Contents:
If you want to get started fast with the analysis of fast-simulated events, you're at the right place.
We currently support two different approaches for fast simulation, Papas and Delphes. For now,
- FCC-ee users are encouraged to use Papas,
- FCC-hh and FCC-eh users should use Delphes.
However, ultimately, all users are encouraged to try both fast simulations and to compare the results.
An analysis ntuple will be produced with heppy, a simple modular event processing framework for high energy physics.
- To configure your environment for the FCC software, just do:
source /afs/cern.ch/exp/fcc/sw/0.8pre/setup.sh
You will need to source this script everytime you want to use the software.
- Then, install heppy
In this tutorial, you will learn how to:
- generate events with pythia8 and write them in the FCC EDM format.
- read these events with heppy to run the papas simulation and the analysis to create an ntuple
- read this ntuple with ROOT to make a few plots
But first, you will set up a working directory for your analysis.
Create a directory anywhere:
mkdir Workdir
cd Workdir
Get a pythia8 card file to generate ZH events. To download this file, on linux, do:
wget https://raw.githubusercontent.com/HEP-FCC/fcc-physics/master/pythia8/ee_ZH_Zmumu_Hbb.txt
On MacOS, do:
curl -O https://raw.githubusercontent.com/HEP-FCC/fcc-physics/master/pythia8/ee_ZH_Zmumu_Hbb.txt
Get the heppy configuration file for a ZH analysis:
cp $HEPPY/test/analysis_ee_ZH_cfg.py .
Here, we decide to use the standalone fcc-physics package to generate events instead of FCCSW. The advantage of the fcc-physics package is that it is supported for several operating systems, and in particular for mac os X and linux while FCCSW only works on lxplus. In other words, you could use fcc-physics to generate events on your notebook. That being said, it is of course possible to generate events with FCCSW.
Generate ee to ZH events with Z to mumu and H to b bbar:
fcc-pythia8-generate ee_ZH_Zmumu_Hbb.txt
You should obtain a file ee_ZH_Zmumu_Hbb.root written in the
FCC EDM format. Let us open it and check the contents:
root ee_ZH_Zmumu_Hbb.root
events->Print()
You're getting a list of the available collections. You can use root to draw the distribution of a variable. For example, the distribution of the charge of the stable generated muons:
events->Draw("GenParticle.core.charge", "abs(GenParticle.core.pdgId)==13 && GenParticle.core.status==1")
exit root:
.q
We are now going to run the papas simulation and to build an ntuple with meaningful variables for the analysis.
Run papas and the analysis in heppy
The ROOT file obtained in the previous section contains for each event all generated particles produced by pythia.
In this section we will run an heppy job to:
- run papas, a fast simulation that processes these generated particles to produce "reconstructed" particles;
- process the papas reconstructed particles and compute various analysis variables;
- store these variables in an ntuple written in an output root file.
The heppy configuration file analysis_ee_ZH_cfg.py describes these steps.
First, produce the display for the first event:
ipython -i analysis_ee_ZH_cfg.py 0
You should get an event display. Move to the next event, and print the event content:
next()
print loop.event
You should get a printout like:
Event: 3
{ 'gen_particles_stable': [ Particle :p1841 :17179871025: pdgid = -13, status = 1, q = 1 e = 55.1, theta = -1.14, phi = 0.59, mass = 0.11,
Particle :p1840 :17179871024: pdgid = 13, status = 1, q = -1 e = 50.2, theta = 0.56, phi = 2.35, mass = 0.11,
Particle :p1962 :17179871146: pdgid = 2112, status = 1, q = 0 e = 16.6, theta = 0.01, phi = -1.89, mass = 0.94,
Particle :p1849 :17179871033: pdgid = 321, status = 1, q = 1 e = 12.9, theta = 1.05, phi = -0.27, mass = 0.49,
Particle :p1961 :17179871145: pdgid = 22, status = 1, q = 0 e = 11.2, theta = 0.03, phi = -1.80, mass = 0.00,
Particle :p1900 :17179871084: pdgid = 22, status = 1, q = 0 e = 5.9, theta = 1.18, phi = -0.42, mass = 0.00,
Particle :p1892 :17179871076: pdgid = -211, status = 1, q = -1 e = 5.1, theta = -0.57, phi = -2.08, mass = 0.14,
Particle :p1867 :17179871051: pdgid = 211, status = 1, q = 1 e = 4.7, theta = -0.03, phi = -1.80, mass = 0.14,
Particle :p1925 :17179871109: pdgid = 22, status = 1, q = 0 e = 4.5, theta = 0.43, phi = 0.87, mass = 0.00,
Particle :p1935 :17179871119: pdgid = -2212, status = 1, q = -1 e = 4.4, theta = 0.01, phi = -1.70, mass = 0.94,
