void TMVAClassificationApplication_TX(TString myMethodList = "" , TString iFileName = "", TString sampleLocation = "", TString outputLocation = "")
{
#ifdef __CINT__
gROOT->ProcessLine( ".O0" ); // turn off optimization in CINT
#endif
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
// --- Cut optimisation
Use["Cuts"] = 0;
Use["CutsD"] = 0;
Use["CutsPCA"] = 0;
Use["CutsGA"] = 0;
Use["CutsSA"] = 0;
//
//
// --- Boosted Decision Trees
Use["BDT"] = 1; // uses Adaptive Boost
std::cout << std::endl;
std::cout << "==> Start TMVAClassificationApplication" << std::endl;
// Select methods (don't look at this code - not of interest)
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod
<< "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
std::cout << it->first << " ";
}
std::cout << std::endl;
return;
}
Use[regMethod] = 1;
}
}
// --------------------------------------------------------------------------------------------------
// --- Create the Reader object
TMVA::Reader *reader = new TMVA::Reader("!Color:!Silent" );
// Create a set of variables and declare them to the reader
// - the variable names MUST corresponds in name and type to those given in the weight file(s) used
Float_t var1, var2, var3, var4, var5, var6, var7, var8, var9, var10, var11, var12, var13, var14, var15, var16, var17, var18, var19, var20, var21, var22, var23, var24, var25, var26, var27, var28, var29;
//reader->AddVariable( "Alt$(jet_pt_singleLepCalc[0],0)", &var1);
//reader->AddVariable( "Alt$(jet_pt_singleLepCalc[1],0)", &var2 );
//reader->AddVariable( "Alt$(jet_pt_singleLepCalc[2],0)", &var3 );
reader->AddVariable( "Alt$(bJetPt_CATopoCalc[0],0)", &var4 );
reader->AddVariable( "Alt$(bJetPt_CATopoCalc[1],0)", &var5 );
//reader->AddVariable( "corr_met_singleLepCalc", &var6 );
//reader->AddVariable( "muon_1_pt_singleLepCalc", &var7 );
//reader->AddVariable( "nBJets_CATopoCalc", &var8 );
//reader->AddVariable( "nSelJets_CommonCalc", &var9 );
//reader->AddVariable( "LeptonJet_DeltaR_LjetsTopoCalcNew", &var10);
reader->AddVariable( "Mevent_LjetsTopoCalcNew", &var11);
//reader->AddVariable( "W_Pt_LjetsTopoCalcNew", &var12 );
reader->AddVariable( "Jet1Jet2_Pt_LjetsTopoCalcNew", &var13 );
//reader->AddVariable( "BestTop_LjetsTopoCalcNew", &var14 );
//reader->AddVariable( "BTagTopMass_LjetsTopoCalcNew", &var15 );
//reader->AddVariable( "Alt$(CAHEPTopJetMass_JetSubCalc[0],0)", &var16 );
//reader->AddVariable( "Alt$(CAWCSVMSubJets_JetSubCalc[0],0)", &var17 );
//reader->AddVariable( "Alt$(CAWCSVLSubJets_JetSubCalc[0],0)", &var18 );
reader->AddVariable( "Alt$(CAWJetPt_JetSubCalc[0],0)", &var19 );
reader->AddVariable( "Alt$(CAWJetMass_JetSubCalc[0],0)", &var20 );
//reader->AddVariable( "Alt$(CAHEPTopJetMass_JetSubCalc[1],0)", &var21 );
//reader->AddVariable( "Hz_LjetsTopoCalcNew", &var22 );
//reader->AddVariable( "Centrality_LjetsTopoCalcNew", &var23 );
reader->AddVariable( "SqrtsT_LjetsTopoCalcNew", &var24 );
reader->AddVariable( "CAMindrBMass_CATopoCalc", &var28 );
reader->AddVariable( "minDRCAtoB_CATopoCalc", &var29 );
//reader->AddVariable( "HT2prime_LjetsTopoCalcNew", &var25 );
reader->AddVariable( "HT2_LjetsTopoCalcNew", &var26 );
//reader->AddVariable( "dphiLepMet_LjetsTopoCalcNew", &var27 );
// Spectator variables declared in the training have to be added to the reader, too
// Float_t spec1,spec2;
// reader->AddSpectator( "spec1 := var1*2", &spec1 );
// reader->AddSpectator( "spec2 := var1*3", &spec2 );
// Float_t Category_cat1, Category_cat2, Category_cat3;
// if (Use["Category"]){
// // Add artificial spectators for distinguishing categories
// reader->AddSpectator( "Category_cat1 := var3<=0", &Category_cat1 );
// reader->AddSpectator( "Category_cat2 := (var3>0)&&(var4<0)", &Category_cat2 );
// reader->AddSpectator( "Category_cat3 := (var3>0)&&(var4>=0)", &Category_cat3 );
// }
// --- Book the MVA methods
// Book method(s)
TString weightFileName = "weights/TMVAClassification_BDT.weights";
reader->BookMVA("BDT method", weightFileName+".xml" );
// Book output histograms
UInt_t nbin = 100;
TH1F *histBdt(0);
histBdt = new TH1F( "MVA_BDT_TX", "MVA_BDT_TX", nbin, -1.0, 1.0);
// Prepare input tree (this must be replaced by your data source)
// in this example, there is a toy tree with signal and one with background events
// we'll later on use only the "signal" events for the test in this example.
//
TFile *input(0);
TString fileName = iFileName;
TString fname = sampleLocation+"/";
fname += fileName;
TString oFileName = fileName;
if (!gSystem->AccessPathName( fname ))
input = TFile::Open( fname ); // check if file in local directory exists
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
// --- Event loop
// Prepare the event tree
// - here the variable names have to corresponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
std::cout << "--- Select signal sample" << std::endl;
TTree* theTree = (TTree*)input->Get("ljmet");
gSystem->mkdir( outputLocation );
TFile *target = new TFile( outputLocation+"/"+oFileName,"RECREATE" );
TTree *newTree = theTree->CloneTree();
Float_t BDT;
TBranch *branchBDT = newTree->Branch("__BDT_TX__",&BDT,"__BDT_TX__/F");
std::vector<Double_t> *vecVar1;
std::vector<Double_t> *vecVar4;
std::vector<Double_t> *vecVar16;
std::vector<Int_t> *vecVar17;
std::vector<Int_t> *vecVar18;
std::vector<Double_t> *vecVar19;
std::vector<Double_t> *vecVar20;
Int_t *intVar5, *intVar8, *intVar9;
Double_t *dVar2, *dVar3, *dVar6, *dVar7, *dVar10, *dVar11, *dVar12, *dVar13, *dVar14, *dVar15, *dVar22, *dVar23, *dVar24, dVar25, *dVar26, *dVar27, *dVar28, *dVar29;
theTree->SetBranchAddress( "jet_pt_singleLepCalc", &vecVar1);
theTree->SetBranchAddress( "bJetPt_CATopoCalc", &vecVar4 );
theTree->SetBranchAddress( "corr_met_singleLepCalc", &dVar6 );
theTree->SetBranchAddress( "muon_1_pt_singleLepCalc", &dVar7 );
theTree->SetBranchAddress( "nBJets_CATopoCalc", &intVar8 );
theTree->SetBranchAddress( "nSelJets_CommonCalc", &intVar9 );
theTree->SetBranchAddress( "LeptonJet_DeltaR_LjetsTopoCalcNew", &dVar10);
theTree->SetBranchAddress( "Mevent_LjetsTopoCalcNew", &dVar11);
theTree->SetBranchAddress( "W_Pt_LjetsTopoCalcNew", &dVar12 );
theTree->SetBranchAddress( "Jet1Jet2_Pt_LjetsTopoCalcNew", &dVar13 );
theTree->SetBranchAddress( "BestTop_LjetsTopoCalcNew", &dVar14 );
theTree->SetBranchAddress( "BTagTopMass_LjetsTopoCalcNew", &dVar15 );
theTree->SetBranchAddress( "CAHEPTopJetMass_JetSubCalc", &vecVar16 );
theTree->SetBranchAddress( "CAWCSVMSubJets_JetSubCalc", &vecVar17 );
theTree->SetBranchAddress( "CAWCSVLSubJets_JetSubCalc", &vecVar18 );
theTree->SetBranchAddress( "CAWJetPt_JetSubCalc", &vecVar19 );
theTree->SetBranchAddress( "CAWJetMass_JetSubCalc", &vecVar20 );
theTree->SetBranchAddress( "Hz_LjetsTopoCalcNew", &dVar22 );
theTree->SetBranchAddress( "Centrality_LjetsTopoCalcNew", &dVar23 );
theTree->SetBranchAddress( "SqrtsT_LjetsTopoCalcNew", &dVar24 );
theTree->SetBranchAddress( "HT2prime_LjetsTopoCalcNew", &dVar25 );
theTree->SetBranchAddress( "HT2_LjetsTopoCalcNew", &dVar26 );
theTree->SetBranchAddress( "dphiLepMet_LjetsTopoCalcNew", &dVar27 );
theTree->SetBranchAddress( "CAMindrBMass_CATopoCalc", &dVar28 );
theTree->SetBranchAddress( "minDRCAtoB_CATopoCalc", &dVar29 );
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
if (ievt%1000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
theTree->GetEntry(ievt);
if(vecVar1->size()>0){
var1 = vecVar1->at(0);
}
if(vecVar1->size()>1){
var2 = vecVar1->at(1);
}
if(vecVar1->size()>2){
var3 = vecVar1->at(2);
}
if(vecVar4->size()>0){
var4 = vecVar4->at(0);
}
if(vecVar4->size()>1){
var5 = vecVar4->at(1);
}
var6 = dVar6;
var7 = dVar7;
var8 = intVar8;
var9 = intVar9;
var10 = dVar10;
var11 = dVar11;
var12 = dVar12;
var13 = dVar13;
var14 = dVar14;
var15 = dVar15;
if(vecVar16->size()>0){
var16 = vecVar16->at(0);
}
else{
var16 = 0;
}
if(vecVar17->size()>0){
var17 = vecVar17->at(0);
}
else{
var18 = 0;
}
if(vecVar19->size()>0){
var19 = vecVar19->at(0);
}
else{
var19 = 0;
}
if(vecVar20->size()>0){
var20 = vecVar20->at(0);
}
else{
var20 = 0;
}
if(vecVar16->size()>1){
var21 = vecVar16->at(1);
}
else{
var21 = 0;
}
var22 = dVar22;
var23 = dVar23;
var24 = dVar24;
var25 = dVar25;
var26 = dVar26;
var27 = dVar27;
var28 = dVar28;
var29 = dVar29; // --- Return the MVA outputs and fill into histograms
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
BDT = reader->EvaluateMVA( "BDT method");
histBdt->Fill(BDT);
branchBDT->Fill();
}
// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// Get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
// --- Write histograms
newTree->Write("",TObject::kOverwrite);
target->Close();
std::cout << "--- Created root file: \""<<oFileName<<"\" containing the MVA output histograms" << std::endl;
delete reader;
std::cout << "==> TMVAClassificationApplication is done!" << endl << std::endl;
}
void TMVAClassificationApplication_new(TString myMethodList = "" , TString iFileName = "", TString bkgSample = "", TString sampleLocation = "", TString massPoint = "", TString oFileLocation = "")
{
#ifdef __CINT__
gROOT->ProcessLine( ".O0" ); // turn off optimization in CINT
#endif
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
// --- Cut optimisation
Use["Cuts"] = 0;
Use["CutsD"] = 0;
Use["CutsPCA"] = 0;
Use["CutsGA"] = 0;
Use["CutsSA"] = 0;
//
//
// --- Boosted Decision Trees
Use["BDT"] = 1; // uses Adaptive Boost
std::cout << std::endl;
std::cout << "==> Start TMVAClassificationApplication" << std::endl;
// Select methods (don't look at this code - not of interest)
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod
<< "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
std::cout << it->first << " ";
}
std::cout << std::endl;
return;
}
Use[regMethod] = 1;
}
}
// --------------------------------------------------------------------------------------------------
// --- Create the Reader object
TMVA::Reader *reader = new TMVA::Reader("!Color:!Silent" );
TString weightTail = "_";
weightTail = weightTail + massPoint;
// Create a set of variables and declare them to the reader
// - the variable names MUST corresponds in name and type to those given in the weight file(s) used
Float_t var1, var2, var3, var4, var5, var6, var7, var8, var9, var10, var11, var12, var13, var14, var15, var16, var17, var18;
reader->AddVariable( "svMass", &var1);
reader->AddVariable( "dRTauTau", &var3 );
reader->AddVariable( "dRJJ", &var4 );
// reader->AddVariable( "svPt", &var5 );
// reader->AddVariable( "dRhh", &var6 );
reader->AddVariable( "met", &var7 );
reader->AddVariable( "mJJ", &var8 );
// reader->AddVariable( "metTau1DPhi", &var9 );
// reader->AddVariable( "metTau2DPhi", &var10);
// reader->AddVariable( "metJ1DPhi", &var11);
// reader->AddVariable( "metJ2DPhi", &var12 );
// reader->AddVariable( "metTauPairDPhi", &var13 );
// reader->AddVariable( "metSvTauPairDPhi", &var14 );
// reader->AddVariable( "metJetPairDPhi", &var15 );
// reader->AddVariable( "CSVJ1", &var16 );
// reader->AddVariable( "CSVJ2", &var17 );
reader->AddVariable( "fMassKinFit", &var2 );
reader->AddVariable( "chi2KinFit2", &var18 );
// Spectator variables declared in the training have to be added to the reader, too
// Float_t spec1,spec2;
// reader->AddSpectator( "spec1 := var1*2", &spec1 );
// reader->AddSpectator( "spec2 := var1*3", &spec2 );
// Float_t Category_cat1, Category_cat2, Category_cat3;
// if (Use["Category"]){
// // Add artificial spectators for distinguishing categories
// reader->AddSpectator( "Category_cat1 := var3<=0", &Category_cat1 );
// reader->AddSpectator( "Category_cat2 := (var3>0)&&(var4<0)", &Category_cat2 );
// reader->AddSpectator( "Category_cat3 := (var3>0)&&(var4>=0)", &Category_cat3 );
// }
// --- Book the MVA methods
// Book method(s)
TString weightFileName = "/nfs_scratch/zmao/test/CMSSW_5_3_15/src/TMVA-v4.2.0/test/weights/TMVAClassification_BDT.weights_";
weightFileName += bkgSample;
weightFileName += weightTail;
reader->BookMVA("BDT method", weightFileName+".xml" );
// Book output histograms
UInt_t nbin = 200;
TH1F *histBdt(0);
histBdt = new TH1F( "MVA_BDT", "MVA_BDT", nbin, -1.0, 1.0);
// Prepare input tree (this must be replaced by your data source)
// in this example, there is a toy tree with signal and one with background events
// we'll later on use only the "signal" events for the test in this example.
//
TFile *input(0);
TString fileName = iFileName;
TString fname = sampleLocation;
fname += fileName;
TString oFileName = oFileLocation;
oFileName += "ClassApp_" + bkgSample;
oFileName += "_";
oFileName += fileName;
if (!gSystem->AccessPathName( fname ))
input = TFile::Open(fname); // check if file in local directory exists
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
// --- Event loop
// Prepare the event tree
// - here the variable names have to corresponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
std::cout << "--- Select signal sample" << std::endl;
TTree* theTree = (TTree*)input->Get("eventTree");
TFile *target = new TFile( oFileName,"RECREATE" );
TTree *newTree = theTree->CloneTree();
Float_t BDT;
TBranch *branchBDT = newTree->Branch("BDT_"+bkgSample,&BDT,"BDT/F");
std::vector<Double_t> *vecVar1;
std::vector<Double_t> *vecVar5;
std::vector<Double_t> *vecVar7;
theTree->SetBranchAddress( "svMass", &vecVar1);
theTree->SetBranchAddress( "dRTauTau", &var3);
theTree->SetBranchAddress( "dRJJ", &var4 );
// theTree->SetBranchAddress( "svPt", &vecVar5 );
// theTree->SetBranchAddress( "dRhh", &var6 );
theTree->SetBranchAddress( "met", &vecVar7 );
theTree->SetBranchAddress( "mJJ", &var8 );
// theTree->SetBranchAddress( "metTau1DPhi", &var9 );
// theTree->SetBranchAddress( "metTau2DPhi", &var10);
// theTree->SetBranchAddress( "metJ1DPhi", &var11);
// theTree->SetBranchAddress( "metJ2DPhi", &var12 );
// theTree->SetBranchAddress( "metTauPairDPhi", &var13 );
// theTree->SetBranchAddress( "metSvTauPairDPhi", &var14 );
// theTree->SetBranchAddress( "metJetPairDPhi", &var15 );
// theTree->SetBranchAddress( "CSVJ1", &var16 );
// theTree->SetBranchAddress( "CSVJ2", &var17 );
theTree->SetBranchAddress( "fMassKinFit", &var2);
theTree->SetBranchAddress( "chi2KinFit2", &var18);
//to get initial pre-processed events
TH1F* cutFlow = (TH1F*)input->Get("preselection");
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
if (ievt%1000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
theTree->GetEntry(ievt);
var1 = vecVar1->at(0);
// var5 = vecVar5->at(0);
var7 = vecVar7->at(0);
// --- Return the MVA outputs and fill into histograms
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
BDT = reader->EvaluateMVA( "BDT method");
histBdt->Fill(BDT);
branchBDT->Fill();
}
// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// Get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
// --- Write histograms
histBdt->Write();
cutFlow->Write();
newTree->Write();
target->Close();
std::cout << "--- Created root file: \""<<oFileName<<"\" containing the MVA output histograms" << std::endl;
delete reader;
std::cout << "==> TMVAClassificationApplication is done!" << endl << std::endl;
}
void TMVAClassificationApplication( TString myMethodList = "" )
{
//---------------------------------------------------------------
// default MVA methods to be trained + tested
// this loads the library
TMVA::Tools::Instance();
std::map<std::string,int> Use;
Use["CutsGA"] = 0; // other "Cuts" methods work identically
// ---
Use["Likelihood"] = 1;
Use["LikelihoodD"] = 0; // the "D" extension indicates decorrelated input variables (see option strings)
Use["LikelihoodPCA"] = 0; // the "PCA" extension indicates PCA-transformed input variables (see option strings)
Use["LikelihoodKDE"] = 0;
Use["LikelihoodMIX"] = 0;
// ---
Use["PDERS"] = 0;
Use["PDERSD"] = 0;
Use["PDERSPCA"] = 0;
Use["PDERSkNN"] = 0; // depreciated until further notice
Use["PDEFoam"] = 0;
// --
Use["KNN"] = 0;
// ---
Use["HMatrix"] = 0;
Use["Fisher"] = 0;
Use["FisherG"] = 0;
Use["BoostedFisher"] = 0;
Use["LD"] = 0;
// ---
Use["FDA_GA"] = 0;
Use["FDA_SA"] = 0;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 0;
Use["FDA_GAMT"] = 0;
Use["FDA_MCMT"] = 0;
// ---
Use["MLP"] = 0; // this is the recommended ANN
Use["MLPBFGS"] = 0; // recommended ANN with optional training method
Use["MLPBNN"] = 0; //
Use["CFMlpANN"] = 0; // *** missing
Use["TMlpANN"] = 0;
// ---
Use["SVM"] = 0;
// ---
Use["BDT"] = 1;
Use["BDTD"] = 0;
Use["BDTG"] = 0;
Use["BDTB"] = 0;
// ---
Use["RuleFit"] = 0;
// ---
Use["Category"] = 0;
// ---
Use["Plugin"] = 0;
// ---------------------------------------------------------------
std::cout << std::endl;
std::cout << "==> Start TMVAClassificationApplication" << std::endl;
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod << "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) std::cout << it->first << " ";
std::cout << std::endl;
return;
}
Use[regMethod] = 0;
}
}
//
// create the Reader object
//
TMVA::Reader *reader = new TMVA::Reader( "!Color:!Silent" );
Float_t Z_rapidity_z;
reader->AddVariable("Z_rapidity_z",&Z_rapidity_z);
Float_t THRUST_2D;
reader->AddVariable("THRUST_2D",&THRUST_2D);
Float_t L1_L2_cosangle;
reader->AddVariable("L1_L2_cosangle",&L1_L2_cosangle);
Float_t TransMass_ZH150_uncl;
reader->AddVariable("TransMass_ZH150_uncl",&TransMass_ZH150_uncl);
Float_t TransMass_ZH150;
reader->AddVariable("TransMass_ZH150",&TransMass_ZH150);
Float_t DeltaPhi_ZH;
reader->AddVariable("DeltaPhi_ZH",&DeltaPhi_ZH);
Float_t DeltaPhi_ZH_uncl;
reader->AddVariable("DeltaPhi_ZH_uncl",&DeltaPhi_ZH_uncl);
Float_t CMAngle;
reader->AddVariable("CMAngle",&CMAngle);
Float_t CS_cosangle;
reader->AddVariable("CS_cosangle",&CS_cosangle);
// create a set of variables and declare them to the reader
// - the variable names must corresponds in name and type to
// those given in the weight file(s) that you use
Float_t var1, var2;
Float_t var3, var4;
// reader->AddVariable( "myvar1 := var1+var2", &var1 );
// reader->AddVariable( "myvar2 := var1-var2", &var2 );
// reader->AddVariable( "var3", &var3 );
// reader->AddVariable( "var4", &var4 );
//Spectator variables declared in the training have to be added to the reader, too
Float_t spec1,spec2;
// reader->AddSpectator( "spec1 := var1*2", &spec1 );
float nonsense =0;
// reader->AddSpectator( "spec2 := var1*3", &spec2 );
float nonsense =0;
Float_t Category_cat1, Category_cat2, Category_cat3;
if (Use["Category"]){
// add artificial spectators for distinguishing categories
// reader->AddSpectator( "Category_cat1 := var3<=0", &Category_cat1 );
float nonsense =0;
// reader->AddSpectator( "Category_cat2 := (var3>0)&&(var4<0)", &Category_cat2 );
float nonsense =0;
// reader->AddSpectator( "Category_cat3 := (var3>0)&&(var4>=0)", &Category_cat3 );
float nonsense =0;
}
//
// book the MVA methods
//
TString dir = "weights/";
TString prefix = "TMVAClassification";
// book method(s)
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString methodName = it->first + " method";
TString weightfile = dir + prefix + "_" + TString(it->first) + ".weights.xml";
reader->BookMVA( methodName, weightfile );
}
}
// example how to use your own method as plugin
if (Use["Plugin"]) {
// the weight file contains a line
// Method : MethodName::InstanceName
// if MethodName is not a known TMVA method, it is assumed to be
// a user implemented method which has to be loaded via the
// plugin mechanism
// for user implemented methods the line in the weight file can be
// Method : PluginName::InstanceName
// where PluginName can be anything
// before usage the plugin has to be defined, which can happen
// either through the following line in .rootrc:
// # plugin handler plugin class library constructor format
// Plugin.TMVA@@MethodBase: PluginName MethodClassName UserPackage "MethodName(DataSet&,TString)"
//
// or by telling the global plugin manager directly
gPluginMgr->AddHandler("TMVA@@MethodBase", "PluginName", "MethodClassName", "UserPackage", "MethodName(DataSet&,TString)");
// the class is then looked for in libUserPackage.so
// now the method can be booked like any other
reader->BookMVA( "User method", dir + prefix + "_User.weights.txt" );
}
// book output histograms
UInt_t nbin = 100;
TH1F *histLk(0), *histLkD(0), *histLkPCA(0), *histLkKDE(0), *histLkMIX(0), *histPD(0), *histPDD(0);
TH1F *histPDPCA(0), *histPDEFoam(0), *histPDEFoamErr(0), *histPDEFoamSig(0), *histKNN(0), *histHm(0);
TH1F *histFi(0), *histFiG(0), *histFiB(0), *histLD(0), *histNn(0),*histNnbfgs(0),*histNnbnn(0), *histNnC(0), *histNnT(0), *histBdt(0), *histBdtG(0), *histBdtD(0);
TH1F *histRf(0), *histSVMG(0), *histSVMP(0), *histSVML(0), *histFDAMT(0), *histFDAGA(0), *histCat(0), *histPBdt(0);
if (Use["Likelihood"]) histLk = new TH1F( "MVA_Likelihood", "MVA_Likelihood", nbin, -1, 1 );
if (Use["LikelihoodD"]) histLkD = new TH1F( "MVA_LikelihoodD", "MVA_LikelihoodD", nbin, -1, 0.9999 );
if (Use["LikelihoodPCA"]) histLkPCA = new TH1F( "MVA_LikelihoodPCA", "MVA_LikelihoodPCA", nbin, -1, 1 );
if (Use["LikelihoodKDE"]) histLkKDE = new TH1F( "MVA_LikelihoodKDE", "MVA_LikelihoodKDE", nbin, -0.00001, 0.99999 );
if (Use["LikelihoodMIX"]) histLkMIX = new TH1F( "MVA_LikelihoodMIX", "MVA_LikelihoodMIX", nbin, 0, 1 );
if (Use["PDERS"]) histPD = new TH1F( "MVA_PDERS", "MVA_PDERS", nbin, 0, 1 );
if (Use["PDERSD"]) histPDD = new TH1F( "MVA_PDERSD", "MVA_PDERSD", nbin, 0, 1 );
if (Use["PDERSPCA"]) histPDPCA = new TH1F( "MVA_PDERSPCA", "MVA_PDERSPCA", nbin, 0, 1 );
if (Use["KNN"]) histKNN = new TH1F( "MVA_KNN", "MVA_KNN", nbin, 0, 1 );
if (Use["HMatrix"]) histHm = new TH1F( "MVA_HMatrix", "MVA_HMatrix", nbin, -0.95, 1.55 );
if (Use["Fisher"]) histFi = new TH1F( "MVA_Fisher", "MVA_Fisher", nbin, -4, 4 );
if (Use["FisherG"]) histFiG = new TH1F( "MVA_FisherG", "MVA_FisherG", nbin, -1, 1 );
if (Use["BoostedFisher"]) histFiB = new TH1F( "MVA_BoostedFisher", "MVA_BoostedFisher", nbin, -2, 2 );
if (Use["LD"]) histLD = new TH1F( "MVA_LD", "MVA_LD", nbin, -2, 2 );
if (Use["MLP"]) histNn = new TH1F( "MVA_MLP", "MVA_MLP", nbin, -1.25, 1.5 );
if (Use["MLPBFGS"]) histNnbfgs = new TH1F( "MVA_MLPBFGS", "MVA_MLPBFGS", nbin, -1.25, 1.5 );
if (Use["MLPBNN"]) histNnbnn = new TH1F( "MVA_MLPBNN", "MVA_MLPBNN", nbin, -1.25, 1.5 );
if (Use["CFMlpANN"]) histNnC = new TH1F( "MVA_CFMlpANN", "MVA_CFMlpANN", nbin, 0, 1 );
if (Use["TMlpANN"]) histNnT = new TH1F( "MVA_TMlpANN", "MVA_TMlpANN", nbin, -1.3, 1.3 );
if (Use["BDT"]) histBdt = new TH1F( "MVA_BDT", "MVA_BDT", nbin, -0.8, 0.8 );
if (Use["BDTD"]) histBdtD = new TH1F( "MVA_BDTD", "MVA_BDTD", nbin, -0.8, 0.8 );
if (Use["BDTG"]) histBdtG = new TH1F( "MVA_BDTG", "MVA_BDTG", nbin, -1.0, 1.0 );
if (Use["RuleFit"]) histRf = new TH1F( "MVA_RuleFit", "MVA_RuleFit", nbin, -2.0, 2.0 );
if (Use["SVM_Gauss"]) histSVMG = new TH1F( "MVA_SVM_Gauss", "MVA_SVM_Gauss", nbin, 0.0, 1.0 );
if (Use["SVM_Poly"]) histSVMP = new TH1F( "MVA_SVM_Poly", "MVA_SVM_Poly", nbin, 0.0, 1.0 );
if (Use["SVM_Lin"]) histSVML = new TH1F( "MVA_SVM_Lin", "MVA_SVM_Lin", nbin, 0.0, 1.0 );
if (Use["FDA_MT"]) histFDAMT = new TH1F( "MVA_FDA_MT", "MVA_FDA_MT", nbin, -2.0, 3.0 );
if (Use["FDA_GA"]) histFDAGA = new TH1F( "MVA_FDA_GA", "MVA_FDA_GA", nbin, -2.0, 3.0 );
if (Use["Category"]) histCat = new TH1F( "MVA_Category", "MVA_Category", nbin, -2., 2. );
if (Use["Plugin"]) histPBdt = new TH1F( "MVA_PBDT", "MVA_BDT", nbin, -0.8, 0.8 );
// PDEFoam also returns per-event error, fill in histogram, and also fill significance
if (Use["PDEFoam"]) {
histPDEFoam = new TH1F( "MVA_PDEFoam", "MVA_PDEFoam", nbin, 0, 1 );
histPDEFoamErr = new TH1F( "MVA_PDEFoamErr", "MVA_PDEFoam error", nbin, 0, 1 );
histPDEFoamSig = new TH1F( "MVA_PDEFoamSig", "MVA_PDEFoam significance", nbin, 0, 10 );
}
// book example histogram for probability (the other methods are done similarly)
TH1F *probHistFi(0), *rarityHistFi(0);
if (Use["Fisher"]) {
probHistFi = new TH1F( "MVA_Fisher_Proba", "MVA_Fisher_Proba", nbin, 0, 1 );
rarityHistFi = new TH1F( "MVA_Fisher_Rarity", "MVA_Fisher_Rarity", nbin, 0, 1 );
}
// Prepare input tree (this must be replaced by your data source)
// in this example, there is a toy tree with signal and one with background events
// we'll later on use only the "signal" events for the test in this example.
