Classes
struct	ConversionInfo

Typedefs
using	TrackRowView = TrackTable::const_iterator::value_type

using	TrackTable = edm::soa::AddColumns< edm::soa::PtEtaPhiTable, TrackTableSpecificColumns >::type

using	TrackTableSpecificColumns = std::tuple< edm::soa::col::Pz, edm::soa::col::PtError, edm::soa::col::MissingInnerHits, edm::soa::col::NumberOfValidHits, edm::soa::col::Charge, edm::soa::col::D0 >

using	TrackTableView = edm::soa::ViewFromTable_t< TrackTable >

Functions
ConversionInfo const &	arbitrateConversionPartnersbyR (const std::vector< ConversionInfo > &convCandidates)

ConversionInfo	findBestConversionMatch (const std::vector< ConversionInfo > &v_convCandidates)

ConversionInfo	findConversion (const reco::GsfElectronCore &gsfElectron, TrackTableView ctfTable, TrackTableView gsfTable, float bFieldAtOrigin, float minFracSharedHits=0.45f)

std::vector< ConversionInfo >	findConversions (const reco::GsfElectronCore &gsfElectron, TrackTableView ctfTable, TrackTableView gsfTable, float bFieldAtOrigin, float minFracSharedHits)

ConversionInfo	getConversionInfo (reco::Track const &el_track, TrackRowView const &track, float bFieldAtOrigin)

Variables
constexpr float	dR2Max = square(0.5f)

constexpr int	maxDeltaMissingHits = 2

constexpr int	maxDeltaMissingHitsForKFtoKF = 3

constexpr float	maxDist2Dcot2 = square(0.05f)

constexpr float	maxDistOrCotForKFtoKF = 0.02f

constexpr float	maxRefPtErrorForKfConv = 0.05f

constexpr float	maxRelGsfPtError = 0.5f

constexpr float	maxRelGsfPtErrorForKfConv = 0.25f

constexpr float	maxRelPtDiffForGsfConv = 0.25f

constexpr float	maxRelPtDiffForKfConv = 0.2f

constexpr int	minNumberOfValidHits = 5

Typedef Documentation

◆ TrackRowView

using egamma::conv::TrackRowView = typedef TrackTable::const_iterator::value_type

Definition at line 58 of file ConversionFinder.h.

◆ TrackTable

using egamma::conv::TrackTable = typedef edm::soa::AddColumns<edm::soa::PtEtaPhiTable, TrackTableSpecificColumns>::type

Definition at line 56 of file ConversionFinder.h.

◆ TrackTableSpecificColumns

using egamma::conv::TrackTableSpecificColumns = typedef std::tuple<edm::soa::col::Pz, edm::soa::col::PtError, edm::soa::col::MissingInnerHits, edm::soa::col::NumberOfValidHits, edm::soa::col::Charge, edm::soa::col::D0>

Definition at line 55 of file ConversionFinder.h.

◆ TrackTableView

using egamma::conv::TrackTableView = typedef edm::soa::ViewFromTable_t<TrackTable>

Definition at line 57 of file ConversionFinder.h.

Function Documentation

◆ arbitrateConversionPartnersbyR()

ConversionInfo const& egamma::conv::arbitrateConversionPartnersbyR ( const std::vector< ConversionInfo > & convCandidates )

Definition at line 209 of file ConversionFinder.cc.

                                                                                                         {
     ConversionInfo const* closestConversion = &convCandidates.front();
  
     if (convCandidates.size() == 1)
       return *closestConversion;
  
     float R = pow(closestConversion->dist, 2) + pow(closestConversion->dcot, 2);
  
     for (auto const& temp : convCandidates) {
       float temp_R = pow(temp.dist, 2) + pow(temp.dcot, 2);
       if (temp_R < R) {
         R = temp_R;
         closestConversion = &temp;
       }
     }
  
     return *closestConversion;
   }

References egamma::conv::ConversionInfo::dcot, egamma::conv::ConversionInfo::dist, funct::pow(), dttmaxenums::R, and groupFilesInBlocks::temp.

Referenced by findBestConversionMatch().

◆ findBestConversionMatch()

ConversionInfo egamma::conv::findBestConversionMatch ( const std::vector< ConversionInfo > & v_convCandidates )

Definition at line 229 of file ConversionFinder.cc.

