| 1 |
kukartse |
1.1 |
/**********************************************************************************
|
| 2 |
|
|
* Project : TMVA - a Root-integrated toolkit for multivariate data analysis *
|
| 3 |
|
|
* Package : TMVA *
|
| 4 |
|
|
* Exectuable: TMVApplication *
|
| 5 |
|
|
* *
|
| 6 |
|
|
* This macro provides a simple example on how to use the trained classifiers *
|
| 7 |
|
|
* within an analysis module *
|
| 8 |
|
|
* *
|
| 9 |
|
|
* ------------------------------------------------------------------------------ *
|
| 10 |
|
|
* see also the alternative (slightly faster) way to retrieve the MVA values in *
|
| 11 |
|
|
* examples/TMVApplicationAlternative.cxx *
|
| 12 |
|
|
* ------------------------------------------------------------------------------ *
|
| 13 |
|
|
**********************************************************************************/
|
| 14 |
|
|
|
| 15 |
|
|
// ---------------------------------------------------------------
|
| 16 |
|
|
// choose MVA methods to be applied
|
| 17 |
|
|
Bool_t Use_Cuts = 0;
|
| 18 |
|
|
Bool_t Use_CutsD = 0;
|
| 19 |
|
|
Bool_t Use_CutsGA = 1;
|
| 20 |
|
|
Bool_t Use_Likelihood = 1;
|
| 21 |
|
|
Bool_t Use_LikelihoodD = 0; // the "D" extension indicates decorrelated input variables (see option strings)
|
| 22 |
|
|
Bool_t Use_LikelihoodPCA = 1; // the "PCA" extension indicates PCA-transformed input variables (see option strings)
|
| 23 |
|
|
Bool_t Use_PDERS = 1;
|
| 24 |
|
|
Bool_t Use_PDERSD = 0;
|
| 25 |
|
|
Bool_t Use_PDERSPCA = 0;
|
| 26 |
|
|
Bool_t Use_KNN = 1;
|
| 27 |
|
|
Bool_t Use_HMatrix = 1;
|
| 28 |
|
|
Bool_t Use_Fisher = 1;
|
| 29 |
|
|
Bool_t Use_FDA_GA = 0;
|
| 30 |
|
|
Bool_t Use_FDA_MT = 1;
|
| 31 |
|
|
Bool_t Use_MLP = 1; // this is the recommended ANN
|
| 32 |
|
|
Bool_t Use_CFMlpANN = 0;
|
| 33 |
|
|
Bool_t Use_TMlpANN = 0;
|
| 34 |
|
|
Bool_t Use_SVM_Gauss = 1;
|
| 35 |
|
|
Bool_t Use_SVM_Poly = 0;
|
| 36 |
|
|
Bool_t Use_SVM_Lin = 0;
|
| 37 |
|
|
Bool_t Use_BDT = 1;
|
| 38 |
|
|
Bool_t Use_BDTD = 0;
|
| 39 |
|
|
Bool_t Use_RuleFit = 1;
|
| 40 |
|
|
// ---------------------------------------------------------------
|
| 41 |
|
|
|
| 42 |
|
|
void TMVApplication( TString myMethodList = "" )
|
| 43 |
|
|
{
|
| 44 |
|
|
cout << endl;
|
| 45 |
|
|
cout << "==> start TMVApplication" << endl;
|
| 46 |
|
|
|
| 47 |
|
|
if (myMethodList != "") {
|
| 48 |
|
|
Use_CutsGA = Use_CutsD = Use_Cuts
|
| 49 |
|
|
= Use_LikelihoodPCA = Use_LikelihoodD = Use_Likelihood
|
| 50 |
|
|
= Use_PDERSPCA = Use_PDERSD = Use_PDERS
|
| 51 |
|
|
= Use_KNN
|
| 52 |
|
|
= Use_MLP = Use_CFMlpANN = Use_TMlpANN
|
| 53 |
|
|
= Use_HMatrix = Use_Fisher = Use_BDTD = Use_BDT = Use_RuleFit
|
| 54 |
|
|
= Use_SVM_Gauss = Use_SVM_Poly = Use_SVM_Lin
