New stable particle in MG/Pythia/Delphes

The best is probably to ask the Delphes author about this.
https://cp3.irmp.ucl.ac.be/projects/delphes/wiki/WikiStart

otherwise, you can try to cheat by assigning to your stable particle the pid of the lightest neutralino in the mssm.
Then I would guess that this will be treated correctly.

Cheers,

Olivier

Thanks, that trick works.

Although MadGraphs miraculously decides to change the name of the particle once I changed the pid, is this wanted behaviour?
It is not a very satisfactory solution though...

Best, Maikel

Hi Maikel,

> Although MadGraphs miraculously decides to change the name of the particle once I changed the pid, is this wanted behaviour?

Yes this is expected. By default all sm/mssm particles have a unique name in MG whatever the name is inside the UFO model.
This allow to have a more uniform behavior between various model develloper.
You can easily bypass this by doing:
import model XXX --modelname
Then you will keep all the name as assigned in the UFO model.

Cheers,

Olivier

Okay the delphes team (rightfully probably) redirected me here again, and there seems to be a problem with MG+Pythia and not with Delphes:

https://cp3.irmp.ucl.ac.be/projects/delphes/ticket/154#comment:8

The unweighted events are still correct, however Pythia converts the particle ids to zero?

Do you know what might cause this? Thanks for considering!

Best, Maikel

Hi,

This is a probably a problem at the pythia level, but before send you to pythia author, it might be good to cross-check
the MG-pythia interface. Johan could you do that?

Cheers,

Olivier

Hello Maikel,

The problem is neither MadGraph, Pythia, nor Delphes, but STDHEP. In the file
pythia-pgs/libraries/PGS4/src/stdhep-dir/src/stdhep/lutran.F
the STDHEP authors did not take into account the possibility of particles in the 9,000,000 range, except for some special exotics in 9,900,000. You can fix this by simply removing (or commenting out) the lines 146-147:

      else
        lutran=0

In fact, I think we will do this in the next release of the MadGraph Pythia-PGS package. So thanks a lot for directing our attention to this.

And a clarification: It seems from your Delphes ticket that there is some confusion regarding which files should be used. As you can see in
https://cp3.irmp.ucl.ac.be/projects/madgraph/wiki/FAQ-General-9
the pythia_events.lhe file is NOT an input to Delphes (or to anything else). This is also clearly stated at the beginning of the file itself:
<!--
# Warning! Never use this file for detector studies!
-->
as well as written to the command line when the file is generated. The ONLY file Delphes cares about is the pythia_events.hep file.

All the best,
Johan

Thanks,

In the Delphes report I raised also the issue of btagging efficiencies and as it seems modifying the Delphes card b-tagging efficiencies do not get transferred to the Delphes program, When I open a set of Delphes lhco events I see:

## Tagging definition
#BTAG_b 67 // b-tag efficiency (%)
#BTAG_mistag_c 1 // mistagging (%)
#BTAG_mistag_l 1 // mistagging (%)

.
.
.

  # # B-tagging efficiency functions *
  # #******************************* *
  # * *
  # * Tagging a "b" as a b-jet: 0.4 + 0*x + 0*y *
  # * Mistagging a c-jet as a b-jet: 0.1 + 0*x + 0*y *
  # * Mistagging a gluon-jet as a b-jet: 0.01 + 0*x + 0*y *
  # * Mistagging a light jet as a b-jet: 0.01 + 0*x + 0*y *

So as it seems to me modifying the cards in the MadGraph Event directory does not seem to work.
Can you comment on this, and maybe give a solution?

I hope it is fine I pose this question in this report.

Cheers, Maikel

Hello Maikel,

What exactly is the problem? Are you saying that when you run Delphes, it doesn't read the information in the Cards/delphes_card.dat file?

The cards copied into the delphes_events.lhco file are simply the cards in the Cards directory - this inclusion is made by MadGraph, not by Delphes, so this doesn't say anything about whether Delphes reads the information in the cards or not.

All the best,
Johan

Hello Johan,

What I am saying is that modifying the Delphes card does not seem to influence the number of b-tagged jets in my LHCO sample generated by Delphes. From what I figured there seems to be a mismatch between those two b-tag definitions above in the same file, but I can't find the second one in any of the Delphes cards.

I hope this is clear now.

Best regards, Maikel

Hello Maikel,

Indeed, the Delphes cards are still the original Delphes v. 1.X cards. I guess we just forgot to upgrade the default cards when we upgraded the Delphes version in the install command. We'll make sure to do this in the next release. The easiest way for now should be to simply replace the cards with the default cards from the Delphes directory and start from these when you do your modifications.

All the best,
Johan

Hello Johan,

Thanks for solving the problem, I will test this later or wait for the update.

Best, Maikel