Puppi MET and MET significance fixes in the MET tool #15565

mmarionncern · 2016-08-23T13:01:20Z

This PR fixes several bugs and inconsistencies in the PAT MET tool :

use of proper JECs for the Puppi MET
fix the MET significance storage when miniAOD are produced
update the MET significance tune
modify some MET constructors (RECO and PAT) to keep the met significance matrix when met objects is copied.

Changes expected :

Puppi MET and its uncertainties
MET significance

The puppi MET results will have sense when #15562 will be integrated.

Packages impacted:

DataFormats/METReco
DataFormats/PatCandidates
PhysicsTools/PatAlgos
PhysicsTools/PatUtils
RecoMET/METProducers

Local tests and matrix tests seems successfull.

I will be out of office from August 27th to September 17th. If comments comes during that period, please contact @mariadalfonso and @zdemirag .

cmsbuild · 2016-08-23T13:02:33Z

A new Pull Request was created by @mmarionncern for CMSSW_8_1_X.

It involves the following packages:

PhysicsTools/PatAlgos
PhysicsTools/PatUtils
RecoMET/METProducers

@cmsbuild, @cvuosalo, @slava77, @montjj, @davidlange6 can you please review it and eventually sign? Thanks.
@TaiSakuma, @imarches, @ahinzmann, @acaudron, @gpetruc, @rappoccio, @jdolen, @nhanvtran, @JyothsnaKomaragiri, @schoef, @ferencek, @mverzett, @mariadalfonso, @pvmulder this is something you requested to watch as well.
@slava77, @smuzaffar you are the release manager for this.

cms-bot commands are list here #13028

slava77 · 2016-08-23T13:25:41Z

PhysicsTools/PatAlgos/python/slimming/puppiForMET_cff.py

+                                         useRefs = cms.bool(True),
+                                         useValueMap = cms.bool(False),
+                                         weight = cms.double(1.0),
+                                         weightsName = cms.InputTag("puppiNoLep")


it would be easier to follow what has changed if parameters were not reshuffled and reindented

slava77 · 2016-08-23T13:28:53Z

@mmarionncern
please provide a link to a presentation in a POG meeting where these changes are described.

Are you planning a backport to 80X?

slava77 · 2016-08-23T13:29:00Z

@cmsbuild please test

cmsbuild · 2016-08-23T13:29:12Z

The tests are being triggered in jenkins.
https://cmssdt.cern.ch/jenkins/job/ib-any-integration/14673/console

cmsbuild · 2016-08-23T15:18:43Z

-1

Tested at: 6b2c1be

You can see the results of the tests here:
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-15565/14673/summary.html

I found follow errors while testing this PR

Failed tests: RelVals

RelVals:

When I ran the RelVals I found an error in the following worklfows:
50202.0 step3

runTheMatrix-results/50202.0_TTbar_13+TTbar_13+DIGIUP15_PU50+RECOUP15_PU50+HARVESTUP15_PU50/step3_TTbar_13+TTbar_13+DIGIUP15_PU50+RECOUP15_PU50+HARVESTUP15_PU50.log

slava77 · 2016-08-23T15:29:18Z

50202 is a 50ns wflow

Exception Message:
No data of type "JME::JetResolutionObject" with label "AK4PFPuppi_pt" in record "JetResolutionRcd"

looks like some condition payloads are missing in the 50ns run2 GT.

mariadalfonso · 2016-08-23T17:39:39Z

@blinkseb can you help to understand why the AK4PFPuppi_pt is not filled ?
I could access the AK4PFPuppi_phi and the AK4PFPuppi_pt from the Spring16_25nsV6_DATA.db for the plot in the JME-16-004

mariadalfonso · 2016-08-23T17:45:13Z

@slava77 these fixes went in the plot for the JME-16-004.
We must indeed backport into the 80X for the re-reco (as soon we are sure we can integrate this one)

cmsbuild · 2016-08-24T08:06:08Z

Pull request #15565 was updated. @cmsbuild, @cvuosalo, @slava77, @montjj, @davidlange6 can you please check and sign again.

cmsbuild · 2016-08-24T09:28:18Z

Pull request #15565 was updated. @cmsbuild, @cvuosalo, @slava77, @montjj, @davidlange6 can you please check and sign again.

