void TMVARegressionApplication( int wMs,int wM, string st,string st2,string option="",TString myMethodList = "" )
{
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
// --- Mutidimensional likelihood and Nearest-Neighbour methods
Use["PDERS"] = 0;
Use["PDEFoam"] = 0;
Use["KNN"] = 0;
//
// --- Linear Discriminant Analysis
Use["LD"] = 0;
//
// --- Function Discriminant analysis
Use["FDA_GA"] = 0;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 0;
Use["FDA_GAMT"] = 0;
//
// --- Neural Network
Use["MLP"] = 0;
//
// --- Support Vector Machine
Use["SVM"] = 0;
//
// --- Boosted Decision Trees
Use["BDT"] = 0;
Use["BDTG"] = 1;
// ---------------------------------------------------------------
std::cout << std::endl;
std::cout << "==> Start TMVARegressionApplication" << 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;
//reader->AddVariable( "var1", &var1 );
//reader->AddVariable( "var2", &var2 );
Float_t pt_AK8MatchedToHbb,eta_AK8MatchedToHbb,nsv_AK8MatchedToHbb,sv0mass_AK8MatchedToHbb,sv1mass_AK8MatchedToHbb,
nch_AK8MatchedToHbb,nmu_AK8MatchedToHbb,nel_AK8MatchedToHbb,muenfr_AK8MatchedToHbb,emenfr_AK8MatchedToHbb;
reader->AddVariable( "pt_AK8MatchedToHbb", &pt_AK8MatchedToHbb );
reader->AddVariable( "eta_AK8MatchedToHbb", &eta_AK8MatchedToHbb );
reader->AddVariable( "nsv_AK8MatchedToHbb", &nsv_AK8MatchedToHbb );
reader->AddVariable( "sv0mass_AK8MatchedToHbb", &sv0mass_AK8MatchedToHbb );
reader->AddVariable( "sv1mass_AK8MatchedToHbb", &sv1mass_AK8MatchedToHbb );
reader->AddVariable( "nch_AK8MatchedToHbb", &nch_AK8MatchedToHbb );
reader->AddVariable( "nmu_AK8MatchedToHbb", &nmu_AK8MatchedToHbb );
reader->AddVariable( "nel_AK8MatchedToHbb", &nel_AK8MatchedToHbb );
reader->AddVariable( "muenfr_AK8MatchedToHbb", &muenfr_AK8MatchedToHbb );
reader->AddVariable( "emenfr_AK8MatchedToHbb", &emenfr_AK8MatchedToHbb );
// Spectator variables declared in the training have to be added to the reader, too
Float_t spec1,spec2;
reader->AddSpectator( "spec1:=n_pv", &spec1 );
reader->AddSpectator( "spec2:=msoftdrop_AK8MatchedToHbb", &spec2 );
// --- Book the MVA methods
TString dir = "weights/";
TString prefix = "TMVARegression";
// 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 );
}
}
TH1* hists[100];
Int_t nhists = -1;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
TH1* h = new TH1F( it->first.c_str(), TString(it->first) + " method", 100, -100, 600 );
if (it->second) hists[++nhists] = h;
}
nhists++;
//1=signal ,0=QCD ,2=data
int nameRoot=1;
if((st2.find("QCD")!= std::string::npos)||
(st2.find("bGen")!= std::string::npos)||
(st2.find("bEnriched")!= std::string::npos))nameRoot=0;
if(st2.find("data")!= std::string::npos)nameRoot=2;
cout<<"nameRoot = "<<nameRoot<<endl;
//option-----------------------------------------------------------
int JESOption=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 *f;
TTree *tree;
int nPass[20]={0};
int total=0;
double fixScaleNum[2]={0};
TH1D* th1=new TH1D("a","a",150,50,200);
string massName[nMass]={"Thea","HCorr","Reg"};
string catName[nCat]={"PP","PF","FP","FF"};
string tau21Name[2]={"withTau21","woTau21"};
string catNameShort[nCat]={"P","F"};
string looseTight[2]={"loose","tight"};
TF1 *fa[nMass][2][2][2];
for(int i=0;i<nMass;i++){
for(int j=0;j<2;j++){
for(int k=0;k<2;k++){
for(int w=0;w<2;w++){
fa[i][j][k][w] = new TF1("fa","[0]+[1]*x+[2]*x*x+[3]*pow(x,3)",-3,3);
ifstream myfile (Form("PFRatio/%s_%s_%s_%s.txt",looseTight[w].data(),massName[i].data(),catNameShort[j].data(),tau21Name[k].data()));
double para[4];
for(int m=0;m<4;m++){
myfile>>para[m];
}
fa[i][j][k][w]->SetParameters(para[0],para[1],para[2],para[3]);
}
}
}
}
/*
TH1D* th2[nMass][nCat][2];
TH1D* th3[nMass][nCat][2];
for(int i=0;i<nMass;i++){
for(int j=0;j<nCat;j++){
for(int k=0;k<2;k++){
th2[i][j][k]=(TH1D*)th1->Clone(Form("loose_%s_%s_%s",massName[i].data(),catName[j].data(),tau21Name[k].data()));
th3[i][j][k]=(TH1D*)th1->Clone(Form("tight_%s_%s_%s",massName[i].data(),catName[j].data(),tau21Name[k].data()));
th2[i][j][k]->Sumw2();
th3[i][j][k]->Sumw2();
}
}
}
*/
TH1D* th2d[14];
th2d[0]=new TH1D("0a","0a",4000,1000,5000);
