Branch: refs/heads/master
Home: https://github.com/Starlink/starlink
Commit: 461a305d236124c4a6e14df84bdc3fa963291143
https://github.com/Starlink/starlink/commit/461a305d236124c4a6e14df84bdc3fa963291143
Author: David Berry <[log in to unmask]>
Date: 2013-08-16 (Fri, 16 Aug 2013)
Changed paths:
M applications/smurf/libsmf/smf_flat_fitpoly.c
Log Message:
-----------
smurf: relax flatfield test for incorrect gradient of poly fit
Previously, a bolometer was flagged as bad if the fast-flat observed
current versus heater power had a temporary downwards wiggle at the
central heater power value. This is because the central value was used a
a reference value. This can happen a lot with noisey fast-flat data such
as is generated by POL-2 when the calibrator is in the beam, resulting in
all (or most) bolometers being rejected.
Now, if the central current value is unusable due to a wiggle, a nearby
good value is used as the reference value instead. This allows more
bolometers to get through to the SNR test, so setting a low value
for parameter FLATSNR will then allow pol-2 data to be flat-fielded.
|