CMS 3D CMS Logo
DPFIsolation.cc
Go to the documentation of this file.
1 /*
2  * \class DPFIsolation
3  *
4  * Deep ParticleFlow tau isolation using Deep NN.
5  *
6  * \author Owen Colegrove, UCSB
7  */
8 
9 #include "RecoTauTag/RecoTau/interface/DeepTauBase.h"
10 
11 namespace {
12 inline int getPFCandidateIndex(const edm::Handle<pat::PackedCandidateCollection>& pfcands,
13  const reco::CandidatePtr& cptr)
14 {
15  for(unsigned int i = 0; i < pfcands->size(); ++i) {
16  if(reco::CandidatePtr(pfcands,i) == cptr)
17  return i;
18  }
19  return -1;
20 }
21 } // anonymous namespace
22 
23 class DPFIsolation : public deep_tau::DeepTauBase {
24 public:
25  static const OutputCollection& GetOutputs()
26  {
27  const size_t tau_index = 0;
28  static const OutputCollection outputs_ = { { "VSall", Output({tau_index}, {}) } };
29  return outputs_;
30  };
31 
32  static unsigned getNumberOfParticles(unsigned graphVersion)
33  {
34  static const std::map<unsigned, unsigned> nparticles { { 0, 60 }, { 1, 36 } };
35  return nparticles.at(graphVersion);
36  }
37 
38  static unsigned GetNumberOfFeatures(unsigned graphVersion)
39  {
40  static const std::map<unsigned, unsigned> nfeatures { { 0, 47 }, { 1, 51 } };
41  return nfeatures.at(graphVersion);
42  }
43 
44  static void fillDescriptions(edm::ConfigurationDescriptions& descriptions)
45  {
46  edm::ParameterSetDescription desc;
47  desc.add<edm::InputTag>("pfcands", edm::InputTag("packedPFCandidates"));
48  desc.add<edm::InputTag>("taus", edm::InputTag("slimmedTaus"));
49  desc.add<edm::InputTag>("vertices", edm::InputTag("offlineSlimmedPrimaryVertices"));
50  desc.add<std::string>("graph_file", "RecoTauTag/TrainingFiles/data/DPFTauId/DPFIsolation_2017v0_quantized.pb");
51  desc.add<unsigned>("version", 0);
52  desc.add<bool>("mem_mapped", false);
53 
54  edm::ParameterSetDescription descWP;
55  descWP.add<std::string>("VVVLoose", "0");
56  descWP.add<std::string>("VVLoose", "0");
57  descWP.add<std::string>("VLoose", "0");
58  descWP.add<std::string>("Loose", "0");
59  descWP.add<std::string>("Medium", "0");
60  descWP.add<std::string>("Tight", "0");
61  descWP.add<std::string>("VTight", "0");
62  descWP.add<std::string>("VVTight", "0");
63  descWP.add<std::string>("VVVTight", "0");
64  desc.add<edm::ParameterSetDescription>("VSallWP", descWP);
65  descriptions.add("DPFTau2016v0", desc);
66  }
67 
68  explicit DPFIsolation(const edm::ParameterSet& cfg,const deep_tau::DeepTauCache* cache) :
69  DeepTauBase(cfg, GetOutputs(), cache),
70  pfcand_token(consumes<pat::PackedCandidateCollection>(cfg.getParameter<edm::InputTag>("pfcands"))),
71  vtx_token(consumes<reco::VertexCollection>(cfg.getParameter<edm::InputTag>("vertices"))),
72  graphVersion(cfg.getParameter<unsigned>("version"))
73  {
74  const auto& shape = cache_->getGraph().node(0).attr().at("shape").shape();
75 
76  if(!(graphVersion == 1 || graphVersion == 0 ))
77  throw cms::Exception("DPFIsolation") << "unknown version of the graph_ file.";
78 
79  if(!(shape.dim(1).size() == getNumberOfParticles(graphVersion) && shape.dim(2).size() == GetNumberOfFeatures(graphVersion)))
80  throw cms::Exception("DPFIsolation") << "number of inputs does not match the expected inputs for the given version";
81 
82  }
83 
84 private:
