示例#1
0
文件: tpc_raw.C 项目: alisw/AliRoot
void renderHLTClusters(TTree* clustersTree)
{
    
    /**************************
     *  Visualization of Clusters
     **************************/
    const Int_t kMaxCl=100*160;
    Int_t fNColorBins = 5;
    
    TEvePointSet* clusters = new TEvePointSet(kMaxCl);
    clusters->SetOwnIds(kTRUE);
    
    TEvePointSetArray * cc = new TEvePointSetArray("TPC Clusters Colorized");
    cc->SetMainColor(kRed);
    cc->SetMarkerStyle(8);
    cc->SetMarkerSize(1.0);
    cc->InitBins("Cluster Charge", fNColorBins, 0., fNColorBins*60.);
    
    cc->GetBin(0)->SetMainColor(kGray);
    cc->GetBin(0)->SetMarkerSize(1.0);
    cc->GetBin(1)->SetMainColor(kBlue);
    cc->GetBin(1)->SetMarkerSize(1.0);
    cc->GetBin(2)->SetMainColor(kCyan);
    cc->GetBin(2)->SetMarkerSize(1.0);
    cc->GetBin(3)->SetMainColor(kGreen);
    cc->GetBin(3)->SetMarkerSize(1.0);
    cc->GetBin(4)->SetMainColor(kYellow);
    cc->GetBin(4)->SetMarkerSize(1.0);
    cc->GetBin(5)->SetMainColor(kRed);
    cc->GetBin(5)->SetMarkerSize(1.0);
    cc->GetBin(6)->SetMainColor(kMagenta);
    cc->GetBin(6)->SetMarkerSize(1.0);
    
    
    // Loop over clusters
    Int_t nentries = clustersTree->GetEntriesFast();
    
    AliTPCClustersRow *clrow = new AliTPCClustersRow();
    clrow->SetClass("AliTPCclusterMI");
    //clrow->SetArray(kMaxCl);
    clustersTree->SetBranchAddress("Segment", &clrow);
    
    for (Int_t i=0; i<nentries; i++) {
        if (!clustersTree->GetEvent(i)) continue;
        
        TClonesArray *cl = clrow->GetArray();
        Int_t ncl = cl->GetEntriesFast();
        
        while (ncl--)
        {
            AliTPCclusterMI* clusterMI = (AliTPCclusterMI*) cl->At(ncl);
            
            AliCluster *c = (AliCluster*) cl->UncheckedAt(ncl);
            Float_t g[3]; //global coordinates
            c->GetGlobalXYZ(g);
            cout<<g[0]<<"\t"<<g[1]<<"\t"<<g[2]<<endl;
            cc->Fill(g[0], g[1], g[2], clusterMI->GetQ());
            clusters->SetNextPoint(g[0], g[1], g[2]);
            AliCluster *atp = new AliCluster(*clusterMI);
            clusters->SetPointId(atp);
            
        }
        cl->Clear();
    }
    
//    delete clrow;
    
    clusters->SetName("TPC Clusters");
    clusters->SetTitle(Form("N=%d", clusters->Size()));
    
//    const TString viz_tag("REC Clusters TPC"); // to be changed
//    clusters->ApplyVizTag(viz_tag, "Clusters");
    
    cc->SetRnrSelf(kTRUE);
    
    gEve->AddElement(cc);
    
