Inheritance diagram for L1NNTauProducer:

Public Member Functions
	L1NNTauProducer (const edm::ParameterSet &, const tensorflow::SessionCache *)

	~L1NNTauProducer () override

Public Member Functions inherited from edm::stream::EDProducer< edm::GlobalCache< tensorflow::SessionCache > >
	EDProducer ()=default

	EDProducer (const EDProducer &)=delete

bool	hasAbilityToProduceInBeginLumis () const final

bool	hasAbilityToProduceInBeginProcessBlocks () const final

bool	hasAbilityToProduceInBeginRuns () const final

bool	hasAbilityToProduceInEndLumis () const final

bool	hasAbilityToProduceInEndProcessBlocks () const final

bool	hasAbilityToProduceInEndRuns () const final

const EDProducer &	operator= (const EDProducer &)=delete

Static Public Member Functions
static void	fillDescriptions (edm::ConfigurationDescriptions &descriptions)

static void	globalEndJob (const tensorflow::SessionCache *)

static std::unique_ptr< tensorflow::SessionCache >	initializeGlobalCache (const edm::ParameterSet &)

Private Member Functions
void	addTau (const l1t::PFCandidate &iCand, const l1t::PFCandidateCollection &iParts, std::unique_ptr< PFTauCollection > &outputTaus)

void	makeTau_HW (const l1t::PFCandidate &seed, l1t::PFCandidateCollection &parts, std::unique_ptr< l1t::PFTauCollection > &iTaus)

void	process_HW (const l1t::PFCandidateCollection &parts, std::unique_ptr< l1t::PFTauCollection > &iTaus)

void	process_SW (const l1t::PFCandidateCollection &parts, std::unique_ptr< l1t::PFTauCollection > &iTaus)

void	produce (edm::Event &iEvent, const edm::EventSetup &iSetup) override

Private Attributes
double	fConeSize_

const bool	fDebug

const bool	fEMSeed

const bool	fHW

edm::EDGetTokenT< vector< l1t::PFCandidate > >	fL1PFToken_

int	fMaxTaus_

int	fNParticles_

double	fSeedPt_

std::unique_ptr< TauNNId >	fTauNNId_

std::unique_ptr< TauNNIdHW >	fTauNNIdHW_

double	fTauSize_

Additional Inherited Members
Public Types inherited from edm::stream::EDProducer< edm::GlobalCache< tensorflow::SessionCache > >
using	CacheTypes = CacheContexts< T... >

using	GlobalCache = typename CacheTypes::GlobalCache

using	HasAbility = AbilityChecker< T... >

using	InputProcessBlockCache = typename CacheTypes::InputProcessBlockCache

using	LuminosityBlockCache = typename CacheTypes::LuminosityBlockCache

using	LuminosityBlockContext = LuminosityBlockContextT< LuminosityBlockCache, RunCache, GlobalCache >

using	LuminosityBlockSummaryCache = typename CacheTypes::LuminosityBlockSummaryCache

using	RunCache = typename CacheTypes::RunCache

using	RunContext = RunContextT< RunCache, GlobalCache >

using	RunSummaryCache = typename CacheTypes::RunSummaryCache

Detailed Description

Definition at line 22 of file L1NNTauProducer.cc.

Constructor & Destructor Documentation

◆ L1NNTauProducer()

L1NNTauProducer::L1NNTauProducer	(	const edm::ParameterSet &	cfg,
		const tensorflow::SessionCache *	cache
	)

explicit

Definition at line 60 of file L1NNTauProducer.cc.

References utilities::cache(), looper::cfg, fHW, fNParticles_, fTauNNId_, fTauNNIdHW_, and AlCaHLTBitMon_QueryRunRegistry::string.

     : fSeedPt_(cfg.getParameter<double>("seedpt")),
       fConeSize_(cfg.getParameter<double>("conesize")),
       fTauSize_(cfg.getParameter<double>("tausize")),
       fMaxTaus_(cfg.getParameter<int>("maxtaus")),
       fNParticles_(cfg.getParameter<int>("nparticles")),
       fHW(cfg.getParameter<bool>("HW")),
       fEMSeed(cfg.getParameter<bool>("emseed")),
       fDebug(cfg.getParameter<bool>("debug")),
       fL1PFToken_(consumes<vector<l1t::PFCandidate>>(cfg.getParameter<edm::InputTag>("L1PFObjects"))) {
   std::string lNNFile = cfg.getParameter<std::string>("NNFileName");  //,"L1Trigger/Phase2L1Taus/data/tau_3layer.pb");
   if (fHW) {
     fTauNNIdHW_ = std::make_unique<TauNNIdHW>();
     fTauNNIdHW_->initialize("input_1:0", fNParticles_);
   } else {
     fTauNNId_ = std::make_unique<TauNNId>(lNNFile.find("v0") == std::string::npos ? "input_1:0" : "dense_1_input:0",
                                           cache->getSession(),
                                           lNNFile,
                                           fNParticles_);
   }
   produces<l1t::PFTauCollection>("L1PFTausNN");
 }

