| [CLOSED] DPM-Direct doesn't have Coulomb el.scattering part [message #15461]
|Sat, 07 September 2013 17:55
Registered: May 2010
Location: Mainz, Germany
I would like to use PndDpmDirect for my studies with DPM. Such studies I did many times before with so called DPM stand-alone.
As far as I understand PndDpmDirect is just interface to DPM generator, which runs PndDpmGenerator+FairPrimaryGenerator.
So generated results, if I run compiled executable, or if I run task with PndDpmDirect, shouldn't differ in terms of distributions of true (MC) variables.
But they have significant and very important difference, especially for the Luminosity Detector. Basically, with current configuration, Coulomb part for elastic scattering is not simulated with PndDpmDirect, even if theta_min is set to very low value!
Please, find attached two plots with theta distribution of simulated tracks and note the difference between them close to 0 and 1.5 rad (in both cases Pbeam=1.5 GeV/c, number of simulated events=1e4, theta_min=0.1 degree).
Well, my standalone version is executable which I have to compile by hand. To do so one need to copy attached GNUmakefile in $PANDAroot/pgenerators/DpmEvtGen
Then in this directory type make and to resolve error
pgenerators/DpmEvtGen/main.cc:246: undefined reference to `dpm_gen__'
change dpm_gen__ to dpm_gen_ (x2 times in this main.cc file)
Afterwards executable created in $PANDAroot/pgenerators/DpmEvtGen
As you can see only one change here is dpm_gen__ "correction". I don't know why it's needed for standalone compilation and isn't needed for common compilation together with other packages in pandaroot. But as far as I remember this dpm_gen__ to dpm_gen_ was an annoying issue while ago.
Anyway all other lines/files in DPM stay exactly the same as in repository.
Any ideas why PndDpmDirect just ignore Coulomb part? Also it would be nice if somebody else could cross-check DPM spectrum to confirm problem which I describe above.
[Updated on: Sun, 15 December 2013 13:50]
Report message to a moderator