Modify the matrix element after 'output' command

Asked by Marco H.

Hello!

Is it possible to modify the matrix element that is used for the calculation before running the launch command, i.e. after the output command?

One reason I ask is because I have to calculate the phase-space for a 2->3 process with a constant matrix element. The latter comes from a manually calculate loop-process, which cannot be calculated in MadGraph itself. In a different scenario, I have to multiply the final matrix element by an additional momentum-dependent factor.

For the first problem, my current idea is to generate some process that has the correct initial and final states (but a non-constant matrix element), and afterwards simply set the matrix element to 1, as this should give me the phase-space contribution.

Is there an easy way of performing this modification of the matrix element? I know that there are files like matrix1_orig.f and myamp.f, which I can modify in principle. However, I am worried that this is not enough and may lead to inconsistent results.

Or is there maybe a general way to access the phase-space sampler of MadGraph with a custom matrix element?

Thanks a lot in advance for your help!

Best,
Marco

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,

> Is it possible to modify the matrix element that is used for the calculation before running the launch command, i.e. after the output command?

Possible yes for sure, just edit the fortran file.

> For the first problem, my current idea is to generate some process that has the correct initial and final states (but a non-constant matrix element), and afterwards simply set the matrix element to 1, as this should give me the phase-space contribution.

This is likely not going to work since the phase-space integrator is configured for integrating matrix element.
Removing the matrix-element will therefore lead to huge issue which means that the result can not be guarantee.

> The latter comes from a manually calculate loop-process, which cannot be calculated in MadGraph itself. In a different scenario, I have to multiply the final matrix element by an additional momentum-dependent factor.

The cleanest here is likely that you define a bias function (or that you use form-factor within your model)

Cheers,

Olivier

> On 3 Jun 2021, at 14:30, Marco H. <email address hidden> wrote:
>
> New question #697373 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/697373
>
> Hello!
>
> Is it possible to modify the matrix element that is used for the calculation before running the launch command, i.e. after the output command?
>
> One reason I ask is because I have to calculate the phase-space for a 2->3 process with a constant matrix element. The latter comes from a manually calculate loop-process, which cannot be calculated in MadGraph itself. In a different scenario, I have to multiply the final matrix element by an additional momentum-dependent factor.
>
> For the first problem, my current idea is to generate some process that has the correct initial and final states (but a non-constant matrix element), and afterwards simply set the matrix element to 1, as this should give me the phase-space contribution.
>
> Is there an easy way of performing this modification of the matrix element? I know that there are files like matrix1_orig.f and myamp.f, which I can modify in principle. However, I am worried that this is not enough and may lead to inconsistent results.
>
> Or is there maybe a general way to access the phase-space sampler of MadGraph with a custom matrix element?
>
> Thanks a lot in advance for your help!
>
> Best,
> Marco
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Marco H. (hep-hufnagel) said :
#2

Hi,

> This is likely not going to work since the phase-space integrator is configured for integrating matrix element.
Removing the matrix-element will therefore lead to huge issue which means that the result can not be guarantee.

Ok, in this case I will just calculate the phase-space by hand. ;P

> The cleanest here is likely that you define a bias function (or that you use form-factor within your model)

Thanks a lot. This might indeed be the best solution.

Best,
Marco

Revision history for this message
Marco H. (hep-hufnagel) said :
#3

Thanks Olivier Mattelaer, that solved my question.