Fix doxygen warnings.

include/plssvm/data_set/classification_data_set.hpp

@@ -41,6 +41,12 @@ namespace plssvm {
 // forward declare C-SVC class
 class csvc;
 
+/**
+ * @brief Encapsulate all necessary data that is needed for training or predicting using an C-SVC.
+ * @details May or may not contain labels!
+ *          Internally, saves all data using [`std::shared_ptr`](https://en.cppreference.com/w/cpp/memory/shared_ptr) to make a plssvm::classification_data_set relatively cheap to copy!
+ * @tparam U the label type of the data (must be an arithmetic type or `std::string`; default: `int`)
+ */
 template <typename U = int>
 class classification_data_set : public data_set<U> {
     // make sure only valid template types are used
@@ -53,6 +59,7 @@ class classification_data_set : public data_set<U> {
     template <typename>
     friend class classification_model;
 
+    /// The base data set class.
     using base_data_set = data_set<U>;
 
     using base_data_set::data_ptr_;

include/plssvm/data_set/data_set.hpp

@@ -220,10 +220,12 @@ class data_set {
     data_set(data_set &&) noexcept = default;
     /**
      * @brief Default copy assignment operator.
+     * @return `*this`
      */
     data_set &operator=(const data_set &) = default;
     /**
      * @brief Default move assignment operator.
+     * @return `*this`
      */
     data_set &operator=(data_set &&) noexcept = default;
 
@@ -300,8 +302,7 @@ class data_set {
         data_ptr_{ std::make_shared<soa_matrix<real_type>>() } { }
 
     /**
-     * @brief Create the mapping between the provided labels and the internally used indices.
-     * @param[in] classes the list of different labels used to create the index mapping
+     * @brief Create the mapping between the provided labels and the internally used values.
      * @throws plssvm::data_set_exception any exception of the plssvm::data_set::label_mapper class
      */
     virtual void map_label() = 0;

include/plssvm/data_set/regression_data_set.hpp

@@ -37,7 +37,13 @@ namespace plssvm {
 // forward declare C-SVR class
 class csvr;
 
-template <typename U = int>
+/**
+ * @brief Encapsulate all necessary data that is needed for training or predicting using an C-SVR.
+ * @details May or may not contain labels!
+ *          Internally, saves all data using [`std::shared_ptr`](https://en.cppreference.com/w/cpp/memory/shared_ptr) to make a plssvm::regression_data_set relatively cheap to copy!
+ * @tparam U the label type of the data (must be an arithmetic type, except boolean or character types; default: `real_type`)
+ */
+template <typename U = real_type>
 class regression_data_set : public data_set<U> {
     // make sure only valid template types are used
     static_assert(detail::tuple_contains_v<U, detail::supported_label_types_regression>,
@@ -49,6 +55,7 @@ class regression_data_set : public data_set<U> {
     template <typename>
     friend class regression_model;
 
+    /// The base data set class.
     using base_data_set = data_set<U>;
 
     using base_data_set::data_ptr_;

include/plssvm/detail/io/regression_libsvm_model_parsing.hpp

@@ -422,10 +422,8 @@ inline void write_libsvm_model_header_regression(fmt::ostream &out, const plssvm
  * @tparam label_type the type of the labels (any arithmetic type, except bool, or std::string)
  * @param[in] filename the file to write the LIBSVM model to
  * @param[in] params the SVM parameters
- * @param[in] classification the used multi-class classification strategy
  * @param[in] rho the rho value resulting from the hyperplane learning
  * @param[in] alpha the weights learned by the SVM
- * @param[in] index_sets index sets containing the SV indices per class
  * @param[in] data the data used to create the model
  * @attention The PLSSVM model file is only compatible with LIBSVM for the one vs. one classification type.
  */

include/plssvm/detail/type_traits.hpp

@@ -347,6 +347,7 @@ constexpr bool is_container_v = is_sequence_container_v<T> || is_associative_con
  */
 template <typename T, typename... Types>
 struct is_one_type_of {
+    /// Set to `true` if @p T is in the type set @p Types.
     constexpr static bool value = (std::is_same_v<T, Types> || ...);
 };
 

include/plssvm/model/classification_model.hpp

@@ -32,7 +32,6 @@
 #include <cstddef>   // std::size_t
 #include <memory>    // std::shared_ptr, std::make_shared
 #include <numeric>   // std::iota
-#include <optional>  // std::optional
 #include <string>    // std::string
 #include <tuple>     // std::tie
 #include <utility>   // std::move
@@ -41,12 +40,7 @@
 namespace plssvm {
 
 /**
- * @example model_examples.cpp
- * @brief A few examples regarding the plssvm::model class.
- */
-
-/**
- * @brief Implements a class encapsulating the result of a call to the SVM fit function. A model is used to predict the labels of a new data set.
+ * @brief Implements a class encapsulating the result of a call to the C-SVC fit function. A model is used to predict the labels of a new data set.
  * @tparam U the type of the used labels (must be an arithmetic type or `std:string`; default: `int`)
  */
 template <typename U = int>
@@ -59,6 +53,7 @@ class classification_model : public model<U> {
     friend class csvm;
     friend class csvc;
 