    return;
}
示例#2
0
文件: diganal.C 项目: alisw/AliRoot
void tpccfind(Int_t threshold, Float_t thr = 2,
	      Int_t i1 = 2, Int_t i2 = 2, 
	      Int_t tharea =20, Int_t thmax=100)
{  
  ///////////////////////GRAPHICS//DECLARATION//////////////////////
  TCanvas  * c1 = new TCanvas("padcluster","Cluster finder 1",700,900);
  c1->cd();
  TCanvas  * c2 = new TCanvas("padcluster2","Cluster finder 2",700,900);
  c2->cd();
  c1->cd();
  TPad * pad11 = new TPad("pad11","",0.01,0.76,0.48,0.95,21);
  pad11->Draw();
  TPad * pad12 = new TPad("pad12","",0.51,0.76,0.95,0.95,21);
  pad12->Draw();
  TPad * pad21 = new TPad("pad21","",0.01,0.56,0.49,0.74,21);
  pad21->Draw();
  TPad * pad22 = new TPad("pad22","",0.51,0.56,0.95,0.74,21);
  pad22->Draw();
  TPad * pad31 = new TPad("pad31","",0.01,0.36,0.49,0.54,21);
  pad31->Draw();
  TPad * pad32 = new TPad("pad32","",0.51,0.36,0.95,0.54,21);
  pad32->Draw();
  TPad * pad41 = new TPad("pad41","",0.01,0.16,0.49,0.34,21);
  pad41->Draw();
  TPad * pad42 = new TPad("pad42","",0.51,0.16,0.95,0.34,21);
  pad42->Draw();

  c2->cd();
  TPad * pad11_2 = new TPad("pad11_2","",0.01,0.76,0.48,0.95,21);
  pad11_2->Draw();
  TPad * pad12_2 = new TPad("pad12_2","",0.51,0.76,0.95,0.95,21);
  pad12_2->Draw();
  /////////////////////HISTOGRAMS///DECLARATION///////////////////////
  pad11->cd();
  TH1F * hsx = new TH1F("hsx","Sigma of distribution in time",40,0,2);
  pad12->cd();
  TH1F * hsy = new TH1F("hsy","Sigma of distribution in pads",40,0,2);

  pad21->cd();
  TProfile * hsx2 = new TProfile("hsx2","Sigma of distribution in time",
			 20,100,500);
  pad22->cd();
  TProfile * hsy2 = new TProfile("hsy2","Sigma of distribution in pads",
			 20,100,500);

  pad31->cd();
  TH1F * harea = new TH1F("harea","Area of the peak",26,0,52);
  pad32->cd();
  TH1F * hmax = new TH1F("hmax","Maximal amplitude in peak",30,0,150);
 

  pad41->cd();  
  TProfile * harea2= new TProfile("harea2","Area dependence z coordinata",
			 20,100,500);
  pad42->cd();
  TProfile * hmax2 = new TProfile("hmax2","Maximal amplitude dependence",
			 20,100,500);
  pad41->cd();  

  pad11_2->cd();
  TProfile * harea2p= new TProfile("harea2p","Area dependence on pad coordinata",
			 20,0,100);
  pad12_2->cd();
  TProfile * hmax2p = new TProfile("hmax2p","Maximal amplitude dependence on pad",
			 20,0,50);
  //////////////////CALCULATION//////////////////////////////////////////

  for (Int_t k = i1;k <=i2; k++)
    {
      tpcanal(1,k,10,0,kFALSE);
      TClusterFinder * cf=new TClusterFinder(0,0,threshold,1);
      cf->GetHisto(&gtpc.GetHis1());
      TClonesArray * arr = cf->FindClusters();
      cf->Delete();
      Int_t size = arr->GetEntries();
      
      if ( size>0 )   
	for (Int_t i=0 ; i<size;i++)
	  {
	    Int_t index;
	    TCluster *c=(TCluster*)arr->UncheckedAt(i);
	    hsx->Fill(TMath::Sqrt(c.fSigmaX2));
	    hsy->Fill(TMath::Sqrt(c.fSigmaY2));    
	    if  (TMath::Sqrt(c.fSigmaX2)<thr)
	      hsx2->Fill(c.fX,TMath::Sqrt(c.fSigmaX2),1);
	    if  (TMath::Sqrt(c.fSigmaY2)<thr)
	      hsy2->Fill(c.fX,TMath::Sqrt(c.fSigmaY2),1);       
	    hmax->Fill(c.fMax);
	    harea->Fill(c.fArea);
	    if (c.fArea<tharea)  harea2->Fill(c.fX,c.fArea,1);
            if (c.fMax<thmax) hmax2->Fill(c.fX,c.fMax,1);
	    if (c.fArea<tharea)  harea2p->Fill(c.fY,c.fArea,1);
            if (c.fMax<thmax) hmax2p->Fill(c.fY,c.fMax,1);
	    
