MadGraph5_aMC@NLO

Create 6 lepton final state?

Asked by Sara Aumiller

Hi!

This is partially a follow-up question of https://answers.launchpad.net/mg5amcnlo/+question/821199

I am interested in generating a ee-> 6l (and ee-> 4l 2j) background for my ee->ZH, H->ZZ 6 lepton final state signal.

I've tried this with whizard, but even with l=e, mu only, generating the events takes more than 100 days.

That's why I had a look at creating the background via the Zs (see https://answers.launchpad.net/mg5amcnlo/+question/821199 ) but getting the right cross section with all the combinatorics is a bit tedious.

Could you give me an advice on how to generate ee->6 (and ee->4l 2j) excluding any higgs so that the production takes max 1 week? Are there any cuts you can recommend doing?

Thank you so much for your intuition on this, your help is very much appreciated!

Best,
Sara

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
Olivier Mattelaer (olivier-mattelaer) said : 		#1

Hi,

It is difficult to help without knowing what is your exact signal definition.
In particular, how do you handle the interference term between (ZH) and the 6lepton case?
Is that part of your signal? part of your background? Or just neglected?

Second important point is what is the constraint on the invariant mass for your signal?
Do you request exactly two onshell Z? or at least two onshell Z (or no constraint,...)

One fast way to compute this is:
e+ e- > z z z / h (or $h), z > l+ l-
and maybe enlarge a bit the z window
But this might not be enough depending of your cut.

Then you might need do another sample (the exactly two resonant one)
e+ e- > l+ l- z z / h (or $h) $z, z > l+ l-
and multiply the cross-section of that sample by three.

I guess that you will not need the exactly one single resonant one (but maybe you do).

But I would advise that you check on small sample that you do have the correct combinatorial since this can be tricky...

Revision history for this message
Sara Aumiller (saraaumiller22) said : 		#2

Hi Olivier!

Thanks for your fast reply.

1) We just neglect the interference term between the ZH and 6 lepton case for now.
2) We will have 60 (20) events for l=e, mu, tau (l=e, mu). Until now counting 6 leptons with correct flavor and charge was enough to eliminate other background. But we have also tried to have a loose cut on the invariant mass of one on shell Z and one of shell Z.
3) This will not work, because as mentioned in my follow-up question here (https://answers.launchpad.net/mg5amcnlo/+question/821199) we run at 240 GeV which is not enough to create 3 on-shell Zs.
4) so you suggest doing the combinatorics? Is there no way to generate ee->6l in a reasonable time with madgraph?

Revision history for this message
Best Olivier Mattelaer (olivier-mattelaer) said : 		#3

The combinatoric is certainly something that is manageable in a very quick way.
(i.e. couple of minuts for 10k events)

Now 6 leptons has quite a lot of diagrams and is indeed seemed quite complex to evaluate.
I never done such computation so I have no idea how fast/slow it will be.

Cheers,

Olivier

> On 13 Mar 2025, at 16:01, Sara Aumiller <email address hidden> wrote:
>
> Question #821225 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/821225
>
> Status: Answered => Open
>
> Sara Aumiller is still having a problem:
> Hi Olivier!
>
> Thanks for your fast reply.
>
> 1) We just neglect the interference term between the ZH and 6 lepton case for now.
> 2) We will have 60 (20) events for l=e, mu, tau (l=e, mu). Until now counting 6 leptons with correct flavor and charge was enough to eliminate other background. But we have also tried to have a loose cut on the invariant mass of one on shell Z and one of shell Z.
> 3) This will not work, because as mentioned in my follow-up question here (https://answers.launchpad.net/mg5amcnlo/+question/821199) we run at 240 GeV which is not enough to create 3 on-shell Zs.
> 4) so you suggest doing the combinatorics? Is there no way to generate ee->6l in a reasonable time with madgraph?
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Sara Aumiller (saraaumiller22) said : 		#4

Thanks Olivier Mattelaer, that solved my question.