[Rivet] b-tagging with q-quarks in Rivet

[Andy Buckley]

On 20/11/14 21:21, Tancredi Carli wrote:
> 
>  Dear Andy,

Hi Tancredi! Rivet list now in CC.

>   I am working on fixed order NLO calculations (on the ttbar+njets
> measurement)
>  using Sherpa. We would like to modify the Rivet routine written by A.
> Grohsjean
>  to run on fixed order calculations (with a decayed top). It seems to work
>  for the jets (by default the final state is done for all particle that
> are not leptons ??),

By default the final state includes all stable particles -- you need to
use vetoed final state to remove specific leptons and/or a
VisibleFinalState to exclude neutrinos etc.

I'm not sure what status codes are being translated from the fixed-order
events, but Rivet will assume the HepMC standard.

>  but we have a problem with the b-tagging, since obviously there are no
>  B-hadrons in the event record.
> 
>  Is there and easy way to tell Rivet that it should use b-quarks as b-tag ?
>  Of course, we could match the b-quarks by ourselves...

When we moved to doing ghost-association tagging we lost the previous
fallback to quarks that was permitted by the "parentage tagging"
approach. Since MPI and hadronization don't have standard event record
representations, tagging by parentage in the pre-hadronization event
record was too inclusive and sometimes just wrong.

We would like to generalise the ghost association tagging to allow any
four-momentum-like object to be ghosted into the clustering, but it's
not there yet. So I think for now if you need quarks then it'll best to
find them explicitly and probably dR match to jets -- except in very
close, unisolated jets, it should give the same answer as
ghost-association in anti-kT jet building.

Cheers,
Andy

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