+    /// The base model class.
     using base_model = model<U>;
 
     // TODO: better?

include/plssvm/model/regression_model.hpp

@@ -13,44 +13,33 @@
 #define PLSSVM_MODEL_REGRESSION_MODEL_HPP_
 #pragma once
 
-#include "plssvm/classification_types.hpp"                           // plssvm::classification_type
-#include "plssvm/constants.hpp"                                      // plssvm::real_type
-#include "plssvm/data_set/data_set.hpp"                              // plssvm::data_set
-#include "plssvm/data_set/regression_data_set.hpp"                   // plssvm::regression_data_set
-#include "plssvm/detail/assert.hpp"                                  // PLSSVM_ASSERT
-#include "plssvm/detail/io/classification_libsvm_model_parsing.hpp"  // plssvm::detail::io::{parse_libsvm_model_header, parse_libsvm_model_data, write_libsvm_model_data}
-#include "plssvm/detail/io/file_reader.hpp"                          // plssvm::detail::io::file_reader
-#include "plssvm/detail/io/regression_libsvm_model_parsing.hpp"      //
-#include "plssvm/detail/logging.hpp"                                 // plssvm::detail::log
-#include "plssvm/detail/tracking/performance_tracker.hpp"            // PLSSVM_DETAIL_TRACKING_PERFORMANCE_TRACKER_ADD_TRACKING_ENTRY, plssvm::detail::tracking::tracking_entry
-#include "plssvm/detail/type_list.hpp"                               // plssvm::detail::{supported_label_types, tuple_contains_v}
-#include "plssvm/matrix.hpp"                                         // plssvm::soa_matrix, plssvm::aos_matrix
-#include "plssvm/model/model.hpp"                                    // plssvm::model
-#include "plssvm/parameter.hpp"                                      // plssvm::parameter
-#include "plssvm/verbosity_levels.hpp"                               // plssvm::verbosity_level
-
-#include <chrono>    // std::chrono::{time_point, steady_clock, duration_cast, milliseconds}
-#include <cstddef>   // std::size_t
-#include <memory>    // std::shared_ptr, std::make_shared
-#include <numeric>   // std::iota
-#include <optional>  // std::optional
-#include <string>    // std::string
-#include <tuple>     // std::tie
-#include <utility>   // std::move
-#include <vector>    // std::vector
+#include "plssvm/constants.hpp"                                  // plssvm::real_type
+#include "plssvm/data_set/data_set.hpp"                          // plssvm::data_set
+#include "plssvm/data_set/regression_data_set.hpp"               // plssvm::regression_data_set
+#include "plssvm/detail/io/file_reader.hpp"                      // plssvm::detail::io::file_reader
+#include "plssvm/detail/io/regression_libsvm_model_parsing.hpp"  // plssvm::detail::io::{parse_libsvm_model_header_regression, parse_libsvm_model_data_regression, write_libsvm_model_data_regression}
+#include "plssvm/detail/logging.hpp"                             // plssvm::detail::log
+#include "plssvm/detail/tracking/performance_tracker.hpp"        // PLSSVM_DETAIL_TRACKING_PERFORMANCE_TRACKER_ADD_TRACKING_ENTRY, plssvm::detail::tracking::tracking_entry
+#include "plssvm/detail/type_list.hpp"                           // plssvm::detail::{supported_label_types, tuple_contains_v}
+#include "plssvm/matrix.hpp"                                     // plssvm::soa_matrix, plssvm::aos_matrix
+#include "plssvm/model/model.hpp"                                // plssvm::model
+#include "plssvm/parameter.hpp"                                  // plssvm::parameter
+#include "plssvm/verbosity_levels.hpp"                           // plssvm::verbosity_level
+
+#include <chrono>   // std::chrono::{time_point, steady_clock, duration_cast, milliseconds}
+#include <cstddef>  // std::size_t
+#include <memory>   // std::shared_ptr, std::make_shared
+#include <string>   // std::string
+#include <tuple>    // std::tie
+#include <utility>  // std::move
 
 namespace plssvm {
 
 /**
- * @example model_examples.cpp
- * @brief A few examples regarding the plssvm::model class.
+ * @brief Implements a class encapsulating the result of a call to the C-SVR fit function. A model is used to predict the labels of a new data set.
+ * @tparam U the type of the used labels (must be an arithmetic type, except boolean or character types; default: `real_type`)
  */
-
-/**
- * @brief Implements a class encapsulating the result of a call to the SVM fit function. A model is used to predict the labels of a new data set.
- * @tparam U the type of the used labels (must be an arithmetic type or `std:string`; default: `int`)
- */
-template <typename U = int>
+template <typename U = real_type>
 class regression_model : public model<U> {
     // make sure only valid template types are used
     static_assert(detail::tuple_contains_v<U, detail::supported_label_types_regression>,
@@ -60,6 +49,7 @@ class regression_model : public model<U> {
     friend class csvm;
     friend class csvr;
 
+    /// The base model class.
     using base_model = model<U>;
 
