Adding photons to tau jet cones #128

msessini · 2022-09-16T08:32:19Z

No description provided.

shedprog · 2022-09-27T15:36:43Z

Production/plugins/TauTupleProducer.cc

+    template<typename Gamma, int cmssw_version>
+    struct GetGammaVer;
+
+    //CMSSW 10 --> 12
+    //hcalOverEcal()   -->  hadronicOverEm()
+    //hcalOverEcalBc() -->  hadTowOverEm()
+    template<typename Gamma>
+    struct GetGammaVer<Gamma, 12> {
+        static float hcalDepth1OverEcal(const Gamma* photon) { return photon->hadronicOverEm(1); }
+        static float hcalDepth2OverEcal(const Gamma* photon) { return photon->hadronicOverEm(2); }
+        static float hcalDepth1OverEcalBc(const Gamma* photon) { return photon->hadTowOverEm(1); }
+        static float hcalDepth2OverEcalBc(const Gamma* photon) { return photon->hadTowOverEm(2); }
+        static float full5x5_hcalDepth1OverEcal(const Gamma* photon) { return photon->full5x5_hadronicOverEm(1); }
+        static float full5x5_hcalDepth2OverEcal(const Gamma* photon) { return photon->full5x5_hadronicOverEm(2); }
+        static float full5x5_hcalDepth1OverEcalBc(const Gamma* photon) { return photon->full5x5_hadTowOverEm(1); }
+        static float full5x5_hcalDepth2OverEcalBc(const Gamma* photon) { return photon->full5x5_hadTowOverEm(2); }
+    };
+
+    template<typename Gamma>
+    struct GetGammaVer<Gamma, 10> {
+        static float hcalDepth1OverEcal(const Gamma* photon) { return photon->hadronicDepth1OverEm(); }
+        static float hcalDepth2OverEcal(const Gamma* photon) { return photon->hadronicDepth2OverEm(); }
+        static float hcalDepth1OverEcalBc(const Gamma* photon) { return photon->hadTowDepth1OverEm(); }
+        static float hcalDepth2OverEcalBc(const Gamma* photon) { return photon->hadTowDepth2OverEm(); }
+        static float full5x5_hcalDepth1OverEcal(const Gamma* photon) { return -999.; }
+        static float full5x5_hcalDepth2OverEcal(const Gamma* photon) { return -999.; }
+        static float full5x5_hcalDepth1OverEcalBc(const Gamma* photon) { return -999.; }
+        static float full5x5_hcalDepth2OverEcalBc(const Gamma* photon) { return -999.; }
+    };
+
+


@msessini please resolve the existing conflict in this part of the code.

And add overloading for version == 11 like we do for Electrons:

template<typename Elec> struct GetElecVer<Elec, 11> : GetElecVer<Elec, 10> {};

Actually I forgot to raise one point: do we want to keep compatibility with releases before CMSSW_12_4? UL MiniAODs can be processed with 12_4, so should be safe from that point. I think we can drop compatibility with previous releases, including python 2 (finally!)

It would make sense in general, but I still use UL for displaced taus. Also: what is the release for phase 2, isn't it 11__?

@shedprog - for UL it is safe to move to 12_4_X for miniAOD->TauTuple step. Actually we need to do this to include DeepTau v2p5 in the next ntuples.
@DennRoy - is it ok to move to >= 12_X for the future Phase 2 studies?

There are at least two changes required to make Phase2 production work with CMSSW 12. These are discussed here:
https://cms-talk.web.cern.ch/t/getting-muon-station-value-0/15765/1

thank you @DennRoy ! I think this can be implemented in this PR and then we can drop compatibility for CMSSW < 12, if there are no other objections.

@DennRoy, which CMSSW_12_4_X release would be best for phase2 production as replacement of CMSSW_11_2_5 ?

I'm not sure about any "best" release. I've used 12_4_0 and 12_4_8 before, which seems to have worked equally well. But I've also been told that 12_6_0_pre3 MIGHT fix the low tau reco efficiency in the Phase2 endcap (to be confirmed).