|
| 55 |
|
|
= Use_FDA_GA = Use_FDA_MT
|
| 56 |
|
|
= 0;
|
| 57 |
|
|
|
| 58 |
|
|
TList* mlist = TMVA::Tools::ParseFormatLine( myMethodList, " :," );
|
| 59 |
|
|
|
| 60 |
|
|
if (mlist->FindObject( "Cuts" ) != 0) Use_Cuts = 1;
|
| 61 |
|
|
if (mlist->FindObject( "CutsD" ) != 0) Use_CutsD = 1;
|
| 62 |
|
|
if (mlist->FindObject( "CutsGA" ) != 0) Use_CutsGA = 1;
|
| 63 |
|
|
if (mlist->FindObject( "Likelihood" ) != 0) Use_Likelihood = 1;
|
| 64 |
|
|
if (mlist->FindObject( "LikelihoodD" ) != 0) Use_LikelihoodD = 1;
|
| 65 |
|
|
if (mlist->FindObject( "LikelihoodPCA" ) != 0) Use_LikelihoodPCA = 1;
|
| 66 |
|
|
if (mlist->FindObject( "PDERS" ) != 0) Use_PDERS = 1;
|
| 67 |
|
|
if (mlist->FindObject( "PDERSD" ) != 0) Use_PDERSD = 1;
|
| 68 |
|
|
if (mlist->FindObject( "PDERSPCA" ) != 0) Use_PDERSPCA = 1;
|
| 69 |
|
|
if (mlist->FindObject( "KNN" ) != 0) Use_KNN = 1;
|
| 70 |
|
|
if (mlist->FindObject( "HMatrix" ) != 0) Use_HMatrix = 1;
|
| 71 |
|
|
if (mlist->FindObject( "Fisher" ) != 0) Use_Fisher = 1;
|
| 72 |
|
|
if (mlist->FindObject( "MLP" ) != 0) Use_MLP = 1;
|
| 73 |
|
|
if (mlist->FindObject( "CFMlpANN" ) != 0) Use_CFMlpANN = 1;
|
| 74 |
|
|
if (mlist->FindObject( "TMlpANN" ) != 0) Use_TMlpANN = 1;
|
| 75 |
|
|
if (mlist->FindObject( "BDTD" ) != 0) Use_BDTD = 1;
|
| 76 |
|
|
if (mlist->FindObject( "BDT" ) != 0) Use_BDT = 1;
|
| 77 |
|
|
if (mlist->FindObject( "RuleFit" ) != 0) Use_RuleFit = 1;
|
| 78 |
|
|
if (mlist->FindObject( "SVM_Gauss" ) != 0) Use_SVM_Gauss = 1;
|
| 79 |
|
|
if (mlist->FindObject( "SVM_Poly" ) != 0) Use_SVM_Poly = 1;
|
| 80 |
|
|
if (mlist->FindObject( "SVM_Lin" ) != 0) Use_SVM_Lin = 1;
|
| 81 |
|
|
if (mlist->FindObject( "FDA_MT" ) != 0) Use_FDA_MT = 1;
|
| 82 |
|
|
if (mlist->FindObject( "FDA_GA" ) != 0) Use_FDA_GA = 1;
|
| 83 |
|
|
|
| 84 |
|
|
delete mlist;
|
| 85 |
|
|
}
|
| 86 |
|
|
|
| 87 |
|
|
//
|
| 88 |
|
|
// create the Reader object
|
| 89 |
|
|
//
|
| 90 |
|
|
TMVA::Reader *reader = new TMVA::Reader("!Color");
|
| 91 |
|
|
|
| 92 |
|
|
// create a set of variables and declare them to the reader
|
| 93 |
|
|
// - the variable names must corresponds in name and type to
|
| 94 |
|
|
// those given in the weight file(s) that you use
|
| 95 |
|
|
Float_t var1, var2;
|
| 96 |
|
|
Float_t var3, var4;
|
| 97 |
|
|
reader->AddVariable( "var1+var2", &var1 );
|
| 98 |
|
|
reader->AddVariable( "var1-var2", &var2 );
|
| 99 |
|
|
reader->AddVariable( "var3", &var3 );
|
| 100 |
|
|
reader->AddVariable( "var4", &var4 );
|
| 101 |
|
|
|
| 102 |
|
|
//
|
| 103 |
|
|
// book the MVA methods
|
| 104 |
|
|
//
|
| 105 |
|
|
string dir = "weights/";
|
| 106 |
|
|
string prefix = "TMVAnalysis";
|
| 107 |
|
|
|
| 108 |
|
|