//
TFile *input(0);
TString fname = "/tmp/chasco/ORIGINAL//Data_MuEG2011B_1.root";
if (!gSystem->AccessPathName( fname )) {
input = TFile::Open( fname ); // check if file in local directory exists
}
else {
input = TFile::Open( "http://root.cern.ch/files/tmva_class_example.root" ); // if not: download from ROOT server
}
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
//
// prepare the tree
// - here the variable names have to corresponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
TTree* BigTree = (TTree*)input->Get("data");
TFile *tmp = new TFile( "tmp.root","RECREATE" );
TTree* theTree = BigTree->CopyTree("((cat == 1) + (cat == 2))*(ln==0)*(Cosmic==0)*(fabs(Mass_Z - 91.18)<10)*(Pt_Z>30)*(DeltaPhi_metjet>0.5)*(Pt_J1 < 30)*(pfMEToverPt_Z > 0.4)*(pfMEToverPt_Z < 1.8)*((Pt_Jet_btag_CSV_max > 20)*(btag_CSV_max < 0.244) + (1-(Pt_Jet_btag_CSV_max > 20)))*(sqrt(pow(dilepPROJLong + 1.25*recoilPROJLong + 0.0*uncertPROJLong,2)*(dilepPROJLong + 1.25*recoilPROJLong + 0.0*uncertPROJLong > 0) + 1.0*pow(dilepPROJPerp + 1.25*recoilPROJPerp + 0.0*uncertPROJPerp,2)*(dilepPROJPerp + 1.25*recoilPROJPerp + 0.0*uncertPROJPerp > 0)) > 45.0)");
std::cout << "--- Select signal sample" << std::endl;
Float_t userVar1, userVar2;
// theTree->SetBranchAddress( "var1", &userVar1 );
// theTree->SetBranchAddress( "var2", &userVar2 );
// theTree->SetBranchAddress( "var3", &var3 );
// theTree->SetBranchAddress( "var4", &var4 );
theTree->SetBranchAddress( " Z_rapidity_z", &Z_rapidity_z);
theTree->SetBranchAddress( " THRUST_2D", &THRUST_2D);
theTree->SetBranchAddress( " L1_L2_cosangle", &L1_L2_cosangle);
theTree->SetBranchAddress( " TransMass_ZH150_uncl", &TransMass_ZH150_uncl);
theTree->SetBranchAddress( " TransMass_ZH150", &TransMass_ZH150);
theTree->SetBranchAddress( " DeltaPhi_ZH", &DeltaPhi_ZH);
theTree->SetBranchAddress( " DeltaPhi_ZH_uncl", &DeltaPhi_ZH_uncl);
theTree->SetBranchAddress( " CMAngle", &CMAngle);
theTree->SetBranchAddress( " CS_cosangle", &CS_cosangle);
// efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(9); // vector for EvaluateMVA tests
std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
if (ievt%1000 == 0){
std::cout << "--- ... Processing event: " << ievt << std::endl;
}
theTree->GetEntry(ievt);
var1 = userVar1 + userVar2;
var2 = userVar1 - userVar2;
if (ievt <20){
// test the twodifferent Reader::EvaluateMVA functions
// access via registered variables compared to access via vector<float>
// vecVar[0]=var1;
// vecVar[1]=var2;
// vecVar[2]=var3;
// vecVar[3]=var4;
vecVar[0]=Z_rapidity_z;
vecVar[1]=THRUST_2D;
vecVar[2]=L1_L2_cosangle;
vecVar[3]=TransMass_ZH150_uncl;
vecVar[4]=TransMass_ZH150;
vecVar[5]=DeltaPhi_ZH;
vecVar[6]=DeltaPhi_ZH_uncl;
vecVar[7]=CMAngle;
vecVar[8]=CS_cosangle;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString mName = it->first + " method";
Double_t mva1 = reader->EvaluateMVA( mName);
Double_t mva2 = reader->EvaluateMVA( vecVar, mName);
if (mva1 != mva2) {
std::cout << "++++++++++++++ ERROR in "<< mName <<", comparing different EvaluateMVA results val1=" << mva1 << " val2="<<mva2<<std::endl;
}
}
}
// now test that the inputs do matter
TRandom3 rand(0);
// vecVar[0]=rand.Rndm();
// vecVar[1]=rand.Rndm();
// vecVar[2]=rand.Rndm();
// vecVar[3]=rand.Rndm();
vecVar[0]=rand.Rndm();
vecVar[1]=rand.Rndm();
vecVar[2]=rand.Rndm();
vecVar[3]=rand.Rndm();
vecVar[4]=rand.Rndm();
vecVar[5]=rand.Rndm();
vecVar[6]=rand.Rndm();
vecVar[7]=rand.Rndm();
vecVar[8]=rand.Rndm();
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString mName = it->first + " method";
Double_t mva1 = reader->EvaluateMVA( mName);
Double_t mva2 = reader->EvaluateMVA( vecVar, mName);
if (mva1 == mva2) {
std::cout << "++++++++++++++ ERROR in "<< mName <<", obtaining idnetical output for different inputs" <<std::endl;
}
}
}
}
//
// return the MVAs and fill to histograms
//
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
if (Use["Likelihood" ]) histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) );
if (Use["LikelihoodD" ]) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) );
if (Use["LikelihoodPCA"]) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) );
if (Use["LikelihoodKDE"]) histLkKDE ->Fill( reader->EvaluateMVA( "LikelihoodKDE method" ) );
if (Use["LikelihoodMIX"]) histLkMIX ->Fill( reader->EvaluateMVA( "LikelihoodMIX method" ) );
if (Use["PDERS" ]) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) );
if (Use["PDERSD" ]) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) );
if (Use["PDERSPCA" ]) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) );
if (Use["KNN" ]) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) );
if (Use["HMatrix" ]) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) );
if (Use["Fisher" ]) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) );
if (Use["FisherG" ]) histFiG ->Fill( reader->EvaluateMVA( "FisherG method" ) );
if (Use["BoostedFisher"]) histFiB ->Fill( reader->EvaluateMVA( "BoostedFisher method" ) );
if (Use["LD" ]) histLD ->Fill( reader->EvaluateMVA( "LD method" ) );
if (Use["MLP" ]) histNn ->Fill( reader->EvaluateMVA( "MLP method" ) );
if (Use["MLPBFGS" ]) histNnbfgs ->Fill( reader->EvaluateMVA( "MLPBFGS method" ) );
if (Use["MLPBNN" ]) histNnbnn ->Fill( reader->EvaluateMVA( "MLPBNN method" ) );
if (Use["CFMlpANN" ]) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) );
if (Use["TMlpANN" ]) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) );
if (Use["BDT" ]) histBdt ->Fill( reader->EvaluateMVA( "BDT method" ) );
if (Use["BDTD" ]) histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) );
if (Use["BDTG" ]) histBdtG ->Fill( reader->EvaluateMVA( "BDTG method" ) );
if (Use["RuleFit" ]) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) );
if (Use["SVM_Gauss" ]) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) );
if (Use["SVM_Poly" ]) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) );
if (Use["SVM_Lin" ]) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) );
if (Use["FDA_MT" ]) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) );
if (Use["FDA_GA" ]) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) );
if (Use["Category" ]) histCat ->Fill( reader->EvaluateMVA( "Category method" ) );
if (Use["Plugin" ]) histPBdt ->Fill( reader->EvaluateMVA( "P_BDT method" ) );
// retrieve also per-event error
if (Use["PDEFoam"]) {
Double_t val = reader->EvaluateMVA( "PDEFoam method" );
Double_t err = reader->GetMVAError();
histPDEFoam ->Fill( val );
histPDEFoamErr->Fill( err );
histPDEFoamSig->Fill( val/err );
}
// retrieve probability instead of MVA output
if (Use["Fisher"]) {
probHistFi ->Fill( reader->GetProba ( "Fisher method" ) );
rarityHistFi->Fill( reader->GetRarity( "Fisher method" ) );
}
}
// get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
//
// write histograms
//
TFile *target = new TFile( "TMVApp.root","RECREATE" );
if (Use["Likelihood" ]) histLk ->Write();
if (Use["LikelihoodD" ]) histLkD ->Write();
if (Use["LikelihoodPCA"]) histLkPCA ->Write();
if (Use["LikelihoodKDE"]) histLkKDE ->Write();
if (Use["LikelihoodMIX"]) histLkMIX ->Write();
if (Use["PDERS" ]) histPD ->Write();
if (Use["PDERSD" ]) histPDD ->Write();
if (Use["PDERSPCA" ]) histPDPCA ->Write();
if (Use["KNN" ]) histKNN ->Write();
if (Use["HMatrix" ]) histHm ->Write();
if (Use["Fisher" ]) histFi ->Write();
if (Use["FisherG" ]) histFiG ->Write();
if (Use["BoostedFisher"]) histFiB ->Write();
if (Use["LD" ]) histLD ->Write();
if (Use["MLP" ]) histNn ->Write();
if (Use["MLPBFGS" ]) histNnbfgs ->Write();
if (Use["MLPBNN" ]) histNnbnn ->Write();
if (Use["CFMlpANN" ]) histNnC ->Write();
if (Use["TMlpANN" ]) histNnT ->Write();
if (Use["BDT" ]) histBdt ->Write();
if (Use["BDTD" ]) histBdtD ->Write();
if (Use["BDTG" ]) histBdtG ->Write();
if (Use["RuleFit" ]) histRf ->Write();
if (Use["SVM_Gauss" ]) histSVMG ->Write();
if (Use["SVM_Poly" ]) histSVMP ->Write();
if (Use["SVM_Lin" ]) histSVML ->Write();
if (Use["FDA_MT" ]) histFDAMT ->Write();
if (Use["FDA_GA" ]) histFDAGA ->Write();
if (Use["Category" ]) histCat ->Write();
if (Use["Plugin" ]) histPBdt ->Write();
// write also error and significance histos
if (Use["PDEFoam"]) { histPDEFoam->Write(); histPDEFoamErr->Write(); histPDEFoamSig->Write(); }
// write also probability hists
if (Use["Fisher"]) { if (probHistFi != 0) probHistFi->Write(); if (rarityHistFi != 0) rarityHistFi->Write(); }
target->Close();
std::cout << "--- Created root file: \"TMVApp.root\" containing the MVA output histograms" << std::endl;
delete reader;
std::cout << "==> TMVAClassificationApplication is done!" << endl << std::endl;
}
void TMVAClassificationApplication( TString myMethodList = "" )
{
#ifdef __CINT__
gROOT->ProcessLine( ".O0" ); // turn off optimization in CINT
#endif
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
Use["MLP"] = 0; // Recommended ANN
Use["BDT"] = 1; // uses Adaptive Boost
std::cout << std::endl;
std::cout << "==> Start TMVAClassificationApplication" << std::endl;
// --- Create the Reader object
TMVA::Reader *reader = new TMVA::Reader( "!Color:!Silent" );
// Create a set of variables and declare them to the reader
// - the variable names MUST corresponds in name and type to those given in the weight file(s) used
Float_t var1, var2;
Float_t var3, var4;
reader->AddVariable( "myvar1 := var1+var2", &var1 );
reader->AddVariable( "myvar2 := var1-var2", &var2 );
reader->AddVariable( "var3", &var3 );
reader->AddVariable( "var4", &var4 );
// Spectator variables declared in the training have to be added to the reader, too
Float_t spec1,spec2;
reader->AddSpectator( "spec1 := var1*2", &spec1 );
reader->AddSpectator( "spec2 := var1*3", &spec2 );
Float_t Category_cat1, Category_cat2, Category_cat3;
if (Use["Category"]){
// Add artificial spectators for distinguishing categories
reader->AddSpectator( "Category_cat1 := var3<=0", &Category_cat1 );
reader->AddSpectator( "Category_cat2 := (var3>0)&&(var4<0)", &Category_cat2 );
reader->AddSpectator( "Category_cat3 := (var3>0)&&(var4>=0)", &Category_cat3 );
}
// --- Book the MVA methods
TString dir = "weights/";
TString prefix = "TMVAClassification";
// Book method(s)
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString methodName = TString(it->first) + TString(" method");
TString weightfile = dir + prefix + TString("_") + TString(it->first) + TString(".weights.xml");
reader->BookMVA( methodName, weightfile );
}
}
// Book output histograms
UInt_t nbin = 100;
TH1F *histLk(0), *histLkD(0), *histLkPCA(0), *histLkKDE(0), *histLkMIX(0), *histPD(0), *histPDD(0);
TH1F *histPDPCA(0), *histPDEFoam(0), *histPDEFoamErr(0), *histPDEFoamSig(0), *histKNN(0), *histHm(0);
TH1F *histFi(0), *histFiG(0), *histFiB(0), *histLD(0), *histNn(0),*histNnbfgs(0),*histNnbnn(0);
TH1F *histNnC(0), *histNnT(0), *histBdt(0), *histBdtG(0), *histBdtD(0), *histRf(0), *histSVMG(0);
TH1F *histSVMP(0), *histSVML(0), *histFDAMT(0), *histFDAGA(0), *histCat(0), *histPBdt(0);
if (Use["Likelihood"]) histLk = new TH1F( "MVA_Likelihood", "MVA_Likelihood", nbin, -1, 1 );
if (Use["LikelihoodD"]) histLkD = new TH1F( "MVA_LikelihoodD", "MVA_LikelihoodD", nbin, -1, 0.9999 );
if (Use["LikelihoodPCA"]) histLkPCA = new TH1F( "MVA_LikelihoodPCA", "MVA_LikelihoodPCA", nbin, -1, 1 );
if (Use["LikelihoodKDE"]) histLkKDE = new TH1F( "MVA_LikelihoodKDE", "MVA_LikelihoodKDE", nbin, -0.00001, 0.99999 );
if (Use["LikelihoodMIX"]) histLkMIX = new TH1F( "MVA_LikelihoodMIX", "MVA_LikelihoodMIX", nbin, 0, 1 );
if (Use["PDERS"]) histPD = new TH1F( "MVA_PDERS", "MVA_PDERS", nbin, 0, 1 );
if (Use["PDERSD"]) histPDD = new TH1F( "MVA_PDERSD", "MVA_PDERSD", nbin, 0, 1 );
if (Use["PDERSPCA"]) histPDPCA = new TH1F( "MVA_PDERSPCA", "MVA_PDERSPCA", nbin, 0, 1 );
if (Use["KNN"]) histKNN = new TH1F( "MVA_KNN", "MVA_KNN", nbin, 0, 1 );
if (Use["HMatrix"]) histHm = new TH1F( "MVA_HMatrix", "MVA_HMatrix", nbin, -0.95, 1.55 );
if (Use["Fisher"]) histFi = new TH1F( "MVA_Fisher", "MVA_Fisher", nbin, -4, 4 );
if (Use["FisherG"]) histFiG = new TH1F( "MVA_FisherG", "MVA_FisherG", nbin, -1, 1 );
if (Use["BoostedFisher"]) histFiB = new TH1F( "MVA_BoostedFisher", "MVA_BoostedFisher", nbin, -2, 2 );
if (Use["LD"]) histLD = new TH1F( "MVA_LD", "MVA_LD", nbin, -2, 2 );
if (Use["MLP"]) histNn = new TH1F( "MVA_MLP", "MVA_MLP", nbin, -1.25, 1.5 );
if (Use["MLPBFGS"]) histNnbfgs = new TH1F( "MVA_MLPBFGS", "MVA_MLPBFGS", nbin, -1.25, 1.5 );
if (Use["MLPBNN"]) histNnbnn = new TH1F( "MVA_MLPBNN", "MVA_MLPBNN", nbin, -1.25, 1.5 );
if (Use["CFMlpANN"]) histNnC = new TH1F( "MVA_CFMlpANN", "MVA_CFMlpANN", nbin, 0, 1 );
if (Use["TMlpANN"]) histNnT = new TH1F( "MVA_TMlpANN", "MVA_TMlpANN", nbin, -1.3, 1.3 );
if (Use["BDT"]) histBdt = new TH1F( "MVA_BDT", "MVA_BDT", nbin, -0.8, 0.8 );
if (Use["BDTD"]) histBdtD = new TH1F( "MVA_BDTD", "MVA_BDTD", nbin, -0.8, 0.8 );
if (Use["BDTG"]) histBdtG = new TH1F( "MVA_BDTG", "MVA_BDTG", nbin, -1.0, 1.0 );
if (Use["RuleFit"]) histRf = new TH1F( "MVA_RuleFit", "MVA_RuleFit", nbin, -2.0, 2.0 );
if (Use["SVM_Gauss"]) histSVMG = new TH1F( "MVA_SVM_Gauss", "MVA_SVM_Gauss", nbin, 0.0, 1.0 );
if (Use["SVM_Poly"]) histSVMP = new TH1F( "MVA_SVM_Poly", "MVA_SVM_Poly", nbin, 0.0, 1.0 );
if (Use["SVM_Lin"]) histSVML = new TH1F( "MVA_SVM_Lin", "MVA_SVM_Lin", nbin, 0.0, 1.0 );
if (Use["FDA_MT"]) histFDAMT = new TH1F( "MVA_FDA_MT", "MVA_FDA_MT", nbin, -2.0, 3.0 );
if (Use["FDA_GA"]) histFDAGA = new TH1F( "MVA_FDA_GA", "MVA_FDA_GA", nbin, -2.0, 3.0 );
if (Use["Category"]) histCat = new TH1F( "MVA_Category", "MVA_Category", nbin, -2., 2. );
if (Use["Plugin"]) histPBdt = new TH1F( "MVA_PBDT", "MVA_BDT", nbin, -0.8, 0.8 );
// PDEFoam also returns per-event error, fill in histogram, and also fill significance
if (Use["PDEFoam"]) {
histPDEFoam = new TH1F( "MVA_PDEFoam", "MVA_PDEFoam", nbin, 0, 1 );
histPDEFoamErr = new TH1F( "MVA_PDEFoamErr", "MVA_PDEFoam error", nbin, 0, 1 );
histPDEFoamSig = new TH1F( "MVA_PDEFoamSig", "MVA_PDEFoam significance", nbin, 0, 10 );
}
// Book example histogram for probability (the other methods are done similarly)
TH1F *probHistFi(0), *rarityHistFi(0);
if (Use["Fisher"]) {
probHistFi = new TH1F( "MVA_Fisher_Proba", "MVA_Fisher_Proba", nbin, 0, 1 );
rarityHistFi = new TH1F( "MVA_Fisher_Rarity", "MVA_Fisher_Rarity", nbin, 0, 1 );
}
// Prepare input tree (this must be replaced by your data source)
// in this example, there is a toy tree with signal and one with background events
// we'll later on use only the "signal" events for the test in this example.
//
TFile *input(0);
TString fname = "./tmva_example.root";
if (!gSystem->AccessPathName( fname ))
input = TFile::Open( fname ); // check if file in local directory exists
else
input = TFile::Open( "http://root.cern.ch/files/tmva_class_example.root" ); // if not: download from ROOT server
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
// --- Event loop
// Prepare the event tree
// - here the variable names have to corresponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
std::cout << "--- Select signal sample" << std::endl;
TTree* theTree = (TTree*)input->Get("TreeS");
Float_t userVar1, userVar2;
theTree->SetBranchAddress( "var1", &userVar1 );
theTree->SetBranchAddress( "var2", &userVar2 );
theTree->SetBranchAddress( "var3", &var3 );
theTree->SetBranchAddress( "var4", &var4 );
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
if (ievt%1000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
theTree->GetEntry(ievt);
var1 = userVar1 + userVar2;
var2 = userVar1 - userVar2;
// --- Return the MVA outputs and fill into histograms
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
if (Use["Likelihood" ]) histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) );
if (Use["LikelihoodD" ]) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) );
if (Use["LikelihoodPCA"]) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) );
if (Use["LikelihoodKDE"]) histLkKDE ->Fill( reader->EvaluateMVA( "LikelihoodKDE method" ) );
if (Use["LikelihoodMIX"]) histLkMIX ->Fill( reader->EvaluateMVA( "LikelihoodMIX method" ) );
if (Use["PDERS" ]) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) );
if (Use["PDERSD" ]) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) );
if (Use["PDERSPCA" ]) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) );
if (Use["KNN" ]) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) );
if (Use["HMatrix" ]) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) );
if (Use["Fisher" ]) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) );
if (Use["FisherG" ]) histFiG ->Fill( reader->EvaluateMVA( "FisherG method" ) );
if (Use["BoostedFisher"]) histFiB ->Fill( reader->EvaluateMVA( "BoostedFisher method" ) );
if (Use["LD" ]) histLD ->Fill( reader->EvaluateMVA( "LD method" ) );
if (Use["MLP" ]) histNn ->Fill( reader->EvaluateMVA( "MLP method" ) );
if (Use["MLPBFGS" ]) histNnbfgs ->Fill( reader->EvaluateMVA( "MLPBFGS method" ) );
if (Use["MLPBNN" ]) histNnbnn ->Fill( reader->EvaluateMVA( "MLPBNN method" ) );
if (Use["CFMlpANN" ]) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) );
if (Use["TMlpANN" ]) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) );
if (Use["BDT" ]) histBdt ->Fill( reader->EvaluateMVA( "BDT method" ) );
if (Use["BDTD" ]) histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) );
if (Use["BDTG" ]) histBdtG ->Fill( reader->EvaluateMVA( "BDTG method" ) );
if (Use["RuleFit" ]) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) );
if (Use["SVM_Gauss" ]) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) );
if (Use["SVM_Poly" ]) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) );
if (Use["SVM_Lin" ]) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) );
if (Use["FDA_MT" ]) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) );
if (Use["FDA_GA" ]) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) );
if (Use["Category" ]) histCat ->Fill( reader->EvaluateMVA( "Category method" ) );
if (Use["Plugin" ]) histPBdt ->Fill( reader->EvaluateMVA( "P_BDT method" ) );
// Retrieve also per-event error
if (Use["PDEFoam"]) {
Double_t val = reader->EvaluateMVA( "PDEFoam method" );
Double_t err = reader->GetMVAError();
histPDEFoam ->Fill( val );
histPDEFoamErr->Fill( err );
if (err>1.e-50) histPDEFoamSig->Fill( val/err );
}
// Retrieve probability instead of MVA output
if (Use["Fisher"]) {
probHistFi ->Fill( reader->GetProba ( "Fisher method" ) );
rarityHistFi->Fill( reader->GetRarity( "Fisher method" ) );
}
}
// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// Get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
// --- Write histograms
TFile *target = new TFile( "TMVApp.root","RECREATE" );
if (Use["Likelihood" ]) histLk ->Write();
if (Use["LikelihoodD" ]) histLkD ->Write();
if (Use["LikelihoodPCA"]) histLkPCA ->Write();
if (Use["LikelihoodKDE"]) histLkKDE ->Write();
if (Use["LikelihoodMIX"]) histLkMIX ->Write();
if (Use["PDERS" ]) histPD ->Write();
if (Use["PDERSD" ]) histPDD ->Write();
if (Use["PDERSPCA" ]) histPDPCA ->Write();
if (Use["KNN" ]) histKNN ->Write();
if (Use["HMatrix" ]) histHm ->Write();
if (Use["Fisher" ]) histFi ->Write();
if (Use["FisherG" ]) histFiG ->Write();
if (Use["BoostedFisher"]) histFiB ->Write();
if (Use["LD" ]) histLD ->Write();
if (Use["MLP" ]) histNn ->Write();
if (Use["MLPBFGS" ]) histNnbfgs ->Write();
if (Use["MLPBNN" ]) histNnbnn ->Write();
if (Use["CFMlpANN" ]) histNnC ->Write();
if (Use["TMlpANN" ]) histNnT ->Write();
if (Use["BDT" ]) histBdt ->Write();
if (Use["BDTD" ]) histBdtD ->Write();
if (Use["BDTG" ]) histBdtG ->Write();
if (Use["RuleFit" ]) histRf ->Write();
if (Use["SVM_Gauss" ]) histSVMG ->Write();
if (Use["SVM_Poly" ]) histSVMP ->Write();
if (Use["SVM_Lin" ]) histSVML ->Write();
if (Use["FDA_MT" ]) histFDAMT ->Write();
if (Use["FDA_GA" ]) histFDAGA ->Write();
if (Use["Category" ]) histCat ->Write();
if (Use["Plugin" ]) histPBdt ->Write();
// Write also error and significance histos
if (Use["PDEFoam"]) { histPDEFoam->Write(); histPDEFoamErr->Write(); histPDEFoamSig->Write(); }
// Write also probability hists
if (Use["Fisher"]) { if (probHistFi != 0) probHistFi->Write(); if (rarityHistFi != 0) rarityHistFi->Write(); }
target->Close();
std::cout << "--- Created root file: \"TMVApp.root\" containing the MVA output histograms" << std::endl;
delete reader;
std::cout << "==> TMVAClassificationApplication is done!" << endl << std::endl;
}
void Classify_HWW( TString myMethodList = "" )
{
#ifdef __CINT__
gROOT->ProcessLine( ".O0" ); // turn off optimization in CINT
#endif
//--------------------------------------------------------------------
// path to weights dir (this is where MVA training info is stored)
// output root file will be stored at [path]/output
//--------------------------------------------------------------------
TString path = "Trainings/v5/H160_WW_10vars_dphi10/";
//TString path = "./";
//-----------------------------------
// select samples to run over
//-----------------------------------
char* babyPath = "/tas/cerati/HtoWWmvaBabies/latest";
int mH = 160; // choose Higgs mass
vector<char*> samples;
samples.push_back("WWTo2L2Nu");
samples.push_back("GluGluToWWTo4L");
samples.push_back("WZ");
samples.push_back("ZZ");
samples.push_back("TTJets");
samples.push_back("tW");
samples.push_back("WJetsToLNu");
samples.push_back("DY");
//samples.push_back("WJetsFO3");
if ( mH == 130 ) samples.push_back("Higgs130");
else if( mH == 160 ) samples.push_back("Higgs160");
else if( mH == 200 ) samples.push_back("Higgs200");
else{
cout << "Error, unrecognized Higgs mass " << mH << " GeV, quitting" << endl;
exit(0);
}
//--------------------------------------------------------------------------------
// IMPORTANT: set the following variables to the same set used for MVA training!!!
//--------------------------------------------------------------------------------
std::map<std::string,int> mvaVar;
mvaVar[ "lephard_pt" ] = 1;
mvaVar[ "lepsoft_pt" ] = 1;
mvaVar[ "dil_dphi" ] = 1;
mvaVar[ "dil_mass" ] = 1;
mvaVar[ "event_type" ] = 0;
mvaVar[ "met_projpt" ] = 1;
mvaVar[ "met_pt" ] = 0;
mvaVar[ "mt_lephardmet" ] = 1;
mvaVar[ "mt_lepsoftmet" ] = 1;
mvaVar[ "mthiggs" ] = 1;
mvaVar[ "dphi_lephardmet" ] = 1;
mvaVar[ "dphi_lepsoftmet" ] = 1;
mvaVar[ "lepsoft_fbrem" ] = 0;
mvaVar[ "lepsoft_eOverPIn" ] = 0;
mvaVar[ "lepsoft_qdphi" ] = 0;
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
// --- Cut optimisation
Use["Cuts"] = 1;
Use["CutsD"] = 1;
Use["CutsPCA"] = 0;
Use["CutsGA"] = 0;
Use["CutsSA"] = 0;
//
// --- 1-dimensional likelihood ("naive Bayes estimator")
Use["Likelihood"] = 1;
Use["LikelihoodD"] = 0; // the "D" extension indicates decorrelated input variables (see option strings)
Use["LikelihoodPCA"] = 1; // the "PCA" extension indicates PCA-transformed input variables (see option strings)
Use["LikelihoodKDE"] = 0;
Use["LikelihoodMIX"] = 0;
//
// --- Mutidimensional likelihood and Nearest-Neighbour methods
Use["PDERS"] = 1;
Use["PDERSD"] = 0;
Use["PDERSPCA"] = 0;
Use["PDEFoam"] = 1;
Use["PDEFoamBoost"] = 0; // uses generalised MVA method boosting
Use["KNN"] = 1; // k-nearest neighbour method
//
// --- Linear Discriminant Analysis
Use["LD"] = 1; // Linear Discriminant identical to Fisher
Use["Fisher"] = 0;
Use["FisherG"] = 0;
Use["BoostedFisher"] = 0; // uses generalised MVA method boosting
Use["HMatrix"] = 0;
//
// --- Function Discriminant analysis
Use["FDA_GA"] = 1; // minimisation of user-defined function using Genetics Algorithm
Use["FDA_SA"] = 0;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 0;
Use["FDA_GAMT"] = 0;
Use["FDA_MCMT"] = 0;
//
// --- Neural Networks (all are feed-forward Multilayer Perceptrons)
Use["MLP"] = 0; // Recommended ANN
Use["MLPBFGS"] = 0; // Recommended ANN with optional training method
Use["MLPBNN"] = 1; // Recommended ANN with BFGS training method and bayesian regulator
Use["CFMlpANN"] = 0; // Depreciated ANN from ALEPH
Use["TMlpANN"] = 0; // ROOT's own ANN
//
// --- Support Vector Machine
Use["SVM"] = 1;
//
// --- Boosted Decision Trees
Use["BDT"] = 1; // uses Adaptive Boost
Use["BDTG"] = 0; // uses Gradient Boost
Use["BDTB"] = 0; // uses Bagging
Use["BDTD"] = 0; // decorrelation + Adaptive Boost
//
// --- Friedman's RuleFit method, ie, an optimised series of cuts ("rules")
Use["RuleFit"] = 1;
// ---------------------------------------------------------------
Use["Plugin"] = 0;
Use["Category"] = 0;
Use["SVM_Gauss"] = 0;
Use["SVM_Poly"] = 0;
Use["SVM_Lin"] = 0;
std::cout << std::endl;
std::cout << "==> Start TMVAClassificationApplication" << std::endl;
// Select methods (don't look at this code - not of interest)
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod
<< "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
std::cout << it->first << " ";
}
std::cout << std::endl;
return;
}
Use[regMethod] = 1;
}
}
// --------------------------------------------------------------------------------------------------
const unsigned int nsamples = samples.size();
for( unsigned int i = 0 ; i < nsamples ; ++i ){
// --- Create the Reader object
TMVA::Reader *reader = new TMVA::Reader( "!Color:!Silent" );
// Create a set of variables and declare them to the reader
// - the variable names MUST corresponds in name and type to those given in the weight file(s) used
// Float_t var1, var2;
// Float_t var3, var4;
// reader->AddVariable( "myvar1 := var1+var2", &var1 );
// reader->AddVariable( "myvar2 := var1-var2", &var2 );
// reader->AddVariable( "var3", &var3 );
// reader->AddVariable( "var4", &var4 );
Float_t lephard_pt;
Float_t lepsoft_pt;
Float_t dil_dphi;
Float_t dil_mass;
Float_t event_type;
Float_t met_projpt;
Float_t met_pt;
Float_t mt_lephardmet;
Float_t mt_lepsoftmet;
Float_t mthiggs;
Float_t dphi_lephardmet;
Float_t dphi_lepsoftmet;
Float_t lepsoft_fbrem;
Float_t lepsoft_eOverPIn;
Float_t lepsoft_qdphi;
if( mvaVar["lephard_pt"]) reader->AddVariable( "lephard_pt" , &lephard_pt );
if( mvaVar["lepsoft_pt"]) reader->AddVariable( "lepsoft_pt" , &lepsoft_pt );
if( mvaVar["dil_dphi"]) reader->AddVariable( "dil_dphi" , &dil_dphi );
if( mvaVar["dil_mass"]) reader->AddVariable( "dil_mass" , &dil_mass );
if( mvaVar["event_type"]) reader->AddVariable( "event_type" , &event_type );
if( mvaVar["met_projpt"]) reader->AddVariable( "met_projpt" , &met_pt );
if( mvaVar["met_pt"]) reader->AddVariable( "met_pt" , &met_pt );
if( mvaVar["mt_lephardmet"]) reader->AddVariable( "mt_lephardmet" , &mt_lephardmet );
if( mvaVar["mt_lepsoftmet"]) reader->AddVariable( "mt_lepsoftmet" , &mt_lepsoftmet );
if( mvaVar["mthiggs"]) reader->AddVariable( "mthiggs" , &mthiggs );
if( mvaVar["dphi_lephardmet"]) reader->AddVariable( "dphi_lephardmet" , &dphi_lephardmet );
if( mvaVar["dphi_lepsoftmet"]) reader->AddVariable( "dphi_lepsoftmet" , &dphi_lepsoftmet );
if( mvaVar["lepsoft_fbrem"]) reader->AddVariable( "lepsoft_fbrem" , &lepsoft_fbrem );
if( mvaVar["lepsoft_eOverPIn"]) reader->AddVariable( "lepsoft_eOverPIn" , &lepsoft_eOverPIn );
if( mvaVar["lepsoft_qdphi"]) reader->AddVariable( "lepsoft_q * lepsoft_dPhiIn" , &lepsoft_qdphi );
// Spectator variables declared in the training have to be added to the reader, too
// Float_t spec1,spec2;
// reader->AddSpectator( "spec1 := var1*2", &spec1 );
// reader->AddSpectator( "spec2 := var1*3", &spec2 );
Float_t Category_cat1, Category_cat2, Category_cat3;
if (Use["Category"]){
// Add artificial spectators for distinguishing categories
// reader->AddSpectator( "Category_cat1 := var3<=0", &Category_cat1 );
// reader->AddSpectator( "Category_cat2 := (var3>0)&&(var4<0)", &Category_cat2 );
// reader->AddSpectator( "Category_cat3 := (var3>0)&&(var4>=0)", &Category_cat3 );
}
// --- Book the MVA methods
//--------------------------------------------------------------------------------------
// tell Classify_HWW where to find the weights dir, which contains the trained MVA's.
// In this example, the weights dir is located at [path]/[dir]
// and the output root file is written to [path]/[output]
//--------------------------------------------------------------------------------------
TString dir = path + "weights/";
TString outdir = path + "output/";
TString prefix = "TMVAClassification";
// Book method(s)
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString methodName = TString(it->first) + TString(" method");
TString weightfile = dir + prefix + TString("_") + TString(it->first) + TString(".weights.xml");
reader->BookMVA( methodName, weightfile );
}
}
// Book output histograms
UInt_t nbin = 1000;
TH1F *histLk(0), *histLkD(0), *histLkPCA(0), *histLkKDE(0), *histLkMIX(0), *histPD(0), *histPDD(0);
TH1F *histPDPCA(0), *histPDEFoam(0), *histPDEFoamErr(0), *histPDEFoamSig(0), *histKNN(0), *histHm(0);
TH1F *histFi(0), *histFiG(0), *histFiB(0), *histLD(0), *histNn(0),*histNnbfgs(0),*histNnbnn(0);
TH1F *histNnC(0), *histNnT(0), *histBdt(0), *histBdtG(0), *histBdtD(0), *histRf(0), *histSVMG(0);
TH1F *histSVMP(0), *histSVML(0), *histFDAMT(0), *histFDAGA(0), *histCat(0), *histPBdt(0);
if (Use["Likelihood"]) histLk = new TH1F( "MVA_Likelihood", "MVA_Likelihood", nbin, -1, 1 );
if (Use["LikelihoodD"]) histLkD = new TH1F( "MVA_LikelihoodD", "MVA_LikelihoodD", nbin, -1, 0.9999 );
if (Use["LikelihoodPCA"]) histLkPCA = new TH1F( "MVA_LikelihoodPCA", "MVA_LikelihoodPCA", nbin, -1, 1 );
if (Use["LikelihoodKDE"]) histLkKDE = new TH1F( "MVA_LikelihoodKDE", "MVA_LikelihoodKDE", nbin, -0.00001, 0.99999 );
if (Use["LikelihoodMIX"]) histLkMIX = new TH1F( "MVA_LikelihoodMIX", "MVA_LikelihoodMIX", nbin, 0, 1 );
if (Use["PDERS"]) histPD = new TH1F( "MVA_PDERS", "MVA_PDERS", nbin, 0, 1 );
if (Use["PDERSD"]) histPDD = new TH1F( "MVA_PDERSD", "MVA_PDERSD", nbin, 0, 1 );
if (Use["PDERSPCA"]) histPDPCA = new TH1F( "MVA_PDERSPCA", "MVA_PDERSPCA", nbin, 0, 1 );
if (Use["KNN"]) histKNN = new TH1F( "MVA_KNN", "MVA_KNN", nbin, 0, 1 );
if (Use["HMatrix"]) histHm = new TH1F( "MVA_HMatrix", "MVA_HMatrix", nbin, -0.95, 1.55 );
if (Use["Fisher"]) histFi = new TH1F( "MVA_Fisher", "MVA_Fisher", nbin, -4, 4 );
if (Use["FisherG"]) histFiG = new TH1F( "MVA_FisherG", "MVA_FisherG", nbin, -1, 1 );
if (Use["BoostedFisher"]) histFiB = new TH1F( "MVA_BoostedFisher", "MVA_BoostedFisher", nbin, -2, 2 );
if (Use["LD"]) histLD = new TH1F( "MVA_LD", "MVA_LD", nbin, -2, 2 );
if (Use["MLP"]) histNn = new TH1F( "MVA_MLP", "MVA_MLP", nbin, -1.25, 1.5 );
if (Use["MLPBFGS"]) histNnbfgs = new TH1F( "MVA_MLPBFGS", "MVA_MLPBFGS", nbin, -1.25, 1.5 );
if (Use["MLPBNN"]) histNnbnn = new TH1F( "MVA_MLPBNN", "MVA_MLPBNN", nbin, -1.25, 1.5 );
if (Use["CFMlpANN"]) histNnC = new TH1F( "MVA_CFMlpANN", "MVA_CFMlpANN", nbin, 0, 1 );
if (Use["TMlpANN"]) histNnT = new TH1F( "MVA_TMlpANN", "MVA_TMlpANN", nbin, -1.3, 1.3 );
if (Use["BDT"]) histBdt = new TH1F( "MVA_BDT", "MVA_BDT", nbin, -1. , 1. );
if (Use["BDTD"]) histBdtD = new TH1F( "MVA_BDTD", "MVA_BDTD", nbin, -0.8, 0.8 );
if (Use["BDTG"]) histBdtG = new TH1F( "MVA_BDTG", "MVA_BDTG", nbin, -1.0, 1.0 );
if (Use["RuleFit"]) histRf = new TH1F( "MVA_RuleFit", "MVA_RuleFit", nbin, -2.0, 2.0 );
if (Use["SVM_Gauss"]) histSVMG = new TH1F( "MVA_SVM_Gauss", "MVA_SVM_Gauss", nbin, 0.0, 1.0 );
if (Use["SVM_Poly"]) histSVMP = new TH1F( "MVA_SVM_Poly", "MVA_SVM_Poly", nbin, 0.0, 1.0 );
if (Use["SVM_Lin"]) histSVML = new TH1F( "MVA_SVM_Lin", "MVA_SVM_Lin", nbin, 0.0, 1.0 );
if (Use["FDA_MT"]) histFDAMT = new TH1F( "MVA_FDA_MT", "MVA_FDA_MT", nbin, -2.0, 3.0 );
if (Use["FDA_GA"]) histFDAGA = new TH1F( "MVA_FDA_GA", "MVA_FDA_GA", nbin, -2.0, 3.0 );
if (Use["Category"]) histCat = new TH1F( "MVA_Category", "MVA_Category", nbin, -2., 2. );
if (Use["Plugin"]) histPBdt = new TH1F( "MVA_PBDT", "MVA_BDT", nbin, -0.8, 0.8 );
if (Use["Likelihood"]) histLk ->Sumw2();
if (Use["LikelihoodD"]) histLkD ->Sumw2();
if (Use["LikelihoodPCA"]) histLkPCA ->Sumw2();
if (Use["LikelihoodKDE"]) histLkKDE ->Sumw2();
if (Use["LikelihoodMIX"]) histLkMIX ->Sumw2();
if (Use["PDERS"]) histPD ->Sumw2();
if (Use["PDERSD"]) histPDD ->Sumw2();
if (Use["PDERSPCA"]) histPDPCA ->Sumw2();
if (Use["KNN"]) histKNN ->Sumw2();
if (Use["HMatrix"]) histHm ->Sumw2();
if (Use["Fisher"]) histFi ->Sumw2();
if (Use["FisherG"]) histFiG ->Sumw2();
if (Use["BoostedFisher"]) histFiB ->Sumw2();
if (Use["LD"]) histLD ->Sumw2();
if (Use["MLP"]) histNn ->Sumw2();
if (Use["MLPBFGS"]) histNnbfgs ->Sumw2();
if (Use["MLPBNN"]) histNnbnn ->Sumw2();
if (Use["CFMlpANN"]) histNnC ->Sumw2();
if (Use["TMlpANN"]) histNnT ->Sumw2();
if (Use["BDT"]) histBdt ->Sumw2();
if (Use["BDTD"]) histBdtD ->Sumw2();
if (Use["BDTG"]) histBdtG ->Sumw2();
if (Use["RuleFit"]) histRf ->Sumw2();
if (Use["SVM_Gauss"]) histSVMG ->Sumw2();
if (Use["SVM_Poly"]) histSVMP ->Sumw2();
if (Use["SVM_Lin"]) histSVML ->Sumw2();
if (Use["FDA_MT"]) histFDAMT ->Sumw2();
if (Use["FDA_GA"]) histFDAGA ->Sumw2();
if (Use["Category"]) histCat ->Sumw2();
if (Use["Plugin"]) histPBdt ->Sumw2();
// PDEFoam also returns per-event error, fill in histogram, and also fill significance
if (Use["PDEFoam"]) {
histPDEFoam = new TH1F( "MVA_PDEFoam", "MVA_PDEFoam", nbin, 0, 1 );
histPDEFoamErr = new TH1F( "MVA_PDEFoamErr", "MVA_PDEFoam error", nbin, 0, 1 );
histPDEFoamSig = new TH1F( "MVA_PDEFoamSig", "MVA_PDEFoam significance", nbin, 0, 10 );
}
// Book example histogram for probability (the other methods are done similarly)
TH1F *probHistFi(0), *rarityHistFi(0);
if (Use["Fisher"]) {
probHistFi = new TH1F( "MVA_Fisher_Proba", "MVA_Fisher_Proba", nbin, 0, 1 );
rarityHistFi = new TH1F( "MVA_Fisher_Rarity", "MVA_Fisher_Rarity", nbin, 0, 1 );
}
// Prepare input tree (this must be replaced by your data source)
// in this example, there is a toy tree with signal and one with background events
// we'll later on use only the "signal" events for the test in this example.