mmarionncern · 2016-08-24T09:30:32Z

Two extra commits touching the data formats have been added to keep the MET significance matrix when the MET is copied (both RECO and PAT level)
Description is updated

mmusich · 2016-08-25T21:27:31Z

@mariadalfonso missing conditions are in the 50ns Run2 MC Global Tag. The Spring16_25nsV6 set of payloads has not been propagated to 50ns (which is not 25ns ...). I don't think it makes too much sense change the 50ns GT using the latest 25ns JECs. Maybe just adding the labels for the resolutions: what do you think?
Side note: testing the PR with runTheMatrix.py -l limited before submitting is always a good practice.

slava77 · 2016-08-26T19:39:00Z

@mmarionncern @mariadalfonso @mmusich
has there been progress in getting the AK4PFPuppi_pt payolads in all the right places?

… and corrMET as input (constructor used by CorrectedPATMETProducer)

cmsbuild · 2016-09-05T19:48:25Z

Pull request #15565 was updated. @cmsbuild, @cvuosalo, @slava77, @monttj, @davidlange6 can you please check and sign again.

mariadalfonso · 2016-09-05T19:49:38Z

@slava77 rebased interactively picking the right veelken commit,

slava77 · 2016-09-05T20:22:55Z

@cmsbuild please test

@mariadalfonso it looks like we are almost there.
I will check this at run time now.

cmsbuild · 2016-09-05T20:23:20Z

The tests are being triggered in jenkins.
https://cmssdt.cern.ch/jenkins/job/ib-any-integration/14952/console

cmsbuild · 2016-09-05T20:59:15Z

-1

Tested at: 7f89d67

You can see the results of the tests here:
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-15565/14950/summary.html

I found follow errors while testing this PR

Failed tests: RelVals

RelVals:

When I ran the RelVals I found an error in the following worklfows:
4.22 step1

DAS Error

slava77 · 2016-09-05T22:03:19Z

@cmsbuild please test
retry, hoping to get rid of the DAS error

cmsbuild · 2016-09-05T22:03:39Z

The tests are being triggered in jenkins.
https://cmssdt.cern.ch/jenkins/job/ib-any-integration/14957/console

cmsbuild · 2016-09-05T23:47:41Z

+1
Tested at: b742cc1
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-15565/14952/summary.html

cmsbuild · 2016-09-06T00:29:02Z

+1
Tested at: b742cc1
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-15565/14957/summary.html

cmsbuild · 2016-09-06T01:09:49Z

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-15565/14957/summary.html

cmsbuild · 2016-09-06T01:38:13Z

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-15565/14952/summary.html

slava77 · 2016-09-06T04:43:29Z

+1

for #15565 b742cc1

code changes are in line with the description and the follow up review. Main changes are to appropriately select jet correctors in Puppi corrected MET and also to set MET significance (with updated parameters to compute the significance).
jenkins tests pass and comparisons with baseline show differences in miniAOD METs
local tests confirm the changes observed in jenkins tests: there are minor changes in slimmedMETsPuppi p3 (due to updates to T1 correctors), and also the significance variable in the MET product is now filled (instead of -1) in slimmedMETs and slimmedMETsPuppi
- net CPU use is roughly unchanged
- updates in miniAOD show ~0.1% increase in file size (based on 70 events in ttbar PU35, which is probably an overestimate, given the small number of events tested)

cmsbuild added this to the Next CMSSW_8_1_X milestone Aug 23, 2016

cmsbuild added reconstruction-pending analysis-pending pending-signatures tests-pending orp-pending comparison-pending labels Aug 23, 2016

slava77 reviewed Aug 23, 2016
View reviewed changes

cmsbuild added tests-started and removed tests-pending labels Aug 23, 2016

cmsbuild added tests-rejected and removed tests-started labels Aug 23, 2016

cmsbuild added tests-pending and removed tests-rejected labels Aug 24, 2016

mmarionncern mentioned this pull request Aug 25, 2016

Puppi MET and MET significance fixes in the MET tool (80X) #15600

Merged

davidlange6 mentioned this pull request Aug 26, 2016

possible 80X backports (terminated on 8_0_20) slava77/cmssw#93

Closed

mariadalfonso and others added 2 commits September 5, 2016 21:45

revert particleFlowDisplacedVertex for Puppi per slava request

2a1978e

set MET significance matrix elements in constructor that takes srcMET…

b742cc1

… and corrMET as input (constructor used by CorrectedPATMETProducer)