85  tensorflow::Tensor getPredictions(edm::Event& event, const edm::EventSetup& es,
86  edm::Handle<TauCollection> taus) override
87  {
88  edm::Handle<pat::PackedCandidateCollection> pfcands;
89  event.getByToken(pfcand_token, pfcands);
90 
91  edm::Handle<reco::VertexCollection> vertices;
92  event.getByToken(vtx_token, vertices);
93 
94  tensorflow::Tensor tensor(tensorflow::DT_FLOAT, {1,
95  static_cast<int>(getNumberOfParticles(graphVersion)), static_cast<int>(GetNumberOfFeatures(graphVersion))});
96 
97  tensorflow::Tensor predictions(tensorflow::DT_FLOAT, { static_cast<int>(taus->size()), 1});
98 
99  std::vector<tensorflow::Tensor> outputs_;
100 
101  float pfCandPt, pfCandPz, pfCandPtRel, pfCandPzRel, pfCandDr, pfCandDEta, pfCandDPhi, pfCandEta, pfCandDz,
102  pfCandDzErr, pfCandD0, pfCandD0D0, pfCandD0Dz, pfCandD0Dphi, pfCandPuppiWeight,
103  pfCandPixHits, pfCandHits, pfCandLostInnerHits, pfCandPdgID, pfCandCharge, pfCandFromPV,
104  pfCandVtxQuality, pfCandHighPurityTrk, pfCandTauIndMatch, pfCandDzSig, pfCandD0Sig, pfCandD0Err,
105  pfCandPtRelPtRel, pfCandDzDz, pfCandDVx_1, pfCandDVy_1, pfCandDVz_1, pfCandD_1;
106  float pvx = !vertices->empty() ? (*vertices)[0].x() : -1;
107  float pvy = !vertices->empty() ? (*vertices)[0].y() : -1;
108  float pvz = !vertices->empty() ? (*vertices)[0].z() : -1;
109 
110  bool pfCandIsBarrel;
111 
112  // These variables define ranges further used for standardization
113  static constexpr float pfCandPt_max = 500.f;
114  static constexpr float pfCandPz_max = 1000.f;
115  static constexpr float pfCandPtRel_max = 1.f;
116  static constexpr float pfCandPzRel_max = 100.f;
117  static constexpr float pfCandPtRelPtRel_max = 1.f;
118  static constexpr float pfCandD0_max = 5.f;
119  static constexpr float pfCandDz_max = 5.f;
120  static constexpr float pfCandDVx_y_z_1_max = 0.05f;
121  static constexpr float pfCandD_1_max = 0.1f;
122  static constexpr float pfCandD0_z_Err_max = 1.f;
123  static constexpr float pfCandDzSig_max = 3.f;
124  static constexpr float pfCandD0Sig_max = 1.f;
125  static constexpr float pfCandDr_max = 0.5f;
126  static constexpr float pfCandEta_max = 2.75f;
127  static constexpr float pfCandDEta_max = 0.5f;
128  static constexpr float pfCandDPhi_max = 0.5f;
129  static constexpr float pfCandPixHits_max = 7.f;
130  static constexpr float pfCandHits_max = 30.f;
131 
132  for(size_t tau_index = 0; tau_index < taus->size(); tau_index++) {
133  pat::Tau tau = taus->at(tau_index);
134  bool isGoodTau = false;
135  const float lepRecoPt = tau.pt();
136  const float lepRecoPz = std::abs(tau.pz());
137  const float lepRecoEta = tau.eta();
138  const float lepRecoPhi = tau.phi();
139 
140  if(lepRecoPt >= 30 && std::abs(lepRecoEta) < 2.3 && tau.isTauIDAvailable("againstMuonLoose3") &&
141  tau.isTauIDAvailable("againstElectronVLooseMVA6")) {
142  isGoodTau = (tau.tauID("againstElectronVLooseMVA6") && tau.tauID("againstMuonLoose3") );
143  }
144 
145  if (!isGoodTau) {
146  predictions.matrix<float>()(tau_index, 0) = -1;
147  continue;
148  }
149 
150  std::vector<unsigned int> signalCandidateInds;
151 
152  for(const auto c : tau.signalCands())
153  signalCandidateInds.push_back(getPFCandidateIndex(pfcands,c));
154 
155  // Use of setZero results in warnings in eigen library during compilation.
