Installation

Getting COMPHEP

I get COMPHEP from the official page: http://comphep.sinp.msu.ru/. The code is available from downloads - although free registration is required. This will allow you to download via your web browser (you would have to work your authentication into wget, and I don't know how to do that).

Once you have the version you want, download the tarball. The COMPHEP tarball follows the usual practice. Once you untar the code, you should have a directory of form comphep-[Version Number]. Enter that directory, read the INSTALL text file for last minute instructions, and then run ./configure and ./make:

tar -zxvf comphep-4.4.93.tgz
cd comphep-4.4.93
./configure
make

./configure should return a message telling you that you are ready to compile. ./make should print a great deal to the screen, ending in:

******************************************************************
*        CompHEP-4.4.93 has been successfully built          *
*                                                                *
*  Please create a user working directory using the command      *
*         make setup WDIR=path_to_your_user_work_dir             *
*  NOTE: Do not use '~' to refer to home directory               *
*        Use environment variable HOME instead of                *
******************************************************************

After this, I follow the instructions to make a work directory:

make setup WDIR=/home/users/norman/MCGen/Comphep/work/

This appears to create a new directory (as it asks if you want to overwrite an existing directory):

**************************************************************
 The directory WDIR=/home/users/norman/MCGen/Comphep/work/ exists!

 Overwrite?
 Press [Enter] to continue or ^C to stop.
**************************************************************

...


*****************************************************************
 Setup of the user directory for CompHEP-4.2 has been
 successfully performed

 To start CompHEP go to the user directory /home/users/norman/MCGen/Comphep/work/
 and enter the command './comphep'
*****************************************************************

No, I don't know why it says that the setup for CompHEP? -4.2 has been created. I just don't ask.

Basic Operations

Step-By-Step: W+ -> e+ nu, gamma at LHC

First, to run CompHEP, you must have X enabled, as it will open in a new window.

• Go to your work directory.
• Start CompHEP (./comphep). This should open a new window.
• Unless you really want to read the License Agreement, press enter.
• This gets you to the Abstract window. Please select a gauge. This does profound things to the physics of the CompHEP? generator that, quite frankly, we don't really understand. If you are not a theorist you can either select a gauge at random, or pick one of the two Standard Model gauges. Haifeng claims that they both produce the same results.
• Now you have to pick what kind of process you are simulating. For most of your physics career you will probably be doing scattering processes from colliders. Choose that option.
• This takes you to the window in which you choose the initial state of your collider.
  • You have to specify, in this order, the first beam component, the first beam energy, the second beam component, the second beam energy, and the pdf
  • For simplicity, I choose the beam types to be p1s, a proton without s,c, or b contributions from the sea.
  • Use the default pdf for now (12 - cteq6l1)
• Now we pick the particles to decay. We can either pick in terms of initial particles (such as the H), intermediate particles (ZZ or WW for example), or the final state (4 lepton, etc.). For this tutorial, we will use final state particles, e+,v,and gamma. Do this by entering E,ne,A into the line.
  • Do not exclude any special diagrams at this time. You can press enter for the next two querries.
• View the diagrams to make sure that we are producing what we want, choosing either of the two subprocesses. There should be four Feynman diagrams. Press ESC to exit that page when you are happy.
• Square the diagrams. This should be profoundly uninteresting, but there should now be a great many diagrams to look at if you are really crazy.
• Choose Symbolic Calculations and hit ENTER.
• Choose C-Compiler and hit ENTER. This should open a new xterm that will build the executable you have just specified. When it finishes, it asks you to hit ENTER. Do so.
• You should now have a shiny new xwindow. This is where all the real physics takes place.
• You can now examine the parameters you have entered.
  • The collider parameters are under Initial State
  • The Standard Model parameters are under Model Parameters.
• Let's go straight to cuts. We want to put a 10 GeV? pT cut on the photon, and and a 0.35 restriction on the angle between lepton and photon. To do this, we will need to know how to specify both variables and particles. Variables are discussed in the CompHEP Variables section of this page, but for right now we only need two, T (Pt) and J (angle).
  • The particles are numbered in the same order as they are in the (sub)process window. The first particle is 1, and they go up from there. The two initial partons should be number 1 and 2. In this case we have u,D -> A,ne,E. Here, u is 1, D is 2, A is 3, ne is 4, and E is 5. To limit the Pt (variable T) of the photon (particle 3), we constrain T3 to have a minimum of 10.0, and leave the upper bound blank (enter T3 in the first column, and press TAB to get to the next columns. Press enter at the end).
  • To set a lower bound on the angle between photon and positron, we set a limit on J35, with a minimum bound of 0.35.
  • When done, press ESC to exit this section.
• Change the Width Scheme to Running by selecting it and choosing from the menu.
• Let's go down and find the cross-section. Go to Vegas and select it with ENTER.
• Here, we get to choose what we are going to look at. I want to look at the pT of the photon and the lepton. Do this by entering two lines, one for T3 and one for T5. I set both to have bounds between 10.0 and 500.0.
• Hit Start Integration.
  • With Width Scheme set to Running, it's possible to get bad results. If the final result of integration is radically different from the numbers in the same column on preceeding lines, hit ESC, go back and re-select Running for Width Scheme (for some reason, this resets the phase-space random number).
  • If you run the Integration again, it appears to accumulate the statistics, giving a more accurate answer. I got 24.6 for this process.

Useful Documentation

CompHEP Variables

-- MatthewNorman - 09 Oct 2007