APOLLO data reduction scheme
- Establish log file to record progress and parameters relating to
- Check for instances of CPS (counts per second) errors and log
exceptions (also terminal warning)
- Check for consistency between TWS and FRC values for every event to
make sure no clock ticks are missed. Log exceptions and warn on terminal.
- Perform polynomial prediction check against TDC prediction value for
each lunar gate. Log exceptions and warn on terminal.
- Check for missing FID fast-photodiode records & log exceptions,
summarize as fractional loss.
- Report laser dropouts (lack of FID gate at expected time); summarize
clusters of dropouts by when they happen and effective laser frequency
(clusters on >10 sec timescales)
- Sort fiducial records into alternating classes, consistent with
diffuser vs. attenuator. Use motorrps parameter to accurately predict
expected fiducial times. May need to build in self-adjustment/correction.
- Fiducial arrays to be 2-d: channel and shot.
- Plot fiducial fast-photodiode raw TDC distribution.
- Report maximum recorded TDC values for both FID gates and LUN gates, as
a way to track effect of gate width on TDC effective coverage.
- Plot distribution of non-corner-cube FID TDC records.
- Plot distribution of non-lunar LUN TDC records.
- Fit quadratic function to latest TDC calibration data.
- Apply TDC calibration to all calculations of all time residuals.
- For each APD channel and FID type, find clipped mean, mode, and
Gaussian fit parameters for the fiducial records. Record parameter arrays
in log file (worth subsequent study).
- Explore behavior of above characterizations as a function of bin size.
- Plot the fiducial distributions for each channel, along with the
Gaussian fit and mean, mode.
- Report the bias for each channel: the difference between the diffuser
fiducial and the attenuated fiducial measurements.
- Process lunar measurements as before, but with the full machinery of
fresh fiducials and TDC calibration.
- Perform linear fit to lunar range residuals, as before, using minimum
window, taking care to place window symmetrically for noise immunity.
- Explore fit of lunar profile using methods similar to those used for
the fiducials. Perhaps even a correlation method that is insensitive to
functional shape (e.g., asymmetry, diffusion tails, double-pulsing).
- Plot histograms of photon number per shot for both fiducials and
- Provide mechanism for dealing with strong-signal bias, and explore
magnitude of bias as a function of high-signal cutoff.