[Rivet] yodamerge for Histo1D

[Frank Siegert]

Short update on the simple run merging with yodamerge:

Apart from the plotting issues, which are solved with recent trunk of
yoda and rivet, the problem in my example is the normalisation. As far
as I can tell, the yodamerge does not merge the ScaledBy property at
all -- it simply uses the one of the last yoda file on the command
line, right? So this isn't going to give the correct answer yet.

I wonder how we can do a pragmatic merging of ScaledBy -- in
histograms it should typically be
someFactor*crossSection()/sumOfWeights(), but we don't know
sumOfWeights() and could in principle also allow for someFactor \neq
1.0 which we don't know how to merge.

Am I missing something?

Cheers,
Frank

>>> Actually, we can do simple run merging at the YODA level now, thanks to
>>> Dave M putting together the remaining Python += operators for YODA
>>> histograms, and a yodamerge script which uses them to combine multiple
>>> runs into one. It's very simple at the moment, but we will add a command
>>> line way to specify weights for each input run: will that be enough to
>>> combine *different* processes by cross-section? Please try it out... it
>>> *seems* to be working nicely, given a simple test.
>
> I have tested this now using three runs with 500, 1000 and 2000
> events, but it doesn't seem to be working for me:
>
> http://fsiegert.web.cern.ch/fsiegert/tmp/yodatest/plots/MC_ZINC/index.html
>
> I analysed, merged and plotted the three samples using
>
>  $ rivet -a MC_ZINC -H Rivet.${nevt}.yoda ${nevt}.hepmc
>  $ yodamerge -o Rivet.yoda Rivet.*.yoda
>  $ rivet-mkhtml Rivet.yoda Rivet.*.yoda
>
> Does that sound correct? Now this is my first test after switching to
> hg and it's completely possible that I screwed something up. This is
> with the hg repo from two days ago (I have yet to understand the full
> update workflow in hg ;-))
>
> I've uploaded the event files to
> http://fsiegert.web.cern.ch/fsiegert/tmp/yodatest/{500,1000,2000}.hepmc
> if somebody wants to give it a try.
>
> Cheers,
> Frank