Switching off reconstruction of so called "NullTaus"

[mbluj]

Hi All,

it is a copy of #17706 rebased to the main branch (91X). Description of the original RP cited below for reference. The #17706 will be closed in a minute.

Best,
Michał

Hi all,

this is a simple configuration switch to remove the so called "null-taus" from the tau reconstruction. Null taus are jets seeding tau candidates, which actually do not fulfill any good tau reconstruction criterion. In the past these had been kept in the reconstruction for debugging purposes, since it allows a 1:1 correspondence of reconstructed tau candidates with seeding jets. In the meantime this should be switched to False when running in standard reconstruction. For more details refer to @mbluj.

NB:
we have split this change off from #17620 (90X) for clarity reasons. See also discussion in the conversation there.

Cheers,
Roger

[cmsbuild]

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

It involves the following packages:

RecoTauTag/RecoTau

@cmsbuild, @cvuosalo, @slava77, @davidlange6 can you please review it and eventually sign? Thanks.
@Muzaffar, @davidlange6, @smuzaffar you are the release manager for this.

cms-bot commands are listed here #13028

[slava77]

@cmsbuild please test

[cmsbuild]

The tests are being triggered in jenkins.
https://cmssdt.cern.ch/jenkins/job/ib-any-integration/18132/console Started: 2017/03/03 18:52

[cmsbuild]

+1
Tested at: b09eb1b
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-17767/18132/summary.html

The following merge commits were also included on top of IB + this PR after doing git cms-merge-topic:
6518b6b
You can see more details here:
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-17767/18132/git-log-recent-commits
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-17767/18132/git-merge-result

[cmsbuild]

Comparison job queued.

[cmsbuild]

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-17767/18132/summary.html

[slava77]

@davidlange6
why was this merged without a reco signature?

[davidlange6]

err- thats a good question. I think I clicked on the wrong PR. Shall i unmerge it?

[slava77]

@davidlange6
I'll check it today and will let you know if it should be reverted.

[slava77]

@mbluj @roger-wolf
in quite a few workflow comparisons there are differences in DQM like this (from wf 136.731, SinglePh2016B data)

Are these plots missing a basic tau quality preselection or did this update remove more than necessary?

[mbluj]

Yes, this plot is done without checking if a tau candidate passes basic quality criteria as presence of leading track. Differences are expected: it should be less tau candidates with low values of isolation sum.

[slava77]

On 3/14/17 9:38 AM, mbluj wrote: Yes, this plot is done without checking if a tau candidate passes basic quality criteria as presence of leading track. Differences are expected: it should be less tau candidates with low values of isolation sum.

something to put on the list to fix in DQM at some point?

…

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#17767 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEdcbr7hnRWlSq80vYmem93AGtbZy_Olks5rlsKCgaJpZM4MSS9f>.

[mbluj]

> Yes, this plot is done without checking if a tau candidate passes basic > quality criteria as presence of leading track. Differences are expected: > it should be less tau candidates with low values of isolation sum. something to put on the list to fix in DQM at some point?

Yes, right we should think on fixing DQM to not look on very poor taus. But, such poor taus will not be produced when this PR is merged, so DQM will not see them.

[slava77]

+1

for #17767 b09eb1b

• post-merge "+1"
• changes are as described, affecting tau reco configuration
• jenkins tests pass and comparisons with baseline show differences in tau objects
• local test with modified validate.C shows that differences in recoPFTaus_hpsPFTauProducer go away after a requirement that leadPFChargedHadrCandsignedSipt_ is non-NaN, the differences in discriminants go away after requiring the discriminant values (non-binary cases) are non-zero
• either jenkins or local tests show no differences in miniAOD tau objects, confirming that after basic selections there are no changes as expected.

[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 requires discussion in the ORP meeting before it's merged. @Muzaffar, @davidlange6, @smuzaffar