MadGraph5_aMC@NLO

MadSpin on partly decayed processes

Asked by Julien Maurer

  Dear MadSpin authors,

I am generating events featuring gluino pair production followed by a 3-body decay of the gluino into a t+tbar+LSP triplet which is a very common benchmark model for SUSY searches at LHC:
import model mssm
generate p p > go go, go > t t~ n1

I was hoping to further decay the events (tops and W bosons) with MadSpin, gluinos being decayed at MadEvent level since 3-body decays are not supported by MadSpin, as I understood. But I encounter several issues doing so, that I bypassed by modifying my local version of the code. Going further however, the last crash origin makes me wonder if what I want to do is supported at all, hence my question: is it possible to use MadSpin on events that are already partly decayed with MadEvent?

To give a clearer and less exotic aspect to this issue, I can reproduce the same crash (and for the same reasons) when generating SM ttbar events that I partly decay with MadEvent:
generate g g > t t~, t > w+ b

the origin of the MadSpin crash is that the load_events() function in MadSpin's decay.py looks for a production tag ((21, 21), (-6, 5, 24)) in the matrix elements, which do not contain such a key but only the following: [((21, 21), (-6, 6))]. There is therefore a mismatch between on one side a key built from initial and final state particles, and on the other side a key built with the underlying production process.

Many thanks in advance for any input on that subject...

Best regards,
Julien Maurer

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Olivier Mattelaer
Solved:
Last query:
Last reply:
Revision history for this message
Best Olivier Mattelaer (olivier-mattelaer) said : 		#1

Hi Julien,

Indeed this is not supported. The reason is that this starts to be extremely difficult to handle in all generality.

Cheers,

Olivier

On Dec 2, 2013, at 12:11 PM, Julien Maurer <email address hidden> wrote:

> New question #240157 on MadGraph5:
> https://answers.launchpad.net/madgraph5/+question/240157
>
> Dear MadSpin authors,
>
> I am generating events featuring gluino pair production followed by a 3-body decay of the gluino into a t+tbar+LSP triplet which is a very common benchmark model for SUSY searches at LHC:
> import model mssm
> generate p p > go go, go > t t~ n1
>
> I was hoping to further decay the events (tops and W bosons) with MadSpin, gluinos being decayed at MadEvent level since 3-body decays are not supported by MadSpin, as I understood. But I encounter several issues doing so, that I bypassed by modifying my local version of the code. Going further however, the last crash origin makes me wonder if what I want to do is supported at all, hence my question: is it possible to use MadSpin on events that are already partly decayed with MadEvent?
>
> To give a clearer and less exotic aspect to this issue, I can reproduce the same crash (and for the same reasons) when generating SM ttbar events that I partly decay with MadEvent:
> generate g g > t t~, t > w+ b
>
> the origin of the MadSpin crash is that the load_events() function in MadSpin's decay.py looks for a production tag ((21, 21), (-6, 5, 24)) in the matrix elements, which do not contain such a key but only the following: [((21, 21), (-6, 6))]. There is therefore a mismatch between on one side a key built from initial and final state particles, and on the other side a key built with the underlying production process.
>
> Many thanks in advance for any input on that subject...
>
>
> Best regards,
> Julien Maurer
>
> --
> You received this question notification because you are a member of
> MadTeam, which is an answer contact for MadGraph5.

Revision history for this message
Julien Maurer (jln-maurer) said : 		#2

  Dear Olivier,

Thanks for your answer. For my own curiosity, may I ask you to comment on the following:

Again with this half-decayed ttbar example, I modified the input LHE file by hand to pretend it had been generated in the following way, the goal being to suppress (from MadSpin point of view) the intermediate decays that are problematic:
generate p p > w+ b t~ / b~
of course, the events were in fact generated with p p > t t~, t w+ b as said before.

