|
[Rivet] Rivet with asymmetric beamsEike von Seggern jan.eike.von.seggern at physik.hu-berlin.deWed Dec 9 22:26:46 GMT 2009
On Wed, Dec 09, 2009 at 14:52 +0000, Andy Buckley wrote: > Eike von Seggern wrote: > > On Fri, Dec 04, 2009 at 18:21 +0000, Andy Buckley wrote: > >> Eike von Seggern wrote: > >>> Hi Andy et al., > >>> > >>> following up on a thread a few days ago. I set up a directory on > >>> hepforge with an example data file. The problem is that with the > >>> PDG_HADRON_MULTIPLICITIES analysis 23 events of 10000 are processed. The > >>> problem is (afaik) that the _io->fill_next_event(_evt.get()) call in > >>> Run::readEvent() fails. So I'm not sure if it's a HepMC problem or a > >>> rivet bug. > >> Hi Eike, > >> > >> You need to let me know *where* on HepForge you've put your example file > >> ;) This is certainly one to get sorted before 1.2.0 (i.e. SOON!) > > > > Hi Andy et al., > > > > I think, the problem is in HepMC. My suspect is now that the > > > > *m_istr >> *evt; > > > > call on line 132 in HepMC-2.05.00/src/IO_GenEvent.cc fails. And > > _io->error_message() gives me: > > > > "input stream encountered invalid data" > > > > Has any of you encountered this earlier? > > No... that seems very odd. Can you mock up a very small test example > that reproduces the problem (check the HepMC ASCII produced by Sherpa > for any bugs, and see if a mocked-up asymmetric event shows the same > problem), and submit it as a HepMC bug on the LCG Savannah tracker, > linked from the HepMC web page. Something problematic like this should > be back-ported to production versions of HepMC. > > Andy > > -- > Dr Andy Buckley > SUPA Advanced Research Fellow > Particle Physics Experiment Group, University of Edinburgh > _______________________________________________ > Rivet mailing list > Rivet at projects.hepforge.org > http://www.hepforge.org/lists/listinfo/rivet > So the bug was actually that the cross-section error wasn't correctly passed to HepMC in Sherpa. That's fixed now. I couldn't find a ticket. If there was one for this issue, it can be closed. Cheers, eike
More information about the Rivet mailing list |