[daip] DFTPL

[Eric Greisen]

On 02/13/2017 02:38 PM, Farhad Zadeh wrote:
> Hi Eric,
>
> I have a big data set with 64 IFs and 64 channels over 8 GHz at Ka band.
> I have done DFTPL for individual
>
> IFs and it works fine as it should. The mean flux changes because of
> frequency dependence of the Sgr A*.  However, when I use all
>
> IFs, I dont see a range of flux expected between  the low and high
> frequencies. It does averaging  but not all light curves are included.
>
> I think this used to work. Am I doing something wrong?
>
> Thanks,
>
> Farhad
>
> PS I which we could have waterfall presentation of light curves for
> different frequencies in DFTPL.


I found a serious error in DFTPL - it effectively ignores all IFs > BIF
or at least 3/4 of the data since it strides through the data 4 times 
faster than it should.  This has now been fixed in 31DEC16 and 31DEC17.

I also write a new task DFTIM in 31DEC17 only which makes the waterfall 
image you requested (and an optional error image).  The X axis is 
spectral channel (and you can average channels within IFs) and the Y 
axis is time with the same time averaging as DFTPL.

I cleaned up some things in both - you cannot do cross-hands RL and LR 
since they are not Hermitian which DFTPL and DFTIM assume.  So I added 
STOKES and ignore BPARM(1).  This lets you do I, Q, U, V with RR/LL data 
sets which I think was forbidden before.

Enjoy

Eric