2009 Spin QA: 200GeV Part II

The purpose of this follow-up analysis to the Spin QA for the 2009 200GeV run is to investigate fills in which production or zdc_polarimetry runs were discarded and determine why they had to be discarded and if it is possible to assign them a spin pattern. To help with this task, I created two new tickle scripts, spinBitCrawler.tcl and crawlerComp.tcl. The spinBitCrawler script starts at the beginning of a fill and quiries the cdev database every N seconds and prints out the polarization pattern as well as generating a .v124 file which contains information about the intended fill pattern and polarization pattern. The crawlerComp script looks at the .v124 files and determines which sequential pairs do not match. Using these two scripts, I can pinpoint any wierd behavior in cdev which may cause a fill to be bad.

This analysis will focus on 14 fills: F10763, F10754, F10783, F10791, F10850, F10960, F10642, F10651, F10688, F10695, F10781, F10785, F10825, and F10962. See the original QA page as well as Jim's QA page for details.

Most of the runs I had to remove from the above fills where zdc_polarimetry runs which came very close to the end of the fill. Often, the timing of these runs were such that with the additional 10 min. added by the spin script, the last cdev entry would extend into the next fill meaning that the run would pick up the polarization pattern of the following fill and be marked as incompatable with the fill it was actually in.

F10763:

Fill 10763 contains 29 runs (10138073 - 10139022) and spans times 21:10:54 (5/18) to 06:54:53 (5/19). In the previous analysis I found I had to remove 2 runs from the end of the fill to obtain a consistent spin pattern. These runs were: 10139021 (zdc_pol) and 10139022 (fms). Run 10139021 starts at 06:27:46 and ends at 06:32:34. The output of the spinBitCrawler script shows that the change in polarization pattern occurs between 06:39:34 and 06:44:36. The polarization pattern for time stamp 06:44:36 is that of fill 10764. The RICH scalars show that the change in polarization pattern happens after the beam is dumped. It also shows that run 10139021 ends concurrently with the beam dump. With this evidence it is clear that run 10139021 can be included with fill 10763 and that the end fill time can be set as 2009-05-19 10:32:34 GMT (1242729154 Unix time).

F10754:

Fill 10754 contains 25 runs (10136005 - 10136045) and spans times 00:59:06 (5/16) to 11:27:57 (5/16). In the previous analysis I found I had to remove 6 runs to obtain a consistent spin pattern. These runs were: 10136005 (fms), 10136006 (fms), 10136041 (production), 10136042 (production), 10136044 (zdc), and 10136045 (fms). The output of the spinBitCrawler script shows that there is a change in polarization pattern between 01:17:40 and 01:22:43, this corresponds to the two fms runs at the beginning of the fill and does not concern us. There is another change in polarization pattern between 10:47:37 and 10:52:41. The polarization pattern at 10:52:41 is that of fill F10755. The RICH scalars show that the change in polarization pattern occurs during the fill. The scalars also show that the fill ends concurrently with the end of run 10136044 (ends at 11:17:20). Although the polarization pattern in cdev changes during the fill, I could not find any other evidence to suggest that anything changed the beam polarization. Therefore, I believe it is reasonable to include runs 10136041, 042, and 044 in fill 10754 and extend the end fill time to 2009-05-16 15:17:20 GMT (1242487040 Unix time).

F10783:

Fill 10783 contains 27 runs (10143070 - 10144008) and spans times 15:13:05 (5/23) to 02:17:34 (5/24). In the previous analysis I found I had to remove 4 runs to obtain a consistent spin pattern. These runs were: 10143070 (fms), 10144005 (zdc), 10144006 (fms), 10144008 (fms). Run 10144005 starts at 01:58:55 and ends at 02:04:21.The output of the spinBitCrawler script shows that there is a change in the polarization pattern between 15:28:42 and 15:33:44, this corresponds to the fms run at the beginning and does not concern us. There is another change in polarization pattern between 02:07:28 and 02:12:31. The polarization pattern at 02:12:31 is that of fill 10784. The RICH scalars show that the change in polarization pattern happens after the beam is dumped. The scalars also show that run 10144005 ends concurrently with the beam dump. With this evidence it is clear that run 10144005 can be included in fill 10783 and that the end fill time can be extended to 2009-05-24 06:04:21 GMT (1243145061 Unix time).

