Further preparations of deep tau ID for hlt phase2 #32676

swozniewski · 2021-01-18T09:11:16Z

PR description:

Modifications of the DeepTauID code for phase II tau trigger studies:

add auxiliary functionality like a check of input IDs, file dump of input quantities.
add options disable_CellIndex_workaround, disable_hcalFraction_workaround to turn off workarounds which won't be needed in the future.
add option enableHGCalWorkaround to PFRecoTauDiscriminationByIsolation serving as input to DeepTauID to enable a workaround needed for phase II.
Changed default value for variables related to number of hits to 0 instead of -999
Change the puppiweight to have an additional argument to define its return value for AOD. This allows the puppiweights to be defined separately for different blocks. Additionally split between weights for inner and outer blocks. All values have been extracted from the training.

Clean up of defaults in DeepTauID functions introduced during last preparation of phase II trigger studies (PR #31744 ):

Remove the default_value constant from all deepTau v2 code
Removed default_value argument from certain candfunctions and instead use conditionals where needed in the places these functions are called.

Modifications/defaults preserve the current DeepTauID for PAT level and no physics changes are intended there.
DeepTauID on AOD as introduced recently in PR #31744 is developing and changes are intended.

PR validation:

Basic check by the developers: Reran config file with update from deepTauAtHLT.py and everything ran smoothly. Running RecoTauTag/RecoTau/test/runDeepTauIDsOnMiniAOD.py also ran fine. Upon brief inspection, no strange behaviour observed.
code checks and format applied
unit tests pass
limited matrix tests pass

cmsbuild · 2021-01-18T09:18:40Z

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-32676/20773

This PR adds an extra 60KB to repository

cmsbuild · 2021-01-18T09:19:05Z

A new Pull Request was created by @swozniewski for master.

It involves the following packages:

RecoTauTag/RecoTau

@perrotta, @jpata, @cmsbuild, @slava77 can you please review it and eventually sign? Thanks.
@silviodonato, @dpiparo, @qliphy you are the release manager for this.

cms-bot commands are listed here

kandrosov · 2021-01-18T09:41:31Z

RecoTauTag/RecoTau/plugins/DeepTauId.cc

@@ -415,6 +423,22 @@ namespace {
    }
  }  // namespace dnn_inputs_2017_v2

+  float getTauID(const pat::Tau& tau, const std::string& tauID, float default_value = -999.) {
+    std::lock_guard<std::mutex> guard(g_mutex);


guard can be placed inside else section to not block execution for the cases when tau id is available.

mutex can be defined as static inside getTauID instead of using the global namespace

A better solution would be to use something like tbb::concurrent_unordered_set<std::string> for isFirstWarning, and check with something like

if (isFirstWarning.insert(tauID).second) { edm::LogWarning("TauID") << ... }

which doesn't require any explicit locks.

cmsbuild · 2021-01-18T10:01:51Z

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-32676/20775

This PR adds an extra 60KB to repository

cmsbuild · 2021-01-18T10:02:13Z

Pull request #32676 was updated. @perrotta, @jpata, @cmsbuild, @slava77 can you please check and sign again.

slava77 · 2021-01-18T13:42:44Z

@cmsbuild please test

cmsbuild · 2021-01-18T16:16:47Z

+1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-0b3e31/12320/summary.html
COMMIT: 1ab315e
CMSSW: CMSSW_11_3_X_2021-01-17-2300/slc7_amd64_gcc900

Comparison Summary

Summary:

No significant changes to the logs found
Reco comparison results: 0 differences found in the comparisons
DQMHistoTests: Total files compared: 37
DQMHistoTests: Total histograms compared: 2716961
DQMHistoTests: Total failures: 1
DQMHistoTests: Total nulls: 0
DQMHistoTests: Total successes: 2716938
DQMHistoTests: Total skipped: 22
DQMHistoTests: Total Missing objects: 0
DQMHistoSizes: Histogram memory added: 0.0 KiB( 36 files compared)
Checked 156 log files, 37 edm output root files, 37 DQM output files

cmsbuild · 2021-01-19T14:30:41Z

-code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-32676/20787

This PR adds an extra 64KB to repository

Code check has found code style and quality issues which could be resolved by applying following patch(s)

code-format:
https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-32676/20787/code-format.patch
e.g. curl https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-32676/20787/code-format.patch | patch -p1
You can also run scram build code-format to apply code format directly

swozniewski · 2021-02-05T10:42:14Z

apparently I was a bit naive about what small changes require different formatting. To be fixed shortly...

cmsbuild · 2021-02-05T10:51:58Z

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-32676/21025

This PR adds an extra 180KB to repository

cmsbuild · 2021-02-05T10:52:20Z

Pull request #32676 was updated. @perrotta, @jpata, @cmsbuild, @slava77 can you please check and sign again.

slava77 · 2021-02-05T13:17:17Z

@cmsbuild please test

cmsbuild · 2021-02-05T15:54:00Z

+1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-0b3e31/12734/summary.html
COMMIT: 036963d
CMSSW: CMSSW_11_3_X_2021-02-05-1100/slc7_amd64_gcc900
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week0/cms-sw/cmssw/32676/12734/install.sh to create a dev area with all the needed externals and cmssw changes.