                                                                                             {
     using namespace std;
  
     if (v_convCandidates.empty())
       return {};
  
     if (v_convCandidates.size() == 1)
       return v_convCandidates.at(0);
  
     vector<ConversionInfo> v_0;
     vector<ConversionInfo> v_1;
     vector<ConversionInfo> v_2;
     vector<ConversionInfo> v_3;
     //loop over the candidates
  
     for (auto const& temp : v_convCandidates) {
       if (temp.radiusOfConversion <= -2)
         continue;
  
       if (temp.flag == 0) {
         if ((std::abs(temp.dist) < maxDistOrCotForKFtoKF && std::abs(temp.dcot) < maxDistOrCotForKFtoKF &&
              temp.deltaMissingHits < maxDeltaMissingHitsForKFtoKF) ||
             (pow(temp.dist, 2) + pow(temp.dcot, 2) < maxDist2Dcot2 && temp.deltaMissingHits < maxDeltaMissingHits))
           v_0.push_back(temp);
       }
  
       if (temp.flag == 1) {
         if (square(temp.dist) + square(temp.dcot) < maxDist2Dcot2 && temp.deltaMissingHits < maxDeltaMissingHits)
           v_1.push_back(temp);
       }
       if (temp.flag == 2) {
         if (square(temp.dist) + square(temp.dcot) < maxDist2Dcot2 && temp.deltaMissingHits < maxDeltaMissingHits)
           v_2.push_back(temp);
       }
       if (temp.flag == 3) {
         if (square(temp.dist) + square(temp.dcot) < maxDist2Dcot2 && temp.deltaMissingHits < maxDeltaMissingHits)
           v_3.push_back(temp);
       }
  
     }  //candidate conversion loop
  
     //now do some arbitration
  
     //give preference to conversion partners found in the CTF collection
     //using the electron's CTF track
     if (!v_0.empty())
       return arbitrateConversionPartnersbyR(v_0);
  
     if (!v_1.empty())
       return arbitrateConversionPartnersbyR(v_1);
  
     if (!v_2.empty())
       return arbitrateConversionPartnersbyR(v_2);
  
     if (!v_3.empty())
       return arbitrateConversionPartnersbyR(v_3);
  
     //if we get here, we didn't find a candidate conversion partner that
     //satisfied even the loose selections
     //return the the closest partner by R
     return arbitrateConversionPartnersbyR(v_convCandidates);
   }

References funct::abs(), arbitrateConversionPartnersbyR(), maxDeltaMissingHits, maxDeltaMissingHitsForKFtoKF, maxDist2Dcot2, maxDistOrCotForKFtoKF, funct::pow(), square(), and groupFilesInBlocks::temp.

Referenced by findConversion().

◆ findConversion()

ConversionInfo egamma::conv::findConversion	(	const reco::GsfElectronCore &	gsfElectron,
		TrackTableView	ctfTable,
		TrackTableView	gsfTable,
		float	bFieldAtOrigin,
		float	minFracSharedHits = `0.45f`
	)

inline

Definition at line 71 of file ConversionFinder.h.

                                                                         {
     return findBestConversionMatch(findConversions(gsfElectron, ctfTable, gsfTable, bFieldAtOrigin, minFracSharedHits));
   }

References findBestConversionMatch(), and findConversions().

Referenced by GsfElectronAlgo::createElectron().

◆ findConversions()

std::vector< ConversionInfo > egamma::conv::findConversions	(	const reco::GsfElectronCore &	gsfElectron,
		TrackTableView	ctfTable,
		TrackTableView	gsfTable,
		float	bFieldAtOrigin,
		float	minFracSharedHits
	)

Definition at line 29 of file ConversionFinder.cc.

                                                                        {
     using namespace reco;
     using namespace std;
     using namespace edm;
     using namespace edm::soa::col;
  
     //get the references to the gsf and ctf tracks that are made
     //by the electron
     const reco::TrackRef eleCtfTk = gsfElectron.ctfTrack();
     const reco::GsfTrackRef& eleGsfTk = gsfElectron.gsfTrack();
  
     float eleGsfPt = eleGsfTk->pt();
     float eleGsfEta = eleGsfTk->eta();
     float eleGsfPhi = eleGsfTk->phi();
  
     const bool useEleCtfTrack = eleCtfTk.isNonnull() && gsfElectron.ctfGsfOverlap() > minFracSharedHits;
  
     std::optional<float> eleCtfPt = std::nullopt;
     std::optional<float> eleCtfEta = std::nullopt;
     std::optional<float> eleCtfPhi = std::nullopt;
  