156  //tensor.flat<float>().setZero();
157  const unsigned n_inputs = getNumberOfParticles(graphVersion) * GetNumberOfFeatures(graphVersion);
158  for(unsigned input_idx = 0; input_idx < n_inputs; ++input_idx)
159  tensor.flat<float>()(input_idx) = 0;
160 
161  unsigned int iPF = 0;
162  const unsigned max_iPF = getNumberOfParticles(graphVersion);
163 
164  std::vector<unsigned int> sorted_inds(pfcands->size());
165  std::size_t n = 0;
166  std::generate(std::begin(sorted_inds), std::end(sorted_inds), [&]{ return n++; });
167 
168  std::sort(std::begin(sorted_inds), std::end(sorted_inds),
169  [&](int i1, int i2) { return pfcands->at(i1).pt() > pfcands->at(i2).pt(); } );
170 
171  for(size_t pf_index = 0; pf_index < pfcands->size() && iPF < max_iPF; pf_index++) {
172  pat::PackedCandidate p = pfcands->at(sorted_inds.at(pf_index));
173  float deltaR_tau_p = deltaR(p.p4(),tau.p4());
174 
175  if (p.pt() < 0.5) continue;
176  if (p.fromPV() < 0) continue;
177  if (deltaR_tau_p > 0.5) continue;
178  if (p.fromPV() < 1 && p.charge() != 0) continue;
179  pfCandPt = p.pt();
180  pfCandPtRel = p.pt()/lepRecoPt;
181 
182  pfCandDr = deltaR_tau_p;
183  pfCandDEta = std::abs(lepRecoEta - p.eta());
184  pfCandDPhi = std::abs(deltaPhi(lepRecoPhi, p.phi()));
185  pfCandEta = p.eta();
186  pfCandIsBarrel = (std::abs(pfCandEta) < 1.4);
187  pfCandPz = std::abs(std::sinh(pfCandEta)*pfCandPt);
188  pfCandPzRel = pfCandPz/lepRecoPz;
189  pfCandPdgID = std::abs(p.pdgId());
190  pfCandCharge = p.charge();
191  pfCandDVx_1 = p.vx() - pvx;
192  pfCandDVy_1 = p.vy() - pvy;
193  pfCandDVz_1 = p.vz() - pvz;
194 
195  pfCandD_1 = std::sqrt(pfCandDVx_1*pfCandDVx_1 + pfCandDVy_1*pfCandDVy_1 + pfCandDVz_1*pfCandDVz_1);
196 
197  if (pfCandCharge != 0 and p.hasTrackDetails()){
198  pfCandDz = p.dz();
199  pfCandDzErr = p.dzError();
200  pfCandDzSig = (std::abs(p.dz()) + 0.000001)/(p.dzError() + 0.00001);
201  pfCandD0 = p.dxy();
202  pfCandD0Err = p.dxyError();
203  pfCandD0Sig = (std::abs(p.dxy()) + 0.000001)/ (p.dxyError() + 0.00001);
204  pfCandPixHits = p.numberOfPixelHits();
205  pfCandHits = p.numberOfHits();
206  pfCandLostInnerHits = p.lostInnerHits();
207  } else {
208  float disp = 1;
209  int psudorand = p.pt()*1000000;
210  if (psudorand%2 == 0) disp = -1;
211  pfCandDz = 5*disp;
212  pfCandDzErr = 0;
213  pfCandDzSig = 0;
214  pfCandD0 = 5*disp;
215  pfCandD0Err = 0;
216  pfCandD0Sig = 0;
217  pfCandPixHits = 0;
218  pfCandHits = 0;
219  pfCandLostInnerHits = 2.;
220  pfCandDVx_1 = 1;
221  pfCandDVy_1 = 1;
222  pfCandDVz_1 = 1;
223  pfCandD_1 = 1;
224  }
225 
226  pfCandPuppiWeight = p.puppiWeight();
227  pfCandFromPV = p.fromPV();
228  pfCandVtxQuality = p.pvAssociationQuality();
229  pfCandHighPurityTrk = p.trackHighPurity();
230  float pfCandTauIndMatch_temp = 0;
231 
232  for (auto i : signalCandidateInds) {
233  if (i == sorted_inds.at(pf_index)) pfCandTauIndMatch_temp = 1;
234  }
235 
236  pfCandTauIndMatch = pfCandTauIndMatch_temp;
237  pfCandPtRelPtRel = pfCandPtRel*pfCandPtRel;
238  pfCandPt = std::min(pfCandPt, pfCandPt_max);
239  pfCandPt = pfCandPt/pfCandPt_max;
240 
241  pfCandPz = std::min(pfCandPz, pfCandPz_max);
242  pfCandPz = pfCandPz/pfCandPz_max;
243 
