[Rivet] Transverse Mass for WFinder

[Gavin Hesketh]

Hi Andy,
I think that works. It's also calculated in MC_WJETS.cc for example

const double transM = sqrt(ET*ET - px*px - py*py);

cheers,
	Gavin


On 25/07/11 18:02, Andy Buckley wrote:
> On 25/07/11 16:46, Andy Buckley wrote:
>> On 25/07/11 16:40, Gavin Hesketh wrote:
>>> Hi,
>>> the WFinder in Rivet currently allows you to set a min and max mass.
>>> These should (in 99% of W analyses) be cuts on the transverse mass, not
>>> the full invariant mass.
>>> Can these mass cuts be changed by default to operate on the transverse
>>> mass? Or add some flag to the WFinder to switch between transverse and
>>> full mass? Or even just add the transverse mass as a method for the
>>> InvariantMassFinalState?
>>
>> Hi Gavin,
>>
>> Thanks for the pointer... I'll change the WFinder to make the cuts on
>> the transverse mass.
>>
>> I don't want to mess with other projection interfaces too much before
>> version 2.0, since I want to introduce a general set of "momentum
>> selecting" objects to replace the current (incomplete) plethora of
>> different constructor arguments, properties, methods, etc. ... you'll
>> see what I mean when it appears ;)
>
> Hmm, on looking at this in more detail, I couldn't find a less invasive
> way of doing this than to add a useTransverseMass flag to the
> InvMassFinalState and to always use that in the WFinder implementation.
> It's in the Rivet trunk now.
>
> I also added massT and massT2 methods to the Rivet FourMomentum, for
> getting hold of the transverse mass... this will force a complete
> rebuild of any checked-out Rivets, unfortunately! Just to be completely
> sure, Gavin, by "transverse mass" you mean mT = sqrt(E^2 - |p|^2) *
> sin(theta) where theta is the usual polar angle w.r.t. the z-axis, cf. Et?
>
> Cheers,
> Andy
>