[Rivet] FastJets problems

[Andy Buckley]

Hendrik Hoeth wrote:
> Hi,
> 
> I'm trying to get the STAR_2009 UE analysis to do what STAR does. This
> includes running SIScone with a split-merge parameter of 0.75. In our
> FastJets projection we hard-code it to be 0.5. That's not the
> recommended value, but it has been used in many analyses and we don't
> bother passing it as parameter.
> 
> So instead of simply calling
>   addProjection(FastJets(jfs, FastJets::SISCONE, 0.7), "AllJets");
> I tried
>   addProjection(FastJets(jfs, fastjet::SISConePlugin(0.7, 0.75)), "AllJets");
> This yields a segfault during the analysis initialisation.
> 
> What's the recommended way of accessing Fastjet plugins directly? Or
> shall I introduce the split-merge parameter as an argument to FastJets(...)?

The latter, please!

Not sure about the details, but I suspect that the analysis 
initialisation is trying to clone and delete your locally-assigned 
fastjets plugin: we need a smart pointer in FastJets to handle this. 
Don't ask ;)

Andy

-- 
Dr Andy Buckley
Particle Physics Experiment Group, University of Edinburgh