Add mu beamspot constrained pT. Increase mu eta and phi precision to 16 bits #42646
Conversation
|
enable nano |
|
+code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-42646/36700
|
|
A new Pull Request was created by @namapane (Nicola Amapane) for master. It involves the following packages:
@cmsbuild, @simonepigazzini, @vlimant can you please review it and eventually sign? Thanks. cms-bot commands are listed here |
|
please test |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-a0bcfa/34453/summary.html Comparison SummarySummary:
NANO Comparison SummarySummary:
Nano size comparison Summary:
|
|
I wonder if |
cmsbuild
added
tests-pending
code-checks-pending
and removed
tests-approved
code-checks-approved
labels
|
+code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-42646/36710
|
|
Pull request #42646 was updated. @cmsbuild, @simonepigazzini, @vlimant can you please check and sign again. |
I have no strong opinion, but from a quick look I'd say that, in case, possible destinations could be RecoMuon/GlobalTrackingTools/plugins or RecoMuon/MuonIdentification/plugins. Both of these already include value map producers. Please let me know if I should move it.
Done. |
|
Pull request #42646 was updated. @cmsbuild, @simonepigazzini, @mandrenguyen, @clacaputo, @vlimant can you please check and sign again. |
|
please test |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-a0bcfa/34703/summary.html Comparison SummarySummary:
NANO Comparison SummarySummary:
Nano size comparison Summary:
|
|
spurious (unrelated) changes to eta/phi of muons, likely to be due to rounding effect |
|
+1 |
That's actually intended: as mentioned in the PR description, the precision of muon eta/phi is increased to 16 bits instead of 12 (code here). |
|
+1 |
cmsbuild
added
reconstruction-approved
fully-signed
and removed
reconstruction-pending
pending-signatures
labels
|
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. @rappoccio, @antoniovilela, @sextonkennedy (and backports should be raised in the release meeting by the corresponding L2) |
|
type muon |
|
+1 |
PR description:
Following discussions between H4l and Hmumu analysis teams, we would like to put forward a proposal to add the beamspot-constrained muon pT to nanoAOD v13.
Some background:
A BS constraint applied on top of PAT:Muons has been already adopted for the mH measurement in 4l (HIG-21-019) after discussion/blessing in the muon POG [talk1]. It was shown to significantly improve the mH resolution, especially since 2018 when the BS widths are of the order of ~6 microns, and was preferred for such an high profile analysis to ad-hoc parametrizations (eg GeoFit) as it is a natural step in reconstruction rather than an heuristic a-posteriori correction, that requires tuning. There are also ongoing studies for using this for Hmumu.
At the time of discussion in the muon POG, including this in standard sequences was not considered, as this constrain can be easily applied by the user while running on miniAODs.
However, it is not possible to apply the constraint on top of NanoAODs, and this limitation is becoming apparent now. The statement that for mH4l we may be forced to stay with a miniAOD framework for this reason sparked some discussion within the ATTF framework surveys [talk2 - see last page of the discussion document at the end of the agenda]; there were several suggestions to investigate possible ways to include the required information in nanoAODs. My reply at the time was that the implementation of the "VX+BS" approach used in the H4l analysis includes both a BS constraint and a common vertex constraint among the 4 selected muons; this makes it impossible to define a general-purpose constrained pT variable.
But actually, for 2018 and Run3, the BS width is so small that essentially all improvement comes from the BS constraint, not the common vertex constraint, so constraining each muon individually would be enough.
The present implementation is based on SingleTrackVertexConstraint. It essentially just adds 3 variables: the updated pt, ptError (precision=6) and chi2 (precision=6) We checked that the variation of eta and phi (and of their errors) are negligible so it would not be necessary to store these.
One more small update included is the increase of the precision of Muon_eta and Muon_phi from 12 bits to 16 (code here). This is because with the current precision we have a small rounding error on the reconstructed invariant masses, of about ~3.5 MeV (RMS) for H4mu [talk3]. Although this is a very small effect from the statistical point of view, authors feel it is somewhat unsatisfactory for the world's most precise single-channel measurement of mH to add a rounding smearing after they been fighting to reduce the systematic down to 17 MeV (HIG-21-019, expected, 4mu only), especially if that can be avoided by just adding as little as 8 bits per muon.
PR validation:
Both changes have been tested on 2018 (using a backport) and 2022 data/MC.