========================= Run and Fill selections ========================= Condition : From the plots showing the number of dicarded bunches only for the measurements are flat top, the threshold is determined as, Number of bad bunches < 8 --> Keep Number of bad bunches >=8 --> Discard for Chi^2/ndf>5 distribution. The decision was made to cut the bump structure at n=10. c.f. If we want to assign the threshold at the point where the 90% of the measurements are kept, following numbers can be the reference. Chisq/ndf >3 ... n<=9 rate=0.89 n<=10 rate=0.92 Chisq/ndf >5 ... n<=4 rate=0.86 n<=5 rate=0.90 Chisq/ndf >10... n<=1 rate=0.83 n<=2 rate=0.93 For instance, if we apply the cut chisq/ndf>3, and use the CNI measurements which contains less than 10 bad bunches, 89% of the whole measurements are kept. -------------------------------------------------------------- fill# CNI run#(Ring) comment 2140 465B,467B only 469B is availabe for this fill 2151 494B discard the fill 2153 499B 2162 541B 2193 607B discard the fill 2196 614Y discard the fill 2207 634B 2208 650Y,656Y,658Y 2211 667B discard the fill 2212 670B 2218 688B discard the fill 2226 713B discard the fill 2244 735Y,737Y 2246 741B,746B 2251 750Y 2268 808B discard the fill 2269 812B,814B,815B discard the fill 2289 858Y,859B 2290 865B 2320 946Y