|
|
[Rivet] b-tagging in Rivet 2.x.x
Andy Buckley
andy.buckley at cern.ch
Wed Mar 4 15:46:39 GMT 2015
Hi Christian,
That's very reasonable. I'll first look into what logic is for the cuts
not operating in the intuitive way.
By the way, note that we're now using the "Cuts" system by preference to
(double, double, double, ...) type argument lists for kinematic cuts.
This is already available in 2.2.0, and Cut combining will be a bit
nicer to use in the next release. If we add optional args, it'll be a
Cut rather than doubles.
Andy
On 04/03/15 10:37, Christian Gütschow wrote:
> Hi Andy,
>
> just coming back to the HeavyHadrons projection. I'm currently upgrading
> the Zb(b) routine to Rivet 2.2.0 and wanted to use this projection
> instead of looping over all GenParticles and checking their PDG IDs with
> a custom-made list of B-hadron IDs.
>
> I didn't get quite the same answer though when I started out with
>
> addProjection(HeavyHadrons(-2.5, 2.5, 5.0*GeV), "BHadrons");
> ...
> const Particles& stableBs = applyProjection<HeavyHadrons>(e,
> "BHadrons").bHadrons();
>
>
> because this includes Bs from all over the place and not just within the
> fiducial regions defined in the constructor. I haven't fully traced this
> back in the source code, but my understanding is that the projection
> also checks for children and what not, such that the resulting container
> of Bs also contains particles with pT < 5GeV and |eta| > 2.5.
>
> That's a bit counterintuitive, but fair enough, I can still get the Bs
> within my fiducial region using something like:
>
> addProjection(HeavyHadrons(), "BHadrons");
> ...
> const Particles& allBs = applyProjection<HeavyHadrons>(e,
> "BHadrons").bHadrons(5.0*GeV);
> Particles stableBs;
> foreach(Particle p, allBs) {
> if(p.abseta() < 2.5) stableBs += p;
> }
>
>
> I was wondering whether we can modify the bHadrons method, such that the
> min and max eta can be passed along with the min pT in order to avoid
> the extra loop? That'd be much cleaner I think...
>
> Cheers,
> Chris
>
>
>
> On 16 February 2015 at 10:52, Andy Buckley <andy.buckley at cern.ch
> <mailto:andy.buckley at cern.ch>> wrote:
>
> Ah, I think HERWIG does not use the standard status=2 code for unstable
> hadrons: it's one of the reasons to move away from using it. We had
> implemented a status-rewriting feature in the ATLAS Herwig_i interface
> to correct this, but I have a feeling that people were scared of change
> and it never got switched on. I think it'll probably work for ~any other
> generator, since HERWIG is the only one I know of that doesn't follow
> the HepMC status scheme.
>
> Andy
>
>
> On 16/02/15 09:49, Christian Gütschow wrote:
> > Hi Andy,
> >
> > ooh, fancy that! I am liking this option best. :-)
> >
> > Unfortunately, it doesn't seem to work for me either, i.e. both the
> > bHadrons vector is empty and jet.bTagged() also returns false for all my
> > selected jets (at least that's consistent), while the approach via the
> > GenParticle container does find some b-jets.
> >
> > I was using an Alpgen+Herwig sample for this test, namely a file from
> > the mc11_7TeV.126733.AlpgenHerwigEvtGenWbb2p.evgen.EVNT.e1155 container.
> >
> > Cheers,
> > Chris
> >
> >
> > On 15 February 2015 at 23:10, Andy Buckley <andy.buckley at cern.ch <mailto:andy.buckley at cern.ch>
> > <mailto:andy.buckley at cern.ch <mailto:andy.buckley at cern.ch>>> wrote:
> >
> > Hi Christian,
> >
> > The best way to do it now is if (jet.bTagged()) ;-) This uses
> ghost
> > association to hadrons.
> >
> > The HeavyHadrons way should work, though, so I'm very puzzled
> that it
> > doesn't. What generator are you using? Is the bHadrons vector
> non-empty?
> > Could you try printing the dRs in that comparison? (By the
> way, you
> > shouldn't need to re-apply the > 5 GeV cut in the loop, since
> that's one
> > of your b-hadron selection criteria!)
> >
> > Andy
> >
> > PS. I will get back to you about the merging issue. I didn't
> get any
> > chance in the last 2 weeks, but have a "free" week coming up
> so will do
> > my best. Prod me if I don't!
> >
> >
> > On 15/02/15 14:05, Christian Gütschow wrote:
> > > Dear Rivet authors,
> > >
> > > I'm just trying to implement a Rivet routine for a W+b ATLAS
> > measurement
> > > and was wondering what the preferred way of doing particle-level
> > > b-tagging in Rivet is?