mariadalfonso force-pushed the MM_PuppiFix810_100816 branch from 290abad to b742cc1 Compare September 5, 2016 19:47

cmsbuild added tests-started and removed tests-pending labels Sep 5, 2016

cmsbuild added tests-pending tests-started and removed tests-started tests-pending labels Sep 5, 2016

cmsbuild added tests-approved and removed tests-started labels Sep 5, 2016

cmsbuild added comparison-available and removed comparison-pending labels Sep 6, 2016

cmsbuild added reconstruction-approved and removed reconstruction-pending labels Sep 6, 2016

davidlange6 merged commit ed88e5a into cms-sw:CMSSW_8_1_X Sep 6, 2016

This was referenced Sep 6, 2016

Update to fix late validation plots from PR 15562 #15669

Merged

Fix posftix and scheduled mode issue in the runMETUncertainty tool (Copy of the original #14606) #15616

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Puppi MET and MET significance fixes in the MET tool #15565

Puppi MET and MET significance fixes in the MET tool #15565

mmarionncern commented Aug 23, 2016 •

edited

Loading

cmsbuild commented Aug 23, 2016

slava77 Aug 23, 2016

slava77 commented Aug 23, 2016

slava77 commented Aug 23, 2016

cmsbuild commented Aug 23, 2016 •

edited

Loading

cmsbuild commented Aug 23, 2016

slava77 commented Aug 23, 2016

mariadalfonso commented Aug 23, 2016

mariadalfonso commented Aug 23, 2016

cmsbuild commented Aug 24, 2016

cmsbuild commented Aug 24, 2016

mmarionncern commented Aug 24, 2016

mmusich commented Aug 25, 2016

slava77 commented Aug 26, 2016

cmsbuild commented Sep 5, 2016

mariadalfonso commented Sep 5, 2016

slava77 commented Sep 5, 2016

cmsbuild commented Sep 5, 2016 •

edited

Loading

cmsbuild commented Sep 5, 2016

slava77 commented Sep 5, 2016

cmsbuild commented Sep 5, 2016 •

edited

Loading

cmsbuild commented Sep 5, 2016

cmsbuild commented Sep 6, 2016

cmsbuild commented Sep 6, 2016

cmsbuild commented Sep 6, 2016

slava77 commented Sep 6, 2016

Puppi MET and MET significance fixes in the MET tool #15565

Puppi MET and MET significance fixes in the MET tool #15565

Conversation

mmarionncern commented Aug 23, 2016 • edited Loading

cmsbuild commented Aug 23, 2016

slava77 Aug 23, 2016

Choose a reason for hiding this comment

slava77 commented Aug 23, 2016

slava77 commented Aug 23, 2016

cmsbuild commented Aug 23, 2016 • edited Loading

cmsbuild commented Aug 23, 2016

slava77 commented Aug 23, 2016

mariadalfonso commented Aug 23, 2016

mariadalfonso commented Aug 23, 2016

cmsbuild commented Aug 24, 2016

cmsbuild commented Aug 24, 2016

mmarionncern commented Aug 24, 2016

mmusich commented Aug 25, 2016

slava77 commented Aug 26, 2016

cmsbuild commented Sep 5, 2016

mariadalfonso commented Sep 5, 2016

slava77 commented Sep 5, 2016

cmsbuild commented Sep 5, 2016 • edited Loading

cmsbuild commented Sep 5, 2016

slava77 commented Sep 5, 2016

cmsbuild commented Sep 5, 2016 • edited Loading

cmsbuild commented Sep 5, 2016

cmsbuild commented Sep 6, 2016

cmsbuild commented Sep 6, 2016

cmsbuild commented Sep 6, 2016

slava77 commented Sep 6, 2016

mmarionncern commented Aug 23, 2016 •

edited

Loading

cmsbuild commented Aug 23, 2016 •

edited

Loading

cmsbuild commented Sep 5, 2016 •

edited

Loading

cmsbuild commented Sep 5, 2016 •

edited

Loading