This HOW-TO shows how one can do a simple database trend analysis from the awe-prompt. In general to get to a result you need to do the following steps:
Question 1: Make a plot of the bias level of all raw biases of a CCD as a
function of modified julian date of observation.
awe> q = (RawBiasFrame.chip.name == 'ccd50') awe> biases = list(q) awe> x = [b.MJD_OBS for b in biases] awe> y = [b.imstat.median for b in biases] awe> pylab.scatter(x,y,s=0.5)
Question 2: Look for raw biases for ccd50 (WFI) in 2004 for which the level of the trim section differs significantly from the level of the overscan.
awe> q = (RawBiasFrame.filename.like('WFI.2004*_1.fits')) awe> len(q) 419 awe> biases = list(q) awe> x = [b.MJD_OBS for b in biases] awe> y = [b.imstat.median-b.overscan_x_stat.median for b in biases] awe> pylab.scatter(x,y,s=0.5)
awe> frames = [b for b in biases if b.imstat.median-b.overscan_x_stat.median > 4] awe> len(frames) 2 awe> for f in frames: print f.filename ... WFI.2004-10-15T15:10:02.248_1.fits WFI.2004-10-15T15:11:52.384_1.fits awe> for f in frames: f.retrieve() ...
It turns out there are in fact two frames of this kind. The images seem to have
an uncharacteristic bright region in them; something was obviously wrong during