From PREX Wiki
Meeting room at JLab: No meeting room at JLab
BlueJeans calling instructions: Toll-Free Number (U.S.& Canada): 888-240-2560 International toll number: 408-740-7256 Bluejeans CODE: 385 600 867 Bluejeans link: https://bluejeans.com/385600867
Updates from last week
- Weibin was going to work on surveying the respin2 outputs for bcm_an_us glitches we hadn't recognized yet.
- Paul was going to look at what we might do to cut BCM glitch events for CREX.
- Robert was going look at the four runs discussed in HAPLog 4132 (runs 3853, 4013, 4015, and 4174) to see what impact the beam fluctuations have on the corrected asymmetries. This would be to decide if we needed to add event cuts to those runs and reanalyze.
- Victoria was preparing a set of dithering corrections for the CREX transverse period
PREX respin2 issues
- The six runs with two data files need to be reanalyzed with the "--chainfiles" argument to the qwparity calls so that "short miniruns" are properly recognized. Tao set up the special submission script to do this and Sakib was preparing to submit the jobs.
- The tape archiving of outputs was halted when the above problem was noticed. Other than this, which directories are ready or not ready?
- japanOutput/ --- ready
- LogFiles/ --- ready
- SummaryText/ --- ready
- LRBoutput/ --- ready
- respin2 webplots --- wait for any additional aggregation plots for a bit
- ditherOutput/ --- ready
- LagrangeOutput/ --- I will check with Tao about this and the next two directories to see if they are ready to be archived, or not yet.
CREX respin1 planning
- Transverse respin
- Once the dithering corrections have been committed, we will start setting up for a CREX transverse-only respin
- Progress on dithering correction terms
- Work is continuing to evaluate the dithereing for the first part of CREX and this may impact the choices for the rest of CREX as well
- Looking for split-file runs
- Only one run in the CREX prompt slug lists has a split-file: run 6647 has two data files
- There are 57 runs in the CREX run range with two or more file-splits, we'll need to check if some of these are calibration runs that we want to replay. Two runs have more than 16 splits; 5501 has 29 and 6414 has 39. For runs with lots of splits, just using "chainfiles" will end up causing problems too. If we need to analyze runs like that we'll need to think about how to do ti.
- I discovered a failure mode if a correlator/corrector asks for a data_handler variable which doesn't exist, it really gets messed up. Should add better checks and a way that it can progress smoothly with a missing variable. I have not yet added this as a github issue, but will shortly.
- Possibly work to streamline the "minirun conbination" process instead of running pass0 and pass1 to get the LRB slopes. Cameron had done some start on this work, but it isn't done.