pythia8 produces corrupt hepmc file with MG2.7.2

[Michaela Queitsch-Maitland]

Hello,

I'm not sure if this is a MG or other (Pythia8/HepMC) bug, but with MG v2.7.2 I observe that Pythia8 seems to produce a 'corrupt' HepMC file on lxplus that cannot be read by other HepMC interpreters (e.g. Rivet). The problem seems to be the duplication of the header START_EVENT_LISTING, e.g.:

HepMC::Version 2.06.09
HepMC::IO_GenEvent-START_EVENT_LISTING
HepMC::IO_GenEvent-START_EVENT_LISTING
E 0 -1 -1.0000000000000000e+00 -1.0000000000000000e+00 -1.0000000000000000e+00 9999 0 230 1 2 0 1 8.4300000000000000e-02

This seems to be environment/compiler dependent, as I see this problem on lxplus and the NAF (also CentOS 7). I use the "vanilla" lxplus environment (gfortran 4.8.5, python 2.7.5). On my laptop (macOS, gfortran 9.3.0) I don't see this.

I can't reproduce this exactly right now because of the pythia8 installation error reported in https://bugs.launchpad.net/mg5amcnlo/+bug/1875436, but essentially all I did was: 1) fresh install MG 2.7.2, 2) install pythia8 using the build-in installer, 3) generate p p < mu+ mu- with showering=pythia8.

Using the same setup/installation and commands with MG v2.6.7 I do not see the duplicated line in the HepMC file. Let me know in case any further information is helpful.

Thanks,
Michaela

[Olivier Mattelaer (olivier-mattelaer)]

Hi,

My guess is that it should be related to this part of the code:
https://bazaar.launchpad.net/~mg5core1/mg5amcnlo/2.7.3/revision/285.1.2

so it will be interesting to know if you support negative number for the "-n" option of the head command.

Cheers,

Olivier