MadGraph5_aMC@NLO

Difference between simulate all process in one generate or use add process

Asked by Francisco Casalinho

Hello,

What is the difference between generating all processes with one command or using add process to add all the other processes?
In my case, I'm simulating the Higgs production via the s- and t-channel using collisions between e+ and e- and for that, I simulated in two ways:

1) generate e+ e- > z > e+ e- h ( s-channel)
     add process e+ e- > e+ e- h $$z. ( t-channel)
getting a cross-section of = 0.00648 ± 1.829*10^-5 [pb]

2) generate e+ e- > e+ e- h ( s- and t- channel )

getting a cross-section of = 0.00673 ± 1.8*10^-5 [pb]

Despite the fact that booth cross-section values are close, they have more than 1 sigma of difference ( more precisely, they have 13.6 sigmas of difference). Why? Is possible that the first method does not include some constructive interference between the s-channel and the t-channel?

 Also, I did not modify my param.card but I get the warning: ( this happens when I use the MadAnalysis4 or MA5)
"Failed to update dependent parameter. This might create trouble for external program (like MadSpin/shower/...) "
it's possible it's the MadGraph version?

Cheers,

Francisco

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,

Yes your first syntax does compute S channel and T channel completely independently (as you can see from the HTML/info.html page and therefore no interference term is included.

Addtionally if you use dynamical_scale_choice =-1 (often the default) both computation will be assigned to a different dynamical scale choice. Therefore even if the interference was exactly 0 you could not compare those two number up to statistical uncertainty, you also have to add the theoretical uncertainty.
You can use another dynamical_scale_choice (or a fix scale choice) for such comparison obviously.

Cheers,

Olivier

> On 2 Mar 2023, at 23:45, Francisco Casalinho <email address hidden> wrote:
>
> New question #705691 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/705691
>
> Hello,
>
> What is the difference between generating all processes with one command or using add process to add all the other processes?
> In my case, I'm simulating the Higgs production via the s- and t-channel using collisions between e+ and e- and for that, I simulated in two ways:
>
> 1) generate e+ e- > z > e+ e- h ( s-channel)
> add process e+ e- > e+ e- h $$z. ( t-channel)
> getting a cross-section of = 0.00648 ± 1.829*10^-5 [pb]
>
> 2) generate e+ e- > e+ e- h ( s- and t- channel )
>
> getting a cross-section of = 0.00673 ± 1.8*10^-5 [pb]
>
> Despite the fact that booth cross-section values are close, they have more than 1 sigma of difference ( more precisely, they have 13.6 sigmas of difference). Why? Is possible that the first method does not include some constructive interference between the s-channel and the t-channel?
>
> Also, I did not modify my param.card but I get the warning: ( this happens when I use the MadAnalysis4 or MA5)
> "Failed to update dependent parameter. This might create trouble for external program (like MadSpin/shower/...) "
> it's possible it's the MadGraph version?
>
>
> Cheers,
>
> Francisco
>
>
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Francisco Casalinho (francisco27) said : 		#2

Thanks Olivier Mattelaer, that solved my question.