244  pfCandPtRel = std::min(pfCandPtRel, pfCandPtRel_max);
245  pfCandPzRel = std::min(pfCandPzRel, pfCandPzRel_max);
246  pfCandPzRel = pfCandPzRel/pfCandPzRel_max;
247  pfCandDr = pfCandDr/pfCandDr_max;
248  pfCandEta = pfCandEta/pfCandEta_max;
249  pfCandDEta = pfCandDEta/pfCandDEta_max;
250  pfCandDPhi = pfCandDPhi/pfCandDPhi_max;
251  pfCandPixHits = pfCandPixHits/pfCandPixHits_max;
252  pfCandHits = pfCandHits/pfCandHits_max;
253 
254  pfCandPtRelPtRel = std::min(pfCandPtRelPtRel, pfCandPtRelPtRel_max);
255 
256  pfCandD0 = std::clamp(pfCandD0, -pfCandD0_max, pfCandD0_max);
257  pfCandD0 = pfCandD0/pfCandD0_max;
258 
259  pfCandDz = std::clamp(pfCandDz, -pfCandDz_max, pfCandDz_max);
260  pfCandDz = pfCandDz/pfCandDz_max;
261 
262  pfCandD0Err = std::min(pfCandD0Err, pfCandD0_z_Err_max);
263  pfCandDzErr = std::min(pfCandDzErr, pfCandD0_z_Err_max);
264  pfCandDzSig = std::min(pfCandDzSig, pfCandDzSig_max);
265  pfCandDzSig = pfCandDzSig/pfCandDzSig_max;
266 
267  pfCandD0Sig = std::min(pfCandD0Sig, pfCandD0Sig_max);
268  pfCandD0D0 = pfCandD0*pfCandD0;
269  pfCandDzDz = pfCandDz*pfCandDz;
270  pfCandD0Dz = pfCandD0*pfCandDz;
271  pfCandD0Dphi = pfCandD0*pfCandDPhi;
272 
273  pfCandDVx_1 = std::clamp(pfCandDVx_1, -pfCandDVx_y_z_1_max, pfCandDVx_y_z_1_max);
274  pfCandDVx_1 = pfCandDVx_1/pfCandDVx_y_z_1_max;
275 
276  pfCandDVy_1 = std::clamp(pfCandDVy_1, -pfCandDVx_y_z_1_max, pfCandDVx_y_z_1_max);
277  pfCandDVy_1 = pfCandDVy_1/pfCandDVx_y_z_1_max;
278 
279  pfCandDVz_1 = std::clamp(pfCandDVz_1, -pfCandDVx_y_z_1_max, pfCandDVx_y_z_1_max);
280  pfCandDVz_1 = pfCandDVz_1/pfCandDVx_y_z_1_max;
281 
282  pfCandD_1 = std::clamp(pfCandD_1, -pfCandD_1_max, pfCandD_1_max);
283  pfCandD_1 = pfCandD_1/ pfCandD_1_max;
284 
285  if (graphVersion == 0) {
286  tensor.tensor<float,3>()( 0, 60-1-iPF, 0) = pfCandPt;
287  tensor.tensor<float,3>()( 0, 60-1-iPF, 1) = pfCandPz;
288  tensor.tensor<float,3>()( 0, 60-1-iPF, 2) = pfCandPtRel;
289  tensor.tensor<float,3>()( 0, 60-1-iPF, 3) = pfCandPzRel;
290  tensor.tensor<float,3>()( 0, 60-1-iPF, 4) = pfCandDr;
291  tensor.tensor<float,3>()( 0, 60-1-iPF, 5) = pfCandDEta;
292  tensor.tensor<float,3>()( 0, 60-1-iPF, 6) = pfCandDPhi;
293  tensor.tensor<float,3>()( 0, 60-1-iPF, 7) = pfCandEta;
294  tensor.tensor<float,3>()( 0, 60-1-iPF, 8) = pfCandDz;
295  tensor.tensor<float,3>()( 0, 60-1-iPF, 9) = pfCandDzSig;
296  tensor.tensor<float,3>()( 0, 60-1-iPF, 10) = pfCandD0;
297  tensor.tensor<float,3>()( 0, 60-1-iPF, 11) = pfCandD0Sig;
298  tensor.tensor<float,3>()( 0, 60-1-iPF, 12) = pfCandDzErr;
299  tensor.tensor<float,3>()( 0, 60-1-iPF, 13) = pfCandD0Err;
300  tensor.tensor<float,3>()( 0, 60-1-iPF, 14) = pfCandD0D0;
301  tensor.tensor<float,3>()( 0, 60-1-iPF, 15) = pfCandCharge==0;
302  tensor.tensor<float,3>()( 0, 60-1-iPF, 16) = pfCandCharge==1;
303  tensor.tensor<float,3>()( 0, 60-1-iPF, 17) = pfCandCharge==-1;
304  tensor.tensor<float,3>()( 0, 60-1-iPF, 18) = pfCandPdgID>22;
305  tensor.tensor<float,3>()( 0, 60-1-iPF, 19) = pfCandPdgID==22;
306  tensor.tensor<float,3>()( 0, 60-1-iPF, 20) = pfCandDzDz;
307  tensor.tensor<float,3>()( 0, 60-1-iPF, 21) = pfCandD0Dz;
308  tensor.tensor<float,3>()( 0, 60-1-iPF, 22) = pfCandD0Dphi;
309  tensor.tensor<float,3>()( 0, 60-1-iPF, 23) = pfCandPtRelPtRel;
