2018-10-10 workshop¶

Tracker¶

Resolution/bias
- Requires field integration to improve bias; Chris H seems resigned to it.
Dead channels see below
- Dead channel mapper is probably busted -> Tom L?
Glue density measurement
- ACTION: To do - find summer student/etc?
Tracker efficiency improvement
- Tom is making plots of digit distributions in each plane which don't form a track.
- Chris H has been playing with efficiency;
  - TKU 98.8 % efficiency before fixes
  - TKD 96.6 % efficiency; compared with table 18 and 20 from Emittance Evolution note; tables are consistent with ~ 97 % efficiency
  - Low pt tracks are likely to be reconstructed poorly
- NPE cut - throwing out "too early", do it at space point? Looks like where we are is okay
- ACTION: Durga and Chris H to sit together and look at how to modify the recon to analyse the events that are not reconstructed
- ACTION: Durga to check the bad channel loading
  - Chris R showed plots of doublets/triplets which are used during track finding; MC vs data; it looked reasonably good i.e. dead channels map appears to work reasonably well
- ACTION: Chris H to look at noise
- ACTION: Chris R to look into the clusters thing - where are they going/coming from

Viktor has dug down to look through the TOF calibrations; but missing calibration runs from < middle of summer
- Calibration data is on the GRID somewhere
For more recent data; need lots of reconstructed data with new calibration constants; ACTION: Durga will turn it around

Paul J is trying to understand discrepancy in measured phase advance and larmor angle
- Use measured ensemble beta function and alpha function; deduce measured single particle phase advance and Larmor angle
- It's a bit weird to look at ensemble properties mixed with single particle properties
- Sources of uncertainty in particular MC vs data:
  - magnet misalignment, windows and apertures, x-y asymmetry of beam
Paul J is doing Hamiltonian in solenoid and emittance growth
- Lie algebra looks challenging; pointed Paul at COSY

Tanaz will do reverse emittance exchange; focus for the thesis
Tanaz has looked at the data vs MC truth; low pt hole is still an issue
Reverse emittance exchange uses MICE data directly with natural dispersion in the beam
- Makes longitudinal heating and enhanced transverse cooling
Emittance exchange makes longitudinal cooling and less/no transverse cooling
- Emittance exchange requires reweighting the beam; struggled to reweight from lower dimension to higher dimension
  Steps in analysis
Sample selection follows standard cuts except no chi2 cuts yet; TOF; momentum cuts us and ds (but momentum cut is wider than "standard" analysis); tracker fiducial cut
Use transverse position and momentum to calculate KDE algorithm to generate density and

Outstanding actions

Select for a dispersive beam
- Chris H looked at using Voronoi density estimators; needs to be adapted for use at global track stations
- Tanaz looked at using KDE density estimators; Tanaz is tidying things up a bit
- Need selection at the absorber; will be a standalone analysis tool
- ACTION: Craig to pick up some bits of code
Try to split the transverse 4d emittance into two 2d emittances; look at evolution of the uncoupled emittances and demonstrate wedge plane is more significant
- Need to demonstrate that we understand Larmor angle between absorber and TKD

Introduce MC truth versus reconstructed early; talk about how the simulation was done
Describe TKU and TKD; define reference surface
Plot ADCs of two SP TOF events
Compare TOF-seeded tracker tracks and tracker tracks
Say a bit more about how the efficiency plots were generated
Francois - working on systems paper; started working yesterday on implementing
the TOF reconstruction into MAUS; ETA end of next week.
Comments on paper:
Eqn 1/2 units are wrong
Eqn 7 "when A_i is \epsilon_n"
Statistical uncertainty -> check

Files (0)