F10791:

Fill 10791contains 11 runs (10146059 - 10146093) and spans times 11:29:41 (5/26) to 18:56:44 (5/26). In the previous analysis I found I had to remove 2 runs to obtain a consistent spin pattern. These runs were: 10146059 (fms) and 10146093 (zdc). Run 10146093 starts at 18:37:27 and ends at 18:46:44. The output of the spinBitCrawler script shows that there is a change in the polarization pattern between 11:59:37 and 12:04:39, this corresponds to the fms run at the beginning and does not concern us. There is another change in polarization pattern between 18:53:14 and 18:58:16. The RICH scalars show that the change in polarization pattern happens after the beam is dumped. The scalars also show that run 10146093 ends before the beam is dumped. With this evidence it is clear that run 10146093 can be included in fill 10791 and that the end fill time can be extended to 2009-05-26 22:46:44 GMT (1243378004 Unix time).

F10850:

Fill 10850 contains 15 runs (10153119 - 10154004) and spans times 18:39:47 (6/2) to 02:51:58 (6/3). In the previous analysis I found I had to remove 3 runs to obtain a consistent spin pattern. These runs were: 10153119 (fms), 10153120 (fms), and 10154004 (zdc). Run 10154004 starts at 02:32:12 and ends at 02:41:58. The output of the spinBitCrawler script shows that there is a change in the polarization pattern between 02:40:38 and 02:45:40. The RICH scalars show that the change in polarization patterns happens during the fill. They also show that run 10154004 ends well before the beam is dumped. Even though the polarization pattern in cdev changed during the fill, I could find no other evidence to suggest that anything changed the polarization pattern of the beam. Therefore, I believe it is reasonable to include run 10154004 in fill 10850 and extend the end fill time to 2009-06-03 06:41:58 GMT (1244011318 Unix time).

F10960:

Fill 10960 contains 12 runs (10171026 - 10171050) and spans times 09:50:41 (6/20) to 18:36:44 (6/20). In the previous analysis I found I had to remove 2 runs to obtain a consistent spin pattern. These runs were: 10171026 (production) and 10171050 (fms). Run 10171026 starts at 10:10:41 and ends at 12:03:10. The RICH scalars show that time 10:10:41 is in fill 10959 and that time 12:03:10 is at the start of fill 10960. So run 10171026 spans the two fills. Looking at the shift log, it appears that cybersecurity blocked access to daqman for a while so the shift crew was not able to controle the run. I will include run 10171026 in fill 10959 and extend the time of valid spin information to 11:10. This is approximately where the beam is dumped according to the RICH scalars. So even though run 10171026 will continue past this point, there will be no valid spin information after 2009-06-20 15:10 GMT (1245510600 Unix time).

F10642:

Fill 10642 contains 4 runs (10112016 - 10112037) and spans times 06:18:05 (4/22) to 11:14:11 (4/22). In the previous analysis I found I had to remove 2 runs to obtain a consistent spin pattern. These runs were: 10112016 (zdc) and 10112017 (zdc). Run 10112016 starts at 06:38:05 and ends at 06:40:08. Run 10112017 starts at 06:41:10 and ends at 06:46:11. The output of the spinBitCrawler script shows that there is a gap in the cdev records. A querey time of 06:18:05 returns a cdev record at 07:44:45, so there is no cdev information for the first part of this fill. The RICH scalars show that these two runs are part of fill 10642. This fill consists only of these two runs and two other vernier scans between ~ 10:40 and 11:00.

F10651:

Fill 10651 contains 32 runs (10113053 - 10114005) and spans times 15:57:30 (4/23) to 01:38:37 (4/24). In the previous analysis I found I had to remove run 10114005 (zdc) to obtain a consistent spin pattern. Run 10114005 starts at 01:24:33 and ends at 01:28:37. The output of the spinBitCrawler script shows that there is a change in the polarization pattern between 01:37:33 and 01:42:35. The RICH scalars show that the change in polarization pattern occurs after the beam is dumped. They also show that run 10114005 is completely within fill 10651. With this evidence it is clear that run 10114005 can be included in fill 10651 and the fill end time can be extended to 2009-04-24 05:28:27 (1240550917 Unix time).

F10688:

Fill 10688 contains 20 runs (10122004 - 10122033) and spans times 23:50:39 (5/1) to 09:09:09 (5/2). In my previous analysis I found I had to remove run 10122033 (production) to obtain a consistent spin pattern. Run 10122033 starts at 08:28:19 and ends at 08:59:09. The output of the spinBitCrawler script shows that there is a change in the polarization pattern between 09:08:36 and 09:13:39. The RICH scalars show that the change in polarization pattern occurs right after the beam is dumped. They also show that run 10122033 is completely within fill 10688. With this evidence it is clear that run 10122033 can be included in fill 10688 and the fill end time can be extended to 2009-05-02 12:59:09 GMT (1241269149 Unix time).