//
TChain *ch = new TChain("Events");
if( strcmp( samples.at(i) , "DY" ) == 0 ){
ch -> Add( Form("%s/DYToMuMuM20_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/DYToMuMuM10To20_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/DYToEEM20_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/DYToEEM10To20_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/DYToTauTauM20_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/DYToTauTauM10To20_PU_testFinal_baby.root",babyPath) );
}
if( strcmp( samples.at(i) , "WJetsFO3" ) == 0 ){
ch -> Add( Form("%s/WJetsToLNu_FOv3_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/WToLNu_FOv3_testFinal_baby.root",babyPath) );
}
else if( strcmp( samples.at(i) , "Higgs130" ) == 0 ){
ch -> Add( Form("%s/HToWWTo2L2NuM130_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/HToWWToLNuTauNuM130_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/HToWWTo2Tau2NuM130_PU_testFinal_baby.root",babyPath) );
}
else if( strcmp( samples.at(i) , "Higgs160" ) == 0 ){
ch -> Add( Form("%s/HToWWTo2L2NuM160_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/HToWWToLNuTauNuM160_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/HToWWTo2Tau2NuM160_PU_testFinal_baby.root",babyPath) );
}
else if( strcmp( samples.at(i) , "Higgs200" ) == 0 ){
ch -> Add( Form("%s/HToWWTo2L2NuM200_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/HToWWToLNuTauNuM200_PU_testFinal_baby.root",babyPath) );
ch -> Add( Form("%s/HToWWTo2Tau2NuM200_PU_testFinal_baby.root",babyPath) );
}
else{
ch -> Add( Form("%s/%s_PU_testFinal_baby.root",babyPath,samples.at(i)) );
}
// --- Event loop
// Prepare the event tree
// - here the variable names have to corresponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
TTree *theTree = (TTree*) ch;
std::cout << "--- Using input files: -------------------" << std::endl;
TObjArray *listOfFiles = ch->GetListOfFiles();
TIter fileIter(listOfFiles);
TChainElement* currentFile = 0;
while((currentFile = (TChainElement*)fileIter.Next())) {
std::cout << currentFile->GetTitle() << std::endl;
}
Float_t lephard_pt_;
Float_t lepsoft_pt_;
Float_t lepsoft_fr_;
Float_t dil_dphi_;
Float_t dil_mass_;
Float_t event_type_;
Float_t met_projpt_;
Int_t jets_num_;
Int_t extralep_num_;
Int_t lowptbtags_num_;
Int_t softmu_num_;
Float_t event_scale1fb_;
Float_t met_pt_;
Int_t lepsoft_passTighterId_;
Float_t mt_lephardmet_;
Float_t mt_lepsoftmet_;
Float_t mthiggs_;
Float_t dphi_lephardmet_;
Float_t dphi_lepsoftmet_;
Float_t lepsoft_fbrem_;
Float_t lepsoft_eOverPIn_;
Float_t lepsoft_q_;
Float_t lepsoft_dPhiIn_;
theTree->SetBranchAddress( "lephard_pt_" , &lephard_pt_ );
theTree->SetBranchAddress( "lepsoft_pt_" , &lepsoft_pt_ );
theTree->SetBranchAddress( "lepsoft_fr_" , &lepsoft_fr_ );
theTree->SetBranchAddress( "dil_dphi_" , &dil_dphi_ );
theTree->SetBranchAddress( "dil_mass_" , &dil_mass_ );
theTree->SetBranchAddress( "event_type_" , &event_type_ );
theTree->SetBranchAddress( "met_projpt_" , &met_projpt_ );
theTree->SetBranchAddress( "jets_num_" , &jets_num_ );
theTree->SetBranchAddress( "extralep_num_" , &extralep_num_ );
theTree->SetBranchAddress( "lowptbtags_num_" , &lowptbtags_num_ );
theTree->SetBranchAddress( "softmu_num_" , &softmu_num_ );
theTree->SetBranchAddress( "event_scale1fb_" , &event_scale1fb_ );
theTree->SetBranchAddress( "lepsoft_passTighterId_" , &lepsoft_passTighterId_ );
theTree->SetBranchAddress( "met_pt_" , &met_pt_ );
theTree->SetBranchAddress( "mt_lephardmet_" , &mt_lephardmet_ );
theTree->SetBranchAddress( "mt_lepsoftmet_" , &mt_lepsoftmet_ );
theTree->SetBranchAddress( "mthiggs_" , &mthiggs_ );
theTree->SetBranchAddress( "dphi_lephardmet_" , &dphi_lephardmet_ );
theTree->SetBranchAddress( "dphi_lepsoftmet_" , &dphi_lepsoftmet_ );
theTree->SetBranchAddress( "lepsoft_fbrem_" , &lepsoft_fbrem_ );
theTree->SetBranchAddress( "lepsoft_eOverPIn_" , &lepsoft_eOverPIn_ );
theTree->SetBranchAddress( "lepsoft_q_" , &lepsoft_q_ );
theTree->SetBranchAddress( "lepsoft_dPhiIn_" , &lepsoft_dPhiIn_ );
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
int npass = 0;
float yield = 0.;
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
if (ievt%1000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
theTree->GetEntry(ievt);
//-------------------------------------------------------
// event selection
//-------------------------------------------------------
if( dil_dphi_ > 1. ) continue;
//em
if( event_type_ > 0.5 && event_type_ < 2.5 ){
if( met_projpt_ < 20. ) continue;
}
//ee/mm
if( event_type_ < 0.5 || event_type_ > 2.5 ){
if( met_projpt_ < 35. ) continue;
}
if( lephard_pt_ < 20. ) continue;
if( jets_num_ > 0 ) continue;
if( extralep_num_ > 0 ) continue;
if( lowptbtags_num_ > 0 ) continue;
if( softmu_num_ > 0 ) continue;
if( dil_mass_ < 12. ) continue;
if( lepsoft_passTighterId_ == 0 ) continue;
//if( event_type_ < 1.5 ) continue;
//if( event_type > 1.5 && lepsoft_pt_ < 15. ) continue;
//mH-dependent selection
if( mH == 130 ){
if( lepsoft_pt_ < 10. ) continue;
if( dil_mass_ > 90. ) continue;
}
else if( mH == 160 ){
if( lepsoft_pt_ < 20. ) continue;
if( dil_mass_ > 100. ) continue;
}
else if( mH == 200 ){
if( lepsoft_pt_ < 20. ) continue;
if( dil_mass_ > 130. ) continue;
}
float weight = event_scale1fb_ * lepsoft_fr_ * 0.5;
//--------------------------------------------------------
// important: here we associate branches to MVA variables
//--------------------------------------------------------
lephard_pt = lephard_pt_;
lepsoft_pt = lepsoft_pt_;
dil_mass = dil_mass_;
dil_dphi = dil_dphi_;
event_type = event_type_;
met_pt = met_pt_;
met_projpt = met_projpt_;
mt_lephardmet = mt_lephardmet_;
mt_lepsoftmet = mt_lepsoftmet_;
mthiggs = mthiggs_;
dphi_lephardmet = dphi_lephardmet_;
dphi_lepsoftmet = dphi_lepsoftmet_;
lepsoft_fbrem = lepsoft_fbrem_;
lepsoft_eOverPIn = lepsoft_eOverPIn_;
lepsoft_qdphi = lepsoft_q_ * lepsoft_dPhiIn_;
npass++;
yield+=weight;
// var1 = userVar1 + userVar2;
// var2 = userVar1 - userVar2;
// --- Return the MVA outputs and fill into histograms
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
if (Use["Likelihood" ]) histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) , weight);
if (Use["LikelihoodD" ]) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) , weight);
if (Use["LikelihoodPCA"]) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) , weight);
if (Use["LikelihoodKDE"]) histLkKDE ->Fill( reader->EvaluateMVA( "LikelihoodKDE method" ) , weight);
if (Use["LikelihoodMIX"]) histLkMIX ->Fill( reader->EvaluateMVA( "LikelihoodMIX method" ) , weight);
if (Use["PDERS" ]) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) , weight);
if (Use["PDERSD" ]) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) , weight);
if (Use["PDERSPCA" ]) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) , weight);
if (Use["KNN" ]) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) , weight);
if (Use["HMatrix" ]) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) , weight);
if (Use["Fisher" ]) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) , weight);
if (Use["FisherG" ]) histFiG ->Fill( reader->EvaluateMVA( "FisherG method" ) , weight);
if (Use["BoostedFisher"]) histFiB ->Fill( reader->EvaluateMVA( "BoostedFisher method" ) , weight);
if (Use["LD" ]) histLD ->Fill( reader->EvaluateMVA( "LD method" ) , weight);
if (Use["MLP" ]) histNn ->Fill( reader->EvaluateMVA( "MLP method" ) , weight);
if (Use["MLPBFGS" ]) histNnbfgs ->Fill( reader->EvaluateMVA( "MLPBFGS method" ) , weight);
if (Use["MLPBNN" ]) histNnbnn ->Fill( reader->EvaluateMVA( "MLPBNN method" ) , weight);
if (Use["CFMlpANN" ]) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) , weight);
if (Use["TMlpANN" ]) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) , weight);
if (Use["BDT" ]) histBdt ->Fill( reader->EvaluateMVA( "BDT method" ) , weight);
if (Use["BDTD" ]) histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) , weight);
if (Use["BDTG" ]) histBdtG ->Fill( reader->EvaluateMVA( "BDTG method" ) , weight);
if (Use["RuleFit" ]) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) , weight);
if (Use["SVM_Gauss" ]) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) , weight);
if (Use["SVM_Poly" ]) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) , weight);
if (Use["SVM_Lin" ]) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) , weight);
if (Use["FDA_MT" ]) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) , weight);
if (Use["FDA_GA" ]) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) , weight);
if (Use["Category" ]) histCat ->Fill( reader->EvaluateMVA( "Category method" ) , weight);
if (Use["Plugin" ]) histPBdt ->Fill( reader->EvaluateMVA( "P_BDT method" ) , weight);
// Retrieve also per-event error
if (Use["PDEFoam"]) {
Double_t val = reader->EvaluateMVA( "PDEFoam method" );
Double_t err = reader->GetMVAError();
histPDEFoam ->Fill( val );
histPDEFoamErr->Fill( err );
if (err>1.e-50) histPDEFoamSig->Fill( val/err , weight);
}
// Retrieve probability instead of MVA output
if (Use["Fisher"]) {
probHistFi ->Fill( reader->GetProba ( "Fisher method" ) , weight);
rarityHistFi->Fill( reader->GetRarity( "Fisher method" ) , weight);
}
}
std::cout << npass << " events passing selection, yield " << yield << std::endl;
// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// Get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
// --- Write histograms
cout << "dir " << dir << endl;
char* mydir = outdir;
TFile *target = new TFile( Form("%s/%s.root",mydir,samples.at(i) ) ,"RECREATE" );
cout << "Writing to file " << Form("%s/%s.root",mydir,samples.at(i) ) << endl;
if (Use["Likelihood" ]) histLk ->Write();
if (Use["LikelihoodD" ]) histLkD ->Write();
if (Use["LikelihoodPCA"]) histLkPCA ->Write();
if (Use["LikelihoodKDE"]) histLkKDE ->Write();
if (Use["LikelihoodMIX"]) histLkMIX ->Write();
if (Use["PDERS" ]) histPD ->Write();
if (Use["PDERSD" ]) histPDD ->Write();
if (Use["PDERSPCA" ]) histPDPCA ->Write();
if (Use["KNN" ]) histKNN ->Write();
if (Use["HMatrix" ]) histHm ->Write();
if (Use["Fisher" ]) histFi ->Write();
if (Use["FisherG" ]) histFiG ->Write();
if (Use["BoostedFisher"]) histFiB ->Write();
if (Use["LD" ]) histLD ->Write();
if (Use["MLP" ]) histNn ->Write();
if (Use["MLPBFGS" ]) histNnbfgs ->Write();
if (Use["MLPBNN" ]) histNnbnn ->Write();
if (Use["CFMlpANN" ]) histNnC ->Write();
if (Use["TMlpANN" ]) histNnT ->Write();
if (Use["BDT" ]) histBdt ->Write();
if (Use["BDTD" ]) histBdtD ->Write();
if (Use["BDTG" ]) histBdtG ->Write();
if (Use["RuleFit" ]) histRf ->Write();
if (Use["SVM_Gauss" ]) histSVMG ->Write();
if (Use["SVM_Poly" ]) histSVMP ->Write();
if (Use["SVM_Lin" ]) histSVML ->Write();
if (Use["FDA_MT" ]) histFDAMT ->Write();
if (Use["FDA_GA" ]) histFDAGA ->Write();
if (Use["Category" ]) histCat ->Write();
if (Use["Plugin" ]) histPBdt ->Write();
// Write also error and significance histos
if (Use["PDEFoam"]) { histPDEFoam->Write(); histPDEFoamErr->Write(); histPDEFoamSig->Write(); }
// Write also probability hists
if (Use["Fisher"]) { if (probHistFi != 0) probHistFi->Write(); if (rarityHistFi != 0) rarityHistFi->Write(); }
target->Close();
delete reader;
std::cout << "==> TMVAClassificationApplication is done with sample " << samples.at(i) << endl << std::endl;
}
}
void ZTMVAClassificationApplication( string filename, TString myMethodList = "" )
{
#ifdef __CINT__
gROOT->ProcessLine( ".O0" ); // turn off optimization in CINT
#endif
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
// --- Cut optimisation
Use["Cuts"] = 0;
Use["CutsD"] = 0;
Use["CutsPCA"] = 0;
Use["CutsGA"] = 0;
Use["CutsSA"] = 0;
//
// --- 1-dimensional likelihood ("naive Bayes estimator")
Use["Likelihood"] = 0;
Use["LikelihoodD"] = 0; // the "D" extension indicates decorrelated input variables (see option strings)
Use["LikelihoodPCA"] = 0; // the "PCA" extension indicates PCA-transformed input variables (see option strings)
Use["LikelihoodKDE"] = 0;
Use["LikelihoodMIX"] = 0;
//
// --- Mutidimensional likelihood and Nearest-Neighbour methods
Use["PDERS"] = 0;
Use["PDERSD"] = 0;
Use["PDERSPCA"] = 0;
Use["PDEFoam"] = 0;
Use["PDEFoamBoost"] = 0; // uses generalised MVA method boosting
Use["KNN"] = 0; // k-nearest neighbour method
//
// --- Linear Discriminant Analysis
Use["LD"] = 0; // Linear Discriminant identical to Fisher
Use["Fisher"] = 0;
Use["FisherG"] = 0;
Use["BoostedFisher"] = 0; // uses generalised MVA method boosting
Use["HMatrix"] = 0;
//
// --- Function Discriminant analysis
Use["FDA_GA"] = 0; // minimisation of user-defined function using Genetics Algorithm
Use["FDA_SA"] = 0;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 0;
Use["FDA_GAMT"] = 0;
Use["FDA_MCMT"] = 0;
//
// --- Neural Networks (all are feed-forward Multilayer Perceptrons)
Use["MLP"] = 1; // Recommended ANN
Use["MLPBFGS"] = 0; // Recommended ANN with optional training method
Use["MLPBNN"] = 0; // Recommended ANN with BFGS training method and bayesian regulator
Use["CFMlpANN"] = 0; // Depreciated ANN from ALEPH
Use["TMlpANN"] = 0; // ROOT's own ANN
//
// --- Support Vector Machine
Use["SVM"] = 0;
//
// --- Boosted Decision Trees
Use["BDT"] = 1; // uses Adaptive Boost
Use["BDTG"] = 1; // uses Gradient Boost
Use["BDTB"] = 0; // uses Bagging
Use["BDTD"] = 1; // decorrelation + Adaptive Boost
//
// --- Friedman's RuleFit method, ie, an optimised series of cuts ("rules")
Use["RuleFit"] = 0;
// ---------------------------------------------------------------
Use["Plugin"] = 0;
Use["Category"] = 0;
Use["SVM_Gauss"] = 0;
Use["SVM_Poly"] = 0;
Use["SVM_Lin"] = 0;
std::cout << std::endl;
std::cout << "==> Start TMVAClassificationApplication" << std::endl;
// Select methods (don't look at this code - not of interest)
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod
<< "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
std::cout << it->first << " ";
}
std::cout << std::endl;
return;
}
Use[regMethod] = 1;
}
}
// --------------------------------------------------------------------------------------------------
// --- Create the Reader object
TMVA::Reader *reader = new TMVA::Reader( "!Color:!Silent" );
Float_t B_s0_ln_FDCHI2; reader->AddVariable("B_s0_ln_FDCHI2", &B_s0_ln_FDCHI2 );
Float_t B_s0_ln_IPCHI2; reader->AddVariable("B_s0_ln_IPCHI2", &B_s0_ln_IPCHI2 );
Float_t B_s0_ln_EVCHI2; reader->AddVariable("B_s0_ln_EVCHI2", &B_s0_ln_EVCHI2 );
Float_t B_s0_PT_fiveGeV;reader->AddVariable("B_s0_PT_fiveGeV",&B_s0_PT_fiveGeV);
Float_t B_s0_Eta; reader->AddVariable("B_s0_Eta", &B_s0_Eta );
Float_t minK_PT_GeV; reader->AddVariable("minK_PT_GeV", &minK_PT_GeV );
Float_t minK_ln_IPCHI2; reader->AddVariable("minK_ln_IPCHI2", &minK_ln_IPCHI2 );
Float_t Category_cat1, Category_cat2, Category_cat3;
if (Use["Category"]){
// Add artificial spectators for distinguishing categories
reader->AddSpectator( "Category_cat1 := var3<=0", &Category_cat1 );
reader->AddSpectator( "Category_cat2 := (var3>0)&&(var4<0)", &Category_cat2 );
reader->AddSpectator( "Category_cat3 := (var3>0)&&(var4>=0)", &Category_cat3 );
}
// --- Book the MVA methods
TString dir = "weights/";
TString prefix = "TMVAClassification";
// Book method(s)
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString methodName = TString(it->first) + TString(" method");
TString weightfile = dir + prefix + TString("_") + TString(it->first) + TString(".weights.xml");
reader->BookMVA( methodName, weightfile );
}
}
// Book output histograms
UInt_t nbin = 100;
TH1F *histLk(0), *histLkD(0), *histLkPCA(0), *histLkKDE(0), *histLkMIX(0), *histPD(0), *histPDD(0);
TH1F *histPDPCA(0), *histPDEFoam(0), *histPDEFoamErr(0), *histPDEFoamSig(0), *histKNN(0), *histHm(0);
TH1F *histFi(0), *histFiG(0), *histFiB(0), *histLD(0), *histNn(0),*histNnbfgs(0),*histNnbnn(0);
TH1F *histNnC(0), *histNnT(0), *histBdt(0), *histBdtG(0), *histBdtD(0), *histRf(0), *histSVMG(0);
TH1F *histSVMP(0), *histSVML(0), *histFDAMT(0), *histFDAGA(0), *histCat(0), *histPBdt(0);
if (Use["Likelihood"]) histLk = new TH1F( "MVA_Likelihood", "MVA_Likelihood", nbin, -1, 1 );
if (Use["LikelihoodD"]) histLkD = new TH1F( "MVA_LikelihoodD", "MVA_LikelihoodD", nbin, -1, 0.9999 );
if (Use["LikelihoodPCA"]) histLkPCA = new TH1F( "MVA_LikelihoodPCA", "MVA_LikelihoodPCA", nbin, -1, 1 );
if (Use["LikelihoodKDE"]) histLkKDE = new TH1F( "MVA_LikelihoodKDE", "MVA_LikelihoodKDE", nbin, -0.00001, 0.99999 );
if (Use["LikelihoodMIX"]) histLkMIX = new TH1F( "MVA_LikelihoodMIX", "MVA_LikelihoodMIX", nbin, 0, 1 );
if (Use["PDERS"]) histPD = new TH1F( "MVA_PDERS", "MVA_PDERS", nbin, 0, 1 );
if (Use["PDERSD"]) histPDD = new TH1F( "MVA_PDERSD", "MVA_PDERSD", nbin, 0, 1 );
if (Use["PDERSPCA"]) histPDPCA = new TH1F( "MVA_PDERSPCA", "MVA_PDERSPCA", nbin, 0, 1 );
if (Use["KNN"]) histKNN = new TH1F( "MVA_KNN", "MVA_KNN", nbin, 0, 1 );
if (Use["HMatrix"]) histHm = new TH1F( "MVA_HMatrix", "MVA_HMatrix", nbin, -0.95, 1.55 );
if (Use["Fisher"]) histFi = new TH1F( "MVA_Fisher", "MVA_Fisher", nbin, -4, 4 );
if (Use["FisherG"]) histFiG = new TH1F( "MVA_FisherG", "MVA_FisherG", nbin, -1, 1 );
if (Use["BoostedFisher"]) histFiB = new TH1F( "MVA_BoostedFisher", "MVA_BoostedFisher", nbin, -2, 2 );
if (Use["LD"]) histLD = new TH1F( "MVA_LD", "MVA_LD", nbin, -2, 2 );
if (Use["MLP"]) histNn = new TH1F( "MVA_MLP", "MVA_MLP", nbin, -1.25, 1.5 );
if (Use["MLPBFGS"]) histNnbfgs = new TH1F( "MVA_MLPBFGS", "MVA_MLPBFGS", nbin, -1.25, 1.5 );
if (Use["MLPBNN"]) histNnbnn = new TH1F( "MVA_MLPBNN", "MVA_MLPBNN", nbin, -1.25, 1.5 );
if (Use["CFMlpANN"]) histNnC = new TH1F( "MVA_CFMlpANN", "MVA_CFMlpANN", nbin, 0, 1 );
if (Use["TMlpANN"]) histNnT = new TH1F( "MVA_TMlpANN", "MVA_TMlpANN", nbin, -1.3, 1.3 );
if (Use["BDT"]) histBdt = new TH1F( "MVA_BDT", "MVA_BDT", nbin, -0.8, 0.8 );
if (Use["BDTD"]) histBdtD = new TH1F( "MVA_BDTD", "MVA_BDTD", nbin, -0.8, 0.8 );
if (Use["BDTG"]) histBdtG = new TH1F( "MVA_BDTG", "MVA_BDTG", nbin, -1.0, 1.0 );
if (Use["RuleFit"]) histRf = new TH1F( "MVA_RuleFit", "MVA_RuleFit", nbin, -2.0, 2.0 );
if (Use["SVM_Gauss"]) histSVMG = new TH1F( "MVA_SVM_Gauss", "MVA_SVM_Gauss", nbin, 0.0, 1.0 );
if (Use["SVM_Poly"]) histSVMP = new TH1F( "MVA_SVM_Poly", "MVA_SVM_Poly", nbin, 0.0, 1.0 );
if (Use["SVM_Lin"]) histSVML = new TH1F( "MVA_SVM_Lin", "MVA_SVM_Lin", nbin, 0.0, 1.0 );
if (Use["FDA_MT"]) histFDAMT = new TH1F( "MVA_FDA_MT", "MVA_FDA_MT", nbin, -2.0, 3.0 );
if (Use["FDA_GA"]) histFDAGA = new TH1F( "MVA_FDA_GA", "MVA_FDA_GA", nbin, -2.0, 3.0 );
if (Use["Category"]) histCat = new TH1F( "MVA_Category", "MVA_Category", nbin, -2., 2. );
if (Use["Plugin"]) histPBdt = new TH1F( "MVA_PBDT", "MVA_BDT", nbin, -0.8, 0.8 );
// PDEFoam also returns per-event error, fill in histogram, and also fill significance
if (Use["PDEFoam"]) {
histPDEFoam = new TH1F( "MVA_PDEFoam", "MVA_PDEFoam", nbin, 0, 1 );
histPDEFoamErr = new TH1F( "MVA_PDEFoamErr", "MVA_PDEFoam error", nbin, 0, 1 );
histPDEFoamSig = new TH1F( "MVA_PDEFoamSig", "MVA_PDEFoam significance", nbin, 0, 10 );
}
// Book example histogram for probability (the other methods are done similarly)
TH1F *probHistFi(0), *rarityHistFi(0);
if (Use["Fisher"]) {
probHistFi = new TH1F( "MVA_Fisher_Proba", "MVA_Fisher_Proba", nbin, 0, 1 );
rarityHistFi = new TH1F( "MVA_Fisher_Rarity", "MVA_Fisher_Rarity", nbin, 0, 1 );
}
//TFile * input = new TFile("../cloosepid.root"); // this is the signal
//TFile * input_Background = new TFile("Z4430Files/merged_ntuple_jpsi_s17.root"); // this is the background
//TFile * input = new TFile("../output/MCBsphif0_after_transform.root"); // this is the signal
TFile * input_Background;
input_Background = new TFile(filename.c_str());
//std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
std::cout << "--- TMVAClassificationApp : Using input file: " << input_Background->GetName() << std::endl;
// --- Event loop
//save the results here
/*
TTree* results= new TTree("results", "results");
Float_t Method_Likelihood; Float_t bdt;
results->Branch("Method_Likelihood",&Method_Likelihood ,"Method_Likelihood/D" );
results->Branch("BDTG method", &bdt, "BDTG method" );
*/
std::cout << "--- Select signal sample" << std::endl;
//TTree* theTree = (TTree*)input->Get("MCtree");
TTree* theTree = (TTree*)input_Background->Get("DecayTree");
//if(mode<3) TCut cut = TCut("time1>0. && abs(phi_mass-1019.455)<15");
// else TCut cut = TCut("time1>0. && abs(phi_mass-1019.455)<12 && abs(phi1_mass-1019.455)<12");
//else TCut cut = TCut("time1>0. && abs(phi_mass-1019.455)<15 && abs(phi1_mass-1019.455)<15");
//TFile* f_out =new TFile("../output/MCBsphif0_after_bdt.root","RECREATE");
TFile* f_out;
string oldLabel="mvaVars_vetoes";
string newLabel="mva";
filename.replace(filename.find(oldLabel), oldLabel.length(), newLabel);
f_out = new TFile(filename.c_str(),"RECREATE");
//TTree* smalltree = theTree->CopyTree(cut);
TTree* newtree = theTree->CloneTree(-1);
//TTree* smalltree = theTree->CloneTree(-1);
//TTree* newtree = theTree->CloneTree(-1);
float bdtg;
TBranch* b_bdtg = newtree->Branch("bdtg", &bdtg,"bdtg/F");
float bdt;
TBranch* b_bdt = newtree->Branch("bdt", &bdt,"bdt/F");
float bdtd;
TBranch* b_bdtd = newtree->Branch("bdtd", &bdtd,"bdtd/F");
float mlp;
TBranch* b_mlp = newtree->Branch("mlp", &mlp,"mlp/F");
// Float_t userptsum, userpionpt, userptj, userdmj , uservchi2dof;
// Float_t usermaxdphi; Float_t userptAsym;
theTree->SetBranchAddress("B_s0_ln_FDCHI2", &B_s0_ln_FDCHI2 );
theTree->SetBranchAddress("B_s0_ln_IPCHI2", &B_s0_ln_IPCHI2 );
theTree->SetBranchAddress("B_s0_ln_EVCHI2", &B_s0_ln_EVCHI2 );
theTree->SetBranchAddress("B_s0_PT_fiveGeV",&B_s0_PT_fiveGeV);
theTree->SetBranchAddress("B_s0_Eta", &B_s0_Eta );
theTree->SetBranchAddress("minK_PT_GeV", &minK_PT_GeV );
theTree->SetBranchAddress("minK_ln_IPCHI2", &minK_ln_IPCHI2 );
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
Int_t num_entries = newtree->GetEntries();
std::cout << "--- Processing: " << num_entries << " events" << std::endl;
TStopwatch sw;
sw.Start();
for (Long64_t ievt=0; ievt<num_entries;ievt++) {
// if (ievt%10000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
newtree->GetEntry(ievt);
// var1 = userVar1 + userVar2;
// var2 = userVar1 - userVar2;
// --- Return the MVA outputs and fill into histograms
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
if (Use["Likelihood" ]) {
histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) );
// Method_Likelihood = reader->EvaluateMVA( "Likelihood method" ) ;
//std::cout << Method_Likelihood << std::endl;
// results->Fill();
}
if (Use["LikelihoodD" ]) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) );
if (Use["LikelihoodPCA"]) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) );
if (Use["LikelihoodKDE"]) histLkKDE ->Fill( reader->EvaluateMVA( "LikelihoodKDE method" ) );
if (Use["LikelihoodMIX"]) histLkMIX ->Fill( reader->EvaluateMVA( "LikelihoodMIX method" ) );
if (Use["PDERS" ]) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) );
if (Use["PDERSD" ]) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) );
if (Use["PDERSPCA" ]) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) );
if (Use["KNN" ]) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) );
if (Use["HMatrix" ]) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) );
if (Use["Fisher" ]) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) );
if (Use["FisherG" ]) histFiG ->Fill( reader->EvaluateMVA( "FisherG method" ) );
if (Use["BoostedFisher"]) histFiB ->Fill( reader->EvaluateMVA( "BoostedFisher method" ) );
if (Use["LD" ]) histLD ->Fill( reader->EvaluateMVA( "LD method" ) );
if (Use["MLP" ]) {
mlp = reader->EvaluateMVA( "MLP method" ) ;
b_mlp->Fill();
histNn ->Fill( reader->EvaluateMVA( "MLP method" ) );
}
if (Use["MLPBFGS" ]) histNnbfgs ->Fill( reader->EvaluateMVA( "MLPBFGS method" ) );
if (Use["MLPBNN" ]) histNnbnn ->Fill( reader->EvaluateMVA( "MLPBNN method" ) );
if (Use["CFMlpANN" ]) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) );
if (Use["TMlpANN" ]) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) );
if (Use["BDT" ]) {
histBdt ->Fill( reader->EvaluateMVA( "BDT method" ) );
bdt = reader->EvaluateMVA( "BDT method" ) ;
b_bdt->Fill();
}
if (Use["BDTD" ]) {
histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) );
bdtd = reader->EvaluateMVA( "BDTD method" );
b_bdtd->Fill();
}
if (Use["BDTG" ]) {
histBdtG ->Fill( reader->EvaluateMVA( "BDTG method" ) );
bdtg = reader->EvaluateMVA( "BDTG method" );
b_bdtg->Fill();
//cout << reader->EvaluateMVA( "BDTG method" ) <<endl;
}
if (Use["RuleFit" ]) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) );
if (Use["SVM_Gauss" ]) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) );
if (Use["SVM_Poly" ]) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) );
if (Use["SVM_Lin" ]) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) );
if (Use["FDA_MT" ]) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) );
if (Use["FDA_GA" ]) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) );
if (Use["Category" ]) histCat ->Fill( reader->EvaluateMVA( "Category method" ) );
if (Use["Plugin" ]) histPBdt ->Fill( reader->EvaluateMVA( "P_BDT method" ) );
// Retrieve also per-event error
if (Use["PDEFoam"]) {
Double_t val = reader->EvaluateMVA( "PDEFoam method" );
Double_t err = reader->GetMVAError();
histPDEFoam ->Fill( val );
histPDEFoamErr->Fill( err );
if (err>1.e-50) histPDEFoamSig->Fill( val/err );
}
// Retrieve probability instead of MVA output
if (Use["Fisher"]) {
probHistFi ->Fill( reader->GetProba ( "Fisher method" ) );
rarityHistFi->Fill( reader->GetRarity( "Fisher method" ) );
}
}
// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// Get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
newtree->Write();
f_out->Close();
// --- Write histograms
TFile *target = new TFile( "TMVApp.root","RECREATE" );
if (Use["Likelihood" ]) histLk ->Write();
if (Use["LikelihoodD" ]) histLkD ->Write();
if (Use["LikelihoodPCA"]) histLkPCA ->Write();
if (Use["LikelihoodKDE"]) histLkKDE ->Write();
if (Use["LikelihoodMIX"]) histLkMIX ->Write();
if (Use["PDERS" ]) histPD ->Write();
if (Use["PDERSD" ]) histPDD ->Write();
if (Use["PDERSPCA" ]) histPDPCA ->Write();
if (Use["KNN" ]) histKNN ->Write();
if (Use["HMatrix" ]) histHm ->Write();
if (Use["Fisher" ]) histFi ->Write();
if (Use["FisherG" ]) histFiG ->Write();
if (Use["BoostedFisher"]) histFiB ->Write();
if (Use["LD" ]) histLD ->Write();
if (Use["MLP" ]) histNn ->Write();
if (Use["MLPBFGS" ]) histNnbfgs ->Write();
if (Use["MLPBNN" ]) histNnbnn ->Write();
if (Use["CFMlpANN" ]) histNnC ->Write();
if (Use["TMlpANN" ]) histNnT ->Write();
if (Use["BDT" ]) histBdt ->Write();
if (Use["BDTD" ]) histBdtD ->Write();
if (Use["BDTG" ]) histBdtG ->Write();
if (Use["RuleFit" ]) histRf ->Write();
if (Use["SVM_Gauss" ]) histSVMG ->Write();
if (Use["SVM_Poly" ]) histSVMP ->Write();
if (Use["SVM_Lin" ]) histSVML ->Write();
if (Use["FDA_MT" ]) histFDAMT ->Write();
if (Use["FDA_GA" ]) histFDAGA ->Write();
if (Use["Category" ]) histCat ->Write();
if (Use["Plugin" ]) histPBdt ->Write();
// results->Write();
// Write also error and significance histos
if (Use["PDEFoam"]) { histPDEFoam->Write(); histPDEFoamErr->Write(); histPDEFoamSig->Write(); }
// Write also probability hists
if (Use["Fisher"]) { if (probHistFi != 0) probHistFi->Write(); if (rarityHistFi != 0) rarityHistFi->Write(); }
target->Close();
std::cout << "--- Created_prob.root file: \"TMVApp.root\" containing the MVA output histograms" << std::endl;
delete reader;
std::cout << "==> TMVAClassificationApplication is done!" << endl << std::endl;
}
void DYPtZ_HF_BDTCut( TString myMethodList = "" )
{
#ifdef __CINT__
gROOT->ProcessLine( ".O0" ); // turn off optimization in CINT
#endif
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
// --- Cut optimisation
Use["Cuts"] = 0;
Use["CutsD"] = 0;
Use["CutsPCA"] = 0;
Use["CutsGA"] = 0;
Use["CutsSA"] = 0;
//
// --- 1-dimensional likelihood ("naive Bayes estimator")
Use["Likelihood"] = 0;
Use["LikelihoodD"] = 0; // the "D" extension indicates decorrelated input variables (see option strings)
Use["LikelihoodPCA"] = 0; // the "PCA" extension indicates PCA-transformed input variables (see option strings)
Use["LikelihoodKDE"] = 0;
Use["LikelihoodMIX"] = 0;
//
// --- Mutidimensional likelihood and Nearest-Neighbour methods
Use["PDERS"] = 0;
Use["PDERSD"] = 0;
Use["PDERSPCA"] = 0;
Use["PDEFoam"] = 0;
Use["PDEFoamBoost"] = 0; // uses generalised MVA method boosting
Use["KNN"] = 0; // k-nearest neighbour method
//
// --- Linear Discriminant Analysis
Use["LD"] = 0; // Linear Discriminant identical to Fisher
Use["Fisher"] = 0;
Use["FisherG"] = 0;
Use["BoostedFisher"] = 0; // uses generalised MVA method boosting
Use["HMatrix"] = 0;
//
// --- Function Discriminant analysis
Use["FDA_GA"] = 0; // minimisation of user-defined function using Genetics Algorithm
Use["FDA_SA"] = 0;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 0;
Use["FDA_GAMT"] = 0;
Use["FDA_MCMT"] = 0;
//
// --- Neural Networks (all are feed-forward Multilayer Perceptrons)
Use["MLP"] = 0; // Recommended ANN
Use["MLPBFGS"] = 0; // Recommended ANN with optional training method
