Hi Daniele,
the plot you want is in
http://www.slac.stanford.edu/~rfaccini/phys/vub/testbreco.ps
(hist: generator, full dots no cut, open dots breco).
The open dots are worse than Alessio's because there is no cut on purity
and mes. I checked that cutting we go in the right direction, but I did
not run enough events to make a decent plot.
Anyhow this is not relevant. My point is:
- the distributions that Alessio shows are used as input to
calculate the weights and disagree
- the disagreement is due to the cut on Breco
- we take into account this effect in our measurement because we
have the difference in tag efficiency accounted for and then we compute
epsilon_x relative to the Breco sample which is the correct thing to do.
- the only weak point of the measurement we are publishing is that
the weights for the hybrid etc etc ... are done at ntuple level while we
should have done it at generator level. This seems to have little impact
given the fact that we see onyl a 2.5% shift and <10% variation in the sys
error (reduction so at most we are conservative).
- I think we can go ahead with the studies with the 3-D weights
ciao
ric
On Fri, 13 Jun 2003, Daniele del Re wrote:
>
> > Hello,
> > I think I understood the difference between the 1d weights I compute at
> > generator's level and the one alessio computes on our ntuples AFTER
> > requiring the presence of a Breco. The
> > difference was already found to be related to the different mx shape
> > on non-resonant Vub MC
> > ( http://www.slac.stanford.edu/~asarti/recoil/mxq2/TestRicgen.eps )
> >
> > I run on non-resonant Vub MC dumping all the evnts and produced the mx
> > distribution with no cuts (top) , requiring the existence of a Breco
> > (middle) and compared them with the plot I produced at generator's level
> > (bottom):
> > http://www.slac.stanford.edu/~rfaccini/phys/vub/testbreco.ps
>
> I am very confused.
> May you overlap the plots? the overlap of the middle and the bottom ones
> should give the same plot that Alessio showed before (
> http://www.slac.stanford.edu/~asarti/recoil/mxq2/TestRicgen.eps )
> , correct?
> Here the effect seems to be really larger.
>
> Top and bottom ones should be identical, correct? They look different...
>
> Daniele
>
|