if (Use_Cuts) reader->BookMVA( "Cuts method", dir + prefix + "_Cuts.weights.txt" );
|
| 109 |
|
|
if (Use_CutsD) reader->BookMVA( "CutsD method", dir + prefix + "_CutsD.weights.txt" );
|
| 110 |
|
|
if (Use_CutsGA) reader->BookMVA( "CutsGA method", dir + prefix + "_CutsGA.weights.txt" );
|
| 111 |
|
|
if (Use_Likelihood) reader->BookMVA( "Likelihood method", dir + prefix + "_Likelihood.weights.txt" );
|
| 112 |
|
|
if (Use_LikelihoodD) reader->BookMVA( "LikelihoodD method", dir + prefix + "_LikelihoodD.weights.txt" );
|
| 113 |
|
|
if (Use_LikelihoodPCA) reader->BookMVA( "LikelihoodPCA method", dir + prefix + "_LikelihoodPCA.weights.txt" );
|
| 114 |
|
|
if (Use_PDERS) reader->BookMVA( "PDERS method", dir + prefix + "_PDERS.weights.txt" );
|
| 115 |
|
|
if (Use_PDERSD) reader->BookMVA( "PDERSD method", dir + prefix + "_PDERSD.weights.txt" );
|
| 116 |
|
|
if (Use_PDERSPCA) reader->BookMVA( "PDERSPCA method", dir + prefix + "_PDERSPCA.weights.txt" );
|
| 117 |
|
|
if (Use_KNN) reader->BookMVA( "KNN method", dir + prefix + "_KNN.weights.txt" );
|
| 118 |
|
|
if (Use_HMatrix) reader->BookMVA( "HMatrix method", dir + prefix + "_HMatrix.weights.txt" );
|
| 119 |
|
|
if (Use_Fisher) reader->BookMVA( "Fisher method", dir + prefix + "_Fisher.weights.txt" );
|
| 120 |
|
|
if (Use_MLP) reader->BookMVA( "MLP method", dir + prefix + "_MLP.weights.txt" );
|
| 121 |
|
|
if (Use_CFMlpANN) reader->BookMVA( "CFMlpANN method", dir + prefix + "_CFMlpANN.weights.txt" );
|
| 122 |
|
|
if (Use_TMlpANN) reader->BookMVA( "TMlpANN method", dir + prefix + "_TMlpANN.weights.txt" );
|
| 123 |
|
|
if (Use_BDT) reader->BookMVA( "BDT method", dir + prefix + "_BDT.weights.txt" );
|
| 124 |
|
|
if (Use_BDTD) reader->BookMVA( "BDTD method", dir + prefix + "_BDTD.weights.txt" );
|
| 125 |
|
|
if (Use_RuleFit) reader->BookMVA( "RuleFit method", dir + prefix + "_RuleFitTMVA.weights.txt" );
|
| 126 |
|
|
if (Use_SVM_Gauss) reader->BookMVA( "SVM_Gauss method", dir + prefix + "_SVM_Gauss.weights.txt" );
|
| 127 |
|
|
if (Use_SVM_Poly) reader->BookMVA( "SVM_Poly method", dir + prefix + "_SVM_Poly.weights.txt" );
|
| 128 |
|
|
if (Use_SVM_Lin) reader->BookMVA( "SVM_Lin method", dir + prefix + "_SVM_Lin.weights.txt" );
|
| 129 |
|
|
if (Use_FDA_MT) reader->BookMVA( "FDA_MT method", dir + prefix + "_FDA_MT.weights.txt" );
|
| 130 |
|
|
if (Use_FDA_GA) reader->BookMVA( "FDA_GA method", dir + prefix + "_FDA_GA.weights.txt" );
|
| 131 |
|
|
|
| 132 |
|
|
// book output histograms
|
| 133 |
|
|
UInt_t nbin = 100;
|
| 134 |
|
|
TH1F *histLk, *histLkD, *histLkPCA, *histPD, *histPDD, *histPDPCA, *histKNN, *histHm, *histFi;
|
| 135 |
|
|
TH1F *histNn, *histNnC, *histNnT, *histBdt, *histBdtD, *histRf;
|
| 136 |
|
|
TH1F *histSVMG, *histSVMP, *histSVML;