310  tensor.tensor<float,3>()( 0, 60-1-iPF, 24) = pfCandPixHits;
311  tensor.tensor<float,3>()( 0, 60-1-iPF, 25) = pfCandHits;
312  tensor.tensor<float,3>()( 0, 60-1-iPF, 26) = pfCandLostInnerHits==-1;
313  tensor.tensor<float,3>()( 0, 60-1-iPF, 27) = pfCandLostInnerHits==0;
314  tensor.tensor<float,3>()( 0, 60-1-iPF, 28) = pfCandLostInnerHits==1;
315  tensor.tensor<float,3>()( 0, 60-1-iPF, 29) = pfCandLostInnerHits==2;
316  tensor.tensor<float,3>()( 0, 60-1-iPF, 30) = pfCandPuppiWeight;
317  tensor.tensor<float,3>()( 0, 60-1-iPF, 31) = (pfCandVtxQuality == 1);
318  tensor.tensor<float,3>()( 0, 60-1-iPF, 32) = (pfCandVtxQuality == 5);
319  tensor.tensor<float,3>()( 0, 60-1-iPF, 33) = (pfCandVtxQuality == 6);
320  tensor.tensor<float,3>()( 0, 60-1-iPF, 34) = (pfCandVtxQuality == 7);
321  tensor.tensor<float,3>()( 0, 60-1-iPF, 35) = (pfCandFromPV == 1);
322  tensor.tensor<float,3>()( 0, 60-1-iPF, 36) = (pfCandFromPV == 2);
323  tensor.tensor<float,3>()( 0, 60-1-iPF, 37) = (pfCandFromPV == 3);
324  tensor.tensor<float,3>()( 0, 60-1-iPF, 38) = pfCandIsBarrel;
325  tensor.tensor<float,3>()( 0, 60-1-iPF, 39) = pfCandHighPurityTrk;
326  tensor.tensor<float,3>()( 0, 60-1-iPF, 40) = pfCandPdgID==1;
327  tensor.tensor<float,3>()( 0, 60-1-iPF, 41) = pfCandPdgID==2;
328  tensor.tensor<float,3>()( 0, 60-1-iPF, 42) = pfCandPdgID==11;
329  tensor.tensor<float,3>()( 0, 60-1-iPF, 43) = pfCandPdgID==13;
330  tensor.tensor<float,3>()( 0, 60-1-iPF, 44) = pfCandPdgID==130;
331  tensor.tensor<float,3>()( 0, 60-1-iPF, 45) = pfCandPdgID==211;
332  tensor.tensor<float,3>()( 0, 60-1-iPF, 46) = pfCandTauIndMatch;
333  }
334 
335  if (graphVersion == 1) {
336  tensor.tensor<float,3>()( 0, 36-1-iPF, 0) = pfCandPt;
337  tensor.tensor<float,3>()( 0, 36-1-iPF, 1) = pfCandPz;
338  tensor.tensor<float,3>()( 0, 36-1-iPF, 2) = pfCandPtRel;
339  tensor.tensor<float,3>()( 0, 36-1-iPF, 3) = pfCandPzRel;
340  tensor.tensor<float,3>()( 0, 36-1-iPF, 4) = pfCandDr;
341  tensor.tensor<float,3>()( 0, 36-1-iPF, 5) = pfCandDEta;
342  tensor.tensor<float,3>()( 0, 36-1-iPF, 6) = pfCandDPhi;
343  tensor.tensor<float,3>()( 0, 36-1-iPF, 7) = pfCandEta;
344  tensor.tensor<float,3>()( 0, 36-1-iPF, 8) = pfCandDz;
345  tensor.tensor<float,3>()( 0, 36-1-iPF, 9) = pfCandDzSig;
346  tensor.tensor<float,3>()( 0, 36-1-iPF, 10) = pfCandD0;
347  tensor.tensor<float,3>()( 0, 36-1-iPF, 11) = pfCandD0Sig;
348  tensor.tensor<float,3>()( 0, 36-1-iPF, 12) = pfCandDzErr;
349  tensor.tensor<float,3>()( 0, 36-1-iPF, 13) = pfCandD0Err;
350  tensor.tensor<float,3>()( 0, 36-1-iPF, 14) = pfCandD0D0;
351  tensor.tensor<float,3>()( 0, 36-1-iPF, 15) = pfCandCharge==0;
352  tensor.tensor<float,3>()( 0, 36-1-iPF, 16) = pfCandCharge==1;
353  tensor.tensor<float,3>()( 0, 36-1-iPF, 17) = pfCandCharge==-1;
354  tensor.tensor<float,3>()( 0, 36-1-iPF, 18) = pfCandPdgID>22;
355  tensor.tensor<float,3>()( 0, 36-1-iPF, 19) = pfCandPdgID==22;
356  tensor.tensor<float,3>()( 0, 36-1-iPF, 20) = pfCandDVx_1;
357  tensor.tensor<float,3>()( 0, 36-1-iPF, 21) = pfCandDVy_1;
358  tensor.tensor<float,3>()( 0, 36-1-iPF, 22) = pfCandDVz_1;
359  tensor.tensor<float,3>()( 0, 36-1-iPF, 23) = pfCandD_1;
360  tensor.tensor<float,3>()( 0, 36-1-iPF, 24) = pfCandDzDz;
