[Rivet] IMPORTANT: consistency between Rivet reference data and HepData records

[Andy Buckley]

Hi all,

As either official LHC experiment contacts for Rivet analysis submission,
or prominent submitters, I want to draw your attention to an important
update to our submission guide, here:
https://rivet.hepforge.org/trac/wiki/SubmittingAnalyses

​Graeme watt from HepData pointed out that we didn't make any explicit
reference to HepData in our guidelines, and I think the original vision​ of
Rivet being "synchronisable" with HepData (thus picking up dataset bugfixes
as and when they happen) got lost somewhere in the last 10 years of Rivet's
existence.

We have many cases where the reference YODA file in an experiment Rivet
analysis either does not use the same dataset IDs, etc. as in the HepData
record for that analysis, or even cases where the numerical data exists in
Rivet but not HepData!! Both these are sad, for different reasons: to make
Rivet scalable as the LHC anaysis preservation toolkit, we need to divide
and conquer, with HepData in charge of the data curation and Rivet the
code/logic part. Ideally we'd be resyncing Rivet with HepData every time we
make a release, but at the moment that's not possible and we need to get
there somehow.

So to clarify, HepData is absolutely the primary repository for LHC "plot
data" of the sort used by Rivet. And please make sure that the Rivet
routines you submit use the YODA files exported by HepData: with the new
sandboxing/preview features in hepdata.net it should be possible to get
these files before the public release of the analysis... I'm sure Graeme
and the Rivet team will all be happy to help if necessary.

​Thanks​ for all your hard work, and best wishes :-)
​Andy​


-- 
Dr Andy Buckley, Lecturer / Royal Society University Research Fellow
Particle Physics Expt Group, University of Glasgow
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.hepforge.org/lists-archive/rivet/attachments/20171206/924d9cbd/attachment.html>