete[] indptr_;
+}
+
+NominalFeatureVector::value_iterator NominalFeatureVector::values_begin() {
+    return values_;
+}
+
+NominalFeatureVector::value_iterator NominalFeatureVector::values_end() {
+    return &values_[numValues_];
+}
+
+NominalFeatureVector::value_const_iterator NominalFeatureVector::values_cbegin() const {
+    return values_;
+}
+
+NominalFeatureVector::value_const_iterator NominalFeatureVector::values_cend() const {
+    return &values_[numValues_];
+}
+
+NominalFeatureVector::index_iterator NominalFeatureVector::indices_begin(uint32 index) {
+    return &indices_[indptr_[index]];
+}
+
+NominalFeatureVector::index_iterator NominalFeatureVector::indices_end(uint32 index) {
+    return &indices_[indptr_[index + 1]];
+}
+
+NominalFeatureVector::index_const_iterator NominalFeatureVector::indices_cbegin(uint32 index) const {
+    return &indices_[indptr_[index]];
+}
+
+NominalFeatureVector::index_const_iterator NominalFeatureVector::indices_cend(uint32 index) const {
+    return &indices_[indptr_[index + 1]];
+}
+
+NominalFeatureVector::indptr_iterator NominalFeatureVector::indptr_begin() {
+    return indptr_;
+}
+
+NominalFeatureVector::indptr_iterator NominalFeatureVector::indptr_end() {
+    return &indptr_[numValues_];
+}
+
+int32 NominalFeatureVector::getMinorityValue() const {
+    return minorityValue_;
+}
+
+uint32 NominalFeatureVector::getNumElements() const {
+    return numValues_;
+}