Dear All, I have tried to see the changes in asymmetry values after throwing out those known bad bunches. Please take a look at following plots, (a) http://ccjsun.riken.go.jp/~jinnai3/polarimeter/bunch_sel/physics_sel_0-25.ps (b) http://ccjsun.riken.go.jp/~jinnai3/polarimeter/bunch_sel/physics_sel_-13-0.ps (c) http://ccjsun.riken.go.jp/~jinnai3/polarimeter/bunch_sel/yprojection.ps (a)(b) compare the several cut (bunch selection) conditions, for - X physics (BLUE and YELLOW) - Y physics (BLUE and YELLOW) these vlaues are plotted as a function of time. (a) is for the run period after Jan.1st, and (b) is for before Jan.1st. 4 conditions are compared (note that the noise correction are performed on every plots). (i) No bunch selection (ii) Anomalies are discarded (i.e. 1st, 11th, two groups of bad 5-bunches), and the bunches with Chi^2/ndf >=10 are discarded. (iii) Chi^2>=5 and the rests are the same (iv) Chi^2>=3 and the rests are same. Let's focus on the Yellow Y component plots. The anomaly at Day=12,16.5 are reduced (but not disappear) especially at bottom plot. Generally the changes are not drastic. (c) The plojection of plots (a)(b) to Y-axis (Y component only). The asymmetries are divided by statistical error and show the siginificance of the deviation from zero. I think now is a good time to convine the good & bad bunch info from CNI , Phenix and STAR (other than the 1st,11th and two groups). As I showed at the last report, CNI can only provide the chi^2 info but the speicific luminosity info. I guess I can prepair the Chi^2 info at any minutes. Who should I ask for the specific luminosity info from experiments? Could anybody nominate the persons for STAR and PHENIX? -- Osamu From: Osamu Jinnouchi Subject: Re: notes on RHIC polarimeter meeting Date: Thu, 08 Aug 2002 18:09:20 +0900 > > Dear all, > > I have done with the specific luminosity issues, and I have just > made a web page for this special theme. The page is based on the recipe > which Gerry sent out soon after the meeting. Please take a look at, > > http://ccjsun.riken.go.jp/~jinnai3/polarimeter/bunch_sel/ > > The plots and notes are available there. From the study, several things > are understood. > o The criteria Chi^2/ndf <10 has a good margin (most of the case they > are within 5), and with the value 10 we can discard the bunches which > are truly bad. > o Special bunch groups (16-20th, 36-40th) are more or less behaving > strangely in many runs. We can confirm this with the specific luminosity > distribution. Then discarding those bunches from all the runs sounds > reasonable. > > PLAN > For now, I will try to calculate the asymmetry for whole runs with the > following bunch discarding criteria, > - 1st,11th bunches > - 16-20th, 36-40th bunches > - bunches CHI^2/ndf > 10 > If things goes well, I will be able to show you the asymmetry result > tomorrow. > > Please take a look at the plots and give me the comments, suggestions. > regards, > Osamu >