SMEFT_sim very slow to generate VBF-Z diagrams at NP^2==1

Asked by Noah Clarke Hall

Hello,

Am running MadGraph5_aMC@NLO V3.2.0.

Following:

# import model /Users/noahclarkehall/Documents/SMEFTsim_U35_MwScheme_UFO

# generate p p > j j mu+ mu- QCD=0 NP^2==1

it is very slow in generating the diagrams, 48hrs so far and counting. Comparable processes (ie p p > j j mu+ mu- QCD=0 NP=0, p p > j j mu+ mu- QCD=0 NP==1, p p > mu+ mu- QCD=0 NP^2==1) generate quickly, on the order of seconds.

What is causing this and how do I fix it?

Many thanks

Noah

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

interference term are obviously harder to generate. since we need to generate a huge amount of amplitude and then
check to which it can combines. So you will have a lot of filtering to use and this likely slows down the computation

My advise here is to start by simplifying your model by using a restriction model: https://answers.launchpad.net/mg5amcnlo/+faq/2312
and remove all the coupling that you are not interested in.

This will limit the number of possible diagram/combination and should allow to speed-up both the code generation but also the running time after that.

Cheers,

Olivier

Revision history for this message
Noah Clarke Hall (noahch) said :
#2

Thanks for getting back so quickly!

When you say remove couplings are you referring to setting specific Wilson coefficients to zero?

Cheers

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

I do not know that model very well, so I did not wanted to be too precise.
The idea is everything that you do not use can be removed.
so yes Wilson coefficient is a perfect fit but also term modifying the propagator, ....

Cheers,

Olivier

> On 2 Nov 2021, at 12:31, Noah Clarke Hall <email address hidden> wrote:
>
> Question #699311 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/699311
>
> Status: Answered => Open
>
> Noah Clarke Hall is still having a problem:
> Thanks for getting back so quickly!
>
> When you say remove couplings are you referring to setting specific
> Wilson coefficients to zero?
>
> Cheers
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Noah Clarke Hall (noahch) said :
#4

Thanks Olivier Mattelaer, that solved my question.