Use["MLPBNN"] = 0; // Recommended ANN with BFGS training method and bayesian regulator
Use["CFMlpANN"] = 0; // Depreciated ANN from ALEPH
Use["TMlpANN"] = 0; // ROOT's own ANN
//
// --- Support Vector Machine
Use["SVM"] = 0;
//
// --- Boosted Decision Trees using this
Use["BDT"] = 1; // uses Adaptive Boost
Use["BDTG"] = 0; // uses Gradient Boost
Use["BDTB"] = 0; // uses Bagging
Use["BDTD"] = 0; // decorrelation + Adaptive Boost
//
// --- Friedman's RuleFit method, ie, an optimised series of cuts ("rules")
Use["RuleFit"] = 0;
// ---------------------------------------------------------------
Use["Plugin"] = 0;
Use["Category"] = 0;
Use["SVM_Gauss"] = 0;
Use["SVM_Poly"] = 0;
Use["SVM_Lin"] = 0;
std::cout << std::endl;
std::cout << "==> Start DYPtZ_HF_BDTCut" << std::endl;
// Select methods (don't look at this code - not of interest)
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod
<< "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
std::cout << it->first << " ";
}
std::cout << std::endl;
return;
}
Use[regMethod] = 1;
}
}
// --------------------------------------------------------------------------------------------------
// --- Create the Reader object
TMVA::Reader *reader = new TMVA::Reader( "!Color:!Silent" );
// Create a set of variables and declare them to the reader
// - the variable names MUST corresponds in name and type to those given in the weight file(s) used
Float_t Hmass, Emumass;
Float_t Hpt, Zpt;
Float_t CSV0, CSV1;
Float_t DeltaPhiHV, DetaJJ;
Int_t nJets, eventFlavor, Naj;
Float_t BDTvalue, Trigweight, B2011PUweight, A2011PUweight, btag2CSF, MET;
Float_t alpha_j, qtb1, jetPhi0, jetPhi1, jetEta0, jetEta1, Zphi, Hphi;
Float_t Ht, EvntShpCircularity, jetCHF0, jetCHF1;
Float_t EtaStandDev, lep_pfCombRelIso0, lep_pfCombRelIso1, EvntShpIsotropy;
Float_t lep0pt, lep1pt, UnweightedEta, DphiJJ, RMS_eta, EvntShpSphericity;
Float_t PtbalZH, EventPt, AngleEMU, Centrality, EvntShpAplanarity;
Float_t UnweightedEta, lep0pt;
Int_t naJets, nSV;
Float_t Mte, Mtmu, dPhiHMET, Dphiemu, delRjj, delRemu, DphiZMET, DeltaPhijetMETmin;
Float_t MassEleb0, DphiEleMET, PtbalMETH,dphiEleMET, dEtaJJ, dphiZMET, ScalarSumPt;
Float_t ZmassSVD, AngleHemu, ProjVisT, topMass, topPt, ZmassSVDnegSol, Mt, Zmass, ZmassNegInclu;
Float_t dphiEMU, dphiZMET;
reader->AddVariable( "Hmass", &Hmass );
//reader->AddVariable( "Naj", &Naj );
reader->AddVariable( "CSV0", &CSV0 );
reader->AddVariable( "Emumass", &Emumass );
reader->AddVariable( "DeltaPhiHV", &DeltaPhiHV );
reader->AddVariable( "Mt", &Mt );
reader->AddVariable( "dPhiHMET", &dPhiHMET );
reader->AddVariable( "dphiEMU := abs(Dphiemu)", &dphiEMU );
reader->AddVariable( "dphiZMET:=abs(DphiZMET)", &dphiZMET );
reader->AddVariable( "PtbalMETH", &PtbalMETH );
reader->AddVariable( "EtaStandDev", &EtaStandDev );
reader->AddVariable( "jetCHF0", &jetCHF0 );
reader->AddVariable( "ProjVisT", &ProjVisT );
// Spectator variables declared in the training have to be added to the reader, too
// reader->AddSpectator( "UnweightedEta", &UnweightedEta );
// reader->AddSpectator( "lep0pt", &lep0pt );
/* Float_t Category_cat1, Category_cat2, Category_cat3;
if (Use["Category"]){
// Add artificial spectators for distinguishing categories
reader->AddSpectator( "Category_cat1 := var3<=0", &Category_cat1 );
reader->AddSpectator( "Category_cat2 := (var3>0)&&(var4<0)", &Category_cat2 );
reader->AddSpectator( "Category_cat3 := (var3>0)&&(var4>=0)", &Category_cat3 );
}
*/
// --- Book the MVA methods
TString dir = "weights/";
TString prefix = "TMVAClassification";
// Book method(s)
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString methodName = TString(it->first) + TString(" method");
TString weightfile = dir + prefix + TString("_") + TString(it->first) + TString(".weights.xml");
reader->BookMVA( methodName, weightfile );
}
}
// Book output histograms
TH1F* hCHFb0_OpenSelection= new TH1F ("hCHFb0_OpenSelection", "charged Hadron Energy Fraction b1", 40, 0.0, 1.2);
TH1F* hCHFb1_OpenSelection= new TH1F ("hCHFb1_OpenSelection", "charged Hadron Energy Fraction b2", 40, 0.0, 1.2);
TH1F* hPtjj_OpenSelection= new TH1F ("hPtjj_OpenSelection","Pt of two b jets with highest CSV ", 50, 0.0, 400);
TH1F* hPtmumu_OpenSelection= new TH1F ("hPtmumu_OpenSelection","Pt of two muons with highest pt ", 50, 0.0, 400);
TH1F* hPtbalZH_OpenSelection= new TH1F ("hPtbalZH_OpenSelection", "Pt balance of Z and H", 40, -80, 80);
TH1F* hPtmu0_OpenSelection= new TH1F ("hPtmu0_OpenSelection","Pt of muon with highest pt ", 30, 0.0, 300);
TH1F* hPtmu1_OpenSelection= new TH1F ("hPtmu1_OpenSelection","Pt of muon with second highest pt ", 30, 0.0, 300);
TH1F* hPFRelIsomu0_OpenSelection= new TH1F ("hPFRelIsomu0_OpenSelection", "PF Rel Iso of muon with highest Pt", 40, 0., 0.2);
TH1F* hPFRelIsomu1_OpenSelection= new TH1F ("hPFRelIsomu1_OpenSelection", "PF Rel Iso of muon with second highest Pt", 40, 0., 0.2);
TH1F* hCSV0_OpenSelection= new TH1F ("hCSV0_OpenSelection","Jet with highest CSV ", 40, 0, 1.5);
TH1F* hCSV1_OpenSelection= new TH1F ("hCSV1_OpenSelection","Jet with second highest CSV ", 40, 0, 1.5);
TH1F* hdphiVH_OpenSelection= new TH1F ("hdphiVH_OpenSelection","Delta phi between Z and Higgs ", 50, -0.1, 4.5);
TH1F* hdetaJJ_OpenSelection= new TH1F ("hdetaJJ_OpenSelection","Delta eta between two jets ", 60, -4, 4);
TH1F* hNjets_OpenSelection= new TH1F ("hNjets_OpenSelection", "Number of Jets", 13, -2.5, 10.5);
TH1F* hMjj_OpenSelection = new TH1F ("hMjj_OpenSelection", "Invariant Mass of two Jets ", 50, 0, 300);
TH1F* hMmumu_OpenSelection = new TH1F ("hMmumu_OpenSelection", "Invariant Mass of two muons ", 75, 0, 200);
TH1F* hRMSeta_OpenSelection= new TH1F ("hRMSeta_OpenSelection", "RMS Eta", 30, 0, 3);
TH1F* hStaDeveta_OpenSelection= new TH1F ("hStaDeveta_OpenSelection", "Standard Deviation Eta", 30, 0, 3);
TH1F* hUnweightedEta_OpenSelection= new TH1F ("hUnweightedEta_OpenSelection", "Unweighted Eta ", 50, 0, 15);
TH1F* hdphiJJ_vect_OpenSelection= new TH1F ("hdphiJJ_vect_OpenSelection", "Delta phi between two jets", 30, -3.5, 4);
TH1F* hCircularity_OpenSelection= new TH1F("hCircularity_OpenSelection","EventShapeVariables circularity", 30, 0.0, 1.2);
TH1F* hHt_OpenSelection= new TH1F("hHt_OpenSelection","scalar sum of pt of four particles", 50, 0.0, 500);
TH1F* hCentrality_OpenSelection= new TH1F ("hCentrality_OpenSelection", "Centrality", 40, 0.0, 0.8);
TH1F* hEventPt_OpenSelection= new TH1F ("hEventPt_OpenSelection", "Pt of HV system", 50, 0.0, 100);
TH1F* hAngleEMU_OpenSelection= new TH1F ("hAngleEMU_OpenSelection", "AngleEMU between H and Z", 45, 0, 4.5);
TH1F* hSphericity_OpenSelection= new TH1F ("hSphericity_OpenSelection", "EventShapeVariables sphericity", 50, 0.0, 1);
TH1F* hAplanarity_OpenSelection= new TH1F ("hAplanarity_OpenSelection", "EventShapeVariables Aplanarity", 50, -0.1, .4);
TH1F* hIsotropy_OpenSelection= new TH1F ("hIsotropy_OpenSelection", "EventShapeVariables isotropy", 30, 0.0, 1.3);
TH2F* hDphiDetajj_OpenSelection= new TH2F ("hDphiDetajj_OpenSelection", "#Delta#phi vs #Delta#eta JJ", 25, -5, 5, 25, -5, 5);
TH1F* hMtmu_OpenSelection= new TH1F ("hMtmu_OpenSelection", "Mt with respect to Muon", 101, -0.1, 200);
TH1F* hMte_OpenSelection= new TH1F ("hMte_OpenSelection", "Mt with respect to Electron", 101, -0.1, 200);
TH1F* hdPhiHMET_OpenSelection= new TH1F ("hdPhiHMET_OpenSelection", "Delta phi between MET and Higgs", 50, -0.1, 4.5);
TH1F* hDphiemu_OpenSelection= new TH1F ("hDphiemu_OpenSelection", "Delta phi between e and muon", 50, -3.5, 4.5);
TH1F* hdelRjj_OpenSelection= new TH1F ("hdelRjj_OpenSelection", "Delta R jj", 55, 0, 5.5);
TH1F* hdelRemu_OpenSelection= new TH1F ("hdelRemu_OpenSelection", "Delta R emu", 55, 0, 5.5);
TH1F* hDphiZMET_OpenSelection= new TH1F ("hDphiZMET_OpenSelection", "Delta phi between Z and MET", 71, -3.5, 4);
TH1F* hDeltaPhijetMETmin_OpenSelection= new TH1F ("hDeltaPhijetMETmin_OpenSelection", "Delta phi between MET and nearest jet", 50, -0.1, 4.5);
TH1F* hAngleHemu_OpenSelection= new TH1F ("hAngleHemu_OpenSelection", "Angle between H and Z", 30, 0, 3.5);
TH1F* hProjVisT_OpenSelection= new TH1F ("hProjVisT_OpenSelection", "Transverse componenet of Projection of Z onto bisector", 80, 0, 200);
TH1F* htopMass_OpenSelection= new TH1F ("htopMass_OpenSelection", "Top Mass single lepton", 100, 75, 375);
TH1F* htopPt_OpenSelection= new TH1F ("htopPt_OpenSelection", "Pt of Top", 125, 0, 250);
TH1F* hVMt_OpenSelection= new TH1F ("hVMt_OpenSelection", "VMt", 75, 0, 150);
TH1F* hZmassSVD_OpenSelection= new TH1F ("hZmassSVD_OpenSelection", "Invariant Mass of two Leptons corrected SVD", 75, 0, 150);
TH1F* hZmassSVDnegSol_OpenSelection= new TH1F ("hZmassSVDnegSol_OpenSelection", "Invariant Mass of two Leptons corrected SVD", 100, -50, 200);
TH1F* hZmass_OpenSelection= new TH1F ("hZmass_OpenSelection", "Zmass ", 5, 0, 150);
TH1F* hZmassNegInclu_OpenSelection= new TH1F ("hZmassNegInclu_OpenSelection", "Invariant Mass of two Leptons corrected SVD", 100, -50, 200);
TTree *treeWithBDT = new TTree("treeWithBDT","Tree wiht BDT output");
treeWithBDT->SetDirectory(0);
treeWithBDT->Branch("nJets",&nJets, "nJets/I");
treeWithBDT->Branch("naJets",&naJets, "naJets/I");
treeWithBDT->Branch("eventFlavor",&eventFlavor, "eventFlavor/I");
treeWithBDT->Branch("CSV0",&CSV0, "CSV0/F");
treeWithBDT->Branch("CSV1",&CSV1, "CSV1/F");
treeWithBDT->Branch("Emumass",&Emumass, "Emumass/F");
treeWithBDT->Branch("Hmass",&Hmass, "Hmass/F");
treeWithBDT->Branch("DeltaPhiHV",&DeltaPhiHV, "DeltaPhiHV/F");
treeWithBDT->Branch("Hpt",&Hpt, "Hpt/F");
treeWithBDT->Branch("Zpt",&Zpt, "Zpt/F");
treeWithBDT->Branch("lep0pt",&lep0pt, "lep0pt/F");
treeWithBDT->Branch("Ht",&Ht, "Ht/F");
treeWithBDT->Branch("EtaStandDev",&EtaStandDev, "EtaStandDev/F");
treeWithBDT->Branch("UnweightedEta",&UnweightedEta, "UnweightedEta/F");
treeWithBDT->Branch("EvntShpCircularity",&EvntShpCircularity, "EvntShpCircularity/F");
treeWithBDT->Branch("alpha_j",&alpha_j, "alpha_j/F");
treeWithBDT->Branch("qtb1",&qtb1, "qtb1/F");
treeWithBDT->Branch("nSV",&nSV, "nSV/I");
treeWithBDT->Branch("Trigweight",&Trigweight, "Trigweight/F");
treeWithBDT->Branch("B2011PUweight",&B2011PUweight, "B2011PUweight/F");
treeWithBDT->Branch("A2011PUweight",&A2011PUweight, "A2011PUweight/F");
treeWithBDT->Branch("btag2CSF",&btag2CSF, "btag2CSF/F");
treeWithBDT->Branch("DetaJJ",&DetaJJ, "DetaJJ/F");
treeWithBDT->Branch("jetCHF0",&jetCHF0, "jetCHF0/F");
treeWithBDT->Branch("jetCHF1",&jetCHF1, "jetCHF1/F");
treeWithBDT->Branch("jetPhi0",&jetPhi0, "jetPhi0/F");
treeWithBDT->Branch("jetPhi1",&jetPhi1, "jetPhi1/F");
treeWithBDT->Branch("jetEta0",&jetEta0, "jetEta0/F");
treeWithBDT->Branch("jetEta1",&jetEta1, "jetEta1/F");
treeWithBDT->Branch("lep1pt",&lep1pt, "lep1pt/F");
treeWithBDT->Branch("lep_pfCombRelIso0",&lep_pfCombRelIso0, "lep_pfCombRelIso0/F");
treeWithBDT->Branch("lep_pfCombRelIso1",&lep_pfCombRelIso1, "lep_pfCombRelIso1/F");
treeWithBDT->Branch("DphiJJ",&DphiJJ, "DphiJJ/F");
treeWithBDT->Branch("RMS_eta",&RMS_eta, "RMS_eta/F");
treeWithBDT->Branch("PtbalZH",&PtbalZH, "PtbalZH/F");
treeWithBDT->Branch("EventPt",&EventPt, "EventPt/F");
treeWithBDT->Branch("AngleEMU",&AngleEMU, "AngleEMU/F");
treeWithBDT->Branch("Centrality",&Centrality, "Centrality/F");
treeWithBDT->Branch("MET",&MET, "MET/F");
treeWithBDT->Branch("EvntShpAplanarity",&EvntShpAplanarity, "EvntShpAplanarity/F");
treeWithBDT->Branch("EvntShpSphericity",&EvntShpSphericity, "EvntShpSphericity/F");
treeWithBDT->Branch("EvntShpIsotropy",&EvntShpIsotropy, "EvntShpIsotropy/F");
treeWithBDT->Branch("Zphi",&Zphi, "Zphi/F");
treeWithBDT->Branch("Hphi",&Hphi, "Hphi/F");
treeWithBDT->Branch("Mte",&Mte, "Mte/F");
treeWithBDT->Branch("Mtmu",&Mtmu, "Mtmu/F");
treeWithBDT->Branch("dPhiHMET",&dPhiHMET, "dPhiHMET/F");
treeWithBDT->Branch("Dphiemu",&Dphiemu, "Dphiemu/F");
treeWithBDT->Branch("delRjj",&delRjj, "delRjj/F");
treeWithBDT->Branch("delRemu",&delRemu, "delRemu/F");
treeWithBDT->Branch("DphiZMET",&DphiZMET, "DphiZMET/F");
treeWithBDT->Branch("DeltaPhijetMETmin",&DeltaPhijetMETmin, "DeltaPhijetMETmin/F");
treeWithBDT->Branch("BDTvalue",&BDTvalue, "BDTvalue/F");
treeWithBDT->Branch("AngleHemu",&AngleHemu, "AngleHemu/F");
treeWithBDT->Branch("ProjVisT",&ProjVisT, "ProjVisT/F");
treeWithBDT->Branch("topMass",&topMass, "topMass/F");
treeWithBDT->Branch("topPt",&topPt, "topPt/F");
treeWithBDT->Branch("ZmassSVD",&ZmassSVD, "ZmassSVD/F");
treeWithBDT->Branch("ZmassSVDnegSol",&ZmassSVDnegSol, "ZmassSVDnegSol/F");
treeWithBDT->Branch("Zmass",&Zmass, "Zmass/F");
treeWithBDT->Branch("ZmassNegInclu",&ZmassNegInclu, "ZmassNegInclu/F");
treeWithBDT->Branch("topMass",&topMass, "topMass/F");
treeWithBDT->Branch("Mt",&Mt, "Mt/F");
UInt_t nbin = 15;
TH1F *histLk(0), *histLkD(0), *histLkPCA(0), *histLkKDE(0), *histLkMIX(0), *histPD(0), *histPDD(0);
TH1F *histPDPCA(0), *histPDEFoam(0), *histPDEFoamErr(0), *histPDEFoamSig(0), *histKNN(0), *histHm(0);
TH1F *histFi(0), *histFiG(0), *histFiB(0), *histLD(0), *histNn(0),*histNnbfgs(0),*histNnbnn(0);
TH1F *histNnC(0), *histNnT(0), *histBdt(0), *histBdtG(0), *histBdtD(0), *histRf(0), *histSVMG(0);
TH1F *histSVMP(0), *histSVML(0), *histFDAMT(0), *histFDAGA(0), *histCat(0), *histPBdt(0);
if (Use["Likelihood"]) histLk = new TH1F( "MVA_Likelihood", "MVA_Likelihood", nbin, -1, 1 );
if (Use["LikelihoodD"]) histLkD = new TH1F( "MVA_LikelihoodD", "MVA_LikelihoodD", nbin, -1, 0.9999 );
if (Use["LikelihoodPCA"]) histLkPCA = new TH1F( "MVA_LikelihoodPCA", "MVA_LikelihoodPCA", nbin, -1, 1 );
if (Use["LikelihoodKDE"]) histLkKDE = new TH1F( "MVA_LikelihoodKDE", "MVA_LikelihoodKDE", nbin, -0.00001, 0.99999 );
if (Use["LikelihoodMIX"]) histLkMIX = new TH1F( "MVA_LikelihoodMIX", "MVA_LikelihoodMIX", nbin, 0, 1 );
if (Use["PDERS"]) histPD = new TH1F( "MVA_PDERS", "MVA_PDERS", nbin, 0, 1 );
if (Use["PDERSD"]) histPDD = new TH1F( "MVA_PDERSD", "MVA_PDERSD", nbin, 0, 1 );
if (Use["PDERSPCA"]) histPDPCA = new TH1F( "MVA_PDERSPCA", "MVA_PDERSPCA", nbin, 0, 1 );
if (Use["KNN"]) histKNN = new TH1F( "MVA_KNN", "MVA_KNN", nbin, 0, 1 );
if (Use["HMatrix"]) histHm = new TH1F( "MVA_HMatrix", "MVA_HMatrix", nbin, -0.95, 1.55 );
if (Use["Fisher"]) histFi = new TH1F( "MVA_Fisher", "MVA_Fisher", nbin, -4, 4 );
if (Use["FisherG"]) histFiG = new TH1F( "MVA_FisherG", "MVA_FisherG", nbin, -1, 1 );
if (Use["BoostedFisher"]) histFiB = new TH1F( "MVA_BoostedFisher", "MVA_BoostedFisher", nbin, -2, 2 );
if (Use["LD"]) histLD = new TH1F( "MVA_LD", "MVA_LD", nbin, -2, 2 );
if (Use["MLP"]) histNn = new TH1F( "MVA_MLP", "MVA_MLP", nbin, -1.25, 1.5 );
if (Use["MLPBFGS"]) histNnbfgs = new TH1F( "MVA_MLPBFGS", "MVA_MLPBFGS", nbin, -1.25, 1.5 );
if (Use["MLPBNN"]) histNnbnn = new TH1F( "MVA_MLPBNN", "MVA_MLPBNN", nbin, -1.25, 1.5 );
if (Use["CFMlpANN"]) histNnC = new TH1F( "MVA_CFMlpANN", "MVA_CFMlpANN", nbin, 0, 1 );
if (Use["TMlpANN"]) histNnT = new TH1F( "MVA_TMlpANN", "MVA_TMlpANN", nbin, -1.3, 1.3 );
if (Use["BDT"]) {
histMattBdt = new TH1F( "Matt_BDT", "Matt_BDT", 15, -1.1, 0.35 );
histTMVABdt = new TH1F( "TMVA_BDT", "TMVA_BDT", 36, -1.0, -0.1 );
}
if (Use["BDTD"]) histBdtD = new TH1F( "MVA_BDTD", "MVA_BDTD", nbin, -0.8, 0.8 );
if (Use["BDTG"]) histBdtG = new TH1F( "MVA_BDTG", "MVA_BDTG", nbin, -1.0, 1.0 );
if (Use["RuleFit"]) histRf = new TH1F( "MVA_RuleFit", "MVA_RuleFit", nbin, -2.0, 2.0 );
if (Use["SVM_Gauss"]) histSVMG = new TH1F( "MVA_SVM_Gauss", "MVA_SVM_Gauss", nbin, 0.0, 1.0 );
if (Use["SVM_Poly"]) histSVMP = new TH1F( "MVA_SVM_Poly", "MVA_SVM_Poly", nbin, 0.0, 1.0 );
if (Use["SVM_Lin"]) histSVML = new TH1F( "MVA_SVM_Lin", "MVA_SVM_Lin", nbin, 0.0, 1.0 );
if (Use["FDA_MT"]) histFDAMT = new TH1F( "MVA_FDA_MT", "MVA_FDA_MT", nbin, -2.0, 3.0 );
if (Use["FDA_GA"]) histFDAGA = new TH1F( "MVA_FDA_GA", "MVA_FDA_GA", nbin, -2.0, 3.0 );
if (Use["Category"]) histCat = new TH1F( "MVA_Category", "MVA_Category", nbin, -2., 2. );
if (Use["Plugin"]) histPBdt = new TH1F( "MVA_PBDT", "MVA_BDT", nbin, -0.8, 0.8 );
// PDEFoam also returns per-event error, fill in histogram, and also fill significance
if (Use["PDEFoam"]) {
histPDEFoam = new TH1F( "MVA_PDEFoam", "MVA_PDEFoam", nbin, 0, 1 );
histPDEFoamErr = new TH1F( "MVA_PDEFoamErr", "MVA_PDEFoam error", nbin, 0, 1 );
histPDEFoamSig = new TH1F( "MVA_PDEFoamSig", "MVA_PDEFoam significance", nbin, 0, 10 );
}
// Book example histogram for probability (the other methods are done similarly)
TH1F *probHistFi(0), *rarityHistFi(0);
if (Use["Fisher"]) {
probHistFi = new TH1F( "MVA_Fisher_Proba", "MVA_Fisher_Proba", nbin, 0, 1 );
rarityHistFi = new TH1F( "MVA_Fisher_Rarity", "MVA_Fisher_Rarity", nbin, 0, 1 );
}
// Prepare input tree (this must be replaced by your data source)
// in this example, there is a toy tree with signal and one with background events
// we'll later on use only the "signal" events for the test in this example.
//
TFile *input(0);
TString fname = "/home/hep/wilken/taus/CMSSW_4_4_2_patch8/src/UserCode/wilken/V21/DY_PtZ.root";
double lumi = 4.457;
Double_t DY_PtZ_weight = lumi/(lumiZJH/2.0); //WW_TuneZ2_7TeV_pythia6_tauola
if (!gSystem->AccessPathName( fname ))
input = TFile::Open( fname ); // check if file in local directory exists
else
input = TFile::Open( "http://root.cern.ch/files/tmva_class_example.root" ); // if not: download from ROOT server
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
// --- Event loop
// Prepare the event tree
// - here the variable names have to corresponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
std::cout << "--- Select signal sample" << std::endl;
TTree* theTree = (TTree*)input->Get("BDT_btree");
theTree->SetBranchAddress( "Hmass", &Hmass );
theTree->SetBranchAddress( "Emumass", &Emumass );
theTree->SetBranchAddress( "Hpt", &Hpt );
theTree->SetBranchAddress( "Zpt", &Zpt );
theTree->SetBranchAddress( "CSV0", &CSV0 );
theTree->SetBranchAddress( "CSV1", &CSV1 );
theTree->SetBranchAddress( "DeltaPhiHV", &DeltaPhiHV );
theTree->SetBranchAddress( "DetaJJ", &DetaJJ );
theTree->SetBranchAddress( "UnweightedEta", &UnweightedEta );
theTree->SetBranchAddress( "lep0pt", &lep0pt );
theTree->SetBranchAddress( "Ht", &Ht );
theTree->SetBranchAddress( "EvntShpCircularity", &EvntShpCircularity );
theTree->SetBranchAddress( "nJets", &nJets );
theTree->SetBranchAddress( "naJets",&naJets);
theTree->SetBranchAddress( "nSV", &nSV );
theTree->SetBranchAddress( "lep1pt", &lep1pt );
theTree->SetBranchAddress( "lep0pt", &lep0pt );
theTree->SetBranchAddress( "EtaStandDev", &EtaStandDev );
theTree->SetBranchAddress( "UnweightedEta", &UnweightedEta );
theTree->SetBranchAddress( "jetCHF0", &jetCHF0 );
theTree->SetBranchAddress( "jetCHF1", &jetCHF1 );
theTree->SetBranchAddress( "lep_pfCombRelIso0", &lep_pfCombRelIso0 );
theTree->SetBranchAddress( "lep_pfCombRelIso1", &lep_pfCombRelIso1 );
theTree->SetBranchAddress( "DphiJJ", &DphiJJ );
theTree->SetBranchAddress( "RMS_eta", &RMS_eta );
theTree->SetBranchAddress( "PtbalZH", &PtbalZH );
theTree->SetBranchAddress( "EventPt", &EventPt );
theTree->SetBranchAddress( "AngleEMU", &AngleEMU );
theTree->SetBranchAddress( "Centrality", &Centrality );
theTree->SetBranchAddress( "EvntShpAplanarity", &EvntShpAplanarity );
theTree->SetBranchAddress( "EvntShpSphericity", &EvntShpSphericity );
theTree->SetBranchAddress( "EvntShpIsotropy", &EvntShpIsotropy );
theTree->SetBranchAddress( "Trigweight", &Trigweight );
theTree->SetBranchAddress( "B2011PUweight", &B2011PUweight );
theTree->SetBranchAddress( "A2011PUweight", &A2011PUweight );
theTree->SetBranchAddress( "btag2CSF", &btag2CSF );
theTree->SetBranchAddress( "MET", &MET );
theTree->SetBranchAddress( "Mte" , &Mte );
theTree->SetBranchAddress( "Mtmu" , &Mtmu );
theTree->SetBranchAddress( "dPhiHMET" , &dPhiHMET );
theTree->SetBranchAddress( "DeltaPhijetMETmin" , &DeltaPhijetMETmin );
theTree->SetBranchAddress( "delRjj" , &delRjj );
theTree->SetBranchAddress( "delRemu" , &delRemu );
theTree->SetBranchAddress( "DphiZMET" , &DphiZMET );
theTree->SetBranchAddress( "Dphiemu" , &Dphiemu );
theTree->SetBranchAddress( "DeltaPhijetMETmin" , &DeltaPhijetMETmin );
theTree->SetBranchAddress("AngleHemu",&AngleHemu);
theTree->SetBranchAddress("ProjVisT",&ProjVisT);
theTree->SetBranchAddress("topMass",&topMass);
theTree->SetBranchAddress("topPt",&topPt);
theTree->SetBranchAddress("ZmassSVD",&ZmassSVD);
theTree->SetBranchAddress("ZmassSVDnegSol",&ZmassSVDnegSol);
theTree->SetBranchAddress("Zmass",&Zmass);
theTree->SetBranchAddress("ZmassNegInclu",&ZmassNegInclu);
theTree->SetBranchAddress("topMass",&topMass);
theTree->SetBranchAddress("Mt",&Mt);
theTree->SetBranchAddress( "eventFlavor", &eventFlavor );
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
TTree* TMVATree = (TTree*)input->Get("TMVA_tree");
int NumInTree = theTree->GetEntries();
int NumberTMVAtree = TMVATree->GetEntries();
DY_PtZ_weight = DY_PtZ_weight* ( NumInTree /float(NumberTMVAtree)) ;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
int Nevents = 0, NpassBDT = 0;
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
Nevents++;
if (ievt%1000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
theTree->GetEntry(ievt);
// var1 = userVar1 + userVar2;
// var2 = userVar1 - userVar2;
// --- Return the MVA outputs and fill into histograms
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
if (Use["Likelihood" ]) histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) );
if (Use["LikelihoodD" ]) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) );
if (Use["LikelihoodPCA"]) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) );
if (Use["LikelihoodKDE"]) histLkKDE ->Fill( reader->EvaluateMVA( "LikelihoodKDE method" ) );
if (Use["LikelihoodMIX"]) histLkMIX ->Fill( reader->EvaluateMVA( "LikelihoodMIX method" ) );
if (Use["PDERS" ]) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) );
if (Use["PDERSD" ]) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) );
if (Use["PDERSPCA" ]) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) );
if (Use["KNN" ]) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) );
if (Use["HMatrix" ]) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) );
if (Use["Fisher" ]) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) );
if (Use["FisherG" ]) histFiG ->Fill( reader->EvaluateMVA( "FisherG method" ) );
if (Use["BoostedFisher"]) histFiB ->Fill( reader->EvaluateMVA( "BoostedFisher method" ) );
if (Use["LD" ]) histLD ->Fill( reader->EvaluateMVA( "LD method" ) );
if (Use["MLP" ]) histNn ->Fill( reader->EvaluateMVA( "MLP method" ) );
if (Use["MLPBFGS" ]) histNnbfgs ->Fill( reader->EvaluateMVA( "MLPBFGS method" ) );
if (Use["MLPBNN" ]) histNnbnn ->Fill( reader->EvaluateMVA( "MLPBNN method" ) );
if (Use["CFMlpANN" ]) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) );
if (Use["TMlpANN" ]) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) );
if (Use["BDT" ]) {
BDTvalue = reader->EvaluateMVA( "BDT method" );
histMattBdt ->Fill( BDTvalue,DY_PtZ_weight*Trigweight*B2011PUweight );
histTMVABdt ->Fill( BDTvalue,DY_PtZ_weight*Trigweight*B2011PUweight );
}
if (Use["BDTD" ]) histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) );
if (Use["BDTG" ]) histBdtG ->Fill( reader->EvaluateMVA( "BDTG method" ) );
if (Use["RuleFit" ]) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) );
if (Use["SVM_Gauss" ]) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) );
if (Use["SVM_Poly" ]) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) );
if (Use["SVM_Lin" ]) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) );
if (Use["FDA_MT" ]) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) );
if (Use["FDA_GA" ]) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) );
if (Use["Category" ]) histCat ->Fill( reader->EvaluateMVA( "Category method" ) );
if (Use["Plugin" ]) histPBdt ->Fill( reader->EvaluateMVA( "P_BDT method" ) );
// Retrieve also per-event error
if (Use["PDEFoam"]) {
Double_t val = reader->EvaluateMVA( "PDEFoam method" );
Double_t err = reader->GetMVAError();
histPDEFoam ->Fill( val );
histPDEFoamErr->Fill( err );
if (err>1.e-50) histPDEFoamSig->Fill( val/err );
}
// Retrieve probability instead of MVA output
if (Use["Fisher"]) {
probHistFi ->Fill( reader->GetProba ( "Fisher method" ) );
rarityHistFi->Fill( reader->GetRarity( "Fisher method" ) );
}
// std::cout << "Ht is "<< Ht << endl;
if(BDTvalue>-1.50){
NpassBDT++;
hMjj_OpenSelection->Fill(Hmass,DY_PtZ_weight*Trigweight*B2011PUweight );
hMmumu_OpenSelection->Fill(Emumass,DY_PtZ_weight*Trigweight*B2011PUweight );
hPtjj_OpenSelection->Fill(Hpt,DY_PtZ_weight*Trigweight*B2011PUweight );
hPtmumu_OpenSelection->Fill(Zpt,DY_PtZ_weight*Trigweight*B2011PUweight );
hCSV0_OpenSelection->Fill(CSV0,DY_PtZ_weight*Trigweight*B2011PUweight );
hCSV1_OpenSelection->Fill(CSV1,DY_PtZ_weight*Trigweight*B2011PUweight );
hdphiVH_OpenSelection->Fill(DeltaPhiHV,DY_PtZ_weight*Trigweight*B2011PUweight );
hdetaJJ_OpenSelection->Fill(DetaJJ,DY_PtZ_weight*Trigweight*B2011PUweight );
hUnweightedEta_OpenSelection->Fill(UnweightedEta,DY_PtZ_weight*Trigweight*B2011PUweight );
hPtmu0_OpenSelection->Fill(lep0pt,DY_PtZ_weight*Trigweight*B2011PUweight );
hHt_OpenSelection->Fill(Ht,DY_PtZ_weight*Trigweight*B2011PUweight );
hCircularity_OpenSelection->Fill(EvntShpCircularity,DY_PtZ_weight*Trigweight*B2011PUweight );
hCHFb0_OpenSelection->Fill(jetCHF0, DY_PtZ_weight*Trigweight*B2011PUweight );
hCHFb1_OpenSelection->Fill(jetCHF1, DY_PtZ_weight*Trigweight*B2011PUweight );
hPtbalZH_OpenSelection->Fill(PtbalZH, DY_PtZ_weight*Trigweight*B2011PUweight );
hPtmu1_OpenSelection->Fill(lep1pt, DY_PtZ_weight*Trigweight*B2011PUweight );
hPFRelIsomu0_OpenSelection->Fill(lep_pfCombRelIso0, DY_PtZ_weight*Trigweight*B2011PUweight );
hPFRelIsomu1_OpenSelection->Fill(lep_pfCombRelIso1, DY_PtZ_weight*Trigweight*B2011PUweight );
hNjets_OpenSelection->Fill(nJets, DY_PtZ_weight*Trigweight*B2011PUweight );
hRMSeta_OpenSelection->Fill(RMS_eta, DY_PtZ_weight*Trigweight*B2011PUweight );
hStaDeveta_OpenSelection->Fill(EtaStandDev, DY_PtZ_weight*Trigweight*B2011PUweight );
hdphiJJ_vect_OpenSelection->Fill(DphiJJ, DY_PtZ_weight*Trigweight*B2011PUweight );
hCentrality_OpenSelection->Fill(Centrality, DY_PtZ_weight*Trigweight*B2011PUweight );
hEventPt_OpenSelection->Fill(EventPt, DY_PtZ_weight*Trigweight*B2011PUweight );
hAngleEMU_OpenSelection->Fill(AngleEMU, DY_PtZ_weight*Trigweight*B2011PUweight );
hSphericity_OpenSelection->Fill(EvntShpSphericity, DY_PtZ_weight*Trigweight*B2011PUweight );
hAplanarity_OpenSelection->Fill(EvntShpAplanarity, DY_PtZ_weight*Trigweight*B2011PUweight );
hIsotropy_OpenSelection->Fill(EvntShpIsotropy, DY_PtZ_weight*Trigweight*B2011PUweight );
hDphiDetajj_OpenSelection->Fill(DphiJJ, DetaJJ, DY_PtZ_weight*Trigweight*B2011PUweight );
hMte_OpenSelection->Fill(Mte, DY_PtZ_weight*Trigweight*B2011PUweight );
hMtmu_OpenSelection->Fill(Mtmu, DY_PtZ_weight*Trigweight*B2011PUweight );
hdPhiHMET_OpenSelection->Fill(dPhiHMET, DY_PtZ_weight*Trigweight*B2011PUweight );
hDphiemu_OpenSelection->Fill(DeltaPhijetMETmin, DY_PtZ_weight*Trigweight*B2011PUweight );
hdelRjj_OpenSelection->Fill(delRjj, DY_PtZ_weight*Trigweight*B2011PUweight );
hdelRemu_OpenSelection->Fill(delRemu, DY_PtZ_weight*Trigweight*B2011PUweight );
hDphiZMET_OpenSelection->Fill(DphiZMET, DY_PtZ_weight*Trigweight*B2011PUweight );
hDphiemu_OpenSelection->Fill(Dphiemu, DY_PtZ_weight*Trigweight*B2011PUweight );
hDeltaPhijetMETmin_OpenSelection->Fill(DeltaPhijetMETmin, DY_PtZ_weight*Trigweight*B2011PUweight );
hAngleHemu_OpenSelection->Fill(AngleHemu, DY_PtZ_weight*Trigweight*B2011PUweight );
hProjVisT_OpenSelection->Fill(ProjVisT, DY_PtZ_weight*Trigweight*B2011PUweight );
htopMass_OpenSelection->Fill(topMass, DY_PtZ_weight*Trigweight*B2011PUweight );
htopPt_OpenSelection->Fill(topPt, DY_PtZ_weight*Trigweight*B2011PUweight );
hVMt_OpenSelection->Fill(Mt, DY_PtZ_weight*Trigweight*B2011PUweight );
hZmassSVD_OpenSelection->Fill(ZmassSVD, DY_PtZ_weight*Trigweight*B2011PUweight );
hZmassSVDnegSol_OpenSelection->Fill(ZmassSVDnegSol, DY_PtZ_weight*Trigweight*B2011PUweight );
hZmass_OpenSelection->Fill(Zmass, DY_PtZ_weight*Trigweight*B2011PUweight );
hZmassNegInclu_OpenSelection->Fill(ZmassNegInclu, DY_PtZ_weight*Trigweight*B2011PUweight );
}
treeWithBDT->Fill();
}//end event loop
// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
std::cout << "Number of Events: "<< Nevents << " Events passed BDT " << NpassBDT<< endl;
// Get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
// --- Write histograms
TFile *target = new TFile( "BDTCut_DYPtZ_HF.root","RECREATE" );
if (Use["Likelihood" ]) histLk ->Write();
if (Use["LikelihoodD" ]) histLkD ->Write();
if (Use["LikelihoodPCA"]) histLkPCA ->Write();
if (Use["LikelihoodKDE"]) histLkKDE ->Write();
if (Use["LikelihoodMIX"]) histLkMIX ->Write();
if (Use["PDERS" ]) histPD ->Write();
if (Use["PDERSD" ]) histPDD ->Write();
if (Use["PDERSPCA" ]) histPDPCA ->Write();
if (Use["KNN" ]) histKNN ->Write();
if (Use["HMatrix" ]) histHm ->Write();
if (Use["Fisher" ]) histFi ->Write();
if (Use["FisherG" ]) histFiG ->Write();
if (Use["BoostedFisher"]) histFiB ->Write();
