[Rivet] Rivet 2 validation nearly done, and 2.0.0b3 tarball

[Andy Buckley]

On 15/08/13 11:41, Frank Siegert wrote:
> Hi Andy,
> 
> thanks for the update and work!

That's ok... I just want this damn thing out so we can start thinking
about improving physics content, funkier stats stuff, etc.! Can't
pretend that the validation has been a lot of fun, but at least it's
nearly done and I think it's been a very healthy thing for Rivet that we
were forced to take a close look at so many analyses.

> Quick question:
> 
>> The JADE_OPAL jet rates analysis also has some differences w.r.t. the
>> AIDA version, because we changed the definition of which x-values are
>> used to define the integral upper limit. Hendrik and I looked at this at
>> Les Houches and what's implemented in the YODA branch is the best
>> choice, we think. The difference between using bin centers, foci, or
>> lower edges to define the integral limit is quite substantial, so a
>> second set of eyes on this analysis code would be very welcome. I
>> mention it particularly because I think this analysis was used for
>> shower tuning: David's H++ validation events look worse with the new
>> definition with the old. Perhaps/hopefully the new version is more
>> compatible with other data... worth cross-checking.
> 
> Does this only affect the integrated distributions (_h_R_Jade,
> _h_R_Durham) or also the differential ones (_h_y_Durham)? I think for
> tuning typically the latter are being used(?) and they should
> hopefully not be affected.

For some reason I only have the differential plots for the YODA version
-- the AIDA plots from David's reference run don't seem to include the
differential ones. But the code/algorithm ambiguity about the correct
"cut" value to be used is for the integral histograms only, so the
differentials should be unaffected. Agreed that the differentials (also
for jet shapes) are the better choice for tuning!

Andy

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