Hi Ric,

  as you know, purity at root code level and purity at Beta
level did not match in release 10. I fixed that for new skims and new
tables in beta should be as the ones used in bad 540.
As a consequence the cut you are talking about should be already in place.
if you remove sb4 you will take only modes that satisfy cut at page
82 of BAD 540.

rate on data went down to 1.6% (from 4%) and to below 7% for gene
MC.

Urs could you take into account this rates and recompute your numbers as
far as disk needed?

Daniele

On Fri, 16 Apr 2004, Riccardo Faccini wrote:

> Hi,
> I guess you are right
> my next suggestion is to cut on purity from the start, i.e. consider only
> events with a higher purity of the indivual modes.
> http://www.slac.stanford.edu/~rfaccini/lost+found/purcut.eps
> shows the distribution of the purity for generic MC events and the cuts
> that we apply at analysis level (we require the events to have a purity
> higher than a given one).
> It looks like we can save quite some time and space if we request for
> istance that all events have at least one candidate with a 8% purity.
>
> 	opinions?
> 	ciao
> 	ric
>
> On Fri, 16 Apr 2004, Oliver Buchmueller wrote:
>
> >
> >
> >
> > It is certainly worth to test whether or not the normalization to
> > semileptonic events will work and perhaps even improve the
> > extraction of the BR(bsg). However, given the fact that there
> > is at least one thesis a stake I would still vote for a production
> > without cuts - unless someone can proof that an alternative way will work
> > as well.
> >
> > my two cents ...
> >
> > Oliver
> >
> >
> >
> > On Fri, 16 Apr 2004, Riccardo Faccini wrote:
> >
> > > hmmm, I see the problem ( ehem and I was among those who thought
> > > Fabio's thesis...).
> > > You can normalize to the number of semileptonic events. In this case you
> > > will be affected by the systematics on the cut on the lepton, but I think
> > > it will balance the systmatics on the mes fit which will be reduced.
> > >
> > > What do people think?
> > > 	ciao
> > > 	ric
> > >
> > > ______________________________________________________
> > > Riccardo Faccini
> > > Universita' "La Sapienza" & I.N.F.N. Roma
> > > tel  +39/06/49914798 Fax.: +39/06/4957697
> > > http://www.slac.stanford.edu/~rfaccini
> > > Univ. La Sapienza. 2,Ple Aldo Moro, I-00185 Roma Dipartimento di Fisica
> > >
> > > "I don't understand what you say, but I believe I disagree"
> > >
> > > On Fri, 16 Apr 2004, Henning Ulrik Flaecher wrote:
> > >
> > > >
> > > >
> > > > Hello,
> > > >
> > > > I just noticed that the filtering on leptons and photons above a certain
> > > > energy cut is (most likely) not an option for the bsg analysis as for the
> > > > branching fraction measurement we need to normalise to an unbiased B
> > > > sample and so the full Breco sample.
> > > > This is how it has been done for Fabio's thesis.
> > > > Requiring a lepton or photon with a certain energy will most likely bias
> > > > our normalisation sample, e.g. all B->charged hadron decays would be
> > > > lost, a fraction of the SL decays etc.
> > > > The reason why the b->ulv analysis can live with this cut is because they
> > > > measure a double ratio of branching fractions, so they can normalise to a
> > > > sample with the same cuts applied.
> > > > At the moment I can't see how we can get around this but appreciate any
> > > > ideas!
> > > >
> > > > Cheers,
> > > > Henning
> > > >
> > >
> >
>