[Rivet] MissingMomentum and WFinder problem & plans for 2.4.0

[Andy Buckley]

Hi all,

A summer student working with Chris and I spotted last week that the W 
returned from WFinder is still nonsensical.

The cuts are ok, and the lepton is also fine, but the use of the full 4D 
missing ET vector to make the "pseudo-neutrino" messes up because that 
vector has an energy component equal to the energy sum of all the 
in-acceptance particles used to calculated the *visible* momentum 
against which the missing one is balanced.

One option would be to reduce the MissingMomentum to the transverse 
vector only, but that seems too far because it should also be usable in 
e+e- colliders. It also seems wrong to give the missing momentum vector 
a null mass, because while that's fine for neutrinos it's wrong for BSM 
invisibles.

I suggest that we change the MissingMomentum interface (again... I 
already made some enhancements for 2.4.0) to only expose a 3-momentum. 
The E component is really a misleading accident. What do you think? Any 
objections or alternative suggestions? (A user-specified missing 
momentum mass target, maybe?)

Similarly, we can't just restrict the WFinder to use transverse mass due 
to e+e- compatibility. But there we do have a definite hypothesis that 
the missing momentum vector should be a neutrino, so we could set the 
pseudo-nu mass to zero, and feed that into the pseudo-W mass in the case 
that full rather than transverse mass is used. Again, thoughts?

Andy

PS. Still waiting for analysis integration volunteers. I'll start 
forcibly assigning them if I don't hear soon (this is the procedure we 
agreed on!), so this is your chance to take a quick look at the tarballs 
and pick a couple of easy ones ;-)

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