25% of events vetoed for inclusive process

Hello Rakhi,

This is perfectly correct. Whenever you have QCD radiated jets (as opposed to jets from decays e.g.), there is an associated Sudakov form factor due to required non-radiation of jets between the factorization scale (2->2 scale) and the scale of the softest jet. This is correctly taken into account if you run with ickkw=1, and results in a vetoing of events where the shower produces jets that are harder than the softest matrix element jet. In fact, you should always run with ickkw=1 whenever there are radiated jets in the process, whether or not you include multiple parton multiplicities.

All the best,
Johan

Dear Johan,

I'm not sure I understand: this is the result for the pure 4j hard process, which is my highest multiplicity sample, so I'm setting this to shower inclusively in pythia. In this case I don't want any additional jets from the shower to be vetoed, no matter how hard, so there should be no sudakov suppression. There are no 3j or 2j processes in this sample (I'm running them separately, and do expect vetoes in those cases).

Shouldn't the parton shower be unitary for the highest multiplicity process?

Cheers,
Rakhi

Hello Rakhi,

The parton shower starts from the central (2->2) scale. This means that there is a probability (given by the Sudakov form factor) that it produces shower emissions that are harder than the softest of the parton-level jets in the same event (the 4th jet). Such an event double-counts the corresponding event where this (harder) jet is generated by the matrix element. Due to this double-counting, the parton shower should only be allowed to generate jets below the scale of the softest jet in the matrix element. The events that are rejected in this procedure corresponds precisely to the Sudakov form factor (non-shower probability) from the central (2->2) scale to the scale of the softest jet (event by event). It is *only* in the leading order process (for multijets, the 2->2 process, for W production, the 2->1 process, etc) that the matching procedure should be unitary.

But in any case, if you don't want the Sudakov suppression, you can simply set the ickkw flag in the run_card to 0. Then, there is no vetoing and hence no such correction.

All the best,
Johan

Sorry, I still don't understand, because I'm not generating a sample with a harder jet from the matrix element, this is really my highest multiplicity sample.  I am in the process of doing exactly what you said, which is turning ickkw off, and I think that should tell me whether the problem is with the madgraph veto, or with pythia...

In every other example where I have a Born level process with a fixed number of partons in the final state, and I did matching with *inclusive* showering in pythia, the cross section has been the same post-Pythia, i.e. no sudakov suppression and unitary parton shower.

Alternatively, what is the 2->2 scale you're referring to in our case, where the only process is 2->4?

Cheers,

Rakhi

=========================
Chemin de l'Esplanade 26
1214 Vernier
Suisse

Tel: (+41) 22 341 29 60
Mob: (+41) 78 320 41 46
==========================
==========================

"One does not discover new lands without consenting to lose
sight of the shore for a very long time." - Andre Gide

________________________________
 From: Johan Alwall <email address hidden>
To: <email address hidden>
Sent: Thursday, June 21, 2012 4:31 PM
Subject: Re: [Question #200925]: 25% of events vetoed for inclusive process

Your question #200925 on MadGraph5 changed:
https://answers.launchpad.net/madgraph5/+question/200925

    Status: Open => Answered

Johan Alwall proposed the following answer:
Hello Rakhi,

The parton shower starts from the central (2->2) scale. This means that
there is a probability (given by the Sudakov form factor) that it
produces shower emissions that are harder than the softest of the
parton-level jets in the same event (the 4th jet). Such an event double-
counts the corresponding event where this (harder) jet is generated by
the matrix element. Due to this double-counting, the parton shower
should only be allowed to generate jets below the scale of the softest
jet in the matrix element. The events that are rejected in this
procedure corresponds precisely to the Sudakov form factor (non-shower
probability) from the central (2->2) scale to the scale of the softest
jet (event by event). It is *only* in the leading order process (for
multijets, the 2->2 process, for W production, the 2->1 process, etc)
that the matching procedure should be unitary.

But in any case, if you don't want the Sudakov suppression, you can
simply set the ickkw flag in the run_card to 0. Then, there is no
vetoing and hence no such correction.

All the best,
Johan

--
If this answers your question, please go to the following page to let us
know that it is solved:
https://answers.launchpad.net/madgraph5/+question/200925/+confirm?answer_id=2

If you still need help, you can reply to this email or go to the
following page to enter your feedback:
https://answers.launchpad.net/madgraph5/+question/200925

You received this question notification because you asked the question.

