S-channel stop pair production

Hi,

Diagram filtering (which you are using by stating that you want only some diagram and not all of them)
can indeed be unphysical since they often break gauge invariance (and consequently lorentz invariance)
Even if you do not break gauge invariance, only the sum of the piece might be physical (including the interference).

Obviously they are case where diagram filtering dos make sense, especially if you do impose cuts to avoid parts of the phase-space where the interference are large, gauge invariance is problematic, ...

So the issue with the syntax
> p p > X Y Z > t1 t1~

is indeed that such syntax is not safe and that if you use it you have to ensure that you can use it.
(which in general you can not)

We typically prefer to suggest syntax like
p p > X , X > t1 t1~
since such syntax are based on the narrow-width approximation and is associated to cut to ensure that the NWA is valid (i.e. a cut on the invariant mass of the X particle)

Now I would not be able to answer your question since I do not know the benmark and the analysis cut that you are going to use. This is crucial to assess how much valid the diagram filtering is going to be. But if your question is how to enforce particle in S-channel and that I can assume that all physics consideration are fine. Then yes the syntax p p > X Y Z > t1 t1~ is a good start. If you need more advanced diagram filtering, you can also check our FAQ that has one section on how to define a python function to do such filtering (with the same caveat that you have to understand/assume the implication in term of physical meaning).

Cheers,

Olivier

> On 1 Nov 2023, at 17:30, Fan Ka Yan <email address hidden> wrote:
>
> New question #708343 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/708343
>
> Dear experts,
>
> I am trying to produce the s-channel pair production with
>
> import model MSSM_SLHA2
> generate p p > t1 t1~
> .
>
> It is expected that there are both s and t channels in the output Feynman diagrams.
> I read some documents that I could use
> generate p p > ... > t1 t1~ in which ... are some propagators. But the MG_aMC@NLO paper mentioned that this result will be non physical.
>
> May I ask is there other ways to just produce the s-channel one with correct physical results, e.g. cross-sections?
>
> Thanks,
> KaYan
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.