◆ ~L1NNTauProducer()

L1NNTauProducer::~L1NNTauProducer ( )

override

Definition at line 335 of file L1NNTauProducer.cc.

335 {}

Member Function Documentation

◆ addTau()

void L1NNTauProducer::addTau	(	const l1t::PFCandidate &	iCand,
		const l1t::PFCandidateCollection &	iParts,
		std::unique_ptr< PFTauCollection > &	outputTaus
	)

private

Definition at line 138 of file L1NNTauProducer.cc.

References l1t::PFCandidate::ChargedHadron, reco::deltaR2(), PVValHelper::dxy, l1t::PFCandidate::Electron, fConeSize_, fSeedPt_, fTauNNId_, fTauSize_, mps_fire::i, dqmiolumiharvest::j, mkfit::NN, l1t::PFCandidate::Photon, jetUpdater_cfi::sort, and HLTMuonOfflineAnalyzer_cfi::z0.

Referenced by process_SW().

                                                                             {
   l1t::PFCandidateCollection pfTauCands;
   math::PtEtaPhiMLorentzVector lTot(0, 0, 0, 0);
   math::PtEtaPhiMLorentzVector lCand(0, 0, 0, 0);
   int lId = 0;
   float z0 = 0;
   float dxy = 0;
   for (const auto& l1PFCand : iParts) {
     if (reco::deltaR2(iCand, l1PFCand) > fConeSize_ * fConeSize_)
       continue;
     math::PtEtaPhiMLorentzVector pVec(l1PFCand.pt(), l1PFCand.eta(), l1PFCand.phi(), 0);
     lTot += pVec;
     if (reco::deltaR2(iCand, l1PFCand) < fTauSize_ * fTauSize_ &&
         (l1PFCand.id() == l1t::PFCandidate::Electron || l1PFCand.id() == l1t::PFCandidate::ChargedHadron ||
          l1PFCand.id() == l1t::PFCandidate::Photon)) {
       lId++;
       lCand += pVec;
       if (z0 == 0 && l1PFCand.id() == l1t::PFCandidate::ChargedHadron) {
         z0 = l1PFCand.z0();
         dxy = l1PFCand.dxy();
       }
     }
     pfTauCands.push_back(l1PFCand);
   }
   if (lTot.Pt() < fSeedPt_)
     return;
   std::sort(
       pfTauCands.begin(), pfTauCands.end(), [](l1t::PFCandidate i, l1t::PFCandidate j) { return (i.pt() > j.pt()); });
   float NN = fTauNNId_->compute(iCand, pfTauCands);
   float* lNNVector = fTauNNId_->NNVectorVar();
   math::PtEtaPhiMLorentzVector tempP4(lCand.Pt(), lCand.Eta(), lCand.Phi(), lCand.M() * lCand.M());
   l1t::PFTau l1PFTau(tempP4, lNNVector, NN, 0, lId);
   l1PFTau.setZ0(z0);
   l1PFTau.setDxy(dxy);
   outputTaus->push_back(l1PFTau);
 }

◆ fillDescriptions()

void L1NNTauProducer::fillDescriptions ( edm::ConfigurationDescriptions & descriptions )

static

Definition at line 176 of file L1NNTauProducer.cc.

References edm::ConfigurationDescriptions::add(), submitPVResolutionJobs::desc, ProducerED_cfi::InputTag, and AlCaHLTBitMon_QueryRunRegistry::string.

                                                                                  {
   // L1NNTauProducer
   edm::ParameterSetDescription desc;
   desc.add<std::string>("NNFileName", "L1Trigger/Phase2L1ParticleFlow/data/tau_3layer.pb");
   desc.add<double>("tausize", 0.1);
   desc.add<int>("maxtaus", 5);
   desc.add<int>("nparticles", 10);
   desc.add<double>("conesize", 0.4);
   desc.add<double>("seedpt", 20);
   desc.add<bool>("HW", true);
   desc.add<bool>("emseed", true);
   desc.add<bool>("debug", false);
   desc.add<edm::InputTag>("L1PFObjects", edm::InputTag("L1PFProducer", "l1pfCandidates"));
   descriptions.add("L1NNTauProducer", desc);
 }

◆ globalEndJob()

static void L1NNTauProducer::globalEndJob ( const tensorflow::SessionCache * )

inlinestatic

Definition at line 29 of file L1NNTauProducer.cc.

29 {};

◆ initializeGlobalCache()

std::unique_ptr< tensorflow::SessionCache > L1NNTauProducer::initializeGlobalCache ( const edm::ParameterSet & cfg )

static

Definition at line 83 of file L1NNTauProducer.cc.

References looper::cfg, edm::FileInPath::fullPath(), HLT_2024v14_cff::graphPath, and AlCaHLTBitMon_QueryRunRegistry::string.

                                                                                                        {
   std::string graphPath = edm::FileInPath(cfg.getParameter<std::string>("NNFileName")).fullPath();
   return std::make_unique<tensorflow::SessionCache>(graphPath);
 }

◆ makeTau_HW()

void L1NNTauProducer::makeTau_HW	(	const l1t::PFCandidate &	seed,
		l1t::PFCandidateCollection &	parts,
		std::unique_ptr< l1t::PFTauCollection > &	iTaus
	)

private

Definition at line 192 of file L1NNTauProducer.cc.

References l1t::PFCandidate::ChargedHadron, PVValHelper::dxy, l1t::PFCandidate::Electron, PVValHelper::eta, L1TauEmu::etaphi_base, l1gt::Scales::floatEta(), l1gt::Scales::floatPhi(), l1gt::Scales::floatPt(), fSeedPt_, fTauNNIdHW_, fTauSize_, l1ct::Tau::hwCharge, l1ct::Tau::hwEta, l1ct::Tau::hwPhi, l1ct::Tau::hwPt, l1ct::Tau::hwRawId, l1ct::Tau::hwSeedPt, l1ct::Tau::hwSeedZ0, l1ct::Tau::hwType, L1TauEmu::inCone(), l1ct::Scales::makeGlbEta(), l1ct::Scales::makeGlbPhi(), EgHLTOffHistBins_cfi::mass, Tau_NN_Result::nn_id, Tau_NN_Result::nn_pt_correction, l1gt::Tau::pack(), phi, l1t::PFCandidate::Photon, DiDispStaMuonMonitor_cfi::pt, fileCollector::seed, l1t::PFTau::set_encodedTau(), l1t::PFTau::setDxy(), l1t::PFTau::setZ0(), l1ct::Tau::toGT(), and HLTMuonOfflineAnalyzer_cfi::z0.

Referenced by process_HW().

                                                                            {
   // Seed Cone Jet algorithm with ap_fixed types and hardware emulation
   L1TauEmu::detaphi_t rCone2 =
       L1TauEmu::detaphi_t(fTauSize_ * fTauSize_ * L1TauEmu::etaphi_base * L1TauEmu::etaphi_base);
   unsigned lId = 0;
 
   input2_t p1_tot = 0;
   input2_t p1x_tot = 0;
   input2_t p1y_tot = 0;
   input2_t p1z_tot = 0;
 
   input_t e1ta_1 = seed.eta();
   input_t p1hi_1 = seed.phi();
   L1TauEmu::pt_t pt = 0;
   L1TauEmu::z0_t z0 = 0;
   L1TauEmu::dxy_t dxy = 0;
 
   // Reconstruct the Tau Cone
   for (unsigned i0 = 0; i0 < parts.size(); i0++) {
     if (L1TauEmu::inCone(seed, (parts[i0]), rCone2)) {
       if (parts[i0].id() == l1t::PFCandidate::Electron || parts[i0].id() == l1t::PFCandidate::ChargedHadron ||
           parts[i0].id() == l1t::PFCandidate::Photon) {
         lId++;
         pt = pt + L1TauEmu::pt_t(parts[i0].pt());
 
         input2_t d1eta = input_t(parts[i0].eta()) - e1ta_1;
         input2_t d1phi = input_t(parts[i0].phi()) - p1hi_1;
         input2_t d1r2 = d1eta * d1eta + d1phi * d1phi;
         input2_t tmppt = input_t(parts[i0].pt());
         input2_t half = 0.5;
         p1z_tot = p1z_tot + tmppt * (1 - d1r2 * half);
         p1y_tot = p1y_tot + tmppt * d1phi;
         p1x_tot = p1x_tot + tmppt * d1eta;
         p1_tot = p1_tot + tmppt;
 
         if (z0 == 0 && parts[i0].id() == l1t::PFCandidate::ChargedHadron) {
           z0 = parts[i0].hwZ0();
           dxy = parts[i0].hwDxy();
         }
       }
     }
   }
 
