example_ensembledecisiontree2.cpp

Programa para el testeo de las funciones.Para compilar o código example_ensembledecisiontree2.cpp:

g++ -static -o example_ensembledecisiontree2 example_ensembledecisiontree2.cpp -lpdsmlmm -lpdsramm -lpdsspmm

Para executar o programa:

./example_ensembledecisiontree2

Retornando por consola:

Pds::DecisionTree::Counter: 71
Trained bag   0 with 2000 samples obtaining  49 decision.
Trained bag   1 with 2000 samples obtaining  63 decision.
Trained bag   2 with 2000 samples obtaining  65 decision.
Trained bag   3 with 2000 samples obtaining  59 decision.
Trained bag   4 with 2000 samples obtaining  61 decision.
Trained bag   5 with 2000 samples obtaining  57 decision.
Trained bag   6 with 2000 samples obtaining  49 decision.
Trained bag   7 with 2000 samples obtaining  43 decision.
Trained bag   8 with 2000 samples obtaining  63 decision.
Trained bag   9 with 2000 samples obtaining  61 decision.
Trained bag  10 with 2000 samples obtaining  47 decision.
Trained bag  11 with 2000 samples obtaining  45 decision.
Trained bag  12 with 2000 samples obtaining  49 decision.
Trained bag  13 with 2000 samples obtaining  61 decision.
Trained bag  14 with 2000 samples obtaining  51 decision.
Trained bag  15 with 2000 samples obtaining  49 decision.

Código example_ensembledecisiontree2.cpp:

 
#include <iostream>
#include <Pds/Ra>
#include <Pds/Ml>
    
int main(void)
{
    Pds::Octave::XLabel="x_1";
    Pds::Octave::YLabel="x_2";
    Pds::Vector Yp;
    Pds::IterationConf Conf;
    Conf.Show=false; Conf.SetMinError(1e-07);    Conf.SetAlpha(0.1);
    
    // Load training data
    unsigned int L=2000;
    Pds::Matrix X;
    Pds::Vector Y;
    Pds::DataSet::Classification::LoadDataYinYang(L,X,Y);
 
    Pds::Octave::Plot::PointsX2DY(X,Y,"testando.m","example_ensembledecisiontree2_data_yinyang.png");
    
    unsigned int StopSamples=3;
    unsigned int StopDeep=64;
    
    // Training DecisionTree
    Pds::DecisionTree::Counter=0;
    Pds::DecisionTree Arbol("k2means",Conf,X,Y,0.8,StopSamples,StopDeep);
    std::cout<<"Pds::DecisionTree::Counter: "<<Pds::DecisionTree::Counter<<std::endl;
    
    Arbol.ExportDotFile("arbol.dot");
    
    // Predict training data
    Yp=Arbol.Predict(X);
    std::cout<<"Acc:"<<Pds::Accuracy(Y,Yp,0.5)<<"\n";
    
    Pds::Octave::Plot::PointsX2DY(X,Yp,"testando.m","example_ensembledecisiontree2_simple_yinyang.png");
    
    std::cout<<Pds::Ra::Repeat("*",128)<<"\n\n";
    
    // Training Ensemble Decisiontree
    Pds::DecisionTree::Counter=0;
    Pds::EnsembleDecisionTree Arbol2(Conf,X,Y,16,0.5,0.8,StopSamples,StopDeep);
    std::cout<<"Pds::DecisionTree::Counter: "<<Pds::DecisionTree::Counter<<std::endl;
    Arbol2.GetStackingClassifier().Print("Arbol2.GetStackingClassifier():\n");
    
    Arbol2.ExportFile("EnsembleDecisionTree.txt");
    
    // Predict testing data
    Yp=Arbol2.Predict(X);
    std::cout<<"Acc:"<<Pds::Accuracy(Y,Yp,0.5)<<"\n";
    
    Pds::Octave::Plot::PointsX2DY(X,Yp,"testando.m","example_ensembledecisiontree2_ensemble_yinyang.png");
    
    std::cout<<Pds::Ra::Repeat("*",128)<<"\n\n";
    
    // Ensemble DecisionTree
    Pds::EnsembleDecisionTree Arbol3("EnsembleDecisionTree.txt");
    Arbol3.GetStackingClassifier().Print("Arbol3.GetStackingClassifier():\n");
    
    // Predict testing data
    Yp=Arbol3.Predict(X);
    std::cout<<"Acc:"<<Pds::Accuracy(Y,Yp,0.5)<<"\n";
    
    Pds::Octave::Plot::PointsX2DY(X,Yp,"testando.m","example_ensembledecisiontree2_import_yinyang.png");
    
    return 0;
}