final_solutions_error.cpp

/****************************************************************************************************************/
/*                                                                                                              */
/*   OpenNN: Open Neural Networks Library                                                                       */
/*   www.artelnics.com/opennn                                                                                   */
/*                                                                                                              */
/*   F I N A L   S O L U T I O N S   E R R O R   C L A S S                                                      */
/*                                                                                                              */
/*   Roberto Lopez                                                                                              */
/*   Artelnics - Making intelligent use of data                                                                 */
/*   [email protected]                                                                                 */
/*                                                                                                              */
/****************************************************************************************************************/

// OpenNN includes

#include "final_solutions_error.h"

namespace OpenNN
{

// DEFAULT CONSTRUCTOR


FinalSolutionsError::FinalSolutionsError(void) 
 : PerformanceTerm()
{
   construct_numerical_differentiation();

   set_default();
}


// NEURAL NETWORK CONSTRUCTOR


FinalSolutionsError::FinalSolutionsError(NeuralNetwork* new_neural_network_pointer)
: PerformanceTerm(new_neural_network_pointer)
{
   construct_numerical_differentiation();

   set_default();
}


// MATHEMATICAL MODEL CONSTRUCTOR


FinalSolutionsError::FinalSolutionsError(MathematicalModel* new_mathematical_model_pointer)
: PerformanceTerm(new_mathematical_model_pointer)
{
   construct_numerical_differentiation();

   set_default();
}


// NEURAL NETWORK AND MATHEMATICAL MODEL CONSTRUCTOR


FinalSolutionsError::FinalSolutionsError(NeuralNetwork* new_neural_network_pointer, MathematicalModel* new_mathematical_model_pointer)
: PerformanceTerm(new_neural_network_pointer, new_mathematical_model_pointer)
{
   construct_numerical_differentiation();

   set_default();
}


// XML CONSTRUCTOR


FinalSolutionsError::FinalSolutionsError(const tinyxml2::XMLDocument& final_state_error_document)
 : PerformanceTerm(final_state_error_document)
{
   construct_numerical_differentiation();

   set_default();

   from_XML(final_state_error_document);
}


// COPY CONSTRUCTOR


FinalSolutionsError::FinalSolutionsError(const FinalSolutionsError& other_final_solutions_error)
 : PerformanceTerm()
{
//   set(other_final_solutions_error);

    neural_network_pointer = other_final_solutions_error.neural_network_pointer;

    data_set_pointer = other_final_solutions_error.data_set_pointer;

    mathematical_model_pointer = other_final_solutions_error.mathematical_model_pointer;

    if(other_final_solutions_error.numerical_differentiation_pointer)
    {
       numerical_differentiation_pointer = new NumericalDifferentiation(*other_final_solutions_error.numerical_differentiation_pointer);
    }

    display = other_final_solutions_error.display;

    final_solutions_errors_weights = other_final_solutions_error.final_solutions_errors_weights;

    target_final_solutions = other_final_solutions_error.target_final_solutions;

}


// DESTRUCTOR


FinalSolutionsError::~FinalSolutionsError(void)
{
}


// ASSIGNMENT OPERATOR

// FinalSolutionsError& operator = (const FinalSolutionsError&) method


FinalSolutionsError& FinalSolutionsError::operator = (const FinalSolutionsError& other_final_solutions_error)
{
   if(this != &other_final_solutions_error) 
   {
      *neural_network_pointer = *other_final_solutions_error.neural_network_pointer;
      *data_set_pointer = *other_final_solutions_error.data_set_pointer;
      *mathematical_model_pointer = *other_final_solutions_error.mathematical_model_pointer;
      *numerical_differentiation_pointer = *other_final_solutions_error.numerical_differentiation_pointer;
      display = other_final_solutions_error.display;

      final_solutions_errors_weights = other_final_solutions_error.final_solutions_errors_weights;
      target_final_solutions = other_final_solutions_error.target_final_solutions;
   }

   return(*this);
}


// EQUAL TO OPERATOR

// bool operator == (const FinalSolutionsError&) const method


bool FinalSolutionsError::operator == (const FinalSolutionsError& other_final_solutions_error) const
{
   if(*neural_network_pointer == *other_final_solutions_error.neural_network_pointer
   && *mathematical_model_pointer == *other_final_solutions_error.mathematical_model_pointer
   && *numerical_differentiation_pointer == *other_final_solutions_error.numerical_differentiation_pointer
   && display == other_final_solutions_error.display    
   && final_solutions_errors_weights == other_final_solutions_error.final_solutions_errors_weights
   && target_final_solutions == other_final_solutions_error.target_final_solutions)
   {
      return(true);
   }
   else
   {
      return(false);  
   }
}


