FourTopNAOD

UC Riverside Four Top Analysis in NanoAOD format (2016 validation, 2017 and 2018 data analysis)

RDataFrame Skimming code

Initial setup (do once from an area where you have access to /cvmfs)

export SCRAM_ARCH=slc7_amd64_gcc830
cmsrel CMSSW_10_2_26
cd CMSSW_10_2_26/src
cmsenv
git clone https://github.com/cms-nanoAOD/nanoAOD-tools.git PhysicsTools/NanoAODTools
git clone https://github.com/NJManganelli/FourTopNAOD.git FourTopNAOD
scram b -j4
source FourTopNAOD/Kai/standalone/env_standalone.sh build
source FourTopNAOD/RDF/standalone/env_standalone.sh build

RDF Usage (do every time when using anything in the FourTopNAOD/RDF directory; recommend a .bashrc function)

source /cvmfs/cms.cern.ch/cmsset_default.sh
export SCRAM_ARCH=slc7_amd64_gcc830
source /cvmfs/sft.cern.ch/lcg/views/LCG_99/x86_64-centos7-gcc8-opt/setup.sh
#Get authentication tokens where necessary, e.g. voms and kerberos
voms-proxy-init -voms cms --valid 192:00
kinit
cd CMSSW_10_2_26/src/FourTopNAOD/RDF
source standalone/env_standalone.sh
cd skimming
#Run a script that uses the FourTopNAOD/RDF/scripts/nanoframe.py skimming code.
python ../scripts/nanoframe.py --help
sh skimming_example.sh

PostProcessor code

export SCRAM_ARCH=slc7_amd64_gcc700
cmsrel CMSSW_10_2_18
cd CMSSW_10_2_18/src
cmsenv
git clone https://github.com/cms-nanoAOD/nanoAOD-tools.git PhysicsTools/NanoAODTools
git clone https://github.com/NJManganelli/FourTopNAOD.git FourTopNAOD
scram b -j4

Getting files with dasgoclient

voms-proxy-init -voms cms
dasgoclient --query="file dataset=/TTToSemiLepton_HT500Njet9_TuneCP5_PSweights_13TeV-powheg-pythia8/RunIIFall17NanoAODv5-PU2017_12Apr2018_Nano1June2019_102X_mc2017_realistic_v7-v1/NANOAODSIM"

Choose a file from those printed, add an xrootd redirector such as root://cms-xrd-global.cern.ch/

xrdcp root://cms-xrd-global.cern.ch//store/mc/RunIIFall17NanoAODv5/TTToSemiLepton_HT500Njet9_TuneCP5_PSweights_13TeV-powheg-pythia8/NANOAODSIM/PU2017_12Apr2018_Nano1June2019_102X_mc2017_realistic_v7-v1/120000/0E27F419-ADDE-1E4C-AC0A-130DA36C1FA6.root testfile.root

Running Lepton Scale Factor insertion code

cd FourTopNAOD/Kai
mkdir test
cd test
python ../scripts/LepSFProcessor.py --help

For example, we can add the Loose ID SF's for muons and electrons, using the LooseRelIso/LooseID Isolation SF's, with the following ommand line code. If path/to/inputfile_*.root is a globbable path to multipe inputs (or a single input), and output directory 'somedirectory' exists

python ../scripts/LepSFProcessor.py 'path/to/testfile.root' --outDirectory . --muon_ID LooseID --muon_ISO LooseRelIso/LooseID --electron_ID LooseID --era 2017 --maxEntries 2500

Running Stitching code for flagging

cd FourTopNAOD/Kai
mkdir test
cd test
python ../scripts/StitchingProcessor.py --help

Run over testfile with dilepton nominal and single lepton filtered setting

python ../scripts/StitchFlagProcessor.py testfile.root --outDirectory . --era 2017 --channel DL --mode Flag --maxEntries 3000 --source Nominal --fillHists --weight 0.27
python ../scripts/StitchFlagProcessor.py testfile.root --outDirectory . --era 2017 --channel SL --mode Flag --maxEntries 3000 --source Filtered --fillHists --weight 0.15

New instructions

Setting up NanoAODv5 Production

export SCRAM_ARCH=slc7_amd64_gcc700
mkdir NanoProduction && cd NanoProduction
cmsrel CMSSW_10_2_15
cd CMSSW_10_2_15/src
cmsenv
git cms-init
git cms-merge-topic -u pastika:AddAxis1_1026
git clone https://github.com/susy2015/TopTagger.git
git clone https://github.com/cms-nanoAOD/nanoAOD-tools.git PhysicsTools/NanoAODTools
git clone https://github.com/NJManganelli/FourTopNAOD.git FourTopNAOD
scram b -j4
mkdir -p ${CMSSW_BASE}/src/TopTagger/TopTagger/data
getTaggerCfg.sh -o -n -t DeepResolved_DeepCSV_GR_noDisc_Release_v1.0.0 -d $CMSSW_BASE/src/TopTagger/TopTagger/data

