Author: Alfredo Gurrola (Vanderbilt University), Anirban Saha (INFN), Stephan Lammel (FNAL)

export SCRAM_ARCH=slc6_amd64_gcc491 (or setenv SCRAM_ARCH slc6_amd64_gcc491)
cmsrel CMSSW_7_4_4
cd CMSSW_7_4_4/src
cmsenv
source /cvmfs/cms.cern.ch/crab3/crab.sh
git clone git@github.com:cms-sw/genproductions.git Configuration/GenProduction
cp Configuration/GenProduction/python/ThirteenTeV/Hadronizer_TuneCUETP8M1_13TeV_generic_LHE_pythia8_cff.py .
git clone https://github.com/gurrola/SampleProducer
mv SampleProducer/*.csh .
mv SamplerProducer/MadGraph5_v1.5.13.tar.gz .
mv SampleProducer/Configuration/GenProduction/python/ThirteenTeV/* Configuration/GenProduction/python/ThirteenTeV
rm -rf SampleProducer
tar -zxvf MadGraph5_v1.5.13.tar.gz
open the relevant .csh file and change the number of events to generate:
--no_exec -n 10 change to --no_exec -n 1
- above will set the number of events to generate to 1
scramv1 b
To run:

FullSim Zprime with PU (using Pythia8 as the generator)
A. ./script.csh
B. cmsRun STEP1_GEN-SIM.py > GenOutput.log
C. cmsRun STEP2.py
D. cmsRun STEP3.py
E. cmsRun STEP4.py

FastSim HeavyNu without PU (using Pythia6 as the generator)
A. ./fastsim_heavyNu_pythia6_noPU.csh
B. ./aodToMiniaod.csh

FastSim Zprime without PU (using Pythia8 as the generator)
A. ./fastsim_zprime.csh
B. ./aodToMiniaod.csh

FastSim using an input LHE file
A. ./fastsim_inputLHE.csh
B. ./aodToMiniaod.csh

Let's look at: Configuration/GenProduction/python/ThirteenTeV/ZprimeToTauTau_M_4500_TuneCUETP8M1_tauola_13TeV_pythia8_cfi.py

specifies the "generator":
generator = cms.EDFilter("Pythia8GeneratorFilter",
specifies the center-of-mass energy:
comEnergy = cms.double(13000.0),
dumps information for only 1 event:
maxEventsToPrint = cms.untracked.int32(1),
tau decays are handled by another package (tauola):
ExternalDecays = cms.PSet(Tauola = cms.untracked.PSet(TauolaPolar, TauolaDefaultInputCards ),
specifies the type of process (i.e. feynman diagram. In this case it's fermion+antifermion production a new boson)
'NewGaugeBoson:ffbar2gmZZprime = on',
the new gauge boson is a Z':
'Zprime:gmZmode = 3',
specifies the mass of the Z' (4500 GeV in this case):
'32:m0 = 4500',
force the Z' to decay to a pair of tau leptons
'32:onIfAny = 15',

Let's look at the cmsDriver commands for step #1:

uses the file "ZprimeToTauTau_M_4500_TuneCUETP8M1_tauola_13TeV_pythia8_cfi.py" as a template:
cmsDriver.py Configuration/GenProduction/python/ThirteenTeV/ZprimeToTauTau_M_4500_TuneCUETP8M1_tauola_13TeV_pythia8_cfi.py
name of the output file from step #1 (generation step):
--fileout GENSIM.root
specifies the detector conditions that will be used for the RAW+SIM step ... need to specify this is Monte Carlo:
--customise SLHCUpgradeSimulations/Configuration/postLS1Customs.customisePostLS1
--conditions auto:run2_mc
specify the magnetic field strength:
--magField 38T_PostLS1
produces a python file with all the above information that is ready for the user to run:
--python_filename STEP1_GEN-SIM.py
how many events to generate?:
--no_exec -n 10

Let's look at the cmsDriver commands for step #2:

specify input file from step #1
--filein file:GENSIM.root
name of the output root file from step #2 (detector simulation):
--fileout RAWSIM.root
energy from pileup is added "on top" of the "hard scatter" generation. This is taken from a separate file:
--pileup_input dbs:/MinBias_TuneA2MB_13TeV-pythia8/Fall13-POSTLS162_V1-v1/GEN-SIM
--pileup AVE_20_BX_25ns
*** you need a grid certificate for this part to work ... otherwise, remove the above two lines from the cmsDriver command

Let's look at the cmsDriver commands for step #3:

specify the "CMS sequences"/"algorithms" used to produce "objects" (electrons, muons, taus, jets, etc.):
--step RAW2DIGI,L1Reco,RECO,EI

Let's look at the cmsDriver commands for step #4:

PerilousApricot/SampleProducer