	  }
    } 
  gStyle->SetOptStat(1);
  pad11->cd();
  hsx->Draw();
  pad12->cd();
  hsy->Draw();

  pad21->cd();
  hsx2->Draw();
  pad22->cd();
  hsy2->Draw();

  pad31->cd();
  harea->Draw();
  pad32->cd();
  hmax->Draw();

  pad41->cd();
  harea2->Draw();
  pad42->cd();
  hmax2->Draw();


  c1->cd();
  TPaveText * comment = new TPaveText(0.05,0.02,0.95,0.14,"NDC");
  comment->SetTextAlign(12);
  comment->SetFillColor(42);   
  comment->ReadFile("clusters.txt");
  comment->Draw();
  c2->cd();
  pad11_2->cd();
  harea2p->Draw();
  pad12_2->cd();
  hmax2p->Draw();


}
示例#3
0
void MakeTRDFullMisAlignment(){
  // Create TClonesArray of full misalignment objects for TRD
  // Expects to read objects for FRAME
  // 
  TClonesArray *array = new TClonesArray("AliAlignObjParams",1000);
  const char* macroname = "MakeTRDFullMisAlignment.C";

  // Activate CDB storage and load geometry from CDB
  AliCDBManager* cdb = AliCDBManager::Instance();
  if(!cdb->IsDefaultStorageSet()) cdb->SetDefaultStorage("local://$ALICE_ROOT/OCDB");
  cdb->SetRun(0);
  
  AliCDBStorage* storage;
  TString Storage;
  
  if( TString(gSystem->Getenv("TOCDB")) == TString("kTRUE") ){
    Storage = gSystem->Getenv("STORAGE");
    if(!Storage.BeginsWith("local://") && !Storage.BeginsWith("alien://")) {
      Error(macroname,"STORAGE variable set to %s is not valid. Exiting\n",Storage.Data());
      return;
    }
    storage = cdb->GetStorage(Storage.Data());
    if(!storage){
      Error(macroname,"Unable to open storage %s\n",Storage.Data());
      return;
    }
    AliCDBPath path("GRP","Geometry","Data");
    AliCDBEntry *entry = storage->Get(path.GetPath(),cdb->GetRun());
    if(!entry) Fatal(macroname,"Could not get the specified CDB entry!");
    entry->SetOwner(0);
    TGeoManager* geom = (TGeoManager*) entry->GetObject();
    AliGeomManager::SetGeometry(geom);
  }else{
    AliGeomManager::LoadGeometry(); //load geom from default CDB storage
  }    
		  
  // load FRAME full misalignment objects (if needed, the macro
  // for FRAME has to be ran in advance) and apply them to geometry
  AliCDBPath fpath("GRP","Align","Data");
  AliCDBEntry *eFrame;
  if( TString(gSystem->Getenv("TOCDB")) == TString("kTRUE") ){
    Info(macroname,"Loading FRAME alignment objects from CDB storage %s",
      Storage.Data());
    eFrame = storage->Get(fpath.GetPath(),cdb->GetRun());
  }else{
    eFrame = cdb->Get(fpath.GetPath());
  }
  if(!eFrame) Fatal(macroname,"Could not get the specified CDB entry!");
  TClonesArray* arFrame = (TClonesArray*) eFrame->GetObject();
  arFrame->Sort();
  Int_t nvols = arFrame->GetEntriesFast();
  Bool_t flag = kTRUE;
  for(Int_t j=0; j<nvols; j++)
  {
    AliAlignObj* alobj = (AliAlignObj*) arFrame->UncheckedAt(j);
    if (alobj->ApplyToGeometry() == kFALSE) flag = kFALSE;
  }
  if(!flag) Fatal(macroname,"Error in the application of FRAME objects");

