[Rivet] Support for explicit ghosts?

[Andy Buckley]

Hi James,

Sounds like FastJets should use the is_pure_ghost() check to veto such 
particles when building lists of jet constituents. Could you take a look 
in FastJets.cc and provide a patch / suggest which functions need that 
vetoing?

If this can be done quickly it would be nice to get this fix into 
version 2.4.0, i.e. in the next ~week. (Reminder: analysis integration 
volunteers still needed... step up, people.)

My one concern is that it could screw up the ghosts that we currently 
use for flavour tagging, if not done carefully. Do you know what 
criterion FastJet uses to decide if a PseudoJet is a "pure ghost"? A 
momentum threshold check maybe, or an explicit bool flag?

Thanks,
Andy


On 19/09/15 21:13, James Monk wrote:
> Hi everyone,
>
> I am trying to run trimmed jets with active area subtraction in Rivet.  Apparently explicit ghosts are needed by fastjet to do this because it needs the ghosts to persist long enough to calculate the areas of the sub jets.  On the other hand, this appears to trip up Rivet’s FastJets projection when it attempts to match the jet’s constituents with the list of particles that were input to the projection, specifically on line 163 of FastJets.cc.  This happens because the jet’s constituents includes all the ghosts, which the projection knows nothing about, so it can’t find them in the _particles map.
>
> Is this known, or am I trying to do things wrong?  I am having to use an uncached version of the FastJet projection because I am feeding it a modified list of input particles after applying some pile-up mitigation, but I don’t think that affects things.
>
> PseudoJet has a method called is_pure_ghost, so I think the solution is for Rivet to use that to ignore any constituent that is a pure ghost - this method returns true if all the constituents of a jet are ghosts, but I presume it works on the ghosts themselves.
>
> cheers,
>
> James
>
> _______________________________________________
> Rivet mailing list
> Rivet at projects.hepforge.org
> https://www.hepforge.org/lists/listinfo/rivet
>


-- 
Dr Andy Buckley, Lecturer / Royal Society University Research Fellow
Particle Physics Expt Group, University of Glasgow