Pile-UP studies
Notes
- Pileup Jet Identification - 2013. (1)
- Jet energy scale and resolution in the CMS experiment - 2017 (2)
- Performance of pileup jet identification in the 2016 run (3)
Jet Analysys wiki pages
Code
- For studies described in (3)
3. Basic EDAnalyzers for making flat ROOT trees from MC and associated configuration files. They were exercised in 2017 MC in CMSSW_9_4_1. This piece of code needs to go under some project area (in my case it was src/KKousour/JMEAnalysis/) and in the corresponding /plugins directory to be compiled with scram. /afs/cern.ch/user/k/kkousour/public/pileupJetId/ After you manage to make your own flat ROOT trees you need to write some script to make ROOT trees out of jets (rather than events), which should then be given to TMVA for the training. If you don’t manage, I will be able to provide you this code too. Please let me know if you have any questions. Regards Kostas (Konstantinos Kousouris <Konstantinos.Kousouris@cern.ch>)
Build at cms0
- cd /cdata; rsync -av serguei@lxplus.cern.ch:/afs/cern.ch/user/k/kkousour/public/pileupJetId ./;
- mkdir /cdata/pileUPcostas; cd pileUPcostas
- cmset; cmsrel CMSSW_9_4_9; cd CMSSW_9_4_9/
- cd src/; cmsenv; mkdir SUpileup; cd SUpileup;
- mkedanlzer PileUpAnalyser; cd PileUpAnalyser/plugins;
- cp -rfvp /cdata/serguei/pileupJetId/* ./
- Edit EventCounter.cc and JMEFlatTreeProducer.cc; replace "#include "KKousour/JMEAnalysis.." with "#include "SUpileup/PileUpAnalyzer.."
- create /CMSSW_9_4_9/src/SUpileup/PileUpAnalyzer/plugins/BuildFile.xml
- [serguei@pcms0 PileUpAnalyzer]\$ scram b -j 12
Samples
From: Konstantinos Kousouris <Konstantinos.Kousouris@cern.ch> Sent: Thursday, November 1, 2018 11:53 AM To: Pushpalatha C Bhat Cc: Konstantinos Kousouris; Harrison B Prosper Subject: Re: Info for pile-up studies
Dear Pushpa,
Below is the information that you need.
1. Lates presentation of our work at the JMAR meeting
2. AN note that describes the techniques (exercised in 2016 data & MC)
http://cms.cern.ch/iCMS/jsp/openfile.jsp?tp=draft&files=AN2017_150_v2.pdf
3. Basic EDAnalyzers for making flat ROOT trees from MC and associated configuration files. They were exercised in 2017 MC in CMSSW_9_4_1. This piece of code needs to go under some project area (in my case it was src/KKousour/JMEAnalysis/) and in the corresponding /plugins directory to be compiled with scram.
/afs/cern.ch/user/k/kkousour/public/pileupJetId/
After you manage to make your own flat ROOT trees you need to write some script to make ROOT trees out of jets (rather than events), which should then be given to TMVA for the training. If you don’t manage, I will be able to provide you this code too.
Please let me know if you have any questions.
Regards Kostas
> On 31 Oct 2018, at 18:53, Pushpalatha C Bhat <pushpa@fnal.gov> wrote: > > Dear Kostas, > > A gentle reminder to send us the info and code to get us started with the pile-up studies! > > Many thanks, > Pushpa >
Hi Sergei, Mike, Vishnu, Ramanpreet,
It'd be great if you can look at the JME notes Andreas cites. We can follow his suggestions along with those by Kostas and produce some plots and results on JEC and resolution ASAP. Hope we can discuss this at the Wednesday meeting.
Mike, if you are at the Lab today or tomorrow, please stop by.
Pushpa
From: Andreas Hinzmann <andreas.hinzmann@cern.ch> Sent: Monday, November 5, 2018 3:36 AM To: Pushpalatha C Bhat Cc: Ramanpreet Singh; Vishnu Zutshi; Mike Eads; Zeynep Demiragli Subject: Re: Pile-up studies for forward jets
Dear Pusha,
(adding also Zeynep)
as a first step I would suggest to quantify the jet energy resolution and pileup jet rate in data and simulation and then see how it depends on the depth profile. Since HCAL is known to be rather mis-calibrated in 2018 w.r.t. simulation in the endcap, looking at both data and simulation will be crucial. The jet energy resolution and pileup jet rate can be looked at with a Z+jet selection. An example for studying jet energy resolution using gamma+jets can be found in JME-13-004. Z+jet with a similar approach should allow to go lower in pT and also into the forward region of the detector. An example for studying pileup jet rate with such a selection can be found here in JME-13-005.
Best, Andreas
@]