Hi,

 here some material for the discussion:

>
> Agenda:

>   - Status of run 2 production

for all seeds ~20fb-1 done.

* Riccardo and I submitted the remaning D0's. Around 1 week to finish.
* Alessio is finishing Dstar0 at Lyon (how long does it take, Alessio?)
* I can submit Dstar's with different privilegies but probably I' ll need
  help
* Dc: help wanted

>   - Status of BAD
>   - yield differences between cpf and ibu; bugfixes?

Riccardo fixed the bugs. We are waiting for the results on reduced
collections with new tags and bugfixes. Once we will have them we will say
more. In principle we could gain a factor 10%.

In the meanwhile I prepared the code to optimize S/S+B  (Mx<1.5GeV) as a
function of the cut on the integrated purity. I need the new root files on
reduced collections and on superblock3 to complete these studies.

>   - Generic vs. Cocktail MC

I started the study of the breco components in vubmix-breco MC. I don't
any enhanced channel but I will study much in detail the problem
comparing with generic MC.

>   - Fitting with vubmix

Vubmix has a lower efficiency with respect to the not resonant MC (~30%
lower).

I tried to optimize the cut on Mv^2 with this MC separately for Qtot=0 and
Qtot=+-1.
Here the plots. You find the err(Vub)/Vub in the Y axis for a sample of
100fb-1 (I mixed properly breco components both for Vcb and Vub
Montecarlo).

- Qtot=0, simmetric cut on Mv:

http://www.slac.stanford.edu/~daniele/vub/qeq0simm.eps


- Qtot=0, asimmetric cut on Mv (no cut on the left, cut on the right):

http://www.slac.stanford.edu/~daniele/vub/qeq0asimm.eps


- Qtot=+-1, asimmetric cut on Mv (no cut on the left, cut on the right):

http://www.slac.stanford.edu/~daniele/vub/qeq1asimm.eps


I selected an asymmetric cut at .8 for both (Q=0 and Q=+-1). The error is

ratioBR = 0.0172 +- 0.0033  (9.7% on Vub)


I tried to study the total error (statistical + theoretical) as a function
of Mx cut (now it is 1.5).

http://www.slac.stanford.edu/~daniele/vub/mxcutstudy.eps

As you see, we can go to higher values of Mx cut but we have to understand
how to handle the systematic error that comes from the vcb tail.

I will repeate this study with the full coherent sample (230fb-1).

Daniele



>   - neutrals (pi0, gammas)
>   - KL ???
>   - tape/diskspace
>
> Cheers,
> --U.
>