[daip] MSORT question

Amy Mioduszewski amiodusz at nrao.edu
Thu May 31 18:51:15 EDT 2007


Hi Jim,

You are right, the cookbook does say that FITLD will automatically load the 
passes separately.  This is wrong.  You have to do it manually.  I.e., load the 
one set of files then rename the output and load the second.

The reason MSORT hung was because both passes were in the same file.  It should 
work fine if they are in separate files.  In fact, MSORT might not even be needed.

Amy

Jim Braatz wrote:
> Hi Amy,
> 
>   The cookbook suggests that data from multiple correlator passes will 
> automatically be written into two uvdata files, so that's why I was 
> surprised I got only one uvdata file for this data set.  So, how do I 
> tell FITLD to make two uvdata files?  Is it a matter of knowing which 
> IFs were correlated in each pass, and then setting BIF and EIF 
> appropriately?  Or, is it simply that one tape contains data from 
> correlator pass 1 and the other tape has pass 2?  In either of these 
> cases, I know how to do it.
> 
>   If there's some documentation on how to handle data from multiple 
> passes, please let me know.  (I'm happy to read documentation and stop 
> the pestering!)
> 
>   Once I've got the data in two uvdata files, will the MSORT happen 
> quicker?
> 
>     Thanks again,
>        Jim
> 
> Amy Mioduszewski wrote:
>> Hi Jim,
>>
>> You need to load them into two separate files.  Then run VLBAFIX on 
>> each file, and then run VBGLU to "glue" them together.
>>
>> Amy
>>
>> P.S.  For your other issue, VLBALOAD and VLBAFIX both do table merging 
>> so you don't need to run MERGCAL if you have run either of these two.
>>
>> Jim Braatz wrote:
>>> Hi Amy,
>>>
>>>   The printout enclosed with the tapes indicates that two correlator 
>>> passes were necessary to handle the 512 data rate.  I was somewhat 
>>> expecting two uvdata files, but only got one.
>>>
>>>    Jim
>>>
>>> Amy Mioduszewski wrote:
>>>> Hi Jim,
>>>>
>>>> Kill MSORT.  Do you have multiple passes?  If yes you need to deal 
>>>> with the data differently than VLBAFIX.  I will send more after lunch.
>>>>
>>>> Cheers,
>>>>
>>>> Amy
>>>
> 




More information about the Daip mailing list