#Karoly Makonyi 12/20/14
#edited by K Makonyi 12/13/16
- Structure
  - sim_radmap
    full simulation of p-p-bar event with DPM event generator
             usage: 
                   [path-to-the-executable]/sim_rad.exe\
                     -o [name of the outout file (blablabla.root)]\
                     -n [number of events to simulate (2)]\
                     -t [used tracker [Geant4]]
  - PndRadMapBoxMesh
    The class that is used to make the maps.
    One can find examples for its usage in the plot_radmap.C file
  - plot_radmap
    The code that demonstrates the abilities of the Mesh class.
    It produces the 'RadMap_Out.root' file where the required histos are stored.
    It asssumes that the simulated file (RadMap_Out_Sim.root' is in the directory
    where the plot_radmap is run.
  - histodivider.C
    It calculates the quantities for the required running time nominal run
    and - optionally - removes the low-statistic bins.
    The output of this code is 'figures.pdf' and 'figures.root'.
    TODO: add the file name as a command line parameter
    One have to compile it and run the resulting executable
    In a separate textfile (the name must be "names.dat") some details of the histos can be given:
    #name of the histo/directory    title   Xlabel  Ylabel  norm
                   - name of the directory - the histos are arranged in directories
                                             one must give the name of the directory where the
                                               'Histo' and the 'StatHisto' histograms can be found
                   - title                 - title of the histogram taht will appear in the figures.* files
                   - Xlabel                - the label on the X-axis
                   - Ylabel                - the label on the Y-axis
                   - norm                  - the scaling factor on the histos (1 -> quantity per event)
                   the fields should be separated by semi-comma (;)
    have to be compiled separatedly in the _src_ directory like 'Makefile histodivider'
  - Makefile    
    The old Makefile produces the executables
    Presently it is usable for the compilation of the 'histodivider' only
  - Makefile.Panda
    old collection of the libs/headers
  - job_submit.sh
    Sends the jobs (sinulation or plot_radmap.exe)to the _old_ (prometheus) cluster.
    Presently this script is pretty dumb and can only process
    the files in a directory structure which was created by the
    dirmaker.sh. 
    But 
    - it can run the simulation in every directory
    or
    - runs the plot_radmap.exe in every directory 
    - 'hadd's the resulting files
    command line parameters:
    - -b - [b]asename -- name of the the dir where the 'dirmaker' produced the directory structure
    - -s - [s]imulation -- the script will run simulation in every directories in the 'basename' directory
    - -n - [n]umber -- number of simulated events per file
    - -a - [a]nalysis -- the script will run analysis
    - -h - [h]elp -- print out a help screen
  - job_submit_kronos.sh - same as the job_submit.sh but runs on the kronos cluster
  - Scripts/dirmaker.sh
    produces 10x10x10 directories (this is the
    directory-structure that I used
  - Scripts/Run[01234].sh - helper scripts to the job_submit(_kronos).sh
- Workflow
  1. Make a directory
       where the data will be stored, for example
        'mkdir Data'.
  2. Go to this directory:
         'cd Data'.
  3. Copy the 'dirmaker.sh' to this directory:
         'cp ../dirmaker.sh .'
  4. Run it:
         './dirmaker.sh'
       -> This will take some time, but finally the director structures (1000
          directories) will be produced
  5. go up by one directory:
         'cd ..'
  6. run the simulation
         'job_submit_kronos.sh -s -b Data -n 100'
       this will run 100 events in one job, and will run a separated job
        for every directory (1000) in the 'Data' (produced by the 'dirmaked')
        parallel, using the cluster ... some output will be produced
  7. run the 'plot_radmap.exe'
         'job_submit.sh -a -b Data'
       this will run 'plot_radmap.exe' on every simulated files in the 'Data'
         directory parallel, using the cluster ... some output will be produced.
         If every simulated files are processed the results will be added into
         one file (RadMap_Out_Final_Geant4.15GeV.root).
  8. rescale the results to one nominal year operation
       run the histodivider.C
  the 1-6 points have to be done once, the last two points
  have to be done every time when new map have to be created