Difference between revisions of "December 12, 2016 - Mott Group Meeting"
Jump to navigation
Jump to search
(Created page with " === Outstanding rates issues === *Fits of Run I HI fail our method for determining MEAN and SIGMA (exponential fit fails). We took LO test runs for all of the Run I HI. W...") |
|||
| (2 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
| − | + | === Outstanding rates issues === | |
| − | *Fits of Run I HI fail our method for determining MEAN and SIGMA (exponential fit fails). | + | *Fits of Run I HI fail our method for determining MEAN and SIGMA (exponential fit fails). |
| + | ** Proposed solution - Use LO test runs corresponding to HI thickness runs to determine MEAN and SIGMA. Apply MEAN and SIGMA to Run I HI runs (flag in code?). Success will be that the Run I LO and HI data sets are no longer separated. | ||
*Once Run I LO/HI discrepancy solved we will have Run I rates and Run II rates from analysis where the uncertainty calculation is defined [https://wiki.jlab.org/ciswiki/images/e/ef/Rates.pdf here]. | *Once Run I LO/HI discrepancy solved we will have Run I rates and Run II rates from analysis where the uncertainty calculation is defined [https://wiki.jlab.org/ciswiki/images/e/ef/Rates.pdf here]. | ||
| + | ** Is the absolute uncertainty removed from calculation shown in file??? | ||
* To this we have additional systematic uncertainty | * To this we have additional systematic uncertainty | ||
| Line 10: | Line 12: | ||
* What else? | * What else? | ||
| + | |||
| + | === Background subtraction === | ||
| + | |||
| + | * There are two ways to determine the background event | ||
| + | ** Method I - Use one detector, learned that applying background works for one detector, but not other - Riad points out asymmetry flips sign for one detector | ||
| + | ** Method II - Use two detectors, super-ratio. I propose asymmetry should be similar (ideally equal), so how does this solve problem. Doesn't. Still need to apply +/- asymmetry in solution, or will background flip sign (+/- solution) ? | ||
Latest revision as of 12:43, 12 December 2016
Outstanding rates issues
- Fits of Run I HI fail our method for determining MEAN and SIGMA (exponential fit fails).
- Proposed solution - Use LO test runs corresponding to HI thickness runs to determine MEAN and SIGMA. Apply MEAN and SIGMA to Run I HI runs (flag in code?). Success will be that the Run I LO and HI data sets are no longer separated.
- Once Run I LO/HI discrepancy solved we will have Run I rates and Run II rates from analysis where the uncertainty calculation is defined here.
- Is the absolute uncertainty removed from calculation shown in file???
- To this we have additional systematic uncertainty
- Fluctuation of rates within a run (this will be the systematic over period of days e.g. stability of injector setup, slow drifts in the instrumentation), looks to be <1%
- Fluctuation of rates between runs (this will be the systematic over period of months e.g. new calibration, new beam setup), looks to be ~5%
- What else?
Background subtraction
- There are two ways to determine the background event
- Method I - Use one detector, learned that applying background works for one detector, but not other - Riad points out asymmetry flips sign for one detector
- Method II - Use two detectors, super-ratio. I propose asymmetry should be similar (ideally equal), so how does this solve problem. Doesn't. Still need to apply +/- asymmetry in solution, or will background flip sign (+/- solution) ?