[Rivet] sneaky Rivet question

[Andy Buckley]

On 17/10/13 22:24, Stephen Mrenna wrote:
> Hi,
> 
> I am sneaking in a Rivet question.
> Is it natural to be able to run Rivet for
> analyses that are performed at different center-of-mass
> energies, or is it expected that the user will handle
> the different MC runs and put them together?

Hi Steve (and I've CC'd the Rivet developer list for wider input),

We've got a mechanism in Rivet's init() stage that looks at the first
event to identify the beam particles, CoM energy, etc. and assumes that
those will be representative of the whole run. So for analyses that
include multiple energies we usually book and fill only the set of
histograms which are relevant for that energy.

Lots of these analyses plot the same observables at each energy, and
typically each histogram is only made up of events at one energy, so
this can be very efficient in terms of code: the analyze() (per-event)
part can be identical and it's only the init() that makes
energy-dependent booking choices.

What we can't really do at the moment is handle histograms which e.g.
plot a trend as a function of sqrt(s). They can be assembled, but it's
not "automatic". With some of the plans that we have for improving
multi-run combination, this could be improved but it's not the main
driver of that development.

So, in short, the user has to handle the run splitting and combination,
but Rivet does try to help a bit... particularly in Rivet version 2,
which should _finally_ get a stable release this week.

Hope that helps,
Andy

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