Let me take as a clearer example W production. The Born process is (with W = W+/-)
p p > W
If you generate, say,
p p > W j
then the jet is generated from QCD radiation, meaning that it is generated from ggg or gqqbar vertices (well, in this case only gqqbzr of course). Imagine that we have an event with a jet pt of 20 GeV. The factorization scale of the process is related to the transverse mass of the W, m_{TW} = sqrt(80^2+20^2)=82 GeV. This is the starting scale for the parton shower. However, we can't allow any parton shower jets above 20 GeV (since that would correspond to a different matrix element event, and therefore give double counting), giving a Sudakov suppression corresponding to \Delta(82 GeV, 20 GeV) (please see my lectures
https://server06.fynu.ucl.ac.be/projects/madgraph/attachment/wiki/SchoolNTU/NTU-MLM-lectures.pdf
for notation and many more details). This Sudakov suppression is precisely what's needed to regulate the matrix element soft/collinear divergence as p_{Tj} -> 0, and is generated in MLM matching by vetoing events where the parton shower generates too hard jets.

In your case, the Born process is p p > j j, and the p p > j j j j process is correspondingly regulated by the matching prescription. Please read the lectures above, as well as some of the literature on matrix element-parton shower matching, e.g. http://arxiv.org/abs/arXiv:0810.5350, http://arxiv.org/abs/hep-ph/0205283.

All the best,
Johan

Hey Johan,

I think I finally understood your previous emails - for some reason I was consistently misinterpreting your statements in the light of my bias, which was that double-counting could only occur between samples of different multiplicities, and not within a sample of a single multiplicity.

I agree now that there should be events that are vetoed when I shower my highest-multiplicity (MadEvent-generated) sample in Pythia with the inclusive flag turned on, but shouldn't the cross section still be the same post-showering (i.e. shouldn't the reduced cross section after the veto be rescaled to match the original cross section for the hard process)?  This does not seem to be happening, since the post-shower cross section in the pythia log exactly reflects the 25% of events that are thrown away by the veto (i.e. the cross-section after showering is 25% lower than the hard cross section).

Cheers,

Rakhi

=========================
Chemin de l'Esplanade 26
1214 Vernier
Suisse

Tel: (+41) 22 341 29 60
Mob: (+41) 78 320 41 46
==========================
==========================

"One does not discover new lands without consenting to lose
sight of the shore for a very long time." - Andre Gide

________________________________
 From: Johan Alwall <email address hidden>
To: <email address hidden>
Sent: Friday, June 22, 2012 4:01 AM
Subject: Re: [Question #200925]: 25% of events vetoed for inclusive process

Your question #200925 on MadGraph5 changed:
https://answers.launchpad.net/madgraph5/+question/200925

    Status: Open => Answered

Johan Alwall proposed the following answer:
Let me take as a clearer example W production. The Born process is (with W = W+/-)
p p > W
If you generate, say,
p p > W j
then the jet is generated from QCD radiation, meaning that it is generated from ggg or gqqbar vertices (well, in this case only gqqbzr of course). Imagine that we have an event with a jet pt of 20 GeV. The factorization scale of the process is related to the transverse mass of the W, m_{TW} = sqrt(80^2+20^2)=82 GeV. This is the starting scale for the parton shower. However, we can't allow any parton shower jets above 20 GeV (since that would correspond to a different matrix element event, and therefore give double counting), giving a Sudakov suppression corresponding to \Delta(82 GeV, 20 GeV) (please see my lectures
https://server06.fynu.ucl.ac.be/projects/madgraph/attachment/wiki/SchoolNTU/NTU-MLM-lectures.pdf
for notation and many more details). This Sudakov suppression is precisely what's needed to regulate the matrix element soft/collinear divergence as p_{Tj} -> 0, and is generated in MLM matching by vetoing events where the parton shower generates too hard jets.

In your case, the Born process is p p > j j, and the p p > j j j j
process is correspondingly regulated by the matching prescription.
Please read the lectures above, as well as some of the literature on
matrix element-parton shower matching, e.g.
http://arxiv.org/abs/arXiv:0810.5350, http://arxiv.org/abs/hep-
ph/0205283.

All the best,
Johan

--
If this answers your question, please go to the following page to let us
know that it is solved:
https://answers.launchpad.net/madgraph5/+question/200925/+confirm?answer_id=4

If you still need help, you can reply to this email or go to the
following page to enter your feedback:
https://answers.launchpad.net/madgraph5/+question/200925

You received this question notification because you asked the question.

Hello Rakhi,

The cross section should go down - this is the effect of the Sudakov form factor (i.e., the fact that we need to forbid radiation between the factorization scale and the lowest jet scale). Again - it is strictly speaking only for Born processes without any jet radiation diagrams that the cross section is fully physical - whenever a cut is needed to shield from collinear and soft divergencies, the cross section is physical only after application of the Sudakov factor. This is what's done by the vetoing. Please see the references in my previous post.

All the best,
Johan