python.doc

---

`Python' Analysis Mode

---

Python is an interpreted, interactive, object-oriented programming language. It is often compared to Tcl, Perl, Scheme or Java.

The file example.py shows an example of a Python program which can be processed interactively as: "python -i example.py". If you only need to execute the script and exit (⇒ no graphics, no extra Python commands), you may just type: "example.py".

You can find simple introductions to Python in: http://www.python.org/doc/Intros.html . Python is naturally available in ROOT via PyROOT .

---

The Python mode has been tested to work only with ROOT versions ≥ 4.01

---

Interpreted analyses are well suited for small or medium size data samples, for which computer time is not the critical issue, while you want to code just few instructions, maybe interactively.

Some AMS/PyROOT specific issues follow:

• All necessary initializations concerning AMS classes are performed by the script ams_utils.py . The script can be easily rebuilt with create_ams_utils.py . It is enough to start your script with the line:
  import ams_utils
  
  See the README file in the python directory for more details.

• We can add several Root files with the AMSChain::Add("file") method. The AMSChain class inherits all the methods of the TChain class, but simplifies the life by hiding some AMS internal operations. Example:
  ams = AMSChain()
  ams.Add("../files/test.root")
  ams.Add("rfio:my_castor_file.root")
  
  Thanks to the AMS utilities, the total number of entries in the chain "ams" can be determined as:
  len(ams)
  
  

• The analysis is performed on AMSEventR objects:
  
  ev = ams.GetEvent()
  
  
  ev = ams.GetEvent(entry_number)
  
  
  ev = ams.GetEvent(run_number, event_number)
  
  The current event can always be accesed as:
  ev = ams.pEvent()
  
  Thanks to the AMS utilities, your can loop over events by just doing:
  for ev in ams:
  
  

• In order to access in a simple way all relevant pieces of information many useful definitions are implemented. For example, the number of hits in ECAL is given by::
  
  print "#Ehits: %s" % ev.nEcalHit()
  
  The "i" Hit in ECAL may be accessed indiferently in any of the following ways:
  
  hit = ev.EcalHit(i)
  hit = ev.pEcalHit(i)
  
  Thanks to the AMS utilities, your can loop over hits by just doing:
  for hit in ev.EcalHitList():
  
  The full list of AMS classes and attributes can be found HERE . For instance, in EcalHitR the first member is "Status", which can be accessed as:
  
  stat = hit.Status
  
  
  

#!/usr/bin/env python
### If the default python version is too old, try for instance:
###    /usr/bin/env /afs/cern.ch/asis/i386_redhat72/usr.local/bin/python2.2
import ams_utils
from ROOT import AMSChain, AMSEventList, TFile, TH1F

ams = AMSChain()
ams.Add("../files/test.root")
#ams.Add("http://pcamsf0.cern.ch/f2dah1/MC/AMS02/2004A/protons/el.pl1.10200/738197524.0000001.root")
#ams.Add("rfio:/castor/cern.ch/ams/MC/AMS02/2004A/protons/el.pl1.10200/738197524.0000001.root")

hfile = TFile("amstest.root", "RECREATE")

list = AMSEventList()

hrig = TH1F("hrig", "Momentum (GeV)", 50, -10.0, 10.0)

for ev in ams:
      for part in ev.ParticleList():
            hrig.Fill(part.Momentum)
            if ev.nVertex()>0:
                  list.Add(ev)  ## Add to list of selected events
                  ev.Fill()     ## write it into output ROOT file

hrig.Draw()
hfile.Write()
list.Write("select.list")

print "We have processed %d events" % len(ams)
print "Histograms saved in '%s'" % hfile.GetName()

---