 |
 |
Hi, there are two things 1) We need to document what has been done so far. 2) Concezio pointed me to something which we likely have not treated correctly in this study, namely, the error on S/P in the data- depleted sample. Since S and P are obtained from a mES fit, both are correlated. As a consequence, the error propagation formula to calculate the uncertainty on S/P(data-depl.) needs to be modified taking into account this correlation. Defining R=S/P(data-depl.), the relative error on R is given by: (sigma_R/R)^2 = (sigma_S/S)^2 + (sigma_P/P)^2 - 2 cov(S,P)/(S*P) That is, the error up-to-now is overestimated. Cheers, Heiko On Tue, 3 Oct 2006, CHUKWUDI KWEKU CLARKE wrote: > Hi all, I'm not quite sure how to proceed with respect to the S/P studies > as Antonio has obtained results for the full weighting and they appear to > be unstable? > C.C. > > On Mon, 2 Oct 2006, Antonio Petrella wrote: > > > Hi Chukwudi, > > actually I don't know how we want to proceed on S/P studies: it seems to me > > that from the systematic studies the situation is not stable. > > For Mx-q^2 and P+ analysis, before leaving Italy I was setting up also the > > fittest.C macro to be ready. But probably fixing S/P will not help too much > > (?) > > > > Cheers, > > Antonio > > > > > > > > CHUKWUDI KWEKU CLARKE wrote: > >> Hi Antonio, you forgot to tell me if about the work being done for the > >> Mx-Q^2 and P+ S/P studies. I'll look into that later today. > >> Cheers, C.C. > >> > >> On Sun, 1 Oct 2006, Antonio Petrella wrote: > >> > >>> Hi Chukwudi, > >>> now I'm at SLAC. I don't know how much time I will spend on analysis since > >>> I will have to attend to IFR software as well. > >>> > >>> Anyway, when are you going to be at SLAC again? > >>> > >>> ciao > >>> Antonio > >>> > >>> CHUKWUDI KWEKU CLARKE wrote: > >>>> Hi Antonio, I'm sorry for being absent for several days I had to get my > >>>> things in order before I flew back to the UK. I'm now back for a couple > >>>> of weeks. When will you be going to SLAC? Do you need help, what's being > >>>> done for the Mx-Q^2 and P+ S/P studies? With alittle help I could look > >>>> into these analyses if you want? > >>>> Cheers, C.C. > >>>> P.S. I'm still abit jet lagged but don't hesitate to ask for my help. > >>>> > >>>> On Wed, 27 Sep 2006, Antonio Petrella wrote: > >>>> > >>>>> Hi Heiko > >>>>> the MX fit plots are available at: > >>>>> > >>>>> > >>>>> http://www.slac.stanford.edu/~petrella/tmp/scra/Ibuzero_nofit_ip020/zero_nofit_ip020fitresults.eps > >>>>> for the ENRICHED SAMPLE > >>>>> > >>>>> and at: > >>>>> > >>>>> http://www.slac.stanford.edu/~petrella/tmp/scra/Ibuzero_nofit_ip020_depl/zero_nofit_ip020_deplfitresults.eps > >>>>> for the DEPLETED SAMPLE. > >>>>> I have used all reweightings and set the D**+other/(D*lv+Dlv) ratio to > >>>>> 0.428585. > >>>>> > >>>>> The third bin (mx=[1.9,2.2] GeV) changes significantly. > >>>>> > >>>>> Ciao > >>>>> Antonio > >>>>> > >>>>> Heiko Lacker ha scritto: > >>>>>> Hi Antonio, > >>>>>> > >>>>>> can we also have a look at the plots for the MX fits, both for > >>>>>> signal-enriched and signal-depleted? > >>>>>> > >>>>>> I'm curious, after all re-weightings discussed so far, > >>>>>> how the fit performs above 2 GeV. > >>>>>> > >>>>>> Cheers, > >>>>>> Heiko > >>>>>> > >>>>>> On Tue, 26 Sep 2006, Antonio Petrella wrote: > >>>>>> > >>>>>>> Hi, > >>>>>>> on the web page where I posted results of systematics due to > >>>>>>> randomization of S/P: > >>>>>>> > >>>>>>> http://www.slac.stanford.edu/~petrella/systsp.html > >>>>>>> > >>>>>>> you can see that this systematic error is not stable, for example when > >>>>>>> cutting on integrated purtity. > >>>>>>> > >>>>>>> Now I'm trying to look at other results from these jobs to see if I > >>>>>>> can find what makes this errors so large, but probably this is also > >>>>>>> due to the S/P ratio and its error. > >>>>>>> > >>>>>>> For example if you look at the correction factors for IP > 0.50 > >>>>>>> (http://www.slac.stanford.edu/~petrella/tmp/SP_allrew/SPallweights/ip050_allrew/corrallwip050pol1.eps) > >>>>>>> you can see that the first bin has a large error (the exact value of > >>>>>>> the correction factor for this bin is S/P = 5.67 +- 5.34) > >>>>>>> > >>>>>>> These numbers (they're on the spreadsheet at > >>>>>>> http://www.slac.stanford.edu/~petrella/tmp/SP_allrew/SPallweights/SoverPFullRew.sxc) > >>>>>>> come from the double ratio of S/P on MC (0.74 +- 0.13) times the S/P > >>>>>>> ratio on data depleted sample > >>>>>>> > >>>>>>> (http://www.slac.stanford.edu/~petrella/tmp/SP_allrew/SPallweights/ip050_allrew/data_depl_AC_intp0.50_0.001.55.eps) > >>>>>>> On data depleted sample the signal component (fitted) is 291 +- 31 and > >>>>>>> the background component (fitted) is 38 +- 35, so the error on the > >>>>>>> final S/P ratio is driven by the background component on data depleted > >>>>>>> sample... and cutting on purity (and having less background) will give > >>>>>>> roughly higher errors on background component (at least the > >>>>>>> statistical error). > >>>>>>> > >>>>>>> For the data depleted sample we get S +- dS and P +- dP as they come > >>>>>>> out from the fit and then we compute the quantity S/P +- d(S/P). But > >>>>>>> these errors are correlated, aren't they? > >>>>>>> > >>>>>>> ciao, > >>>>>>> Antonio > >>>>>>> > >>>>> > >>> > > >