GEGELATI
Public Member Functions | Protected Attributes | List of all members
Learn::EvaluationResult Class Reference

Base class for storing all result of a policy evaluation within a LearningEnvironment. More...

#include <evaluationResult.h>

Inheritance diagram for Learn::EvaluationResult:
Learn::AdversarialEvaluationResult Learn::ClassificationEvaluationResult

Public Member Functions

 EvaluationResult ()=delete
 Deleted default constructor.
 
virtual ~EvaluationResult ()=default
 Virtual destructor for polymorphism.
 
 EvaluationResult (const double &res, const size_t &nbEval)
 Construct a result from a simple double value. More...
 
virtual double getResult () const
 Virtual method to get the default double equivalent of the EvaluationResult. More...
 
virtual size_t getNbEvaluation () const
 Virtual method to get the default number of evaluation of the EvaluationResult.
 
virtual EvaluationResultoperator+= (const EvaluationResult &other)
 Polymorphic addition assignement operator for EvaluationResult. More...
 

Protected Attributes

double result
 Double value for the result.
 
size_t nbEvaluation
 Number of evaluation leading to this result.
 

Detailed Description

Base class for storing all result of a policy evaluation within a LearningEnvironment.

To enable generic learning with the default LearningAgent, all policy evaluation must be convertible to a simple double value. For more complex LearningAgent behavior, like for classification purposes, specific child class can be created.

Constructor & Destructor Documentation

◆ EvaluationResult()

Learn::EvaluationResult::EvaluationResult ( const double &  res,
const size_t &  nbEval 
)
inline

Construct a result from a simple double value.

Parameters
[in]resthe double value representing the result of an evaluation.
[in]nbEvalInteger value representing the number of evaluation leading to the recorded score.

Member Function Documentation

◆ getResult()

double Learn::EvaluationResult::getResult ( ) const
virtual

Virtual method to get the default double equivalent of the EvaluationResult.

Copyright or © or Copr. IETR/INSA - Rennes (2019 - 2020) :

Karol Desnos kdesn.nosp@m.os@i.nosp@m.nsa-r.nosp@m.enne.nosp@m.s.fr (2019 - 2020)

GEGELATI is an open-source reinforcement learning framework for training artificial intelligence based on Tangled Program Graphs (TPGs).

This software is governed by the CeCILL-C license under French law and abiding by the rules of distribution of free software. You can use, modify and/ or redistribute the software under the terms of the CeCILL-C license as circulated by CEA, CNRS and INRIA at the following URL "http://www.cecill.info".

As a counterpart to the access to the source code and rights to copy, modify and redistribute granted by the license, users are provided only with a limited warranty and the software's author, the holder of the economic rights, and the successive licensors have only limited liability.

In this respect, the user's attention is drawn to the risks associated with loading, using, modifying and/or developing or reproducing the software by the user in light of its specific status of free software, that may mean that it is complicated to manipulate, and that also therefore means that it is reserved for developers and experienced professionals having in-depth computer knowledge. Users are therefore encouraged to load and test the software's suitability as regards their requirements in conditions enabling the security of their systems and/or data to be ensured and, more generally, to use and operate it in the same conditions as regards security.

The fact that you are presently reading this means that you have had knowledge of the CeCILL-C license and that you accept its terms.

Reimplemented in Learn::AdversarialEvaluationResult.

◆ operator+=()

Learn::EvaluationResult & Learn::EvaluationResult::operator+= ( const EvaluationResult other)
virtual

Polymorphic addition assignement operator for EvaluationResult.

Exceptions
std::runtime_errorin case the other EvaluationResult and this have a different typeid.

Reimplemented in Learn::AdversarialEvaluationResult, and Learn::ClassificationEvaluationResult.


The documentation for this class was generated from the following files: