NLO Pythia8 shower module drops the last event

[Sung Hak Lim]

Dear MG5 team,

I found that Pythia8 generated HepMC file always miss the last event in the LHE input.

I also compared momentum of parton information in LHE and HepMC, and missing event is the last one.

MG5 version: 2.5.4
Pythia version: 8.226

I think this is a bug due to following code:

[MG5]/template/NLO/MCatNLO/srcPythia8/Pythia82_hep.cc

around line 72

    if (nSelected >= iEventshower) break;

This breaks event saving loop when nSelected = iEventshower. However, nSelected starting from 1 so this will break loop at the last event without saving it.

If I modify it into

    if (nSelected > iEventshower) break;

Then I got following additional Pythia abort message in mcatnlo_run.log file.

 Pythia::next(): 10000 events have been generated
 PYTHIA Abort from Pythia::next: reached end of Les Houches Events File

Therefore, LHE events are fully parsed now.

If we grep "E " for both HepMC output, I got 9999 for current published code and 10000 for my new code.

Could you confirm this?

Best wishes,
Sung Hak Lim

[Paolo Torrielli (paolo-torrielli)]

Dear Sung,
thanks very much for your feedback.
I conform this behaviour and the solution you propose.
I’ll include this in the next 2.5.6 release.
Thanks again.
Cheers.
Paolo

> On 26 Jul 2017, at 08:24, Olivier Mattelaer <email address hidden> wrote:
>
> ** Changed in: mg5amcnlo
> Assignee: (unassigned) => Paolo Torrielli (paolotorriell)
>
> --
> You received this bug notification because you are a member of aMC@NLO,
> which is subscribed to MadGraph5_aMC@NLO.
> Matching subscriptions: aMCatNLO_bug_reports
> https://bugs.launchpad.net/bugs/1706507
>
> Title:
> NLO Pythia8 shower module drops the last event
>
> Status in MadGraph5_aMC@NLO:
> New
>
> Bug description:
> Dear MG5 team,
>
> I found that Pythia8 generated HepMC file always miss the last event
> in the LHE input.
>
> I also compared momentum of parton information in LHE and HepMC, and
> missing event is the last one.
>
> MG5 version: 2.5.4
> Pythia version: 8.226
>
> I think this is a bug due to following code:
>
> [MG5]/template/NLO/MCatNLO/srcPythia8/Pythia82_hep.cc
>
> around line 72
>
> if (nSelected >= iEventshower) break;
>
> This breaks event saving loop when nSelected = iEventshower. However,
> nSelected starting from 1 so this will break loop at the last event
> without saving it.
>
> If I modify it into
>
> if (nSelected > iEventshower) break;
>
> Then I got following additional Pythia abort message in
> mcatnlo_run.log file.
>
> Pythia::next(): 10000 events have been generated
> PYTHIA Abort from Pythia::next: reached end of Les Houches Events File
>
> Therefore, LHE events are fully parsed now.
>
> If we grep "E " for both HepMC output, I got 9999 for current
> published code and 10000 for my new code.
>
> Could you confirm this?
>
> Best wishes,
> Sung Hak Lim
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/mg5amcnlo/+bug/1706507/+subscriptions