MCmasses for leading order generation?

At LO, there is the option 'gridpack' in the run_card.dat. Setting it to '.true.' for the first run sets up a gridpack tarball that can be moved around and run independently on any system with the same architecture to generate events.

At NLO, such an option is not provided. However, before generating a process and outputting it, one can set the MG5_aMC option 'output_dependencies' to 'internal' by typically typing the command

MG5_aMC > set output_dependencies internal; save options;

This will force MG5_aMC to output all the sources of the extra dependencies of the NLO run (such as CutTools and StdHep) in the Source directory.
At this stage, the process directory output can be gzipped and moved around anywhere (assuming of course no absolute path was specified for lhapdf or fastjet). One should steer the execution by starting the MG5_aMC interface with the 'aMCatNLO' script *placed in the bin directory of the process output*. Notice that with this setup, the process output can in principle be compiled on a different machine than the one that generated it.

It is also possible to create the equivalent of an NLO gridpack by running the code a first time while specifying in 'run_card.dat' that 0 events must be generated and a required accuracy 'req_acc' equal to something like one over the squared root of the total number of events intended to be generated. Notice that when generating a very large number of events (i.e. > 100M) then one can afford to have the grids setup with less precision than suggested by the rule of thumbs I just gave.
Once the code has been compiled and the grid setup with the run above, one can perform the event generation only by setting the target number of events to be generated in the run_card and running

MG5_aMC > generate_events -x -o

from the MG5_aMC interface. The '-x' options skips the compilation and '-o' options reuses the grid and jumps directly to event generation. Notice one can use the option '-f' on top to skip the launch questions and reuse the current values in the cards.

Please disregard the above as it should have been posted as an answer to Question #250606.

Hi Josh,

I have take a look and this will need a deep re-factorization of the code. So this is not something that can be done in a hour or less. In top of that this is linked to part of the code where only Tim and Fabio knows all the details about.

I will then assign this question to Fabio to see if it considers important to investigate.

Cheers,

Olivier

Hi Olivier, Fabio,
In the meantime can someone comment on the present understanding and best practices for running in LO mode with pythia 8? In particular is it still considered problematic to simply pass massless leptons and/or b-quarks to pythia?

For leptons we could for LO generation use strictly models with massive leptons, but for b-quarks we still need to use models with massless b's for five-flavour generation of course.

Thanks,
Josh

Ciao Josh,

I have asked Stefan an opinion, which you find below. For me the most reasonable option is to keep the elctron and muon mass to zero in the mem generation and to reshuffle the momenta right before giving them to pythia. Actually, ideally and if possible it should be pythia taking care of this...

Fabio
Inviato da iPad

> Il giorno 31/lug/2014, alle ore 14:35, "Stefan Prestel" <email address hidden> ha scritto:
>
> Hi Fabio,
>
> The main issue in Pythia8 is with massless leptons: The QED radiation
> cut-off is very small in Pythia8, since the soft divergence of photon
> emissions would normally be regulated by the lepton masses. That's the
> reason why it is good to have massive leptons.
>
> The QED cut is however only marginally tuned (if at all, I'm honestly not
> sure which data could help to pin it down). Also, the problem is not as
> bad as it sounds: Very rarely, the QED shower will run into numerical
> problems, which then lead to a re-trial of the event. The code will thus
> split out warnings, but still perform okay. Of course, it's nice to avoid
> this issue.
>
> As for the bottom, the same basically applies. However, the QCD radiation
> off the bottom is so much more dominant than QED that I've never seen a
> massless bottom quark act up. (Initial bottoms will anyway be converted,
> so the soft limit is never realised, and final state bottoms radiate
> enough so that QED is very, very rare.).
>
> When running Pythia8 with massless b's, it's of course always helpful to
> tell it (5:m0 = 0.0 in the input card). Otherwise, it will print warnings
> :).
>
> So my suggestion would be massive leptons, massless bottoms - although
> that does sound a bit funny :).
>
> What I actually do is a bit more evil. I simply run the aMC@LNLO momentum
> reshuffling also on tree-level samples. Then, I don't need to worry about
> mass instabilities. However, the PS input kinematics do not match the ME
> evaluation kinematics exactly... so you see where I admit to being a
> plumber :D.
>
> Hope I could help.
>
> Cheers,
> Stefan
>
> PS: It could also be interesting to check if we actually need such a small
> QED cut-off - but I simply don't know where to look. So I'm happy if you
> have suggestions.
>
>
>> On Wed, July 30, 2014 9:55 am, Fabio Maltoni wrote:
>> Ciao Stefan, Simon,
>>
>> I have got this message from Josh. Can you help?
>>
>> Fabio
>>
>>
>> Begin forwarded message:
>>
>> From: Josh Bendavid
>> <<email address hidden><mailto:<email address hidden>>>
>> Subject: Re: [Question #250699]: MCmasses for leading order generation?
>> Date: 29 Jul 2014 17:41:41 GMT+2
>> To: <<email address hidden><mailto:<email address hidden>>>
>> Reply-To:
>> <<email address hidden><mailto:<email address hidden>>>
>>
>> Question #250699 on MadGraph5_aMC@NLO changed:
>> https://answers.launchpad.net/mg5amcnlo/+question/250699
>>
>> Status: Answered => Open
>>
>> Josh Bendavid is still having a problem:
>> Hi Olivier, Fabio,
>> In the meantime can someone comment on the present understanding and best
>> practices for running in LO mode with pythia 8? In particular is it still
>> considered problematic to simply pass massless leptons and/or b-quarks to
>> pythia?
>>
>> For leptons we could for LO generation use strictly models with massive
>> leptons, but for b-quarks we still need to use models with massless b's
>> for five-flavour generation of course.
>>
>> Thanks,
>> Josh
>>
>> --
>> You received this question notification because you are the assignee for
>> this question.
>
>

Thanks Fabio,
Indeed my original question was whether the momentum shuffling already implemented for the NLO case would make sense also for LO.

Given that such reshuffling is not present in pythia at the moment, we are trying to avoid as much as possible messing with the kinematics in "extra" scripts sitting in between.

We're running some tests as well, but if the conclusion is indeed that the massless b's are ok, but massive leptons are preferred for input to pythia, probably the direction we'll go is to use massive leptons in the ME for LO (where there shouldn't be numerical stability issues and we can afford the loss in speed) and simply leave the b's as they are massive/massless for 4F/5F scheme.

Josh