UInt_t jobs = 4;

TMVA::Tools::Instance();

TFile *input(0);
TString fname = "./tmva_class_example.root";
if (!gSystem->AccessPathName(fname)) {
   input = TFile::Open(fname); // check if file in local directory exists
} else {
   TFile::SetCacheFileDir(".");
   input = TFile::Open("http://root.cern/files/tmva_class_example.root", "CACHEREAD");
}
if (!input) {
   std::cout << "ERROR: could not open data file" << std::endl;
   exit(1);
}

TTree *signalTree = (TTree *)input->Get("TreeS");
TTree *background = (TTree *)input->Get("TreeB");

TMVA::DataLoader *dataloader = new TMVA::DataLoader("dataset");

dataloader->AddVariable("myvar1 := var1+var2", 'F');
dataloader->AddVariable("myvar2 := var1-var2", "Expression 2", "", 'F');
dataloader->AddVariable("var3", "Variable 3", "units", 'F');
dataloader->AddVariable("var4", "Variable 4", "units", 'F');

dataloader->AddSpectator("spec1 := var1*2", "Spectator 1", "units", 'F');
dataloader->AddSpectator("spec2 := var1*3", "Spectator 2", "units", 'F');

Double_t signalWeight = 1.0;
Double_t backgroundWeight = 1.0;

dataloader->AddSignalTree(signalTree, signalWeight);
dataloader->AddBackgroundTree(background, backgroundWeight);

dataloader->SetBackgroundWeightExpression("weight");
dataloader->PrepareTrainingAndTestTree(
   "", "", "nTrain_Signal=1000:nTrain_Background=1000:SplitMode=Random:NormMode=NumEvents:!V");

TFile *outputFile = TFile::Open("TMVAClass.root", "RECREATE");

TMVA::Experimental::Classification *cl = new TMVA::Experimental::Classification(dataloader, Form("Jobs=%d", jobs));

cl->BookMethod(TMVA::Types::kBDT, "BDTG", "!H:!V:NTrees=2000:MinNodeSize=2.5%:BoostType=Grad:Shrinkage=0.10:"
                                          "UseBaggedBoost:BaggedSampleFraction=0.5:nCuts=20:MaxDepth=2");
cl->BookMethod(TMVA::Types::kSVM, "SVM", "Gamma=0.25:Tol=0.001:VarTransform=Norm");

cl->BookMethod(TMVA::Types::kBDT, "BDTB", "!H:!V:NTrees=2000:BoostType=Bagging:SeparationType=GiniIndex:nCuts=20");

cl->BookMethod(TMVA::Types::kCuts, "Cuts", "!H:!V:FitMethod=MC:EffSel:SampleSize=200000:VarProp=FSmart");

cl->Evaluate(); // Train and Test all methods

auto &results = cl->GetResults();

TCanvas *c = new TCanvas(Form("ROC"));
c->SetTitle("ROC-Integral Curve");

auto mg = new TMultiGraph();
for (UInt_t i = 0; i < results.size(); i++) {
   if (!results[i].IsCutsMethod()) {
      auto roc = results[i].GetROCGraph();
      roc->SetLineColorAlpha(i + 1, 0.1);
      mg->Add(roc);
   }
}
mg->Draw("AL");
mg->GetXaxis()->SetTitle(" Signal Efficiency ");
mg->GetYaxis()->SetTitle(" Background Rejection ");
c->BuildLegend(0.15, 0.15, 0.3, 0.3);
c->Draw();

outputFile->Close();
delete cl;