// METHODS

// const Vector<double>& get_final_solutions_errors_weights(void) const method


const Vector<double>& FinalSolutionsError::get_final_solutions_errors_weights(void) const
{
   return(final_solutions_errors_weights);
}


// const Vector<double>& get_target_final_solutions(void) const method


const Vector<double>& FinalSolutionsError::get_target_final_solutions(void) const
{
   return(target_final_solutions);
}


// void set(void) method


void FinalSolutionsError::set(void)
{
   neural_network_pointer = NULL;
   data_set_pointer = NULL;
   mathematical_model_pointer = NULL;

   delete numerical_differentiation_pointer;

   numerical_differentiation_pointer = NULL;

   set_default();   
}


// void set(NeuralNetwork*) method


void FinalSolutionsError::set(NeuralNetwork*)
{
}


// void set(MathematicalModel*) method


void FinalSolutionsError::set(MathematicalModel*)
{
}


// void set(NeuralNetwork*, MathematicalModel*) method


void FinalSolutionsError::set(NeuralNetwork*, MathematicalModel*)
{
}


// void set(const FinalSolutionsError&) method

// Sets this final solutions error object with the members from another object of the same class.
// @param other_final_solutions_error Final solutions error object to be copied.

//void FinalSolutionsError::set(const FinalSolutionsError& other_final_solutions_error)
//{
//    neural_network_pointer = other_final_solutions_error.neural_network_pointer;

//    data_set_pointer = other_final_solutions_error.data_set_pointer;

//    mathematical_model_pointer = other_final_solutions_error.mathematical_model_pointer;

//    if(other_final_solutions_error.numerical_differentiation_pointer)
//    {
//        numerical_differentiation_pointer = new NumericalDifferentiation(*other_final_solutions_error.numerical_differentiation_pointer);
//    }

//    display = other_final_solutions_error.display;

//    final_solutions_errors_weights = other_final_solutions_error.final_solutions_errors_weights;

//    target_final_solutions = other_final_solutions_error.target_final_solutions;
//}


// void set_mathematical_model_pointer(MathematicalModel*) method


void FinalSolutionsError::set_mathematical_model_pointer(MathematicalModel* new_mathematical_model_pointer)
{
    mathematical_model_pointer = new_mathematical_model_pointer;

    if(mathematical_model_pointer)
    {
       const size_t dependent_variables_number = mathematical_model_pointer->get_dependent_variables_number();

       target_final_solutions.set(dependent_variables_number, 0.0);

       final_solutions_errors_weights.set(dependent_variables_number, 1.0);
    }
}


// void set_final_solutions_errors_weights(const Vector<double>&) method


void FinalSolutionsError::set_final_solutions_errors_weights(const Vector<double>& new_final_solutions_errors_weights)
{
   final_solutions_errors_weights = new_final_solutions_errors_weights;
}


// void set_final_solution_error_weight(const size_t&, const double&) method


void FinalSolutionsError::set_final_solution_error_weight(const size_t& i, const double& new_final_solution_error_weight)
{
   final_solutions_errors_weights[i] = new_final_solution_error_weight;
}


// void set_target_final_solutions(const Vector<double>&) method


void FinalSolutionsError::set_target_final_solutions(const Vector<double>& new_target_final_solutions)
{
   target_final_solutions = new_target_final_solutions;
}


// void set_target_final_solution(const size_t& const double&) method


void FinalSolutionsError::set_target_final_solution(const size_t& i, const double& new_target_final_solution)
{
   target_final_solutions[i] = new_target_final_solution;
}