if (Use["LD" ]) histLD ->Write();
if (Use["MLP" ]) histNn ->Write();
if (Use["MLPBFGS" ]) histNnbfgs ->Write();
if (Use["MLPBNN" ]) histNnbnn ->Write();
if (Use["CFMlpANN" ]) histNnC ->Write();
if (Use["TMlpANN" ]) histNnT ->Write();
if (Use["BDT" ]) {
histMattBdt ->Write();
histTMVABdt ->Write();
}
if (Use["BDTD" ]) histBdtD ->Write();
if (Use["BDTG" ]) histBdtG ->Write();
if (Use["RuleFit" ]) histRf ->Write();
if (Use["SVM_Gauss" ]) histSVMG ->Write();
if (Use["SVM_Poly" ]) histSVMP ->Write();
if (Use["SVM_Lin" ]) histSVML ->Write();
if (Use["FDA_MT" ]) histFDAMT ->Write();
if (Use["FDA_GA" ]) histFDAGA ->Write();
if (Use["Category" ]) histCat ->Write();
if (Use["Plugin" ]) histPBdt ->Write();
// Write also error and significance histos
if (Use["PDEFoam"]) { histPDEFoam->Write(); histPDEFoamErr->Write(); histPDEFoamSig->Write(); }
// Write also probability hists
if (Use["Fisher"]) { if (probHistFi != 0) probHistFi->Write(); if (rarityHistFi != 0) rarityHistFi->Write(); }
hCHFb0_OpenSelection->Write();
hCHFb1_OpenSelection->Write();
hPtbalZH_OpenSelection->Write();
hPtmu1_OpenSelection->Write();
hPFRelIsomu0_OpenSelection->Write();
hPFRelIsomu1_OpenSelection->Write();
hMjj_OpenSelection->Write();
hMmumu_OpenSelection->Write();
hPtjj_OpenSelection->Write();
hPtmumu_OpenSelection->Write();
hCSV0_OpenSelection->Write();
hCSV1_OpenSelection->Write();
hdphiVH_OpenSelection->Write();
hdetaJJ_OpenSelection->Write();
hUnweightedEta_OpenSelection->Write();
hPtmu0_OpenSelection->Write();
hCircularity_OpenSelection->Write();
hHt_OpenSelection->Write();
hNjets_OpenSelection->Write();
hRMSeta_OpenSelection->Write();
hStaDeveta_OpenSelection->Write();
hdphiJJ_vect_OpenSelection->Write();
hCentrality_OpenSelection->Write();
hEventPt_OpenSelection->Write();
hAngleEMU_OpenSelection->Write();
hSphericity_OpenSelection->Write();
hAplanarity_OpenSelection->Write();
hIsotropy_OpenSelection->Write();
hDphiDetajj_OpenSelection->Write();
hMtmu_OpenSelection->Write();
hMte_OpenSelection->Write();
hdPhiHMET_OpenSelection->Write();
hDphiemu_OpenSelection->Write();
hdelRjj_OpenSelection->Write();
hdelRemu_OpenSelection->Write();
hDphiZMET_OpenSelection->Write();
hDeltaPhijetMETmin_OpenSelection->Write();
hAngleHemu_OpenSelection->Write();
hProjVisT_OpenSelection->Write();
htopMass_OpenSelection->Write();
htopPt_OpenSelection->Write();
hVMt_OpenSelection->Write();
hZmassSVD_OpenSelection->Write();
hZmassSVDnegSol_OpenSelection->Write();
hZmass_OpenSelection->Write();
hZmassNegInclu_OpenSelection->Write();
treeWithBDT->Write();
target->Close();
delete reader;
hCHFb0_OpenSelection->Delete();
hCHFb1_OpenSelection->Delete();
hPtbalZH_OpenSelection->Delete();
hPtmu1_OpenSelection->Delete();
hPFRelIsomu0_OpenSelection->Delete();
hPFRelIsomu1_OpenSelection->Delete();
hMjj_OpenSelection->Delete();
hMmumu_OpenSelection->Delete();
hPtjj_OpenSelection->Delete();
hPtmumu_OpenSelection->Delete();
hCSV0_OpenSelection->Delete();
hCSV1_OpenSelection->Delete();
hdphiVH_OpenSelection->Delete();
hdetaJJ_OpenSelection->Delete();
hUnweightedEta_OpenSelection->Delete();
hPtmu0_OpenSelection->Delete();
hCircularity_OpenSelection->Delete();
hHt_OpenSelection->Delete();
hNjets_OpenSelection->Delete();
hRMSeta_OpenSelection->Delete();
hStaDeveta_OpenSelection->Delete();
hdphiJJ_vect_OpenSelection->Delete();
hCentrality_OpenSelection->Delete();
hEventPt_OpenSelection->Delete();
hAngleEMU_OpenSelection->Delete();
hSphericity_OpenSelection->Delete();
hAplanarity_OpenSelection->Delete();
hIsotropy_OpenSelection->Delete();
hDphiDetajj_OpenSelection->Delete();
hMtmu_OpenSelection->Delete();
hMte_OpenSelection->Delete();
hdPhiHMET_OpenSelection->Delete();
hDphiemu_OpenSelection->Delete();
hdelRjj_OpenSelection->Delete();
hdelRemu_OpenSelection->Delete();
hDphiZMET_OpenSelection->Delete();
hDeltaPhijetMETmin_OpenSelection->Delete();
hAngleHemu_OpenSelection->Delete();
hProjVisT_OpenSelection->Delete();
htopMass_OpenSelection->Delete();
htopPt_OpenSelection->Delete();
hVMt_OpenSelection->Delete();
hZmassSVD_OpenSelection->Delete();
hZmassSVDnegSol_OpenSelection->Delete();
hZmass_OpenSelection->Delete();
hZmassNegInclu_OpenSelection->Delete();
if (Use["BDT" ]) {
histMattBdt ->Delete();
histTMVABdt ->Delete();
}
treeWithBDT->Delete();
std::cout << "==> DYPtZ_HF_BDTCut is done!" << endl << std::endl;
gROOT->ProcessLine(".q");
}
void TMVAClassificationApplicationLambda( TString myMethodList = "" )
{
#ifdef __CINT__
gROOT->ProcessLine( ".O0" ); // turn off optimization in CINT
#endif
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
// --- Cut optimisation
Use["Cuts"] = 1;
Use["CutsD"] = 1;
Use["CutsPCA"] = 0;
Use["CutsGA"] = 0;
Use["CutsSA"] = 0;
//
// --- 1-dimensional likelihood ("naive Bayes estimator")
Use["Likelihood"] = 1;
Use["LikelihoodD"] = 0; // the "D" extension indicates decorrelated input variables (see option strings)
Use["LikelihoodPCA"] = 1; // the "PCA" extension indicates PCA-transformed input variables (see option strings)
Use["LikelihoodKDE"] = 0;
Use["LikelihoodMIX"] = 0;
//
// --- Mutidimensional likelihood and Nearest-Neighbour methods
Use["PDERS"] = 1;
Use["PDERSD"] = 0;
Use["PDERSPCA"] = 0;
Use["PDEFoam"] = 1;
Use["PDEFoamBoost"] = 0; // uses generalised MVA method boosting
Use["KNN"] = 1; // k-nearest neighbour method
//
// --- Linear Discriminant Analysis
Use["LD"] = 1; // Linear Discriminant identical to Fisher
Use["Fisher"] = 0;
Use["FisherG"] = 0;
Use["BoostedFisher"] = 0; // uses generalised MVA method boosting
Use["HMatrix"] = 0;
//
// --- Function Discriminant analysis
Use["FDA_GA"] = 1; // minimisation of user-defined function using Genetics Algorithm
Use["FDA_SA"] = 0;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 0;
Use["FDA_GAMT"] = 0;
Use["FDA_MCMT"] = 0;
//
// --- Neural Networks (all are feed-forward Multilayer Perceptrons)
Use["MLP"] = 0; // Recommended ANN
Use["MLPBFGS"] = 0; // Recommended ANN with optional training method
Use["MLPBNN"] = 1; // Recommended ANN with BFGS training method and bayesian regulator
Use["CFMlpANN"] = 0; // Depreciated ANN from ALEPH
Use["TMlpANN"] = 0; // ROOT's own ANN
//
// --- Support Vector Machine
Use["SVM"] = 1;
//
// --- Boosted Decision Trees
Use["BDT"] = 1; // uses Adaptive Boost
Use["BDTG"] = 0; // uses Gradient Boost
Use["BDTB"] = 0; // uses Bagging
Use["BDTD"] = 0; // decorrelation + Adaptive Boost
//
// --- Friedman's RuleFit method, ie, an optimised series of cuts ("rules")
Use["RuleFit"] = 1;
// ---------------------------------------------------------------
Use["Plugin"] = 0;
Use["Category"] = 0;
Use["SVM_Gauss"] = 0;
Use["SVM_Poly"] = 0;
Use["SVM_Lin"] = 0;
std::cout << std::endl;
std::cout << "==> Start TMVAClassificationApplication" << std::endl;
// Select methods (don't look at this code - not of interest)
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod
<< "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
std::cout << it->first << " ";
}
std::cout << std::endl;
return;
}
Use[regMethod] = 1;
}
}
// --------------------------------------------------------------------------------------------------
// --- Create the Reader object
TMVA::Reader *reader = new TMVA::Reader( "!Color:!Silent" );
// Create a set of variables and declare them to the reader
// - the variable names MUST corresponds in name and type to those given in the weight file(s) used
//Float_t var1, var2;
//Float_t var3, var4;
Float_t la_agl, la_dlos, la_dau1_dzos, la_dau1_dxyos, la_dau2_dzos, la_dau2_dxyos;
Float_t la_vtxChi2, la_dau1_nhit, la_dau2_nhit;
reader->AddVariable( "la_agl", &la_agl );
reader->AddVariable( "la_dlos", &la_dlos );
reader->AddVariable( "la_dau1_dzos", &la_dau1_dzos );
reader->AddVariable( "la_dau2_dzos",&la_dau2_dzos);
reader->AddVariable( "la_dau1_dxyos", &la_dau1_dxyos );
reader->AddVariable( "la_dau2_dxyos",&la_dau2_dxyos);
reader->AddVariable( "la_vtxChi2",&la_vtxChi2);
reader->AddVariable( "la_dau1_nhit",&la_dau1_nhit);
reader->AddVariable( "la_dau2_nhit",&la_dau2_nhit);
// Spectator variables declared in the training have to be added to the reader, too
Float_t la_mass;
reader->AddSpectator( "la_mass", &la_mass );
//reader->AddSpectator( "spec2 := var1*3", &spec2 );
/*
Float_t Category_cat1, Category_cat2, Category_cat3;
if (Use["Category"]){
// Add artificial spectators for distinguishing categories
reader->AddSpectator( "Category_cat1 := var3<=0", &Category_cat1 );
reader->AddSpectator( "Category_cat2 := (var3>0)&&(var4<0)", &Category_cat2 );
reader->AddSpectator( "Category_cat3 := (var3>0)&&(var4>=0)", &Category_cat3 );
}
*/
// --- Book the MVA methods
TString dir = "weights/";
TString prefix = "TMVAClassification";
// Book method(s)
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString methodName = TString(it->first) + TString(" method");
TString weightfile = dir + prefix + TString("_") + TString(it->first) + TString(".weights.xml");
reader->BookMVA( methodName, weightfile );
}
}
// Book output histograms
UInt_t nbin = 100;
TH1F *histLk(0), *histLkD(0), *histLkPCA(0), *histLkKDE(0), *histLkMIX(0), *histPD(0), *histPDD(0);
TH1F *histPDPCA(0), *histPDEFoam(0), *histPDEFoamErr(0), *histPDEFoamSig(0), *histKNN(0), *histHm(0);
TH1F *histFi(0), *histFiG(0), *histFiB(0), *histLD(0), *histNn(0),*histNnbfgs(0),*histNnbnn(0);
TH1F *histNnC(0), *histNnT(0), *histBdt(0), *histBdtG(0), *histBdtD(0), *histRf(0), *histSVMG(0);
TH1F *histSVMP(0), *histSVML(0), *histFDAMT(0), *histFDAGA(0), *histCat(0), *histPBdt(0);
if (Use["Likelihood"]) histLk = new TH1F( "MVA_Likelihood", "MVA_Likelihood", nbin, -1, 1 );
if (Use["LikelihoodD"]) histLkD = new TH1F( "MVA_LikelihoodD", "MVA_LikelihoodD", nbin, -1, 0.9999 );
if (Use["LikelihoodPCA"]) histLkPCA = new TH1F( "MVA_LikelihoodPCA", "MVA_LikelihoodPCA", nbin, -1, 1 );
if (Use["LikelihoodKDE"]) histLkKDE = new TH1F( "MVA_LikelihoodKDE", "MVA_LikelihoodKDE", nbin, -0.00001, 0.99999 );
if (Use["LikelihoodMIX"]) histLkMIX = new TH1F( "MVA_LikelihoodMIX", "MVA_LikelihoodMIX", nbin, 0, 1 );
if (Use["PDERS"]) histPD = new TH1F( "MVA_PDERS", "MVA_PDERS", nbin, 0, 1 );
if (Use["PDERSD"]) histPDD = new TH1F( "MVA_PDERSD", "MVA_PDERSD", nbin, 0, 1 );
if (Use["PDERSPCA"]) histPDPCA = new TH1F( "MVA_PDERSPCA", "MVA_PDERSPCA", nbin, 0, 1 );
if (Use["KNN"]) histKNN = new TH1F( "MVA_KNN", "MVA_KNN", nbin, 0, 1 );
if (Use["HMatrix"]) histHm = new TH1F( "MVA_HMatrix", "MVA_HMatrix", nbin, -0.95, 1.55 );
if (Use["Fisher"]) histFi = new TH1F( "MVA_Fisher", "MVA_Fisher", nbin, -4, 4 );
if (Use["FisherG"]) histFiG = new TH1F( "MVA_FisherG", "MVA_FisherG", nbin, -1, 1 );
if (Use["BoostedFisher"]) histFiB = new TH1F( "MVA_BoostedFisher", "MVA_BoostedFisher", nbin, -2, 2 );
if (Use["LD"]) histLD = new TH1F( "MVA_LD", "MVA_LD", nbin, -2, 2 );
if (Use["MLP"]) histNn = new TH1F( "MVA_MLP", "MVA_MLP", nbin, -1.25, 1.5 );
if (Use["MLPBFGS"]) histNnbfgs = new TH1F( "MVA_MLPBFGS", "MVA_MLPBFGS", nbin, -1.25, 1.5 );
if (Use["MLPBNN"]) histNnbnn = new TH1F( "MVA_MLPBNN", "MVA_MLPBNN", nbin, -1.25, 1.5 );
if (Use["CFMlpANN"]) histNnC = new TH1F( "MVA_CFMlpANN", "MVA_CFMlpANN", nbin, 0, 1 );
if (Use["TMlpANN"]) histNnT = new TH1F( "MVA_TMlpANN", "MVA_TMlpANN", nbin, -1.3, 1.3 );
if (Use["BDT"]) histBdt = new TH1F( "MVA_BDT", "MVA_BDT", nbin, -0.8, 0.8 );
if (Use["BDTD"]) histBdtD = new TH1F( "MVA_BDTD", "MVA_BDTD", nbin, -0.8, 0.8 );
if (Use["BDTG"]) histBdtG = new TH1F( "MVA_BDTG", "MVA_BDTG", nbin, -1.0, 1.0 );
if (Use["RuleFit"]) histRf = new TH1F( "MVA_RuleFit", "MVA_RuleFit", nbin, -2.0, 2.0 );
if (Use["SVM_Gauss"]) histSVMG = new TH1F( "MVA_SVM_Gauss", "MVA_SVM_Gauss", nbin, 0.0, 1.0 );
if (Use["SVM_Poly"]) histSVMP = new TH1F( "MVA_SVM_Poly", "MVA_SVM_Poly", nbin, 0.0, 1.0 );
if (Use["SVM_Lin"]) histSVML = new TH1F( "MVA_SVM_Lin", "MVA_SVM_Lin", nbin, 0.0, 1.0 );
if (Use["FDA_MT"]) histFDAMT = new TH1F( "MVA_FDA_MT", "MVA_FDA_MT", nbin, -2.0, 3.0 );
if (Use["FDA_GA"]) histFDAGA = new TH1F( "MVA_FDA_GA", "MVA_FDA_GA", nbin, -2.0, 3.0 );
if (Use["Category"]) histCat = new TH1F( "MVA_Category", "MVA_Category", nbin, -2., 2. );
if (Use["Plugin"]) histPBdt = new TH1F( "MVA_PBDT", "MVA_BDT", nbin, -0.8, 0.8 );
// PDEFoam also returns per-event error, fill in histogram, and also fill significance
if (Use["PDEFoam"]) {
histPDEFoam = new TH1F( "MVA_PDEFoam", "MVA_PDEFoam", nbin, 0, 1 );
histPDEFoamErr = new TH1F( "MVA_PDEFoamErr", "MVA_PDEFoam error", nbin, 0, 1 );
histPDEFoamSig = new TH1F( "MVA_PDEFoamSig", "MVA_PDEFoam significance", nbin, 0, 10 );
}
// Book example histogram for probability (the other methods are done similarly)
TH1F *probHistFi(0), *rarityHistFi(0);
if (Use["Fisher"]) {
probHistFi = new TH1F( "MVA_Fisher_Proba", "MVA_Fisher_Proba", nbin, 0, 1 );
rarityHistFi = new TH1F( "MVA_Fisher_Rarity", "MVA_Fisher_Rarity", nbin, 0, 1 );
}
// Prepare input tree (this must be replaced by your data source)
// in this example, there is a toy tree with signal and one with background events
// we'll later on use only the "signal" events for the test in this example.
//
TFile *input(0);
TString fname = "./tmva_example.root";
if (!gSystem->AccessPathName( fname ))
input = TFile::Open( fname ); // check if file in local directory exists
else
input = TFile::Open( "~/2014Research/ROOT_file/V0reco_PbPb/MCPbPb_central1.root" ); // if not: download from ROOT server
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
// --- Event loop
// Prepare the event tree
// - here the variable names have to corresponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
std::cout << "--- Select signal sample" << std::endl;
TTree* theTree = (TTree*)input->Get("ana/v0_Lambda");
//Float_t userVar1, userVar2;
theTree->SetBranchAddress( "la_agl", &la_agl );
theTree->SetBranchAddress( "la_dlos", &la_dlos );
theTree->SetBranchAddress( "la_dau1_dzos", &la_dau1_dzos );
theTree->SetBranchAddress( "la_dau1_dxyos", &la_dau1_dxyos );
theTree->SetBranchAddress( "la_dau2_dzos",&la_dau2_dzos);
theTree->SetBranchAddress( "la_dau2_dxyos",&la_dau2_dxyos);
theTree->SetBranchAddress( "la_vtxChi2",&la_vtxChi2);
theTree->SetBranchAddress( "la_dau1_nhit",&la_dau1_nhit);
theTree->SetBranchAddress( "la_dau2_nhit",&la_dau2_nhit);
theTree->SetBranchAddress( "la_mass", &la_mass );
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
TFile *target = new TFile( "TMVAppLambda.root","RECREATE" );
TNtuple *n1 = new TNtuple( "n1","n1","Lam_mass:MVA");
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
if (ievt%1000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
theTree->GetEntry(ievt);
// --- Return the MVA outputs and fill into histograms
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
if (Use["Likelihood" ]) histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) );
if (Use["LikelihoodD" ]) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) );
if (Use["LikelihoodPCA"]) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) );
if (Use["LikelihoodKDE"]) histLkKDE ->Fill( reader->EvaluateMVA( "LikelihoodKDE method" ) );
if (Use["LikelihoodMIX"]) histLkMIX ->Fill( reader->EvaluateMVA( "LikelihoodMIX method" ) );
if (Use["PDERS" ]) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) );
if (Use["PDERSD" ]) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) );
if (Use["PDERSPCA" ]) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) );
if (Use["KNN" ]) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) );
if (Use["HMatrix" ]) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) );
if (Use["Fisher" ]) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) );
if (Use["FisherG" ]) histFiG ->Fill( reader->EvaluateMVA( "FisherG method" ) );
if (Use["BoostedFisher"]) histFiB ->Fill( reader->EvaluateMVA( "BoostedFisher method" ) );
if (Use["LD" ]) histLD ->Fill( reader->EvaluateMVA( "LD method" ) );
if (Use["MLP" ]) histNn ->Fill( reader->EvaluateMVA( "MLP method" ) );
if (Use["MLPBFGS" ]) histNnbfgs ->Fill( reader->EvaluateMVA( "MLPBFGS method" ) );
if (Use["MLPBNN" ]) histNnbnn ->Fill( reader->EvaluateMVA( "MLPBNN method" ) );
if (Use["CFMlpANN" ]) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) );
if (Use["TMlpANN" ]) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) );
if (Use["BDT" ]) {
double Lam_mass = la_mass;
histBdt ->Fill( reader->EvaluateMVA( "BDT method" ) );
double MVA = 0.0;
MVA = reader->EvaluateMVA("BDT method");
n1->Fill(Lam_mass,MVA);
// cout << "la mass: " << temp << endl
}
if (Use["BDTD" ]) {
double Lam_mass = la_mass;
histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) );
double MVA = 0.0;
MVA = reader->EvaluateMVA("BDTD method");
n1->Fill(Lam_mass,MVA);
}
if (Use["BDTG" ]) {
double Lam_mass = la_mass;
histBdtG ->Fill( reader->EvaluateMVA( "BDTG method" ) );
double MVA = 0.0;
MVA = reader->EvaluateMVA("BDTG method");
n1->Fill(Lam_mass,MVA);
}
if (Use["RuleFit" ]) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) );
if (Use["SVM_Gauss" ]) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) );
if (Use["SVM_Poly" ]) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) );
if (Use["SVM_Lin" ]) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) );
if (Use["FDA_MT" ]) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) );
if (Use["FDA_GA" ]) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) );
if (Use["Category" ]) histCat ->Fill( reader->EvaluateMVA( "Category method" ) );
if (Use["Plugin" ]) histPBdt ->Fill( reader->EvaluateMVA( "P_BDT method" ) );
// Retrieve also per-event error
if (Use["PDEFoam"]) {
Double_t val = reader->EvaluateMVA( "PDEFoam method" );
Double_t err = reader->GetMVAError();
histPDEFoam ->Fill( val );
histPDEFoamErr->Fill( err );
if (err>1.e-50) histPDEFoamSig->Fill( val/err );
}
// Retrieve probability instead of MVA output
if (Use["Fisher"]) {
probHistFi ->Fill( reader->GetProba ( "Fisher method" ) );
rarityHistFi->Fill( reader->GetRarity( "Fisher method" ) );
}
}
// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// Get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
// --- Write histograms
if (Use["Likelihood" ]) histLk ->Write();
if (Use["LikelihoodD" ]) histLkD ->Write();
if (Use["LikelihoodPCA"]) histLkPCA ->Write();
if (Use["LikelihoodKDE"]) histLkKDE ->Write();
if (Use["LikelihoodMIX"]) histLkMIX ->Write();
if (Use["PDERS" ]) histPD ->Write();
if (Use["PDERSD" ]) histPDD ->Write();
if (Use["PDERSPCA" ]) histPDPCA ->Write();
if (Use["KNN" ]) histKNN ->Write();
if (Use["HMatrix" ]) histHm ->Write();
if (Use["Fisher" ]) histFi ->Write();
if (Use["FisherG" ]) histFiG ->Write();
if (Use["BoostedFisher"]) histFiB ->Write();
if (Use["LD" ]) histLD ->Write();
if (Use["MLP" ]) histNn ->Write();
if (Use["MLPBFGS" ]) histNnbfgs ->Write();
if (Use["MLPBNN" ]) histNnbnn ->Write();
if (Use["CFMlpANN" ]) histNnC ->Write();
if (Use["TMlpANN" ]) histNnT ->Write();
if (Use["BDT" ]){
histBdt ->Write();
n1->Write();
}
if (Use["BDTD" ]){
histBdtD ->Write();
n1->Write();
}
if (Use["BDTG" ]) {
histBdtG ->Write();
n1->Write();
}
if (Use["RuleFit" ]) histRf ->Write();
if (Use["SVM_Gauss" ]) histSVMG ->Write();
if (Use["SVM_Poly" ]) histSVMP ->Write();
if (Use["SVM_Lin" ]) histSVML ->Write();
if (Use["FDA_MT" ]) histFDAMT ->Write();
if (Use["FDA_GA" ]) histFDAGA ->Write();
if (Use["Category" ]) histCat ->Write();
if (Use["Plugin" ]) histPBdt ->Write();
// Write also error and significance histos
if (Use["PDEFoam"]) { histPDEFoam->Write(); histPDEFoamErr->Write(); histPDEFoamSig->Write(); }
// Write also probability hists
if (Use["Fisher"]) { if (probHistFi != 0) probHistFi->Write(); if (rarityHistFi != 0) rarityHistFi->Write(); }
target->Close();
std::cout << "--- Created root file: \"TMVApp.root\" containing the MVA output histograms" << std::endl;
delete reader;
std::cout << "==> TMVAClassificationApplication is done!" << endl << std::endl;
}
void rezamyTMVAClassificationApplication1systematic( TString myMethodList = "" )
{
#ifdef __CINT__
gROOT->ProcessLine( ".O0" ); // turn off optimization in CINT
#endif
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
// --- Cut optimisation
Use["Cuts"] = 0;
Use["CutsD"] = 0;
Use["CutsPCA"] = 0;
Use["CutsGA"] = 0;
Use["CutsSA"] = 0;
//
// --- 1-dimensional likelihood ("naive Bayes estimator")
Use["Likelihood"] = 0;
Use["LikelihoodD"] = 0; // the "D" extension indicates decorrelated input variables (see option strings)
Use["LikelihoodPCA"] = 0; // the "PCA" extension indicates PCA-transformed input variables (see option strings)
Use["LikelihoodKDE"] = 0;
Use["LikelihoodMIX"] = 0;
//
// --- Mutidimensional likelihood and Nearest-Neighbour methods
Use["PDERS"] = 0;
Use["PDERSD"] = 0;
Use["PDERSPCA"] = 0;
Use["PDEFoam"] = 0;
Use["PDEFoamBoost"] = 0; // uses generalised MVA method boosting
Use["KNN"] = 0; // k-nearest neighbour method
//
// --- Linear Discriminant Analysis
Use["LD"] = 0; // Linear Discriminant identical to Fisher
Use["Fisher"] = 0;
Use["FisherG"] = 0;
Use["BoostedFisher"] = 0; // uses generalised MVA method boosting
Use["HMatrix"] = 0;
//
// --- Function Discriminant analysis
Use["FDA_GA"] = 0; // minimisation of user-defined function using Genetics Algorithm
Use["FDA_SA"] = 0;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 0;
Use["FDA_GAMT"] = 0;
Use["FDA_MCMT"] = 0;
//
// --- Neural Networks (all are feed-forward Multilayer Perceptrons)
Use["MLP"] = 0; // Recommended ANN
Use["MLPBFGS"] = 0; // Recommended ANN with optional training method
Use["MLPBNN"] = 0; // Recommended ANN with BFGS training method and bayesian regulator
Use["CFMlpANN"] = 0; // Depreciated ANN from ALEPH
Use["TMlpANN"] = 0; // ROOT's own ANN
//
// --- Support Vector Machine
Use["SVM"] = 0;
//
// --- Boosted Decision Trees
Use["BDT"] = 1; // uses Adaptive Boost
Use["BDTG"] = 0; // uses Gradient Boost
Use["BDTB"] = 0; // uses Bagging
Use["BDTD"] = 0; // decorrelation + Adaptive Boost
//
// --- Friedman's RuleFit method, ie, an optimised series of cuts ("rules")
Use["RuleFit"] = 0;
// ---------------------------------------------------------------
Use["Plugin"] = 0;
Use["Category"] = 0;
Use["SVM_Gauss"] = 0;
Use["SVM_Poly"] = 0;
Use["SVM_Lin"] = 0;
std::cout << std::endl;
std::cout << "==> Start TMVAClassificationApplication" << std::endl;
// Select methods (don't look at this code - not of interest)
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod
<< "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
std::cout << it->first << " ";
}
std::cout << std::endl;
return;
}
Use[regMethod] = 1;
}
}
// --------------------------------------------------------------------------------------------------
// --- Create the Reader object
TMVA::Reader *reader = new TMVA::Reader( "!Color:!Silent" );
Float_t ptphoton,etaphoton,ptmuon,etamuon,ptjet,etajet,masstop,mtw,deltaRphotonjet,deltaRphotonmuon,ht,costopphoton,deltaphiphotonmet,cvsdiscriminant;
Float_t jetmultiplicity,bjetmultiplicity,leptoncharge;
// Create a set of variables and declare them to the reader
// - the variable names MUST corresponds in name and type to those given in the weight file(s) used
reader->AddVariable ("ptphoton", &ptphoton);
// reader->AddVariable ("etaphoton", &etaphoton);
reader->AddVariable ("ptmuon", &ptmuon);
// reader->AddVariable ("etamuon", &etamuon);
reader->AddVariable ("ptjet", &ptjet);
// reader->AddVariable ("etajet", &etajet);
// reader->AddVariable ("masstop", &masstop);
// reader->AddVariable ("mtw", &mtw);
reader->AddVariable ("deltaRphotonjet", &deltaRphotonjet);
reader->AddVariable ("deltaRphotonmuon", &deltaRphotonmuon);
// reader->AddVariable ("ht", &ht);
// reader->AddVariable ("photonmuonmass", &photonmuonmass);
reader->AddVariable ("costopphoton", &costopphoton);
// reader->AddVariable ("topphotonmass", &topphotonmass);
//reader->AddVariable ("pttop", &pttop);
//reader->AddVariable ("etatop", &etatop);
reader->AddVariable ("jetmultiplicity", &jetmultiplicity);
// reader->AddVariable ("bjetmultiplicity", &bjetmultiplicity);
reader->AddVariable ("deltaphiphotonmet", &deltaphiphotonmet);
reader->AddVariable ("cvsdiscriminant", &cvsdiscriminant);
// reader->AddVariable ("leptoncharge", &leptoncharge);
/*
// Spectator variables declared in the training have to be added to the reader, too
reader->AddSpectator( "spec1 := var1*2", &spec1 );
reader->AddSpectator( "spec2 := var1*3", &spec2 );
Float_t Category_cat1, Category_cat2, Category_cat3;
if (Use["Category"]){
// Add artificial spectators for distinguishing categories
reader->AddSpectator( "Category_cat1 := var3<=0", &Category_cat1 );
reader->AddSpectator( "Category_cat2 := (var3>0)&&(var4<0)", &Category_cat2 );
reader->AddSpectator( "Category_cat3 := (var3>0)&&(var4>=0)", &Category_cat3 );
}
*/
// --- Book the MVA methods
TString dir = "weights/";
TString prefix = "TMVA";
// Book method(s)
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString methodName = TString(it->first) + TString(" method");
TString weightfile = dir + prefix + TString("_") + TString(it->first) + TString(".weights.xml");
reader->BookMVA( methodName, weightfile );
}
}
// Book output histograms
UInt_t nbin = 40;
TH1F *histLk(0), *histLkD(0), *histLkPCA(0), *histLkKDE(0), *histLkMIX(0), *histPD(0), *histPDD(0);
TH1F *histPDPCA(0), *histPDEFoam(0), *histPDEFoamErr(0), *histPDEFoamSig(0), *histKNN(0), *histHm(0);
TH1F *histFi(0), *histFiG(0), *histFiB(0), *histLD(0), *histNn(0),*histNnbfgs(0),*histNnbnn(0);
TH1F *histNnC(0), *histNnT(0), *histBdt(0), *histBdtG(0), *histBdtD(0), *histRf(0), *histSVMG(0);
TH1F *histSVMP(0), *histSVML(0), *histFDAMT(0), *histFDAGA(0), *histCat(0), *histPBdt(0);
if (Use["Likelihood"]) histLk = new TH1F( "MVA_Likelihood", "MVA_Likelihood", nbin, -1, 1 );
if (Use["LikelihoodD"]) histLkD = new TH1F( "MVA_LikelihoodD", "MVA_LikelihoodD", nbin, -1, 0.9999 );
if (Use["LikelihoodPCA"]) histLkPCA = new TH1F( "MVA_LikelihoodPCA", "MVA_LikelihoodPCA", nbin, -1, 1 );
if (Use["LikelihoodKDE"]) histLkKDE = new TH1F( "MVA_LikelihoodKDE", "MVA_LikelihoodKDE", nbin, -0.00001, 0.99999 );
if (Use["LikelihoodMIX"]) histLkMIX = new TH1F( "MVA_LikelihoodMIX", "MVA_LikelihoodMIX", nbin, 0, 1 );
if (Use["PDERS"]) histPD = new TH1F( "MVA_PDERS", "MVA_PDERS", nbin, 0, 1 );
if (Use["PDERSD"]) histPDD = new TH1F( "MVA_PDERSD", "MVA_PDERSD", nbin, 0, 1 );
if (Use["PDERSPCA"]) histPDPCA = new TH1F( "MVA_PDERSPCA", "MVA_PDERSPCA", nbin, 0, 1 );
if (Use["KNN"]) histKNN = new TH1F( "MVA_KNN", "MVA_KNN", nbin, 0, 1 );
if (Use["HMatrix"]) histHm = new TH1F( "MVA_HMatrix", "MVA_HMatrix", nbin, -0.95, 1.55 );
if (Use["Fisher"]) histFi = new TH1F( "MVA_Fisher", "MVA_Fisher", nbin, -4, 4 );
if (Use["FisherG"]) histFiG = new TH1F( "MVA_FisherG", "MVA_FisherG", nbin, -1, 1 );
if (Use["BoostedFisher"]) histFiB = new TH1F( "MVA_BoostedFisher", "MVA_BoostedFisher", nbin, -2, 2 );
if (Use["LD"]) histLD = new TH1F( "MVA_LD", "MVA_LD", nbin, -2, 2 );
if (Use["MLP"]) histNn = new TH1F( "MVA_MLP", "MVA_MLP", nbin, -1.25, 1.5 );
if (Use["MLPBFGS"]) histNnbfgs = new TH1F( "MVA_MLPBFGS", "MVA_MLPBFGS", nbin, -1.25, 1.5 );
if (Use["MLPBNN"]) histNnbnn = new TH1F( "MVA_MLPBNN", "MVA_MLPBNN", nbin, -1.25, 1.5 );
if (Use["CFMlpANN"]) histNnC = new TH1F( "MVA_CFMlpANN", "MVA_CFMlpANN", nbin, 0, 1 );
if (Use["TMlpANN"]) histNnT = new TH1F( "MVA_TMlpANN", "MVA_TMlpANN", nbin, -1.3, 1.3 );
if (Use["BDT"]) histBdt = new TH1F( "MVA_BDT", "MVA_BDT", nbin, -0.8, 0.8 );
if (Use["BDTD"]) histBdtD = new TH1F( "MVA_BDTD", "MVA_BDTD", nbin, -0.8, 0.8 );
if (Use["BDTG"]) histBdtG = new TH1F( "MVA_BDTG", "MVA_BDTG", nbin, -1.0, 1.0 );
if (Use["RuleFit"]) histRf = new TH1F( "MVA_RuleFit", "MVA_RuleFit", nbin, -2.0, 2.0 );
if (Use["SVM_Gauss"]) histSVMG = new TH1F( "MVA_SVM_Gauss", "MVA_SVM_Gauss", nbin, 0.0, 1.0 );
if (Use["SVM_Poly"]) histSVMP = new TH1F( "MVA_SVM_Poly", "MVA_SVM_Poly", nbin, 0.0, 1.0 );
if (Use["SVM_Lin"]) histSVML = new TH1F( "MVA_SVM_Lin", "MVA_SVM_Lin", nbin, 0.0, 1.0 );
if (Use["FDA_MT"]) histFDAMT = new TH1F( "MVA_FDA_MT", "MVA_FDA_MT", nbin, -2.0, 3.0 );
if (Use["FDA_GA"]) histFDAGA = new TH1F( "MVA_FDA_GA", "MVA_FDA_GA", nbin, -2.0, 3.0 );
if (Use["Category"]) histCat = new TH1F( "MVA_Category", "MVA_Category", nbin, -2., 2. );
if (Use["Plugin"]) histPBdt = new TH1F( "MVA_PBDT", "MVA_BDT", nbin, -0.8, 0.8 );
// PDEFoam also returns per-event error, fill in histogram, and also fill significance
if (Use["PDEFoam"]) {
histPDEFoam = new TH1F( "MVA_PDEFoam", "MVA_PDEFoam", nbin, 0, 1 );
histPDEFoamErr = new TH1F( "MVA_PDEFoamErr", "MVA_PDEFoam error", nbin, 0, 1 );
histPDEFoamSig = new TH1F( "MVA_PDEFoamSig", "MVA_PDEFoam significance", nbin, 0, 10 );
}
// Book example histogram for probability (the other methods are done similarly)
TH1F *probHistFi(0), *rarityHistFi(0);
if (Use["Fisher"]) {
probHistFi = new TH1F( "MVA_Fisher_Proba", "MVA_Fisher_Proba", nbin, 0, 1 );
rarityHistFi = new TH1F( "MVA_Fisher_Rarity", "MVA_Fisher_Rarity", nbin, 0, 1 );
}
// Prepare input tree (this must be replaced by your data source)
// in this example, there is a toy tree with signal and one with background events
// we'll later on use only the "signal" events for the test in this example.
