[daip] Problems with FITLD

[Eric Greisen]

Francesca D'Arcangelo wrote:
> I have been almost exclusively using the "raw" files, just trying the
> "cal" one once for good measure.  (As a sidenote, I tried
> concatenating the raw files in FITLD, but got persistent "data out of
> order" errors from INDXR even after I used UVSRT...which is why I
> concatenate them in DBCON.)
> 
> The same exact raw files worked fine before my AIPS reinstall, which
> is what makes me think it's something wrong with my AIPS, not the
> files.
> 
> Thanks,
> Frannie
> 
> On Tue, Jun 17, 2008 at 12:31 PM, Eric Greisen <egreisen at nrao.edu> wrote:
>> Francesca D'Arcangelo wrote:
>>> Hello.  I am having issues with FITLD producing corrupted weather
>>> tables when I load in my VLBA data.  I know that this was a problem
>>> before, solved with the patch on 01/18/08.  This worked on my previous
>>> install of 31DEC08.  However, I recently had to wipe my OS and, as a
>>> result, reinstall AIPS.  Ever since, I have been receiving these
>>> weather table errors from APCAL:
>>>
>>> APCAL1: Task APCAL  (release of 31DEC08) begins
>>> APCAL1: REAWX: ERROR  6 SOMETHING WRONG WITH WX TABLE
>>> APCAL1: REAWX: WXINI FINDS     0 ROWS, WX TABLE CORRUPT?
>>> APCAL1: TXTWX: ERR  6 OPENING WX TABLE   1
>>> APCAL1: Purports to die of UNNATURAL causes
>>> APCAL1: agn          31DEC08 TST: Cpu=       0.0  Real=       0
>>>
>>> Now, I reinstalled AIPS another time for good measure, and I ran the
>>> MNJ several times over a number of days, with seemingly no errors.
>>> After more WX errors, I attempted to do the FITLD patch by hand from
>>> the Patches section of the AIPS site, however I received an error when
>>> I used the COMLNK line :
>>>
>>> AIPS environment not defined; do CDTST?
>>>
>>> I'm not sure how to proceed on that.
>>> I have tried to run the program sequence on both the calibrated VLBA
>>> files *and* on the individual raw files from the archive (combined
>>> with DBCON), so that shouldn't be the problem.  I tried reading out
>>> (TBOUT) and in (TBIN) the WX table.  I even went so far as to paste in
>>> a version of FITLD.EXE from a colleague who has had no weather table
>>> problems.  And yet I still get the WX table errors.  Is there anything
>>> else I can do?  Is there a way of editing the table to suit the
>>> "infile" line in APCAL, perhaps?
>>>
>> The data in the archive - when in the form that has had some aips processing
>> - is corrupted.  I had suggested that these data be corrected but so far
>> that has not been done.  You will need to download the raw data - not the
>> files that have had FITLD and other tasks already run on them.  They are
>> called "raw" I believe while the others are called "cal".  You do not need
>> to concatenate the raw files before FITLD - it will read as many
>> (sequentially named) files as you tell it to read.
>>
>> Eric Greisen
>>

Okay - I gave you the first standard answer.  I think that Amy has a bug 
in the code in APCAL that tests for the error that made bad WX tables in 
the archive for a while.  She is running the Summer Schoold for another 
hour or two but then may be able to let you know about this.

Can you tell me a project name that has the FITLD concatenation issue?
It should not happen needless to say.

Eric Greisen