[Rivet] Rivet meeting at 10:00 - minutes

[Andy Buckley]

Two addenda:

  * Riccardo to commit his Fox-Wolram moment projection.
  * Emily to commit something small (memory fade... remind me!)

Andy


Andy Buckley wrote:
> Here's some rough minutes of today's phone meeting: please add more if 
> I've forgotten or misunderstood.
> 
> 
> Release schedule and targets
> ----------------------------
> Aim for 1.1.3 release during or shortly after MC4LHC at CERN. Andy will 
> work with Emily & Frank (at CERN) and James & Alex (at UCL) during that 
> week, esp. on finalising the Tevatron Z dilepton definitions in 3 of the 
> analyses. Once 1.1.3 is out, we start work on overhauling the 
> histogramming for 1.2.0, and writing up the manual as a CPC paper to 
> match that version, since we intend the whole design to then be stable.
> ACTIONS: ALL
> 
> 
> Axis label migration
> --------------------
> It's a bit embarrassing that we don't currently have labels on the axes 
> of plots coming from Rivet, but that's mainly because AIDA doesn't 
> support them. So I've hacked LWH to add them, and provided new histo 
> booking functions which take the x & y axis labels as arguments (after 
> the title arg). Using the no-axes versions of the booking functions will 
> result in lots of runtime warnings: the plan is that these should all be 
> eliminated by the time of the release. Please check your own analyses to 
> make sure that axis labels are specified as the histos are booked, and 
> be a good citizen and opportunistically patch up any others that you 
> happen to see!
> ACTIONS: ALL!
> 
> 
> Migrate TrackJet, Jade & D0ILC jet algs to FastJet
> --------------------------------------------------
> We're happy to migrate to FastJet 2.4.0, since we can ensure that Atlas 
> does the same! Andy will do the conversions. We will still be missing 
> the Durham alg, but hopefully we can now just bundle and build Hendrik's 
> Durham version of his Jade plugin with Rivet, rather than having to rely 
> on users to patch FastJet itself.
> ACTIONS: ANDY (& HENDRIK)
> 
> 
> Check your analyses for sensible norms & use of sumOfWeights
> ------------------------------------------------------------
> Emily has removed the ill-advised auto weight counting from vetoEvent: 
> any analysis which was depending on sumOfWeights returning the sum of 
> event weights *to have passed the cuts* should check their analyses to 
> make sure that they are doing the right thing: you may need to add your 
> own weight counter member variable(s).
> ACTIONS: ALL
> 
> 
> DISKinematics & event rotation
> ------------------------------
> James to implement the fixes we decided on (whatever they were!) for the 
> DISKinematics projection. He will also look into the logistics of 
> rotating events conventionally, and assuming it's reasonably 
> straightforward will implement it for this release. We suspect that 
> assuming a per-collider convention will be sufficient, hence no 
> per-analysis versions of the event should be needed. For object 
> ownership reasons, a deep copy of the incoming GenEvent should probably 
> be made as it "enters" Rivet. Note that rotations should be implemented 
> as rotations (e.g. x->x, y->-y, z->-z) rather than parity 
> transformations (all x_i->-x_i) for proper physical invariance.
> ACTIONS: JAMES
> 
> 
> Mac Genser bootstrap script issue(s)
> ------------------------------------
> Emily encountered problems when using the version of the Genser 
> bootstrap script attached to the MacInstallationIssues wiki page. We'd 
> forgotten about this and it was almost a year out of date! James 
> reported that he uses the main agile-genser-bootstrap script for his 
> Mac, so Emily will move to using that and update the wiki page to 
> reflect reality!
> ACTIONS: EMILY
> 
> 
> Analyses: RHIC, UA1, UA5
> ------------------------
> Andy is implementing min bias analyses at the moment, for LHC QCD tunes. 
> James has a couple of HERA analyses that he's interested in. Frank has a 
> couple of CDF jet analyses almost ready to go: Andy will prod Mike about 
> getting some of their data into HepData. Hendrik will be working on the 
> RHIC analyses that he's currently looking at on his US tour. Emily has a 
> CDF analysis ready but is waiting for the data to pass CDF publication 
> committee. Jon commented that we shouldn't get too UE-focused, but I 
> forget which analyses he thought would be interesting ;)
> ACTIONS: ALL, plus Andy to try and chase down Hw++ authors to implement 
> some of their analyses
> 
> 
> What measurements we'd like from RHIC
> -------------------------------------
> Hendrik has clarified that he wasn't just asked about UE analyses, but 
> from our discussion we identified <pT> vs N_ch, standard UE, UE as a 
> function of |eta|, UE with different transverse region sizes, maybe some 
> Gavin/Matteo type jet density stuff. It would be graet to have 
> measurements that can be directly compared to Atlas/CMS first 
> measurements: N_ch, |eta| distribution, low pT jets (incl, 1st, 2nd, 3rd 
> pT, etc.) with good jet algs (anti-kT, SIScone, kT, 
> Cam-Aachen(filtered)) etc.
> ACTIONS: Everyone to have a think about non-UE observables, bearing in 
> mind excellent tracking and PID (any strangeness observables? there's a 
> LEP(?) paper on chain-lke production of strange baryons that could be 
> maybe repeated at RHIC). Andy to relay the message to MC authors.
> 
> 
> Validation
> ----------
> Given that we've done some infrastructure overhauls, validation against 
> the previous version of Rivet is important. Since the validator is most 
> set up for Sherpa at the moment, we will check the same Sherpa events 
> passed through Rivet 1.1.2 and 1.1.3 before releasing, and make sure 
> that there is some decent documentation Web page with plots as part of 
> the Rivet site.
> ACTIONS: ANDY & FRANK
> 
> 
> Other issues
> ------------
>   * MCnet school: Hendrik is in charge, Andy and Emily (?) to get in 
> touch about organisation.
> 
> _______________________________________________
> Rivet mailing list
> Rivet at projects.hepforge.org
> http://www.hepforge.org/lists/listinfo/rivet