> > >
> > > Naively, I would think that the easiest way to decide whether a
> > jet is a
> > > b-jet is to use the HeavyHadrons projection, e.g. like so
> > >
> > > addProjection(HeavyHadrons(-2.5, 2.5, 5.0*GeV), "BHadrons");
> > > ...
> > > const Particles& bHadrons =
> applyProjection<HeavyHadrons>(event,
> > > "BHadrons").bHadrons();
> > > ...
> > > // for each jet "j":
> > > foreach(const Particle& b, bHadrons) {
> > > if( (b.pT() > 5.0*GeV) && (deltaR(j, b) < 0.3) ) {
> > > // jet matched to B hadron!
> > > ...
> > > break;
> > > }
> > > }
> > >
> > >
> > > Turns out, this doesn't do anything at all because bHadrons
> has always
> > > size 0 for some reason. Is this to be expected or likely to be
> > some sort
> > > of bug (on my side)?
> > >
> > > On the other hand, if I use the approach that was used in
> the Z+b(b)
> > > analysis (ATLAS_2014_I1306294), i.e. something like
> > >
> > > bool hasB = false;
> > >
> > > foreach(const GenParticle* b, particles(event.genEvent())) {
> > > if(b->momentum().perp() < 5.0*GeV) continue;
> > > for(uint k=0; k < stableB.size(); ++k) {
> > > if( fabs(b->pdg_id()) == stableB[k] ) {
> > > if( deltaR(j.momentum().eta(), j.momentum().phi(),
> > > b->momentum().eta(), b->momentum().phi()) < 0.3 ) {
> > > // jet matched to B hadron!
> > > hasB = true;
> > > break;
> > > }
> > > }
> > > }
> > > if(hasB) break;
> > > }
> > >
> > >
> > > where stableB is a custom-made vector of the B-hadron PDG IDs,
> > then I do
> > > find the B hadrons in the event and can successfully match
> my jets
> > to them.
> > >
> > > I think it's safe to say that while the latter option does the
> > trick, it
> > > is a bit of an awkward way to code this up. (Presumably
> that's the
> > > reason why the Z+b(b) analysis is still dwelling in the
> contributions
> > > area...?) Is there a better way of doing this?
> > >
> > > Many thanks for your advice in advance!
> > >
> > > Cheers,
> > > Chris
> > >
> > > --
> > >
> > > Dr. Christian Gütschow
> > >
> > > TU Dresden
> > > Institut für Kern- und Teilchenphysik
> > > Zellescher Weg 19
> > > 01069 Dresden
> > >
> > > > E17, Andreas-Schubert-Bau
> > > > chris.g at cern.ch <mailto:chris.g at cern.ch>
> <mailto:chris.g at cern.ch <mailto:chris.g at cern.ch>>
> > <mailto:chris.g at cern.ch <mailto:chris.g at cern.ch>
> <mailto:chris.g at cern.ch <mailto:chris.g at cern.ch>>>
> > >
> > >
> > >
> > > _______________________________________________
> > > Rivet mailing list
> > > Rivet at projects.hepforge.org
> <mailto:Rivet at projects.hepforge.org>
> <mailto:Rivet at projects.hepforge.org
> <mailto:Rivet at projects.hepforge.org>>
> > > https://www.hepforge.org/lists/listinfo/rivet
> > >
> >
> >
> > --
> > Dr Andy Buckley, Royal Society University Research Fellow
> > Particle Physics Expt Group, University of Glasgow / PH Dept, CERN
> >
> >
> >
> >
> > --
> >
> > Dr. Christian Gütschow
> >
> > TU Dresden
> > Institut für Kern- und Teilchenphysik
> > Zellescher Weg 19
> > 01069 Dresden
> >
> > > E17, Andreas-Schubert-Bau
> > > chris.g at cern.ch <mailto:chris.g at cern.ch>
> <mailto:chris.g at cern.ch <mailto:chris.g at cern.ch>>
> >
>
>
> --
> Dr Andy Buckley, Royal Society University Research Fellow
> Particle Physics Expt Group, University of Glasgow / PH Dept, CERN
>
>
>
>
> --
>
> Dr. Christian Gütschow
>
> TU Dresden
> Institut für Kern- und Teilchenphysik
> Zellescher Weg 19
> 01069 Dresden
>
> > E17, Andreas-Schubert-Bau
> > chris.g at cern.ch <mailto:chris.g at cern.ch>
>
--
Dr Andy Buckley, Lecturer / Royal Society University Research Fellow
Particle Physics Expt Group, University of Glasgow
More information about the Rivet
mailing list
|