Difference between revisions of "BMOD 17Feb2021"
From PREX Wiki
Jump to navigationJump to search (Created page with "BMOD_3Feb2021 << >> BMOD_24Feb2021 == Logistic information == BlueJeans calling instructions: Toll-Free Number (U.S.& Canada): 888-240-2560 International toll...") |
m (→Agenda) |
||
| (3 intermediate revisions by the same user not shown) | |||
| Line 12: | Line 12: | ||
* Done - BMOD Sensitivity calculations | * Done - BMOD Sensitivity calculations | ||
* Done - BPM Eigenvector analysis (Cameron learns some software) | * Done - BPM Eigenvector analysis (Cameron learns some software) | ||
| − | * Evaluating regression vs. dithering, various diagnostics | + | * [https://prex.jlab.org/analysis/crex/agg-respin1/ErrorFlag_slug//slug_list/slug223/grand_signed_100-223.pdf Slug plots of 5bpm regression and BMOD corrections] Evaluating regression vs. dithering, various diagnostics |
| − | * Selecting stable BPM combos for another pass with plain BMOD, then lagrange regression, etc. | + | * [http://ace.phys.virginia.edu/HAPPEX/4472 HAPLOG] - Selecting stable BPM combos for another pass with plain BMOD, then lagrange regression, etc. |
==Notes== | ==Notes== | ||
| + | * The combination definition (coefficient of which BPMs live in orthonormal evMon eigen vectors) plots are not so useful when multiplying by the slope value - recreate these plots with just the orthonormal composition values directly | ||
| + | * Also the 5BPM slopes plots need to be just Draw("ap"), no connecting lines needed | ||
| + | * Lagrange regression would be good to pick up next, for consistency with PREX analysis | ||
| + | * Evaluate the dithering residuals that already were made from initial BMOD analysis and put on the HAPLOG | ||
| + | * Make comparisons of different analysis techniques using the difference in quadrature mean_error's method (per minirun, and per longer time scales) to see how far off two analyses are from each other | ||
| + | * Do a second pass of eigen vector composition stabilizing/resorting to get the remaining bits of noise to go away | ||
| + | |||
| + | Plan for next week: | ||
| + | * Do a segment-wise eigenvector-averaged dithering combo analysis for a direct comparison with eigenvector regression | ||
| + | ** Segment is a stable BMOD sensitivity time period, determined from the simple BPM set initial BMOD data (HAPLOG incoming soon) | ||
| + | ** Check the similarity between evMon regression and BMOD slopes calculated (averaged over various time scales) | ||
| + | ** Also check the composition of the evMon combos vs. time | ||
| + | ** Also check on residual sensitivities calculated using these slopes | ||
== Attendence == | == Attendence == | ||
| + | Cameron Clarke, Kent Paschke, Paul Souder, Tao Ye, Weibin Zhang, Ye Tian | ||
Latest revision as of 02:28, 18 February 2021
BMOD_3Feb2021 << >> BMOD_24Feb2021
Logistic information
BlueJeans calling instructions: Toll-Free Number (U.S.& Canada): 888-240-2560 International toll number: 408-740-7256 Bluejeans CODE: 948942477 Bluejeans link: https://bluejeans.com/948942477
Agenda
- Done - BMOD run period segmentation and outliers
- Done - BMOD Sensitivity calculations
- Done - BPM Eigenvector analysis (Cameron learns some software)
- Slug plots of 5bpm regression and BMOD corrections Evaluating regression vs. dithering, various diagnostics
- HAPLOG - Selecting stable BPM combos for another pass with plain BMOD, then lagrange regression, etc.
Notes
- The combination definition (coefficient of which BPMs live in orthonormal evMon eigen vectors) plots are not so useful when multiplying by the slope value - recreate these plots with just the orthonormal composition values directly
- Also the 5BPM slopes plots need to be just Draw("ap"), no connecting lines needed
- Lagrange regression would be good to pick up next, for consistency with PREX analysis
- Evaluate the dithering residuals that already were made from initial BMOD analysis and put on the HAPLOG
- Make comparisons of different analysis techniques using the difference in quadrature mean_error's method (per minirun, and per longer time scales) to see how far off two analyses are from each other
- Do a second pass of eigen vector composition stabilizing/resorting to get the remaining bits of noise to go away
Plan for next week:
- Do a segment-wise eigenvector-averaged dithering combo analysis for a direct comparison with eigenvector regression
- Segment is a stable BMOD sensitivity time period, determined from the simple BPM set initial BMOD data (HAPLOG incoming soon)
- Check the similarity between evMon regression and BMOD slopes calculated (averaged over various time scales)
- Also check the composition of the evMon combos vs. time
- Also check on residual sensitivities calculated using these slopes
Attendence
Cameron Clarke, Kent Paschke, Paul Souder, Tao Ye, Weibin Zhang, Ye Tian
