Hi Wolfgang,

thanks for the clarification.
I was not sure how to interpret the first row of plots.

The second row shows that the shapes for charged and neutral
B background are different and hence I would have expected
that this should improve the fit when taking into account
this difference in shape.

It is not clear to me how this difference is taken into
account in the fit to the full sample. Can you comment on
that? I'm not sure what to compare with when comparing
chi^2 values.

Cheers,
Heiko

On Mon, 10 Apr 2006, Wolfgang Menges wrote:

> Hi Heiko,
> 
> Here is a short summary. Antonio fitted the MC background sample (Breco
> with wrong Breco reconstruction) seperately for neutrals and charged Bs.
> See http://www.slac.stanford.edu/~petrella/mesfits/mesfits.html, second
> set of plots from 06 - Apr - 2006.
> 
> Both fits converge but show the same behaviour as the fit to the full set
> (first plot in frist set of plots from 06 - Apr - 2006. The peak region is
> not fitted vey well. Here are the chi2 for all three fits:
> 
> 	full sample: 2.7769
> 	neutral Bs: 3.2169
> 	charged Bs: 2.48632
> 
> You can see an improvement in the charged Bs compareed to the full sample,
> but at the same time the chi2 for the neutral Bs worsen.
> 
> Antonio and I concluded that this does not help us to improve the fit
> around the peak and we didn't do any further studies or fits. So we drop
> this option.
> 
> Heiko, this is the opposite conclusion as yours. Can you tell me why you
> think the fit i improved?
> 
> Cheers,
> 
> 	Wolfgang
> 
> On Mon, 10 Apr 2006, Heiko Lacker wrote:
> 
> > Hi,
> >
> > I just realized that Antonio up-dated his web page already last Thursday
> > http://www.slac.stanford.edu/~petrella/mesfits/mesfits.html
> >
> > Antonio, could you please post a short summary?
> >
> > Looking at the plots for the separate fits to charged and neutral B
> > background it seems that a separation for these BG's should improve
> > the fit.
> >
> > If I'm not wrong the plots in the first row show fits when combining
> > charged and neutral B's. Did you try to separate the fits into charged
> > and neutral B's in order to see if the two samples can be described in
> > a better way by the two fits?
> >
> > Alternatively, one could fix the BG parameters for both classes and also
> > the ratio between these two BG classes on the MC in order to see if the
> > combined fit describes the data in a better way..
> >
> > Cheers,
> > Heiko
> >
> 
> -----------------------------------------------------------------------------
> Wolfgang Menges
> Queen Mary, University of London                 SLAC, MS 35
> Mile End Road                                    2575 Sand Hill Road
> London, E1 4NS, UK                               Menlo Park, CA 94025, USA
> ++44 20 7882 3753                                ++1 650 926 8503
>                                                  [log in to unmask]
> -----------------------------------------------------------------------------
>