Very few events are being produced in newer MG5 versions

i am also want to know more about it . i am share this on my site to know about about it .https://reviewires.com/best-monitors-for-eye-strain/

Hi,

Personally, I do get the requested number of events.

Now you can look at 2102.00773 [hep-ph]
where we describe all the new features of 2.9.x for such type of process and how to turn them off (and therefore on how to get back to 2.8.x integration method).

Note also that they are a physical change that you might be sensitive to between 2.7.x and 2.8.x is the definition of the T-channel propagator (2.7.x does include the width while 2.8.x does not). If you are sensitive to that, it likely means that you are sensitive to non gauge invariant piece of the computation and that you will need to use formalism like complex mass-scheme to avoid such issue.

Cheers,

Olivier

Thanks, Olivier, for the comments.

So I have tested the process in MG5 versions 2.6.7, 2.7.3, 2.8.3.2, 2.9.10, and 3.4.0. It could generate all the events up to version 2.8.3.2. But from version 2.9.X it can't generate all the events. I also tried by setting 'zerowidth_tchannel' to 'False' in versions 2.9.10 and 3.4.0. But the same thing. So, I suppose this issue is not related to the width of the t-channel propagator. What do you think?
May I know in which version you got the requested number of events?
The script I used in versions 2.X.Y is as follows:
=================
generate e- e+ > vl vl~ l- l+
output ; launch
done
set ebeam 500
set cut_decays True
set ptl 2
set etal 3.5
set drll 0.2
done
=================

Regards,
Saumyen

Oh sorry, I was running at 100TeV not at 1 TeV.

Let me retry

Olivier

Ok at that lower energy, it seems clear

That you have to force the "old" phase-space integration strategy since it is working much better than the new one.
You can set in the run_card sde_strategy on "1" and you should get back the speed of the 2.7.x/2.8.x.
If you want more detail on that parameter, I would refer you again to the paper that I mentioned before: 2102.00773 [hep-ph]

Cheers,

Olivier

Okay, thanks a lot Olivier. That worked fine.

Just a question. With this choice of sde_strategy change the event configurations? Because a lot of my signal files are already generated with the previous configuration without any issue (I don't want to regenerate). I guess it'll be fine (since it appears at the phase-space integration level), right?

Thanks a lot.
Regards,
Saumyen

Hi,

So this change only the way the phase-space is probed.
So this should not impact the (differential) cross-section.

Now this might not be always True since some algorithm are based on the channel of integration.
(like the determination of the scale and MLM factor or the writting of some non gauge invariant information within the leshouches file)
So for QCD related process (and the default scale) the cross-section will be within scale uncertainty.

So yes, this should be fine (or at least within theoretical uncertainty)

Cheers,

Olivier

> On 26 Jul 2022, at 11:05, Saumyen Kundu <email address hidden> wrote:
>
> Question #702544 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/702544
>
> Saumyen Kundu posted a new comment:
> Okay, thanks a lot Olivier. That worked fine.
>
> Just a question. With this choice of sde_strategy change the event
> configurations? Because a lot of my signal files are already generated
> with the previous configuration without any issue (I don't want to
> regenerate). I guess it'll be fine (since it appears at the phase-space
> integration level), right?
>
> Thanks a lot.
> Regards,
> Saumyen
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Thank you so much, Olivier for the comments. It was helpful.

Regards,
Saumyen