// void set_default(void)


void FinalSolutionsError::set_default(void)
{
    if(mathematical_model_pointer)
    {
       const size_t dependent_variables_number = mathematical_model_pointer->get_dependent_variables_number();

       target_final_solutions.set(dependent_variables_number, 0.0);

       final_solutions_errors_weights.set(dependent_variables_number, 1.0);
    }
    else
    {
       target_final_solutions.set();

       final_solutions_errors_weights.set();   
    }

    display = true;
}


// void check(void) const method


void FinalSolutionsError::check(void) const
{
   std::ostringstream buffer;

   // Neural network stuff

   if(!neural_network_pointer)
   {
      buffer << "OpenNN Exception: FinalSolutionsError class.\n"
             << "void check(void) const method.\n"
             << "Pointer to neural network is NULL.\n";

      throw std::logic_error(buffer.str());   
   }

   const MultilayerPerceptron* multilayer_perceptron_pointer = neural_network_pointer->get_multilayer_perceptron_pointer();

   if(!multilayer_perceptron_pointer)
   {
      buffer << "OpenNN Exception: FinalSolutionsError class.\n"
             << "void check(void) const method.\n"
             << "Pointer to multilayer perceptron is NULL.\n";

      throw std::logic_error(buffer.str());   
   }

   const size_t inputs_number = multilayer_perceptron_pointer->get_inputs_number();
   const size_t outputs_number = multilayer_perceptron_pointer->get_outputs_number();

   if(inputs_number == 0)
   {
      buffer << "OpenNN Exception: FinalSolutionsError class.\n"
             << "void check(void) const method.\n"
             << "Number of inputs in multilayer perceptron object is zero.\n";

      throw std::logic_error(buffer.str());   
   }

   if(outputs_number == 0)
   {
      buffer << "OpenNN Exception: FinalSolutionsError class.\n"
             << "void check(void) const method.\n"
             << "Number of outputs in multilayer perceptron object is zero.\n";

      throw std::logic_error(buffer.str());   
   }

   // Mathematical model stuff

   if(!mathematical_model_pointer)
   {
      buffer << "OpenNN Exception: FinalSolutionsError class.\n"
             << "void check(void) const method.\n"
             << "Pointer to mathematical model is NULL.\n";

      throw std::logic_error(buffer.str());   
   }

   // Final solutions error stuff

   const size_t dependent_variables_number = mathematical_model_pointer->get_dependent_variables_number();

   const size_t target_final_solutions_size = target_final_solutions.size();

   if(target_final_solutions_size != dependent_variables_number)
   {
      std::ostringstream buffer;

      buffer << "OpenNN Exception: FinalSolutionsError class." << std::endl
             << "double calculate_performance(void) const method." << std::endl
             << "Size of target final solutions must be equal to number of dependent variables." << std::endl;

      throw std::logic_error(buffer.str());   
   }

}


// Vector<double> calculate_performance(void) const method

double FinalSolutionsError::calculate_performance(void) const   
{
   // Control sentence

   #ifndef NDEBUG 

   check();

   #endif

   // Final state error stuff

   const size_t independent_variables_number = mathematical_model_pointer->get_independent_variables_number();
   const size_t dependent_variables_number = mathematical_model_pointer->get_dependent_variables_number();

   const Vector<double> final_solutions = mathematical_model_pointer->calculate_final_solutions(*neural_network_pointer);

   const Vector<double> dependent_variables_final_solutions = final_solutions.take_out(independent_variables_number, dependent_variables_number);

   const Vector<double> final_solutions_errors = dependent_variables_final_solutions - target_final_solutions;

   return((final_solutions_errors_weights*final_solutions_errors*final_solutions_errors).calculate_sum());
}


// double calculate_performance(const Vector<double>&) const method

double FinalSolutionsError::calculate_performance(const Vector<double>&) const
{
    return(0.0);
}


// std::string write_performance_term_type(void) const method


std::string FinalSolutionsError::write_performance_term_type(void) const
{
   return("FINAL_SOLUTIONS_ERROR");
}


