From vmartin@lotus.phys.northwestern.edu Fri Aug 13 16:48:50 2004 Date: Mon, 28 Jun 2004 00:40:14 -0500 (CDT) From: Victoria Martin To: cdf-muon-offline@fnal.gov Subject: Minutes Minutes of the muon software meeting 23 June 04 =============================================== 0. News Dale has calculated final values for the CMX drift velocity (see his plot and documentation). The values he obtains from both a low-pt sample and a high-pt sample agree. Also the values obtained for the miniskirt agree with the arches. We were a bit confused about the error bars on the plot, but we concluded these are from a projection of a profile plot. We had two discussion points: A. What parameters can we give to Young Kee for her detector page? Phil suggested: reconstruction efficiency, single-wire hit efficiency, drift velocity, residuals and resolution. These are supposed to be performance-type variables, so we agreed to go with single-wire hit efficiency and resolution. Victoria will email the appropriate people to try to collect all of the information. B. What to do with the good run list for bad miniskirt data. Up to now we've defined a starting good run for the miniskirt. However the miniskirt has not been working well in the last two weeks. We have three options to deal with this: 1. Make a separate good run bit for the miniskirt 2. Mark the CMX bad if the miniskirt is bad 3. Redefine the starting good run for the miniskirt. Option 1 is ruled out as it's a hard call for the shift crew to make online, and because it will cause confusion for users later. Previous to the last two weeks we have had only one run where the miniskirt is bad and there has been no concurrent problem in the arches. In the part two weeks, however, there have been a number of problems with the miniskirt. For all of these runs the CMX has been marked bad. Phil is still working out what the luminosity of these runs is. The miniskirt now appears better. If this is the case, we will probably following option 2. If not we may yet go with option 3. We will re-visit this issue in two weeks. (Currently the starting good run for the miniskirt is after the September 2003 shutdown. I normally quote run 175008. However the miniskirt is not providing CMX triggers until run 182408.) 1. Muons in 5.3.1 Victoria has looked at the ~250 pb-1 of data collected before the compromised COT data. We decided to keep the standard cuts the same for 5.3.1 as for 4.11, with the exception of the track-chi2 cut, which has changed due to a more realistic expectation for the track resolution. We discussed what to recommend to users concerning the miniskirt muons. We agree that they are good as of run 175008 (pending resolution of the above issue). However the trigger was only on from run 182408. People using the CMX trigger will need to be careful about calculating their acceptance. The fiducial requirements can be quite confusing for users to apply so we thought about writing a fidMu function in the vein of the fidEle function. Initial values for the ID, trigger and reconstruction efficiencies were reported. No big difference were found between 4.11 and 5.3.1. Victoria also looked at the tracks belonging to BMU muons in the CMUP+track and CMX+track sample. To do this the number of hit requirement was removed from the second leg. This sample is still biased somewhat due to the CMUP or CMX trigger requirement. Most of the muons were found to come from Outside-in tracks, only a few from inside-out tracks, and almost none from Silicon stand-alone. Phil suggested asking the Plug electron people about hit requirements they used for these kind of tracks. One MC issue for 5.3.2 was discussed. For the run dependent MC we are going to turn on the bluebeam from run 154449, and the miniskirt and keystone from runs 175009. However we only have the ability to turn on all of the keystone when, in reality, only wedge 6 is on. We may still have the chance to patch the needed code in, but this looks unlikely. 2. Good run list We discussed the situation with bad run flagged in the CMP from the DQM module. Four runs are flagged due to 2 channels noisy in the CMP (Top). They appear as a spike in the phi distribution of CMP muons. We don't know if the DQM plots are from triggered CMUP muons or just any CMP. We are oriented to not mark bad these runs, but we'll seek more information about the plots and decide soon. ________________________________________________________________ Victoria Martin phone: (+1) 847 491 2811 (nwu) Northwestern University (+1) 630 840 3054 (fnal) email: vjmartin@fnal.gov (+1) 847 414 4034 (cell)