Change tau trigger bits stored in NanoAOD

[mbluj]

PR description:

The tau trigger bits in NanoAOD never provided a match in 2018 triggers because the 'Hps' part in the name was forgotten. This PR is meant to fix this issue as well as to prepare to accommodate Run-3 tau triggers. We have decided to go with a unified strategy that will work for both Run-2 and Run-3 triggers which have similar structure and naming conventions, hence the addition of extra bits containing DeepTau. It now also contains trigger bits for taus in the PAG VBF and displaced tau triggers. A full overview of all trigger filters in Run-2 and Run-3 was made and is contained in google sheets here

Changes:

• Addition of wildcards to original trigger bits to work for both HPS and non-HPS paths;
• Addition of DeepTau trigger bits;
• Addition of bits 4096 to 32768 to include PAG paths;
• Addition of bits 65536 for completeness, they allow to go a bit more specific in the triggers to select;
• Addition of boosted tau block. Right now the bits are redundant, but we expect more to be added in the future;
• Added bits to non-tau blocks from cross triggers.

PR validation:

Reran NanoAOD for all three Run-2 years and in a Run-3 sample, everything seemed to act as expected

[cmsbuild]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-38472/30686

• This PR adds an extra 20KB to repository

[cmsbuild]

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

It involves the following packages:

• PhysicsTools/NanoAOD (xpog)

@cmsbuild, @mariadalfonso, @gouskos, @fgolf can you please review it and eventually sign? Thanks.
@gpetruc, @swertz this is something you requested to watch as well.
@perrotta, @dpiparo, @qliphy you are the release manager for this.

cms-bot commands are listed here

[mariadalfonso]

please test

[cmsbuild]

+1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-e3dcee/25702/summary.html
COMMIT: e5864f0
CMSSW: CMSSW_12_5_X_2022-06-22-1100/el8_amd64_gcc10
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week0/cms-sw/cmssw/38472/25702/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: 13 differences found in the comparisons
• DQMHistoTests: Total files compared: 50
• DQMHistoTests: Total histograms compared: 3659099
• DQMHistoTests: Total failures: 161
• DQMHistoTests: Total nulls: 1
• DQMHistoTests: Total successes: 3658915
• DQMHistoTests: Total skipped: 22
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.004 KiB( 49 files compared)
• DQMHistoSizes: changed ( 312.0 ): 0.004 KiB MessageLogger/Warnings
• Checked 208 log files, 45 edm output root files, 50 DQM output files
• TriggerResults: no differences found

[mariadalfonso]

@mbluj
as expected there are differences in the [NanoAODDQM]/[TrigObj]
You can have a look here and report on if is all expected
https://gitlab.cern.ch/cms-nanoAOD/nanoAOD-integration/-/issues/155

[mbluj]

@mbluj as expected there are differences in the [NanoAODDQM]/[TrigObj] You can have a look here and report on if is all expected https://gitlab.cern.ch/cms-nanoAOD/nanoAOD-integration/-/issues/155

@mariadalfonso, could you remind me please how to access DQM plots from outside CERN?

[mariadalfonso]

@mbluj here the plots

[mbluj]

For me changes in fired trigger bits seem sensible; distributions describing trigger objects seem to agree between updated and reference, but I am not sure how to interpret the fact that numbers of entries are different - if it is an issue of DQM that different number of input events is compared or it is because of different filtering of trigger with modified trigger bits. Anyway, local tests were successful.
@lwezenbe @kandrosov could you cross check it and comment, please?

[kandrosov]

For tau trigger objects pre-selection is changed, so changes in size are expected. Why size has changed for jets - I don't know... @mariadalfonso - is there some requirement that one trigger object can belong to only one TrigObj nanoAOD type? E.g. if one miniAOD trigger objects pass pre-selection for two nanoAOD trig object types, would it be stored as two nanoAOD TrigObj?

We have introduced a new object type BoostedTau. If exclusivity is enforced, it might reduce the amount of Jet objects.

[mariadalfonso]

For tau trigger objects pre-selection is changed, so changes in size are expected.

indeed this done in
https://github.com/cms-sw/cmssw/pull/38472/files#diff-9b898944f08810913ea5c1605c47bd9a6c99eb91fc20f491ba2767702f4a0184R85

Why size has changed for jets - I don't know... @mariadalfonso - is there some requirement that one trigger object can belong to only one TrigObj nanoAOD type? E.g. if one miniAOD trigger objects pass pre-selection for two nanoAOD trig object types, would it be stored as two nanoAOD TrigObj?

The plot of size of jets seems unchanged, but the plots of the properties changes because we have have more match due to the 4 more bits added
https://github.com/cms-sw/cmssw/pull/38472/files#diff-9b898944f08810913ea5c1605c47bd9a6c99eb91fc20f491ba2767702f4a0184R149

We have introduced a new object type BoostedTau. If exclusivity is enforced, it might reduce the amount of Jet objects.

Let me know what you think.
is good from my side,

@mbluj can you prepare the 12_4 backport ?

[mbluj]

I think it is fine also from Tau side. @kandrosov please comment if it is not true.

Yes, I will prepare a backport to 12_4 soon (hopefully later today).

[kandrosov]

@mariadalfonso in the plot you've posted above, there are differences in _size distribution for TrigObj/Jet, or I'm misreading the plot?
Except that, all the rest looks fine to me.

[mariadalfonso]

The entries are always 9000.

The average size of "matched" increase 2.48 -> 2.72 but this is because of the 4 more bits

Then all the matched go in the other plots.

[mariadalfonso]

+xpog

updates to the trigger bit for both Run2 and Run3 nano.

[cmsbuild]

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

[perrotta]

+1