Issue while using mmll cut in Madgraph

Asked by Mohit sharma on 2021-03-17

Hi Madgraph experts,
I am generating the following process where final decay particles contains only one lepton (excluding the neutrions):

        generate p p > wx, wx > lx vx @1
 add process p p > wx j, wx > lx vx @2
 add process p p > wx j j, wx > lx vx @3

What's the result of using a mmll (say 80 GeV) cut for the above process. I expected the cross section to be reduced to zero, but that's not happening here. Also what does imposing a cut on leptons (e/mu) invariant mass signifies here?
Thanks,
Mohit

Question information

Language:
English Edit question
Status:
Answered
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Last query:
2021-03-17
Last reply:
2021-03-17

Hi,

1) the comment for that cut is:
" min invariant mass of l+l- (same flavour) lepton pair"

So i do not see why it should apply for your process.
To be sure, you can check the run_card, if the mmll cut does not appear in the default card, this means it has no potential impact for your process (need a recent version of the code for that0

2) since you use decay_chain, you need to check the value of the cut_decays, if set on False, a lot of cut will not be applied on the particles comming from the decays.

Cheers,

Olivier

Mohit sharma (mohitsharma98) said : #2

Thanks for the prompt response Olivier. As you correctly said, the mmll cut does not appear as a default cut in the run card for newer versions of Madgraph. But out of curiosity, I added the mmll cut as a hidden parameter to see how it's affecting the cross-section. To my surprise, the cross-section decreased significantly but did not reduce to zero. Can you tell what's going on internally and why the final cross-section is greater than zero.

Hi,

> Can you tell what's going on internally and why the final cross-section
> is greater than zero.

The real question is why the two cross-sections are not the same, the value of that cut is irrelevant for your computation.

I have therefore run the following script:

define wx = w+ w-
define lx = e+ e- mu+ mu-
define vx = ve ve~ vm vm~
generate p p > wx, wx > lx vx @1
add process p p > wx j, wx > lx vx @2
#add process p p > wx j j, wx > lx vx @3
output
launch
launch
set mmll 80

Which keep parameter for a MLM run,
As you can see
1) I bypass the two jet multiplicity to have the code running super fast)
2) I did not run the parton-shower here (so reported cross-section are unphysical for this MLM setup)
3) The run are 100% identical up to the inclusion of the mmll parameter (to 80)

The resulting cross-section are
run 1 (mmll set as default value): 3.686e+04 +- 98
run 2 (mmll set to 80): 3.686e+04 +- 98

So I do not observe any impact of the cut

actually including the 2j sample was fast enought that it finished before me finishing this post.
here is the associated (still unphysical) cross-section

run 1 (mmll set as default value): 3.854e+04 +- 100
run 2 (mmll set to 80): 3.853e+04 +- 84
(again no impact of the cut)

Which is very close to the previous computation (first time I run this one in years, I was expecting bigger difference) looks like the 0j computation is actually very huge here (but likely highly suppresed when the parton-shower will multiply those by the proper sudakov form-factor to have the physical cross-section. Could you confirm me if this is the case or not when you will have the validation plot for your MLM generation?

Cheers,

Olivier

PS: I guess that you change other parameter in the run_card that can explain why your cross-section is not the same.

> On 17 Mar 2021, at 12:31, Mohit sharma <email address hidden> wrote:
>
> Question #696106 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/696106
>
> Mohit sharma posted a new comment:
> Thanks for the prompt response Olivier. As you correctly said, the mmll
> cut does not appear as a default cut in the run card for newer versions
> of Madgraph. But out of curiosity, I added the mmll cut as a hidden
> parameter to see how it's affecting the cross-section. To my surprise,
> the cross-section decreased significantly but did not reduce to zero.
> Can you tell what's going on internally and why the final cross-section
> is greater than zero.
>
> --
> 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 Mohit sharma for more information if necessary.

To post a message you must log in.