  // Sigmas for the chambers
  Double_t smdx    = 0.3; // 3 mm
  Double_t smdy    = 0.3; // 3 mm
  Double_t smdz    = 0.3; // 3 mm
  Double_t smrx    = 0.4 / 1000.0 / TMath::Pi()*180; // 0.4 mrad
  Double_t smry    = 2.0 / 1000.0 / TMath::Pi()*180; // 2.0 mrad
  Double_t smrz    = 0.4 / 1000.0 / TMath::Pi()*180; // 0.4 mrad
  // Truncation for the chambers
  Double_t cutSmdx = 3.0 * smdx;
  Double_t cutSmdy = 3.0 * smdy;
  Double_t cutSmdz = 3.0 * smdz;

  // Sigmas for the chambers
  Double_t chdx    = 0.05;  // 0.5 mm
  Double_t chdy    = 0.1;   // 1.0 mm
  Double_t chdz    = 0.007; // 70  microns
  Double_t chrx    = 0.0005 / 1000.0 / TMath::Pi()*180; // 0 mrad
  Double_t chry    = 0.0005 / 1000.0 / TMath::Pi()*180; // 0 mrad
  Double_t chrz    = 0.3    / 1000.0 / TMath::Pi()*180; // 0.3 mrad
  // Truncation for the chambers
  Double_t cutChdx = 1.0  * chdx;
  Double_t cutChdy = 1.0  * chdy;
  Double_t cutChdz = 0.14 * chdz;

  Int_t sActive[18]={1,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,1};
  Double_t dx,dy,dz,rx,ry,rz;

  Int_t j=0;
  UShort_t volid;
  const char *symname;

  // create the supermodules' alignment objects
  for (int iSect; iSect<18; iSect++) {
    TString sm_symname(Form("TRD/sm%02d",iSect));
    dx = AliMathBase::TruncatedGaus(0.0,smdx,cutSmdx); 
    dy = AliMathBase::TruncatedGaus(0.0,smdy,cutSmdy); 
    dz = AliMathBase::TruncatedGaus(0.0,smdz,cutSmdz); 
    rx = gRandom->Rndm() * 2.0*smrx - smrx;
    ry = gRandom->Rndm() * 2.0*smry - smry;
    rz = gRandom->Rndm() * 2.0*smrz - smrz;
    if( (TString(gSystem->Getenv("REALSETUP")) == TString("kTRUE")) && !sActive[iSect] ) continue;
    new((*array)[j++]) AliAlignObjParams(sm_symname.Data(),0,dx,dy,dz,rx,ry,rz,kFALSE);
  }
  // apply supermodules' alignment objects
  Int_t smCounter=0;
  for(Int_t iSect=0; iSect<18; iSect++){
    if( (TString(gSystem->Getenv("REALSETUP")) == TString("kTRUE")) && !sActive[iSect] ) continue;
    AliAlignObjParams* smobj =
      (AliAlignObjParams*)array->UncheckedAt(smCounter++);
    if(!smobj->ApplyToGeometry()){
      Fatal(macroname,Form("application of full misalignment object for sector %d failed!",iSect));
      return;
    }
  }

  // create the chambers' alignment objects
  ran = new TRandom(4357);
  Int_t chId;
  for (Int_t iLayer = AliGeomManager::kTRD1; iLayer <= AliGeomManager::kTRD6; iLayer++) {
    chId=-1;
    for (Int_t iSect = 0; iSect < 18; iSect++){
      for (Int_t iCh = 0; iCh < 5; iCh++) {
        dx = AliMathBase::TruncatedGaus(0.0,chdx,cutChdx); 
        dy = AliMathBase::TruncatedGaus(0.0,chdy,cutChdy); 
        dz = AliMathBase::TruncatedGaus(0.0,chdz,cutChdz); 
        rx = gRandom->Rndm() * 2.0*chrx - chrx;
        ry = gRandom->Rndm() * 2.0*chry - chry;
        rz = gRandom->Rndm() * 2.0*chrz - chrz;
        chId++;
        if ((iSect==13 || iSect==14 || iSect==15) && iCh==2) continue;
        volid = AliGeomManager::LayerToVolUID(iLayer,chId);
        if( (TString(gSystem->Getenv("REALSETUP")) == TString("kTRUE")) && !sActive[iSect] ) continue;
        symname = AliGeomManager::SymName(volid);
        new((*array)[j++]) AliAlignObjParams(symname,volid,dx,dy,dz,rx,ry,rz,kFALSE);
      }
    }
  }