F10695:

Fill 10695 contains 22 runs (10123083 - 10124029) and spans times 22:37:33 (5/3) to 04:33:43 (5/4). In my previous analysis I found I had to remove 2 runs to obtain a consistent spin pattern. These runs were 10124028 (zdc) and 10124029 (fms). Run 10124028 starts at 04:10:04 and ends at 04:16:07. The output of the spinBitCrawler script shows that there is a change in the polarization pattern between 04:23:38 and 04:28:40. The RICH scalars show that the change in polarization pattern occurs right after the beam is dumped. They also show that run 10124028 is completely within fill 109695. With this evidence it is clear that run 10124028 can be included in fill 10695 and the fill end time can be extended to 2009-05-04 08:16:07 GMT (1241424967 Unix time).

F10781:

Fill 10781 contains 14 runs (10143004 - 10143031) and spans times 23:59:03 (5/22) to 07:08:06 (5/23). In my previous analysis I found I had to remove run 10143031 (zdc) to obtain a consistent spin pattern. Run 10143031 starts at 06:01:14 and ends at 06:58:06. The output of the spinBitCrawler script shows that there is a change in polarization pattern between 06:39:26 and 06:44:29. The RICH scalars show that the change in polarization pattern happens during the fill. They also show that run 10143031 ends concurrently with the beam dump. Even though cdev indicates the polarization pattern changed during the fill, I could find no other evidence to suggest that anything changed the polarization, however there were vernier scans going on at PHENIX during this time. I believe it is reasonable to include run 10143031 in fill 10781 and extend the end fill time to 2009-05-23 10:58:06 GMT (1243076286 Unix time).

F10785:

Fill 10785 contains 25 runs (10144064 - 10144101) and spans times 14:31:52 (5/24) to 23:44:34 (5/24). In my previous analysis I found I had to remove run 10144101 (zdc) to obtain a consistent spin pattern. Run 10144101 starts at 23:28:56 and ends at 23:34:34. The output of the spinBitCrawler script shows that there is a change in polarization pattern between 23:35:48 and 23:40:51. The RICH scalars show that the change in polarization pattern occurs right after the beam is dumped. They also show that run 10144101 is completely within fill 10785. With this evidence it is clear that run 10144101 can be included in fill 10785 and that the end fill time can be extended to 2009-05-25 03:34:34 GMT (1243222474 Unix time).

F10825:

Fill 10825 contains 16 runs (10151029 - 10151052) and spans times 14:37:03 (5/31) to 22:02:57 (5/31). In my previous analysis I found I had to remove 2 runs to obtain a consistent spin pattern. These two runs were 10151050 (production) and 10151052 (fms). Run 10151050 starts at 21:09:07 and ends at 21:35:49. The output of the spinBitCrawler script shows that there is a change in polarization pattern between 21:37:45 and 21:42:48. The RICH scalars show that the change in polarization pattern occurs during the fill. They also show that run 10151050 is completely within fill 10825. Even though the change in polarization pattern occures in the fill, run 10151050 finishes just before this change occurs. Therefore it should be safe to include run 10151050 in fill 10825 and extend the end fill time to the last consistent cdev entry, 2009-06-01 01:37:45 GMT (1243820265 Unix time).

F10962:

Fill 10962 contains 11 runs (10172012 - 10172028) and spans times 03:37:52 (6/21) to 09:25:55 (6/21). In my previous analysis I found I had to remove 5 runs to obtain a consistent spin pattern. These runs were 10172024 - 028 (all production). Run 10172024 starts at 06:45:33 and ends at 07:03:37. Run 10172028 starts at 08:36:41 and ends at 09:15:55. The output of the spinBitCrawler script shows that there is a change in polarization pattern between 06:51:47 and 06:56:50. The RICH scalars show that the change in polarization pattern occurs during the fill (actually during run 10172024). They also show that run 10172028 ends before the fill is over. I couldn't find any other evidence to suggest that the polarization pattern actually changed during the fill. I believe it is reasonable to include runs 10172024 - 028 in fill 10962 and extend the end fill time to 2009-06-21 13:15:55 GMT (1245590155 Unix time).