     //the electron's CTF track must share at least 45% of the inner hits
     //with the electron's GSF track
     std::optional<int> ctfidx = std::nullopt;
     int gsfidx = static_cast<int>(eleGsfTk.key());
  
     if (useEleCtfTrack) {
       eleCtfPt = eleCtfTk->pt();
       eleCtfEta = eleCtfTk->eta();
       eleCtfPhi = eleCtfTk->phi();
       ctfidx = static_cast<int>(eleCtfTk.key());
     }
  
     //these vectors are for those candidate partner tracks that pass our cuts
     std::vector<ConversionInfo> v_candidatePartners;
     //track indices required to make references
     int ctftk_i = 0;
     int gsftk_i = 0;
  
     //loop over the CTF tracks and try to find the partner track
     for (auto ctftkItr = ctfTable.begin(); ctftkItr != ctfTable.end(); ++ctftkItr, ctftk_i++) {
       if (useEleCtfTrack && ctftk_i == ctfidx.value())
         continue;
  
       auto ctftk = *ctftkItr;
  
       //apply quality cuts to remove bad tracks
       if (ctftk.get<PtError>() > maxRefPtErrorForKfConv * ctftk.get<Pt>() ||
           ctftk.get<NumberOfValidHits>() < minNumberOfValidHits)
         continue;
  
       if (useEleCtfTrack && std::abs(ctftk.get<Pt>() - eleCtfPt.value()) < maxRelPtDiffForKfConv * eleCtfPt.value())
         continue;
  
       //use the electron's CTF track, if not null, to search for the partner track
       //look only in a cone of 0.5 to save time, and require that the track is opp. sign
       if (useEleCtfTrack && eleCtfTk->charge() + ctftk.get<Charge>() == 0 &&
           deltaR2(eleCtfEta.value(), eleCtfPhi.value(), ctftk.get<Eta>(), ctftk.get<Phi>()) < dR2Max) {
         ConversionInfo convInfo = getConversionInfo(*eleCtfTk, ctftk, bFieldAtOrigin);
  
         //need to add the track reference information for completeness
         //because the overloaded fnc above does not make a trackRef
         int deltaMissingHits = ctftk.get<MissingInnerHits>() - eleCtfTk->missingInnerHits();
  
         v_candidatePartners.push_back(
             {convInfo.dist, convInfo.dcot, convInfo.radiusOfConversion, ctftk_i, std::nullopt, deltaMissingHits, 0});
  
       }  //using the electron's CTF track
  
       //now we check using the electron's gsf track
       if (eleGsfTk->charge() + ctftk.get<Charge>() == 0 &&
           deltaR2(eleGsfEta, eleGsfPhi, ctftk.get<Eta>(), ctftk.get<Phi>()) < dR2Max &&
           eleGsfTk->ptError() < maxRelGsfPtErrorForKfConv * eleGsfPt) {
         int deltaMissingHits = ctftk.get<MissingInnerHits>() - eleGsfTk->missingInnerHits();
  
         ConversionInfo convInfo = getConversionInfo(*eleGsfTk, ctftk, bFieldAtOrigin);
  
         v_candidatePartners.push_back(
             {convInfo.dist, convInfo.dcot, convInfo.radiusOfConversion, ctftk_i, std::nullopt, deltaMissingHits, 1});
       }  //using the electron's GSF track
  
     }  //loop over the CTF track collection
  
     //------------------------------------------------------ Loop over GSF collection ----------------------------------//
     for (auto gsftkItr = gsfTable.begin(); gsftkItr != gsfTable.end(); ++gsftkItr, gsftk_i++) {
       //reject the electron's own gsfTrack
       if (gsfidx == gsftk_i)
         continue;
  
       auto gsftk = *gsftkItr;
  
       //apply quality cuts to remove bad tracks
       if (gsftk.get<PtError>() > maxRelGsfPtError * gsftk.get<Pt>() ||
           gsftk.get<NumberOfValidHits>() < minNumberOfValidHits)
         continue;
  
       if (std::abs(gsftk.get<Pt>() - eleGsfPt) < maxRelPtDiffForGsfConv * eleGsfPt)
         continue;
  
       //try using the electron's CTF track first if it exists
       //look only in a cone of 0.5 around the electron's track
       //require opposite sign
       if (useEleCtfTrack && eleCtfTk->charge() + gsftk.get<Charge>() == 0 &&
           deltaR2(eleCtfEta.value(), eleCtfPhi.value(), gsftk.get<Eta>(), gsftk.get<Phi>()) < dR2Max) {
         int deltaMissingHits = gsftk.get<MissingInnerHits>() - eleCtfTk->missingInnerHits();
  