  if( TString(gSystem->Getenv("TOCDB")) != TString("kTRUE") ){
    // save on file
    const char* filename = "TRDfullMisalignment.root";
    TFile f(filename,"RECREATE");
    if(!f){
      Error(macroname,"cannot open file for output\n");
      return;
    }
    Info(macroname,"Saving alignment objects to the file %s", filename);
    f.cd();
    f.WriteObject(array,"TRDAlignObjs","kSingleKey");
    f.Close();
  }else{
    // save in CDB storage
    Info(macroname,"Saving alignment objects in CDB storage %s",
	 Storage.Data());
    AliCDBMetaData* md = new AliCDBMetaData();
    md->SetResponsible("Dariusz Miskowiec");
    md->SetComment("Full misalignment for TRD");
    md->SetAliRootVersion(gSystem->Getenv("ARVERSION"));
    AliCDBId id("TRD/Align/Data",0,AliCDBRunRange::Infinity());
    storage->Put(array,id,md);
  }

  array->Delete();
}
示例#4
0
/**
 * 1. Data sample : pp200 W->e nu with  pile-up corresponding to 1 MHz min. bias
 * events, 50 K event y2011, 10 K event y2012.
 *
 * 2. Proof of principal: no pile-up for both PPV and KFV
 *
 *   a.  Reconstructed primary track multiplicity versus corresponding MC
 *   "reconstructable" (i.e. in n STAR acceptance,no. TPC MC hits >= 15)  tracks
 *   multiplicity.
 *
 *   b.  Corrected reconstructed primary track multiplicity (i.e. multiplied by
 *   QA/100.) versus corresponding MC "reconstructable"  (i.e. in n STAR
 *   acceptance,no. TPC MC hits >= 15)  tracks multiplicity.
 *
 *   c.  Efficiency primary vertex reconstruction versus  MC "reconstructable"
 *   tracks multiplicity.
 *
 * 3. With pileup. repeat above (a-c) with old ranking scheme for
 *
 *     I. Any reconstructed primary vertex which is matched with MC trigger
 *     vertex (MC = 1)
 *
 *    II. The best (in sense of ranking) reconstructed primary vertex which is
 *    matched with MC trigger vertex (MC = 1)
 *
 *   III. The best (in sense of ranking) reconstructed primary vertex which is
 *   not matched with MC trigger vertex (MC != 1)
 *
 * 4. With pileup. repeat above (a-c) with new ranking scheme for cases I-III
 */
void MuMcPrVKFV2012(Long64_t nevent, const char *file, const std::string& outFile, bool fillNtuple)
{
#ifdef __TMVA__
   boost::replace_last(outFile, ".root", "");
   outFile += ".TMVArank.root";

   // 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
   TString separator(":");
   TString Vnames(vnames);
   TObjArray *array = Vnames.Tokenize(separator);

   std::vector<std::string> inputVars;
   TIter next(array);
   TObjString *objs;

   while ((objs = (TObjString *) next())) {
      std::cout << objs->GetString() << std::endl;
   }

   inputVars.push_back("beam");
   inputVars.push_back("postx");
   inputVars.push_back("prompt");
   inputVars.push_back("cross");
   inputVars.push_back("tof");
   inputVars.push_back("notof");
   inputVars.push_back("EEMC");
   inputVars.push_back("noEEMC");
   inputVars.push_back("chi2");

   std::vector<double> *inputVec = new std::vector<double>( inputVars.size() );
   IClassifierReader *classReader = new ReadBDT( inputVars );

#endif /* __TMVA__ */

   TFile *fOut = TFile::Open(outFile.c_str(), "recreate");
   data_t data;

