TMVAExpert Class Reference

Expert for the TMVA MVA method. More...

#include <TMVA.h>

Inheritance diagram for TMVAExpert:

Collaboration diagram for TMVAExpert:

Public Member Functions
virtual void	load (Weightfile &weightfile) override
	Load the expert from a Weightfile. More...
virtual std::vector< float >	apply (Dataset &test_data) const =0
	Apply this expert onto a dataset. More...
virtual std::vector< float >	apply (Dataset &test_data, const unsigned int classID) const
	Apply this expert onto a dataset. More...

Protected Attributes
std::unique_ptr< TMVA::Reader >	m_expert
	TMVA::Reader pointer.
std::vector< float >	m_input_cache
	Input Cache for TMVA::Reader: Otherwise we would have to set the branch addresses in each apply call.
std::vector< float >	m_spectators_cache
	Spectators Cache for TMVA::Reader: Otherwise we would have to set the branch addresses in each apply call.
GeneralOptions	m_general_options
	General options loaded from the weightfile.

Detailed Description

Expert for the TMVA MVA method.

Definition at line 297 of file TMVA.h.

Member Function Documentation

apply() [1/2]

virtual std::vector<float> apply ( Dataset &  test_data ) const

pure virtualinherited

Apply this expert onto a dataset.

Parameters

test_data

dataset

Implemented in TMVAExpertRegression, TMVAExpertMulticlass, TMVAExpertClassification, RegressionExpert< BaseClassifierExpert, RegressionClassifierOptions >, FastBDTExpert, PythonExpert, FANNExpert, PDFExpert, ReweighterExpert, TrivialExpert, and CombinationExpert.

apply() [2/2]

virtual std::vector<float> apply ( Dataset &  test_data, const unsigned int  classID  ) const

inlinevirtualinherited

Apply this expert onto a dataset.

Multi-class mode signature. Not pure virtual, since not all derived classes need to re-implement this.

Parameters

test_data	dataset
classID	class identifier.

Reimplemented in TMVAExpertMulticlass.

Definition at line 59 of file Expert.h.

load()

void load ( Weightfile &  weightfile )

overridevirtual

Load the expert from a Weightfile.

Parameters

weightfile

containing all information necessary to build the expert

Implements Expert.

Reimplemented in TMVAExpertRegression, TMVAExpertMulticlass, and TMVAExpertClassification.

Definition at line 446 of file TMVA.cc.

                                                                            {
        m_expert->AddVariable(Belle2::makeROOTCompatible(general_options.m_variables[i]), &m_input_cache[i]);
      }
      m_spectators_cache.resize(general_options.m_spectators.size(), 0);
      for (unsigned int i = 0; i < general_options.m_spectators.size(); ++i) {
        m_expert->AddSpectator(Belle2::makeROOTCompatible(general_options.m_spectators[i]), &m_spectators_cache[i]);
      }
 
      if (weightfile.containsElement("TMVA_Logfile")) {
        std::string custom_weightfile = weightfile.generateFileName("logfile");
        weightfile.getFile("TMVA_Logfile", custom_weightfile);
      }
 
    }
 
    void TMVAExpertClassification::load(Weightfile& weightfile)
    {
 
      weightfile.getOptions(specific_options);
      if (specific_options.transform2probability) {
        expert_signalFraction = weightfile.getSignalFraction();
      }

---

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

• mva/methods/include/TMVA.h
• mva/methods/src/TMVA.cc