[Rivet] Is it possible? Rivet and truth look different

[Liron Barak]

Hi Andy,

On Fri, Nov 8, 2013 at 2:36 PM, Andy Buckley <andy.buckley at cern.ch> wrote:

> Ok, well the two sets of curves do look a bit different. Maybe this
> could come from different eta acceptance and changes in Sherpa's
> handling of beam remnant dipoles, etc.

But then, wouldn't it also be seen in the first plot?

> In the Rivet plot it looks to me
> like Sherpas 1 and 2 could be ~consistent in the high-stats part, and
> maybe the deviations elsewhere are just statistical jitter: can you make
> the plot again with the --mc-errs option flag, to show the MC error bars?
>
It is attached... but surprised me is not that the 2 sherpa options are
different, that is kind of expected, no? What worries me is why the Rivet
gave different (actually opposite) answers than simple draw of the truth
jets? If it is because the eta/pt are different, what should I "believe" to?

Thanks
Liron

>
> Thanks,
> Andy
>
> PS. Please reply to the list, not just me! "Reply all" is usually a good
> default model for work email... no-one wants to suddenly have all the
> responsibility just because they were the first to reply ;-)
>
Sure, sorry, in the first email (when you added the list) there was no
option of reply-all, just replay, I assumed it will be sent to an email
which is your name but actually the list :) the second one has reply -all
but I didn't bother to check that time... my mistake, sorry!

>
>
> On 07/11/13 20:17, Liron Barak wrote:
> > Hi Andy,
> >
> > Sorry... I sent the file from the draw command and I thought you can see
> > the html file of the rivet.....
> > Now it is attached...
> > I compared in each (the rivet and the draw command) the 2 sherpa
> > versions (1.4 and 2)...
> >
> > Best
> > Liron
> >
> > On Thu, Nov 7, 2013 at 7:16 PM, Andy Buckley <andy.buckley at cern.ch
> > <mailto:andy.buckley at cern.ch>> wrote:
> >
> >     I guess so, but am not sure. And the default ATLAS truth jets
> probably
> >     apply some restrictions on the clustering inputs such as excluding
> muons
> >     and neutrinos. I'd be surprised if that made a big difference,
> though.
> >
> >     I'm also not sure if ATLAS eta/y acceptance cuts are used in the DPD
> >     default output, but that would not be unreasonable. In which case
> Rivet
> >     would find more jets.
> >
> >     However, I just realised that the curves labelled Sherpa1 and Sherpa2
> >     are not cryptically labelled ATLAS truth jets vs. Rivet, but really
> >     Sherpa1 and Sherpa2... right? So we can't actually see your Rivet
> output
> >     because you just sent the index.html file, no plots!
> >
> >     Andy
> >
> >
> >     On 07/11/13 13:52, Liron Barak wrote:
> >     > Hi Andy,
> >     >
> >     > The "ROOT" analysis.... is only making the truthD3PDs from evgens
> >     with:
> >     > Reco_trf.py preExec='from D3PDMakerConfig.D3PDMakerFlags import
> >     >
> >
> D3PDMakerFlags;D3PDMakerFlags.TruthWriteHadrons.set_Value_and_Lock(True)'
> >     inputEVNTFile=1111.pool.root
> >     > outputNTUP_TRUTHFile=11111.truth.1.root
> autoConfiguration=everything
> >     >
> >     > Then I simply do stupid draw command:
> >     > sherpa1.Draw('jet_AntiKt4TruthJets_n','','hist&&norm')
> >     > sherpa2.Draw('jet_AntiKt4TruthJets_n','','hist&&norm&&same')
> >     >
> >     > So I didn't apply any isolation here.....  I would guess it is
> >     also all
> >     > the jets, no?
> >     >
> >     > Thanks
> >     > Liron
> >     >
> >     >
> >     >
> >     > On Thu, Nov 7, 2013 at 1:05 PM, 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:
> >     >
> >     >     On 07/11/13 09:48, Liron Barak wrote:
> >     >     > Hi Andy,
> >     >     >
> >     >     > Now that Rivet is working for me, thanks!
> >     >     > I see something weird....
> >     >     > I took the official sherpa samples (made by sherpa 1) and
> >     compared
> >     >     to my
> >     >     > private sherpa2 samples... I used MC_TTBAR in river to
> >     compare (with
> >     >     > evgens) and simple Draw of the truth jet branch with the
> >     truthD3PDs I
> >     >     > made...
> >     >     > Please see the attached.... the Jet_N for both samples
> >     behave opposite
> >     >     > in the 2 plots, no?
> >     >
> >     >     Hi Liron,
> >     >
> >     >     Can you check if there is lepton/jet isolation in the ROOT
> >     analysis? In
> >     >     Rivet we make jets from *everything* provided to the FastJets,
> >     so hard
> >     >     leptons will count as jets unless they are identified and
> removed.
> >     >
> >     >     This is something which would be nice to improve, as is the
> >     definition
> >     >     of jet inputs, but unlike the ATLAS truth jets we need to be
> >     generic
> >     >     rather than expt-specific. I hope we'll provide an AtlasJets
> >     projection
> >     >     to make this easier at some point, but even ATLAS itself
> >     doesn't have a
> >     >     single truth-jet constituent definition yet!
> >     >
> >     >     Andy
> >     >
> >     >     --
> >     >     Dr Andy Buckley, Royal Society University Research Fellow
> >     >     Particle Physics Expt Group, University of Glasgow / PH Dept,
> CERN
> >     >
> >     >
> >     >
> >     >
> >     > _______________________________________________
> >     > Rivet mailing list
> >     > Rivet at projects.hepforge.org <mailto:Rivet at projects.hepforge.org>
> >     > http://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 Andy Buckley, Royal Society University Research Fellow
> Particle Physics Expt Group, University of Glasgow / PH Dept, CERN
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.hepforge.org/lists-archive/rivet/attachments/20131109/99e75ca5/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: MC_TTBAR_jet_mult.pdf
Type: application/pdf
Size: 11016 bytes
Desc: not available
URL: <http://www.hepforge.org/lists-archive/rivet/attachments/20131109/99e75ca5/attachment.pdf>