MadGraph5_aMC@NLO

Integrated run of Delphes+Pythia on NLO computation

Asked by David Englert

Dear MadGraph Developers,

I am wondering about the following:
The MG5 menu automatically offers the option to switch ON/OFF the running of Delphes & Pythia right after the MG5 event sample is generated, for events involving tree-level and loop-induced diagrams ('[noborn=QCD]'). However this option is not there for events involving general NLO diagrams ( i.e. adding the option '[QCD]' )

Is this purely because it's difficult to implement the automatic integration of Pythia & Delphes in this scenario, or there is some deeper underlying reason?
As there is no feature to do this seamlessly, is there any nice & easy way to run Pythia + Delphes on MG5 NLO events manually?

Thank you for your help.

Best regards,
David Englert

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 David,

Running a shower is indeed not trivial at NLO, since you have a double counting between the real emission and with the first parton-shower
emission coming from the Born.
So we use the MC@NLO method to remove this double counting but this as the impact that we need to know in advance which parton shower you are going to use such that we can apply the correct counter-term.
The specification of which parton-shower you want to use is part of the run_card (parameter called parton_shower).

Now, it is actually pretty simple to run a shower at NLO. Since instead of a flag PYTHIA=ON/OFF
you have at NLO: shower=ON/OFF.
If put it on ON (the default), then we run the directly the parton-shower (if the code is installed obviously)

If you put it on OFF, the event file produced needs to be shower by the program that you specify in the run_card.
Be carefull that running with any other parton shower will not be physical correct.

Concerning Delphes, they are no good reason anymore why it is not interfaced.
The easiest is to run the Delphes code in standalone on the stdhep or hepmc file created by the parton-shower.

Cheers,

olivier

> On Nov 11, 2016, at 12:53, David Englert <email address hidden> wrote:
>
> New question #403984 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/403984
>
> Dear MadGraph Developers,
>
> I am wondering about the following:
> The MG5 menu automatically offers the option to switch ON/OFF the running of Delphes & Pythia right after the MG5 event sample is generated, for events involving tree-level and loop-induced diagrams ('[noborn=QCD]'). However this option is not there for events involving general NLO diagrams ( i.e. adding the option '[QCD]' )
>
> Is this purely because it's difficult to implement the automatic integration of Pythia & Delphes in this scenario, or there is some deeper underlying reason?
> As there is no feature to do this seamlessly, is there any nice & easy way to run Pythia + Delphes on MG5 NLO events manually?
>
>
> Thank you for your help.
>
> Best regards,
> David Englert
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
David Englert (ganzilion) said : 		#2

Thanks Olivier Mattelaer, that solved my question.