   // Book histograms
   const int nMcRecMult = 75;
   TArrayD xMult(nMcRecMult + 1);
   xMult[0] = -0.5;

   for (int i = 1; i <= nMcRecMult; i++) {
      if      (xMult[i - 1] <  50) xMult[i] = xMult[i - 1] +   1; //  1 - 50
      else if (xMult[i - 1] < 100) xMult[i] = xMult[i - 1] +   2; // 51 - 75
      else if (xMult[i - 1] < 200) xMult[i] = xMult[i - 1] +  10; // 76 - 85
      else                         xMult[i] = xMult[i - 1] + 100; // 86 -100
   }

   TH1D *McRecMulT = new TH1D("McRecMulT", "Reconstructable multiplicity for trigger Mc Vertex", nMcRecMult, xMult.GetArray());
   struct Name_t {
      const Char_t *Name;
      const Char_t *Title;
   };
   const Name_t HCases[3] = {
      {"Any", "Any vertex matched with MC == 1"},
      {"Good", "The best rank vertex with MC == 1"},
      {"Bad", "The best rank vertex with MC != 1"}
   };
   const Name_t Plots[4] = {
      {"Mult"    , "the reconstructed (uncorrected) track multiplicity versus Reconstructable multiplicity"},
      {"MultQA"  , "the reconstructed (corrected for QA) track multiplicity versus Reconstructable multiplicity"},
      {"McRecMul", "Reconstructable multiplicity"},
      {"YvsX"    , "Bad versus Good value"}
   };
   /*         h  p  */
   TH1 *hists[3][4];

   for (int h = 0; h < 3; h++) {
      for (int p = 0; p < 4; p++) {
         TString Name(Plots[p].Name); Name += HCases[h].Name;
         TString Title(Plots[p].Title); Title += " for "; Title += HCases[h].Title; Title += " vertex";

         if      (p <  2)  hists[h][p] = new TH2D(Name, Title, nMcRecMult, xMult.GetArray(), nMcRecMult, xMult.GetArray());
         else if (p == 2)  hists[h][p] = new TH1D(Name, Title, nMcRecMult, xMult.GetArray());
      }
   }

   TNtuple *VertexG = new TNtuple("VertexG", "good vertex & global params info", vnames);
   TNtuple *VertexB = new TNtuple("VertexB", "bad  vertex & global params info", vnames);
   // ----------------------------------------------
   StMuDstMaker *maker = new StMuDstMaker(0, 0, "", file, "st:MuDst.root", 1e9); // set up maker in read mode
   //                       0,0                        this mean read mode
   //                           dir                    read all files in this directory
   //                               file               bla.lis real all file in this list, if (file!="") dir is ignored
   //                                    filter        apply filter to filenames, multiple filters are separated by ':'
   //                                          10      maximum number of file to read
   maker->SetStatus("*", 0);

   std::vector<std::string> activeBranchNames = {
      "MuEvent",
      "PrimaryVertices",
      "StStMuMcVertex",
      "StStMuMcTrack"
   };

   // Set Active braches
   for (const auto& branchName : activeBranchNames)
      maker->SetStatus(branchName.c_str(), 1);

   TChain *tree = maker->chain();
   Long64_t nentries = tree->GetEntries();
   nevent = TMath::Min(nevent, nentries);
   std::cout << nentries << " events in chain " << nevent << " will be read." << std::endl;
   tree->SetCacheSize(-1);        //by setting the read cache to -1 we set it to the AutoFlush value when writing
   tree->SetCacheLearnEntries(1); //one entry is sufficient to learn
   tree->SetCacheEntryRange(0, nevent);

   for (Long64_t ev = 0; ev < nevent; ev++) {
      if (maker->Make()) break;

      StMuDst *muDst = maker->muDst();   // get a pointer to the StMuDst class, the class that points to all the data
      StMuEvent *muEvent = muDst->event(); // get a pointer to the class holding event-wise information
      int referenceMultiplicity = muEvent->refMult(); // get the reference multiplicity