th2d[1]=new TH1D("0c","0c",4000,1000,5000);
th2d[2]=new TH1D("1a","1a",4000,1000,5000);
th2d[3]=new TH1D("1c","1c",4000,1000,5000);
th2d[4]=new TH1D("2a","2a",4000,1000,5000);
th2d[5]=new TH1D("2b","2b",4000,1000,5000);
th2d[6]=new TH1D("2d","2d",4000,1000,5000);
th2d[7]=new TH1D("0aL","0aL",4000,1000,5000);
th2d[8]=new TH1D("0cL","0cL",4000,1000,5000);
th2d[9]=new TH1D("1aL","1aL",4000,1000,5000);
th2d[10]=new TH1D("1cL","1cL",4000,1000,5000);
th2d[11]=new TH1D("2aL","2aL",4000,1000,5000);
th2d[12]=new TH1D("2bL","2bL",4000,1000,5000);
th2d[13]=new TH1D("2dL","2dL",4000,1000,5000);
//int nWidth=5,nBmin=11;
int width [nWidth]={25,30,35,40};
int bmin[nBmin]={100,105,110,115};
TH1D* th3d[14][nWidth][nBmin][2];
TH1D* th3f[14][nWidth][nBmin][2];
TH1D* th3v[14][nWidth][nBmin][2];
for(int i=0;i<nWidth;i++){
for(int j=0;j<nBmin;j++){
for(int k=0;k<2;k++){
for(int l=0;l<14;l++){
th3d[l][i][j][k]=(TH1D*) th2d[l]->Clone(Form("%s_%d_%d_%s",th2d[l]->GetTitle(),bmin[j],width[i]+bmin[j],tau21Name[k].data()));
th3f[l][i][j][k]=(TH1D*) th2d[l]->Clone(Form("fill_%s_%d_%d_%s",th2d[l]->GetTitle(),bmin[j],width[i]+bmin[j],tau21Name[k].data()));
th3v[l][i][j][k]=(TH1D*) th2d[l]->Clone(Form("valid_%s_%d_%d_%s",th2d[l]->GetTitle(),bmin[j],width[i]+bmin[j],tau21Name[k].data()));
th3d[l][i][j][k]->Sumw2();
th3f[l][i][j][k]->Sumw2();
th3v[l][i][j][k]->Sumw2();
}
}
}
}
for (int w=wMs;w<wM;w++){
if(w%20==0)cout<<w<<endl;
if (nameRoot!=1)f = TFile::Open(Form("%s%d.root",st.data(),w));
else f = TFile::Open(st.data());
if (!f || !f->IsOpen())continue;
TDirectory * dir;
if (nameRoot!=1)dir = (TDirectory*)f->Get(Form("%s%d.root:/tree",st.data(),w));
else dir = (TDirectory*)f->Get(Form("%s:/tree",st.data()));
dir->GetObject("treeMaker",tree);
//tree=(TTree*)f->Get("treeMaker");
TreeReader data(tree);
total+=data.GetEntriesFast();
for(Long64_t jEntry=0; jEntry<data.GetEntriesFast() ;jEntry++){
data.GetEntry(jEntry);
Int_t nVtx = data.GetInt("nVtx");
//0. has a good vertex
if(nVtx<1)continue;
nPass[0]++;
//1.trigger
std::string* trigName = data.GetPtrString("hlt_trigName");
vector<bool> &trigResult = *((vector<bool>*) data.GetPtr("hlt_trigResult"));
bool passTrigger=false;
for(int it=0; it< data.GetPtrStringSize();it++){
std::string thisTrig= trigName[it];
bool results = trigResult[it];
if( ((thisTrig.find("HLT_PFHT800")!= std::string::npos||
thisTrig.find("HLT_AK8DiPFJet300_200_TrimMass30_BTagCSV_p20")!= std::string::npos
) && results==1)){
passTrigger=true;
break;
}
}
if(!passTrigger && nameRoot==2)continue;
nPass[1]++;
const int nAK8Jet=data.GetInt("AK8PuppinJet");
//2.nJets
if(nAK8Jet<2)continue;nPass[2]++;
int* AK8PuppinSubSDJet=data.GetPtrInt("AK8PuppinSubSDJet");
if(AK8PuppinSubSDJet[0]!=2||AK8PuppinSubSDJet[1]!=2)continue;
TClonesArray* AK8PuppijetP4 = (TClonesArray*) data.GetPtrTObject("AK8PuppijetP4");
float* AK8PuppijetCorrUncUp = data.GetPtrFloat("AK8PuppijetCorrUncUp");
float* AK8PuppijetCorrUncDown = data.GetPtrFloat("AK8PuppijetCorrUncDown");
TLorentzVector* thisJet ,* thatJet;
thisJet=(TLorentzVector*)AK8PuppijetP4->At(0);
thatJet=(TLorentzVector*)AK8PuppijetP4->At(1);
//3. Pt
if(thisJet->Pt()>99998 ||thatJet->Pt()>99998 )continue;
if(thisJet->Pt()<300)continue;
if(thatJet->Pt()<300)continue;
nPass[3]++;
//4tightId-----------------------------------------
vector<bool> &AK8PuppijetPassIDTight = *((vector<bool>*) data.GetPtr("AK8PuppijetPassIDTight"));
if(AK8PuppijetPassIDTight[0]==0)continue;
if(AK8PuppijetPassIDTight[1]==0)continue;
Float_t* AK8PuppijetCEmEF = data.GetPtrFloat("AK8PuppijetCEmEF");
Float_t* AK8PuppijetMuoEF = data.GetPtrFloat("AK8PuppijetMuoEF");
if(AK8PuppijetMuoEF[0]>0.8)continue;
if(AK8PuppijetCEmEF[0]>0.9)continue;
if(AK8PuppijetMuoEF[1]>0.8)continue;
if(AK8PuppijetCEmEF[1]>0.9)continue;
nPass[4]++;
//5. Eta-----------------------------------------
if(fabs(thisJet->Eta())>2.4)continue;
if(fabs(thatJet->Eta())>2.4)continue;
nPass[5]++;
//6. DEta-----------------------------------------
float dEta = fabs(thisJet->Eta()-thatJet->Eta());
if(dEta>1.3)continue;
nPass[6]++;
//7. Mjj-----------------------------------------
