[daip] CLCAL's CUTOFF: suggested tweak

[Michael Rupen]

Dear DAIP,
  the new CUTOFF parameter in CLCAL is a lovely idea, but doesn't quite work
as expected.  I believe this is a bug rather than a mis-interpretation on my
part but I may be wrong...anyhow here goes.

  I have a data set with a number of scans:

    I1746-32B      07:54:20-07:57:00
    1751-253       07:57:23-07:58:20
    I1746-32B      07:58:43-08:01:40

    SS433          09:16:40-09:19:20
    1929+0507      09:19:40-09:20:40
    SS433          09:21:00-09:23:50

    SS433          10:04:40-10:07:20
    1929+0507      10:07:40-10:08:40
    SS433          10:09:00-10:11:40

    1950+081       10:12:13-10:13:10
    J1912+0518     10:13:43-10:14:50
    1944+548       10:16:33-10:17:30
    2202+422       10:18:50-10:20:00
    0137+331       10:22:57-10:24:10
    0319+415       10:34:20-10:35:53

I wish to calibrate I1746-32B wrt. 1751-253, and SS433 wrt. 1929+0507.
Since I've observed SS433 in two blocks separated by some 40mins., I want to
use CUTOFF=10 to restrict the interpolation to +/-10mins. from the nearest
calibrator.

Using CUTOFF=10 works perfectly for I1746-32B, and for the 1st and last
scans on SS433: the gains are correctly copied from the appropriate
calibrator scan.  The 2nd & 3rd SS433 scans (09:21-09:24, 10:04-10:08) are
NOT calibrated at all, presumably because they're each more than 10 minutes
from _one of_ the adjacent calibrator scans. 

While this is understandable, it's not really what is wanted.  CLCAL should
I believe simply discard the more distant calibrator scan and copy the
gains from the nearer one, only doing a real interpolation when there are
*two* surrounding "good" scans (i.e., both less than CUTOFF minutes away).
In other words, CUTOFF should be thought of as a window in time, with stuff
outside that window ignored entirely, rather than being used (effectively)
for flagging.  

I append the inputs to CLCAL (for SS433/1929) below; I can give you all
sorts of outputs if you want them, but for this particular purpose that
doesn't seem very useful.  CUTOFF and DOBTWEEN are lovely extensions to
CLCAL; I think this tweak will make them even more useful.

  Cheers,

             Michael

p.s. I'd be happy to write this data set to disk if that's useful.

On verdi (Linux box), version 31DEC03:
> inp CLCAL
AIPS 1: CLCAL     Task to manage SN and CL calibration tables
AIPS 1: Adverbs     Values                 Comments
AIPS 1: ----------------------------------------------------------------
AIPS 1: INNAME     '20030716    '          Input UV file name (name)
AIPS 1: INCLASS    'L BAND'                Input UV file name (class)
AIPS 1: INSEQ         2                    Input UV file name (seq. #)
AIPS 1: INDISK        1                    Input UV file disk unit #
AIPS 1: SOURCES    'SS433           '      Source list to calibrate
AIPS 1:            'J1912+0518      '  *rest ' '
AIPS 1: SOUCODE    '    '                  Source "Cal codes"
AIPS 1: CALSOUR    '1929+0507       '      Cal sources for calibration
AIPS 1:            *rest ' '
AIPS 1: QUAL         -1                    Source qualifier -1=>all
AIPS 1: CALCODE    '    '                  Calibrator code '    '=>all
AIPS 1: TIMERANG   *all 0                  Time range to calibrate
AIPS 1: SUBARRAY      0                    Subarray, 0=>all,
AIPS 1: ANTENNAS   *all 0                  Antennas selected, 0=> all
AIPS 1: SELBAND      -1                    Bandwidth to select (kHz)
AIPS 1: SELFREQ      -1                    Frequency to select (MHz)
AIPS 1: FREQID        1                    Freq. ID to select.
AIPS 1: OPCODE     '    '                  Operation 'MERG','CALI',
AIPS 1:                                    'CALP'; ' ' => 'CALI'
AIPS 1: INTERPOL   'SIMP'                  Interpolation function,
AIPS 1:                                    choices are: '2PT','SIMP',
AIPS 1:                                    'AMBG','CUBE','SELF','POLY'
AIPS 1:                                    see HELP for more details.
AIPS 1: CUTOFF       10                    Interpolation limit in
AIPS 1:                                    time (min); 0=> no limit.
AIPS 1: SAMPTYPE   '    '                  Smoothing function
AIPS 1: BPARM      *all 0                  Smoothing parameters
AIPS 1: ICUT          0.1                  Cutoff for functional forms
AIPS 1: DOBLANK       1                    Blanked value interpolation
AIPS 1: DOBTWEEN     -1                    > 0 -> smooth all sources
AIPS 1:                                    together; else separate them
AIPS 1: SMOTYPE    '    '                  Data to smooth
AIPS 1: SNVER         0                    Input SN table, 0=>all.
AIPS 1: INVERS        0                    Upper SN table vers in a
AIPS 1:                                    range.  0=>SNVER
AIPS 1: GAINVER       1                    Input Cal table 0=>high
AIPS 1: GAINUSE       2                    Output CAL table 0=>high+1
AIPS 1: REFANT       20                    Reference antenna 0=>pick.
AIPS 1: BADDISK    *all 0                  Disks to avoid for scratch

================================================================================