//
TFile *input(0);
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
std::vector<string> samples_;
std::vector<string> datasamples_;
std::vector<TH1F*> datahists;
std::vector<TH1F*> revDATAhists;
float scales[] = {0.0978,1.491,0.0961,0.0253,0.0224,0.0145,0.0125,0.0160,0.0158,0.0341,0.0341,0.0341,0.020,0.0017,0.0055,0.0032,0.00084,0.02,19.145*0.0169,19.145*0.0169,19.145*0.0169,19.145*0.0169,19.145*0.0169};
//samples_.push_back("WJET.root");
samples_.push_back("ZJET.root");
samples_.push_back("PHJET200400.root");
samples_.push_back("WPHJET.root");
samples_.push_back("T-W-CH.root");
samples_.push_back("TBAR-W-CH.root");
samples_.push_back("T-S-CH.root");
samples_.push_back("TBAR-S-CH.root");
samples_.push_back("T-T-CH.root");
samples_.push_back("TBAR-T-CH.root");
samples_.push_back("TTBAR1.root");
samples_.push_back("TTBAR2.root");
samples_.push_back("TTBAR3.root");
samples_.push_back("TTG.root");
samples_.push_back("WWG.root");
samples_.push_back("WW.root");
samples_.push_back("WZ.root");
samples_.push_back("ZZ.root");
samples_.push_back("ZGAMMA.root");
samples_.push_back("SIGNAL.root");
samples_.push_back("SIGNAL.root");
samples_.push_back("SIGNAL.root");
samples_.push_back("SIGNAL.root");
samples_.push_back("SIGNAL.root");
datasamples_.push_back("REALDATA1.root");
datasamples_.push_back("REALDATA2.root");
datasamples_.push_back("REALDATA3.root");
std::vector<string> datasamplesreverse_;
datasamplesreverse_.push_back("etarev/REALDATA1.root");
datasamplesreverse_.push_back("etarev/REALDATA2.root");
datasamplesreverse_.push_back("etarev/REALDATA3.root");
std::vector<string> systematics;
//systematics.push_back("__JES__plus");
//systematics.push_back("__JES__minus");
//systematics.push_back("__JER__plus");
//systematics.push_back("__JER__minus");
systematics.push_back("__PU__plus");
systematics.push_back("__PU__minus");
systematics.push_back("__TRIG__plus");
systematics.push_back("__TRIG__minus");
systematics.push_back("__BTAG__plus");
systematics.push_back("__BTAG__minus");
systematics.push_back("__MISSTAG__plus");
systematics.push_back("__MISSTAG__minus");
systematics.push_back("__MUON__plus");
systematics.push_back("__MUON__minus");
systematics.push_back("__PHOTON__plus");
systematics.push_back("__PHOTON__minus");
systematics.push_back("");
map<string, double> eventwight;
TList* hList = new TList(); // list of histograms to store
std::vector<TFile*> files;
for(unsigned int idx=0; idx<samples_.size(); ++idx){
files.push_back(new TFile(samples_[idx].c_str()));
}
std::vector<TFile*> datafiles;
for(unsigned int idx=0; idx<datasamples_.size(); ++idx){
datafiles.push_back(new TFile(datasamples_[idx].c_str()));
}
for(unsigned int phi=0; phi<systematics.size(); ++phi){
std::vector<TH1F*> hists;
TH1F *wphjethist(0), *zjethist(0) , *phjethist(0), *wjethist(0), *twchhist(0), *tbarwhist(0), *tschhist(0), *tbarschhist(0), *ttchhist(0), *tbartchhist(0), *tt1hist(0) ,*tt2hist(0), *tt3hist(0), *ttphhist(0), *wwphhist(0), *wwhist(0), *wzhist(0), *zzhist(0), *zgammahist(0), *signalhist5(0) , *signalhist10(0), *signalhist20(0), *signalhist30(0), *signalhist40(0);
TH1F *data1histrev(0), *data2histrev(0) ,*data3histrev(0);
wphjethist = new TH1F( std::string("BDT__wphjethist").append(systematics[phi]).c_str(), std::string("BDT__wphjethist").append(systematics[phi]).c_str() , nbin, -1, 1 );
zjethist = new TH1F( std::string("BDT__zjethist").append(systematics[phi]).c_str(), std::string("BDT__zjethist").append(systematics[phi]).c_str(), nbin, -1, 1 );
phjethist = new TH1F( std::string("BDT__phjethist").append(systematics[phi]).c_str(),std::string("BDT__phjethist").append(systematics[phi]).c_str() , nbin, -1, 1 );
//wjethist = new TH1F( std::string("BDT__wjethist").append(systematics[phi]).c_str(),std::string("BDT__wjethist").append(systematics[phi]).c_str() , nbin, -0.8, 0.8 );
twchhist = new TH1F( std::string("BDT__twchhist").append(systematics[phi]).c_str(),std::string("BDT__twchhist").append(systematics[phi]).c_str() ,nbin, -1, 1 );
tbarwhist = new TH1F( std::string("BDT__tbarwhist").append(systematics[phi]).c_str(),std::string("BDT__tbarwhist").append(systematics[phi]).c_str() ,nbin, -1, 1 );
tschhist = new TH1F( std::string("BDT__tschhist").append(systematics[phi]).c_str(), std::string("BDT__tschhist").append(systematics[phi]).c_str(), nbin, -1, 1 );
tbarschhist = new TH1F( std::string("BDT__tbarschhist").append(systematics[phi]).c_str(),std::string("BDT__tbarschhist").append(systematics[phi]).c_str(), nbin, -1, 1 );
ttchhist = new TH1F( std::string("BDT__ttchhist").append(systematics[phi]).c_str(),std::string("BDT__ttchhist").append(systematics[phi]).c_str(), nbin, -1, 1 );
tbartchhist = new TH1F( std::string("BDT__tbartchhist").append(systematics[phi]).c_str(), std::string("BDT__tbartchhist").append(systematics[phi]).c_str(), nbin, -1, 1 );
tt1hist = new TH1F( std::string("BDT__tt1hist").append(systematics[phi]).c_str(), std::string("BDT__tt1hist").append(systematics[phi]).c_str(), nbin, -1, 1 );
tt2hist = new TH1F( std::string("BDT__tt2hist").append(systematics[phi]).c_str(), std::string("BDT__tt2hist").append(systematics[phi]).c_str(), nbin, -1, 1 );
tt3hist = new TH1F( std::string("BDT__tt3hist").append(systematics[phi]).c_str(),std::string("BDT__tt3hist").append(systematics[phi]).c_str() , nbin, -1, 1 );
ttphhist = new TH1F( std::string("BDT__ttphhist").append(systematics[phi]).c_str(),std::string("BDT__ttphhist").append(systematics[phi]).c_str() ,nbin, -1, 1 );
wwphhist = new TH1F( std::string("BDT__wwphhist").append(systematics[phi]).c_str(),std::string("BDT__wwphhist").append(systematics[phi]).c_str(), nbin, -1, 1 );
wwhist = new TH1F( std::string("BDT__wwhist").append(systematics[phi]).c_str(),std::string("BDT__wwhist").append(systematics[phi]).c_str() ,nbin, -1, 1 );
wzhist = new TH1F( std::string("BDT__wzhist").append(systematics[phi]).c_str(),std::string("BDT__wzhist").append(systematics[phi]).c_str(), nbin, -1, 1 );
zzhist = new TH1F( std::string("BDT__zzhist").append(systematics[phi]).c_str(),std::string("BDT__zzhist").append(systematics[phi]).c_str() ,nbin, -1, 1 );
zgammahist = new TH1F( std::string("BDT__zgammahist").append(systematics[phi]).c_str(),std::string("BDT__zgammahist").append(systematics[phi]).c_str() ,nbin, -1, 1 );
signalhist5 = new TH1F( std::string("BDT__signal5").append(systematics[phi]).c_str(),std::string("BDT__signal5").append(systematics[phi]).c_str() ,nbin, -1, 1 );
signalhist10 = new TH1F( std::string("BDT__signal10").append(systematics[phi]).c_str(),std::string("BDT__signal10").append(systematics[phi]).c_str() ,nbin, -1, 1 );
signalhist20 = new TH1F( std::string("BDT__signal20").append(systematics[phi]).c_str(),std::string("BDT__signal20").append(systematics[phi]).c_str() ,nbin, -1, 1 );
signalhist30 = new TH1F( std::string("BDT__signal30").append(systematics[phi]).c_str(),std::string("BDT__signal30").append(systematics[phi]).c_str() ,nbin, -1, 1 );
signalhist40 = new TH1F( std::string("BDT__signal40").append(systematics[phi]).c_str(),std::string("BDT__signal40").append(systematics[phi]).c_str() ,nbin, -1, 1 );
hists.push_back(zjethist);
hists.push_back(phjethist);
hists.push_back(wphjethist);
//hists.push_back(wjethist);
hists.push_back(twchhist);
hists.push_back(tbarwhist);
hists.push_back(tschhist);
hists.push_back(tbarschhist);
hists.push_back(ttchhist);
hists.push_back(tbartchhist);
hists.push_back(tt1hist);
hists.push_back(tt2hist);
hists.push_back(tt3hist);
hists.push_back(ttphhist);
hists.push_back(wwphhist);
hists.push_back(wwhist);
hists.push_back(wzhist);
hists.push_back(zzhist);
hists.push_back(zgammahist);
hists.push_back(signalhist5);
hists.push_back(signalhist10);
hists.push_back(signalhist20);
hists.push_back(signalhist30);
hists.push_back(signalhist40);
for(unsigned int idx=0; idx<samples_.size(); ++idx){
TFile *input(0);
TString fname =samples_[idx];
if (!gSystem->AccessPathName( fname )) input = TFile::Open( fname ); // check if file in local directory exists
else
input = TFile::Open( "http://root.cern.ch/files/tmva_class_example.root" ); // if not: download from ROOT server
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
// --- Event loop
// Prepare the event tree
// - here the variable names have to corres[1]ponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
//Double_t myptphoton,myetaphoton,myptmuon,myetamuon,myptjet,myetajet,mymasstop,mymtw,mydeltaRphotonjet,mydeltaRphotonmuon,myht,mycostopphoton,mydeltaphiphotonmet,mycvsdiscriminant,myjetmultiplicity,mybjetmultiplicity,myleptoncharge;
std::vector<double> *myptphoton=0;
std::vector<double> *myetaphoton=0;
std::vector<double> *myptmuon=0;
std::vector<double> *myetamuon=0;
std::vector<double> *myptjet=0;
std::vector<double> *myetajet=0;
std::vector<double> *mymasstop=0;
//std::vector<double> *mymtw=0;
std::vector<double> *mydeltaRphotonjet=0;
std::vector<double> *mydeltaRphotonmuon=0;
//std::vector<double> *myht=0;
std::vector<double> *mycostopphoton=0;
std::vector<double> *mydeltaphiphotonmet=0;
std::vector<double> *mycvsdiscriminant=0;
std::vector<double> *myjetmultiplicity=0;
//std::vector<double> *mybjetmultiplicity=0;
//std::vector<double> *myleptoncharge=0;
std::vector<double> *myweight=0;
std::vector<double> *mybtagSF=0;
std::vector<double> *mybtagSFup=0;
std::vector<double> *mybtagSFdown=0;
std::vector<double> *mymistagSFup=0;
std::vector<double> *mymistagSFdown=0;
std::vector<double> *mytriggerSF=0;
std::vector<double> *mytriggerSFup=0;
std::vector<double> *mytriggerSFdown=0;
std::vector<double> *myphotonSF=0;
std::vector<double> *myphotonSFup=0;
std::vector<double> *myphotonSFdown=0;
std::vector<double> *mypileupSF=0;
std::vector<double> *mypileupSFup=0;
std::vector<double> *mypileupSFdown=0;
std::vector<double> *mymuonSFup=0;
std::vector<double> *mymuonSFdown=0;
std::vector<double> *mymuonSF=0;
std::cout << "--- Select signal sample" << std::endl;
TTree* theTree = (TTree*)input->Get("analyzestep2/atq");
// Int_t myjetmultiplicity, mybjetmultiplicity , myleptoncharge;
// Float_t userVar1, userVar2;
theTree->SetBranchAddress("ptphoton", &myptphoton );
theTree->SetBranchAddress( "etaphoton", &myetaphoton );
theTree->SetBranchAddress( "ptmuon", &myptmuon );
theTree->SetBranchAddress( "etamuon", &myetamuon );
theTree->SetBranchAddress( "ptjet", &myptjet );
theTree->SetBranchAddress( "etajet", &myetajet );
theTree->SetBranchAddress( "masstop", &mymasstop );
// theTree->SetBranchAddress( "mtw", &mymtw );
theTree->SetBranchAddress( "deltaRphotonjet", &mydeltaRphotonjet );
theTree->SetBranchAddress( "deltaRphotonmuon", &mydeltaRphotonmuon );
// theTree->SetBranchAddress( "ht", &myht );
theTree->SetBranchAddress( "costopphoton", &mycostopphoton );
theTree->SetBranchAddress( "jetmultiplicity", &myjetmultiplicity );
// theTree->SetBranchAddress( "bjetmultiplicity", &mybjetmultiplicity );
theTree->SetBranchAddress( "deltaphiphotonmet", &mydeltaphiphotonmet );
theTree->SetBranchAddress( "cvsdiscriminant", &mycvsdiscriminant );
// theTree->SetBranchAddress( "leptoncharge", &myleptoncharge );
theTree->SetBranchAddress( "weight", &myweight);
theTree->SetBranchAddress( "btagSF", &mybtagSF);
theTree->SetBranchAddress( "btagSFup", &mybtagSFup);
theTree->SetBranchAddress( "btagSFdown", &mybtagSFdown);
theTree->SetBranchAddress( "mistagSFup", &mymistagSFup);
theTree->SetBranchAddress( "mistagSFdown", &mymistagSFdown);
theTree->SetBranchAddress( "triggerSF", &mytriggerSF);
theTree->SetBranchAddress( "triggerSFup", &mytriggerSFup);
theTree->SetBranchAddress( "triggerSFdown", &mytriggerSFdown);
theTree->SetBranchAddress( "photonSF", &myphotonSF);
theTree->SetBranchAddress( "photonSFup", &myphotonSFup);
theTree->SetBranchAddress( "photonSFdown", &myphotonSFdown);
theTree->SetBranchAddress( "muonSF", &mymuonSF);
theTree->SetBranchAddress( "muonSFup", &mymuonSFup);
theTree->SetBranchAddress( "muonSFdown", &mymuonSFdown);
theTree->SetBranchAddress( "pileupSF", &mypileupSF);
theTree->SetBranchAddress( "pileupSFup", &mypileupSFup);
theTree->SetBranchAddress( "pileupSFdown", &mypileupSFdown);
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
// std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
// std::cout << "--- ... Processing event: " << ievt << std::endl;
double finalweight;
if (ievt%1000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
theTree->GetEntry(ievt);
//for (int l=0;l<sizeof(myptphoton);l++){
//std::cout << "--- ... reza: " << myptphoton[l] <<std::endl;
//}
//std::cout << "--- ......................."<< (*mycvsdiscriminant)[0]<<std::endl;
// --- Return the MVA outputs and fill into histograms
ptphoton=(float)(*myptphoton)[0];
etaphoton=(float)(*myetaphoton )[0];
ptmuon=(float)(*myptmuon )[0];
etamuon=(float)(*myetamuon )[0];
ptjet=(float)(*myptjet )[0];
etajet=(float)(*myetajet )[0];
masstop=(float)(*mymasstop )[0];
//mtw=(float)(*mymtw )[0];
deltaRphotonjet=(float)(*mydeltaRphotonjet )[0];
deltaRphotonmuon=(float)(*mydeltaRphotonmuon )[0];
//ht=(float)(*myht )[0];
costopphoton=(float)(*mycostopphoton )[0];
jetmultiplicity=(float)(*myjetmultiplicity )[0];
//bjetmultiplicity=(float)(*mybjetmultiplicity )[0];
deltaphiphotonmet=(float)(*mydeltaphiphotonmet )[0];
cvsdiscriminant=(float)(*mycvsdiscriminant)[0];
//leptoncharge=(float)(*myleptoncharge )[0];
finalweight=(*myweight)[0];
//cout<<(*myweight)[0]<<endl;
eventwight["__PU__plus"]=(*myweight)[0]*(*mypileupSFup)[0]/(*mypileupSF)[0];
eventwight["__PU__minus"]=(*myweight)[0]*(*mypileupSFdown)[0]/(*mypileupSF)[0];
eventwight["__TRIG__plus"]=(*myweight)[0]*(*mytriggerSFup)[0]/(*mytriggerSF)[0];
eventwight["__TRIG__minus"]=(*myweight)[0]*(*mytriggerSFdown)[0]/(*mytriggerSF)[0];
eventwight["__BTAG__plus"]=(*myweight)[0]*(*mybtagSFup)[0]/(*mybtagSF)[0];
eventwight["__BTAG__minus"]=(*myweight)[0]*(*mybtagSFdown)[0]/(*mybtagSF)[0];
eventwight["__MISSTAG__plus"]=(*myweight)[0]*(*mymistagSFup)[0]/(*mybtagSF)[0];
eventwight["__MISSTAG__minus"]=(*myweight)[0]*(*mymistagSFdown)[0]/(*mybtagSF)[0];
eventwight["__MUON__plus"]=(*myweight)[0]*(*mymuonSFup)[0]/(*mymuonSF)[0];
eventwight["__MUON__minus"]=(*myweight)[0]*(*mymuonSFdown)[0]/(*mymuonSF)[0];
eventwight["__PHOTON__plus"]=(*myweight)[0]*(*myphotonSFup)[0]/(*myphotonSF)[0];
eventwight["__PHOTON__minus"]=(*myweight)[0]*(*myphotonSFdown)[0]/(*myphotonSF)[0];
eventwight[""]=(*myweight)[0];
finalweight=eventwight[systematics[phi].c_str()];
if (samples_[idx]=="SIGNAL.root") finalweight=(*myweight)[0];
//if (samples_[idx]=="WPHJET") finalweight=(*mypileupSF)[0]*(*mytriggerSF)[0]*(*mybtagSF)[0]*(*mymuonSF)[0]*(*myphotonSF)[0];
//if (finalweight<0) finalweight=30;
//cout<<"negative event weight"<<finalweight<<" "<<ptphoton<<endl;
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
hists[idx] ->Fill( reader->EvaluateMVA( "BDT method" ),finalweight* scales[idx] );
//cout<<reader->EvaluateMVA( "BDT method")<<" "<<finalweight<<endl;
//cout<<(*myweight)[0]<<endl;
// Retrieve also per-event error
if (Use["PDEFoam"]) {
Double_t val = reader->EvaluateMVA( "PDEFoam method" );
Double_t err = reader->GetMVAError();
histPDEFoam ->Fill( val );
histPDEFoamErr->Fill( err );
if (err>1.e-50) histPDEFoamSig->Fill( val/err );
}
// Retrieve probability instead of MVA output
if (Use["Fisher"]) {
probHistFi ->Fill( reader->GetProba ( "Fisher method" ) );
rarityHistFi->Fill( reader->GetRarity( "Fisher method" ) );
}
//delete finalweight;
}
// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// Get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
delete myptphoton;
delete myetaphoton;
delete myptmuon;
delete myetamuon;
delete myptjet;
delete myetajet;
delete mymasstop;
//delete mymtw;
delete mydeltaRphotonjet;
delete mydeltaRphotonmuon;
//delete myht;
delete mycostopphoton;
delete mydeltaphiphotonmet;
delete mycvsdiscriminant;
delete myjetmultiplicity;
//delete mybjetmultiplicity;
//delete myleptoncharge;
//delete myplot;
delete mybtagSF;
delete mybtagSFup;
delete mybtagSFdown;
delete mymistagSFup;
delete mytriggerSF;
delete mytriggerSFup;
delete mytriggerSFdown;
delete myphotonSF;
delete myphotonSFup;
delete myphotonSFdown;
delete mypileupSF;
delete mypileupSFup;
delete mypileupSFdown;
delete input;
if (idx==samples_.size()-5) hists[idx]->Scale(5/hists[idx]->Integral());
if (idx==samples_.size()-4) hists[idx]->Scale(10/hists[idx]->Integral());
if (idx==samples_.size()-3) hists[idx]->Scale(20/hists[idx]->Integral());
if (idx==samples_.size()-2) hists[idx]->Scale(30/hists[idx]->Integral());
if (idx==samples_.size()-1) hists[idx]->Scale(40/hists[idx]->Integral());
if (samples_[idx]=="WPHJET.root") hists[idx]->Scale(3173/hists[idx]->Integral());
if (samples_[idx]=="TTBAR2.root") hists[idx]->Add(hists[idx-1]);
if (samples_[idx]=="TTBAR3.root") hists[idx]->Add(hists[idx-1]);
if (!(samples_[idx]=="TTBAR1.root" || samples_[idx]=="TTBAR2.root")) hList->Add(hists[idx]);
}
}
TH1F *data1hist(0), *data2hist(0) ,*data3hist(0);
data1hist = new TH1F( "mu_BDT__data1hist", "mu_BDT__data1hist", nbin, -1, 1 );
data2hist = new TH1F( "mu_BDT__data2hist", "mu_BDT__data2hist", nbin, -1, 1 );
data3hist = new TH1F( "BDT__DATA", "BDT__DATA", nbin, -1, 1 );
datahists.push_back(data1hist);
datahists.push_back(data2hist);
datahists.push_back(data3hist);
for(unsigned int idx=0; idx<datasamples_.size(); ++idx){
TFile *input(0);
TString fname =datasamples_[idx];
if (!gSystem->AccessPathName( fname )) input = TFile::Open( fname ); // check if file in local directory exists
else
input = TFile::Open( "http://root.cern.ch/files/tmva_class_example.root" ); // if not: download from ROOT server
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
// --- Event loop
// Prepare the event tree
// - here the variable names have to corres[1]ponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
std::vector<double> *myptphoton=0;
std::vector<double> *myetaphoton=0;
std::vector<double> *myptmuon=0;
std::vector<double> *myetamuon=0;
std::vector<double> *myptjet=0;
std::vector<double> *myetajet=0;
std::vector<double> *mymasstop=0;
//std::vector<double> *mymtw=0;
std::vector<double> *mydeltaRphotonjet=0;
std::vector<double> *mydeltaRphotonmuon=0;
//std::vector<double> *myht=0;
std::vector<double> *mycostopphoton=0;
std::vector<double> *mydeltaphiphotonmet=0;
std::vector<double> *mycvsdiscriminant=0;
std::vector<double> *myjetmultiplicity=0;
//std::vector<double> *mybjetmultiplicity=0;
//std::vector<double> *myleptoncharge=0;
std::cout << "--- Select signal sample" << std::endl;
TTree* theTree = (TTree*)input->Get("analyzestep2/atq");
// Int_t myjetmultiplicity, mybjetmultiplicity , myleptoncharge;
// Float_t userVar1, userVar2;
theTree->SetBranchAddress("ptphoton", &myptphoton );
theTree->SetBranchAddress( "etaphoton", &myetaphoton );
theTree->SetBranchAddress( "ptmuon", &myptmuon );
theTree->SetBranchAddress( "etamuon", &myetamuon );
theTree->SetBranchAddress( "ptjet", &myptjet );
theTree->SetBranchAddress( "etajet", &myetajet );
theTree->SetBranchAddress( "masstop", &mymasstop );
// theTree->SetBranchAddress( "mtw", &mymtw );
theTree->SetBranchAddress( "deltaRphotonjet", &mydeltaRphotonjet );
theTree->SetBranchAddress( "deltaRphotonmuon", &mydeltaRphotonmuon );
// theTree->SetBranchAddress( "ht", &myht );
theTree->SetBranchAddress( "costopphoton", &mycostopphoton );
theTree->SetBranchAddress( "jetmultiplicity", &myjetmultiplicity );
// theTree->SetBranchAddress( "bjetmultiplicity", &mybjetmultiplicity );
theTree->SetBranchAddress( "deltaphiphotonmet", &mydeltaphiphotonmet );
theTree->SetBranchAddress( "cvsdiscriminant", &mycvsdiscriminant );
// theTree->SetBranchAddress( "leptoncharge", &myleptoncharge );
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
// std::cout << "--- ... Processing event: " << ievt << std::endl;
if (ievt%1000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
theTree->GetEntry(ievt);
//for (int l=0;l<sizeof(myptphoton);l++){
//std::cout << "--- ... reza: " << myptphoton[l] <<std::endl;
//}
//std::cout << "--- ......................."<< (*mycvsdiscriminant)[0]<<std::endl;
// --- Return the MVA outputs and fill into histograms
ptphoton=(float)(*myptphoton)[0];
etaphoton=(float)(*myetaphoton )[0];
ptmuon=(float)(*myptmuon )[0];
etamuon=(float)(*myetamuon )[0];
ptjet=(float)(*myptjet )[0];
etajet=(float)(*myetajet )[0];
masstop=(float)(*mymasstop )[0];
//mtw=(float)(*mymtw )[0];
deltaRphotonjet=(float)(*mydeltaRphotonjet )[0];
deltaRphotonmuon=(float)(*mydeltaRphotonmuon )[0];
//ht=(float)(*myht )[0];
costopphoton=(float)(*mycostopphoton )[0];
jetmultiplicity=(float)(*myjetmultiplicity )[0];
//bjetmultiplicity=(float)(*mybjetmultiplicity )[0];
deltaphiphotonmet=(float)(*mydeltaphiphotonmet )[0];
cvsdiscriminant=(float)(*mycvsdiscriminant)[0];
//leptoncharge=(float)(*myleptoncharge )[0];
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
if (Use["Likelihood" ]) histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) );
if (Use["LikelihoodD" ]) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) );
if (Use["LikelihoodPCA"]) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) );
if (Use["LikelihoodKDE"]) histLkKDE ->Fill( reader->EvaluateMVA( "LikelihoodKDE method" ) );
if (Use["LikelihoodMIX"]) histLkMIX ->Fill( reader->EvaluateMVA( "LikelihoodMIX method" ) );
if (Use["PDERS" ]) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) );
if (Use["PDERSD" ]) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) );
if (Use["PDERSPCA" ]) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) );
if (Use["KNN" ]) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) );
if (Use["HMatrix" ]) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) );
if (Use["Fisher" ]) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) );
if (Use["FisherG" ]) histFiG ->Fill( reader->EvaluateMVA( "FisherG method" ) );
if (Use["BoostedFisher"]) histFiB ->Fill( reader->EvaluateMVA( "BoostedFisher method" ) );
if (Use["LD" ]) histLD ->Fill( reader->EvaluateMVA( "LD method" ) );
if (Use["MLP" ]) histNn ->Fill( reader->EvaluateMVA( "MLP method" ) );
if (Use["MLPBFGS" ]) histNnbfgs ->Fill( reader->EvaluateMVA( "MLPBFGS method" ) );
if (Use["MLPBNN" ]) histNnbnn ->Fill( reader->EvaluateMVA( "MLPBNN method" ) );
if (Use["CFMlpANN" ]) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) );
if (Use["TMlpANN" ]) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) );
if (Use["BDT" ]) histBdt ->Fill( reader->EvaluateMVA( "BDT method" ) );
if (Use["BDTD" ]) histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) );
if (Use["BDTG" ]) histBdtG ->Fill( reader->EvaluateMVA( "BDTG method" ) );
if (Use["RuleFit" ]) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) );
if (Use["SVM_Gauss" ]) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) );
if (Use["SVM_Poly" ]) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) );
if (Use["SVM_Lin" ]) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) );
if (Use["FDA_MT" ]) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) );
if (Use["FDA_GA" ]) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) );
if (Use["Category" ]) histCat ->Fill( reader->EvaluateMVA( "Category method" ) );
if (Use["Plugin" ]) histPBdt ->Fill( reader->EvaluateMVA( "P_BDT method" ) );
datahists[idx] ->Fill( reader->EvaluateMVA( "BDT method" ) );
}
// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
delete myptphoton;
delete myetaphoton;
delete myptmuon;
delete myetamuon;
delete myptjet;
delete myetajet;
delete mymasstop;
//delete mymtw;
delete mydeltaRphotonjet;
delete mydeltaRphotonmuon;
//delete myht;
delete mycostopphoton;
delete mydeltaphiphotonmet;
delete mycvsdiscriminant;
delete myjetmultiplicity;
//delete mybjetmultiplicity;
//delete myleptoncharge;
//delete myplot;
delete input;
}
for(unsigned int idx=1; idx<datasamples_.size(); ++idx){
datahists[idx]->Add(datahists[idx-1]);
}
hList->Add(datahists[2]);
TH1F *data1histrev(0), *data2histrev(0) ,*data3histrev(0);
data1histrev = new TH1F( "BDT__data1histrev", "BDT__data1histrev", nbin, -1, 1 );
data2histrev = new TH1F( "BDT__data2histrev", "BDT__data2histrev", nbin, -1, 1 );
data3histrev = new TH1F( "BDT__wjet", "BDT__wjet", nbin, -1, 1 );
revDATAhists.push_back(data1histrev);
revDATAhists.push_back(data2histrev);
revDATAhists.push_back(data3histrev);
for(unsigned int idx=0; idx<datasamplesreverse_.size(); ++idx){
TFile *input(0);
TString fname =datasamplesreverse_[idx];
if (!gSystem->AccessPathName( fname )) input = TFile::Open( fname ); // check if file in local directory exists
else
input = TFile::Open( "http://root.cern.ch/files/tmva_class_example.root" ); // if not: download from ROOT server
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
// --- Event loop
//
// // Prepare the event tree
// // - here the variable names have to corres[1]ponds to your tree
// // - you can use the same variables as above which is slightly faster,
// // but of course you can use different ones and copy the values inside the event loop
// //
// Double_t myptphoton,myetaphoton,myptmuon,myetamuon,myptjet,myetajet,mymasstop,mymtw,mydeltaRphotonjet,mydeltaRphotonmuon,myht,mycostopphoton,mydeltaphiphotonmet,mycvsdiscriminant,myjetmultiplicity,mybjetmultiplicity,myleptoncharge;
std::vector<double> *myptphoton=0;
std::vector<double> *myetaphoton=0;
std::vector<double> *myptmuon=0;
std::vector<double> *myetamuon=0;
std::vector<double> *myptjet=0;
std::vector<double> *myetajet=0;
std::vector<double> *mymasstop=0;
//std::vector<double> *mymtw=0;
std::vector<double> *mydeltaRphotonjet=0;
std::vector<double> *mydeltaRphotonmuon=0;
//std::vector<double> *myht=0;
std::vector<double> *mycostopphoton=0;
std::vector<double> *mydeltaphiphotonmet=0;
std::vector<double> *mycvsdiscriminant=0;
std::vector<double> *myjetmultiplicity=0;
//std::vector<double> *mybjetmultiplicity=0;
//std::vector<double> *myleptoncharge=0;
TTree* theTree = (TTree*)input->Get("analyzestep2/atq");
theTree->SetBranchAddress("ptphoton", &myptphoton );
theTree->SetBranchAddress( "etaphoton", &myetaphoton );
theTree->SetBranchAddress( "ptmuon", &myptmuon );
theTree->SetBranchAddress( "etamuon", &myetamuon );
theTree->SetBranchAddress( "ptjet", &myptjet );
theTree->SetBranchAddress( "etajet", &myetajet );
theTree->SetBranchAddress( "masstop", &mymasstop );
// theTree->SetBranchAddress( "mtw", &mymtw );
theTree->SetBranchAddress( "deltaRphotonjet", &mydeltaRphotonjet );
theTree->SetBranchAddress( "deltaRphotonmuon", &mydeltaRphotonmuon );
// theTree->SetBranchAddress( "ht", &myht );
theTree->SetBranchAddress( "costopphoton", &mycostopphoton );
theTree->SetBranchAddress( "jetmultiplicity", &myjetmultiplicity );
// theTree->SetBranchAddress( "bjetmultiplicity", &mybjetmultiplicity );
theTree->SetBranchAddress( "deltaphiphotonmet", &mydeltaphiphotonmet );
theTree->SetBranchAddress( "cvsdiscriminant", &mycvsdiscriminant );
// theTree->SetBranchAddress( "leptoncharge", &myleptoncharge );
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
// std::cout << "--- ... Processing event: " << ievt << std::endl;
if (ievt%1000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
theTree->GetEntry(ievt);
//for (int l=0;l<sizeof(myptphoton);l++){
//std::cout << "--- ... reza: " << myptphoton[l] <<std::endl;
//}
// std::cout << "--- ......................."<< (*mycvsdiscriminant)[0]<<std::endl;
// --- Return the MVA outputs and fill into histograms
ptphoton=(float)(*myptphoton)[0];
etaphoton=(float)(*myetaphoton )[0];
ptmuon=(float)(*myptmuon )[0];
etamuon=(float)(*myetamuon )[0];
ptjet=(float)(*myptjet )[0];
etajet=(float)(*myetajet )[0];
masstop=(float)(*mymasstop )[0];
//mtw=(float)(*mymtw )[0];
deltaRphotonjet=(float)(*mydeltaRphotonjet )[0];
deltaRphotonmuon=(float)(*mydeltaRphotonmuon )[0];
//ht=(float)(*myht )[0];
costopphoton=(float)(*mycostopphoton )[0];
jetmultiplicity=(float)(*myjetmultiplicity )[0];
//bjetmultiplicity=(float)(*mybjetmultiplicity )[0];
deltaphiphotonmet=(float)(*mydeltaphiphotonmet )[0];
cvsdiscriminant=(float)(*mycvsdiscriminant)[0];
//leptoncharge=(float)(*myleptoncharge )[0];
if (Use["Likelihood" ]) histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) );
if (Use["LikelihoodD" ]) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) );
if (Use["LikelihoodPCA"]) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) );
if (Use["LikelihoodKDE"]) histLkKDE ->Fill( reader->EvaluateMVA( "LikelihoodKDE method" ) );
if (Use["LikelihoodMIX"]) histLkMIX ->Fill( reader->EvaluateMVA( "LikelihoodMIX method" ) );
if (Use["PDERS" ]) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) );
if (Use["PDERSD" ]) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) );
if (Use["PDERSPCA" ]) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) );
if (Use["KNN" ]) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) );
if (Use["HMatrix" ]) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) );
if (Use["Fisher" ]) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) );
if (Use["FisherG" ]) histFiG ->Fill( reader->EvaluateMVA( "FisherG method" ) );
if (Use["BoostedFisher"]) histFiB ->Fill( reader->EvaluateMVA( "BoostedFisher method" ) );
if (Use["LD" ]) histLD ->Fill( reader->EvaluateMVA( "LD method" ) );
if (Use["MLP" ]) histNn ->Fill( reader->EvaluateMVA( "MLP method" ) );
if (Use["MLPBFGS" ]) histNnbfgs ->Fill( reader->EvaluateMVA( "MLPBFGS method" ) );
if (Use["MLPBNN" ]) histNnbnn ->Fill( reader->EvaluateMVA( "MLPBNN method" ) );
if (Use["CFMlpANN" ]) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) );
if (Use["TMlpANN" ]) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) );
if (Use["BDT" ]) histBdt ->Fill( reader->EvaluateMVA( "BDT method" ) );
if (Use["BDTD" ]) histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) );
if (Use["BDTG" ]) histBdtG ->Fill( reader->EvaluateMVA( "BDTG method" ) );
if (Use["RuleFit" ]) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) );
if (Use["SVM_Gauss" ]) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) );
if (Use["SVM_Poly" ]) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) );
if (Use["SVM_Lin" ]) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) );
if (Use["FDA_MT" ]) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) );
if (Use["FDA_GA" ]) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) );
if (Use["Category" ]) histCat ->Fill( reader->EvaluateMVA( "Category method" ) );
if (Use["Plugin" ]) histPBdt ->Fill( reader->EvaluateMVA( "P_BDT method" ) );
revDATAhists[idx]->Fill( reader->EvaluateMVA( "BDT method" ) );}
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
delete myptphoton;
delete myetaphoton;
delete myptmuon;
delete myetamuon;
delete myptjet;
delete myetajet;
delete mymasstop;
//delete mymtw;
delete mydeltaRphotonjet;
delete mydeltaRphotonmuon;
//delete myht;
delete mycostopphoton;
delete mydeltaphiphotonmet;
delete mycvsdiscriminant;
delete myjetmultiplicity;
////delete mybjetmultiplicity;
////delete myleptoncharge;
////delete myplot;
//
delete input;
}
for(unsigned int idx=1; idx<datasamplesreverse_.size(); ++idx){
revDATAhists[idx]->Add(revDATAhists[idx-1]);
}
revDATAhists[2]->Scale(620.32/revDATAhists[2]->Integral());
hList->Add(revDATAhists[2]);
cout<<revDATAhists[2]->Integral()<<endl;
TFile *target = new TFile( "MVApp.root","RECREATE" );
hList->Write();
target->Close();
}
void TMVAClassification( std::string selectionName, std::string charge, TString myMethodList = "" )
{
// The explicit loading of the shared libTMVA is done in TMVAlogon.C, defined in .rootrc
// if you use your private .rootrc, or run from a different directory, please copy the
// corresponding lines from .rootrc
// methods to be processed can be given as an argument; use format:
//
// mylinux~> root -l TMVAClassification.C\(\"myMethod1,myMethod2,myMethod3\"\)
//
// if you like to use a method via the plugin mechanism, we recommend using
//
// mylinux~> root -l TMVAClassification.C\(\"P_myMethod\"\)
// (an example is given for using the BDT as plugin (see below),
// but of course the real application is when you write your own
// method based)
// this loads the library
TMVA::Tools::Instance();
//---------------------------------------------------------------
// default MVA methods to be trained + tested
std::map<std::string,int> Use;
Use["Cuts"] = 1;
Use["CutsD"] = 0;
Use["CutsPCA"] = 0;
Use["CutsGA"] = 0;
Use["CutsSA"] = 0;
// ---
Use["Likelihood"] = 0;
Use["LikelihoodD"] = 0; // the "D" extension indicates decorrelated input variables (see option strings)
Use["LikelihoodPCA"] = 0; // the "PCA" extension indicates PCA-transformed input variables (see option strings)
Use["LikelihoodKDE"] = 0;
Use["LikelihoodMIX"] = 0;
// ---
Use["PDERS"] = 0;
Use["PDERSD"] = 0;
Use["PDERSPCA"] = 0;
Use["PDERSkNN"] = 0; // depreciated until further notice
Use["PDEFoam"] = 0;
// --
Use["KNN"] = 0;
// ---
Use["HMatrix"] = 0;
Use["Fisher"] = 0;
Use["FisherG"] = 0;
Use["BoostedFisher"] = 0;
Use["LD"] = 0;
// ---
Use["FDA_GA"] = 0;
Use["FDA_SA"] = 0;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 0;
Use["FDA_GAMT"] = 0;
Use["FDA_MCMT"] = 0;
// ---
Use["MLP"] = 0; // this is the recommended ANN
Use["MLPBFGS"] = 0; // recommended ANN with optional training method
Use["CFMlpANN"] = 0; // *** missing
Use["TMlpANN"] = 0;
// ---
Use["SVM"] = 0;
// ---
Use["BDT"] = 0;
Use["BDTD"] = 0;
Use["BDTG"] = 0;
Use["BDTB"] = 0;
// ---
Use["RuleFit"] = 0;
// ---
Use["Plugin"] = 0;
// ---------------------------------------------------------------
std::cout << std::endl;
std::cout << "==> Start TMVAClassification" << std::endl;
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = TMVA::gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod << "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) std::cout << it->first << " ";
std::cout << std::endl;
return;
}
Use[regMethod] = 1;
}
}
// Create a new root output file.
