 |
 |
Hi On Wed, 13 Sep 2006, Antonio Petrella wrote: > Hi Heiko, > > Heiko Lacker ha scritto: >> Hi, >> >> to finalize the S/P fixing study we need: >> >> 1. Considering better fitting functions to describe the S/P functional >> dependence, e.g. a Gaussian fit and determine the systematic uncertainty >> Antonio, what is your estimate: Can we do this until Friday? >> > Yes, I will let it work by Friday. > > >> 2. Do this also on higher purity cuts. Time estimate for that? > > once it works it's only a matter of job submission with different cuts. > I need to know: do we want to produce these numbers for the all reweighting > schema (that is: use Wolfgang's latest tag and also the numbers on D** > computed by Chukwudi)? In this case, before fitting the S/P ratio with a > gaussian I have to re-dump weighted dataset and recompute the S/P ratio > itself. At this point, since the talks are over I believe that it will be best if we move to Wolfgang's tag. That is to say recreate the datasets and I'll do the all weighting corr factors then get the latest Dss ratios (which may be more correct). ***** BTW as discussed earlier there may be a bug in the code as the corr factors produced with Dss in the datasets is the say as those without Remember? ********* > ciao > Antonio > >> 3. Apply all reweightings >> * Chukwudi has produced with the correct run-wise MC >> reweighting new numbers for D**lnu/(Dlnu+D*lnu) >> If I interprete the numbers correctly, it is now >> O(0.65) instead of 0.55. >> >> * We need to incorporate in the full reweighting also >> the run-wise MC reweighting using Wolfgang's VVF tag. >> >> Chukwudi & Antonio: What is your time estimate for that? >> >> Beginning of next week would great since we need to >> document it until the end of next week! >> >> 4. Wolfgang: I know that you will not read this before next >> week but are you able to continue on the fallback >> solution next week (18.9.-22.9.)? >> >> * The study on the MC would be very helpful. >> * In addition, the fits on data should be repeated with >> all reweightings applied. >> * If we could also make a comparison of the signal yields >> bin-by-bin between the fallback solution and the (S/P) >> fixing strategy for equal purity cuts that would give >> us also more confidence in case of agreement. >> >> >> Cheers, >> Heiko >> >