//float mjjRed = (*thisJet+*thatJet).M()+250-thisJet->M()-thatJet->M();
//if(mjjRed<1000)continue;
nPass[7]++;
//8. fatjetPRmassL2L3Corr-----------------------------------------
nPass[8]++;
//9.-----------------------------------------
Float_t* AK8Puppijet_DoubleSV = data.GetPtrFloat("AK8Puppijet_DoubleSV");
int looseStat=-1;
int tightStat=-1;
if(AK8Puppijet_DoubleSV[0]>0.3 && AK8Puppijet_DoubleSV[1]>0.3)looseStat=0;
else if(AK8Puppijet_DoubleSV[0]>0.3 && AK8Puppijet_DoubleSV[1]<0.3)looseStat=1;
else if(AK8Puppijet_DoubleSV[0]<0.3 && AK8Puppijet_DoubleSV[1]>0.3)looseStat=2;
else looseStat=3;
if(AK8Puppijet_DoubleSV[0]>0.8 && AK8Puppijet_DoubleSV[1]>0.8)tightStat=0;
else if(AK8Puppijet_DoubleSV[0]>0.8 && AK8Puppijet_DoubleSV[1]<0.8)tightStat=1;
else if(AK8Puppijet_DoubleSV[0]<0.3 && AK8Puppijet_DoubleSV[1]>0.8)tightStat=2;
else if(AK8Puppijet_DoubleSV[0]<0.3 && AK8Puppijet_DoubleSV[1]<0.8)tightStat=3;
else tightStat=-1;
double varTemp[2];
Float_t* AK8PuppijetSDmass = data.GetPtrFloat("AK8PuppijetSDmass");
if(AK8PuppijetSDmass[0]<50||AK8PuppijetSDmass[1]<50)continue;
Int_t* AK8Puppijet_nSV=data.GetPtrInt("AK8Puppijet_nSV");
vector<float> *AK8Puppijet_SVMass = data.GetPtrVectorFloat("AK8Puppijet_SVMass");
int nEle= data.GetInt("nEle");
int nMu=data.GetInt("nMu");
Float_t* AK8PuppijetEleEF = data.GetPtrFloat("AK8PuppijetEleEF");
//Float_t* AK8PuppijetMuoEF = data.GetPtrFloat("AK8PuppijetMuoEF");
Int_t* AK8PuppijetCMulti=data.GetPtrInt("AK8PuppijetCMulti");
Int_t* AK8PuppijetEleMulti=data.GetPtrInt("AK8PuppijetEleMulti");
Int_t* AK8PuppijetMuoMulti=data.GetPtrInt("AK8PuppijetMuoMulti");
for(int i=0; i<2;i++){
TLorentzVector* thisAK8Jet ;
if(i==1)thisAK8Jet=thatJet;
else thisAK8Jet=thisJet;
pt_AK8MatchedToHbb=thisAK8Jet->Pt();
eta_AK8MatchedToHbb=thisAK8Jet->Eta();
nsv_AK8MatchedToHbb=AK8Puppijet_nSV[i];
sv0mass_AK8MatchedToHbb=AK8Puppijet_SVMass[i][0];
sv1mass_AK8MatchedToHbb=AK8Puppijet_SVMass[i][1];
nmu_AK8MatchedToHbb=AK8PuppijetMuoMulti[i];
nel_AK8MatchedToHbb=AK8PuppijetEleMulti[i];
muenfr_AK8MatchedToHbb=AK8PuppijetMuoEF[i];
nch_AK8MatchedToHbb=AK8PuppijetCMulti[i];
emenfr_AK8MatchedToHbb=AK8PuppijetEleEF[i];
spec1=nVtx;
spec2=AK8PuppijetSDmass[i];
Float_t val ;
for (Int_t ih=0; ih<nhists; ih++) {
TString title = hists[ih]->GetTitle();
val= (reader->EvaluateRegression( title ))[0];
}
varTemp[i]=val;
}
double PUPPIweight[2]={0};
PUPPIweight[0]=getPUPPIweight(thisJet->Pt(),thisJet->Eta());
PUPPIweight[1]=getPUPPIweight(thatJet->Pt(),thatJet->Eta());
double PUPPIweightThea[2]={0};
PUPPIweightThea[0]=getPUPPIweight_o(thisJet->Pt(),thisJet->Eta());
PUPPIweightThea[1]=getPUPPIweight_o(thatJet->Pt(),thatJet->Eta());
double Mjja= ((*thisJet)+(*thatJet)).M()+250
-((*thisJet)).M()-((*thatJet)).M();
TLorentzVector thisJetReg, thatJetReg;
thisJetReg=(*thisJet)*varTemp[0];
thatJetReg=(*thatJet)*varTemp[1];
double Mjjb= (thisJetReg+thatJetReg).M()+250
-(thisJetReg).M()-(thatJetReg).M();
double PUPPIweightOnRegressed[2]={0};
PUPPIweightOnRegressed[0]=getPUPPIweightOnRegressed(thisJetReg.Pt(),thisJetReg.Eta());
PUPPIweightOnRegressed[1]=getPUPPIweightOnRegressed(thatJetReg.Pt(),thatJetReg.Eta());
vector<float> *subjetSDPx = data.GetPtrVectorFloat("AK8PuppisubjetSDPx");
vector<float> *subjetSDPy = data.GetPtrVectorFloat("AK8PuppisubjetSDPy");
vector<float> *subjetSDPz = data.GetPtrVectorFloat("AK8PuppisubjetSDPz");
vector<float> *subjetSDE = data.GetPtrVectorFloat("AK8PuppisubjetSDE");
vector<float> *AK8PuppisubjetSDRawFactor = data.GetPtrVectorFloat("AK8PuppisubjetSDRawFactor");
TLorentzVector thisSDJet, thatSDJet;
TLorentzVector* subjetP4[2][2];
for(int i=0;i<2;i++){
for(int j=0;j<2;j++){
subjetP4[i][j]=new TLorentzVector(0,0,0,0);
subjetP4[i][j]->SetPxPyPzE(subjetSDPx[i][j],subjetSDPy[i][j],subjetSDPz[i][j],subjetSDE[i][j]);
// subjetP4[i][j]*=AK8PuppisubjetSDRawFactor[i][j];
}
}
thisSDJet=(*subjetP4[0][0])*AK8PuppisubjetSDRawFactor[0][0]+(*subjetP4[0][1])*AK8PuppisubjetSDRawFactor[0][1];
thatSDJet=(*subjetP4[1][0])*AK8PuppisubjetSDRawFactor[1][0]+(*subjetP4[1][1])*AK8PuppisubjetSDRawFactor[1][1];
//thatSDJet=(*subjetP4[1][0])+(*subjetP4[1][1]);
TLorentzVector thisSDJetReg, thatSDJetReg;