|
| 137 |
|
|
TH1F *histFDAMT, *histFDAGA;
|
| 138 |
|
|
|
| 139 |
|
|
if (Use_Likelihood) histLk = new TH1F( "MVA_Likelihood", "MVA_Likelihood", nbin, 0, 1 );
|
| 140 |
|
|
if (Use_LikelihoodD) histLkD = new TH1F( "MVA_LikelihoodD", "MVA_LikelihoodD", nbin, 0.000001, 0.9999 );
|
| 141 |
|
|
if (Use_LikelihoodPCA) histLkPCA = new TH1F( "MVA_LikelihoodPCA", "MVA_LikelihoodPCA", nbin, 0, 1 );
|
| 142 |
|
|
if (Use_PDERS) histPD = new TH1F( "MVA_PDERS", "MVA_PDERS", nbin, 0, 1 );
|
| 143 |
|
|
if (Use_PDERSD) histPDD = new TH1F( "MVA_PDERSD", "MVA_PDERSD", nbin, 0, 1 );
|
| 144 |
|
|
if (Use_PDERSPCA) histPDPCA = new TH1F( "MVA_PDERSPCA", "MVA_PDERSPCA", nbin, 0, 1 );
|
| 145 |
|
|
if (Use_KNN) histKNN = new TH1F( "MVA_KNN", "MVA_KNN", nbin, 0, 1 );
|
| 146 |
|
|
if (Use_HMatrix) histHm = new TH1F( "MVA_HMatrix", "MVA_HMatrix", nbin, -0.95, 1.55 );
|
| 147 |
|
|
if (Use_Fisher) histFi = new TH1F( "MVA_Fisher", "MVA_Fisher", nbin, -4, 4 );
|
| 148 |
|
|
if (Use_MLP) histNn = new TH1F( "MVA_MLP", "MVA_MLP", nbin, -0.25, 1.5 );
|
| 149 |
|
|
if (Use_CFMlpANN) histNnC = new TH1F( "MVA_CFMlpANN", "MVA_CFMlpANN", nbin, 0, 1 );
|
| 150 |
|
|
if (Use_TMlpANN) histNnT = new TH1F( "MVA_TMlpANN", "MVA_TMlpANN", nbin, -1.3, 1.3 );
|
| 151 |
|
|
if (Use_BDT) histBdt = new TH1F( "MVA_BDT", "MVA_BDT", nbin, -0.8, 0.8 );
|
| 152 |
|
|
if (Use_BDTD) histBdtD = new TH1F( "MVA_BDTD", "MVA_BDTD", nbin, -0.4, 0.6 );
|
| 153 |
|
|
if (Use_RuleFit) histRf = new TH1F( "MVA_RuleFitTMVA", "MVA_RuleFitTMVA", nbin, -2.0, 2.0 );
|
| 154 |
|
|
if (Use_SVM_Gauss) histSVMG = new TH1F( "MVA_SVM_Gauss", "MVA_SVM_Gauss", nbin, 0.0, 1.0 );
|
| 155 |
|
|
if (Use_SVM_Poly) histSVMP = new TH1F( "MVA_SVM_Poly", "MVA_SVM_Poly", nbin, 0.0, 1.0 );
|
| 156 |
|
|
if (Use_SVM_Lin) histSVML = new TH1F( "MVA_SVM_Lin", "MVA_SVM_Lin", nbin, 0.0, 1.0 );
|
| 157 |
|
|
if (Use_FDA_MT) histFDAMT = new TH1F( "MVA_FDA_MT", "MVA_FDA_MT", nbin, -2.0, 3.0 );
|
| 158 |
|
|
if (Use_FDA_GA) histFDAGA = new TH1F( "MVA_FDA_GA", "MVA_FDA_GA", nbin, -2.0, 3.0 );
|
| 159 |
|
|
|
| 160 |
|
|
// book examsple histogram for probability (the other methods are done similarly)
|
| 161 |
|
|
TH1F *probHistFi, *rarityHistFi;
|
| 162 |
|
|
if (Use_Fisher) {
|
| 163 |
|
|
probHistFi = new TH1F( "PROBA_MVA_Fisher", "PROBA_MVA_Fisher", nbin, 0, 1 );
|
| 164 |
|
|
rarityHistFi = new TH1F( "RARITY_MVA_Fisher", "RARITY_MVA_Fisher", nbin, 0, 1 );
|
| 165 |
|
|
}
|
| 166 |
|
|
|
| 167 |
|
|
// Prepare input tree (this must be replaced by your data source)
|
| 168 |
|
|
// in this example, there is a toy tree with signal and one with background events
|
| 169 |
|
|
// we'll later on use only the "signal" events for the test in this example.