// std::string write_information(void) const method

std::string FinalSolutionsError::write_information(void) const
{
   std::ostringstream buffer;

   const size_t independent_variables_number = mathematical_model_pointer->get_independent_variables_number();
   const size_t dependent_variables_number = mathematical_model_pointer->get_dependent_variables_number();

   const Vector<double> final_solutions = mathematical_model_pointer->calculate_final_solutions(*neural_network_pointer);

   const Vector<double> dependent_variables_final_solutions = final_solutions.take_out(independent_variables_number, dependent_variables_number);

   const Vector<double> final_solutions_errors = dependent_variables_final_solutions - target_final_solutions;

   const double performance = (final_solutions_errors_weights*final_solutions_errors*final_solutions_errors).calculate_sum();

   buffer << "Final solutions error\n"
          << "Target final solutions: " << target_final_solutions << "\n"
          << "Final solutions: " << dependent_variables_final_solutions << "\n"
          << "performance: " << performance << "\n";

   return(buffer.str());
}


// tinyxml2::XMLDocument* to_XML(void) method method 


tinyxml2::XMLDocument* FinalSolutionsError::to_XML(void) const
{
   std::ostringstream buffer;

   tinyxml2::XMLDocument* document = new tinyxml2::XMLDocument;

   // Final solutions error

   tinyxml2::XMLElement* final_solutions_error_element = document->NewElement("FinalSolutionsError");

   document->InsertFirstChild(final_solutions_error_element);

   // Numerical differentiation
   {
      if(numerical_differentiation_pointer)
      {
         tinyxml2::XMLElement* element = numerical_differentiation_pointer->to_XML()->FirstChildElement();

         if(element)
         {
            final_solutions_error_element->LinkEndChild(element);
         }
      }
   }

   // Final solutions errors weights
   {
      tinyxml2::XMLElement* element = document->NewElement("FinalSolutionsErrorsWeights");
      final_solutions_error_element->LinkEndChild(element);

      buffer.str("");
      buffer << final_solutions_errors_weights;

      tinyxml2::XMLText* text = document->NewText(buffer.str().c_str());
      element->LinkEndChild(text);
   }

   // Target final solution
   {
      tinyxml2::XMLElement* element = document->NewElement("TargetFinalSolution");
      final_solutions_error_element->LinkEndChild(element);

      buffer.str("");
      buffer << target_final_solutions;

      tinyxml2::XMLText* text = document->NewText(buffer.str().c_str());
      element->LinkEndChild(text);
   }

   // Display
   {
      tinyxml2::XMLElement* display_element = document->NewElement("Display");
      final_solutions_error_element->LinkEndChild(display_element);

      buffer.str("");
      buffer << display;

      tinyxml2::XMLText* display_text = document->NewText(buffer.str().c_str());
      display_element->LinkEndChild(display_text);
   }

   return(document);
}


// void from_XML(const tinyxml2::XMLDocument&) method


void FinalSolutionsError::from_XML(const tinyxml2::XMLDocument& document)
{
    const tinyxml2::XMLElement* root_element = document.FirstChildElement("FinalSolutionsError");

    if(!root_element)
    {
        std::ostringstream buffer;

        buffer << "OpenNN Exception: FinalSolutionsError class.\n"
               << "void from_XML(const tinyxml2::XMLDocument&) method.\n"
               << "Final solutions error element is NULL.\n";

        throw std::logic_error(buffer.str());
    }

  // Display
  {
     const tinyxml2::XMLElement* display_element = root_element->FirstChildElement("Display");

     if(display_element)
     {
        const std::string new_display_string = display_element->GetText();

        try
        {
           set_display(new_display_string != "0");
        }
        catch(const std::logic_error& e)
        {
           std::cout << e.what() << std::endl;
        }
     }
  }
}


}


// OpenNN: Open Neural Networks Library.
// Copyright (c) 2005-2015 Roberto Lopez.
//
// This library is free software; you can redistribute it and/or
// modify it under the terms of the GNU Lesser General Public
// License as published by the Free Software Foundation; either
// version 2.1 of the License, or any later version.
//
// This library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
// Lesser General Public License for more details.
// You should have received a copy of the GNU Lesser General Public
// License along with this library; if not, write to the Free Software
// Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA