reconstruction/tracking_efficiency/TrackingEfficiency_processor.cc

0001 #include "TrackingEfficiency_processor.h"
0002
0003 #include <Acts/Definitions/TrackParametrization.hpp>
0004 #include <Acts/EventData/MultiTrajectoryHelpers.hpp>
0005 #include <ActsExamples/EventData/Trajectories.hpp>
0006 #include <JANA/JApplication.h>
0007 #include <JANA/JApplicationFwd.h>
0008 #include <JANA/JEvent.h>
0009 #include <JANA/Services/JGlobalRootLock.h>
0010 #include <Math/GenVector/Cartesian3D.h>
0011 #include <Math/GenVector/PxPyPzM4D.h>
0012 #include <Rtypes.h>
0013 #include <edm4eic/ReconstructedParticleCollection.h>
0014 #include <edm4hep/MCParticleCollection.h>
0015 #include <edm4hep/Vector3d.h>
0016 #include <edm4hep/Vector3f.h>
0017 #include <fmt/core.h>
0018 #include <spdlog/logger.h>
0019 #include <Eigen/Core>
0020 #include <cmath>
0021 #include <cstddef>
0022 #include <iterator>
0023 #include <map>
0024 #include <optional>
0025 #include <string>
0026 #include <vector>
0027
0028 #include "services/log/Log_service.h"
0029 #include "services/rootfile/RootFile_service.h"
0030
0031 //------------------
0032 // Init
0033 //------------------
0034 void TrackingEfficiency_processor::Init() {
0035   std::string plugin_name = ("tracking_efficiency");
0036
0037   // Get JANA application
0038   auto* app = GetApplication();
0039
0040   // Ask service locator a file to write histograms to
0041   auto root_file_service = app->GetService<RootFile_service>();
0042
0043   // Get TDirectory for histograms root file
0044   auto globalRootLock = app->GetService<JGlobalRootLock>();
0045   globalRootLock->acquire_write_lock();
0046   auto* file = root_file_service->GetHistFile();
0047   globalRootLock->release_lock();
0048
0049   // Create a directory for this plugin. And subdirectories for series of histograms
0050   m_dir_main = file->mkdir(plugin_name.c_str());
0051
0052   // Get logger
0053   m_log = app->GetService<Log_service>()->logger(plugin_name);
0054 }
0055
0056 //------------------
0057 // Process
0058 //------------------
0059 void TrackingEfficiency_processor::Process(const std::shared_ptr<const JEvent>& event) {
0060   using namespace ROOT;
0061
0062   // EXAMPLE I
0063   // This is access to for final result of the calculation/data transformation of central detector CFKTracking:
0064   const auto reco_particles =
0065       event->Get<edm4eic::ReconstructedParticle>("ReconstructedChargedParticles");
0066
0067   m_log->debug("Tracking reconstructed particles N={}: ", reco_particles.size());
0068   m_log->debug("   {:<5} {:>8} {:>8} {:>8} {:>8} {:>8}", "[i]", "[px]", "[py]", "[pz]", "[P]",
0069                "[P*3]");
0070
0071   for (std::size_t i = 0; i < reco_particles.size(); i++) {
0072     const auto& particle = *(reco_particles[i]);
0073
0074     double px = particle.getMomentum().x;
0075     double py = particle.getMomentum().y;
0076     double pz = particle.getMomentum().z;
0077     // ROOT::Math::PxPyPzM4D p4v(px, py, pz, particle.getMass());
0078     ROOT::Math::Cartesian3D p(px, py, pz);
0079     m_log->debug("   {:<5} {:>8.2f} {:>8.2f} {:>8.2f} {:>8.2f} {:>8.2f}", i, px, py, pz, p.R(),
0080                  p.R() * 3);
0081   }
0082
0083   // EXAMPLE II
0084   // This gets access to more direct ACTS results from CFKTracking
0085   auto acts_results = event->Get<ActsExamples::Trajectories>("CentralCKFActsTrajectories");
0086   m_log->debug("ACTS Trajectories( size: {} )", std::size(acts_results));
0087   m_log->debug("{:>10} {:>10}  {:>10} {:>10} {:>10} {:>10} {:>12} {:>12} {:>12} {:>8}", "[loc 0]",
0088                "[loc 1]", "[phi]", "[theta]", "[q/p]", "[p]", "[err phi]", "[err th]", "[err q/p]",
0089                "[chi2]");
0090
0091   // Loop over the trajectories
0092   for (const auto& traj : acts_results) {
0093
0094     // Get the entry index for the single trajectory
0095     // The trajectory entry indices and the multiTrajectory
0096     const auto& mj        = traj->multiTrajectory();
0097     const auto& trackTips = traj->tips();
0098     if (trackTips.empty()) {
0099       m_log->debug("Empty multiTrajectory.");
0100       continue;
0101     }
0102
0103     const auto& trackTip = trackTips.front();
0104
0105     // Collect the trajectory summary info
0106     auto trajState = Acts::MultiTrajectoryHelpers::trajectoryState(mj, trackTip);
0107     if (traj->hasTrackParameters(trackTip)) {
0108       const auto& boundParam = traj->trackParameters(trackTip);
0109       const auto& parameter  = boundParam.parameters();
0110       const auto& covariance = *boundParam.covariance();
0111       m_log->debug("{:>10.2f} {:>10.2f}  {:>10.2f} {:>10.3f} {:>10.4f} {:>10.3f} {:>12.4e} "
0112                    "{:>12.4e} {:>12.4e} {:>8.2f}",
0113                    parameter[Acts::eBoundLoc0], parameter[Acts::eBoundLoc1],
0114                    parameter[Acts::eBoundPhi], parameter[Acts::eBoundTheta],
0115                    parameter[Acts::eBoundQOverP], 1.0 / parameter[Acts::eBoundQOverP],
0116                    sqrt(covariance(Acts::eBoundPhi, Acts::eBoundPhi)),
0117                    sqrt(covariance(Acts::eBoundTheta, Acts::eBoundTheta)),
0118                    sqrt(covariance(Acts::eBoundQOverP, Acts::eBoundQOverP)), trajState.chi2Sum);
0119     }
0120   }
0121
0122   // EXAMPLE III
0123   // Loop over MC particles
0124   auto mc_particles = event->Get<edm4hep::MCParticle>("MCParticles");
0125   m_log->debug("MC particles N={}: ", mc_particles.size());
0126   m_log->debug("   {:<5} {:<6} {:<7} {:>8} {:>8} {:>8} {:>8}", "[i]", "status", "[PDG]", "[px]",
0127                "[py]", "[pz]", "[P]");
0128   for (std::size_t i = 0; i < mc_particles.size(); i++) {
0129     const auto* particle = mc_particles[i];
0130
0131     // GeneratorStatus() == 1 - stable particles from MC generator. 0 - might be added by Geant4
0132     if (particle->getGeneratorStatus() != 1) {
0133       continue;
0134     }
0135
0136     double px = particle->getMomentum().x;
0137     double py = particle->getMomentum().y;
0138     double pz = particle->getMomentum().z;
0139     ROOT::Math::PxPyPzM4D p4v(px, py, pz, particle->getMass());
0140     ROOT::Math::Cartesian3D p(px, py, pz);
0141     if (p.R() < 1) {
0142       continue;
0143     }
0144
0145     m_log->debug("   {:<5} {:<6} {:<7} {:>8.2f} {:>8.2f} {:>8.2f} {:>8.2f}", i,
0146                  particle->getGeneratorStatus(), particle->getPDG(), px, py, pz, p.R());
0147   }
0148 }
0149
0150 //------------------
0151 // Finish
0152 //------------------
0153 void TrackingEfficiency_processor::Finish() {
0154   fmt::print("OccupancyAnalysis::Finish() called\n");
0155
0156   // Next we want to create several pretty canvases (with histograms drawn on "same")
0157   // But we don't want those canvases to pop up. So we set root to batch mode
0158   // We will restore the mode afterwards
0159   //bool save_is_batch = gROOT->IsBatch();
0160   //gROOT->SetBatch(true);
0161
0162   // 3D hits distribution
0163   //      auto th3_by_det_canvas = new TCanvas("th3_by_det_cnv", "Occupancy of detectors");
0164   //      dir_main->Append(th3_by_det_canvas);
0165   //      for (auto& kv : th3_by_detector->GetMap()) {
0166   //              auto th3_hist = kv.second;
0167   //              th3_hist->Draw("same");
0168   //      }
0169   //      th3_by_det_canvas->GetPad(0)->BuildLegend();
0170   //
0171   //      // Hits Z by detector
0172   //
0173   //      // Create pretty canvases
0174   //      auto z_by_det_canvas = new TCanvas("z_by_det_cnv", "Hit Z distribution by detector");
0175   //      dir_main->Append(z_by_det_canvas);
0176   //      th1_hits_z->Draw("PLC PFC");
0177   //
0178   //      for (auto& kv : th1_z_by_detector->GetMap()) {
0179   //              auto hist = kv.second;
0180   //              hist->Draw("SAME PLC PFC");
0181   //              hist->SetFillStyle(3001);
0182   //      }
0183   //      z_by_det_canvas->GetPad(0)->BuildLegend();
0184   //
0185   //      gROOT->SetBatch(save_is_batch);
0186 }