[Halld-offline] retrieving information on generated particles

[Kei Moriya]

Hi Mark and offliners,

Regarding the talk I gave yesterday at the physics meeting,
I think obtaining the generated particle information
is a relatively high-priority issue since this affects
every analysis where we want to know the main
background channels.

I therefore propose that this be put on the agenda
for offline issues, and make a collaborative effort
to solve it, or at least have a plan of what to do
in the coming weeks.

I think there are many inter-related issues:
1. We need to maintain a consistent output throughout
the processing flow from pythia/hdgeant/REST/Paul's trees,
which probably all have different experts.
2. The upgrade to GEANT4 in the coming weeks. We don't want
to duplicate the same efforts for different code.
3. There needs to be agreement on which information needs
to be saved and decide which PID scheme to use.

I think part of the confusion comes from the PID scheme,
since pythia uses the PDG PID scheme internally, while
the GlueX PID scheme depends on the GEANT particle scheme,
but has been modified for numbers higher than 45. Sometimes
the conversion between different schemes seems to work,
sometimes they don't.

I think we should try to:
1. see which problems can be fixed rather easily for GEANT3
2. be aware that the problem currently exists while updating to GEANT4

I can try to iron out some of the specific issues with Paul
and others by the next offline meeting next week, but if anybody has any
other input or suggestions, I think it would benefit the discussion.

Below is part of an email I got from Paul, and I agree with
what he has here. One more thing I would like is a mechanism
that tells me what created the particle. As Curtis
mentioned during the meeting, retaining wide states such as
rho mesons would be controversial, but I think we should aim
for something that works for the majority of analyses at this stage,
even if it means sweeping generalizations.

---

It seems like, beyond bug fixes, the best long-term solution for getting 
the thrown topology in ROOT is function(s) that will return you either:

1) A string of the particles
2) A vector of the particle IDs
3) A map of particle_id to #-of-particles

These functions would have options like:

1) Include final-state particles only.
2) Include long-lived-decaying (LLD) (e.g. Lambda) particles and 
final-state particles, excluding final-state particles that decayed from 
LLD particles.
3) Include all "known" decaying particles (e.g. rho resonances, exclude 
diquarks), excluding final-state particles that decayed from 
known-decaying particles.

---

Thanks,
	Kei