SLD Offline Processing Report for 18-jun-96 Prepared by: RSP from SLDWWW 18 Jun 1996 --------Monitor Evaluation--------- Run # Polar ZSLD #LUM -RTH-- -Ev DSP- -CDC HV- -Overall- ------------------------------------------------------------------------ 34704 .772 19 89 see comments ok ok (?) 34707 .768 110 198 see comments ok ok (?) 34719 .774 201 412 see comments ok ok (?) 34720 .768 189 408 see comments ok ok (?) ------------------------------------------------------------------------ 0.770 519 1107 ------------------------------------------------------------------------ Comments: My comments for these four runs are qualitatively the same as yesterday's. 1) The SDATCHK files give, respectively, 3468, 1940, 1546 and 1403 "lying bastard" (#10) errors. 2) The various quantities checked on the STATS files seem to be ok. 3) The RTH plots look ok except for the phi plot in run 719. Sometimes this is not flat because of poor statistics. However, run 719 seems to have a reasonable sample size. On the other hand, there is no consistent evidence on a run-by-run basis that there is any problem in the phi distributions. 4) The event displays look normal in the sense that I see nothing different, on average, from event displays in the past. The VXD3 display is new and impressive. However, not all is perfect, and for the record let me note what I see as imperfect. a) The CDC often has extra vectored hits. These fall into two categories: "random noise" which is characterized by clusters of hits near each other within a given layer and strings of hits that should have been fit but were not. There are no obvious "hot spots" in the CDC, so I am not worried about occasional noise. However, it is possible that the tracking algorithm can be improved to capture more real tracks. Let me cite a few events for an expert to check. Event 130 in run 34704, event 275 in run 34707, event 139 in run 34719. b) There appear to be cases where CDC tracks don't link with the VXD3 (not surprisingly). As indicated in the summary plots, the linking inefficiency seems to be about 12%. I wonder if the linking can be improved and cite a few example events where it seems that links did not occur. From run 34704, events 130, 1379 and 830. From run 34707, events 275 and 351. 5) As noted yesterday, something is peculiar about the average beam position in y as recorded in the statistics plots. YPRI is not a gaussian on the plot (I made a typo yesterday and said XPRI when I meant YPRI). The histogram labelled VEP (vertex inefficiency vs momentum) still "contains no data." The other two inefficiency histograms, one vs cos(theta) and the other vs phi, still show uniform linking inefficiency of about 12%. I queried yesterday if 12% is ok.