PGS Particle Type 5

Is that MadAnalysis5 or 4?

Cheers,

Olivier

I am using MadAnalysis 5 (version 1.1.9)
I already checked the MA-code, the problem is in the file tools/SampleAnalyzer/Reader/LHCOReader.cpp, where type 5 is not handled.

Cheers,
Markus

This still poses the question to me if it's just a madanalysis problem or if this case actually should not occur at all.

Hi Markus, (cc John, Benj, Eric, …)

I actually don't know the answer.
The only document explaining the lhco format is the following one:
http://madgraph.phys.ucl.ac.be/Manual/lhco.html
and indeed they are not "5" in that documents.

I've put John Conway (PGS author) and the MA5 Team in cc of this email.
Since this is clearly a topic independent of MadGraph.

Cheers,

Olivier

PS for John: You can see all the (short) discussion on the following web-page:
https://answers.launchpad.net/madgraph5/+question/239651

On Nov 22, 2013, at 7:36 PM, Markus Ebert <email address hidden> wrote:

> Question #239651 on MadGraph5 changed:
> https://answers.launchpad.net/madgraph5/+question/239651
>
> Status: Solved => Open
>
> Markus Ebert is still having a problem:
> This still poses the question to me if it's just a madanalysis problem
> or if this case actually should not occur at all.
>
> --
> You received this question notification because you are a member of
> MadTeam, which is an answer contact for MadGraph5.

Hi all,

We have implemented the LHCO support following the html page indicated by Olivier. This is why this "5" particles are leading to warning messages. We can of course extend the support. However, I would like to know what this "5" means. In principle, it should be part of the met, shouldn't it?

Cheers,

Benj

Hi all,

I just hit this problem myself when using a FeynRules->MadEvent->Pythia->PGS->MadAnalysis toolchain. In the produced lhco file I have entries of the type

  # typ eta phi pt jmas ntrk btag had/em dum1 dum2
  [...]
  5 5 0.095 2.067 58.34 0.00 0.0 0.0 0.00 0.0 0.0
  [...]
  4 5 -0.402 0.457 64.86 0.00 -0.3 0.0 0.00 0.0 0.0

Looking about in the PGS code I found these definitions:

pythia-pgs/libraries/PGS4/src/pgs.inc:342
     . typobj(nobjmx), ! reconstructed type: 0 = photon
                                              ! 1 = electron
                                              ! 2 = muon
                                              ! 3 = tau (hadronic)
                                              ! 4 = jet
                                              ! 5 = heavy charged

Now, jmas (invariant mass) = 0 for a heavy charged particle seems counter-intuitive, and in addition the model I'm using doesn't introduce any heavy charged particles. (The model's a simple EFT with 'chi chi q q'- operator where chi is a real scalar field.)

I'm at a loss on how to solve this conundrum. I can, of course, provide relevant files and follow instructions.

Cheers,
-- Sebastian Liem

Hi all,

I've hit the same problem as well. The following one sample of such event.

 0 6442 3587
  1 2 0.447 4.448 77.74 0.11 -1.0 5.0 0.04 0.0 0.0
  2 4 -1.757 2.220 291.38 17.28 9.0 0.0 2.40 0.0 0.0
  3 4 3.718 5.260 127.23 7.45 2.0 0.0 0.33 0.0 0.0
  4 4 3.193 0.851 87.37 9.76 2.0 0.0 1.50 0.0 0.0
  5 4 0.740 0.204 36.12 6.86 7.0 0.0 1.86 0.0 0.0
  6 5 -0.085 5.349 55.40 0.00 -1.0 0.0 0.00 0.0 0.0
  7 5 -0.275 4.968 41.92 0.00 1.0 0.0 0.00 0.0 0.0
  8 6 0.000 5.007 110.13 0.00 0.0 0.0 0.00 0.0 0.0

I do have a H++ particle in my UFO model. However this particle has been decayed, which I can see from the Pythia file.
I can provide relevant files.

Regards,
Tathagata

Hi,

It is hard to find the proper documentation on the LHCO format. Is the "5" tag specific to PGS or standard? If PGS-specific, then the problem is a PGS problem. If not, then the problem is a documentation problem as the "5" is not part of the LHCO documentation which is btw no more available on the older official website (what is implemented in madanalysis 5 however follows: http://madgraph.phys.ucl.ac.be/Manual/lhco.html) . If anyone has any reference, please do not hesitate to share it with us. Note that I am not reluctant of extending the support but only if this "5"-stuff is a standard.

Cheers,

Benjamin

Hi,

I took another look into the problem:
1)
I scanned the PGS-code delivered with MadGraph for occurences of typobj(..)=5.
It appears exactly once, namely in pythia-pgs/libraries/PGS4/src/pgslib.f at line 4415 (in routine pgs_find_heavy).
Interestingly, at this point the routine pgs_heavy_eff is called, which according to its descriptions "estimates the efficiency for a highly ionizing track in the CDF Run II detector".
Thus it seems like this piece of code is strongly related to CDF simulations and should not be run for LHC simulations at all.
However, this leaves unanswered what kind of particles are actually handled here.
In my case, where new neutral particles are introduced, this leaves me puzzled, as the code explicitely checks for nonzero charge, which excludes my particles. But without them (i.e. SM background), this problem never occurs.

2)
I ran DELPHES for some of my samples.
The results are very similar to PGS, although there are actually clearly visible differences.
Anyhow, these seem to be also the case for samples without the problem, so it doesn't give me any hint on how to treat those strange particles.
Sebastian, Tathagata: Have you tried the same? Since you have different models, maybe you find some hints.
(Note that evaluating DELPHES results does not produce any warnings).

3)
LHCO seems to give MET particles always eta=0.
For type 5 particles, this is not the case, so treating them simply as additional part of MET doesn't look very convincing to me.

Cheers,
Markus

This question was expired because it remained in the 'Open' state without activity for the last 15 days.