[daip] AIPS troubleshooting, multiple ASDM configurations

[Wen-fai Fong]

Hi Eric, thanks for the quick and informative reply.

Yes, I had 9 data sets through this program taken within a couple of weeks
of each other and this was the only one with two configurations. It is
indeed odd. I found nothing in the log files to denote why this was the
case.

No, there is no real reason to have 7 1-min scans, I was just seeing if I
could preserve the original scan structure. It's not a big issue though and
I think the reduction steps are all working as normal. I just wanted to
make sure I hadn't overlooked a better routine to do the combine. Thanks
again!

Wen-fai

On Thu, Oct 22, 2015 at 4:37 PM, Eric Greisen <egreisen at nrao.edu> wrote:

> On 10/22/2015 03:10 PM, Wen-fai Fong wrote:
>
>> Dear Eric,
>>
>> I'm working with a VLA continuum data set in AIPS. BDFLIST shows that I
>> have 2 configurations (0 and 1) which are virtually identical in
>> frequency, number of spectral windows, etc. Configuration 0 refers to
>> the flux calibrator, while configuration 1 refers to the phase
>> calibrator and my source, which comes right after the observation of the
>> flux cal. Ideally I would be able to combine the two and preserve the
>> time stamps and all scans.
>>
>> I understand that I can't load multiple configurations in a single run
>> of BDF2AIPS, so I've run it twice to load each configuration separately.
>> Then I used DBCON to combine the two data sets. I noticed when running
>> listr that the scans go from ~1 min long to ~7 min long, so it seems to
>> want to preserve the number of scans.
>>
>> Is this the best way to go about this? And is there a way to preserve
>> the scan length of ~1 min?
>>
>> Thanks,
>>
>> Wen-fai Fong
>>
>
> If DBCON is willing to combine the 2 data sets then I find it odd that
> they were 2 configurations to begin with.  It should be fine to combine
> the 2 in that case.
>
> DBCON, unlike many AIPS tasks, goes to some work to ignore the input index
> tables.  It then makes the "natural" scans found in your data.  Is there
> some reason to have 7 consecutive 1 minute scans?  I can not think of a
> reason in AIPS calibration to require that.  You can force the CL table
> interval to be short (we have users that use 1 sec CL table intervals for
> special reasons!).  I would worry that scheduling such short scans would
> incur overhead during the observing and might well have a few bad samples
> in each scan at the start.
>
> It would be rather some work to get DBCON to manage NX tables to preserve
> the scan structure and it might not even be possible in some cases but it
> might work in others.  I will consider this for the future but it will not
> happen immediately.
>
> Eric Greisen
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listmgr.nrao.edu/pipermail/daip/attachments/20151022/0ba31a59/attachment.html>