'...',
Particle :p1902 :17179871086: pdgid = 22, status = 1, q = 0 e = 0.1, theta = -0.28, phi = 2.41, mass = 0.00],
'higgses': [ Resonance2 : pdgid = 25, status = 3, q = 0 e = 124.3, theta = 0.65, phi = -0.94, mass = 116.25],
'higgses_legs': [ Jet : e = 66.8, theta = 0.87, phi = 0.63, mass = 39.62, tags=,
Jet : e = 57.5, theta = -0.28, phi = -1.73, mass = 26.31, tags=],
'rec_particles': [ Particle :p1969 :17179871153: pdgid = -13, status = 1, q = 1 e = 54.8, theta = -1.14, phi = 0.59, mass = 0.10,
Particle :p1971 :17179871155: pdgid = 13, status = 1, q = -1 e = 51.4, theta = 0.56, phi = 2.35, mass = 0.11,
Particle :r2445 :21474838925: pdgid = 211, status = 1, q = 1 e = 12.8, theta = 1.05, phi = -0.27, mass = 0.14,
Particle :r2417 :21474838897: pdgid = 22, status = 1, q = 0 e = 11.2, theta = 1.33, phi = 1.39, mass = -0.00,
Particle :r2442 :21474838922: pdgid = 130, status = 1, q = 0 e = 8.2, theta = 0.01, phi = -1.89, mass = 0.50,
Particle :r2398 :21474838878: pdgid = 22, status = 1, q = 0 e = 7.5, theta = 1.17, phi = -0.42, mass = 0.00,
Particle :r2403 :21474838883: pdgid = 22, status = 1, q = 0 e = 7.5, theta = 0.42, phi = 0.87, mass = -0.00,
Particle :r2414 :21474838894: pdgid = 130, status = 1, q = 0 e = 6.0, theta = -0.40, phi = -0.68, mass = 0.50,
Particle :r2440 :21474838920: pdgid = -211, status = 1, q = -1 e = 5.0, theta = -0.57, phi = -2.08, mass = 0.14,
Particle :r2416 :21474838896: pdgid = 22, status = 1, q = 0 e = 4.9, theta = -0.61, phi = -2.03, mass = -0.00,
'...',
Particle :r2422 :21474838902: pdgid = 22, status = 1, q = 0 e = 0.3, theta = 0.33, phi = 1.81, mass = 0.00],
'recoil': Particle : pdgid = 0, status = 1, q = 0 e = 133.8, theta = 0.46, phi = -1.30, mass = 123.97,
'zeds': [ Resonance2 : pdgid = 23, status = 3, q = 0 e = 106.2, theta = -0.46, phi = 1.84, mass = 93.44],
'zeds_legs': [ Particle :p1969 :17179871153: pdgid = -13, status = 1, q = 1 e = 54.8, theta = -1.14, phi = 0.59, mass = 0.10,
Particle :p1971 :17179871155: pdgid = 13, status = 1, q = -1 e = 51.4, theta = 0.56, phi = 2.35, mass = 0.11]}
Locate the zed. If you have none, call next() and
print loop.event until you do. Check its mass. Then, locate the
recoil (momentum of the particles recoiling against the zed), and check its mass.
Exit ipython:
quit()
Process the whole input file:
heppy_loop.py OutDir analysis_ee_ZH_cfg.py -N 1000
For your analysis, you will be able to use parallel processing, either on a multicore machine or on the CERN LSF cluster.
You get an output directory OutDir . Check its contents and the
contents of its subdirectories. In particular, the following root file
contains an ntuple with the variables we need:
OutDir/ee_ZH_Zmumu_Hbb/heppy.analyzers.examples.zh.ZHTreeProducer.ZHTreeProducer_1/tree.root
Open the root file containing the ntuple in root:
root OutDir/ee_ZH_Zmumu_Hbb/heppy.analyzers.examples.zh.ZHTreeProducer.ZHTreeProducer_1/tree.root
Make a few plots:
-
reconstructed Z mass:
events->Draw("zed_m") -
recoil mass:
events->Draw("recoil_m", "zed_m>80") -
dijet mass (note that jets have not been calibrated, using raw jets here):
events->Draw("higgs_m", "zed_m>80")
In this tutorial, you will learn how to:
- generate events with Pythia8, process them through Delphes with FCCSW and write them in the FCC EDM format.
- read these events with heppy to perfom basic selection and create an ntuple
- read this ntuple with ROOT to make a few plots
But first, you will set up a working directory for your analysis.
Start by getting the master branch of FCCSW{.twikiCurrentTopicLink .twikiAnchorLink} and proceed with installation. If not already in %FCCSW%:
cd $FCCSW
If %FCCSW% not initialized:
source ./init.sh
Now you are ready to produce 100TeV ttbar events with Pythia, process them through Delphes and store them in the FCC-EDM :
./run gaudirun.py Sim/SimDelphesInterface/options/PythiaDelphes_config.py
you should obtain a file called FCCDelphesOutput.root
Edit the ttbar example of heppy:
heppy/test/analysis_hh_ttbar_cfg.py
and link the file you produced running FCCSW previously
FCCDelphesOutput.root in the files list line 14. If
not the example will use files already produced and stored on eos, so
you will need to setup eos:
export EOS_MGM_URL="root://eospublic.cern.ch"
source /afs/cern.ch/project/eos/installation/client/etc/setup.sh
Now you are ready to run the ttbar example:
cd test
heppy_loop.py myoutput analysis_hh_ttbar_cfg.py
Look at the output files