thisSDJetReg= thisSDJet*varTemp[0]*PUPPIweightOnRegressed[0];
thatSDJetReg= thatSDJet*varTemp[1]*PUPPIweightOnRegressed[1];
//double Mjjc= ((thisSDJet)+(thatSDJet)).M()+250
// -((thisSDJet)).M()-((thatSDJet)).M();
double Mjjd= ((thisSDJet)+(thatSDJet)).M()+250
-((thisSDJet)).M()-((thatSDJet)).M();
Float_t* AK8PuppijetTau1 = data.GetPtrFloat("AK8PuppijetTau1");
Float_t* AK8PuppijetTau2 = data.GetPtrFloat("AK8PuppijetTau2");
double puppiTau21[2];
puppiTau21[0]=(AK8PuppijetTau2[0]/AK8PuppijetTau1[0]),puppiTau21[1]=(AK8PuppijetTau2[1]/AK8PuppijetTau1[1]);
double mass_j0,mass_j1,MjjLoop;
int massCat;
for(int k=0;k<7;k++){
if(k==0||k==2||k==4){
if(thisJet->Pt()<300)continue;
if(thatJet->Pt()<300)continue;
}
else if (k==1){
if((thisSDJet*PUPPIweightThea[0]).Pt()<300)continue;
if((thatSDJet*PUPPIweightThea[1]).Pt()<300)continue;
}
else if (k==3){
if((thisSDJet*PUPPIweight[0]).Pt()<300)continue;
if((thatSDJet*PUPPIweight[1]).Pt()<300)continue;
}
else if (k==5){
if(thisJetReg.Pt()<300)continue;
if(thatJetReg.Pt()<300)continue;
}
else{
if(thisSDJetReg.Pt()<300)continue;
if(thatSDJetReg.Pt()<300)continue;
}
if(k==0||k==1){
mass_j0=AK8PuppijetSDmass[0]*PUPPIweightThea[0];
mass_j1=AK8PuppijetSDmass[1]*PUPPIweightThea[1];
massCat=0;
}
else if (k==2||k==3){
mass_j0=AK8PuppijetSDmass[0]*PUPPIweight[0];
mass_j1=AK8PuppijetSDmass[1]*PUPPIweight[1];
massCat=1;
}
else{
mass_j0=AK8PuppijetSDmass[0]*varTemp[0]*PUPPIweightOnRegressed[0];
mass_j1=AK8PuppijetSDmass[1]*varTemp[1]*PUPPIweightOnRegressed[1];
massCat=2;
}
if(k==0||k==2||k==4)MjjLoop=Mjja;
else if (k==1)MjjLoop=((thisSDJet)*PUPPIweightThea[0]+(thatSDJet)*PUPPIweightThea[1]).M()+250-((thisSDJet)*PUPPIweightThea[0]).M()-((thatSDJet)*PUPPIweightThea[1]).M();
else if (k==3)MjjLoop=((thisSDJet)*PUPPIweight[0]+(thatSDJet)*PUPPIweight[1]).M()+250-((thisSDJet)*PUPPIweight[0]).M()-((thatSDJet)*PUPPIweight[1]).M();
else if (k==5)MjjLoop=Mjjb;
else MjjLoop=Mjjd;
//cout<<mass_j0<<","<<mass_j1<<",k="<<k<<endl;
for(int i=0;i<nWidth;i++){
for(int j=0;j<nBmin;j++){
if(mass_j0<bmin[j] ||mass_j0>width[i]+bmin[j]
||mass_j1<bmin[j] ||mass_j1>width[i]+bmin[j] )continue;
for(int m=0;m<2;m++){
if(m==0 && (puppiTau21[0]>0.6 || puppiTau21[1]>0.6))continue;
double tightPFRatio=0,loosePFRatio=0;
tightPFRatio=fa[massCat][0][m][1]->Eval(mass_j0);
loosePFRatio=fa[massCat][0][m][0]->Eval(mass_j0);
if(tightStat==0){
th3d[k][i][j][m]->Fill(MjjLoop);
}
else if (tightStat==1){
th3f[k][i][j][m]->Fill(MjjLoop);
}
else if(tightStat==3){
tightPFRatio=fa[massCat][1][m][1]->Eval(mass_j0);
th3v[k][i][j][m]->Fill(MjjLoop,tightPFRatio);
}
if(looseStat==0 && tightStat!=0){
th3d[k+7][i][j][m]->Fill(MjjLoop);
}
else if (looseStat==1){
th3f[k+7][i][j][m]->Fill(MjjLoop);
}
else if(looseStat==3){
loosePFRatio=fa[massCat][1][m][0]->Eval(mass_j0);
th3v[k+7][i][j][m]->Fill(MjjLoop,loosePFRatio);
}
}
}
}
}
}
}
for(int i=0;i<10;i++)cout<<"npass["<<i<<"]="<<nPass[i]<<endl;
TFile* outFile;//= new TFile(Form("PFRatio/%s.root",st2.data()),"recreate");
outFile= new TFile(Form("MjjVC/%s.root",st2.data()),"recreate");
for(int i=0;i<nWidth;i++){
for(int j=0;j<nBmin;j++){
for(int k=0;k<2;k++){
for(int l=0;l<14;l++){
th3d[l][i][j][k]->Write();
th3f[l][i][j][k]->Write();
th3v[l][i][j][k]->Write();
}
}
}
}
outFile->Close();
for(int i=0;i<nWidth;i++){
for(int j=0;j<nBmin;j++){
for(int k=0;k<2;k++){
for(int l=0;l<14;l++){
delete th3d[l][i][j][k];
delete th3f[l][i][j][k];
delete th3v[l][i][j][k];
}
}
}
}
delete reader;
}
void TMVARegressionApplication( TString myMethodList = "" )
{
//---------------------------------------------------------------
// This loads the library
TMVA::Tools::Instance();
// Default MVA methods to be trained + tested
std::map<std::string,int> Use;
// --- Mutidimensional likelihood and Nearest-Neighbour methods
Use["PDERS"] = 0;
Use["PDEFoam"] = 1;
Use["KNN"] = 1;
//
// --- Linear Discriminant Analysis
Use["LD"] = 1;
//
// --- Function Discriminant analysis
Use["FDA_GA"] = 1;
Use["FDA_MC"] = 0;
Use["FDA_MT"] = 0;
Use["FDA_GAMT"] = 0;
//
// --- Neural Network
Use["MLP"] = 1;
Use["DNN_CPU"] = 0;
//
// --- Support Vector Machine
