Gluex_MC.py not defining DMCThrown object in hddm file

[Shannen Graham-Howard]

Hi all,

I'm currently working on a generator for exclusive pi0 events. It produces sensible values when I run gluex_MC.py over the hddm files that I'm producing, but they resulting files don't seem to have the DMCThrown object defined, and both hd_dump and hd_root segfault when I try and call the DMCThrown object. DBCALShowers and DBCALHits are also not defined, but DBCALdigihit is defined and filled. All of the FCal information is defined and it's producing results that look sensible for the energies.

An example of the segfault I get can be found at /w/halld-scshelf2101/home/shannen/events/generators/err.out. When I initially ran everything I was still on version_6.3.0.xml, but I've since gone back and built the newest version of halld_recon with version_7.0.0.xml, so I don't think it's anything to do with my environment. I'm wondering if it's just a flag I'm missing in my jana.config file, or if there's something deeper I'm doing wrong.

Shannen

[alexander.a...@gmail.com]

Hi Shannen,

I can confirm the crash using this generator file:

/w/halld-scshelf2101/home/shannen/events/generators/generatorOUT/paraHom.hddm

I looked at your generator file and found some inconsistencies. Here is one event:

       <product decayVertex="0" id="1" mech="909" parentid="0" pdgtype="1000822080" type="Lead(111)">
          <momentum E="193.974" px="-0.322404" py="0.187523" pz="0.307988" />
        </product>
        <product decayVertex="0" id="2" mech="909" parentid="0" pdgtype="111" type="Pi0(7)">
          <momentum E="5.54856" px="0.322404" py="-0.187523" pz="5.53436" />
        </product>
        <product decayVertex="0" id="0" mech="0" parentid="0" pdgtype="0" type="UnknownParticle(0)">
        </product>

The final state lead nucleus does not have the right mass and geant complains about it:
G4WT3 >    Pb208 found in input MC record, geant3 type 111, PDG type 1000822080 has unphysical mass:
G4WT3 >    expected 193.688 GeV, found 193.973 GeV, difference -285 MeV

In addition, there is an UnknownParticle without a four-vector.

I don't know if these are the reason for the crash, but they should be fixed before investigating further.

Cheers,

Alex

[Shannen Graham-Howard]

Hi Alex,

For the target, I'm trying to model the recoil of the lead (just for completeness). I'm trying to write the target as a Pb atom at rest, and the recoil as a Pb atom with some 4-momentum I calculate using conservation of mass + energy in my generator. What it looks like is happening though is that it's trying to treat the recoil Pb as the target, and then maybe it doesn't know what production mechanism to use for the Pb atom at rest, hence the mystery particle without a four-vector. I'll try and run it without the recoil Pb written to the .hddm file (since it's velocity is very small anyhow, it shouldn't show up anything) and see what happens.

Thanks for looking into this,

Shannen

[Shannen Graham-Howard]

I had added the wrong number of products, causing the issue. Sorry for the mistake, and thanks for your help!

Shannen