361  tensor.tensor<float,3>()( 0, 36-1-iPF, 25) = pfCandD0Dz;
362  tensor.tensor<float,3>()( 0, 36-1-iPF, 26) = pfCandD0Dphi;
363  tensor.tensor<float,3>()( 0, 36-1-iPF, 27) = pfCandPtRelPtRel;
364  tensor.tensor<float,3>()( 0, 36-1-iPF, 28) = pfCandPixHits;
365  tensor.tensor<float,3>()( 0, 36-1-iPF, 29) = pfCandHits;
366  tensor.tensor<float,3>()( 0, 36-1-iPF, 30) = pfCandLostInnerHits==-1;
367  tensor.tensor<float,3>()( 0, 36-1-iPF, 31) = pfCandLostInnerHits==0;
368  tensor.tensor<float,3>()( 0, 36-1-iPF, 32) = pfCandLostInnerHits==1;
369  tensor.tensor<float,3>()( 0, 36-1-iPF, 33) = pfCandLostInnerHits==2;
370  tensor.tensor<float,3>()( 0, 36-1-iPF, 34) = pfCandPuppiWeight;
371  tensor.tensor<float,3>()( 0, 36-1-iPF, 35) = (pfCandVtxQuality == 1);
372  tensor.tensor<float,3>()( 0, 36-1-iPF, 36) = (pfCandVtxQuality == 5);
373  tensor.tensor<float,3>()( 0, 36-1-iPF, 37) = (pfCandVtxQuality == 6);
374  tensor.tensor<float,3>()( 0, 36-1-iPF, 38) = (pfCandVtxQuality == 7);
375  tensor.tensor<float,3>()( 0, 36-1-iPF, 39) = (pfCandFromPV == 1);
376  tensor.tensor<float,3>()( 0, 36-1-iPF, 40) = (pfCandFromPV == 2);
377  tensor.tensor<float,3>()( 0, 36-1-iPF, 41) = (pfCandFromPV == 3);
378  tensor.tensor<float,3>()( 0, 36-1-iPF, 42) = pfCandIsBarrel;
379  tensor.tensor<float,3>()( 0, 36-1-iPF, 43) = pfCandHighPurityTrk;
380  tensor.tensor<float,3>()( 0, 36-1-iPF, 44) = pfCandPdgID==1;
381  tensor.tensor<float,3>()( 0, 36-1-iPF, 45) = pfCandPdgID==2;
382  tensor.tensor<float,3>()( 0, 36-1-iPF, 46) = pfCandPdgID==11;
383  tensor.tensor<float,3>()( 0, 36-1-iPF, 47) = pfCandPdgID==13;
384  tensor.tensor<float,3>()( 0, 36-1-iPF, 48) = pfCandPdgID==130;
385  tensor.tensor<float,3>()( 0, 36-1-iPF, 49) = pfCandPdgID==211;
386  tensor.tensor<float,3>()( 0, 36-1-iPF, 50) = pfCandTauIndMatch;
387  }
388  iPF++;
389  }
390  tensorflow::run(&(cache_->getSession()), { { "input_1", tensor } }, { "output_node0" }, {}, &outputs_);
391  predictions.matrix<float>()(tau_index, 0) = outputs_[0].flat<float>()(0);
392  }
393  return predictions;
394  }
395 
396 private:
397  edm::EDGetTokenT<pat::PackedCandidateCollection> pfcand_token;
398  edm::EDGetTokenT<reco::VertexCollection> vtx_token;
399  unsigned graphVersion;
400 };
401 
402 #include "FWCore/Framework/interface/MakerMacros.h"
403 DEFINE_FWK_MODULE(DPFIsolation);
pat::PackedCandidate::puppiWeight
float puppiWeight() const
Set both weights at once (with option for only full PUPPI)
Definition: PackedCandidate.cc:361
pat::PackedCandidate::dz
virtual float dz(size_t ipv=0) const
dz with respect to the PV[ipv]
Definition: PackedCandidate.h:496
DPFIsolation::pfcand_token
edm::EDGetTokenT< pat::PackedCandidateCollection > pfcand_token
Definition: DPFIsolation.cc:397
pat::PackedCandidate::dxyError
float dxyError() const override
uncertainty on dxy
Definition: PackedCandidate.h:507
mps_fire.i
i
Definition: mps_fire.py:269
reco::LeafCandidate::eta
double eta() const final
momentum pseudorapidity
Definition: LeafCandidate.h:137
DPFIsolation::fillDescriptions
static void fillDescriptions(edm::ConfigurationDescriptions &descriptions)
Definition: DPFIsolation.cc:44
AlCaHLTBitMon_ParallelJobs.p
p
Definition: AlCaHLTBitMon_ParallelJobs.py:152
