Tutorial to make Full Simulation at High Pileup Scenarios

This tutorial illustrates the simulation of neutrino samples with pileup 70 considering upgraded geometry for the pixel detector (4-layers, 3-disks).

Generation of events

We are going to simulate Neutrinos with Pt 2-20 according to this configuration fragment

ssh jruizvar@lxplus5.cern.ch
cmsrel CMSSW_6_1_2_SLHC6_patch1
cd CMSSW_6_1_2_SLHC6_patch1/src
source script.sh

script.sh requires Bash terminal.

Copy the following configure file (in CMSSW_6_1_2_SLHC6_patch1/src)

• Neutrino_Pt2to20_gun_cff_py_GEN_SIM.py

• crab.cfg

scram b -j 8
cmsRun Neutrino_Pt2to20_gun_cff_py_GEN_SIM.py

Check edmFileUtil Neutrino_Pt2to20_gun_cff_py_GEN_SIM.root. This file is only to test and can be removed ( rm Neutrino_Pt2to20_gun_cff_py_GEN_SIM.root)

Submit to crab

crab -create
crab -submit -c Neutrino_Pt2to20_14TeV

It takes around 15 min to generate 10000 events. 0.2M events can be generated in 20 jobs. After crab getoutput, don't forget

crab -publish -c Neutrino_Pt2to20_14TeV

Create Analyzer

We are going to create an standalone analyzer (no need to checkout external libraries) that can be used with any GEN-SIM sample, not only neutrinos.

ssh jruizvar@lxplus5.cern.ch
cmsrel CMSSW_6_1_2_SLHC6_patch1
cd CMSSW_6_1_2_SLHC6_patch1/src
source script.sh

Same script.sh as above. Make a directory with a dummy name, in this case neutrinoPU70. Create an analyzer respecting the name Pxecal

mkdir neutrinoPU70
mkedanlzr Pxecal
cd Pxecal

Copy to this folder the following files

• pxecal_cfg.py
• MinBiasSource.py
• src/Pxecal.cc
• BuildFile.xml
• crab.cfg

scram b -j 8
cmsRun pxecal_cfg.py
root -l Neutrino_PU70_ntuple.root
   [0] NtupleMaker->cd()
   [1] t->Draw("sqrt(ClGx*ClGx+ClGy*ClGy):ClGz")
   [2] .q
rm Neutrino_PU70_ntuple.root

If local run succeeds, you can create and submit the jobs

crab -create
crab -submit 1-500 -c Neutrino_PU70_100K_ntuple
crab -submit 501-1000 -c Neutrino_PU70_100K_ntuple

This step is more time consuming than the generation part. No need to publish because the output are n-tuples stored at T2_BR_SPRACE for private analysis.

Example plots

• Number of pixel cluster for low pileup (neutrino sample with 5000 events)
  

• Number of pixel cluster for high pileup (neutrino sample with 5000 events)
  

Exercise

Make the full simulation for a SingleElectronFlatPt5To50 for three pileup scenarios:

• PU35 (200K events)
• PU70 (100K events)
• PU140 (50K events)

• The GEN-SIM step should be done only once.
• The python configure files should be slightly modified.
• No need to touch the Pxecal.cc source file.

Simulated events for single electron

Full simulation with CMSSW_6_1_X using steps above. Samples are @ T2_BR_SPRACE.

New Simulated events for single electron

Files in the table bellow were produced with CMSSW_6_1_2_SLHC6_patch1 from these official GEN-SIM samples (50K events)

/SingleElectronFlatPt0p2To50/UpgFall13-POSTLS261_V2-v1/GEN-SIM

and using the official MinBias samples (30M events)

/MinBias_TuneZ2star_14TeV-pythia6/UpgFall13-POSTLS261_V2-v1/GEN-SIM

To open a merged file in lxplus , inside root mode, one can use

TFile *file = TFile::Open("dcap://osg-se.sprace.org.br:/pnfs/sprace.org.br/data/cms/store/user/adesouza/SingleElectronFlatPt0p2To50/MergedNtuples/SingleElectron_Pt2to50_noPU_ntuple.root");

Ntuples listed here are located at T2_SPRACE and contain information about both 5x5 and 1x1 EM crystals.

For rejection factor

Updated code available in GIT.

Raw tables with pixel and ecal information

To produce a raw table use the following

root -b SingleElectron_Pt2to50_noPU_ntuple.root
   [0] NtupleMaker->cd() 
   [1] t->Scan("bHitLayer:bHitGx:bHitGy:bHitGz:egEt:egGx:egGy:egGz","event<=10 && bHitLayer==1"); > singleElectron_withoutPileup.log

The resulting table is

* singleElectron_withoutPileup.log

Adding pileup to the event, the resulting table is

* singleElectron_35Pileup.log