This is just to give an update on where I am in the spin QA.

First, I've had to modify the codes since Run 15 used a 24 bunch pattern instead of the previous 8, as well as deal with miscillaneous other changes, but I think that's all sorted.

Since there was an issue with the spin patterns at the beginning of the run, I've decided to start with pp transverse data. The current plan is to move forward throught the run and then at the end go back to the origional longitudinal data to see if anything can be done.

To ensure that the code was working properly, I first started with a short period of a weeks worth of runs. To start with, I included all runs within a fill (except for those marked as junk by RTS) and set the triggers to JP0, JP1, JP2. Note that I recently found an error with the trigger rates (which is why they appear as 0 in the attached document), but I believe these values are just used in diagnosing bad runs.

After preforming an initial sweep of this period, the Fills that were marked as bad are listed below in badListTrialPeriodOriginal.txt. I wil briefly go through what happened with these fills.

First, as was done in previous spin QA, I have removed all fills with two or less runs. The only note I'll make here is that one fill was listed as F100000000 instead of 18753. This same issue is seen in the RunLog Browser Outside of this there were two fills originally listed as bad

F18749-Last run happened close to end of fill, looked into it closer and found that spin patterns changed after last run, but within the 10 minute window after the last run that we search; introduced and exception into the code to set fillEnd soon after the last fill ended

F18760-Fill patterns changed after first run in fill, looked into it and saw that a steveProd run was taken an hour before other runs; removed steveProd run from list and added it to exceptions list for future runs

With these changes, reruning the code produced no errors. And I decided to move on to the full pp Transverse data set.

I've included several files at the bottom of the page (ppTrans*) for reference. Of particular note, there is a list of files at the end with the *fb error code. This indicates there was a problem at the beginning of the run. Looking at the testSummary file, I noticed that most of these files started with a pedAsPhys_tcd_only run, usually an hour before the next run. I removed that run, and will add this run type to the exclude list for future analyses. After removing those runs and fills with 2 or less runs, we are left with the following.

F18781-Fill had much more runs than usual (82 vs ~20). Looking at the runs I noticed that all of them are eithier pedestal runs or cosmics (down period from power loss?); Removed entire run

F18787-Last run in fill (1607748) ended about a minute after rampdown started, causing spin patterns to change to 0.; Removed this run and manually adjusted fill endtime

F18793-The spin patern on the blue beam changed sometime between the start of run 16080006 and the end of the next run; Not sure why this happened with several minutes left in the fill, but removing the last two runs clears things up (should we handle these runs somehow)

**Update from 02/27/16 After talking to others found that change should be entirely an artifact of cdev. Manually changed this entry in final goodLists.txt to include final 2 runs.

F18856-Fill ended soon after the last run; manually set fill end time for this fill.

After making these changes, all bad runs have been taken care of, as can be seen in the final badList.

Open Questions:

1)One of the fills F18750 had an unusual spin patterns in the yellow beam (a combination of old patterns 1 1 2 instead of the declared 1 2 2). However, the pattern remains constant throughout the fill, so it's marked as a good run. What should be done with this?

2)Looking at some old spin QA, there's usually some analysis done on bunch crossing patterns. Should this be done before moving on to pp longitudinal or before?