MadGraph5_aMC@NLO

DECAY interface and MG5

Asked by Julien Maurer

   Hello,

I am generating gluino pairs with MadGraph 1.5.2 (imposed), including gluino decays that are either ~g -> t t~ n1 or ~g -> w+ w- b b~ n1.

In order to preserve part of the spin correlation, I was hoping to decay the tops or Ws with the DECAY module (can't use MadSpin, it's not supporting the upstream process, and would be too slow anyway), with MadGraph 5 (1.5.2, imposed).

I have to run locally (not through the web interface) so inspired by instructions for gridpack mode I have copied the DECAY and HELAS directories (+ make_opts) somewhere, compiled them (gfortran), created two input_*.in files to decay tops, then executed ./decay. This results in a prompt crash after answering the questions:
At line 484 of file decay_coupling.f
Fortran Runtime error: Bad integer for item 1 in list input

Since it is working correctly on

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Julien Maurer
Solved:
Last query:
Last reply:
Revision history for this message
Julien Maurer (jln-maurer) said : 		#1

[sorry for the split message, I did a wrong manipulation]

The crash occurs after specifying the output file in the prompt, and before specifying the particle to decay.

Since it is otherwise working fine on a pp->ttbar test file, I assume this crash has to do with the banner being SLHA-like instead of SM. Is there a way to avoid this issue? I'm not interested in decays of the SUSY particles, only the tops/W.

More generally, I wanted to ask if you still recommand using DECAY with MadGraph5, and whether I am correct in assuming that spin correlation effects would indeed be present (at least better than if passing directly the .lhe to pythia).

Thanks a lot,
Julien

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#2

Hi Julien,

One solution is probably to modify your event file to include a valid SM param_card at the beginning of the file.

>More generally, I wanted to ask if you still recommand using DECAY with MadGraph5

We don't have an official statement on the support of the DECAY package but this is clear that we are not going to develop it any further. This code is not part of the official version anymore and the support for it will therefore be very minimal. So I will recommend to use MadSpin (I can add support for your process actually)
BRIDGE is also an option but he is working only for v4 model and I don't think that he is still supported anymore.

>, and whether I am correct in assuming that
>spin correlation effects would indeed be present (at least better than if passing directly the .lhe to pythia).

You will not have the full spin-correlation, neither off-shell effects, but you will have some (most?) of the spin correlation and actually more than the one handle by pythia.

Cheers,

Olivier

Revision history for this message
Julien Maurer (jln-maurer) said : 		#3

  Hello Olivier,

Thanks a lot for your answers.

Modifying the <slha> block indeed fixed the crash and the rest went fine after that.

As for including the process in MadSpin, the issue when I tried was that there are decays already at MadEvent level, and MadSpin was not very happy with that (i.e. crashing).
To illustrate, the processes I'm interested in are:
generate p p > go go, go >t t~ n1 (for testing)
generate p p > go go, go > w+ w- b b~ n1

The second case is already very time consuming (I also have to add p p > go go j) so I can't add further decays in MadEvent. Processing time would also be an issue with MadSpin I guess, even if it was working on this situation...

Cheers,
Julien