[Halld-offline] hdgeant modifications

[Kei Moriya]

Hi Beni, Paul,

I have been analyzing the new output from bggen and hdgeant,
and here are my thoughts:

1. As advertised, eta', phi, omega are decayed in hdgeant, not bggen

2. All unphysical single quarks, diquarks have myid=0.

3. I don't find any rho's in bggen output, I find rho's only from
    phi -> pi0 rho0 and eta' -> gamma rho0 in hdgeant output,
    and find all rho's in the REST output.

4. Decay chain of eta' is OK, all particles have correct parentid.

I have the hd_dump output for eta', phi decays at
http://dustbunny.physics.indiana.edu/~kmoriya/GlueX/2013-09-27-bggen/

Just in case you want them, I have the full hd_dump logs
also, but they are quite heavy.

Problems:

1. There are particles, mostly neutrinos where the myids are messed up.
If you look at EtaPrime.txt or phiMeson.txt in the above link,
you will find particles with XXX at the end of the line.
These particles have a myid that is less than the myid of previous
entries. This does not mean that they are wrong, but for most neutrino
cases you will see that there is a problem.

For example in phiMeson.txt, if you look at event 1082,
a mu+ neutrino pair are created, but the neutrino has myid 0.

Event 1552 is another case, where a mu+ neutrino pair are created
from parentid 281 (does not exist), and the mu+ (myid 327)
decays to e+ nu nu, with the 2 nu's having strange myids.

Out of 50k events, there were more than 15k neutrinos that had
myid 0, but 34k that had myid not 0. Maybe you guys know what
the distinction is (maybe if 2 neutrinos are produced in a mu
decay, only one is tagged with the correct myid? for charged pi
to mu nu, the nu is never tagged correctly?)

2. The attached plots are created by analyzing Paul's Thrown_Tree.
I take the particles with Thrown__ParentID[i]==-1, and
add the 4-momenta. The file p4_bggen.pdf is when I run
the plugin on the bggen output, p4_REST.pdf is when I run
on the REST output. The situation is much better than before,
but there are still events that are not reconstructing the
generated information correctly, presumably because of 1. above.

Notice also that there is a humongous spike at pz = 0 for
the REST output.

There are also some events with total pz between 0 and 7 GeV.
This seems like most of the generated particles are missing.
The attached file lowPz.txt shows the event numbers for
events with 0 < pz < 7 (75 events) if you are interested in
looking at what they are.

Any input is very welcome.

	Kei

On 9/26/13 1:15 PM, Beni Zihlmann wrote:
> Hi Kei,
>
> I committed a few changes to hdgeant and bggen. Now I hope the relation
> between parent
> and daugther particles is more or less restored. The initial thrown
> particles have an myid
> number which is the track number of the particle and parentid is zero.
> All intermittent
> particles from bggen have myid number zero and parentid zero. If
> parentid is not zero
> than that number is the track id number of the parent which is myid.
> However if you
> can not find a parent particle with that myid that means the parent
> particle was
> generated inside the calorimeters. You will find that rather often you
> see two photons
> being associated with a parent particle that does not show up in the
> list. That means
> the parent pizero was generated inside the calo or just at the edge but
> the decay
> vertex is outside. I think this is an edge effect were in gustep the
> particles being saved
> into an new vertex is slightly different than particles being made
> primary. I will check
> on that. The upshut is at the moment that you will find some particles
> that you can
> not associate with a parent.
>
> cheers,
> Beni
>
>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: p4_bggen.pdf
Type: application/pdf
Size: 27342 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20130927/1e24bcae/attachment-0004.pdf>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: p4_REST.pdf
Type: application/pdf
Size: 27892 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20130927/1e24bcae/attachment-0005.pdf>
-------------- next part --------------
    3395
    3415
    3596
    4422
    5918
    6034
    6040
    7184
    7687
    7738
    7778
    7949
    8288
    8778
    9832
   10832
   10924
   11142
   11431
   11574
   12221
   12869
   12993
   13729
   15322
   16965
   17565
   17949
   18791
   19284
   20650
   20998
   21206
   21580
   22388
   22603
   22754
   23882
   25421
   27030
   27641
   27969
   28675
   29040
   29848
   30755
   31033
   33451
   34000
   34625
   35243
   35545
   35684
   35761
   36231
   37265
   37318
   37857
   37958
   38464
   39121
   41762
   41791
   42332
   42594
   42863
   44232
   44468
   45890
   46951
   47680
   48723
   49433
   49895