      TClonesArray *PrimaryVertices   = muDst->primaryVertices();
      int nPrimaryVertices = PrimaryVertices->GetEntriesFast();

      TClonesArray *MuMcVertices   = muDst->mcArray(0);
      int nMuMcVertices = MuMcVertices->GetEntriesFast();

      TClonesArray *MuMcTracks     = muDst->mcArray(1);
      int nMuMcTracks = MuMcTracks->GetEntriesFast();

      if ( nevent >= 10 && ev % int(nevent*0.1) == 0 )
      {
         std::cout << "Event #" << ev << "\tRun\t" << muEvent->runId()
                   << "\tId: " << muEvent->eventId()
                   << " refMult= " << referenceMultiplicity
                   << "\tPrimaryVertices " << nPrimaryVertices
                   << "\t" << " " << nMuMcVertices
                   << "\t" << " " << nMuMcTracks
                   << std::endl;
      }

      //    const Double_t field = muEvent->magneticField()*kilogauss;
      if (! nMuMcVertices || ! nMuMcTracks || nPrimaryVertices <= 0) {
         std::cout << "Ev. " << ev << " has no MC information ==> skip it" << std::endl;
         std::cout << "OR no reconstructed verticies found" << std::endl;
         continue;
      }

      // Count number of MC tracks at a vertex with TPC reconstructable tracks
      std::multimap<int, int> Mc2McHitTracks;

      for (int m = 0; m < nMuMcTracks; m++) {
         StMuMcTrack *McTrack = (StMuMcTrack *) MuMcTracks->UncheckedAt(m);

         if (McTrack->No_tpc_hit() < 15) continue;

         Mc2McHitTracks.insert(std::pair<int, int>(McTrack->IdVx(), McTrack->Id()));
      }

      // This is the "reconstructable" track multiplicity
      int nMcTracksWithHits = Mc2McHitTracks.count(1);

      // Let's skip events in which we do not expect to reconstruct any tracks
      // (and thus vertex) from the primary vertex
      if (nMcTracksWithHits <= 0) continue;

      // This is our denominator histogram for efficiencies
      McRecMulT->Fill(nMcTracksWithHits);

      // =============  Build map between  Rc and Mc vertices
      std::map<StMuPrimaryVertex *, StMuMcVertex *> reco2McVertices;
      TArrayF vertexRanks(nPrimaryVertices);
      int mcMatchedVertexIndex  = -1; // any vertex with MC==1 and highest reconstrated multiplicity.
      int vertexMaxMultiplicity = -1;

      // First loop over all verticies in this event. There is at least one
      // must be available
      for (int recoVertexIndex = 0; recoVertexIndex < nPrimaryVertices; recoVertexIndex++)
      {
         vertexRanks[recoVertexIndex] = -1e10;

         StMuPrimaryVertex *recoVertex = (StMuPrimaryVertex *) PrimaryVertices->UncheckedAt(recoVertexIndex);

         if ( !AcceptVX(recoVertex) ) continue;

         // Check Mc
         if (recoVertex->idTruth() < 0 || recoVertex->idTruth() > nMuMcVertices) {
            std::cout << "ERROR: Illegal idTruth " << recoVertex->idTruth() << " The track is ignored" << std::endl;
            continue;
         }

         StMuMcVertex *mcVertex = (StMuMcVertex *) MuMcVertices->UncheckedAt(recoVertex->idTruth() - 1);

         if (mcVertex->Id() != recoVertex->idTruth()) {
            std::cout << "ERROR: Mismatched idTruth " << recoVertex->idTruth() << " and mcVertex Id " <<  mcVertex->Id()
                 << " The vertex is ignored" <<  std::endl;
            continue;
         }

         reco2McVertices[recoVertex] = mcVertex;
         vertexRanks[recoVertexIndex] = recoVertex->ranking();

         if (recoVertex->idTruth() == 1 && recoVertex->noTracks() > vertexMaxMultiplicity)
         {
            mcMatchedVertexIndex  = recoVertexIndex;
            vertexMaxMultiplicity = recoVertex->noTracks();
         }

