search

qliphy / MGISR

BSD 3-Clause "New" or "Revised" License
3 stars 1 forks source link

put modification in dummy_fct.f to avoid to have to fully re-write file #1

Closed oliviermattelaer closed 6 years ago

oliviermattelaer commented 6 years ago

Hi,

I propose here a new version of MG5aMC, which should be more stable against modifications of MG5aMC in the future.

Here are some details of the proposed modification:

1) lpp is set by default to "9" (mode dedicated for plugin computation)

2) thanks to that NO modification are needed in genps.f, the definition of shat/x1/x2 is done via a function which is by default dummy (i.e. empty) that are defined in dummy_fct.f. The plugin remove that function and replace it by the required function. All function from "additional_fct.f" are also added into that file.

3) The PDF is set automatically to 1 in presence of lpp=9, so we do not need to edit pdg2pdf.f anymore.

4) I'm not sure why you remove the "reweight.f" file (I see that you change a threshold here). I simply change the default dynamical scale choice to "4" (sqrts) which should be quite correct for such type of computation (and not very important here anyway)

qliphy commented 6 years ago

@oliviermattelaer Thanks a lot! I will have a check asap.

"I'm not sure why you remove the "reweight.f" file (I see that you change a threshold here). " I was checking ee>mumu at Belle2 and validate MG against BabayagaNLO and I found mumu invariant mass cut at 2GeV, and then I traced it to reweight.f. But now if we set dynamical scale choice to "4", it should be passed and thus be fine.

qliphy commented 6 years ago

Thanks @oliviermattelaer I have further commented this line: #cp(pjoin(rootdir, 'reweight.f'), pjoin(self.dir_path, 'SubProcesses'))

Tests on ee>zh works well.