Skip to content

Latest commit

 

History

History
168 lines (108 loc) · 10.4 KB

README.md

File metadata and controls

168 lines (108 loc) · 10.4 KB
Raw

DEMPgen

Event generator for Deep Exclusive Meson Production

Building

  • To build the event generator, the compiler needs acces to both a compiled installation of CERN ROOT and its source code. ROOT version 6.08.06 or later is supported, and must be installed with the MathMore package enabled. Be sure not to change the location of either the ROOT source or compiled files after installation, as this will interfere with ROOT's built in CMake configurators.

  • CMake is also required. CMake 2.8 is the minimum supported version, and CMake 3 has been tested as well.

  • After downloading the source create a build directory and cd to it. Take note of the location of the source directory (where CMakeLists.txt should be stored) and run the commands:

    • mdkir build.
    • cd build
    • cmake ..
    • make -j8

  • As a one liner -

    • mkdir build && cd build && cmake ../ && make -j8

  • The event generator can now be run using the following command in the data/ directory.

    • cd data/
    • ./../build/DEMPgen ../Config.json

Building on the iFarm

  • Building on the JLab iFarm requires you to set up some software versions beforehand, to build successfully, I did the following

    • Comment out any CUE or other initialisation in your .login/.cshrc scripts
    • Login to an ifarm node
    • module load cmake/3.19.4
    • module use /group/halla/modulefiles
    • module load root

  • I then build DEMGen using the one liner above without any issues, you will need to load these modules when running the generator subsequently (this is done by default in the farm job scripts)

Configuration

The file Config.json contains all the configuration options. Use this as a template for other configuration files, which may be given as an argument to the event generator.

Output

The event data is output to the configured file location relative to the build directory. The TTree in this file contains all kinematic data for all particles in the laboratory rest frame. Variables with the prefix "Vert" represent values read at the interaction vertex. Values with the prefix "Lab" represent values read after all correcting effects (multiple scattering, ionization, etc.) have been applied.

Sources and Classes

Particle Class

The Particle class inherits from the root TLorentzVector class so that its functions are immediately accessible through the Particle objects. It's member variables include its mass, charge, and its GEMC compatible id.

DEMPEvent Class

The Event class contains all particles of a single event, in one frame of reference. It includes methods enabling transformation to other frames of reference and coordinate systems.

Asymmetry Class

The Asymmetry class calculates asymmetry amplitudes based on Monte-Carlo data by Goloskokov and Kroll. The asymmetry objects are persistent between events.

Customrand Class

A class to hold various distribution functions for randomly generated variables, such as scattered electron energy. These functions should be persistent between events, and will provide fast random numbers.

ScatteredParticleGen

Stores the kinematic ranges for the scattered electron and generates them with random energy and direction within the range, using sphere point picking.

TargetGen

Generates the target neutron with Fermi momentum (when enabled) and proton for FSI (when enabled, may also have Fermi momentum).

ProductGen

This class reads in the kinematic variables for the incident, target, and scattered particles and uses conservation laws to solve for the remaining two particles. The pion is first given a random direction. A fast root finding algorithm then calculates the pion's momentum magnitude, which is then used to find the proton's momentum. Pion direction may also be passed as an argument for debugging purposes.

SigmaCalc

This class returns the cross sections and weights for the event. It acts as an interface between the current version of the event generator and header files from the old event generator (seen under branch "original"), which contain the parameterization of the cross sections.

TreeBuilder

Manages the ROOT TTree object to be stored in the output .root file. Contains methods to easily add all kinematics data stored in a particle or DEMPEvent object.

Matter Effects

Transforms particle kinematics based on three effects caused by transition through matter: Ionization, Bremsstrahlung, and Multiple Scattering.

FSI

Computes the effect of final state interaction between the produced pion and one of the two protons of the target nucleus. The momentum of the outgoing pion and recoiled proton, as well as the cross section of the interaction is calculated based on elastic pion-nucleon scattering.

Acknowledgments

JsonCpp

This project uses JsonCpp to read in configuration options. The amalgamated sources for JsonCpp are redistributed with this project in compliance with the MIT license.

Processing Scripts and json examples

  • There are several shell scripts in the main directory that can be used to run the generator and produce some output, information on these scripts is provided below
  • To remove clutter from the main directory, example .json scripts were moved to a new folder
    • json_examples

Process_EIC.csh

!!! NOTICE !!!
This script copies Config_EIC.json and formats a new file based upon this, DO NOT MODIFY Config_EIC.json if you want to use this script!
!!! NOTICE !!!

  • To facilitate the submission of batch jobs, I created a csh script to automatically construct .json config files and run them. This script can also be utilised to run the generator manually, without the need to go and edit a json file.

