Fix analyses with assumed picobarn units
There are quite a lot of analyses with xsec-normalised histograms which aren't explicit about the cross-section units:
analyses$ egrep "crossSection\(" plugin*/*.cc | grep -v barn | wc -l
253
This includes a bunch of MC_
analyses, which should exhibit best practice and hence definitely need fixing. Of the remainder, there are a lot of multiple offences within one analysis, which should be relatively easy fixes.
If this goes smoothly enough, we can also try applying the same approach to e.g. fill()
calls with energy-unit variables that don't have an explicit division by GeV
.
Targetting for release 3.2, since wide-ranging changes in a lot of analysis files are a good way to generate clashes when merging to the main branch.