     // TODO: better?
@@ -92,6 +82,13 @@ class regression_model : public model<U> {
     void save(const std::string &filename) const override;
 
   private:
+    /**
+     * @brief Create a new model using the SVM parameter @p params and the @p data.
+     * @details Default initializes the weights, i.e., no weights have currently been learned.
+     * @note This constructor may only be used in the befriended base C-SVM class!
+     * @param[in] params the SVM parameters used to learn this model
+     * @param[in] data the data used to learn this model
+     */
     regression_model(parameter params, regression_data_set<label_type> data);
 };
 

include/plssvm/regression_report.hpp

@@ -214,7 +214,7 @@ std::ostream &operator<<(std::ostream &out, const regression_report &report);
  * @param[in] metric the metric
  * @return the output-stream
  */
-std::ostream &operator<<(std::ostream &out, const regression_report::metric &m);
+std::ostream &operator<<(std::ostream &out, const regression_report::metric &metric);
 
 }  // namespace plssvm
 

include/plssvm/svm/csvc.hpp

@@ -46,6 +46,10 @@
 
 namespace plssvm {
 
+/**
+ * @brief Base class for all C-SVC backends.
+ * @details This class implements all features shared between all C-SVC backends. It defines the whole public API of a C-SVC.
+ */
 class csvc : virtual public csvm {
   public:
     /// The type of the model returned by a call to the `fit` function and used in the `predict` and `score` functions.
@@ -422,6 +426,14 @@ class csvc : virtual public csvm {
         return predicted_labels;
     }
 
+    /**
+     * @brief Calculate the accuracy of the @p model.
+     * @details Uses the one vs. all (OAA) for the multi-class classification task.
+     * @tparam label_type the type of the label (an arithmetic type or `std::string`)
+     * @param[in] model a previously learned model
+     * @throws plssvm::exception any exception thrown in the respective backend's implementation of `plssvm::csvm::predict_values`
+     * @return the accuracy of the model (`[[nodiscard]]`)
+     */
     template <typename label_type>
     [[nodiscard]] real_type score(const classification_model<label_type> &model) const {
         return this->score(model, model.data_);

include/plssvm/svm/csvm.hpp

@@ -76,7 +76,7 @@ namespace plssvm {
 
 /**
  * @brief Base class for all C-SVM backends.
- * @details This class implements all features shared between all C-SVM backends. It defines the whole public API of a C-SVM.
+ * @details This class implements all features shared between all C-SVM backends.
  */
 class csvm {
   public:

include/plssvm/svm/csvr.hpp

@@ -39,6 +39,10 @@
 
 namespace plssvm {
 
+/**
+ * @brief Base class for all C-SVR backends.
+ * @details This class implements all features shared between all C-SVR backends. It defines the whole public API of a C-SVR.
+ */
 class csvr : virtual public csvm {
   public:
     /// The type of the model returned by a call to the `fit` function and used in the `predict` and `score` functions.
@@ -128,8 +132,7 @@ class csvr : virtual public csvm {
     //*************************************************************************************************************************************//
     /**
      * @brief Predict the labels for the @p data set using the @p model.
-     * @details Uses the one vs. all (OAA) for the multi-class classification task.
-     * @tparam label_type the type of the label (an arithmetic type or `std::string`)
+     * @tparam label_type the type of the label
      * @param[in] model a previously learned model
      * @param[in] data the data to predict the labels for
      * @throws plssvm::invalid_parameter_exception if the number of features in the @p model's support vectors don't match the number of features in the @p data set
@@ -189,21 +192,27 @@ class csvr : virtual public csvm {
     }
 
     // TODO: not possible since the LIBSVM SVR model loses the original label values?!
+    // /**
+    //  * @brief Calculate the regression loss of the @p model.
+    //  * @tparam label_type the type of the label
+    //  * @param[in] model a previously learned model
+    //  * @throws plssvm::exception any exception thrown in the respective backend's implementation of `plssvm::csvm::predict_values`
+    //  * @return the regression loss of the model (`[[nodiscard]]`)
+    //  */
     // template <typename label_type>
     // [[nodiscard]] real_type score(const regression_model<label_type> &model) const {
     //     return this->score(model, model.data_);
     // }
 
     /**
-     * @brief Calculate the accuracy of the labeled @p data set using the @p model.
-     * @details Uses the one vs. all (OAA) for the multi-class classification task.
-     * @tparam label_type the type of the label (an arithmetic type or `std::string`)
+     * @brief Calculate the regression loss of the labeled @p data set using the @p model.
+     * @tparam label_type the type of the label
      * @param[in] model a previously learned model
      * @param[in] data the labeled data set to score
      * @throws plssvm::invalid_parameter_exception if the @p data to score has no labels
      * @throws plssvm::invalid_parameter_exception if the number of features in the @p model's support vectors don't match the number of features in the @p data set
      * @throws plssvm::exception any exception thrown in the respective backend's implementation of `plssvm::csvm::predict_values`
-     * @return the accuracy of the labeled @p data (`[[nodiscard]]`)
+     * @return the regression loss of the labeled @p data (`[[nodiscard]]`)
      */
     template <typename label_type>
     [[nodiscard]] real_type score(const regression_model<label_type> &model, const regression_data_set<label_type> &data) const {