add run3_miniAOD_12X era modifier #42740
Conversation
|
@cms-sw/jetmet-pog-l2 @cms-sw/tau-pog-l2 is there anything needed from your side to make a consistent update of taus from 12_4/12_6 AOD to 13_0 MINIAOD? |
|
enable nano |
|
please test |
|
Enabling nano just to have a convenient way to check that the modifier is "harmless" when not active |
|
+code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-42740/36833
|
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-12984e/34670/summary.html Comparison SummarySummary:
NANO Comparison SummarySummary:
Nano size comparison Summary:
|
|
Few plots comparing jets from 126X AOD (estd.Jet_* variables) vs those reclustered using v17 in the PAT sequence activating the modifier introduced in this PR (Jet_* variables). The plots are made from NANOAODs produced with the same cmsDriver on top. All other objects present in the NANOAOD show no difference as expected. The plots are made form a 1000 events from the JetHT dataset (2022D). On average the reclustered jets have larger pt and a larger EM fraction. @nurfikri89 thank you again for pointing out the need for a modifier, can you have a look at the PR and the plots?
|
Thanks for making the PR @simonepigazzini. The plots look as expected and everything is in place to recompute puppi weights and recluster puppi jets and MET at PAT level. |
|
please test |
| @@ -508,6 +508,10 @@ | |||
| workflows[140.112] = ['',['RunCommissioning2022D','HLTDR3_2022','SKIMCOMMISSIONINGRUN3_reHLT_2022','HARVESTRUN3_2022']] | |||
| workflows[140.113] = ['',['RunCosmics2022D','HLTDR3_2022','SKIMCOSMICSRUN3_reHLT_2022','HARVESTRUN3_COS_2022']] | |||
|
|
|||
| ### run3 (2022) reMINIAOD+NANO ### | |||
| workflows[140.201] = ['',['RunJetMET2022D_reMINI', 'REMINIAOD_data2022']] | |||
There was a problem hiding this comment.
Why do we need only MINI workflow, and MINI+Nano+DQM?
There was a problem hiding this comment.
MINI + Validation + DQM:@miniAOD fails with an error of missing HGCal collections. I'm really puzzled by the error and do not have time to debug the DQM side at this stage. The MINI and NANO content are anyway as expected, therefore I think the issue with DQM can be dealt with later
There was a problem hiding this comment.
Yeah, I have that question too on DQM. Since you start from RECO, I am not sure DQM will work as you may miss some transient product which are produced during reconstruction. May I propose to drop the broken workflow for now, for example, you comment it. This will avoid broken in the long matrix test in IB, as workflow in the standard will run. Thx.
There was a problem hiding this comment.
as they are right now both tests run ok, or do you mean that the IB expects a DQM output?
There was a problem hiding this comment.
There was a problem hiding this comment.
Ah, OK. I read it quick and confuse on DQM module.
The DQM of Mini should fail as they need transient output from RECO, but DQM of Nano does not need.
What I don't see the point is why we need both workflows as both produce the same MINI.
There was a problem hiding this comment.
There was some concern (I just unjustified) that running MINI or MINI+NANO in the same job could lead to different results / issues. I agree with you that the MINI only workflow is not strictly needed. I will revise that with a follow up PR (we, xpog, want to revise some validation workflows anyway).
|
@cmsbuild please test workflow 140.201 |
|
-1 Failed Tests: RelVals-INPUT RelVals-INPUT
Comparison SummarySummary:
NANO Comparison SummarySummary:
Nano size comparison Summary:
|
|
please test let's give it another try, seems unrelated to me |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-12984e/34762/summary.html Comparison SummarySummary:
NANO Comparison SummarySummary:
Nano size comparison Summary:
|
|
+operations |
|
+pdmv |
|
+Upgrade The PR adds 2 workflows. One with MINI only, one with MINI+Nano+nano DQM. MiniAOD DQM is not possible as it needs transient products during RECO step. One unclear point to me is why do need both 140.201 and 140.202 as they produce the same mini. But I assume there is a reason behind this decision, which I may skip when reviewed. |
cmsbuild
added
fully-signed
upgrade-approved
and removed
pending-signatures
upgrade-pending
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) |
|
+1 |
PR description:
Add a modifier to reprocess 2022 data/MC from AOD to produce MINIAOD in 13X.
The new modifier is used to activate the PUPPI jets and MET reclustering in the PAT sequence
PR validation:
Run manually on 1000 events with and w/o the modifier. Differences (at the NANOAOD level, produced on top of the MINIAODs) are spotted for all concerned variables (Jets/MET) while other objects are unchanged. The differences are at the level expect. Plots to be added to the PR.
Backport expected for 13_0_X.
A RelVal workflow should be added for data and MC.