Crash with Phase space cuts inside MadSpin card (MG5_aMC_v2.6.5)

Hi,

I fully forget about such options that I never include in the documentation.
The line to handle that option are still within our LTS version but maybe this is a parsing issue...
I will do some test.

Since neither 2.4.2 or 2.6.5 are still under support, the only support that I can provide is some patch for our LTS version (2.9.x) which will be include in our bug-fixing only update (likely 2.9.7) and inside our next short term support release (likely 3.3.1)

So far i have not face any issue with spinmode=none and setting a full run_card.
Do you have more details on that issue? (like the debug file?)

Cheers,

Olivier

> On 24 Nov 2021, at 11:50, Rocco Ardino <email address hidden> wrote:
>
> New question #699615 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/699615
>
> Dear Experts,
>
> I am trying to generate the gridpacks needed to produce NLO MC samples of Higgs dalitz decay to a pair of muons and a photon. The cards used for the purpose can be found here (in particular, the folders *HToMuMuGamma):
> https://github.com/cms-sw/genproductions/tree/master/bin/MadGraph5_aMCatNLO/cards/production/2017/13TeV/Higgs/dalitz
>
> Unfortunately, I am not the original author of the cards and they were created about 3-4 years ago. These cards were written for MG5_aMC_v2.4.2, but now I have to regenerate the gridpacks using necessarily the MG5_aMC_v2.6.5 version.
> However, when running the gridpack generation, the pilotrun fails since it is not possible to apply a 0<m_ll<60 GeV phase space cut inside the madspin card. For instance, the problematic lines of the cards are the following:
> https://github.com/cms-sw/genproductions/blob/master/bin/MadGraph5_aMCatNLO/cards/production/2017/13TeV/Higgs/dalitz/ggH_012j_NLO_FXFX_HtoMuMuGamma/ggH_012j_NLO_FXFX_HtoMuMuGamma_madspin_card.dat#L11-L12
>
> set run_card mmll=0
> set run_card mmllmax=60
>
>
> A long discussion has been done with the gen-contacts of the physics group and it can be found here:
> https://github.com/SMP-GEN-Coordination/mc-requests/issues/12
> where, the problem is fully described.
>
> Apparently, from the version 2.4.2 to 2.6.5, a lot of changes have been done and it seems that this feature (cuts in the madspin card) are no longer supported, from what I understood. A clean fix could not be found for this problem for the version 2.6.5 and the only way to perform a correct and bug-free gridpack generation was to remove those lines. By this way, a higher number of events would be required in order to take into account the loss in efficiency.
> Apparently, a possible way to solve the problem would be to create a second run_card for madspin, but it seems that this solution may lead to bugs that would cause other problems to the gridpack generation.
>
> Thus, I am writing here to ask if there is a clean and simple way to apply this m_ll cut inside the madspin card or somewhere else (like in the run card), avoiding its implementation in the fortran file 'cuts.f'. For a LO generation, this would be simple to implement inside the run_card, since mmll and mmllmax entries are available, but in the NLO card template I could not find a way to set an upper cut for m_ll.
>
> I apologize for the long question and I hope that the description is clear.
> Thanks in advance!
> Best,
> Rocco
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Thanks Olivier Mattelaer, that solved my question.

Looks like that patch is making some gridpack to crash. so will not be include in our LTS (but maybe a revised version will do).