[10_6_X] Add ParticleNet to NanoAOD

[hqucms]

PR description:

Backport of #31096.

This PR adds the ParticleNet tagger to NanoAOD. The following scores will be included in NanoAOD:

• ParticleNet: TvsQCD, WvsQCD, ZvsQCD, HbbvsQCD, HccvsQCD, H4qvsQCD, QCD
• ParticleNet-MD: Xbb, Xcc, Xqq, QCD

The ParticleNet tagger needs to run in the NanoAOD sequence until we have the tagger stored in MiniAOD. The updated training [V01] should be used for both UL and EOY samples.

FYI @camclean @alefisico

[cmsbuild]

A new Pull Request was created by @hqucms (Huilin Qu) for CMSSW_10_6_X.

It involves the following packages:

PhysicsTools/NanoAOD
RecoBTag/ONNXRuntime

@perrotta, @gouskos, @cmsbuild, @fgolf, @slava77, @jpata, @mariadalfonso, @santocch, @peruzzim can you please review it and eventually sign? Thanks.
@emilbols, @smoortga, @JyothsnaKomaragiri, @ferencek, @gpetruc, @andrzejnovak this is something you requested to watch as well.
@silviodonato, @dpiparo, @qliphy you are the release manager for this.

cms-bot commands are listed here

• Backported from Add ParticleNet to NanoAOD #31096

[mariadalfonso]

@hqucms
The use of modifiers is not appropriate we have to use the run2_miniAOD_devel (like for all the other back ports)

if run2_miniAOD_devel:
all will be available in miniAOD

if (~run2_miniAOD_devel) recompute stuff

[hqucms]

@hqucms
The use of modifiers is not appropriate we have to use the run2_miniAOD_devel (like for all the other back ports)

if run2_miniAOD_devel:
all will be available in miniAOD

if (~run2_miniAOD_devel) recompute stuff

@mariadalfonso
Is this for the no-change policy (i.e., the new taggers should only be added when run2_miniAOD_devel is enabled)?

[mariadalfonso]

@hqucms
The use of modifiers is not appropriate we have to use the run2_miniAOD_devel (like for all the other back ports)
if run2_miniAOD_devel:
all will be available in miniAOD
if (~run2_miniAOD_devel) recompute stuff

@mariadalfonso
Is this for the no-change policy (i.e., the new taggers should only be added when run2_miniAOD_devel is enabled)?

You are adding new variables for particleNet, and in previous eras the analyser will not use them so there is no problem if some variable is present in some dataset. For the old tagger you were already re-running them before, so this will result in no change.

The new mini in 10_6 are controlled by the run2_miniAOD_devel, so you can take particleNet from mini and add in nano only when the run2_miniAOD_devel is enabled.

[slava77]

@cmsbuild please test

[cmsbuild]

The tests are being triggered in jenkins.

• CMSSW_10_6_X_2020-08-25-2300/slc7_amd64_gcc700: https://cmssdt.cern.ch/jenkins/job/ib-run-pr-tests/8941/console Started: 2020/08/26 22:38

[cmsbuild]

+1
Tested at: cb85649
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-1fc915/8941/summary.html
CMSSW: CMSSW_10_6_X_2020-08-25-2300
SCRAM_ARCH: slc7_amd64_gcc700

[cmsbuild]

Comparison job queued.

[cmsbuild]

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-1fc915/8941/summary.html

Comparison Summary:

• No significant changes to the logs found
• Reco comparison results: 49 differences found in the comparisons
• DQMHistoTests: Total files compared: 34
• DQMHistoTests: Total histograms compared: 3214614
• DQMHistoTests: Total failures: 1
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 3214279
• DQMHistoTests: Total skipped: 334
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 2.114 KiB( 33 files compared)
• DQMHistoSizes: changed ( 1325.7 ): 2.114 KiB Physics/NanoAODDQM
• Checked 140 log files, 14 edm output root files, 34 DQM output files

[perrotta]

backport of #31096

[cmsbuild]

The tests are being triggered in jenkins.
Test Parameters:

• MATRIX_EXTRAS = 1325.7,1325.6,136.8521,1325.8,136.7722,1329.1,136.7952
• CMSSW_10_6_X_2020-08-27-1100/slc7_amd64_gcc700: https://cmssdt.cern.ch/jenkins/job/ib-run-pr-tests/8969/console Started: 2020/08/28 00:37

[cmsbuild]

+1
Tested at: 9cf15b7
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-1fc915/8969/summary.html
CMSSW: CMSSW_10_6_X_2020-08-27-1100
SCRAM_ARCH: slc7_amd64_gcc700

[cmsbuild]

Comparison job queued.

[cmsbuild]

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-1fc915/8969/summary.html

Comparison Summary:

• No significant changes to the logs found
• Reco comparison results: 49 differences found in the comparisons
• DQMHistoTests: Total files compared: 34
• DQMHistoTests: Total histograms compared: 3214618
• DQMHistoTests: Total failures: 1
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 3214283
• DQMHistoTests: Total skipped: 334
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 2.114 KiB( 33 files compared)
• DQMHistoSizes: changed ( 1325.7 ): 2.114 KiB Physics/NanoAODDQM
• Checked 140 log files, 14 edm output root files, 34 DQM output files

[mariadalfonso]

+xpog
Taggers added in nano directly from mini when minAOD_devel is active otherwise recomputed, new branch "ParticleNet" not present in previous nano will be available.

[slava77]

+1

for #31244 9cf15b7

• changes since the last reco signoff [10_6_X] Add ParticleNet to NanoAOD #31244 (comment) were outside (downstream) of reco, in nanoAOD https://github.com/cms-sw/cmssw/pull/31244/files/cb8564948618bcaa4699399fa339580a6148c3e4..9cf15b74188f058617e5b674a37f3c1f618685ce

[santocch]

+1

[cmsbuild]

This pull request is fully signed and it will be integrated in one of the next CMSSW_10_6_X IBs (tests are also fine) and once validation in the development release cycle CMSSW_11_2_X is complete. This pull request will now be reviewed by the release team before it's merged. @silviodonato, @dpiparo, @qliphy (and backports should be raised in the release meeting by the corresponding L2)

[qliphy]

+1