         FillData(data, recoVertex);

#ifdef __TMVA__
         Float_t *dataArray = &data.beam;

         for (size_t j = 0; j < inputVec->size(); j++)
            (*inputVec)[j] = dataArray[j];

         vertexRanks[recoVertexIndex] = classReader->GetMvaValue( *inputVec );
#endif
      }

      // If we reconstructed a vertex which matches the MC one we fill the
      // numerator of the "Any" efficiency histogram
      if (mcMatchedVertexIndex != -1) {

         StMuPrimaryVertex *recoVertexMatchedMc = (StMuPrimaryVertex*) PrimaryVertices->UncheckedAt(mcMatchedVertexIndex);

         double nTracks = recoVertexMatchedMc->noTracks();
         double nTracksQA = nTracks * recoVertexMatchedMc->qaTruth() / 100.;

         hists[0][0]->Fill(nMcTracksWithHits, nTracks);
         hists[0][1]->Fill(nMcTracksWithHits, nTracksQA);
         hists[0][2]->Fill(nMcTracksWithHits);
      }

      // Now deal with the highest rank vertex
      int maxRankVertexIndex = TMath::LocMax(nPrimaryVertices, vertexRanks.GetArray());

      StMuPrimaryVertex *recoVertexMaxRank = (StMuPrimaryVertex*) PrimaryVertices->UncheckedAt(maxRankVertexIndex);
      StMuMcVertex *mcVertex = reco2McVertices[recoVertexMaxRank];

      double nTracks = recoVertexMaxRank->noTracks();
      double nTracksQA = nTracks * recoVertexMaxRank->qaTruth() / 100.;

      // Fill numerator for "good" and "bad" efficiencies
      int h = ( mcVertex && mcVertex->Id() == 1) ? 1 : 2;

      hists[h][0]->Fill(nMcTracksWithHits, nTracks);
      hists[h][1]->Fill(nMcTracksWithHits, nTracksQA);
      hists[h][2]->Fill(nMcTracksWithHits);


      // Proceed with filling ntuple only if requested by the user
      if ( !fillNtuple ) continue;


      // Second loop over all verticies in this event
      for (int recoVertexIndex = 0; recoVertexIndex < nPrimaryVertices; recoVertexIndex++)
      {
         StMuPrimaryVertex *recoVertex = (StMuPrimaryVertex *) PrimaryVertices->UncheckedAt(recoVertexIndex);

         if ( !AcceptVX(recoVertex) ) continue;

         StMuMcVertex *mcVertex = reco2McVertices[recoVertex];

         if ( !mcVertex ) {
            std::cout << "No Match from RC to MC" << std::endl;
            continue;
         }

         if (vtxeval::gDebugFlag) {
            std::cout << Form("Vx[%3i]", recoVertexIndex) << *recoVertex << " " << *mcVertex;
            int nMcTracksWithHitsatL = Mc2McHitTracks.count(recoVertex->idTruth());
            std::cout << Form("Number of McTkHit %4i rank %8.3f", nMcTracksWithHitsatL, vertexRanks[recoVertexIndex]);
         }

         int IdPar = mcVertex->IdParTrk();

         if (IdPar > 0 && IdPar <= nMuMcTracks) {
            StMuMcTrack *mcTrack = (StMuMcTrack *) MuMcTracks->UncheckedAt(IdPar - 1);

            if (mcTrack && vtxeval::gDebugFlag) std::cout << " " << mcTrack->GeName();
         }

         FillData(data, recoVertex);

         double nTracks = recoVertex->noTracks();

         if (mcVertex->Id() == 1 && nTracks == vertexMaxMultiplicity) {// good
            VertexG->Fill(&data.beam);
         }
         else {   // bad
            VertexB->Fill(&data.beam);
         }
      }

      if ( !gROOT->IsBatch() ) {
         if (vtxeval::ask_user()) return;
      }
      else { vtxeval::gDebugFlag = false; }
   }

   fOut->Write();
}