Use["SVM"] = 0;
//
// --- Boosted Decision Trees
Use["BDT"] = 0;
Use["BDTG"] = 1;
// ---------------------------------------------------------------
std::cout << std::endl;
std::cout << "==> Start TMVARegressionApplication" << 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;
reader->AddVariable( "var1", &var1 );
reader->AddVariable( "var2", &var2 );
// 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 );
// --- Book the MVA methods
TString dir = "dataset/weights/";
TString prefix = "TMVARegression";
// 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 );
}
}
// Book output histograms
TH1* hists[100];
Int_t nhists = -1;
for (std::map<std::string,int>::iterator it = Use.begin(); it != Use.end(); it++) {
TH1* h = new TH1F( it->first.c_str(), TString(it->first) + " method", 100, -100, 600 );
if (it->second) hists[++nhists] = h;
}
nhists++;
// 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_reg_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.ch/files/tmva_reg_example.root", "CACHEREAD"); // if not: download from ROOT server
}
if (!input) {
std::cout << "ERROR: could not open data file" << std::endl;
exit(1);
}
std::cout << "--- TMVARegressionApp : Using input file: " << input->GetName() << std::endl;
// --- Event loop
// 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* theTree = (TTree*)input->Get("TreeR");
std::cout << "--- Select signal sample" << std::endl;
theTree->SetBranchAddress( "var1", &var1 );
theTree->SetBranchAddress( "var2", &var2 );
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);
// Retrieve the MVA target values (regression outputs) and fill into histograms
// NOTE: EvaluateRegression(..) returns a vector for multi-target regression
for (Int_t ih=0; ih<nhists; ih++) {
TString title = hists[ih]->GetTitle();
Float_t val = (reader->EvaluateRegression( title ))[0];
hists[ih]->Fill( val );
}
}
sw.Stop();
std::cout << "--- End of event loop: "; sw.Print();
// --- Write histograms
TFile *target = new TFile( "TMVARegApp.root","RECREATE" );
for (Int_t ih=0; ih<nhists; ih++) hists[ih]->Write();
target->Close();
std::cout << "--- Created root file: \"" << target->GetName()
<< "\" containing the MVA output histograms" << std::endl;
delete reader;
std::cout << "==> TMVARegressionApplication is done!" << std::endl << std::endl;
}
////////////////////////////////////////////////////////////////////////////////
/// Main ///
////////////////////////////////////////////////////////////////////////////////
void GrowTree(TString process, std::string regMethod="BDTG", Long64_t beginEntry=0, Long64_t endEntry=-1)
{
gROOT->SetBatch(1);
TH1::SetDefaultSumw2(1);
gROOT->LoadMacro("HelperFunctions.h"); //< make functions visible to TTreeFormula
if (!TString(gROOT->GetVersion()).Contains("5.34")) {
std::cout << "INCORRECT ROOT VERSION! Please use 5.34:" << std::endl;
std::cout << "source /uscmst1/prod/sw/cms/slc5_amd64_gcc462/lcg/root/5.34.02-cms/bin/thisroot.csh" << std::endl;
std::cout << "Return without doing anything." << std::endl;
return;
}
const TString indir = "/afs/cern.ch/work/d/degrutto/public/MiniAOD/ZnnHbb_Phys14_PU20bx25/skimV11/";
const TString outdir = "/afs/cern.ch/work/d/degrutto/public/MiniAOD/ZnnHbb_Phys14_PU20bx25/skimV11/step3/";
const TString prefix = "skim_";
const TString suffix = ".root";
TFile *input = TFile::Open(indir + prefix + process + suffix);
if (!input) {
std::cout << "ERROR: Could not open input file." << std::endl;
exit(1);
}
/// Make output directory if it doesn't exist
if (gSystem->AccessPathName(outdir))
gSystem->mkdir(outdir);
std::cout << "--- GrowTree : Using input file: " << input->GetName() << std::endl;
TTree *inTree = (TTree *) input->Get("tree");
TH1F *hcount = (TH1F *) input->Get("Count");
TFile *output(0);
if (beginEntry == 0 && endEntry == -1)
output = TFile::Open(outdir + "Step3_" + process + suffix, "RECREATE");
else
output = TFile::Open(outdir + "Step3_" + process + TString::Format("_%Li_%Li", beginEntry, endEntry) + suffix, "RECREATE");
TTree *outTree = inTree->CloneTree(0); // Do no copy the data yet
/// The clone should not delete any shared i/o buffers.