In that way, MadSpin runs without apparent trouble and indeed decays events. Does this approach make any sense at all (spin correlations should be tracked properly, but I guess it's not so simple...)? It seems not to work fully anyway (the W boson "in the matrix element" is decayed only in cs~ for some reason, while the other one goes in all channels), but I'm just curious now...

Many thanks,
Julien

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

Thanks Olivier Mattelaer, that solved my question.

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#4

Hi,

> Again with this half-decayed ttbar example, I modified the input LHE file by hand to pretend it had been generated in the following way, the goal being to suppress (from MadSpin point of view) the intermediate decays that are problematic:
> generate p p > w+ b t~ / b~
> of course, the events were in fact generated with p p > t t~, t w+ b as said before.

I would expect this to work indeed.

> (the W boson "in the matrix element" is decayed
> only in cs~ for some reason, while the other one goes in all channels),
> but I'm just curious now…

This is suspicious. If I found some time, I will try to understand that point.

Cheers,

Olivier

On Dec 2, 2013, at 2:06 PM, Julien Maurer <email address hidden> wrote:

> Question #240157 on MadGraph5 changed:
> https://answers.launchpad.net/madgraph5/+question/240157
>
> Status: Answered => Open
>
> Julien Maurer is still having a problem:
> Dear Olivier,
>
>
> Thanks for your answer. For my own curiosity, may I ask you to comment on the following:
>
> Again with this half-decayed ttbar example, I modified the input LHE file by hand to pretend it had been generated in the following way, the goal being to suppress (from MadSpin point of view) the intermediate decays that are problematic:
> generate p p > w+ b t~ / b~
> of course, the events were in fact generated with p p > t t~, t w+ b as said before.
>
> In that way, MadSpin runs without apparent trouble and indeed decays
> events. Does this approach make any sense at all (spin correlations
> should be tracked properly, but I guess it's not so simple...)? It seems
> not to work fully anyway (the W boson "in the matrix element" is decayed
> only in cs~ for some reason, while the other one goes in all channels),
> but I'm just curious now...
>
>
> Many thanks,
> Julien
>
> --
> You received this question notification because you are a member of
> MadTeam, which is an answer contact for MadGraph5.

Revision history for this message
Julien Maurer (jln-maurer) said : 		#5

OK, thank you very much for your fast answers...
I think I went too fast on the W->cs~ "issue" (sorry), I had only looked at the last 4 events but looking more there are events with decays in other channels. It still looks suspicious because this channel seems to appear too often, but I will obtain a more quantitative statement now.

Revision history for this message
Julien Maurer (jln-maurer) said : 		#6

Yes, the BRs seem correct, I got mislead by a very unfortunate coincidence. Sorry about that... So, the only issue seems to be that the mother ID of the w boson decayed by MadEvent is not properly set in the final LHE file: it is set to 1. But this is not a concern.

Sorry for the useless noise... and thanks again for your help.

Best regards,
Julien

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#7

Hi Julien,

> Yes, the BRs seem correct, I got mislead by a very unfortunate
> coincidence. Sorry about that…

Thanks to have keep us up-to-date :-)

> So, the only issue seems to be that the
> mother ID of the w boson decayed by MadEvent is not properly set in the
> final LHE file: it is set to 1.

That's probably linked to the changed of identity in the process that you made (this is at least my current guess).

Cheers,

Olivier

On Dec 2, 2013, at 3:11 PM, Julien Maurer <email address hidden> wrote:

> Question #240157 on MadGraph5 changed:
> https://answers.launchpad.net/madgraph5/+question/240157
>
> Julien Maurer posted a new comment:
> Yes, the BRs seem correct, I got mislead by a very unfortunate
> coincidence. Sorry about that... So, the only issue seems to be that the
> mother ID of the w boson decayed by MadEvent is not properly set in the
> final LHE file: it is set to 1. But this is not a concern.
>
> Sorry for the useless noise... and thanks again for your help.
>
> Best regards,
> Julien
>
> --
> You received this question notification because you are a member of
> MadTeam, which is an answer contact for MadGraph5.