Simulated data processing - little guide Version 0.

18 January 2002
  • Introduction
  • Generator
  • Simulation
  • Offline
  • Digitisation details
  • Example
    Introduction:
    In order to analyse Monte Carlo simulated events you need to follow a chain of programs :
    1) you need to generate the events,
    2) run the generated events through the detector simulation,
    3) feed the simulation to Ariane.
    Generator:

    At the moment there are two generators :

    the A2pi_generator by Cibran:
    the instructions are in the README file in the directory.
    /afs/cern.ch/user/d/diracww/public/offline/generator/
    Contact: Cibran Santamarina Rios at Cibran.Santamarina.Rios@cern.ch


    the DIPGEN generator by Mikhail Zhabitsky:
    the link to the DIPGEN homepage is     DIPGEN
    http://www.cern.ch/jabitski/dirac/dipgen/
    Contact:Mikhail Zhabitsky Mikhail.Jabitski@cern.ch


    Both programs produce an output file : "kine_f.dat" that will be the input to the Geant simulation.
    Simulation:
    The latest version of the Geant MonteCarlo is in : /afs/cern.ch/user/z/zrelov/public/geant-dirac/version2.61/dev
    where also the instruction for the installation and usage can be found.

    The user-inputs for the simulation are:
  • "kine_f.dat"
  • "jobcrds.dat"

    • The output of the program is a file "buffer.dat" that can be read by the offline program Ariane.
    Some data already produced are in: /afs/cern.ch/exp/dirac/events7/baselmc and /events14/baselmc

    Contact: Peter Zrelov at zrelov@jinr.ru
    Offline:
    The latest version of the Offline program is in : /afs/cern.ch/user/d/diracww/public/offline/ariane/dev/car
    version 304_12 is the recommended one at the moment to analyse MonteCarlo data.

    The digitisation of the MonteCarlo data is performed in the Ariane code, it takes the position, momentum, energy loss and time information from Geant for each track and produces a TDC/ADC values in the different detectors.
    TDC and ADC are then translated again in physical quantities following the different logic for each detector.

    Two types of Digitisation are available :

  • The Ideal situation, thus the true information from the MC are translated in TDC and ADC without any efficiency correction.
    You should activate the flag IdealSimuMC to T in the FFreadInput to Ariane.


  • The more Realistic situation, thus the true information from the MC are corrected with efficiency, delay and all the detector characteristic distortion of the original signal.
    You should activate the flag DigiSimuMC to T in the FFreadInput file.


  • Some of the detectors produce noise, like electronic noise, thus to the original true information given by the Montecarlo (that are always associated to a track), are added extra hits (not associated with a real track) if you activate the flag BackgroundSimuMC to T.


    • Contact: Daniel Drijard at Daniel.Drijard@cern.ch or Angela Benelli at Angela.Benelli@cern.ch

     Digitisation details:

    The detectors that are already digitised are :

  • MSGC : by J.Saborido, from the energy lost it evaluates the cluster size and ADC value for each hit slab.

  • SCIFI : by F.Takeutchi, from position and momenta of each track, the fibre columns hit are re-calculated, the amount of light simulated and depending on the time interval between tracks the PSC circuit is activated.

  • DEDX : by M. Jabitski, from the position inside the detector is evaluated the hit slab, the TDC and ADC value for each track.

  • VHod and HHod : by C. Schuetz, from the hit position and the original TOF is calculated the time smearing. See internal Note : VHSimDet .

  • DC : by L. Kruglova, from the position and momenta of each track is calculated the Drift Time.

  • PRSH : by M. Pentia, from the original energy lost the realistic one is calculated and the true time of the first particle of a shower is converted in TDC value.


    The detectors that are still not digitised are :

  • CHERENKOV : by I. Mihai, the simulation exists as a separate program is still not included in Ariane, ( - I am working on it, A. Benelli)

  • MUON : by V. Brekhovskikh.

    The detectors that have an evaluation of noise are :

  • VHod and HHod. See internal Note : VHSimDet . : by C. Schuetz
  • SCIFI : by D. Goldin.

    Example:
  • For the generation of events :
    Run the a2pi_generator.
    The file "kine_f.dat" is produced.

  • For the detector simulation :
    Run the geant-dirac with the "kine_f.dat" just produced as input.
    In the "jobcrds.dat" put the swit(3) = 3 and swit(6) = 0
    Remember to set the run number and the number of events you like to simulate.
    The file "buffer.dat" is produced.

  • For the offline analysis :
    Run the ariane code (from version 304_06 onward) using the "buffer.dat" as input.
    In the FFreadInput you usually use for your analysis of real data you add the lines:

    IdealSimuMC F
    DigiSimuMC T
    BackgroundSimuMC F

    This is enough to tell Ariane to perform the digitisation of the simulated data taking into account the efficiency of the detectors, i.e. introducing time smearing, detection efficiency, energy resolution, PSC circuit (for the SciFi), etc..
    But it will not add any extra noise hits in the detectors.

    Remember :
    it's not possible to set to True both IdealSimuMC and DigiSimuMC. BackgroundSimuMC could be used independently with IdealSimuMC or DigiSimuMC.
    There are others flags that should be set differently from real data analysis when running on MonteCarlo events:

    CorrTgyDC 0. 0. 0. 0.
    CorrTgxDC 0. 0. 0. 0.
    CorrSFyDC 0. 0. 0. 0.
    UpDownOff 0. 0.

    PosBeamDC 0. 0.
    CorrTgyDC 0. 0. 0. 0.
    CorrTmDrDC F
    For any questions contact me at :
    Angela.benelli@cern.ch
    or at 73891