ResetDeleteBranches(outTree);
///-- Set branch addresses -------------------------------------------------
EventInfo EVENT;
double hJet_pt[MAXJ], hJet_eta[MAXJ], hJet_phi[MAXJ], hJet_m[MAXJ], hJet_ptRaw[MAXJ], hJet_genPt[MAXJ];
int hJCidx[2];
inTree->SetBranchStatus("*", 1);
inTree->SetBranchStatus("hJCidx",1);
inTree->SetBranchStatus("Jet_*",1);
inTree->SetBranchAddress("hJCidx", &hJCidx);
inTree->SetBranchAddress("Jet_pt", &hJet_pt);
inTree->SetBranchAddress("Jet_eta", &hJet_eta);
inTree->SetBranchAddress("Jet_phi", &hJet_phi);
inTree->SetBranchAddress("Jet_mass", &hJet_m);
inTree->SetBranchAddress("Jet_rawPt", &hJet_ptRaw);
inTree->SetBranchAddress("Jet_mcPt", &hJet_genPt);
///-- Make new branches ----------------------------------------------------
int EVENT_run, EVENT_event; // set these as TTree index?
float lumi_ = lumi, efflumi, efflumi_old,
efflumi_UEPS_up, efflumi_UEPS_down;
float hJet_ptReg[2];
float HptNorm, HptGen, HptReg;
float HmassNorm, HmassGen, HmassReg;
outTree->Branch("EVENT_run", &EVENT_run, "EVENT_run/I");
outTree->Branch("EVENT_event", &EVENT_event, "EVENT_event/I");
outTree->Branch("lumi", &lumi_, "lumi/F");
outTree->Branch("efflumi", &efflumi, "efflumi/F");
outTree->Branch("efflumi_old", &efflumi_old, "efflumi_old/F");
outTree->Branch("efflumi_UEPS_up", &efflumi_UEPS_up, "efflumi_UEPS_up/F");
outTree->Branch("efflumi_UEPS_down", &efflumi_UEPS_down, "efflumi_UEPS_down/F");
outTree->Branch("hJet_ptReg", &hJet_ptReg, "hJet_ptReg[2]/F");
outTree->Branch("HptNorm", &HptNorm, "HptNorm/F");
outTree->Branch("HptGen", &HptGen, "HptGen/F");
outTree->Branch("HptReg", &HptReg, "HptReg/F");
outTree->Branch("HmassNorm", &HmassNorm, "HmassNorm/F");
outTree->Branch("HmassGen", &HmassGen, "HmassGen/F");
outTree->Branch("HmassReg", &HmassReg, "HmassReg/F");
/// Get effective lumis
std::map < std::string, float > efflumis = GetLumis();
efflumi = efflumis[process.Data()];
assert(efflumi > 0);
efflumi_old = efflumi;
efflumi_UEPS_up = efflumi * hcount->GetBinContent(2) / hcount->GetBinContent(3);
efflumi_UEPS_down = efflumi * hcount->GetBinContent(2) / hcount->GetBinContent(4);
TTreeFormula* ttf_lheweight = new TTreeFormula("ttf_lheweight", Form("%f", efflumi), inTree);
#ifdef STITCH
std::map < std::string, std::string > lheweights = GetLHEWeights();
TString process_lhe = process;
if (process_lhe.BeginsWith("WJets") && process_lhe != "WJetsHW")
process_lhe = "WJets";
else if (process_lhe.BeginsWith("ZJets") && process_lhe != "ZJetsHW")
process_lhe = "ZJets";
else
process_lhe = "";
TString lheweight = lheweights[process_lhe.Data()];
if (lheweight != "") {
delete ttf_lheweight;
// Bug fix for ZJetsPtZ100
if (process == "ZJetsPtZ100")
lheweight.ReplaceAll("lheV_pt", "999");
std::cout << "BUGFIX: " << lheweight << std::endl;
ttf_lheweight = new TTreeFormula("ttf_lheweight", lheweight, inTree);
}
#endif
ttf_lheweight->SetQuickLoad(1);
// regression stuff here
///-- Setup TMVA Reader ----------------------------------------------------
TMVA::Tools::Instance(); //< This loads the library
TMVA::Reader * reader = new TMVA::Reader("!Color:!Silent");
/// Get the variables
const std::vector < std::string > & inputExpressionsReg = GetInputExpressionsReg();
const UInt_t nvars = inputExpressionsReg.size();
Float_t readerVars[nvars];
int idx_rawpt = -1, idx_pt = -1, idx_et = -1, idx_mt = -1;
for (UInt_t iexpr = 0; iexpr < nvars; iexpr++) {
const TString& expr = inputExpressionsReg.at(iexpr);
reader->AddVariable(expr, &readerVars[iexpr]);
if (expr.BeginsWith("breg_rawptJER := ")) idx_rawpt = iexpr;
else if (expr.BeginsWith("breg_pt := ")) idx_pt = iexpr;
else if (expr.BeginsWith("breg_et := ")) idx_et = iexpr;
else if (expr.BeginsWith("breg_mt := ")) idx_mt = iexpr;
}
// assert(idx_rawpt!=-1 && idx_pt!=-1 && idx_et!=-1 && idx_mt!=-1);
assert(idx_rawpt!=-1 && idx_pt!=-1 );
/// Setup TMVA regression inputs
const std::vector < std::string > & inputExpressionsReg0 = GetInputExpressionsReg0();
const std::vector < std::string > & inputExpressionsReg1 = GetInputExpressionsReg1();
assert(inputExpressionsReg0.size() == nvars);
assert(inputExpressionsReg1.size() == nvars);
/// Load TMVA weights
TString weightdir = "weights/";
TString weightfile = weightdir + "TMVARegression_" + regMethod + ".testweights.xml";
reader->BookMVA(regMethod + " method", weightfile);
TStopwatch sw;
sw.Start();
/// Create TTreeFormulas
TTreeFormula *ttf = 0;
std::vector < TTreeFormula * >::const_iterator formIt, formItEnd;
std::vector < TTreeFormula * > inputFormulasReg0;
