[daip] spectral windows

[Eric Greisen]

On 01/28/2017 09:58 AM, Farhad Zadeh wrote:
> Hi Eric,
> Sorry for a simple question. I have a casa reduced data  set that is loaded
> into aips. However, the spectral windows are moved around. I have 8GHz band
> with 4x16 IFs at Q band.  So, the frequency of each 16 IFs are not in
> order of increasing frequencies.
> Also, when I selfcalibrate in phase and make a DFTPL, the frequency in
> the plot shows 34 GHz!
> This is a Q band. In any case, it may be that the frequency axis is
> confused. Listr shows that the frequencies are
> all correct but not in order. I forget the aips  command reoders the
> spectral windows.  I appreciate it.

We have tasks that reverse the order of things (FLOPM) but none that do 
a more general sort.  What you will need to do is run UVCOP with BIF-EIF 
selecting IFs in the desired order.  Then VBGLU to put them back 
together.  VBGLU can take up to 4 input data sets so you may need to run 
it more than once depending on the details.

With the EVLA it is possible to have IFs 1-8 (AIPS nomenclature) overlap 
slightly with 9-16 (IFs 8 and 9 overlap).  Usually in that case you want 
the two sets to be in that order so that FRING can work on halves since 
that is the actual hardware IFs.  I don't know what CASA has given or 
why.  I suppose I could write an AIPS task to do this, but the table 
part would be messy for sure and I hope this is not a general need.

Eric Greisen