Fixing RECO in pp_on_AA era jet reconstruction #24655

stepobr · 2018-09-25T14:18:09Z

This pull request is a follow-up of #24380
The issue is partly described in #24587
The ideal scenario for #24380 would be to get rid of all jet algorithms not used for pp_on_AA era. However 3 of them (ak4PF, ak4PFCHS, ak8PFCHS) were kept because otherwise it failed to pass the validation step. But in that case it was later found that the sequences crashes if stopped after RECO step due to the absence of pfNoPileUpJME used by the 3 mentioned algorithms. Trivial solution would be to add back pfNoPileUpJME. But in order to speed-up reconstruction it was replaced by a dummy EDFilter with a very high pt threshold which results in an empty input for these 3 jet modules. The filter does not prevent the modules to be run, but since they have no input objects they do nothing.

@cfmcginn @mandrenguyen

cmsbuild · 2018-09-25T14:18:27Z

The code-checks are being triggered in jenkins.

cmsbuild · 2018-09-25T14:20:56Z

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-24655/6565

cmsbuild · 2018-09-25T14:21:14Z

A new Pull Request was created by @stepobr (Stepan Obraztsov) for master.

It involves the following packages:

RecoHI/HiJetAlgos
RecoJets/Configuration
RecoJets/JetProducers

@perrotta, @cmsbuild, @slava77 can you please review it and eventually sign? Thanks.
@TaiSakuma, @jdolen, @ahinzmann, @schoef, @rappoccio, @jazzitup, @jdamgov, @yslai, @nhanvtran, @yenjie, @gkasieczka, @clelange, @kurtejung, @mariadalfonso, @mandrenguyen, @dgulhan, @seemasharmafnal, @yetkinyilmaz this is something you requested to watch as well.
@davidlange6, @slava77, @fabiocos you are the release manager for this.

cms-bot commands are listed here

slava77 · 2018-09-25T14:33:46Z

RecoJets/JetProducers/python/ak4PFJets_cfi.py

@@ -33,3 +33,6 @@
    doAreaFastjet = cms.bool(True),
    jetPtMin = cms.double(100.0)
    )
+from Configuration.Eras.Modifier_pp_on_AA_2018_cff import pp_on_AA_2018
+pp_on_AA_2018.toModify(ak4PFJets,src = cms.InputTag("pfNoPileUpJMEHI"), inputEtMin = cms.double(9999))
+pp_on_AA_2018.toModify(ak4PFJetsCHS,src = cms.InputTag("pfNoPileUpJMEHI"), inputEtMin = cms.double(9999))


please drop type specifications for all parameters which already exist.
This is a safer way to protect from parameter name mistakes and will also help in possible parameter migrations.

slava77 · 2018-09-25T14:34:26Z

@cmsbuild please test

cmsbuild · 2018-09-25T14:34:46Z

The tests are being triggered in jenkins.
https://cmssdt.cern.ch/jenkins/job/ib-any-integration/30629/console Started: 2018/09/25 16:59

cmsbuild · 2018-09-25T18:13:05Z

+1
Tested at: 934d61f
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-24655/30629/summary.html

cmsbuild · 2018-09-25T18:13:07Z

Comparison job queued.

cmsbuild · 2018-09-25T20:12:21Z

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-24655/30629/summary.html

Comparison Summary:

No significant changes to the logs found
Reco comparison results: 308 differences found in the comparisons
DQMHistoTests: Total files compared: 32
DQMHistoTests: Total histograms compared: 3146876
DQMHistoTests: Total failures: 1934
DQMHistoTests: Total nulls: 0
DQMHistoTests: Total successes: 3144745
DQMHistoTests: Total skipped: 197
DQMHistoTests: Total Missing objects: 0
DQMHistoSizes: Histogram memory added: 0.0 KiB( 31 files compared)
Checked 133 log files, 14 edm output root files, 32 DQM output files

stepobr · 2018-09-29T11:45:33Z

Most of your points are correct, a few comments:

May be that's a typo but from that list ideally I would get rid of only ak4PFJets, ak4PFJetsCHS, ak8PFJetsCHS.
I can't do that not because of DQM, but rather because of the validation step. For example, excluding ak4PFJets from the sequence leads to:
An exception of category 'ProductNotFound' occurred while [0] Processing Event run: 1 lumi: 1 event: 1 stream: 3 [1] Running path 'prevalidation_step' [2] Prefetching for module MultiTrackValidator/'trackValidator' [3] Prefetching for module JetTracksAssociationToTrackRefs/'cutsRecoTracksAK4PFJets' [4] Calling method for module JetTracksAssociatorExplicit/'ak4JetTracksAssociatorExplicitAll' Exception Message: Principal::getByToken: Found zero products matching all criteria Looking for a container with elements of type: reco::Jet Looking for module label: ak4PFJets Looking for productInstanceName:
Yes, this is the correct logic behind using this filter.