Analysis/interface/TauTuple.h

Production/interface/Selectors.h

Production/plugins/TauTupleProducer.cc

env.sh

Production/src/TauJet.cc

msessini · 2022-10-14T09:12:05Z

General question, it seems that EDAnalyzer is now deprecated (??), do you know what should be used instead ?

warning: 'edm::EDAnalyzer::EDAnalyzer()' is deprecated [-Wdeprecated-declarations]

ofivite · 2022-10-14T10:09:08Z

General question, it seems that EDAnalyzer is now deprecated (??), do you know what should be used instead ?

Hmm this is something unusual.. I personally haven't heard about that, so searching for this warning led me to some PRs in CMSSW doing an API change to edm::one::EDAnalyzer, for example: cms-sw/cmssw#35343 , cms-sw/cmssw#37044

And then I found this twiki page: https://twiki.cern.ch/twiki/bin/view/CMSPublic/FWMultithreadedFrameworkOneModuleInterface

It looks like they talk about thread safety, but I would need to read this further. @kandrosov do you think it affects us and we would need to migrate too?

env.sh

shedprog · 2022-10-17T12:55:03Z

Production/plugins/TauTupleProducer.cc

+        for(const auto& sv_ptr : sv_col) {
+            const reco::VertexCompositePtrCandidate* SV = sv_ptr.obj;
+            auto candsIdx = collectCands(*SV);
+            if(candsIdx.size() > 0){
+	        tauTuple().sv_x.push_back(static_cast<float>(SV->position().x()));
+                tauTuple().sv_y.push_back(static_cast<float>(SV->position().y()));
+           	tauTuple().sv_z.push_back(static_cast<float>(SV->position().z()));
+           	tauTuple().sv_t.push_back(static_cast<float>(SV->t()));
+	   	tauTuple().sv_xE.push_back(static_cast<float>(std::sqrt(SV->vertexCovariance(0,0))));
+           	tauTuple().sv_yE.push_back(static_cast<float>(std::sqrt(SV->vertexCovariance(1,1))));
+           	tauTuple().sv_zE.push_back(static_cast<float>(std::sqrt(SV->vertexCovariance(2,2))));
+           	tauTuple().sv_tE.push_back(static_cast<float>(SV->tError()));
+	   	tauTuple().sv_chi2.push_back(static_cast<float>(SV->vertexChi2()));
+           	tauTuple().sv_ndof.push_back(static_cast<float>(SV->vertexNdof()));
+		for(int cand_idx : candsIdx) {
+                    tauTuple().sv_cands_svIdx.push_back(sv_ptr.index);
+                    tauTuple().sv_cands_pfcandIdx.push_back(cand_idx);
+  		}
+	        ++nsv;
+            }


@ofivite correct me if I am wrong, in this way -> SVs that matched to PFCand that selected within cone will be filled only, so then the difference I was thinking to also add the SVs that are within the cone but do not have any pfcands. Also, this structure with indexes is a bit weird, I would have indexes for Pfcands pointing SV then it is just

we have chosen SV<->pfCand mapping as a separate vectors, because we were not sure if pfCand->SV is guaranteed to be unique. Do you know if it is unique by design?

kandrosov · 2022-10-17T13:32:54Z

It looks like they talk about thread safety, but I would need to read this further. @kandrosov do you think it affects us and we would need to migrate too?

it is usually safer to run crab tasks with one thread to reduce memory limit exceptions. I would say it is not urgent to migrate, but we can add it to our todo list to stay aligned with the mainstream CMSSW code.

kandrosov · 2022-10-17T13:38:06Z

Production/src/MuonHitMatch.cc

@@ -47,7 +47,7 @@ void MuonHitMatch::CountMatches(const pat::Muon& muon, CountMap& n_matches)
 {
    for(const auto& segment : muon.matches()) {
        if(segment.segmentMatches.empty() && segment.rpcMatches.empty() && segment.gemMatches.empty() && segment.me0Matches.empty()) continue;
-        if(n_matches.count(segment.detector())) {
+        if(n_matches.count(segment.detector()) && segment.station()!=0) {


do we understand why it is needed? If 0 is a valid station, why it should not be saved?

I don't know if "0" is really valid (My understanding is insufficient to answer this). In order to fix the crash that happens otherwise here, the alternative would be to allow the station 0 here.

I agree that the crush must be fixed. But it would be preferable to make a conscious choice whatever we want to ignore stations "0" or they could be useful, and then modify the code accordingly.

