[Rivet] Some suggestions

[Chris Wymant]

Hi Rivet team,

A few suggestions:

Declaring two different histograms with the same string name, e.g.
      _hist_var1 = bookHistogram1D("Name",30,0,1000);
      _hist_var2 = bookHistogram1D("Name",30,0,1000);
results in the error
'Segmentation fault (core dumped)'
if an event is analysed for which one attempts to fill both histos.
Perhaps this could be caught and a more helpful error message displayed?

Clicking on most of the FourMomentum members (e.g. 'add') at
https://rivet.hepforge.org/code/dev/a01151.html
links to the CMS_2011_S8957746 Class Reference page. I'm not sure if this is intentional; it doesn't seem to be very relevant.

The description of the operator overloading of '+' for the FourMomentum class is "Subtract from this 4-vector. NB time as well as space components are subtracted."

A potentially controversial one - the rivet definition of transverse energy Et() is E*sin(theta). I've never come across this definition in papers (though quite possibly that's just my lack of experience), I've only seen it defined as sqrt(m^2 + pt^2), e.g. for ATLAS H->WW. The two definitions coincide for massless four-vectors but not otherwise. Changing the function Et() I guess is not a good idea since some previous analyses will have consciously chosen it with E*sin(theta) in mind, but perhaps an EtAlternative() could be thrown in for the sake of those unaware that there are two definitions, to prompt them to check they're choosing the right one?

Thanks & best wishes,
Chris Wymant