[Rivet] Additional data removed from rivet

[Andy Buckley]

On 03/06/14 11:45, David Grellscheid wrote:
> Hi Andy,
> 
> Can you please take a look at Rivet a355f9d65ba0?
> 
> That changeset removed a number of histograms (e.g. d02-x01-y03..19)
> that Peter had obtained data for privately. That dataset is not (and
> AFAIK has never been) in Hepdata.

D'oh...

The rest of the data in the HepData tables was unusably mangled because
(again!) the bin widths hadn't been accounted for when the data was
entered. I assumed that no-one had been using it until now, since it was
so obviously wrong!

> I hope we haven't lost data in this manner elsewhere. We also need a
> flag to indicate when the YODA file is not taken 1-to-1 from Hepdata and
> needs to be preserved.

Yes. Or ideally get the missing data into HepData: isn't that the
preferred way? I feel a bit uncomfortable about the idea that data was
sufficiently private that it can't go in HepData, yet somehow it's ok
for it to be in public Rivet releases...

I would far prefer that we point ourselves toward taking everything
exactly as it is in HepData, except (?) in the case that the paper
doesn't have a HepData record at all. Then we could think about
re-dumping all records periodically to benefit from data bugfixes,
format changes, or system improvements without fear of losing important
hacks. Not the current situation, unfortunately!

Andy

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