Dear all,
After long and hard work I have submitted into the cvs a working version
of the e-p elastic event generator with the radiative tail calculated
according to the Mascarad formalism (see Afanasev et al. ). I cannot
promise that the generator is a 100 % ok, but just as any
self-respecting software company I decided to release a "beta" version
and have the users debug it for me. So do so. Here are some details
about it.
1) The generator itself is a combination of the epel generator which
calculates the elastic cross section and the mascarad generator which
only generates the radiative events. The total cross section is then
calculated and the events are generated according to the relative
strength of the cross sections
2) The mascarad cross section is cut off at 10 MeV^{2} to avoid the
infrared divergence as suggested by Cris Crawford. This value is well
below the resolution of our detector. The high cut-off is chosen to be
850 MeV^{2}. 3) The mascarad cross section is generated on the 4-d
grid. Since the cross section peaks very strongly in the phi angle I
integrate in it. I also have to smear out the distribution in tau
variable to avoid problems with the grid (more details in the meeting).
4) The generator returns three momenta : electron, proton and the real
photon. I have not run full Geant with it, but I hope that someone would.
5) Since the angle of the radiated photon peaks very strongly in the
zero angle I only adjust the magnitude of the momenta not the angles.
This might be changed later, but I believe it has little barring on our
kinematics
How to run it:
1) Update your cvs. If it does not compile, talk to Chi. :-) 2) The
new directory will be called MASCARAD in it is the Mascarad.cc and
Mascarad.h programs.
3) update kine.inti file. At the end of it is the definition of the
Mascarad grid. The most important variable is inelasticity, v.
4) run dgen as a stand alone program by doing the following
./dgen 1 1 1 45 0 666 0 11 0 1000000
5) run blastmc with the corresponding kine call.
6) the generator creates two trees: epel.tree and mascarad.tree. Use
those trees to plot the variables together. I submitted a quick macro
into the cvs called "testRadiative.C" which plots the asymmetries in
each sector as function of the Q^{2} with and without the radiative
corrections. You can use it an example if nothing else.
To do:
1) Extend the formalism to the e-d elastic. This will require a long
brainstorming session with Peter K.
2) Extend the method to the quasielastic. Still not sure how I will do
this.
Comments:
Please run the code and send me your corrections and comments. I'm
particularly interested in the full blastmc use of this generator. I'm
also answering questions about the generator over the email, phone and
in person.
Thanks:
I would like to thank Nikolas, Chris, Chi and everyone else who helped
me with this work.
Cheers, Vitaliy
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:32 EST