Difference: VeloMoniWishList (5 vs. 6)

Revision 62008-04-16 - MarcoGersabeck

Line: 1 to 1
 
META TOPICPARENT name="VELOCommissioningSoftware"

VELO Online Monitoring Wish List

This page is meant to be a scratch area for collecting ideas what to monitor and how. So everybody should feel free to add
Line: 58 to 58
 
      • calculate residual (DeVeloSensor)
      • for R sensor: plot residual vs phi (calculating phi from track state)
      • for Phi sensor: plot residual vs r (calculating r from track state) and vs phi (phi of strip at minimum strip radius +/- stereo angle)
Deleted:
<
<
      • update frequency for residual scatter plots: order of 10 seconds
 
      • when histogram has about 10'000 entries at least 1 minute old) create a ProfileX() (root method)
      • this profile should be kept, update can be ensured to be >=1min
      • fit profile vs phi with par_0 * sin(phi) + par_1 * cos(phi) + par_2
Line: 70 to 69
 
Added:
>
>
    • vs beetle readout direction for R sensor x-talk
      • largest cross-talk depends on readout direction and is particularly visible for R sensors
      • Phi sensors are less strongly affected by cross-talk due to different readout pattern
      • for each R sensor plot: residual distribution for each readout direction in each sector, i.e. make 8 plots: 4 sectors, strips 0-127 and 128-511 per sector
      • any non-zero mean is a sign for cross-talk
      • cross-talk overview plot: plot the distribution of the 8 mean values per sensor for all sensors, update frequency: >=1min
    • vs pitch and angle for resolution
      • for each sensor, plot residual vs pitch and angle (projected angle for Phi sensors?) 3-dim histogram
      • for each bin in pitch+angle fit the spread of the residual distribution
      • the fit should only be done for bins with at least 10000 entries
      • it has to be noted that extrapolation errors contribute to the measured spread!
      • projections of resolution (spread) vs pitch/angle can be plotted for various bins in the other variable using colour-coded lines (see Tomasz's plots)
      • residual plots should not be updated too often to ensure necessary statistics, hence update frequency: (>)>=10min
 

Considered

Quantities derived from NZS Data
Line: 89 to 101
 
Higher level quantities
  • tracks (this is very generic, needs to be fleshed out, like e.g. below)
Deleted:
<
<
  • residuals (preferably for small angle tracks with many spacepoints)
    • vs beetle readout direction for R sensor x-talk
    • vs pitch and angle for resolution
    • (more or less all of this code already available in sensor alignment software)
 
  • Alignment monitoring through alignment constants
    • residuals insensitive to less well constrained DOFs
    • rerun alignment algorithms and monitor change in parameters
 
This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback