[daip] CLCAL glitch

[Eric Greisen]

Craig Walker writes:

 > eolus > CLCAL1: SNSMO: Smoothing SN table
 > eolus > CLCAL1: FUNBSM: FUNCTION GAUS TIMES  0.000E+00  0.000E+00 ERROR
 > eolus > CLCAL1: FUNBSM: FUNCTION GAUS TIMES  0.000E+00  0.000E+00 ERROR
 > eolus > CLCAL1: FUNBSM: FUNCTION GAUS TIMES  0.000E+00  0.000E+00 ERROR

    You left out the message that it changed AMPL below to FULL.  It
did that because of the setting of DOBLANK.

    FUNBSM did the correct thing - copying the phase, delay and rate
arrays with no smoothing.  I have suppressed the message.

 > 
 > The output CL table looks ok.

     Probably not - it probably did not smoothing...


 > AIPS 1: DOBLANK       1                    Blanked value interpolation

     This means replace previously blanked values with smoothed
values, leaving all previously good values unchanged.  See the help
file.

 > AIPS 1: SMOTYPE    'AMPL'                  Data to smooth

When blanks are replaced, they must have all data types included not
just amplitude.  But smoothing with 0 support size (no smoothing) is
and should be legal.

Eric Greisen