   //Compute the mass
   input2_t tmpmass1 = (p1_tot * p1_tot - p1x_tot * p1x_tot - p1y_tot * p1y_tot - p1z_tot * p1z_tot);
   if (tmpmass1 < 0)
     tmpmass1 = 0;
   L1TauEmu::pt_t mass = l1ct::pt_t(tmpmass1);
 
   if (pt < fSeedPt_)
     return;
 
   // Tau NN Inference
   Tau_NN_Result NN_ouput = fTauNNIdHW_->compute(seed, parts);
 
   // Needed for making PFTau
   input_t* lNNVector = fTauNNIdHW_->NNVectorVar();
   float pNNVec[80];
   for (unsigned i0 = 0; i0 < 80; i0++)
     pNNVec[i0] = float(lNNVector[i0]);
 
   //Firmware Tau
   l1ct::Tau l1ctTau;
   l1ctTau.hwPt = l1ct::pt_t(pt * NN_ouput.nn_pt_correction);  //l1gt is <16,11> and currently <16,14>
   l1ctTau.hwEta = l1ct::Scales::makeGlbEta(seed.eta());       // seed.eta() and seed.phi() are in physical coordinates
   l1ctTau.hwPhi = l1ct::Scales::makeGlbPhi(seed.phi());
 
   l1ctTau.hwSeedPt = seed.pt();
   l1ctTau.hwSeedZ0 = seed.hwZ0();
   l1ctTau.hwCharge = seed.charge();
 
   l1ctTau.hwType = l1ct::Tau::type_t(lId);
   l1ctTau.hwRawId = ap_uint<10>(NN_ouput.nn_id * 1024);  //NN Output is ap_fixed<16, 6> so need to cast.
 
   //Convert to GT format and pack to encodedTau of PFTau
   l1gt::Tau l1gtTau = l1ctTau.toGT();
   l1gt::PackedTau packed_Tau = l1gtTau.pack();
 
   //Make PFTau
   //Save pt, eta and phi in gt scales
   math::PtEtaPhiMLorentzVector tempP4(l1gt::Scales::floatPt(l1gtTau.v3.pt),
                                       l1gt::Scales::floatEta(l1gtTau.v3.eta),
                                       l1gt::Scales::floatPhi(l1gtTau.v3.phi),
                                       float(mass));
 
   l1t::PFTau l1PFTau(tempP4, pNNVec, NN_ouput.nn_id, 0, lId);
   l1PFTau.setZ0(float(z0) * 0.05);    //L1TauEmu::z0_base);
   l1PFTau.setDxy(float(dxy) * 0.05);  //L1TauEmu::dxy_base);
 
   l1PFTau.set_encodedTau(packed_Tau);
 
   iTaus->push_back(l1PFTau);
 }

◆ process_HW()

void L1NNTauProducer::process_HW	(	const l1t::PFCandidateCollection &	parts,
		std::unique_ptr< l1t::PFTauCollection > &	iTaus
	)

private

Definition at line 288 of file L1NNTauProducer.cc.

References funct::abs(), l1t::PFCandidate::ChargedHadron, l1t::PFCandidate::Electron, l1HPSPFTauEmu::etaphi_base, fConeSize_, fEMSeed, fMaxTaus_, mps_fire::i, L1TauEmu::inCone(), dqmiolumiharvest::j, makeTau_HW(), l1t::PFCandidate::Photon, fileCollector::seed, HLT_2024v14_cff::seeds, jetUpdater_cfi::sort, track_trigger_eta_max, HcalDetIdTransform::transform(), parallelization::uint, and edmMakeDummyCfis::work.

Referenced by produce().

