MadGraph5_aMC@NLO

Gluons from ME and Pythia

Asked by Han

Dear MadGraph5_aMC@NLO team,

I would like to know the difference between below two cases.

1) Explicitly demanding gluons in decay legs on ME module (1 = ickkw, with some xqcut)
generate e+ e- > ul ul~ , ul > u n1 g, ul~ > u~ n1 g
then showering the process using Pythia (with some Qcut; here we set xqcut~ Qcut )

2) No gluon on ME module
generate e+ e- > ul ul~ , ul > u n1, ul~ > u~ n1 (0 = ickkw)
then showering the process using Pythia

I know that ME will generate gluons with a large kT distance with respect to u quarks, while Pythia will generate multiple QCD radiations with a small kT distance.

The method 1 relies the first hard gluon emission on ME module. Then let Pythia take it over.

The method 2 entirely relies QCD radiations on Pythia.

Which method is correctly processed given the current ME-Pythia chain?

Best regards

Jeong Han Kim

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,

parton-shower should be used for low scale radiation while matrix-element should be used for hard-scale radiation.
you can merge both approach by generating multiple multiplicity at the matrix-element level and merge them thanks to a matching scale
(via MLM algorithm or CKKW-L,…)

> 1) Explicitly demanding gluons in decay legs on ME module (1 = ickkw, with some xqcut)
> generate e+ e- > ul ul~ , ul > u n1 g, ul~ > u~ n1 g
> then showering the process using Pythia (with some Qcut; here we set xqcut~ Qcut )

In this case, using ickkw=1/MLM is basically useless since you have only one multiplicity at matrix-element level.

> Which method is correctly processed given the current ME-Pythia chain?

Both are valid in some part of the phase-space and wrong in some other.
Depending of your analysis, you should sometimes use one or the other (or merge them via one of the above algorithm)

I would suggest that you learn more about QCD/matching/merging algorithm.
One example of lecture is the following one:
https://cp3.irmp.ucl.ac.be/projects/madgraph/wiki/Lund2014
(Lecture 5 and 6)

Cheers,

Olivier

> On Dec 18, 2016, at 08:28, Jeong Han Kim <email address hidden> wrote:
>
> New question #407562 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/407562
>
> Dear MadGraph5_aMC@NLO team,
>
> I would like to know the difference between below two cases.
>
> 1) Explicitly demanding gluons in decay legs on ME module (1 = ickkw, with some xqcut)
> generate e+ e- > ul ul~ , ul > u n1 g, ul~ > u~ n1 g
> then showering the process using Pythia (with some Qcut; here we set xqcut~ Qcut )
>
> 2) No gluon on ME module
> generate e+ e- > ul ul~ , ul > u n1, ul~ > u~ n1 (0 = ickkw)
> then showering the process using Pythia
>
> I know that ME will generate gluons with a large kT distance with respect to u quarks, while Pythia will generate multiple QCD radiations with a small kT distance.
>
> The method 1 relies the first hard gluon emission on ME module. Then let Pythia take it over.
>
> The method 2 entirely relies QCD radiations on Pythia.
>
> Which method is correctly processed given the current ME-Pythia chain?
>
> Best regards
>
> Jeong Han Kim
>
>
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Han (jeonghan-kim) said : 		#2

Thanks Olivier Mattelaer, that solved my question.

Revision history for this message
Han (jeonghan-kim) said : 		#3

Dear Olivier,

Oh yes... Thanks. I'm trying

generate e+ e- > ul ul~ , ul > n1 u , ul~ > n1 u~
add process e+ e- > ul ul~ , ul > n1 u g, ul~ > n1 u~
add process e+ e- > ul ul~ , ul > n1 u , ul~ > n1 u~ g
add process e+ e- > ul ul~ , ul > n1 u g , ul~ > n1 u~ g

and varying xqcut and QCUT together to check a continuity of DJR plots.

Best regards

Jeong Han Kim

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#4

Dear Jeong,

We can not run the matching/merging for QCD radiation coming from an on shell decay.
The problem is that you have a non trivial cut on the invariant mass of the particles which forbids to
have a smooth transition between parton-shower mode and matrix-element mode.

So if you are interested in such process you need to replace your process by a syntax without the decay-chain.

Cheers,

Olivier

> On Dec 19, 2016, at 04:48, Jeong Han Kim <email address hidden> wrote:
>
> Question #407562 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/407562
>
> Jeong Han Kim posted a new comment:
> Dear Olivier,
>
> Oh yes... Thanks. I'm trying
>
> generate e+ e- > ul ul~ , ul > n1 u , ul~ > n1 u~
> add process e+ e- > ul ul~ , ul > n1 u g, ul~ > n1 u~
> add process e+ e- > ul ul~ , ul > n1 u , ul~ > n1 u~ g
> add process e+ e- > ul ul~ , ul > n1 u g , ul~ > n1 u~ g
>
> and varying xqcut and QCUT together to check a continuity of DJR plots.
>
> Best regards
>
> Jeong Han Kim
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Han (jeonghan-kim) said : 		#5

Dear Olivier,

Your first sentence `` We can not run the matching/merging for QCD radiation coming from an on shell decay." solved my problem.

I thank you again.

Best regards

Jeong-Han Kim