[daip] Re: BB132B Tapes

[Eric Greisen]

Data Analysts writes:
 > Andreas is having some curious problems loading his data for his VLBA project
 > BB132B which was a phase referencing experiment.  I'm not familar enough with
 > AIPS to answer his queries.  Can you send him some insight?
 > 
 > Jason
 > 
 > ----- Begin Included Message -----
 > 
 > >From abrunthaler at cfa.harvard.edu Thu Apr 19 20:11 MDT 2001
 > Sender: abruntha at rgalp1.harvard.edu
 > Date: Thu, 19 Apr 2001 22:10:57 -0400
 > From: Andreas Brunthaler <abrunthaler at cfa.harvard.edu>
 > X-Mailer: Mozilla 3.04Gold (X11; I; OSF1 V4.0 alpha)
 > MIME-Version: 1.0
 > To: Data Analysts <analysts at zia.aoc.NRAO.EDU>
 > Subject: BB132B Tapes
 > Content-Transfer-Encoding: 7bit
 > X-Lines: 16
 > Status: RO
 > 
 > Dear Analysts,
 > 
 > I recieved yesterday the tapes of the BB132B observation. If I load in
 > the Data into AIPS it is extremely slow. With tape 2 and tape 6 it takes
 > 10-20 minutes to read in a 1 minute scan. I also tried to read in the
 > data on a different computer, but it's also slow there.
 > Another problem is that the source numbers are mixed up on different
 > tapes, e.g source #2 on tape 1 is source #3 on tape 1 etc. This gives me
 > messages like:
 > 
 > FITLD2: Found J0027+5958       at   1/00:47:49 src #   2 was    1
 > 
 > Does this affect the processing of my data?
 > 

No - each file on each tape is likely to mix around the source
numbers, frequency IDs, antenna numbers and the like.  FITLD when
appending can repair and renumber.  AIPS accesses sources by name but
you should always start your processing with LISTR ('SCAN') and PRTAN
to tell you what is the result of all the renumbering, scan indexing
etc

As to the speed, that is much harder to suggest anything.  If the
1-minute data set is in the middle of a file, all the data ahead of it
is read, decoded, processed and finally rejected which takes time.
The uv data are in a table that follows all the other weather,
calibration, etc tables so that all those tables must be processed
first and some of them are large.  You are better off loading all the
data in one go.

Eric Greisen