                                                                            {
   // The fixed point algorithm emulation
   using namespace L1TauEmu;
   std::vector<l1t::PFCandidate> work;
   work.resize(parts.size());
   std::transform(parts.begin(), parts.end(), work.begin(), [](const l1t::PFCandidate& part) { return part; });
   std::sort(work.begin(), work.end(), [](l1t::PFCandidate i, l1t::PFCandidate j) {
     return (l1ct::pt_t(i.pt()) > l1ct::pt_t(j.pt()));
   });
 
   std::vector<l1t::PFCandidate> seeds;
   uint lSeed = l1t::PFCandidate::ChargedHadron;
   if (fEMSeed)
     lSeed = l1t::PFCandidate::Photon;
   std::copy_if(work.begin(), work.end(), std::back_inserter(seeds), [&](const l1t::PFCandidate& part) {
     return ((part.id() == l1t::PFCandidate::Electron || part.id() == l1t::PFCandidate::ChargedHadron ||
              part.id() == lSeed) &&
             std::abs(part.eta()) < track_trigger_eta_max);
   });
   // It would be nice to transform the inputs to the etaphi_base of the FW here, as in the line below
   // However the phi may wrap around if the etaphi_base > 1, so don't do it...
   //std::for_each(work.begin(), work.end(), [](l1t::PFCandidate& x){x.setP4(math::PtEtaPhiMLorentzVector(pt_t(x.pt()), etaphi_t(x.eta()*etaphi_base), etaphi_t(x.phi()*etaphi_base), x.mass()));});
   detaphi_t rCone2 = detaphi_t(fConeSize_ * fConeSize_ * etaphi_base * etaphi_base);
 
   iTaus->reserve(fMaxTaus_);
   while (!seeds.empty() && iTaus->size() < unsigned(fMaxTaus_)) {
     // Take the first (highest pt) candidate as a seed
     l1t::PFCandidate seed = seeds.at(0);
     // Get the particles within a _coneSize of the seed
     std::vector<l1t::PFCandidate> particlesInCone;
     std::copy_if(work.begin(), work.end(), std::back_inserter(particlesInCone), [&](l1t::PFCandidate& part) {
       return inCone(seed, part, rCone2);
     });
     makeTau_HW(seed, particlesInCone, iTaus);
     // remove the clustered particles
     work.erase(std::remove_if(
                    work.begin(), work.end(), [&](const l1t::PFCandidate& part) { return inCone(seed, part, rCone2); }),
                work.end());
 
     seeds.erase(
         std::remove_if(
             seeds.begin(), seeds.end(), [&](const l1t::PFCandidate& part) { return inCone(seed, part, rCone2); }),
         seeds.end());
   }
 }

◆ process_SW()

void L1NNTauProducer::process_SW	(	const l1t::PFCandidateCollection &	parts,
		std::unique_ptr< l1t::PFTauCollection > &	iTaus
	)

private

Definition at line 102 of file L1NNTauProducer.cc.

References funct::abs(), addTau(), l1t::PFCandidate::ChargedHadron, reco::deltaR2(), l1t::PFCandidate::Electron, fConeSize_, fMaxTaus_, mps_fire::i, testProducerWithPsetDescEmpty_cfi::i1, dqmiolumiharvest::j, eostools::move(), boostedElectronIsolation_cff::pfChargedHadrons, jetUpdater_cfi::sort, and track_trigger_eta_max.

Referenced by produce().

                                                                            {
   std::vector<unique_ptr<l1t::PFCandidate>> pfChargedHadrons;
   std::vector<unique_ptr<l1t::PFCandidate>> pfChargedHadrons_sort_v;
   std::vector<unique_ptr<l1t::PFCandidate>> pfChargedHadrons_seeds_v;
   for (const auto& l1PFCand : parts)
     if ((l1PFCand.id() == l1t::PFCandidate::ChargedHadron || l1PFCand.id() == l1t::PFCandidate::Electron) &&
         std::abs(l1PFCand.eta()) < track_trigger_eta_max)
       pfChargedHadrons_sort_v.push_back(std::make_unique<l1t::PFCandidate>(l1PFCand));
 
   if (pfChargedHadrons_sort_v.empty())
     return;
   std::sort(
       pfChargedHadrons_sort_v.begin(),
       pfChargedHadrons_sort_v.end(),
       [](std::unique_ptr<l1t::PFCandidate>& i, std::unique_ptr<l1t::PFCandidate>& j) { return (i->pt() > j->pt()); });
 
   pfChargedHadrons_seeds_v.push_back(std::move(pfChargedHadrons_sort_v[0]));
   for (unsigned int i0 = 1; i0 < pfChargedHadrons_sort_v.size(); i0++) {
     bool pMatch = false;
     for (unsigned int i1 = 0; i1 < pfChargedHadrons_seeds_v.size(); i1++) {
       if (reco::deltaR2(*(pfChargedHadrons_seeds_v[i1]), *(pfChargedHadrons_sort_v[i0])) < fConeSize_ * fConeSize_)
         pMatch = true;
     }
     if (pMatch)
       continue;
     pfChargedHadrons_seeds_v.push_back(std::move(pfChargedHadrons_sort_v[i0]));
     if (int(pfChargedHadrons_seeds_v.size()) > fMaxTaus_ - 1)
       break;
   }
   for (unsigned int i0 = 0; i0 < pfChargedHadrons_seeds_v.size(); i0++) {
     addTau(*(pfChargedHadrons_seeds_v[i0]), parts, iTaus);
   }
 }

