Hi,
> > we have a question concerning your comment about the separate treatment of
> > charged and neutral B. We are using the VirVubFitter code, meaning that
> > the correctionratios for MC are calculated (in VirUtil.icc, functions
> > mixingCorr and chargeCorr) and applied if I am not mistaken.
> > Did your comment mean that we should apply this correction (which is
> > applied)? Shouldn't this take care of possible differences in data and MC?
>
> as far as the BRBR is concerned it is done automatically.
>
> Are you doing the same when calculating your response matrix?
At this point, we do no reweighting related to the B charge when we
calculate the response matrix and the xini spectrum.
This is what I found having a look at the VirVubFitter code:
For signal MC the correctionfactor is correctionratiovub 0.969355, so it
is something like a 3% effect. This factor is used to scale the charged B
events of the signal MC that is used for the BRBR extraction.
What I would do now is to apply this scaling factor to the truth
level charged B events in the signal MC used for the determination of the
response matrix and xini, leaving the neutral B events untouched.
Do you agree that it can be done this way?
We only expect a very small effect, given that the rescaling factors for the
multiplicity categories are much more different from 1 (see Table 3 in
BAD 894). We wonder if this rescaling should be done given the size of the
weight.
Cheers,
Kerstin
|