std::vector < TTreeFormula * > inputFormulasReg1;
std::vector < TTreeFormula * > inputFormulasFJReg0;
std::vector < TTreeFormula * > inputFormulasFJReg1;
std::vector < TTreeFormula * > inputFormulasFJReg2;
for (UInt_t iexpr = 0; iexpr < nvars; iexpr++) {
ttf = new TTreeFormula(Form("ttfreg%i_0", iexpr), inputExpressionsReg0.at(iexpr).c_str(), inTree);
ttf->SetQuickLoad(1);
inputFormulasReg0.push_back(ttf);
ttf = new TTreeFormula(Form("ttfreg%i_1", iexpr), inputExpressionsReg1.at(iexpr).c_str(), inTree);
ttf->SetQuickLoad(1);
inputFormulasReg1.push_back(ttf);
}
///-- Loop over events -----------------------------------------------------
Int_t curTree = inTree->GetTreeNumber();
const Long64_t nentries = inTree->GetEntries();
if (endEntry < 0) endEntry = nentries;
Long64_t ievt = 0;
for (ievt=TMath::Max(ievt, beginEntry); ievt<TMath::Min(nentries, endEntry); ievt++) {
if (ievt % 2000 == 0)
std::cout << "--- ... Processing event: " << ievt << std::endl;
const Long64_t local_entry = inTree->LoadTree(ievt); // faster, but only for TTreeFormula
if (local_entry < 0) break;
inTree->GetEntry(ievt); // same event as received by LoadTree()
if (inTree->GetTreeNumber() != curTree) {
curTree = inTree->GetTreeNumber();
for (formIt=inputFormulasReg0.begin(), formItEnd=inputFormulasReg0.end(); formIt!=formItEnd; formIt++)
(*formIt)->UpdateFormulaLeaves(); // if using TChain
for (formIt=inputFormulasReg1.begin(), formItEnd=inputFormulasReg1.end(); formIt!=formItEnd; formIt++)
(*formIt)->UpdateFormulaLeaves(); // if using TChain
for (formIt=inputFormulasFJReg0.begin(), formItEnd=inputFormulasFJReg0.end(); formIt!=formItEnd; formIt++)
(*formIt)->UpdateFormulaLeaves(); // if using TChain
for (formIt=inputFormulasFJReg1.begin(), formItEnd=inputFormulasFJReg1.end(); formIt!=formItEnd; formIt++)
(*formIt)->UpdateFormulaLeaves(); // if using TChain
for (formIt=inputFormulasFJReg2.begin(), formItEnd=inputFormulasFJReg2.end(); formIt!=formItEnd; formIt++)
(*formIt)->UpdateFormulaLeaves(); // if using TChain
ttf_lheweight->UpdateFormulaLeaves();
}
/// These need to be called when arrays of variable size are used in TTree.
for (formIt=inputFormulasReg0.begin(), formItEnd=inputFormulasReg0.end(); formIt!=formItEnd; formIt++)
(*formIt)->GetNdata();
for (formIt=inputFormulasReg1.begin(), formItEnd=inputFormulasReg1.end(); formIt!=formItEnd; formIt++)
(*formIt)->GetNdata();
for (formIt=inputFormulasFJReg0.begin(), formItEnd=inputFormulasFJReg0.end(); formIt!=formItEnd; formIt++)
(*formIt)->GetNdata();
for (formIt=inputFormulasFJReg1.begin(), formItEnd=inputFormulasFJReg1.end(); formIt!=formItEnd; formIt++)
(*formIt)->GetNdata();
for (formIt=inputFormulasFJReg2.begin(), formItEnd=inputFormulasFJReg2.end(); formIt!=formItEnd; formIt++)
(*formIt)->GetNdata();
ttf_lheweight->GetNdata();
/// Fill branches
EVENT_run = EVENT.run;
EVENT_event = EVENT.event;
#ifdef STITCH
efflumi = ttf_lheweight->EvalInstance();
// efflumi_UEPS_up = efflumi * hcount->GetBinContent(2) / hcount->GetBinContent(3);
//efflumi_UEPS_down = efflumi * hcount->GetBinContent(2) / hcount->GetBinContent(4);
#endif
bool verbose = false;
for (Int_t ihj = 0; ihj < 2; ihj++) {
/// Evaluate TMVA regression output
for (UInt_t iexpr = 0; iexpr < nvars; iexpr++) {
if (ihj==0) {
readerVars[iexpr] = inputFormulasReg0.at(iexpr)->EvalInstance();
} else if (ihj==1) {
readerVars[iexpr] = inputFormulasReg1.at(iexpr)->EvalInstance();
}
}
hJet_ptReg[ihj] = (reader->EvaluateRegression(regMethod + " method"))[0];
if (verbose) std::cout << readerVars[idx_pt] << " " << readerVars[idx_rawpt] << " " << hJet_pt[ihj] << " " << hJet_ptReg[ihj] << " " << hJet_genPt[ihj] << std::endl;
const TLorentzVector p4Zero = TLorentzVector(0., 0., 0., 0.);
// int idx = hJCidx[0] ;
// std::cout << "the regressed pt for jet 0 is " << hJet_ptReg[0] << "; the hJCidx is " << hJCidx[0] << ", hence the origianl pt is " << hJet_pt[idx] << std::endl;
const TLorentzVector& hJet_p4Norm_0 = makePtEtaPhiM(hJet_pt[hJCidx[0]] , hJet_pt[hJCidx[0]], hJet_eta[hJCidx[0]], hJet_phi[hJCidx[0]], hJet_m[hJCidx[0]]);
const TLorentzVector& hJet_p4Norm_1 = makePtEtaPhiM(hJet_pt[hJCidx[1]] , hJet_pt[hJCidx[1]], hJet_eta[hJCidx[1]], hJet_phi[hJCidx[1]], hJet_m[hJCidx[1]]);
const TLorentzVector& hJet_p4Gen_0 = hJet_genPt[hJCidx[0]] > 0 ?
makePtEtaPhiM(hJet_genPt[hJCidx[0]] , hJet_pt[hJCidx[0]], hJet_eta[hJCidx[0]], hJet_phi[hJCidx[0]], hJet_m[hJCidx[0]]) : p4Zero;
const TLorentzVector& hJet_p4Gen_1 = hJet_genPt[hJCidx[1]] > 0 ?
