[daip] [!12819]: AIPS - FITLD DOWEIGHT seems to affect visibility amplitudes as well as weights

[Michael Bietenholz]

              Status: Open (was: Closed)

FITLD DOWEIGHT seems to affect visibility amplitudes as well as weights
-----------------------------------------------------------------------

           Ticket ID: 12819
                 URL: https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/12819
                Name: Michael Bietenholz
       Email address: mbieten at yorku.ca
             Creator: User
          Department: AIPS Data Reduction
       Staff (Owner): Eric Greisen
                Type: Issue
              Status: Open
            Priority: Default
                 SLA: NRAO E2E
      Template group: Default
             Created: 18 October 2018 11:27 AM
             Updated: 19 October 2018 05:53 AM
           Reply due: 23 October 2018 05:53 AM (4d 0h 0m)
      Resolution due: 15 July 2021 12:00 AM (999d 18h 7m)

Hi Eric:

   Thanks as ever for the quick fix.  It will be good
to have weights that are something like 1/sigma**2.

Just to make sure I understand: is it true that the DOWEIGHT=1 will
change the weights only by a global (to a particular data-set)
scale-factor (int_time*bw), so that the relative weights (ie any ratio
w_i/w_j) is independent of DOWEIGHT (provided bw or int_time doesn't
change within a data-set, which might happend with more than
one FREQID)?

                  thanks,   michael


On Thu, 18 Oct 2018, Eric Greisen wrote:

> Yes.  I attempted to make the VLBI weights more meaningful
> but did not understand all the ramifications.  Therefore the
> visibilities got divided by the very much larger new weights.
> I have now fixed it to set the weights only after scaling the
> visibilities.  There is still one mystery that I do not understand
> but I have sent an e-mail to Walter Brisken for clarification.  I
> have put the correction back so tomorrow's MNJ should
> provide improved performance (or DOWEIGHT=-1 should
> now give the old performance).
> 
> Eric Greisen
> 
> 
> Ticket Details
> 
> 

------------------------------------------------------
Staff CP:  https://help.nrao.edu/staff
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listmgr.nrao.edu/pipermail/daip/attachments/20181019/b9e153cb/attachment.html>