TString outfileName( "TMVA.root" );
TFile* outputFile = TFile::Open( outfileName, "RECREATE" );
// Create the factory object. Later you can choose the methods
// whose performance you'd like to investigate. The factory will
// then run the performance analysis for you.
//
// The first argument is the base of the name of all the
// weightfiles in the directory weight/
//
// The second argument is the output file for the training results
// All TMVA output can be suppressed by removing the "!" (not) in
// front of the "Silent" argument in the option string
TMVA::Factory *factory = new TMVA::Factory( "TMVAClassification", outputFile,
"!V:!Silent:Color:DrawProgressBar:Transformations=I;D;P;G,D" );
// If you wish to modify default settings
// (please check "src/Config.h" to see all available global options)
// (TMVA::gConfig().GetVariablePlotting()).fTimesRMS = 8.0;
// (TMVA::gConfig().GetIONames()).fWeightFileDir = "myWeightDirectory";
// Define the input variables that shall be used for the MVA training
// note that you may also use variable expressions, such as: "3*var1/var2*abs(var3)"
// [all types of expressions that can also be parsed by TTree::Draw( "expression" )]
//factory->AddVariable( "MET" , "ME_{T}", "GeV", 'F');
//factory->AddVariable( "TMath::Max(pT1,pT2)" , "Lead Lepton p_{T}", "GeV", 'F');
factory->AddVariable( "HT" , "H_{T}", "GeV", 'F');
//factory->AddVariable( "M3" , "M_{3}", "GeV", 'F');
factory->AddVariable( "TMath::Min(pT1,pT2)" , "Sublead Lepton p_{T}", "GeV", 'F');
//factory->AddVariable( "NbJ" , "N B Jets", "", 'I');
//factory->AddVariable( "NbJmed" , "N B Jets (medium)", "", 'I');
//factory->AddVariable( "NJ" , "N Jets", "", 'I');
// You can add so-called "Spectator variables", which are not used in the MVA training,
// but will appear in the final "TestTree" produced by TMVA. This TestTree will contain the
// input variables, the response values of all trained MVAs, and the spectator variables
//factory->AddSpectator( "spec1:=var1*2", "Spectator 1", "units", 'F' );
//factory->AddSpectator( "spec2:=var1*3", "Spectator 2", "units", 'F' );
// read training and test data
if (ReadDataFromAsciiIFormat) {
// load the signal and background event samples from ascii files
// format in file must be:
// var1/F:var2/F:var3/F:var4/F
// 0.04551 0.59923 0.32400 -0.19170
// ...
TString datFileS = "tmva_example_sig.dat";
TString datFileB = "tmva_example_bkg.dat";
factory->SetInputTrees( datFileS, datFileB );
}
else {
// load the signal and background event samples from ROOT trees
TFile *input(0);
//TString fname = "../macros/tmva_example.root";
//TString fname = "opt_ttW_Apr10_Iso005_NoZVeto_jet20.root";
TString fname = "opt_ttW_Nov20_muDetIso0p05_elDetIso0p05_jet20_withZveto_optimization.root";
//TString fname = "opt_ttW_" + selectionName + ".root";
if (!gSystem->AccessPathName( fname )) {
input = TFile::Open( fname ); // check if file in local directory exists
}
else {
input = TFile::Open( "http://root.cern.ch/files/tmva_class_example.root" ); // if not: download from ROOT server
}
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassification : Using input file: " << input->GetName() << std::endl;
TTree* opt_tree = (TTree*)input->Get("tree_opt");
TFile* signalFile = TFile::Open("/shome/mdunser/workspace/CMSSW_5_2_5/src/DiLeptonAnalysis/NTupleProducer/macros/plots/Nov26_muPFIso0p05_elPFIso0p05_jet20_withZveto/TTbarW_Yields.root");
TTree *signal = (TTree*)signalFile->Get("SigEvents");
TChain* background = new TChain("SigEvents");
background->Add("/shome/mdunser/workspace/CMSSW_5_2_5/src/DiLeptonAnalysis/NTupleProducer/macros/plots/Nov26_muPFIso0p05_elPFIso0p05_jet20_withZveto/TTJets_Yields.root");
background->Add("/shome/mdunser/workspace/CMSSW_5_2_5/src/DiLeptonAnalysis/NTupleProducer/macros/plots/Nov26_muPFIso0p05_elPFIso0p05_jet20_withZveto/WZTo3LNu_Yields.root");
//TTree *background = (TTree*)opt_tree->CopyTree("SName==\"TTJets\" || SName==\"DYJets\" || SName==\"WZTo3LNu\"");
//TTree *background = (TTree*)opt_tree->CopyTree("SName==\"DYJets\"");
// global event weights per tree (see below for setting event-wise weights)
Double_t signalWeight = 1.0;
Double_t backgroundWeight = 1.0;
// ====== register trees ====================================================
//
// the following method is the prefered one:
// you can add an arbitrary number of signal or background trees
factory->AddSignalTree ( signal, signalWeight );
factory->AddBackgroundTree( background, backgroundWeight );
// To give different trees for training and testing, do as follows:
// factory->AddSignalTree( signalTrainingTree, signalTrainWeight, "Training" );
// factory->AddSignalTree( signalTestTree, signalTestWeight, "Test" );
// Use the following code instead of the above two or four lines to add signal and background
// training and test events "by hand"
// NOTE that in this case one should not give expressions (such as "var1+var2") in the input
// variable definition, but simply compute the expression before adding the event
//
// // --- begin ----------------------------------------------------------
// std::vector<Double_t> vars( 4 ); // vector has size of number of input variables
// Float_t treevars[4];
// for (Int_t ivar=0; ivar<4; ivar++) signal->SetBranchAddress( Form( "var%i", ivar+1 ), &(treevars[ivar]) );
// for (Int_t i=0; i<signal->GetEntries(); i++) {
// signal->GetEntry(i);
// for (Int_t ivar=0; ivar<4; ivar++) vars[ivar] = treevars[ivar];
// // add training and test events; here: first half is training, second is testing
// // note that the weight can also be event-wise
// if (i < signal->GetEntries()/2) factory->AddSignalTrainingEvent( vars, signalWeight );
// else factory->AddSignalTestEvent ( vars, signalWeight );
// }
//
// for (Int_t ivar=0; ivar<4; ivar++) background->SetBranchAddress( Form( "var%i", ivar+1 ), &(treevars[ivar]) );
// for (Int_t i=0; i<background->GetEntries(); i++) {
// background->GetEntry(i);
// for (Int_t ivar=0; ivar<4; ivar++) vars[ivar] = treevars[ivar];
// // add training and test events; here: first half is training, second is testing
// // note that the weight can also be event-wise
// if (i < background->GetEntries()/2) factory->AddBackgroundTrainingEvent( vars, backgroundWeight );
// else factory->AddBackgroundTestEvent ( vars, backgroundWeight );
// }
// // --- end ------------------------------------------------------------
//
// ====== end of register trees ==============================================
}
// This would set individual event weights (the variables defined in the
// expression need to exist in the original TTree)
// for signal : factory->SetSignalWeightExpression("weight1*weight2");
// for background: factory->SetBackgroundWeightExpression("weight1*weight2");
//factory->SetSignalWeightExpression("eventWeight");
factory->SetBackgroundWeightExpression("1./SLumi");
// Apply additional cuts on the signal and background samples (can be different)
TCut mycuts;
TCut mycutb;
if( charge == "plus" ) {
mycuts = "Charge==1 && SystFlag==0 && NJ>=3 && pT1>20. && pT2>20. && NbJmed>0 && TLCat==0 && Flavor<3 && PassZVeto==1"; // for example: TCut mycuts = "abs(var1)<0.5 && abs(var2-0.5)<1";
mycutb = "Charge==1 && SystFlag==0 && NJ>=3 && pT1>20. && pT2>20. && NbJmed>0 && TLCat==0 && Flavor<3 && PassZVeto==1"; // for example: TCut mycutb = "abs(var1)<0.5";
} else if( charge == "minus" ) {
mycuts = "Charge==-1 && SystFlag==0 && NJ>=3 && pT1>20. && pT2>20. && NbJmed>0 && TLCat==0 && Flavor<3 && PassZVeto==1"; // for example: TCut mycuts = "abs(var1)<0.5 && abs(var2-0.5)<1";
mycutb = "Charge==-1 && SystFlag==0 && NJ>=3 && pT1>20. && pT2>20. && NbJmed>0 && TLCat==0 && Flavor<3 && PassZVeto==1"; // for example: TCut mycutb = "abs(var1)<0.5";
} else if( charge == "all" ) {
mycuts = " SystFlag==0 && NJ>=3 && pT1>20. && pT2>20. && NbJmed>0 && TLCat==0 && Flavor<3 && PassZVeto==1"; // for example: TCut mycuts = "abs(var1)<0.5 && abs(var2-0.5)<1";
mycutb = " SystFlag==0 && NJ>=3 && pT1>20. && pT2>20. && NbJmed>0 && TLCat==0 && Flavor<3 && PassZVeto==1"; // for example: TCut mycutb = "abs(var1)<0.5";
} else {
std::cout << "only 'plus' and 'minus' and 'all' are allowed for charge." <<std::endl;
return;
}
//if( btagMed_presel_ ) {
// mycuts += "NbJmed>0"; // for example: TCut mycuts = "abs(var1)<0.5 && abs(var2-0.5)<1";
// mycutb += "NbJmed>0"; // for example: TCut mycutb = "abs(var1)<0.5";
//}
// tell the factory to use all remaining events in the trees after training for testing:
factory->PrepareTrainingAndTestTree( mycuts, mycutb,
"nTrain_Signal=0:nTrain_Background=0:SplitMode=Random:NormMode=NumEvents:!V" );
// If no numbers of events are given, half of the events in the tree are used for training, and
// the other half for testing:
// factory->PrepareTrainingAndTestTree( mycut, "SplitMode=random:!V" );
// To also specify the number of testing events, use:
// factory->PrepareTrainingAndTestTree( mycut,
// "NSigTrain=3000:NBkgTrain=3000:NSigTest=3000:NBkgTest=3000:SplitMode=Random:!V" );
// ---- Book MVA methods
//
// please lookup the various method configuration options in the corresponding cxx files, eg:
// src/MethoCuts.cxx, etc, or here: http://tmva.sourceforge.net/optionRef.html
// it is possible to preset ranges in the option string in which the cut optimisation should be done:
// "...:CutRangeMin[2]=-1:CutRangeMax[2]=1"...", where [2] is the third input variable
// Cut optimisation
if (Use["Cuts"]) {
std::string bookConditions;
bookConditions = "H:!V:FitMethod=MC";
bookConditions += ":VarProp[0]=FMax"; //HT
bookConditions += ":VarProp[1]=FMax"; //pt2
bookConditions += ":EffSel:SampleSize=500000000";
//bookConditions += ":EffSel:SampleSize=50000";
factory->BookMethod( TMVA::Types::kCuts, "Cuts", bookConditions.c_str() );
//factory->BookMethod( TMVA::Types::kCuts, "Cuts",
// "!H:!V:FitMethod=MC:EffSel:SampleSize=200000:VarProp=FSmart" );
}
if (Use["CutsD"])
factory->BookMethod( TMVA::Types::kCuts, "CutsD",
"!H:!V:FitMethod=MC:EffSel:SampleSize=200000:VarProp=FSmart:VarTransform=Decorrelate" );
if (Use["CutsPCA"])
factory->BookMethod( TMVA::Types::kCuts, "CutsPCA",
"!H:!V:FitMethod=MC:EffSel:SampleSize=200000:VarProp=FSmart:VarTransform=PCA" );
if (Use["CutsGA"])
factory->BookMethod( TMVA::Types::kCuts, "CutsGA",
"H:!V:FitMethod=GA:CutRangeMin[0]=-10:CutRangeMax[0]=10:VarProp[1]=FMax:EffSel:Steps=30:Cycles=3:PopSize=400:SC_steps=10:SC_rate=5:SC_factor=0.95" );
if (Use["CutsSA"])
factory->BookMethod( TMVA::Types::kCuts, "CutsSA",
"!H:!V:FitMethod=SA:EffSel:MaxCalls=150000:KernelTemp=IncAdaptive:InitialTemp=1e+6:MinTemp=1e-6:Eps=1e-10:UseDefaultScale" );
// Likelihood
if (Use["Likelihood"])
factory->BookMethod( TMVA::Types::kLikelihood, "Likelihood",
"H:!V:!TransformOutput:PDFInterpol=Spline2:NSmoothSig[0]=20:NSmoothBkg[0]=20:NSmoothBkg[1]=10:NSmooth=1:NAvEvtPerBin=50" );
// test the decorrelated likelihood
if (Use["LikelihoodD"])
factory->BookMethod( TMVA::Types::kLikelihood, "LikelihoodD",
"!H:!V:!TransformOutput:PDFInterpol=Spline2:NSmoothSig[0]=20:NSmoothBkg[0]=20:NSmooth=5:NAvEvtPerBin=50:VarTransform=Decorrelate" );
if (Use["LikelihoodPCA"])
factory->BookMethod( TMVA::Types::kLikelihood, "LikelihoodPCA",
"!H:!V:!TransformOutput:PDFInterpol=Spline2:NSmoothSig[0]=20:NSmoothBkg[0]=20:NSmooth=5:NAvEvtPerBin=50:VarTransform=PCA" );
// test the new kernel density estimator
if (Use["LikelihoodKDE"])
factory->BookMethod( TMVA::Types::kLikelihood, "LikelihoodKDE",
"!H:!V:!TransformOutput:PDFInterpol=KDE:KDEtype=Gauss:KDEiter=Adaptive:KDEFineFactor=0.3:KDEborder=None:NAvEvtPerBin=50" );
// test the mixed splines and kernel density estimator (depending on which variable)
if (Use["LikelihoodMIX"])
factory->BookMethod( TMVA::Types::kLikelihood, "LikelihoodMIX",
"!H:!V:!TransformOutput:PDFInterpolSig[0]=KDE:PDFInterpolBkg[0]=KDE:PDFInterpolSig[1]=KDE:PDFInterpolBkg[1]=KDE:PDFInterpolSig[2]=Spline2:PDFInterpolBkg[2]=Spline2:PDFInterpolSig[3]=Spline2:PDFInterpolBkg[3]=Spline2:KDEtype=Gauss:KDEiter=Nonadaptive:KDEborder=None:NAvEvtPerBin=50" );
// test the multi-dimensional probability density estimator
// here are the options strings for the MinMax and RMS methods, respectively:
// "!H:!V:VolumeRangeMode=MinMax:DeltaFrac=0.2:KernelEstimator=Gauss:GaussSigma=0.3" );
// "!H:!V:VolumeRangeMode=RMS:DeltaFrac=3:KernelEstimator=Gauss:GaussSigma=0.3" );
if (Use["PDERS"])
factory->BookMethod( TMVA::Types::kPDERS, "PDERS",
"!H:!V:NormTree=T:VolumeRangeMode=Adaptive:KernelEstimator=Gauss:GaussSigma=0.3:NEventsMin=400:NEventsMax=600" );
if (Use["PDERSkNN"])
factory->BookMethod( TMVA::Types::kPDERS, "PDERSkNN",
"!H:!V:VolumeRangeMode=kNN:KernelEstimator=Gauss:GaussSigma=0.3:NEventsMin=400:NEventsMax=600" );
if (Use["PDERSD"])
factory->BookMethod( TMVA::Types::kPDERS, "PDERSD",
"!H:!V:VolumeRangeMode=Adaptive:KernelEstimator=Gauss:GaussSigma=0.3:NEventsMin=400:NEventsMax=600:VarTransform=Decorrelate" );
if (Use["PDERSPCA"])
factory->BookMethod( TMVA::Types::kPDERS, "PDERSPCA",
"!H:!V:VolumeRangeMode=Adaptive:KernelEstimator=Gauss:GaussSigma=0.3:NEventsMin=400:NEventsMax=600:VarTransform=PCA" );
// Multi-dimensional likelihood estimator using self-adapting phase-space binning
if (Use["PDEFoam"])
factory->BookMethod( TMVA::Types::kPDEFoam, "PDEFoam",
"H:!V:SigBgSeparate=F:TailCut=0.001:VolFrac=0.0333:nActiveCells=500:nSampl=2000:nBin=5:CutNmin=T:Nmin=100:Kernel=None:Compress=T" );
// K-Nearest Neighbour classifier (KNN)
if (Use["KNN"])
factory->BookMethod( TMVA::Types::kKNN, "KNN",
"H:nkNN=20:ScaleFrac=0.8:SigmaFact=1.0:Kernel=Gaus:UseKernel=F:UseWeight=T:!Trim" );
// H-Matrix (chi2-squared) method
if (Use["HMatrix"])
factory->BookMethod( TMVA::Types::kHMatrix, "HMatrix", "!H:!V" );
// Fisher discriminant
if (Use["Fisher"])
factory->BookMethod( TMVA::Types::kFisher, "Fisher", "H:!V:Fisher:CreateMVAPdfs:PDFInterpolMVAPdf=Spline2:NbinsMVAPdf=60:NsmoothMVAPdf=10" );
// Fisher with Gauss-transformed input variables
if (Use["FisherG"])
factory->BookMethod( TMVA::Types::kFisher, "FisherG", "H:!V:VarTransform=Gauss" );
// Composite classifier: ensemble (tree) of boosted Fisher classifiers
if (Use["BoostedFisher"])
factory->BookMethod( TMVA::Types::kFisher, "BoostedFisher", "H:!V:Boost_Num=20:Boost_Transform=log:Boost_Type=AdaBoost:Boost_AdaBoostBeta=0.2");
// Linear discriminant (same as Fisher)
if (Use["LD"])
factory->BookMethod( TMVA::Types::kLD, "LD", "H:!V:VarTransform=None" );
// Function discrimination analysis (FDA) -- test of various fitters - the recommended one is Minuit (or GA or SA)
if (Use["FDA_MC"])
factory->BookMethod( TMVA::Types::kFDA, "FDA_MC",
"H:!V:Formula=(0)+(1)*x0+(2)*x1+(3)*x2+(4)*x3:ParRanges=(-1,1);(-10,10);(-10,10);(-10,10);(-10,10):FitMethod=MC:SampleSize=100000:Sigma=0.1" );
if (Use["FDA_GA"]) // can also use Simulated Annealing (SA) algorithm (see Cuts_SA options])
factory->BookMethod( TMVA::Types::kFDA, "FDA_GA",
"H:!V:Formula=(0)+(1)*x0+(2)*x1+(3)*x2+(4)*x3:ParRanges=(-1,1);(-10,10);(-10,10);(-10,10);(-10,10):FitMethod=GA:PopSize=300:Cycles=3:Steps=20:Trim=True:SaveBestGen=1" );
if (Use["FDA_SA"]) // can also use Simulated Annealing (SA) algorithm (see Cuts_SA options])
factory->BookMethod( TMVA::Types::kFDA, "FDA_SA",
"H:!V:Formula=(0)+(1)*x0+(2)*x1+(3)*x2+(4)*x3:ParRanges=(-1,1);(-10,10);(-10,10);(-10,10);(-10,10):FitMethod=SA:MaxCalls=15000:KernelTemp=IncAdaptive:InitialTemp=1e+6:MinTemp=1e-6:Eps=1e-10:UseDefaultScale" );
if (Use["FDA_MT"])
factory->BookMethod( TMVA::Types::kFDA, "FDA_MT",
"H:!V:Formula=(0)+(1)*x0+(2)*x1+(3)*x2+(4)*x3:ParRanges=(-1,1);(-10,10);(-10,10);(-10,10);(-10,10):FitMethod=MINUIT:ErrorLevel=1:PrintLevel=-1:FitStrategy=2:UseImprove:UseMinos:SetBatch" );
if (Use["FDA_GAMT"])
factory->BookMethod( TMVA::Types::kFDA, "FDA_GAMT",
"H:!V:Formula=(0)+(1)*x0+(2)*x1+(3)*x2+(4)*x3:ParRanges=(-1,1);(-10,10);(-10,10);(-10,10);(-10,10):FitMethod=GA:Converger=MINUIT:ErrorLevel=1:PrintLevel=-1:FitStrategy=0:!UseImprove:!UseMinos:SetBatch:Cycles=1:PopSize=5:Steps=5:Trim" );
if (Use["FDA_MCMT"])
factory->BookMethod( TMVA::Types::kFDA, "FDA_MCMT",
"H:!V:Formula=(0)+(1)*x0+(2)*x1+(3)*x2+(4)*x3:ParRanges=(-1,1);(-10,10);(-10,10);(-10,10);(-10,10):FitMethod=MC:Converger=MINUIT:ErrorLevel=1:PrintLevel=-1:FitStrategy=0:!UseImprove:!UseMinos:SetBatch:SampleSize=20" );
// TMVA ANN: MLP (recommended ANN) -- all ANNs in TMVA are Multilayer Perceptrons
if (Use["MLP"])
factory->BookMethod( TMVA::Types::kMLP, "MLP", "H:!V:NeuronType=tanh:VarTransform=N:NCycles=500:HiddenLayers=N+5:TestRate=10:EpochMonitoring" );
if (Use["MLPBFGS"])
factory->BookMethod( TMVA::Types::kMLP, "MLPBFGS", "H:!V:NeuronType=tanh:VarTransform=N:NCycles=500:HiddenLayers=N+5:TestRate=10:TrainingMethod=BFGS:!EpochMonitoring" );
// CF(Clermont-Ferrand)ANN
if (Use["CFMlpANN"])
factory->BookMethod( TMVA::Types::kCFMlpANN, "CFMlpANN", "!H:!V:NCycles=2000:HiddenLayers=N+1,N" ); // n_cycles:#nodes:#nodes:...
// Tmlp(Root)ANN
if (Use["TMlpANN"])
factory->BookMethod( TMVA::Types::kTMlpANN, "TMlpANN", "!H:!V:NCycles=200:HiddenLayers=N+1,N:LearningMethod=BFGS:ValidationFraction=0.3" ); // n_cycles:#nodes:#nodes:...
