Different xsec between MG and MadGolem

Asked by Alexis Kalogeropoulos

Hello

I am trying the following process

p p > t1 x1- / t2 , with 0,1,2j (and the the charge conjugates, in total 6 processes).

My question is that the xsec I get from MG (LO) is way different from the
one I get from MadGolem. I know that the latter is deprecated as a project, but still, I want
to clarify if it is to be trusted (given that it claims to be NLO) or not.

In particular, I am considering a SMS with the production of fully RH stops while particles other than C1, t1 are decoupled , using the MSSM_SLHA2-no_b_mass model under MG_2.6.5

stop =240 C1 = 140 : MG + pythia8 after matching ~ 41 pb , MadGolem ~ 1.4 pb (NLO)

However, if I just generate only 0 partons, my MG xsec seems to be in better agreement, which is ~ 1.3 (I can attribute the remaining difference to the different pdf).

So, it seems that my MG xsec when considering additional partons after matching is way off from the NLO bare value. Is there an explanation what could cause that ?

Thanks

Alexis

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
Olivier Mattelaer (olivier-mattelaer) said :
#1

Hi,

You might have some non radiative process added by your 1 and/or 2 jets diagrams that are not technically part of the pure QCD radiation. Did you check that.

Cheers,

Olivier

Revision history for this message
Alexis Kalogeropoulos (alkaloge) said :
#2

Hi Olivier

I do see indeed, that the 2j processes

 g g > t1 x1- g b~

g g > t1~ x1+ g b

 give the largest contribution to the xsec. So, what would you suggest to do ? I mean, the xsecs I get from these 2j process do not seem physical.

Thanks

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said :
#3

So now the question is why such cross-section is that big?
(s)Top resonances?

In that case, the cross-section IS physical but not what you call a 2j QCD radiation from your born.
You typically have to forbid the resonances (with the $ syntax in general)

Cheers,

Olivier

> On 20 Mar 2019, at 10:22, Alexis Kalogeropoulos <email address hidden> wrote:
>
> Question #679358 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/679358
>
> Status: Answered => Open
>
> Alexis Kalogeropoulos is still having a problem:
> Hi Olivier
>
> I do see indeed, that the 2j processes
>
> g g > t1 x1- g b~
>
> g g > t1~ x1+ g b
>
> give the largest contribution to the xsec. So, what would you suggest
> to do ? I mean, the xsecs I get from these 2j process do not seem
> physical.
>
> Thanks
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Alexis Kalogeropoulos (alkaloge) said :
#4

Ah, ok, I see what you mean. One last point to clarify.

If using the LHE as they are (ie NOT forbidding the resonances), do you think that would compromise that physics result I have (like skewed kinematics etc)? In principle, I am asking, would it be still alright you think using the existing LHE, or I really need to do everything from scratch (which would be a hustle at this point...) - how severe is not excluding the resonances ?

Thanks

Revision history for this message
Best Olivier Mattelaer (olivier-mattelaer) said :
#5

Well,

I would expect that ~ 96% of your events come from such resonances (40-1.4)/40
So if this is not what you are interested in ...

Cheers,

Olivier

> On 20 Mar 2019, at 10:43, Alexis Kalogeropoulos <email address hidden> wrote:
>
> Question #679358 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/679358
>
> Status: Answered => Open
>
> Alexis Kalogeropoulos is still having a problem:
> Ah, ok, I see what you mean. One last point to clarify.
>
> If using the LHE as they are (ie NOT forbidding the resonances), do you
> think that would compromise that physics result I have (like skewed
> kinematics etc)? In principle, I am asking, would it be still alright
> you think using the existing LHE, or I really need to do everything from
> scratch (which would be a hustle at this point...) - how severe is not
> excluding the resonances ?
>
> Thanks
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Alexis Kalogeropoulos (alkaloge) said :
#6

Thanks Olivier Mattelaer, that solved my question.