I will update the description now.

perrotta · 2018-10-02T08:41:09Z

@cmsbuild please test
(to have back the plots with differences, which have disappeared)

cmsbuild · 2018-10-02T08:41:28Z

The tests are being triggered in jenkins.
https://cmssdt.cern.ch/jenkins/job/ib-any-integration/30841/console Started: 2018/10/02 10:42

cmsbuild · 2018-10-02T11:20:43Z

+1
Tested at: 00abcf2
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-24655/30841/summary.html

cmsbuild · 2018-10-02T11:20:46Z

Comparison job queued.

cmsbuild · 2018-10-02T12:55:07Z

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-24655/30841/summary.html

Comparison Summary:

No significant changes to the logs found
Reco comparison results: 311 differences found in the comparisons
DQMHistoTests: Total files compared: 32
DQMHistoTests: Total histograms compared: 3162800
DQMHistoTests: Total failures: 1567
DQMHistoTests: Total nulls: 0
DQMHistoTests: Total successes: 3161036
DQMHistoTests: Total skipped: 197
DQMHistoTests: Total Missing objects: 0
DQMHistoSizes: Histogram memory added: 0.0 KiB( 31 files compared)
Checked 134 log files, 14 edm output root files, 32 DQM output files

perrotta · 2018-10-02T16:45:29Z

+1

ak4PF, ak4PFCHS, ak8PFCHS are "sterilized" thanks to the pfNoPileUpJME selector, which feeds them with an empty set of PF objects
The implementation follows what is in the PR description
Changes in the jenkins tets outputs are only in the 2018 HI workflow, and they correspond to the expectations (empty ak4PF, ak4PFCHS, ak8PFCHS collections)

cmsbuild · 2018-10-02T16:45:49Z

This pull request is fully signed and it will be integrated in one of the next master IBs (tests are also fine). This pull request will now be reviewed by the release team before it's merged. @davidlange6, @slava77, @smuzaffar, @fabiocos (and backports should be raised in the release meeting by the corresponding L2)

fabiocos · 2018-10-03T15:49:37Z

+1

stepobr added 4 commits September 25, 2018 15:53

add dummy sequence

b84c189

bug fix

36da67f

increase threshold for algos not used in pp_on_AA era

027bb76

fixing pp_on_AA jet sequence

934d61f

cmsbuild added this to the CMSSW_10_3_X milestone Sep 25, 2018

cmsbuild added code-checks-pending comparison-pending orp-pending pending-signatures reconstruction-pending tests-pending labels Sep 25, 2018

cmsbuild added code-checks-approved and removed code-checks-pending labels Sep 25, 2018

slava77 reviewed Sep 25, 2018

View reviewed changes

cmsbuild added tests-started and removed tests-pending labels Sep 25, 2018

cmsbuild added tests-approved and removed tests-started labels Sep 25, 2018

cmsbuild added comparison-available and removed comparison-pending labels Sep 25, 2018

remove type specifications

00abcf2

cmsbuild removed the code-checks-approved label Sep 25, 2018

cmsbuild added comparison-pending tests-pending and removed comparison-available tests-approved labels Oct 2, 2018

cmsbuild added tests-started and removed tests-pending labels Oct 2, 2018

cmsbuild added tests-approved and removed tests-started labels Oct 2, 2018

cmsbuild added comparison-available and removed comparison-pending labels Oct 2, 2018

cmsbuild added fully-signed reconstruction-approved and removed pending-signatures reconstruction-pending labels Oct 2, 2018

cmsbuild added orp-approved and removed orp-pending labels Oct 3, 2018

cmsbuild merged commit e07a28e into cms-sw:master Oct 3, 2018

This was referenced Oct 18, 2018

Raw data remapper sequence #24910

Merged

Bug fixes to Run2_2018_pp_on_AA in EI and DQM #24958

Merged

bi-ran mentioned this pull request May 21, 2020

restore jet collections/use dummy inputs CmsHI/cmssw#289

Merged

stepobr deleted the from-CMSSW_10_3_0_pre4 branch May 21, 2020 17:40

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fixing RECO in pp_on_AA era jet reconstruction #24655

Fixing RECO in pp_on_AA era jet reconstruction #24655

stepobr commented Sep 25, 2018 •

edited

cmsbuild commented Sep 25, 2018

cmsbuild commented Sep 25, 2018

cmsbuild commented Sep 25, 2018

slava77 Sep 25, 2018

slava77 commented Sep 25, 2018

cmsbuild commented Sep 25, 2018 •

edited

cmsbuild commented Sep 25, 2018

cmsbuild commented Sep 25, 2018

cmsbuild commented Sep 25, 2018

stepobr commented Sep 29, 2018

perrotta commented Oct 2, 2018

cmsbuild commented Oct 2, 2018 •

edited

cmsbuild commented Oct 2, 2018

cmsbuild commented Oct 2, 2018

cmsbuild commented Oct 2, 2018

perrotta commented Oct 2, 2018

cmsbuild commented Oct 2, 2018

fabiocos commented Oct 3, 2018

Fixing RECO in pp_on_AA era jet reconstruction #24655

Fixing RECO in pp_on_AA era jet reconstruction #24655

Conversation

stepobr commented Sep 25, 2018 • edited

cmsbuild commented Sep 25, 2018

cmsbuild commented Sep 25, 2018

cmsbuild commented Sep 25, 2018

slava77 Sep 25, 2018

Choose a reason for hiding this comment

slava77 commented Sep 25, 2018

cmsbuild commented Sep 25, 2018 • edited

cmsbuild commented Sep 25, 2018

cmsbuild commented Sep 25, 2018

cmsbuild commented Sep 25, 2018

stepobr commented Sep 29, 2018

perrotta commented Oct 2, 2018

cmsbuild commented Oct 2, 2018 • edited

cmsbuild commented Oct 2, 2018

cmsbuild commented Oct 2, 2018

cmsbuild commented Oct 2, 2018

perrotta commented Oct 2, 2018

cmsbuild commented Oct 2, 2018

fabiocos commented Oct 3, 2018

stepobr commented Sep 25, 2018 •

edited

cmsbuild commented Sep 25, 2018 •

edited

cmsbuild commented Oct 2, 2018 •

edited