makePtEtaPhiM(hJet_genPt[hJCidx[1]] , hJet_pt[hJCidx[1]], hJet_eta[hJCidx[1]], hJet_phi[hJCidx[1]], hJet_m[hJCidx[1]]) : p4Zero;
const TLorentzVector& hJet_p4Reg_0 = makePtEtaPhiM(hJet_ptReg[0] , hJet_pt[hJCidx[0]], hJet_eta[hJCidx[0]], hJet_phi[hJCidx[0]], hJet_m[hJCidx[0]]);
const TLorentzVector& hJet_p4Reg_1 = makePtEtaPhiM(hJet_ptReg[1] , hJet_pt[hJCidx[1]], hJet_eta[hJCidx[1]], hJet_phi[hJCidx[1]], hJet_m[hJCidx[1]]);
HptNorm = (hJet_p4Norm_0 + hJet_p4Norm_1 ).Pt();
HptGen = (hJet_p4Gen_0 + hJet_p4Gen_1 ).Pt();
HptReg = (hJet_p4Reg_0 + hJet_p4Reg_1 ).Pt();
HmassNorm = (hJet_p4Norm_0 + hJet_p4Norm_1 ).M();
HmassGen = (hJet_p4Gen_0 + hJet_p4Gen_1 ).M();
HmassReg = (hJet_p4Reg_0 + hJet_p4Reg_1 ).M();
// std::cout << "HmassReg is " << HmassReg << std::endl;
}
outTree->Fill(); // fill it!
} // end loop over TTree entries
/// Get elapsed time
sw.Stop();
std::cout << "--- End of event loop: ";
sw.Print();
output->cd();
outTree->Write();
output->Close();
input->Close();
delete input;
delete output;
for (formIt=inputFormulasReg0.begin(), formItEnd=inputFormulasReg0.end(); formIt!=formItEnd; formIt++)
delete *formIt;
for (formIt=inputFormulasReg1.begin(), formItEnd=inputFormulasReg1.end(); formIt!=formItEnd; formIt++)
delete *formIt;
for (formIt=inputFormulasFJReg0.begin(), formItEnd=inputFormulasFJReg0.end(); formIt!=formItEnd; formIt++)
delete *formIt;
for (formIt=inputFormulasFJReg1.begin(), formItEnd=inputFormulasFJReg1.end(); formIt!=formItEnd; formIt++)
delete *formIt;
for (formIt=inputFormulasFJReg2.begin(), formItEnd=inputFormulasFJReg2.end(); formIt!=formItEnd; formIt++)
delete *formIt;
delete ttf_lheweight;
std::cout << "==> GrowTree is done!" << std::endl << std::endl;
return;
}
TString useAutoencoder (TString method_name)
{
TMVA::Tools::Instance();
std::cout << "==> Start useAutoencoder" << std::endl;
TMVA::Reader *reader = new TMVA::Reader( "!Color:!Silent" );
Float_t signal = 0.0;
Float_t outSignal = 0.0;
Float_t inSignal = 0.0;
std::vector<std::string> localVariableNames (variableNames+additionalVariableNames);
std::vector<Float_t> variables (localVariableNames.size ());
auto itVar = begin (variables);
for (auto varName : localVariableNames)
{
Float_t* pVar = &(*itVar);
reader->AddVariable(varName.c_str(), pVar);
(*itVar) = 0.0;
++itVar;
}
int idxSignal = std::distance (localVariableNames.begin (),
std::find (localVariableNames.begin (), localVariableNames.end (),std::string ("signal")));
TString dir = "weights/";
TString prefix = "TMVAAutoencoder";
TString weightfile = dir + prefix + TString("_") + method_name + TString(".weights.xml");
TString outPrefix = "transformed";
TString outfilename = pathToData + outPrefix + TString("_") + method_name + TString(".root");
reader->BookMVA( method_name, weightfile );
TFile* outFile = new TFile (outfilename.Data (), "RECREATE");
std::vector<std::string> inputNames = {"training"};
std::map<std::string,std::vector<std::string>> varsForInput;
varsForInput["training"].emplace_back ("id");
varsForInput["training"].emplace_back ("signal");
for (auto inputName : inputNames)
{
std::stringstream outfilename;
outfilename << inputName << "_transformed__" << method_name.Data () << ".root";
std::cout << outfilename.str () << std::endl;
/* return; */
std::stringstream infilename;
infilename << pathToData.Data () << inputName << ".root";
TTree* outTree = new TTree("transformed","transformed");
std::vector<Float_t> outVariables (localVariableNames.size ());
itVar = begin (variables);
auto itOutVar = begin (outVariables);
for (auto varName : localVariableNames)
{
Float_t* pOutVar = &(*itOutVar);
outTree->Branch (varName.c_str (), pOutVar, "F");
(*itOutVar) = 0.0;
++itOutVar;
Float_t* pVar = &(*itVar);
std::stringstream svar;
svar << varName << "_in";
outTree->Branch (svar.str ().c_str (), pVar, "F");
(*itVar) = 0.0;
++itVar;
}
Float_t signal_original = 0.0;
outTree->Branch ("signal_original", &signal_original, "F");
TFile *input(0);
std::cout << "infilename = " << infilename.str ().c_str () << std::endl;
input = TFile::Open (infilename.str ().c_str ());
TTree* tree = (TTree*)input->Get("data");
Int_t ids;
// id field if needed
if (std::find (varsForInput[inputName].begin (), varsForInput[inputName].end (), "id") != varsForInput[inputName].end ())
tree->SetBranchAddress("id", &ids);
// variables for prediction
itVar = begin (variables);
for (auto inputName : localVariableNames)
{
Float_t* pVar = &(*itVar);
tree->SetBranchAddress (inputName.c_str(), pVar);
++itVar;
}
for (Long64_t ievt=0; ievt < tree->GetEntries(); ievt++)
{
tree->GetEntry(ievt);
// predict
signal_original = variables.at (idxSignal);
for (int forcedSignal = 0; forcedSignal <= 1; ++forcedSignal)
{
variables.at (idxSignal) = forcedSignal;
std::vector<Float_t> regressionValues = reader->EvaluateRegression (method_name);
size_t idx = 0;
for (auto it = std::begin (regressionValues), itEnd = std::end (regressionValues); it != itEnd; ++it)
{
outVariables.at (idx) = *it;
++idx;
}
outTree->Fill ();
}
}
outFile->Write ();
input->Close();
}
delete reader;
return outfilename;
}