Jet Clustering and Merging Scale Parameters Questions

Asked by Zachary Warner on 2020-05-29

Hello,

I am currently wanting to run an inclusive dijet production at NLO using madgraph 5 v 2.7.2 and I am showering with PYTHIA8. The specific process I'm running is
p p > j j j [QCD]

My question pertains to the jet merging and clustering set up. It says on the madgraph website to use ickkw = 1 for MLM merging and ickkw = 2 for CKKW merging. However my version of madgraph only offers FXFX merging (ickkw =3), UNLOPS merging (ickkw = 4), and NNLL+NLO jet-veto computation (ickkw = -1). I have noticed that none of these are compatible with the anti-kt jet algorithm, only the kt jet algorithm. Is FXFX merging with the kt algorithm the best way to go about jet matching for inclusive dijets, or should I download a different version of madgraph to obtain MLM merging?

I was also wondering about the merging scale parameter in the showering process. I am using minimum transverse jet momentum (ptj) of 25 GeV, and I read on another madgraph forum that your merging scale should be about 20% higher than your pt min set for jets, but this was for e+e- collisions. Is this true for pp collisions as well? I couldn't find much information on merging scales and their function.

Thanks in advance for any help, and I apologize if these questions are amateurish.

-Zach Warner

Question information

Language:
English Edit question
Status:
Answered
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Last query:
2020-06-10
Last reply:
2020-06-10

HI Zachary,

MLM and CKKW-L are LO merging method. Such method can not be used at NLO.
However you have the equivalent of such method at NLO.
FxFx follows the same line (strength and weakness) of the MLM method but is valid at NLO accuracy.

So if you want to use MLM, you have to drop NLO, or if you want to keep NLO, you have to drop MLM.

However, I do not think that FxFx can handle dijet, as far as I know, FxFx does not support jets at the lower multiplicity. (but you can try if it works or not obviously)
So I guess the best that we can do is MLM
generate p p > j j
add process p p > 3j
add process p p > 4j

Since you have jets at the lowest multiplicity, the setup for this MLM is not trivial,
you certainly need to set
auto_ptj_mjj to False
and setup carefully xqcut and qcut for this process.
You will also need to be carefull with the pythia8 card to check that everything is configure as it needs.

For the scale choice, you have to test different value and check which one works.
In general our guideline is to take the hard-scale of your process (I guess 2*ptcut in your case) and then divide that scale by a number between 2 and 6. So have a cut above the pt cut is not in our guideline, but ok.

Cheers,

Olivier

Zachary Warner (zwarner) said : #2

Thank you for your reply, Olivier. This was helpful. One more question with regards to xqcut. It seems like the choice of this value is not available at NLO were FXFX merging method is the only option. In this case I can only chose the value of ptj_min. What is the relation, if any between xqcut and ptj_min? I have read that xqcut is the minimum jet measure (pT/kT) for QCD partons. So for a kT value chosen to be 1 would xqcut = ptj_min?

Also, when choosing a value of Qcut (the merging scale) in PYTHIA it will not run the showering process unless a value above ptj_min is chosen. Therefore, it is not possible to have a value for the merging scale that is not above the pt cut.

Thanks again for your help.

Zach Warner

Hi,

For the first jet the two variable xq and ptj are the same.
since the only jet direction are the beams one.
When you start to have more jet, then you have additional direction related to the other jets and the definition differs.

Cheers,

Olivier

> On 10 Jun 2020, at 18:20, Zachary Warner <email address hidden> wrote:
>
> Question #691037 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/691037
>
> Status: Answered => Open
>
> Zachary Warner is still having a problem:
> Thank you for your reply, Olivier. This was helpful. One more question
> with regards to xqcut. It seems like the choice of this value is not
> available at NLO were FXFX merging method is the only option. In this
> case I can only chose the value of ptj_min. What is the relation, if any
> between xqcut and ptj_min? I have read that xqcut is the minimum jet
> measure (pT/kT) for QCD partons. So for a kT value chosen to be 1 would
> xqcut = ptj_min?
>
> Also, when choosing a value of Qcut (the merging scale) in PYTHIA it
> will not run the showering process unless a value above ptj_min is
> chosen. Therefore, it is not possible to have a value for the merging
> scale that is not above the pt cut.
>
> Thanks again for your help.
>
> Zach Warner
>
> --
> 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 Zachary Warner for more information if necessary.

To post a message you must log in.