Everything begins here

Log in to lxplus.

ssh -Y [email protected] -o ServerAliveInterval=240
bash -l

First time

Setup the CMSSW release. The code nanoFakes.C fails with (at least) 10_2_0 and 10_6_4.

cd work
mkdir fakes

export SCRAM_ARCH=slc7_amd64_gcc630
cmsrel CMSSW_10_1_0
cd CMSSW_10_1_0/src
cmsenv

Original code.

git clone https://github.com/latinos/FakeRateMeasurement

Updated code.

git clone https://github.com/piedraj/FakeRateMeasurement

Get ready

cd work/fakes/CMSSW_10_1_0/src
cmsenv
cd FakeRateMeasurement

Check before job submission

Check (and edit if needed) the following files.

• nanoFakes.h contains the tight lepton names. They might differ between 2016, 2017, 2018.
• nanoFakes.C contains the triggers and the corresponding prescales. Edit this file if, for example, you need to move from Ele12 to Ele8.
• submitJobs.py contains the data and MC samples names. Verify that they match the current production.

Submit jobs

2022/07/13, open issue. The option -d below does not seem to fully work. It helps (at least for 2017 and 2018) choosing between data and MC, but the full path is read from runNanoFakes.C.

2016_HIPM

python submitJobs.py -d /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Summer20UL16_106x_nAODv9_HIPM_Full2016v9/MCl1loose2016v9__fakeSelKinMC/ -y 2016_HIPM
python submitJobs.py -d /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Run2016_UL2016_nAODv9_HIPM_Full2016v9/DATAl1loose2016v9__fakeSel/ -y 2016_HIPM

Read trees without any mtw1 cut to debug the 2016_HIPM negative fakes.

python submitJobs.py -d /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Summer20UL16_106x_nAODv9_HIPM_Full2016v9/MCl1loose2016v9__MCCorr2016v9/ -y 2016_HIPM
python submitJobs.py -d /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Run2016_UL2016_nAODv9_HIPM_Full2016v9/DATAl1loose2016v9/ -y 2016_HIPM

2016_noHIPM

python submitJobs.py -d /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Summer20UL16_106x_nAODv9_noHIPM_Full2016v9/MCl1loose2016v9__fakeSelKinMC/ -y 2016_noHIPM
python submitJobs.py -d /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Run2016_UL2016_nAODv9_noHIPM_Full2016v9/DATAl1loose2016v9__fakeSel/ -y 2016_noHIPM

2017

python submitJobs.py -d /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Summer20UL17_106x_nAODv9_Full2017v9/MCl1loose2017v9__fakeSelKinMC/ -y 2017
python submitJobs.py -d /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Run2017_UL2017_nAODv9_Full2017v9/DATAl1loose2017v9__fakeSel/ -y 2017

2018

python submitJobs.py -d /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Summer20UL18_106x_nAODv9_Full2018v9/MCl1loose2018v9__fakeSelKinMC/ -y 2018
python submitJobs.py -d /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Run2018_UL2018_nAODv9_Full2018v9/DATAl1loose2018v9__fakeSel -y 2018

Babysit jobs

condor_q
condor_q -hold -af HoldReason

Wrap it up

cd results

hadd -f -k hadd_wjets.root nanoLatino_WJetsToLNu*.root
hadd -f -k hadd_zjets.root nanoLatino_DYJetsToLL*.root
hadd -f -k hadd_data.root  nanoLatino_*_Run201*.root

For 2018 data there are too many files, and the hadd has to be done in two steps.

hadd -f -k hadd_EGamma_Run2018.root nanoLatino_EGamma_Run2018*.root
hadd -f -k hadd_DoubleMuon_Run2018.root nanoLatino_DoubleMuon_Run2018*.root

hadd -f -k hadd_data.root hadd_EGamma_Run2018.root hadd_DoubleMuon_Run2018.root

rm hadd_EGamma_Run2018.root
rm hadd_DoubleMuon_Run2018.root

Move the merged files to their corresponding year directory.

rm nanoLatino*
mkdir <year>
mv hadd* <year>/.

Extract the fake and prompt rates

root -l -b -q getFakeRate.C

Share on the web

cp -r png /afs/cern.ch/user/p/piedra/www/ul_fakerate

pushd /afs/cern.ch/user/p/piedra/www/ul_fakerate
cp ../index.php .
find . -type d -exec cp index.php {} \;
popd

And the results should appear here,

https://piedra.web.cern.ch/piedra/ul_fakerate

Some relevant physics

• The jet pt thresholds for electrons are 35 GeV for 0-jet, 1-jet and 2-jets.
• The jet pt thresholds for muons are 20 GeV for 0-jet, 25 GeV for 1-jet, and 35 GeV for 2-jets.