pat::PackedCandidate::numberOfHits
int numberOfHits() const
Definition: PackedCandidate.h:454
pat::PackedCandidate::pdgId
int pdgId() const override
PDG identifier.
Definition: PackedCandidate.h:565
DPFIsolation::getNumberOfParticles
static unsigned getNumberOfParticles(unsigned graphVersion)
Definition: DPFIsolation.cc:32
deep_tau::DeepTauCache::getGraph
const tensorflow::GraphDef & getGraph() const
Definition: DeepTauBase.h:50
AlCaHLTBitMon_QueryRunRegistry.string
string
Definition: AlCaHLTBitMon_QueryRunRegistry.py:255
DEFINE_FWK_MODULE
#define DEFINE_FWK_MODULE(type)
Definition: MakerMacros.h:17
deep_tau::DeepTauCache::getSession
tensorflow::Session & getSession() const
Definition: DeepTauBase.h:49
deep_tau::DeepTauBase::cache_
const DeepTauCache * cache_
Definition: DeepTauBase.h:101
pat::PackedCandidateCollection
std::vector< pat::PackedCandidate > PackedCandidateCollection
Definition: PackedCandidate.h:791
pat::PackedCandidate::p4
const LorentzVector & p4() const override
four-momentum Lorentz vecto r
Definition: PackedCandidate.h:318
pat::PackedCandidate::dzError
float dzError() const override
uncertainty on dz
Definition: PackedCandidate.h:505
electrons_cff.vertices
vertices
Definition: electrons_cff.py:14
reco::LeafCandidate::pt
double pt() const final
transverse momentum
Definition: LeafCandidate.h:131
reco::VertexCollection
std::vector< Vertex > VertexCollection
collection of Vertex objects
Definition: VertexFwd.h:9
pat::Tau::tauID
float tauID(const std::string &name) const
pat::Tau::signalCands
reco::CandidatePtrVector signalCands() const
pat::PackedCandidate::vz
double vz() const override
z coordinate of vertex position
Definition: PackedCandidate.h:470
deep_tau::DeepTauBase::OutputCollection
std::map< std::string, Output > OutputCollection
Definition: DeepTauBase.h:82
constexpr
#define constexpr
pat
Definition: HeavyIon.h:7
models.generate
def generate(map_blobs=False, class_name=None)
Definition: models.py:187
deep_tau::DeepTauBase
Definition: DeepTauBase.h:58
deep_tau::DeepTauBase::outputs_
OutputCollection outputs_
Definition: DeepTauBase.h:100
deep_tau::DeepTauCache
Definition: DeepTauBase.h:40
pat::PackedCandidate::charge
int charge() const override
electric charge
Definition: PackedCandidate.h:301
pat::PackedCandidate::vy
double vy() const override
y coordinate of vertex position
Definition: PackedCandidate.h:468
DPFIsolation::GetNumberOfFeatures
static unsigned GetNumberOfFeatures(unsigned graphVersion)
Definition: DPFIsolation.cc:38
pat::PackedCandidate::pvAssociationQuality
const PVAssociationQuality pvAssociationQuality() const
Definition: PackedCandidate.h:488
reco::LeafCandidate::pz
double pz() const final
z coordinate of momentum vector
Definition: LeafCandidate.h:129
pat::PackedCandidate::fromPV
const PVAssoc fromPV(size_t ipv=0) const
Definition: PackedCandidate.h:476
mathSSE::sqrt
T sqrt(T t)
Definition: SSEVec.h:18
pat::PackedCandidate::trackHighPurity
bool trackHighPurity() const
true if the track had the highPurity quality bit
Definition: PackedCandidate.h:527
pat::PackedCandidate::lostInnerHits
LostInnerHits lostInnerHits() const
Definition: PackedCandidate.h:538
funct::abs
Abs< T >::type abs(const T &t)
Definition: Abs.h:22
reco::LeafCandidate::p4