|
| 170 |
|
|
//
|
| 171 |
|
|
TFile *input(0);
|
| 172 |
|
|
TString fname = "./tmva_example.root";
|
| 173 |
|
|
if (!gSystem->AccessPathName( fname )) {
|
| 174 |
|
|
// first we try to find tmva_example.root in the local directory
|
| 175 |
|
|
cout << "--- accessing data file: " << fname << endl;
|
| 176 |
|
|
input = TFile::Open( fname );
|
| 177 |
|
|
}
|
| 178 |
|
|
else {
|
| 179 |
|
|
// second we try accessing the file via the web from
|
| 180 |
|
|
// http://root.cern.ch/files/tmva_example.root
|
| 181 |
|
|
cout << "--- accessing tmva_example.root file from http://root.cern.ch/files" << endl;
|
| 182 |
|
|
cout << "--- for faster startup you may consider downloading it into you local directory" << endl;
|
| 183 |
|
|
input = TFile::Open("http://root.cern.ch/files/tmva_example.root");
|
| 184 |
|
|
}
|
| 185 |
|
|
|
| 186 |
|
|
if (!input) {
|
| 187 |
|
|
std::cout << "ERROR: could not open data file: " << fname << std::endl;
|
| 188 |
|
|
exit(1);
|
| 189 |
|
|
}
|
| 190 |
|
|
|
| 191 |
|
|
//
|
| 192 |
|
|
// prepare the tree
|
| 193 |
|
|
// - here the variable names have to corresponds to your tree
|
| 194 |
|
|
// - you can use the same variables as above which is slightly faster,
|
| 195 |
|
|
// but of course you can use different ones and copy the values inside the event loop
|
| 196 |
|
|
//
|
| 197 |
|
|
TTree* theTree = (TTree*)input->Get("TreeS");
|
| 198 |
|
|
cout << "--- select signal sample" << endl;
|
| 199 |
|
|
Float_t userVar1, userVar2;
|
| 200 |
|
|
theTree->SetBranchAddress( "var1", &userVar1 );
|
| 201 |
|
|
theTree->SetBranchAddress( "var2", &userVar2 );
|
| 202 |
|
|
theTree->SetBranchAddress( "var3", &var3 );
|
| 203 |
|
|
theTree->SetBranchAddress( "var4", &var4 );
|
| 204 |
|
|
|
| 205 |
|
|
// efficiency calculator for cut method
|
| 206 |
|
|
Int_t nSelCuts = 0, nSelCutsD = 0, nSelCutsGA = 0;
|
| 207 |
|
|
Double_t effS = 0.7;
|
| 208 |
|
|
|
| 209 |
|
|
cout << "--- processing: " << theTree->GetEntries() << " events" << endl;
|
| 210 |
|
|
TStopwatch sw;
|
| 211 |
|
|
sw.Start();
|
| 212 |
|
|
for (Long64_t ievt=0; ievt<theTree->GetEntries();ievt++) {
|
| 213 |
|
|
|
| 214 |
|
|
if (ievt%1000 == 0)
|
| 215 |
|
|
cout << "--- ... processing event: " << ievt << endl;
|
| 216 |
|
|
|
| 217 |
|
|
theTree->GetEntry(ievt);
|
| 218 |
|
|
|
| 219 |
|
|
var1 = userVar1 + userVar2;
|
| 220 |
|
|
var2 = userVar1 - userVar2;
|
| 221 |
|
|
|
| 222 |
|
|
//
|
| 223 |
|
|
// return the MVAs and fill to histograms
|
| 224 |
|
|
//
|
| 225 |
|
|
if (Use_Cuts) {
|