◆ produce()

void L1NNTauProducer::produce	(	edm::Event &	iEvent,
		const edm::EventSetup &	iSetup
	)

overrideprivate

Definition at line 88 of file L1NNTauProducer.cc.

References fHW, fL1PFToken_, mps_fire::i, iEvent, dqmiolumiharvest::j, eostools::move(), process_HW(), process_SW(), and jetUpdater_cfi::sort.

                                                                            {
   edm::Handle<l1t::PFCandidateCollection> l1PFCandidates;
   iEvent.getByToken(fL1PFToken_, l1PFCandidates);
   auto lTaus = std::make_unique<l1t::PFTauCollection>();
 
   if (fHW) {
     process_HW(*l1PFCandidates, lTaus);
   } else {
     process_SW(*l1PFCandidates, lTaus);
   }
 
   std::sort(lTaus->begin(), lTaus->end(), [](l1t::PFTau i, l1t::PFTau j) { return (i.pt() > j.pt()); });
   iEvent.put(std::move(lTaus), "L1PFTausNN");
 }

Member Data Documentation

◆ fConeSize_

double L1NNTauProducer::fConeSize_

private

Definition at line 48 of file L1NNTauProducer.cc.

Referenced by addTau(), process_HW(), and process_SW().

◆ fDebug

const bool L1NNTauProducer::fDebug

private

Definition at line 54 of file L1NNTauProducer.cc.

◆ fEMSeed

const bool L1NNTauProducer::fEMSeed

private

Definition at line 53 of file L1NNTauProducer.cc.

Referenced by process_HW().

◆ fHW

const bool L1NNTauProducer::fHW

private

Definition at line 52 of file L1NNTauProducer.cc.

Referenced by L1NNTauProducer(), and produce().

◆ fL1PFToken_

edm::EDGetTokenT<vector<l1t::PFCandidate> > L1NNTauProducer::fL1PFToken_

private

Definition at line 55 of file L1NNTauProducer.cc.

Referenced by produce().

◆ fMaxTaus_

int L1NNTauProducer::fMaxTaus_

private

Definition at line 50 of file L1NNTauProducer.cc.

Referenced by process_HW(), and process_SW().

◆ fNParticles_

int L1NNTauProducer::fNParticles_

private

Definition at line 51 of file L1NNTauProducer.cc.

Referenced by L1NNTauProducer().

◆ fSeedPt_

double L1NNTauProducer::fSeedPt_

private

Definition at line 47 of file L1NNTauProducer.cc.

Referenced by addTau(), and makeTau_HW().

◆ fTauNNId_

std::unique_ptr<TauNNId> L1NNTauProducer::fTauNNId_

private

Definition at line 29 of file L1NNTauProducer.cc.

Referenced by addTau(), and L1NNTauProducer().

◆ fTauNNIdHW_

std::unique_ptr<TauNNIdHW> L1NNTauProducer::fTauNNIdHW_

private

Definition at line 34 of file L1NNTauProducer.cc.

Referenced by L1NNTauProducer(), and makeTau_HW().

◆ fTauSize_

double L1NNTauProducer::fTauSize_

private

Definition at line 49 of file L1NNTauProducer.cc.

Referenced by addTau(), and makeTau_HW().

Public Member Functions

Static Public Member Functions

Private Member Functions

Private Attributes

Additional Inherited Members

Detailed Description

Constructor & Destructor Documentation

◆ L1NNTauProducer()

◆ ~L1NNTauProducer()

Member Function Documentation

◆ addTau()

◆ fillDescriptions()

◆ globalEndJob()

◆ initializeGlobalCache()

◆ makeTau_HW()

◆ process_HW()

◆ process_SW()

◆ produce()

Member Data Documentation

◆ fConeSize_

◆ fDebug

◆ fEMSeed

◆ fHW

◆ fL1PFToken_

◆ fMaxTaus_

◆ fNParticles_

◆ fSeedPt_

◆ fTauNNId_

◆ fTauNNIdHW_

◆ fTauSize_