(Much) Fewer events than requested when calculating the cross section

Asked by Qiji Xin on 2019-09-16

Dear MadGraph Experts,

I got (much) Fewer events than requested when calculating the cross section of a process.

So what could be the reason of the issue and how to fix it? Is it due to divergence or due to the inefficiency of the phase space integration?

Thanks so much!

Question information

Language:
English Edit question
Status:
Answered
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Last query:
2019-09-18
Last reply:
2019-09-20

Hi,

Yes I do believe that you have singularity for this process (namely when the photon is soft)
Additionally, The lepton mass acts as a regulator for the colinear singularity but in your case the lepton mass are so small that it is not surprising if you still have numerical issue with the regularization.

My advice here would be to at least have a cut preventing the photon to be soft and if this is not enough you also need to have a cut preventing the photon to be colinear with the leptons.

Cheers,

Olivier

On 17 Sep 2019, at 02:47, Qiji Xin <<email address hidden><mailto:<email address hidden>>> wrote:

generate a ve > e- W+ a

Qiji Xin (xinqijisuper) said : #4

A correction to what I said. There are 7 diagrams in total, instead of 10 diagrams.
Thanks!

Looks like I miss your original answer here.

Cheers,

Olivier

Hi,

Looks like (some of your) email are not put online in the thread... no sure why but it does not matter too much.

> 1) the cross sections I get depends on the photon energy that I set

Yes indeed

> 2) the "fewer events" issue still happens in CM energy is very large. Actually, the lower energy I cut on the photon, the lower CM energy that has the "fewer events" issue.

This fully makes sense as well.

> Or include the vertex correction diagrams to
> cancell the singularity?

In that case you will need to compute the cross-section of a + ν → l W
at NLO accuracy (in the weak coupling). In the best case scenario this can be allowed in the beta version 3.0.1 but it will be only allowed at fNLO and therefore you will not have any events in that case.

Additionally you are computing the cross-section of a + ν → l W
and not the one of a + ν → l W a.
So if you are considering this then maybe the best solution in your case is to compute the LO cross-section of a + ν → l W.
And trust the parton shower if you want to look at the photon spectrum

Cheers,

Olivier

> On 18 Sep 2019, at 23:07, Qiji Xin <email address hidden> wrote:
>
> Question #683970 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/683970
>
> Qiji Xin gave more information on the question:
> Below is my original answer. Thanks!
>
> =========================================================================================================
> Hi Olivier,
>
> Thanks for your reply!
> I did cuts on the photon energy (set ea XXX), and it can generated the number of events I request, mostly.
>
> However, the issue is not totally fixed.
> 1) the cross sections I get depends on the photon energy that I set
> 2) the "fewer events" issue still happens in CM energy is very large. Actually, the lower energy I cut on the photon, the lower CM energy that has the "fewer events" issue.
>
>> From the QFT book I know that, for the infrared singularity, the photon-
> emission diagrams should be cancelled by the vertex correction diagrams.
>
> My process, a + ν → l W a, at tree level, has both the singularity
> diagrams and non-singularity diagrams (10 diagrams in total), as
> generated by MadGraph.
>
> So how to make my cross section results physical? Choose an appropriate
> photon energy to cut? Or include the vertex correction diagrams to
> cancell the singularity? or?
>
> Thanks so much!
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Can you help with this problem?

Provide an answer of your own, or ask Qiji Xin for more information if necessary.

To post a message you must log in.