| 226 |
|
|
// Cuts is a special case: give the desired signal efficienciy
|
| 227 |
|
|
Bool_t passed = reader->EvaluateMVA( "Cuts method", effS );
|
| 228 |
|
|
if (passed) nSelCuts++;
|
| 229 |
|
|
}
|
| 230 |
|
|
if (Use_CutsD) {
|
| 231 |
|
|
// Cuts is a special case: give the desired signal efficienciy
|
| 232 |
|
|
Bool_t passed = reader->EvaluateMVA( "CutsD method", effS );
|
| 233 |
|
|
if (passed) nSelCutsD++;
|
| 234 |
|
|
}
|
| 235 |
|
|
if (Use_CutsGA) {
|
| 236 |
|
|
// Cuts is a special case: give the desired signal efficienciy
|
| 237 |
|
|
Bool_t passed = reader->EvaluateMVA( "CutsGA method", effS );
|
| 238 |
|
|
if (passed) nSelCutsGA++;
|
| 239 |
|
|
}
|
| 240 |
|
|
|
| 241 |
|
|
if (Use_Likelihood ) histLk ->Fill( reader->EvaluateMVA( "Likelihood method" ) );
|
| 242 |
|
|
if (Use_LikelihoodD ) histLkD ->Fill( reader->EvaluateMVA( "LikelihoodD method" ) );
|
| 243 |
|
|
if (Use_LikelihoodPCA) histLkPCA ->Fill( reader->EvaluateMVA( "LikelihoodPCA method" ) );
|
| 244 |
|
|
if (Use_PDERS ) histPD ->Fill( reader->EvaluateMVA( "PDERS method" ) );
|
| 245 |
|
|
if (Use_PDERSD ) histPDD ->Fill( reader->EvaluateMVA( "PDERSD method" ) );
|
| 246 |
|
|
if (Use_PDERSPCA ) histPDPCA ->Fill( reader->EvaluateMVA( "PDERSPCA method" ) );
|
| 247 |
|
|
if (Use_KNN ) histKNN ->Fill( reader->EvaluateMVA( "KNN method" ) );
|
| 248 |
|
|
if (Use_HMatrix ) histHm ->Fill( reader->EvaluateMVA( "HMatrix method" ) );
|
| 249 |
|
|
if (Use_Fisher ) histFi ->Fill( reader->EvaluateMVA( "Fisher method" ) );
|
| 250 |
|
|
if (Use_MLP ) histNn ->Fill( reader->EvaluateMVA( "MLP method" ) );
|
| 251 |
|
|
if (Use_CFMlpANN ) histNnC ->Fill( reader->EvaluateMVA( "CFMlpANN method" ) );
|
| 252 |
|
|
if (Use_TMlpANN ) histNnT ->Fill( reader->EvaluateMVA( "TMlpANN method" ) );
|
| 253 |
|
|
if (Use_BDT ) histBdt ->Fill( reader->EvaluateMVA( "BDT method" ) );
|
| 254 |
|
|
if (Use_BDTD ) histBdtD ->Fill( reader->EvaluateMVA( "BDTD method" ) );
|
| 255 |
|
|
if (Use_RuleFit ) histRf ->Fill( reader->EvaluateMVA( "RuleFit method" ) );
|
| 256 |
|
|
if (Use_SVM_Gauss ) histSVMG ->Fill( reader->EvaluateMVA( "SVM_Gauss method" ) );
|
| 257 |
|
|
if (Use_SVM_Poly ) histSVMP ->Fill( reader->EvaluateMVA( "SVM_Poly method" ) );
|
| 258 |
|
|
if (Use_SVM_Lin ) histSVML ->Fill( reader->EvaluateMVA( "SVM_Lin method" ) );
|
| 259 |
|
|
if (Use_FDA_MT ) histFDAMT ->Fill( reader->EvaluateMVA( "FDA_MT method" ) );
|
| 260 |
|
|
if (Use_FDA_GA ) histFDAGA ->Fill( reader->EvaluateMVA( "FDA_GA method" ) );
|
| 261 |
|
|
|
| 262 |
|
|
