[Rivet] Thoughts on a Rivet 2.2.2 release?

[Leif Lönnblad]

On 2015-05-10 13:52, David Grellscheid wrote:
> 1) no automatic guessing
> 2) no physics in YODA (which makes (1) almost impossible anyway)

Let me reiterate what I've advocated several times. The filling of 
histograms in analyze() Rivet should be made with special "fill-only" 
histograms. finalize() should then be forced to only read these 
histograms and produce separate histograms (or data point sets) for 
plotting. The adding of different runs should be made in Rivet by 
manipulating the fill-only histograms (using eg. the integrated 
luminosity of each run as weight), followed by a separate finalize() 
call to produce the new plots.

(The special fill-only histograms could then also be made to work with 
NLO counter events and such stuff as we discussed in Copenhagen)

Of course, I do not advocate doing this for v2.2.2.

/Leif