[evlatests] Efficacy of system flagging

[Rick Perley]

    I noted yesterday three distinct phenomena in my 6cm data:

    1) Slow recovery to full amplitude upon change of band
    2) Scans with low -- but not zero -- amplitudes
    3) Scans with zero amplitudes. 

    The system flagging catches two of these three phenomena:

    1) All 'slow recoveries' are properly flagged.  Surely this is a 
subreflector slowness issue.
    2) All but one of these 'low amplitude' events are properly 
flagged.  The single exception was on antenna 27.   I offer a guess that 
these are subreflector out of position issues...
    3) None of the 'zero amplitude' scans was flagged.   A bandswitch or 
LO issue? 

    It's also worth noting here that there are numerous instances where 
data appear perfectly normal, but have been flagged by the system.  
Antennas 7 and 27 seem particularly susceptible to this. 

    Overall, I'd say that more 'good' data have been flagged as bad than 
there are bad data flagged as good (i.e. not flagged at all).