  • The script requires 8 arguments (which is a lot, I know), but in the K+ case, it expects 9. They are as follows -

    • Arg 1 - FileNum -> For batch running, we typically run X files of Y events, this argument is just X, if you're running manually as a test, just input 1 or whatever you fancy
    • Arg 2 - NumEvents -> The number of events thrown for this file, set this to whatever you want to run. For reference, with the Pi+/K+ generator, 1B files takes ~1 hour
    • Arg 3 - EBeamE -> The electron beam energy, set this to whatever you want, typically, we use 5, 10 or 18 (the nominal max for the EIC)
    • Arg 4 - HBeamE -> The hadron beam energy, again, set this to whatevr you want. Typically we use 41, 100 or 275 (41 and 275 being the nominal min/max)
    • Arg 5 - RandomSeed -> The random seed, self explanatory. Set this however you like, the batch submission job randomly generates a random seed to feed in here
    • Arg 6 - OutputType -> The format of the output file, select from LUND, Pythia6 (for Fun4All) or HEPMC3 (for ATHENA), the default is Pythia6 if your choice is invalid
    • Arg 7 - InteractionPoint -> The interaction point, choose from ip6 or ip8. The default is ip6 if your choice is invalid
    • Arg 8 - Particle -> The produced particle (meson) in the reaction, choose from omega, pi+, pi0 or K+
    • Arg 9 - Hadron -> OPTIONAL - This only matters if you select K+ as the particle, in this case, choose from Lambda or Sigma0 here. If your choice is invalid (or you don't specify arg9), the default is Lambda

  • So as an example if you executed the following -

    • ./Process_EIC.csh 1 100000 18 275 24432 HEPMC3 ip6 K+ Lambda

  • You would run the generator for 18 GeV e- on 275 protons for ip6, throwing 100000 events with the K+/Lambda generator.

Batch_Submission_EIC.sh

  • This script creates and submits batch jobs. It is designed for use with the torque queueing system on Lark at the University of Regina. However, it could quickly be adapted for use on the JLab iFarm for example if desired. The jobs the script creates and submits all execute the Process_EIC.csh script described above. This script requries a very similar set of arguments -

    • Arg 1 - NumFiles -> The batch script is designed to run X jobs of Y events, this number is just X, the number of files you want to run
    • Arg 2 - NumEvents -> The number of events thrown for this file, set this to whatever you want to run. For reference, with the Pi+/K+ generator, 1B files takes ~1 hour
    • Arg 3 - EBeamE -> The electron beam energy, set this to whatever you want, typically, we use 5, 10 or 18 (the nominal max for the EIC)
    • Arg 4 - HBeamE -> The hadron beam energy, again, set this to whatevr you want. Typically we use 41, 100 or 275 (41 and 275 being the nominal min/max)
    • Arg 5 - OutputType -> The format of the output file, select from LUND, Pythia6 (for Fun4All) or HEPMC3 (for ATHENA), the default is Pythia6 if your choice is invalid
    • Arg 6 - InteractionPoint -> The interaction point, choose from ip6 or ip8. The default is ip6 if your choice is invalid
    • Arg 7 - Particle -> The produced particle (meson) in the reaction, choose from omega, pi+, pi0 or K+
    • Arg 8 - Hadron -> OPTIONAL - This only matters if you select K+ as the particle, in this case, choose from Lambda or Sigma0 here. If your choice is invalid (or you don't specify arg9), the default is Lambda

  • The script automatically generates a random seed itself using the /dev/urandom function

Process_EIC_iFarm.csh

  • This version is for use on the JLab iFarm/Farm
  • This script uses the same arguments as Process_EIC.csh
  • If you are processing interactively (i.e. on the iFarm), you will need to make sure that you have executed -
    • module use /group/halla/modulefiles
    • module load root
  • You should also check the path set at the top looks OK
    • By default, /eic/users/${USER}/DEMPGen is assumed

JLab_Batch_Submission.sh

  • This version should be used to submit jobs to the Farm batch queueing system (swif2)
  • It uses the same arguments as Batch_Submission_EIC.sh
  • Before running, ensure you have a folder or sym link for LundFiles
    • data/LundFiles
    • This must exist for output!
  • You should also check the paths set throughout the script look ok (for example, in the COMMAND: ... line)
    • By default, /eic/users/${USER}/DEMPGen is assumed

json_examples

  • There were several .json files clogging up the main directory, many of these were very outdated. As such, I've moved them all to a subfolder -
    • json_examples
  • This folder has .json config files for a variety of different conditions.
    • Due to several of them being quite outdated, the Config_EIC.json file in the main (the directory of this README) directory should be consulted to see the options that are actually availble.