// retrieve probability instead of MVA output
|
| 263 |
|
|
if (Use_Fisher ) {
|
| 264 |
|
|
probHistFi ->Fill( reader->GetProba ( "Fisher method" ) );
|
| 265 |
|
|
rarityHistFi->Fill( reader->GetRarity( "Fisher method" ) );
|
| 266 |
|
|
}
|
| 267 |
|
|
}
|
| 268 |
|
|
sw.Stop();
|
| 269 |
|
|
cout << "--- end of event loop: "; sw.Print();
|
| 270 |
|
|
// get elapsed time
|
| 271 |
|
|
if (Use_Cuts) cout << "--- efficiency for Cuts method : " << double(nSelCuts)/theTree->GetEntries()
|
| 272 |
|
|
<< " (for a required signal efficiency of " << effS << ")" << endl;
|
| 273 |
|
|
if (Use_CutsD) cout << "--- efficiency for CutsD method : " << double(nSelCutsD)/theTree->GetEntries()
|
| 274 |
|
|
<< " (for a required signal efficiency of " << effS << ")" << endl;
|
| 275 |
|
|
if (Use_CutsGA) cout << "--- efficiency for CutsGA method: " << double(nSelCutsGA)/theTree->GetEntries()
|
| 276 |
|
|
<< " (for a required signal efficiency of " << effS << ")" << endl;
|
| 277 |
|
|
|
| 278 |
|
|
//
|
| 279 |
|
|
// write histograms
|
| 280 |
|
|
//
|
| 281 |
|
|
TFile *target = new TFile( "TMVApp.root","RECREATE" );
|
| 282 |
|
|
if (Use_Likelihood ) histLk ->Write();
|
| 283 |
|
|
if (Use_LikelihoodD ) histLkD ->Write();
|
| 284 |
|
|
if (Use_LikelihoodPCA) histLkPCA ->Write();
|
| 285 |
|
|
if (Use_PDERS ) histPD ->Write();
|
| 286 |
|
|
if (Use_PDERSD ) histPDD ->Write();
|
| 287 |
|
|
if (Use_PDERSPCA ) histPDPCA ->Write();
|
| 288 |
|
|
if (Use_KNN ) histKNN ->Write();
|
| 289 |
|
|
if (Use_HMatrix ) histHm ->Write();
|
| 290 |
|
|
if (Use_Fisher ) histFi ->Write();
|
| 291 |
|
|
if (Use_MLP ) histNn ->Write();
|
| 292 |
|
|
if (Use_CFMlpANN ) histNnC ->Write();
|
| 293 |
|
|
if (Use_TMlpANN ) histNnT ->Write();
|
| 294 |
|
|
if (Use_BDT ) histBdt ->Write();
|
| 295 |
|
|
if (Use_BDTD ) histBdtD ->Write();
|
| 296 |
|
|
if (Use_RuleFit ) histRf ->Write();
|
| 297 |
|
|
if (Use_SVM_Gauss ) histSVMG ->Write();
|
| 298 |
|
|
if (Use_SVM_Poly ) histSVMP ->Write();
|
| 299 |
|
|
if (Use_SVM_Lin ) histSVML ->Write();
|
| 300 |
|
|
if (Use_FDA_MT ) histFDAMT ->Write();
|
| 301 |
|
|
if (Use_FDA_GA ) histFDAGA ->Write();
|
| 302 |
|
|
|
| 303 |
|
|
if (Use_Fisher ) { probHistFi->Write(); rarityHistFi->Write(); }
|
| 304 |
|
|
target->Close();
|
| 305 |
|
|
|
| 306 |
|
|
cout << "--- created root file: \"TMVApp.root\" containing the MVA output histograms" << endl;
|
| 307 |
|
|
|
| 308 |
|
|
delete reader;
|
| 309 |
|
|
|
| 310 |
|
|
cout << "==> TMVApplication is done!" << endl << endl;
|
| 311 |
|
|
}
|