Attempt to fix resolvedTagger jet source

in TopTagger/TopTagger/python/resolvedTagger_cff.py, replace instances of slimmedJets with upatedJets, like so:

inputJetCollection = process.slimmedJetsWithUserData.src

with

inputJetCollection = process.updatedJetsWithUserData.src

cmsDriver commands for NanoAOD Production

2017:

cmsDriver.py myNanoProdMc2017 -s NANO --mc --eventcontent NANOAODSIM --datatier NANOAODSIM  --no_exec  --conditions 102X_mc2017_realistic_v7 --era Run2_2017,run2_nanoAOD_94XMiniAODv2 --customise_commands="process.add_(cms.Service('InitRootHandlers', EnableIMT = cms.untracked.bool(False)))"

cmsDriver.py myNanoProdData2017 -s NANO --data --eventcontent NANOAOD --datatier NANOAOD  --no_exec  --conditions 102X_dataRun2_v11 --era Run2_2017,run2_nanoAOD_94XMiniAODv2 --customise_commands="process.add_(cms.Service('InitRootHandlers', EnableIMT = cms.untracked.bool(False)))"

Customizing with Resolved TopTagger (WIP)

Add the following lines in the custoization section of the cmsRun configuration produced by cmsDriver:

from TopTagger.TopTagger.resolvedTagger_cff import customizeResolvedTagger, customizeResolvedTaggerAllCanidiates, customizeResolvedTaggerAllCanidiates, customizeResolvedTaggerAndVariables, customizeResolvedTaggerAllCanidiatesAndVariables

#call to customisation function customizeResolvedTagger imported from TopTagger.TopTagger.resolvedTagger_cff
process = customizeResolvedTagger(process)
#process = customizeResolvedTaggerAllCanidiates(process)

Old Notes and Instructions

Please read here: https://twiki.cern.ch/twiki/bin/view/CMSPublic/WorkBookNanoAOD#How_to_check_out_the_code_and_pr cmsDriver.py can be used to create barebones configurations, as in documentation above, for producing NanoAOD from Mini

Centrally Produced NanoAOD Campaigns: 80X MC + data: https://cmsweb.cern.ch/das/request?view=list&limit=50&instance=prod%2Fglobal&input=%2F*%2F*05Feb2018*%2FNANOAOD* 80X Run2016 data: https://cmsweb.cern.ch/das/request?view=list&limit=50&instance=prod%2Fglobal&input=%2F*%2F*05Feb2018*%2FNANOAOD 94X MC: https://cmsweb.cern.ch/das/request?view=list&limit=50&instance=prod%2Fglobal&input=%2F*%2F*12Apr2018*%2FNANOAODSIM 94X data: https://cmsweb.cern.ch/das/request?view=list&limit=50&instance=prod%2Fglobal&input=%2F*%2F*31Mar2018*%2FNANOAOD 94X data not necessarily the most recently processed!

Note about GitHub source code: There are two collections of NanoAOD source code; one is maintained within cms-sw, and the other within cms-nanoAOD. The latter is the collection of tools for processing NanoAOD data in a postprocessor framework. https://github.com/cms-sw/cmssw/tree/master/PhysicsTools/NanoAOD https://github.com/cms-nanoAOD/nanoAOD-tools.git

Root/Python or Mixed Toolchain: (DO ONCE:) (Intall Python 2.7 and recent ROOT release OR CMSSW Release) (Case: Standalone) git clone https://github.com/cms-nanoAOD/nanoAOD-tools.git NanoAODTools #The postprocessing tools! git clone https://github.com/NJManganelli/FourTopNAOD.git FourTopNAOD #this code cd NanoAODTools bash standalone/env_standalone.sh build source standalone/env_standalone.sh (Case: CMSSW) source /cvmfs/cms.cern.ch/cmsset_default.sh #LXPLUS source $VO_CMS_SW_DIR/cmsset_default.sh #T2_BE MX Machines cmsrel CMSSW_X_Y_Z (9_4_9 tested for 2017 data) cd CMSSW_X_Y_Z/src cmsenv git clone https://github.com/cms-nanoAOD/nanoAOD-tools.git NanoAODTools #The postprocessing tools! git clone https://github.com/NJManganelli/FourTopNAOD.git FourTopNAOD #this code scram b -j10 (or package instructions!)

(DO EVERY NEW SESSION:) (Case: Standalone) cd NanoAODTools source standalone/env-standalone.sh (Case: CMSSW) cd CMSSW_X_Y_Z/src cmsenv

####################################

Using Python Notebooks

#################################### A python notebook can be started by opening another terminal, and executing the command: ssh -L localhost:4444:localhost:4444 @lxplus.cern.ch Establishing cmsenv (new session) as above Getting grid proxy (if accessing root files stored on servers, using CRAB, etc): voms-proxy-init -voms cms Navigate to: /path/to/notebook/directory execute: jupyter-notebook --no-browser --port=4444 --ip=127.0.0.1 When it produces a token, paste in your webbrowser to connect to homepage, and open one of the notebooks or start a new one

########################

Stable Tag

######################## A stable tag of this release can be grabbed by, following the clone of this repository, checking out a stable tag and simultaneously creating a branch

git checkout tags/ -b branchedFrom

Tag-Versions: v0.2 - demonstrates accessing collections, triggers, MET filters, etc. in basicAccessAndLoop.ipynb and dataTrigAndJson.ipynb

vicha-w / fourtopnaod Goto Github PK

fourtopnaod's Introduction