Production/src/TauJet.cc

env.sh

Production/src/TauJet.cc

shedprog · 2022-10-18T17:37:39Z

Production/plugins/TauTupleProducer.cc

+	        tauTuple().sv_x.push_back(static_cast<float>(SV->position().x()));
+                tauTuple().sv_y.push_back(static_cast<float>(SV->position().y()));
+           	tauTuple().sv_z.push_back(static_cast<float>(SV->position().z()));
+           	tauTuple().sv_t.push_back(static_cast<float>(SV->t()));


I would also store pt, eta, phi, mass

Production/plugins/TauTupleProducer.cc

Production/src/TauJet.cc

Production/plugins/TauTupleProducer.cc

env.sh

Adding photons variables to jet cones

5fb4ee8

shedprog requested changes Sep 27, 2022

View reviewed changes

Adding secondary vertices and more photons variables

0ee030c

ofivite requested changes Oct 3, 2022

View reviewed changes

Analysis/interface/TauTuple.h Show resolved Hide resolved

Analysis/interface/TauTuple.h Show resolved Hide resolved

Production/interface/Selectors.h Outdated Show resolved Hide resolved

Production/plugins/TauTupleProducer.cc Outdated Show resolved Hide resolved

msessini added 2 commits October 6, 2022 00:46

Adding all pf candidates associated to a SV

5c085f2

Resolving conflicts

704a17e

ofivite requested changes Oct 10, 2022

View reviewed changes

Production/plugins/TauTupleProducer.cc Outdated Show resolved Hide resolved

env.sh Outdated Show resolved Hide resolved

Production/src/TauJet.cc Outdated Show resolved Hide resolved

Production/src/TauJet.cc Outdated Show resolved Hide resolved

Update for SV filling

57a7ca9

shedprog reviewed Oct 14, 2022

View reviewed changes

env.sh Outdated Show resolved Hide resolved

shedprog reviewed Oct 17, 2022

View reviewed changes

kandrosov requested changes Oct 17, 2022

View reviewed changes

shedprog reviewed Oct 18, 2022

View reviewed changes

kandrosov requested changes Oct 21, 2022

View reviewed changes

Production/plugins/TauTupleProducer.cc Outdated Show resolved Hide resolved

Production/plugins/TauTupleProducer.cc Outdated Show resolved Hide resolved

Production/src/TauJet.cc Outdated Show resolved Hide resolved

Production/plugins/TauTupleProducer.cc Outdated Show resolved Hide resolved

Update for SV filling

61203e6

kandrosov requested changes Nov 2, 2022

View reviewed changes

Production/plugins/TauTupleProducer.cc Outdated Show resolved Hide resolved

env.sh Outdated Show resolved Hide resolved

Update 03/11

2a3ca28

kandrosov approved these changes Nov 3, 2022

View reviewed changes

ofivite approved these changes Nov 3, 2022

View reviewed changes

Merge branch 'master' into master

108bdf6

shedprog approved these changes Nov 7, 2022

View reviewed changes

shedprog merged commit db4ed90 into cms-tau-pog:master Nov 7, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Adding photons to tau jet cones #128

Adding photons to tau jet cones #128

msessini commented Sep 16, 2022

shedprog Sep 27, 2022

kandrosov Sep 27, 2022

shedprog Sep 27, 2022

kandrosov Oct 3, 2022

DennRoy Oct 3, 2022

kandrosov Oct 3, 2022

msessini Oct 4, 2022 •

edited

DennRoy Oct 4, 2022

msessini commented Oct 14, 2022

ofivite commented Oct 14, 2022

shedprog Oct 17, 2022 •

edited

kandrosov Oct 17, 2022

kandrosov commented Oct 17, 2022

kandrosov Oct 17, 2022

DennRoy Oct 17, 2022

kandrosov Oct 17, 2022 •

edited

shedprog Oct 18, 2022

Adding photons to tau jet cones #128

Adding photons to tau jet cones #128

Conversation

msessini commented Sep 16, 2022

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

msessini Oct 4, 2022 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

msessini commented Oct 14, 2022

ofivite commented Oct 14, 2022

shedprog Oct 17, 2022 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kandrosov commented Oct 17, 2022

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kandrosov Oct 17, 2022 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

msessini Oct 4, 2022 •

edited

shedprog Oct 17, 2022 •

edited

kandrosov Oct 17, 2022 •

edited