const LorentzVector & p4() const final
four-momentum Lorentz vector
Definition: LeafCandidate.h:99
pat::PackedCandidate::pt
double pt() const override
transverse momentum
Definition: PackedCandidate.h:350
deep_tau::DeepTauBase::DeepTauBase
DeepTauBase(const edm::ParameterSet &cfg, const OutputCollection &outputs, const DeepTauCache *cache)
Definition: DeepTauBase.cc:79
end
#define end
Definition: vmac.h:39
min
T min(T a, T b)
Definition: MathUtil.h:58
edm::ParameterSetDescription::add
ParameterDescriptionBase * add(U const &iLabel, T const &value)
Definition: ParameterSetDescription.h:89
pat::PackedCandidate::numberOfPixelHits
int numberOfPixelHits() const
Definition: PackedCandidate.h:453
pat::Tau
Analysis-level tau class.
Definition: Tau.h:55
DPFIsolation::getPredictions
tensorflow::Tensor getPredictions(edm::Event &event, const edm::EventSetup &es, edm::Handle< TauCollection > taus) override
Definition: DPFIsolation.cc:85
helperFunctions.cache
def cache(function)
Definition: helperFunctions.py:152
pat::PackedCandidate::hasTrackDetails
bool hasTrackDetails() const
Return true if a bestTrack can be extracted from this Candidate.
Definition: PackedCandidate.h:523
deltaR
double deltaR(double eta1, double eta2, double phi1, double phi2)
Definition: TreeUtility.cc:17
pat::Tau::isTauIDAvailable
bool isTauIDAvailable(const std::string &name) const
Returns true if a specific ID is available in this pat::Tau.
pat::PackedCandidate::eta
double eta() const override
momentum pseudorapidity
Definition: PackedCandidate.h:376
pat::PackedCandidate::phi
double phi() const override
momentum azimuthal angle
Definition: PackedCandidate.h:352
DPFIsolation::vtx_token
edm::EDGetTokenT< reco::VertexCollection > vtx_token
Definition: DPFIsolation.cc:398
DPFIsolation::DPFIsolation
DPFIsolation(const edm::ParameterSet &cfg, const deep_tau::DeepTauCache *cache)
Definition: DPFIsolation.cc:68
edm::ConfigurationDescriptions::add
void add(std::string const &label, ParameterSetDescription const &psetDescription)
Definition: ConfigurationDescriptions.cc:67
reco
fixed size matrix
Definition: AlignmentAlgorithmBase.h:43
begin
#define begin
Definition: vmac.h:32
edm
HLT enums.
Definition: AlignableModifier.h:17
Output
#define Output(cl)
Definition: vmac.h:194
DPFIsolation::graphVersion
unsigned graphVersion
Definition: DPFIsolation.cc:399
tensorflow::run
void run(Session *session, const NamedTensorList &inputs, const std::vector< std::string > &outputNames, const std::vector< std::string > &targetNodes, std::vector< Tensor > *outputs)
Definition: TensorFlow.cc:210
hiPixelPairStep_cff.deltaPhi
deltaPhi
Definition: hiPixelPairStep_cff.py:87
DPFIsolation::GetOutputs
static const OutputCollection & GetOutputs()
Definition: DPFIsolation.cc:25
pat::PackedCandidate::dxy
virtual float dxy() const
dxy with respect to the PV ref
Definition: PackedCandidate.h:494
reco::LeafCandidate::phi
double phi() const final
momentum azimuthal angle
Definition: LeafCandidate.h:133
nano_cff.taus
taus
Definition: nano_cff.py:37
pat::PackedCandidate::vx
double vx() const override
x coordinate of vertex position
Definition: PackedCandidate.h:466
event
Definition: event.py:1
Generated for CMSSW Reference Manual by doxygen 1.8.11