Comparison Summary

Summary:

No significant changes to the logs found
Reco comparison results: 7 differences found in the comparisons
DQMHistoTests: Total files compared: 37
DQMHistoTests: Total histograms compared: 2751765
DQMHistoTests: Total failures: 12
DQMHistoTests: Total nulls: 1
DQMHistoTests: Total successes: 2751730
DQMHistoTests: Total skipped: 22
DQMHistoTests: Total Missing objects: 0
DQMHistoSizes: Histogram memory added: 0.004 KiB( 36 files compared)
DQMHistoSizes: changed ( 312.0 ): 0.004 KiB MessageLogger/Warnings
Checked 156 log files, 37 edm output root files, 37 DQM output files

swozniewski · 2021-02-05T16:01:48Z

The quantities where the seven differences show up are not known to me and do not seem to be tau related. Please let me know if you think that these differences are in fact due to the latest commit.

slava77 · 2021-02-05T16:21:50Z

+1

for #32676 036963d

code changes are in line with the PR description and the follow up review; the changes are technical in the context of the use cases in miniAOD
jenkins tests pass and comparisons with the baseline show no relevant differences
I managed to run the "online" variant as described in Further preparations of deep tau ID for hlt phase2 #32676 (comment) somewhat blindly in CMSSW_11_3_X_2021-02-04-1100 (after a small fix in config of PFBlockProducer, a change since the 11_2_0 version of GRun)

cmsbuild · 2021-02-05T16:22:12Z

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. @silviodonato, @dpiparo, @qliphy (and backports should be raised in the release meeting by the corresponding L2)

qliphy · 2021-02-05T16:38:32Z

+1

cmsbuild added this to the CMSSW_11_3_X milestone Jan 18, 2021

cmsbuild added code-checks-pending orp-pending pending-signatures reconstruction-pending tests-pending labels Jan 18, 2021

cmsbuild added code-checks-approved and removed code-checks-pending labels Jan 18, 2021

kandrosov reviewed Jan 18, 2021

View reviewed changes

cmsbuild added code-checks-pending and removed code-checks-approved labels Jan 18, 2021

cmsbuild added code-checks-approved and removed code-checks-pending labels Jan 18, 2021

cmsbuild added tests-started and removed tests-pending labels Jan 18, 2021

cmsbuild added tests-approved code-checks-pending tests-pending and removed tests-started code-checks-approved tests-approved labels Jan 18, 2021

cmsbuild added code-checks-rejected and removed code-checks-pending labels Jan 19, 2021

cmsbuild added code-checks-rejected and removed code-checks-pending labels Feb 5, 2021

fix code format

036963d

cmsbuild added code-checks-pending and removed code-checks-rejected labels Feb 5, 2021

cmsbuild added code-checks-approved and removed code-checks-pending labels Feb 5, 2021

cmsbuild added tests-started and removed tests-pending labels Feb 5, 2021

cmsbuild added tests-approved and removed tests-started labels Feb 5, 2021

cmsbuild added fully-signed reconstruction-approved and removed pending-signatures reconstruction-pending labels Feb 5, 2021

cmsbuild added orp-approved and removed orp-pending labels Feb 5, 2021

cmsbuild merged commit acf500a into cms-sw:master Feb 5, 2021

This was referenced Feb 5, 2021

Preparations of deep tau ID for hlt phase2 (backport to CMSSW_11_1_X) #32831

Merged

Further preparations of deep tau ID for hlt phase2 (backport to CMSSW_11_2_X) #32832

Merged

slava77 mentioned this pull request Feb 6, 2021

ASAN problem in DeepTauId #32837

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Further preparations of deep tau ID for hlt phase2 #32676

Further preparations of deep tau ID for hlt phase2 #32676

swozniewski commented Jan 18, 2021

cmsbuild commented Jan 18, 2021

cmsbuild commented Jan 18, 2021

kandrosov Jan 18, 2021

swozniewski Jan 18, 2021

fwyzard Jan 19, 2021

cmsbuild commented Jan 18, 2021

cmsbuild commented Jan 18, 2021

slava77 commented Jan 18, 2021

cmsbuild commented Jan 18, 2021

cmsbuild commented Jan 19, 2021

swozniewski commented Feb 5, 2021

cmsbuild commented Feb 5, 2021

cmsbuild commented Feb 5, 2021

slava77 commented Feb 5, 2021

cmsbuild commented Feb 5, 2021

swozniewski commented Feb 5, 2021

slava77 commented Feb 5, 2021

cmsbuild commented Feb 5, 2021

qliphy commented Feb 5, 2021

Further preparations of deep tau ID for hlt phase2 #32676

Further preparations of deep tau ID for hlt phase2 #32676

Conversation

swozniewski commented Jan 18, 2021

PR description:

PR validation:

cmsbuild commented Jan 18, 2021

cmsbuild commented Jan 18, 2021

kandrosov Jan 18, 2021

Choose a reason for hiding this comment

swozniewski Jan 18, 2021

Choose a reason for hiding this comment

fwyzard Jan 19, 2021

Choose a reason for hiding this comment

cmsbuild commented Jan 18, 2021

cmsbuild commented Jan 18, 2021

slava77 commented Jan 18, 2021

cmsbuild commented Jan 18, 2021

Comparison Summary

cmsbuild commented Jan 19, 2021

swozniewski commented Feb 5, 2021

cmsbuild commented Feb 5, 2021

cmsbuild commented Feb 5, 2021

slava77 commented Feb 5, 2021

cmsbuild commented Feb 5, 2021

Comparison Summary

swozniewski commented Feb 5, 2021

slava77 commented Feb 5, 2021

cmsbuild commented Feb 5, 2021

qliphy commented Feb 5, 2021