         ConversionInfo convInfo = getConversionInfo(*eleCtfTk, gsftk, bFieldAtOrigin);
         //fill the Ref info
         v_candidatePartners.push_back(
             {convInfo.dist, convInfo.dcot, convInfo.radiusOfConversion, std::nullopt, gsftk_i, deltaMissingHits, 2});
       }
  
       //use the electron's gsf track
       if (eleGsfTk->charge() + gsftk.get<Charge>() == 0 &&
           deltaR2(eleGsfEta, eleGsfPhi, gsftk.get<Eta>(), gsftk.get<Phi>()) < dR2Max &&
           (eleGsfTk->ptError() < maxRelGsfPtError * eleGsfPt)) {
         ConversionInfo convInfo = getConversionInfo(*eleGsfTk, gsftk, bFieldAtOrigin);
         //fill the Ref info
  
         int deltaMissingHits = gsftk.get<MissingInnerHits>() - eleGsfTk->missingInnerHits();
  
         v_candidatePartners.push_back(
             {convInfo.dist, convInfo.dcot, convInfo.radiusOfConversion, std::nullopt, gsftk_i, deltaMissingHits, 3});
       }
     }  //loop over the gsf track collection
  
     return v_candidatePartners;
   }

References funct::abs(), reco::GsfElectronCore::ctfGsfOverlap(), reco::GsfElectronCore::ctfTrack(), egamma::conv::ConversionInfo::dcot, HLTMuonOfflineAnalyzer_cfi::deltaR2, egamma::conv::ConversionInfo::dist, dR2Max, edm::Ref< C, T, F >::get(), getConversionInfo(), reco::GsfElectronCore::gsfTrack(), edm::Ref< C, T, F >::isNonnull(), edm::Ref< C, T, F >::key(), maxRefPtErrorForKfConv, maxRelGsfPtError, maxRelGsfPtErrorForKfConv, maxRelPtDiffForGsfConv, maxRelPtDiffForKfConv, minNumberOfValidHits, and egamma::conv::ConversionInfo::radiusOfConversion.

Referenced by findConversion().

◆ getConversionInfo()

ConversionInfo egamma::conv::getConversionInfo	(	reco::Track const &	el_track,
		TrackRowView const &	track,
		float	bFieldAtOrigin
	)

Definition at line 163 of file ConversionFinder.cc.

                                                                                                           {
     using namespace edm::soa::col;
  
     //now calculate the conversion related information
     float rEl = 100.f * ele.pt() / (-0.3f * bFieldAtOrigin * ele.charge());
     float xEl = -1.f * (rEl - ele.d0()) * sin(ele.phi());
     float yEl = (rEl - ele.d0()) * cos(ele.phi());
     rEl = std::abs(rEl);
  
     float rCand = 100.f * track.get<Pt>() / (-0.3f * bFieldAtOrigin * track.get<Charge>());
     float xCand = -1.f * (rCand - track.get<D0>()) * sin(track.get<Phi>());
     float yCand = (rCand - track.get<D0>()) * cos(track.get<Phi>());
     rCand = std::abs(rCand);
  
     float d = sqrt(pow(xEl - xCand, 2) + pow(yEl - yCand, 2));
     float dist = d - (rEl + rCand);
  
     // this is equivalent to `1/tan(theta_1) - 1/tan(theta_2)` but requires less trigonometry
     float dcot = ele.pz() / ele.pt() - track.get<Pz>() / track.get<Pt>();
  
     //get the point of conversion
     float xa1 = xEl + (xCand - xEl) * rEl / d;
     float xa2 = xCand + (xEl - xCand) * rCand / d;
     float ya1 = yEl + (yCand - yEl) * rEl / d;
     float ya2 = yCand + (yEl - yCand) * rCand / d;
  
     float x = .5f * (xa1 + xa2);
     float y = .5f * (ya1 + ya2);
     float rconv = sqrt(pow(x, 2) + pow(y, 2));
     // The z-position of the conversion is unused, but here is how it could be computed if needed:
     // float z = ele.dz() + rEl * ele.pz() * std::acos(1 - pow(rconv, 2) / (2. * pow(rEl, 2))) / ele.pt();
  
     //now assign a sign to the radius of conversion
     float tempsign = ele.px() * x + ele.py() * y;
     tempsign = tempsign / std::abs(tempsign);
     rconv = tempsign * rconv;
  
     //return an instance of ConversionInfo, but with a NULL track refs
     return {dist, dcot, rconv};
   }