[daip] SAD docrt -4 bug

[Eric Greisen]

Mark Reid wrote:
> Eric,
> 
>    We noticed two things in SAD/DOCRT=-4:
> 
> 1) the beam position angle seems to be 90 degrees from the correct value;
> 
> 2) occassionally the flux density uncertainty fills the entire field, so 
> that it "merges" with the previous field.  This is OK if we read the 
> file with a formatted read, but it fails for an unformatted read (which 
> we often now do).  Could you change units, eg, from milliJy to Jy for 
> these cases?  I know we would lose some significance, but these are 
> cases where the signal is weak and the uncertainty large enough that 
> this should not be a problem.  (The same would probably apply to JMFIT.)
> 
> Thanks,
> Mark
> 
> -------------------------------------------------------------
> Mark J. Reid                 Phone: 617-495-7470
> Harvard-Smithsonian CfA      Fax  : 617-495-7345
> 60 Garden Street             Email: reid at cfa.harvard.edu
> Cambridge, MA 02138, USA     Web  : www.cfa.harvard.edu/~reid
> -------------------------------------------------------------
> 
> ------------------------------------------------------------------------
> 
> Subject:
> SAD docrt -4
> From:
> Mayumi Sato <mayumi.sato at gmail.com>
> Date:
> Wed, 20 Jan 2010 10:07:10 -0500
> To:
> Mark Reid <reid at cfa.harvard.edu>
> 
> To:
> Mark Reid <reid at cfa.harvard.edu>
> 
> 
> Hi Mark,
> 
> I came across another problem with SAD docrt -4 (ORFIT format) output
> besides the beam PA +90 deg rotated.
> 
> Here's an example of output lines:
> 
> !Source       Ch   Fnu _Jy Err    E-W Offset (mas)     N-S Offset
> (mas)       RA          DEC      Major(mas)   Minor(mas) PA(deg)
> Beam         RMS  #
> W51CH042  01   1   1.144  0.103       29.752   0.017      22.776
> 0.043 192343.8857 143035.7195   1.16  0.10   0.36  0.03 167  2   1.14
> 0.34  71 102.51m 1
> W51CH042  01   1 961.691102.198m      26.398   0.023      25.238
> 0.042 192343.8857 143035.7195   0.95  0.10   0.47  0.05 166  6   1.14
> 0.34  71 102.51m 2
> W51CH042  01   1 856.215101.674m      28.154   0.024      27.574
> 0.059 192343.8857 143035.7195   1.19  0.14   0.40  0.05 167  4   1.14
> 0.34  71 102.51m 3
> W51CH042  01   1 833.947100.417m      25.583   0.025      23.097
> 0.066 192343.8857 143035.7195   1.32  0.16   0.41  0.05 167  3   1.14
> 0.34  71 102.51m 4
> W51CH042  02   1   8.047  0.194       -2.973   0.005      -4.531
> 0.009 192343.8891 143035.7195   0.94  0.02   0.44  0.01 160  1   1.14
> 0.34  71 193.60m 1
> 
> As you can see we need more digits for intensity and its error,
> otherwise the two numbers could get connected as one for milliJy/beam
> cases.
> 
> I don't know if I have this digit problem with JMFIT docrt -4.
> 
> Mayumi

I have found that the Clean beam position angle was that used internal 
for the pixel image - thus SAD was 90 degrees out and both failed to put 
any rotation of the image into the value.  I have corrected this to use 
the header Clean Beam directly.  I also put better tests for scaling the 
flux and tests on the error bar and on the sign of the peak.  The error 
bar and negative peaks require tests that are 10 times tighter beacuse 
of the need for a blank or negative sign, resp.

Tomorrow's MNJ should fix this (after 10:30 pm Mountain Standard Time 
tonight).

Eric Greisen