// Support Vector Machine
if (Use["SVM"])
factory->BookMethod( TMVA::Types::kSVM, "SVM", "Gamma=0.25:Tol=0.001:VarTransform=Norm" );
// Boosted Decision Trees
if (Use["BDTG"]) // Gradient Boost
factory->BookMethod( TMVA::Types::kBDT, "BDTG",
"!H:!V:NTrees=1000:BoostType=Grad:Shrinkage=0.30:UseBaggedGrad:GradBaggingFraction=0.6:SeparationType=GiniIndex:nCuts=20:NNodesMax=5" );
if (Use["BDT"]) // Adaptive Boost
factory->BookMethod( TMVA::Types::kBDT, "BDT",
"!H:!V:NTrees=400:nEventsMin=400:MaxDepth=3:BoostType=AdaBoost:SeparationType=GiniIndex:nCuts=20:PruneMethod=NoPruning" );
if (Use["BDTB"]) // Bagging
factory->BookMethod( TMVA::Types::kBDT, "BDTB",
"!H:!V:NTrees=400:BoostType=Bagging:SeparationType=GiniIndex:nCuts=20:PruneMethod=NoPruning" );
if (Use["BDTD"]) // Decorrelation + Adaptive Boost
factory->BookMethod( TMVA::Types::kBDT, "BDTD",
"!H:!V:NTrees=400:nEventsMin=400:MaxDepth=3:BoostType=AdaBoost:SeparationType=GiniIndex:nCuts=20:PruneMethod=NoPruning:VarTransform=Decorrelate" );
// RuleFit -- TMVA implementation of Friedman's method
if (Use["RuleFit"])
factory->BookMethod( TMVA::Types::kRuleFit, "RuleFit",
"H:!V:RuleFitModule=RFTMVA:Model=ModRuleLinear:MinImp=0.001:RuleMinDist=0.001:NTrees=20:fEventsMin=0.01:fEventsMax=0.5:GDTau=-1.0:GDTauPrec=0.01:GDStep=0.01:GDNSteps=10000:GDErrScale=1.02" );
// For an example of the category classifier, see: TMVAClassificationCategory
// --------------------------------------------------------------------------------------------------
// As an example how to use the ROOT plugin mechanism, book BDT via
// plugin mechanism
if (Use["Plugin"]) {
//
// first the plugin has to be defined, which can happen either through the following line in the local or global .rootrc:
//
// # plugin handler plugin name(regexp) class to be instanciated library constructor format
// Plugin.TMVA@@MethodBase: ^BDT TMVA::MethodBDT TMVA.1 "MethodBDT(TString,TString,DataSet&,TString)"
//
// or by telling the global plugin manager directly
gPluginMgr->AddHandler("TMVA@@MethodBase", "BDT", "TMVA::MethodBDT", "TMVA.1", "MethodBDT(TString,TString,DataSet&,TString)");
factory->BookMethod( TMVA::Types::kPlugins, "BDT",
"!H:!V:NTrees=400:BoostType=AdaBoost:SeparationType=GiniIndex:nCuts=20:PruneMethod=CostComplexity:PruneStrength=50" );
}
// --------------------------------------------------------------------------------------------------
// ---- Now you can tell the factory to train, test, and evaluate the MVAs
// Train MVAs using the set of training events
factory->TrainAllMethods();
// ---- Evaluate all MVAs using the set of test events
factory->TestAllMethods();
// ----- Evaluate and compare performance of all configured MVAs
factory->EvaluateAllMethods();
if (Use["Cuts"]) {
for( unsigned iEff=1; iEff<11; ++iEff ) {
TMVA::IMethod* method = (TMVA::IMethod*)factory->GetMethod("Cuts");
TMVA::MethodCuts* cuts = dynamic_cast<TMVA::MethodCuts*>(method);
std::string optcutsdir = "optcuts_" + selectionName + "_" + charge;
std::string mkdir_command = "mkdir -p " + optcutsdir;
system(mkdir_command.c_str());
char cutsFileName[500];
sprintf( cutsFileName, "%s/cuts_Seff%d.txt", optcutsdir.c_str(), 10*iEff );
ofstream ofs(cutsFileName);
std::vector<Double_t> cutsMin, cutsMax;
cuts->GetCuts((float)iEff*0.10, cutsMin, cutsMax);
bool found_pT1 = false;
bool found_pT2 = false;
bool found_NJ = false;
bool found_NbJ = false;
bool found_NbJmed = false;
for( unsigned iCut=0; iCut<cutsMin.size(); ++iCut) {
TString varName = factory->DefaultDataSetInfo().GetVariableInfo(iCut).GetInternalName();
if( varName=="TMath_Min_pT1,pT2_") {
ofs << "pT1 " << cutsMin[iCut] << " " << cutsMax[iCut] << std::endl;
ofs << "pT2 " << cutsMin[iCut] << " " << cutsMax[iCut] << std::endl;
found_pT1 = true;
found_pT2 = true;
} else {
ofs << varName << " " << cutsMin[iCut] << " " << cutsMax[iCut] << std::endl;
}
if( varName=="pT1" ) found_pT1 = true;
if( varName=="pT2" ) found_pT2 = true;
if( varName=="NJ" ) found_NJ = true;
if( varName=="NbJ" ) found_NbJ = true;
if( varName=="NbJmed" ) found_NbJmed = true;
}
// preselection cuts (if not optimized):
if( !found_pT1 ) ofs << "pT1 20. 100000." << std::endl;
if( !found_pT2 ) ofs << "pT2 20. 100000." << std::endl;
if( !found_NJ ) ofs << "NJ 3 100000." << std::endl;
if( !found_NbJ ) ofs << "NbJ 1 100000." << std::endl;
if( !found_NbJmed && btagMed_presel_ ) ofs << "NbJmed 1 100000." << std::endl;
if( charge=="plus" )
ofs << "Charge 1 10" << std::endl;
else if( charge=="minus" )
ofs << "Charge -10 0" << std::endl;
ofs.close();
} // for eff
} // if cuts
// --------------------------------------------------------------
// Save the output
outputFile->Close();
std::cout << "==> Wrote root file: " << outputFile->GetName() << std::endl;
std::cout << "==> TMVAClassification is done!" << std::endl;
delete factory;
// Launch the GUI for the root macros
if (!gROOT->IsBatch()) TMVAGui( outfileName );
}
void TMVAClassificationApplication( TString myMethodList = "" , TString decay_mode)
{
#ifdef __CINT__
gROOT->ProcessLine( ".O0" ); // turn off optimization in CINT
#endif
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
// --- Cut optimisation
Use["Cuts"] = 1;
Use["CutsD"] = 1;
Use["CutsPCA"] = 0;
Use["CutsGA"] = 0;
Use["CutsSA"] = 0;
//
// --- 1-dimensional likelihood ("naive Bayes estimator")
Use["Likelihood"] = 1;
Use["LikelihoodD"] = 0; // the "D" extension indicates decorrelated input variables (see option strings)
Use["LikelihoodPCA"] = 1; // the "PCA" extension indicates PCA-transformed input variables (see option strings)
Use["LikelihoodKDE"] = 0;
Use["LikelihoodMIX"] = 0;
//
// --- Mutidimensional likelihood and Nearest-Neighbour methods
Use["PDERS"] = 1;
Use["PDERSD"] = 0;
Use["PDERSPCA"] = 0;
Use["PDEFoam"] = 1;
Use["PDEFoamBoost"] = 0; // uses generalised MVA method boosting
Use["KNN"] = 1; // k-nearest neighbour method
//
// --- Linear Discriminant Analysis
Use["LD"] = 1; // Linear Discriminant identical to Fisher
Use["Fisher"] = 0;
Use["FisherG"] = 0;
Use["BoostedFisher"] = 0; // uses generalised MVA method boosting
Use["HMatrix"] = 0;
//
// --- Function Discriminant analysis
Use["FDA_GA"] = 1; // minimisation of user-defined function using Genetics Algorithm
Use["FDA_SA"] = 0;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 0;
Use["FDA_GAMT"] = 0;
Use["FDA_MCMT"] = 0;
//
// --- Neural Networks (all are feed-forward Multilayer Perceptrons)
Use["MLP"] = 0; // Recommended ANN
Use["MLPBFGS"] = 0; // Recommended ANN with optional training method
Use["MLPBNN"] = 1; // Recommended ANN with BFGS training method and bayesian regulator
Use["CFMlpANN"] = 0; // Depreciated ANN from ALEPH
Use["TMlpANN"] = 0; // ROOT's own ANN
//
// --- Support Vector Machine
Use["SVM"] = 1;
//
// --- Boosted Decision Trees
Use["BDT"] = 1; // uses Adaptive Boost
Use["BDTG"] = 0; // uses Gradient Boost
Use["BDTB"] = 0; // uses Bagging
Use["BDTD"] = 0; // decorrelation + Adaptive Boost
//
// --- Friedman's RuleFit method, ie, an optimised series of cuts ("rules")
Use["RuleFit"] = 1;
// ---------------------------------------------------------------
Use["Plugin"] = 0;
Use["Category"] = 0;
Use["SVM_Gauss"] = 0;
Use["SVM_Poly"] = 0;
Use["SVM_Lin"] = 0;
std::cout << std::endl;
std::cout << "==> Start TMVAClassificationApplication" << std::endl;
// Select methods (don't look at this code - not of interest)
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod
<< "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
std::cout << it->first << " ";
}
std::cout << std::endl;
return;
}
Use[regMethod] = 1;
}
}
// --------------------------------------------------------------------------------------------------
// --- Create the Reader object
TMVA::Reader *reader = new TMVA::Reader( "!Color:!Silent" );
// Create a set of variables and declare them to the reader
// - the variable names MUST corresponds in name and type to those given in the weight file(s) used
Float_t met, mT, mT2W;
reader->AddVariable("met", &met);
reader->AddVariable("mT", &mT);
reader->AddVariable("mT2W",&mT2W);
// Spectator variables declared in the training have to be added to the reader, too
/* Float_t spec1,spec2;
reader->AddSpectator( "spec1 := var1*2", &spec1 );
reader->AddSpectator( "spec2 := var1*3", &spec2 );
Float_t Category_cat1, Category_cat2, Category_cat3;
if (Use["Category"]){
// Add artificial spectators for distinguishing categories
reader->AddSpectator( "Category_cat1 := var3<=0", &Category_cat1 );
reader->AddSpectator( "Category_cat2 := (var3>0)&&(var4<0)", &Category_cat2 );
reader->AddSpectator( "Category_cat3 := (var3>0)&&(var4>=0)", &Category_cat3 );
}
*/
// --- Book the MVA methods
TString dir = "weights/";
TString prefix = "TMVAClassification";
// Book method(s)
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString methodName = TString(it->first) + TString(" method");
TString weightfile = dir + prefix + TString("_") + TString(it->first) + TString(".weights.xml");
reader->BookMVA( methodName, weightfile );
}
}
// Book output histograms
UInt_t nbin = 100;
TH1F *histLk(0), *histLkD(0), *histLkPCA(0), *histLkKDE(0), *histLkMIX(0), *histPD(0), *histPDD(0);
TH1F *histPDPCA(0), *histPDEFoam(0), *histPDEFoamErr(0), *histPDEFoamSig(0), *histKNN(0), *histHm(0);
TH1F *histFi(0), *histFiG(0), *histFiB(0), *histLD(0), *histNn(0),*histNnbfgs(0),*histNnbnn(0);
TH1F *histNnC(0), *histNnT(0), *histBdt(0), *histBdtG(0), *histBdtD(0), *histRf(0), *histSVMG(0);
TH1F *histSVMP(0), *histSVML(0), *histFDAMT(0), *histFDAGA(0), *histCat(0), *histPBdt(0);
if (Use["Likelihood"]) histLk = new TH1F( "MVA_Likelihood", "MVA_Likelihood", nbin, -1, 1 );
if (Use["LikelihoodD"]) histLkD = new TH1F( "MVA_LikelihoodD", "MVA_LikelihoodD", nbin, -1, 0.9999 );
if (Use["LikelihoodPCA"]) histLkPCA = new TH1F( "MVA_LikelihoodPCA", "MVA_LikelihoodPCA", nbin, -1, 1 );
if (Use["LikelihoodKDE"]) histLkKDE = new TH1F( "MVA_LikelihoodKDE", "MVA_LikelihoodKDE", nbin, -0.00001, 0.99999 );
if (Use["LikelihoodMIX"]) histLkMIX = new TH1F( "MVA_LikelihoodMIX", "MVA_LikelihoodMIX", nbin, 0, 1 );
if (Use["PDERS"]) histPD = new TH1F( "MVA_PDERS", "MVA_PDERS", nbin, 0, 1 );
if (Use["PDERSD"]) histPDD = new TH1F( "MVA_PDERSD", "MVA_PDERSD", nbin, 0, 1 );
if (Use["PDERSPCA"]) histPDPCA = new TH1F( "MVA_PDERSPCA", "MVA_PDERSPCA", nbin, 0, 1 );
if (Use["KNN"]) histKNN = new TH1F( "MVA_KNN", "MVA_KNN", nbin, 0, 1 );
if (Use["HMatrix"]) histHm = new TH1F( "MVA_HMatrix", "MVA_HMatrix", nbin, -0.95, 1.55 );
if (Use["Fisher"]) histFi = new TH1F( "MVA_Fisher", "MVA_Fisher", nbin, -4, 4 );
if (Use["FisherG"]) histFiG = new TH1F( "MVA_FisherG", "MVA_FisherG", nbin, -1, 1 );
if (Use["BoostedFisher"]) histFiB = new TH1F( "MVA_BoostedFisher", "MVA_BoostedFisher", nbin, -2, 2 );
if (Use["LD"]) histLD = new TH1F( "MVA_LD", "MVA_LD", nbin, -2, 2 );
if (Use["MLP"]) histNn = new TH1F( "MVA_MLP", "MVA_MLP", nbin, -1.25, 1.5 );
if (Use["MLPBFGS"]) histNnbfgs = new TH1F( "MVA_MLPBFGS", "MVA_MLPBFGS", nbin, -1.25, 1.5 );
if (Use["MLPBNN"]) histNnbnn = new TH1F( "MVA_MLPBNN", "MVA_MLPBNN", nbin, -1.25, 1.5 );
if (Use["CFMlpANN"]) histNnC = new TH1F( "MVA_CFMlpANN", "MVA_CFMlpANN", nbin, 0, 1 );
if (Use["TMlpANN"]) histNnT = new TH1F( "MVA_TMlpANN", "MVA_TMlpANN", nbin, -1.3, 1.3 );
if (Use["BDT"]) histBdt = new TH1F( "MVA_BDT", "MVA_BDT", nbin, -0.8, 0.8 );
if (Use["BDTD"]) histBdtD = new TH1F( "MVA_BDTD", "MVA_BDTD", nbin, -0.8, 0.8 );
if (Use["BDTG"]) histBdtG = new TH1F( "MVA_BDTG", "MVA_BDTG", nbin, -1.0, 1.0 );
if (Use["RuleFit"]) histRf = new TH1F( "MVA_RuleFit", "MVA_RuleFit", nbin, -2.0, 2.0 );
if (Use["SVM_Gauss"]) histSVMG = new TH1F( "MVA_SVM_Gauss", "MVA_SVM_Gauss", nbin, 0.0, 1.0 );
if (Use["SVM_Poly"]) histSVMP = new TH1F( "MVA_SVM_Poly", "MVA_SVM_Poly", nbin, 0.0, 1.0 );
if (Use["SVM_Lin"]) histSVML = new TH1F( "MVA_SVM_Lin", "MVA_SVM_Lin", nbin, 0.0, 1.0 );
if (Use["FDA_MT"]) histFDAMT = new TH1F( "MVA_FDA_MT", "MVA_FDA_MT", nbin, -2.0, 3.0 );
if (Use["FDA_GA"]) histFDAGA = new TH1F( "MVA_FDA_GA", "MVA_FDA_GA", nbin, -2.0, 3.0 );
if (Use["Category"]) histCat = new TH1F( "MVA_Category", "MVA_Category", nbin, -2., 2. );
if (Use["Plugin"]) histPBdt = new TH1F( "MVA_PBDT", "MVA_BDT", nbin, -0.8, 0.8 );
// PDEFoam also returns per-event error, fill in histogram, and also fill significance
if (Use["PDEFoam"]) {
histPDEFoam = new TH1F( "MVA_PDEFoam", "MVA_PDEFoam", nbin, 0, 1 );
histPDEFoamErr = new TH1F( "MVA_PDEFoamErr", "MVA_PDEFoam error", nbin, 0, 1 );
histPDEFoamSig = new TH1F( "MVA_PDEFoamSig", "MVA_PDEFoam significance", nbin, 0, 10 );
}
// Book example histogram for probability (the other methods are done similarly)
TH1F *probHistFi(0), *rarityHistFi(0);
if (Use["Fisher"]) {
probHistFi = new TH1F( "MVA_Fisher_Proba", "MVA_Fisher_Proba", nbin, 0, 1 );
rarityHistFi = new TH1F( "MVA_Fisher_Rarity", "MVA_Fisher_Rarity", nbin, 0, 1 );
}
TString fname;
Double_t event_weight; // events weights are: xsection * efficiency (preselection)
if (decay_mode == "TT_1Lep") {fname = "Example_Rootfiles/ttbar_1l/output/ttbar_1l.root"; event_weight = 0.153;}
if (decay_mode == "TT_2Lep") {fname = "Example_Rootfiles/ttbar_2l/output/ttbar_2l.root"; event_weight = 0.129;}
if (decay_mode == "WJets") {fname = "Example_Rootfiles/wjets_all/output/wjets_all.root"; event_weight = 0.67;}
if (decay_mode == "Others") {fname = "Example_Rootfiles/others_all/output/others_all.root"; event_weight = 0.106;}
if (decay_mode == "T2tt_R1") { fname = "Example_Rootfiles/t2tt_all/R1/output/t2tt_all_R1.root"; event_weight = 1592./24845.;}
if (decay_mode == "T2tt_R2") { fname = "Example_Rootfiles/t2tt_all/R2/output/t2tt_all_R2.root"; event_weight = 1018./24994.;}
if (decay_mode == "T2tt_R3") { fname = "Example_Rootfiles/t2tt_all/R3/output/t2tt_all_R3.root"; event_weight = 306./25006.;}
if (decay_mode == "T2tt_R4") { fname = "Example_Rootfiles/t2tt_all/R4/output/t2tt_all_R4.root"; event_weight = 87./25094.;}
if (decay_mode == "T2tt_R5") { fname = "Example_Rootfiles/t2tt_all/R5/output/t2tt_all_R5.root"; event_weight = 28./25075.;}
if (decay_mode == "T2tt_R6") { fname = "Example_Rootfiles/t2tt_all/R6/output/t2tt_all_R6.root"; event_weight = 9./24863.;}
TFile *input = TFile::Open( fname );
std::cout << "--- TMVAClassification : Using input file: " << input->GetName() << std::endl;
// --- //Event loop
// Prepare the event tree
// - here the variable names have to corresponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
std::cout << "--- Select signal sample" << std::endl;
TTree *theTree = (TTree*)input->Get("BDTtree");
Float_t met, mT, mT2W;
Int_t event;
theTree->SetBranchAddress("met", &met);
theTree->SetBranchAddress("mT", &mT);
theTree->SetBranchAddress("mT2W", &mT2W);
theTree->SetBranchAddress("event", &event);
// Efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::vector<Float_t> vecVar(4); // vector for EvaluateMVA tests
Int_t totevt;
if (string(decay_mode).find("T2tt") != std::string::npos) totevt = 50000; // Takes too long (for exercise) to run over all the signal events
else totevt = theTree->GetEntries();
std::cout << "--- Processing: " << totevt << " events" << std::endl;
TStopwatch sw;
sw.Start();
for (Long64_t ievt=0; ievt<totevt; ievt++) {
theTree->GetEntry(ievt);
if (ievt%1000 == 0) std::cout << "--- ... Processing event: " << ievt << std::endl;
if ( event%2 == 1 ) continue; // remove all odd numbers from our actual analysis
// --- Return the MVA outputs and fill into histograms
// this gives you the mva value per event
// cout << reader->EvaluateMVA("BDT method") << endl;
if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
if (Use["Likelihood" ]) histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) );
if (Use["LikelihoodD" ]) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) );
if (Use["LikelihoodPCA"]) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) );
if (Use["LikelihoodKDE"]) histLkKDE ->Fill( reader->EvaluateMVA( "LikelihoodKDE method" ) );
if (Use["LikelihoodMIX"]) histLkMIX ->Fill( reader->EvaluateMVA( "LikelihoodMIX method" ) );
if (Use["PDERS" ]) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) );
if (Use["PDERSD" ]) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) );
if (Use["PDERSPCA" ]) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) );
if (Use["KNN" ]) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) );
if (Use["HMatrix" ]) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) );
if (Use["Fisher" ]) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) );
if (Use["FisherG" ]) histFiG ->Fill( reader->EvaluateMVA( "FisherG method" ) );
if (Use["BoostedFisher"]) histFiB ->Fill( reader->EvaluateMVA( "BoostedFisher method" ) );
if (Use["LD" ]) histLD ->Fill( reader->EvaluateMVA( "LD method" ) );
if (Use["MLP" ]) histNn ->Fill( reader->EvaluateMVA( "MLP method" ) );
if (Use["MLPBFGS" ]) histNnbfgs ->Fill( reader->EvaluateMVA( "MLPBFGS method" ) );
if (Use["MLPBNN" ]) histNnbnn ->Fill( reader->EvaluateMVA( "MLPBNN method" ) );
if (Use["CFMlpANN" ]) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) );
if (Use["TMlpANN" ]) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) );
if (Use["BDT" ]) histBdt ->Fill( reader->EvaluateMVA( "BDT method" ), event_weight );
//if (Use["BDT" ]) histBdt ->Fill( reader->EvaluateMVA( "BDT method" ));
if (Use["BDTD" ]) histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) );
if (Use["BDTG" ]) histBdtG ->Fill( reader->EvaluateMVA( "BDTG method" ) );
if (Use["RuleFit" ]) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) );
if (Use["SVM_Gauss" ]) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) );
if (Use["SVM_Poly" ]) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) );
if (Use["SVM_Lin" ]) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) );
if (Use["FDA_MT" ]) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) );
if (Use["FDA_GA" ]) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) );
if (Use["Category" ]) histCat ->Fill( reader->EvaluateMVA( "Category method" ) );
if (Use["Plugin" ]) histPBdt ->Fill( reader->EvaluateMVA( "P_BDT method" ) );
// Retrieve also per-event error
if (Use["PDEFoam"]) {
Double_t val = reader->EvaluateMVA( "PDEFoam method" );
Double_t err = reader->GetMVAError();
histPDEFoam ->Fill( val );
histPDEFoamErr->Fill( err );
if (err>1.e-50) histPDEFoamSig->Fill( val/err );
}
// Retrieve probability instead of MVA output
if (Use["Fisher"]) {
probHistFi ->Fill( reader->GetProba ( "Fisher method" ) );
rarityHistFi->Fill( reader->GetRarity( "Fisher method" ) );
}
}
// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// Get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
// --- Write histograms
TFile *target = new TFile( "results_"+decay_mode+".root","RECREATE" );
if (Use["Likelihood" ]) histLk ->Write();
if (Use["LikelihoodD" ]) histLkD ->Write();
if (Use["LikelihoodPCA"]) histLkPCA ->Write();
if (Use["LikelihoodKDE"]) histLkKDE ->Write();
if (Use["LikelihoodMIX"]) histLkMIX ->Write();
if (Use["PDERS" ]) histPD ->Write();
if (Use["PDERSD" ]) histPDD ->Write();
if (Use["PDERSPCA" ]) histPDPCA ->Write();
if (Use["KNN" ]) histKNN ->Write();
if (Use["HMatrix" ]) histHm ->Write();
if (Use["Fisher" ]) histFi ->Write();
if (Use["FisherG" ]) histFiG ->Write();
if (Use["BoostedFisher"]) histFiB ->Write();
if (Use["LD" ]) histLD ->Write();
if (Use["MLP" ]) histNn ->Write();
if (Use["MLPBFGS" ]) histNnbfgs ->Write();
if (Use["MLPBNN" ]) histNnbnn ->Write();
if (Use["CFMlpANN" ]) histNnC ->Write();
if (Use["TMlpANN" ]) histNnT ->Write();
if (Use["BDT" ]) histBdt ->Write();
if (Use["BDTD" ]) histBdtD ->Write();
if (Use["BDTG" ]) histBdtG ->Write();
if (Use["RuleFit" ]) histRf ->Write();
if (Use["SVM_Gauss" ]) histSVMG ->Write();
if (Use["SVM_Poly" ]) histSVMP ->Write();
if (Use["SVM_Lin" ]) histSVML ->Write();
if (Use["FDA_MT" ]) histFDAMT ->Write();
if (Use["FDA_GA" ]) histFDAGA ->Write();
if (Use["Category" ]) histCat ->Write();
if (Use["Plugin" ]) histPBdt ->Write();
// Write also error and significance histos
if (Use["PDEFoam"]) { histPDEFoam->Write(); histPDEFoamErr->Write(); histPDEFoamSig->Write(); }
// Write also probability hists
if (Use["Fisher"]) { if (probHistFi != 0) probHistFi->Write(); if (rarityHistFi != 0) rarityHistFi->Write(); }
target->Close();
std::cout << "--- Created root file: \"results_"+decay_mode+".root\" containing the MVA output histograms" << std::endl;
delete reader;
std::cout << "==> TMVAClassificationApplication is done!" << endl << std::endl;
}
void TMVAClassificationApplication( TString myMethodList = "", TString filename="" )
{
//---------------------------------------------------------------
// default MVA methods to be trained + tested
// this loads the library
TMVA::Tools::Instance();
std::map<std::string,int> Use;
Use["CutsGA"] = 1; // other "Cuts" methods work identically
// ---
Use["Likelihood"] = 1;
Use["LikelihoodD"] = 1; // the "D" extension indicates decorrelated input variables (see option strings)
Use["LikelihoodPCA"] = 0; // the "PCA" extension indicates PCA-transformed input variables (see option strings)
Use["LikelihoodKDE"] = 0;
Use["LikelihoodMIX"] = 0;
// ---
Use["PDERS"] = 1;
Use["PDERSD"] = 0;
Use["PDERSPCA"] = 0;
Use["PDERSkNN"] = 0; // depreciated until further notice
Use["PDEFoam"] = 1;
// --
Use["KNN"] = 1;
// ---
Use["HMatrix"] = 0;
Use["Fisher"] = 1;
Use["FisherG"] = 0;
Use["BoostedFisher"] = 0;
Use["LD"] = 1;
// ---
Use["FDA_GA"] = 0;
Use["FDA_SA"] = 0;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 1;
Use["FDA_GAMT"] = 0;
Use["FDA_MCMT"] = 0;
// ---
Use["MLP"] = 1; // this is the recommended ANN
Use["MLPBFGS"] = 0; // recommended ANN with optional training method
Use["CFMlpANN"] = 0; // *** missing
Use["TMlpANN"] = 0;
// ---
Use["SVM"] = 1;
// ---
Use["BDT"] = 1;
Use["BDTD"] = 0;
Use["BDTG"] = 1;
Use["BDTB"] = 0;
// ---
Use["RuleFit"] = 0;
// ---
Use["Plugin"] = 0;
// ---------------------------------------------------------------
std::cout << std::endl;
std::cout << "==> Start TMVAClassificationApplication" << std::endl;
if (myMethodList != "") {
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) it->second = 0;
std::vector<TString> mlist = gTools().SplitString( myMethodList, ',' );
for (UInt_t i=0; i<mlist.size(); i++) {
std::string regMethod(mlist[i]);
if (Use.find(regMethod) == Use.end()) {
std::cout << "Method \"" << regMethod << "\" not known in TMVA under this name. Choose among the following:" << std::endl;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) std::cout << it->first << " ";
std::cout << std::endl;
return;
}
Use[regMethod] = 1;
}
}
//
// create the Reader object
//
TMVA::Reader *reader = new TMVA::Reader( "!Color:!Silent" );
// create a set of variables and declare them to the reader
// - the variable names must corresponds in name and type to
// those given in the weight file(s) that you use
///Our variables here:
/************************************************************************************************************
* Row * nEvt * sumEt * Et1 * lepeta * MET * dphiMETle * detajet2j * detajet3j *
************************************************************************************************************
mindijetmass
maxdijetmass
*/
Float_t sumEt;
Float_t Et1;
Float_t lepeta;
Float_t MET;
Float_t dphiMETlepton;
Float_t detajet2jet3;
Float_t detajet3jet4;
Float_t mindijetmass;
Float_t maxdijetmass;
reader->AddVariable( "sumEt", &sumEt );
reader->AddVariable( "Et1", &Et1 );
reader->AddVariable( "lepeta", &lepeta );
reader->AddVariable( "MET", &MET );
reader->AddVariable( "dphiMETlepton", &dphiMETlepton );
reader->AddVariable( "detajet2jet3", &detajet2jet3 );
reader->AddVariable( "detajet3jet4", &detajet3jet4 );
reader->AddVariable( "mindijetmass", &mindijetmass );
reader->AddVariable( "maxdijetmass", &maxdijetmass );
TString TreeName = "KinVars";
TFile * input = TFile::Open(filename+".root");
TTree* theTree = (TTree*)input->Get(TreeName);
Float_t userVar1, userVar2;
theTree->SetBranchAddress( "MassOfJets", &MassOfJets );
theTree->SetBranchAddress( "DeltaPhiOfJets", &DeltaPhiOfJets );
theTree->SetBranchAddress( "DeltaPhiLeadingJet", &DeltaPhiLeadingJet );
theTree->SetBranchAddress( "Pbalance", &Pbalance );
theTree->SetBranchAddress( "Sphericity", &Sphericity );
Double_t _Weight=1.0;
if (filename.Contains("hbb_120GeV")) _Weight = 5.09999999999999967e-02;
if (filename.Contains("qcd_15to30Gev")) _Weight = 3.54815799999999990e+03;
if (filename.Contains("qcd_30to50Gev")) _Weight = 3.36987000000000023e+02;
if (filename.Contains("qcd_50to150Gev")) _Weight = 1.08418000000000006e+02;
if (filename.Contains("qcd_150toInfGev")) _Weight = 4.70999999999999996e+00;
if (filename.Contains("data")) _Weight = -1e0;
//
// book the MVA methods
//
TString dir = "weights/";
TString prefix = "TMVAClassification";
// book method(s)
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
if (it->second) {
TString methodName = it->first + " method";
TString weightfile = dir + prefix + "_" + TString(it->first) + ".weights.xml";
reader->BookMVA( methodName, weightfile );
}
}
// example how to use your own method as plugin
if (Use["Plugin"]) {
// the weight file contains a line
// Method : MethodName::InstanceName
// if MethodName is not a known TMVA method, it is assumed to be
// a user implemented method which has to be loaded via the
// plugin mechanism
// for user implemented methods the line in the weight file can be
// Method : PluginName::InstanceName
// where PluginName can be anything
// before usage the plugin has to be defined, which can happen
// either through the following line in .rootrc:
// # plugin handler plugin class library constructor format
// Plugin.TMVA@@MethodBase: PluginName MethodClassName UserPackage "MethodName(DataSet&,TString)"
//
// or by telling the global plugin manager directly
gPluginMgr->AddHandler("TMVA@@MethodBase", "PluginName", "MethodClassName", "UserPackage", "MethodName(DataSet&,TString)");
// the class is then looked for in libUserPackage.so
// now the method can be booked like any other
reader->BookMVA( "User method", dir + prefix + "_User.weights.txt" );
}
// book output histograms
UInt_t nbin = 100;
TH1F *histLk(0), *histLkD(0), *histLkPCA(0), *histLkKDE(0), *histLkMIX(0), *histPD(0), *histPDD(0);
TH1F *histPDPCA(0), *histPDEFoam(0), *histPDEFoamErr(0), *histPDEFoamSig(0), *histKNN(0), *histHm(0);
TH1F *histFi(0), *histFiG(0), *histFiB(0), *histLD(0), *histNn(0), *histNnC(0), *histNnT(0), *histBdt(0), *histBdtG(0), *histBdtD(0);
TH1F *histRf(0), *histSVMG(0), *histSVMP(0), *histSVML(0), *histFDAMT(0), *histFDAGA(0), *histPBdt(0);
if (Use["Likelihood"]) histLk = new TH1F( "MVA_Likelihood", "MVA_Likelihood", nbin, -1, 1 );
if (Use["LikelihoodD"]) histLkD = new TH1F( "MVA_LikelihoodD", "MVA_LikelihoodD", nbin, -1, 0.9999 );
if (Use["LikelihoodPCA"]) histLkPCA = new TH1F( "MVA_LikelihoodPCA", "MVA_LikelihoodPCA", nbin, -1, 1 );
if (Use["LikelihoodKDE"]) histLkKDE = new TH1F( "MVA_LikelihoodKDE", "MVA_LikelihoodKDE", nbin, -0.00001, 0.99999 );
if (Use["LikelihoodMIX"]) histLkMIX = new TH1F( "MVA_LikelihoodMIX", "MVA_LikelihoodMIX", nbin, 0, 1 );
if (Use["PDERS"]) histPD = new TH1F( "MVA_PDERS", "MVA_PDERS", nbin, 0, 1 );
if (Use["PDERSD"]) histPDD = new TH1F( "MVA_PDERSD", "MVA_PDERSD", nbin, 0, 1 );
if (Use["PDERSPCA"]) histPDPCA = new TH1F( "MVA_PDERSPCA", "MVA_PDERSPCA", nbin, 0, 1 );
if (Use["KNN"]) histKNN = new TH1F( "MVA_KNN", "MVA_KNN", nbin, 0, 1 );
if (Use["HMatrix"]) histHm = new TH1F( "MVA_HMatrix", "MVA_HMatrix", nbin, -0.95, 1.55 );
if (Use["Fisher"]) histFi = new TH1F( "MVA_Fisher", "MVA_Fisher", nbin, -4, 4 );
if (Use["FisherG"]) histFiG = new TH1F( "MVA_FisherG", "MVA_FisherG", nbin, -1, 1 );
if (Use["BoostedFisher"]) histFiB = new TH1F( "MVA_BoostedFisher", "MVA_BoostedFisher", nbin, -2, 2 );
if (Use["LD"]) histLD = new TH1F( "MVA_LD", "MVA_LD", nbin, -2, 2 );
if (Use["MLP"]) histNn = new TH1F( "MVA_MLP", "MVA_MLP", nbin, -1.25, 1.5 );
if (Use["CFMlpANN"]) histNnC = new TH1F( "MVA_CFMlpANN", "MVA_CFMlpANN", nbin, 0, 1 );
if (Use["TMlpANN"]) histNnT = new TH1F( "MVA_TMlpANN", "MVA_TMlpANN", nbin, -1.3, 1.3 );
if (Use["BDT"]) histBdt = new TH1F( "MVA_BDT", "MVA_BDT", nbin, -0.8, 0.8 );
if (Use["BDTD"]) histBdtD = new TH1F( "MVA_BDTD", "MVA_BDTD", nbin, -0.8, 0.8 );
if (Use["BDTG"]) histBdtG = new TH1F( "MVA_BDTG", "MVA_BDTG", nbin, -1.0, 1.0 );
if (Use["RuleFit"]) histRf = new TH1F( "MVA_RuleFit", "MVA_RuleFit", nbin, -2.0, 2.0 );
if (Use["SVM_Gauss"]) histSVMG = new TH1F( "MVA_SVM_Gauss", "MVA_SVM_Gauss", nbin, 0.0, 1.0 );
if (Use["SVM_Poly"]) histSVMP = new TH1F( "MVA_SVM_Poly", "MVA_SVM_Poly", nbin, 0.0, 1.0 );
if (Use["SVM_Lin"]) histSVML = new TH1F( "MVA_SVM_Lin", "MVA_SVM_Lin", nbin, 0.0, 1.0 );
if (Use["FDA_MT"]) histFDAMT = new TH1F( "MVA_FDA_MT", "MVA_FDA_MT", nbin, -2.0, 3.0 );
if (Use["FDA_GA"]) histFDAGA = new TH1F( "MVA_FDA_GA", "MVA_FDA_GA", nbin, -2.0, 3.0 );
if (Use["Plugin"]) histPBdt = new TH1F( "MVA_PBDT", "MVA_BDT", nbin, -0.8, 0.8 );
// PDEFoam also returns per-event error, fill in histogram, and also fill significance
if (Use["PDEFoam"]) {
histPDEFoam = new TH1F( "MVA_PDEFoam", "MVA_PDEFoam", nbin, 0, 1 );
histPDEFoamErr = new TH1F( "MVA_PDEFoamErr", "MVA_PDEFoam error", nbin, 0, 1 );
histPDEFoamSig = new TH1F( "MVA_PDEFoamSig", "MVA_PDEFoam significance", nbin, 0, 10 );
}
// book example histogram for probability (the other methods are done similarly)
TH1F *probHistFi(0), *rarityHistFi(0);
if (Use["Fisher"]) {
probHistFi = new TH1F( "MVA_Fisher_Proba", "MVA_Fisher_Proba", nbin, 0, 1 );
rarityHistFi = new TH1F( "MVA_Fisher_Rarity", "MVA_Fisher_Rarity", nbin, 0, 1 );
}
// Prepare input tree (this must be replaced by your data source)
// in this example, there is a toy tree with signal and one with background events
// we'll later on use only the "signal" events for the test in this example.
//
/*TFile *input(0);
TString fname = "./tmva_example.root";
if (!gSystem->AccessPathName( fname )) {
input = TFile::Open( fname ); // check if file in local directory exists
}
else {
input = TFile::Open( "http://root.cern.ch/files/tmva_class_example.root" ); // if not: download from ROOT server
}
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVAClassificationApp : Using input file: " << input->GetName() << std::endl;
*/
//
// prepare the tree
// - here the variable names have to corresponds to your tree
// - you can use the same variables as above which is slightly faster,
// but of course you can use different ones and copy the values inside the event loop
//
// efficiency calculator for cut method
Int_t nSelCutsGA = 0;
Double_t effS = 0.7;
std::cout << "--- Processing: " << theTree->GetEntries() << " events" << std::endl;
TStopwatch sw;
sw.Start();
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
cout<<"Entry # "<<ievt<<endl;
if (ievt%1000 == 0){
std::cout << "--- ... Processing event: " << ievt << std::endl;
}
theTree->GetEntry(ievt);
//
// return the MVAs and fill to histograms
//
/* if (Use["CutsGA"]) {
// Cuts is a special case: give the desired signal efficienciy
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
if (passed) nSelCutsGA++;
}
*/
if (Use["Likelihood" ]) histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) );
if (Use["Likelihood" ]) cout<<"MVA value: "<<reader->EvaluateMVA( "Likelihood method" )<<endl;
continue;
/*
if (Use["LikelihoodD" ]) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) );
if (Use["LikelihoodPCA"]) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) );
if (Use["LikelihoodKDE"]) histLkKDE ->Fill( reader->EvaluateMVA( "LikelihoodKDE method" ) );
if (Use["LikelihoodMIX"]) histLkMIX ->Fill( reader->EvaluateMVA( "LikelihoodMIX method" ) );
if (Use["PDERS" ]) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) );
if (Use["PDERSD" ]) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) );
if (Use["PDERSPCA" ]) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) );
if (Use["KNN" ]) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) );
if (Use["HMatrix" ]) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) );
if (Use["Fisher" ]) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) );
if (Use["FisherG" ]) histFiG ->Fill( reader->EvaluateMVA( "FisherG method" ) );
if (Use["BoostedFisher"]) histFiB ->Fill( reader->EvaluateMVA( "BoostedFisher method" ) );
if (Use["LD" ]) histLD ->Fill( reader->EvaluateMVA( "LD method" ) );
if (Use["MLP" ]) histNn ->Fill( reader->EvaluateMVA( "MLP method" ) );
if (Use["CFMlpANN" ]) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) );
if (Use["TMlpANN" ]) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) );
if (Use["BDT" ]) histBdt ->Fill( reader->EvaluateMVA( "BDT method" ) );
if (Use["BDTD" ]) histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) );
if (Use["BDTG" ]) histBdtG ->Fill( reader->EvaluateMVA( "BDTG method" ) );
if (Use["RuleFit" ]) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) );
if (Use["SVM_Gauss" ]) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) );
if (Use["SVM_Poly" ]) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) );
if (Use["SVM_Lin" ]) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) );
if (Use["FDA_MT" ]) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) );
if (Use["FDA_GA" ]) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) );
if (Use["Plugin" ]) histPBdt ->Fill( reader->EvaluateMVA( "P_BDT method" ) );
// retrieve also per-event error
if (Use["PDEFoam"]) {
Double_t val = reader->EvaluateMVA( "PDEFoam method" );
Double_t err = reader->GetMVAError();
histPDEFoam ->Fill( val );
histPDEFoamErr->Fill( err );
histPDEFoamSig->Fill( val/err );
}
// retrieve probability instead of MVA output
if (Use["Fisher"]) {
probHistFi ->Fill( reader->GetProba ( "Fisher method" ) );
rarityHistFi->Fill( reader->GetRarity( "Fisher method" ) );
}
*/
cout<<"End of processing"<<endl;
}
// get elapsed time
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// get efficiency for cuts classifier
if (Use["CutsGA"]) std::cout << "--- Efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
<< " (for a required signal efficiency of " << effS << ")" << std::endl;
if (Use["CutsGA"]) {
// test: retrieve cuts for particular signal efficiency
// CINT ignores dynamic_casts so we have to use a cuts-secific Reader function to acces the pointer
TMVA::MethodCuts* mcuts = reader->FindCutsMVA( "CutsGA method" ) ;
if (mcuts) {
std::vector<Double_t> cutsMin;
std::vector<Double_t> cutsMax;
mcuts->GetCuts( 0.7, cutsMin, cutsMax );
std::cout << "--- -------------------------------------------------------------" << std::endl;
std::cout << "--- Retrieve cut values for signal efficiency of 0.7 from Reader" << std::endl;
for (UInt_t ivar=0; ivar<cutsMin.size(); ivar++) {
std::cout << "... Cut: "
<< cutsMin[ivar]
<< " < \""
<< mcuts->GetInputVar(ivar)
<< "\" <= "
<< cutsMax[ivar] << std::endl;
}
std::cout << "--- -------------------------------------------------------------" << std::endl;
}
}
//
// write histograms
//
TFile *target = new TFile( filename+"_out"+".root","RECREATE" );
if (Use["Likelihood" ]) {if (_Weight>0)histLk->Scale(_Weight);histLk ->Write();}
if (Use["LikelihoodD" ]) {if (_Weight>0)histLkD->Scale(_Weight);histLkD ->Write();}
if (Use["LikelihoodPCA"]) {if (_Weight>0)histLkPCA->Scale(_Weight);histLkPCA ->Write();}
if (Use["LikelihoodKDE"]) {if (_Weight>0)histLkKDE->Scale(_Weight);histLkKDE ->Write();}
if (Use["LikelihoodMIX"]) {if (_Weight>0)histLkMIX->Scale(_Weight);histLkMIX ->Write();}
if (Use["PDERS" ]) histPD ->Write();
if (Use["PDERSD" ]) histPDD ->Write();
if (Use["PDERSPCA" ]) histPDPCA ->Write();
if (Use["KNN" ]) histKNN ->Write();
if (Use["HMatrix" ]) histHm ->Write();
if (Use["Fisher" ]) histFi ->Write();
if (Use["FisherG" ]) histFiG ->Write();
if (Use["BoostedFisher"]) histFiB ->Write();
if (Use["LD" ]) histLD ->Write();
if (Use["MLP" ]) histNn ->Write();
if (Use["CFMlpANN" ]) histNnC ->Write();
if (Use["TMlpANN" ]) histNnT ->Write();
if (Use["BDT" ]) histBdt ->Write();
if (Use["BDTD" ]) histBdtD ->Write();
if (Use["BDTG" ]) histBdtG ->Write();
if (Use["RuleFit" ]) histRf ->Write();
if (Use["SVM_Gauss" ]) histSVMG ->Write();
if (Use["SVM_Poly" ]) histSVMP ->Write();
if (Use["SVM_Lin" ]) histSVML ->Write();
if (Use["FDA_MT" ]) histFDAMT ->Write();
if (Use["FDA_GA" ]) histFDAGA ->Write();
if (Use["Plugin" ]) histPBdt ->Write();
// write also error and significance histos
if (Use["PDEFoam"]) { histPDEFoam->Write(); histPDEFoamErr->Write(); histPDEFoamSig->Write(); }
// write also probability hists
if (Use["Fisher"]) { if (probHistFi != 0) probHistFi->Write(); if (rarityHistFi != 0) rarityHistFi->Write(); }
target->Close();
std::cout << "--- Created root file: \"TMVApp.root\" containing the MVA output histograms" << std::endl;
delete reader;
std::cout << "==> TMVAClassificationApplication is done!" << endl << std::endl;
}
