[daip] daip Help-- Thanks

[Eric Greisen]

Ching-Li Chen wrote:
> Hi,
> 
> I was trying to combine two data sets with DBCON, and I received the following message  
> that "FQIDS HAVE NOT BEEN RESET AS THEY SHOULD DUE TO FQTOL OR DUE TO THERE BEING NO FQ 
> RANDOM PARAMETER.' However, the task "Appears to have ended successfully" and did 
> produce a combined data set. So I was wondering whether that will cause a problem as I 
> proceed to combine more data, etc. If it will, how should I proceed instead? 
> 
> The VLA datasets I'm trying to combine are AY0152 (CBand,AArray), AM0192(CBand,AArray), 
> AA0068(CBand,BArray), and AJ0162(CBandDArray). Both AY152 and AM192 each have two IFs 
> and four stokes while AA68 and AJ162 each has only one IF and one stoke. 
> 
> Therefore, I first used SPLIT to separate AY152 and AM192 into individual uv-data with 
> only one IF and one stoke then combined those individual datasets using DBCON. (I've 
> only done so with IF 1, not 2 yet. But while I was doing so, I received the error 
> message that the "stokes ranges do not overlap." Therefore,I used "puthead" to change 
> all the coordinate values of the stokes to -1, and that worked.) 
> 
> Subsequently, I attempted to combine the resulting DBCON data with both IF 1 data from 
> AY152 and AM192 of all stokes with the AA68 dataset (AA68 has three frequency IDs, but 
> I'm currently only working with FQID 2), which is when I ran into the error message. 
> 
> I was using the default inputs with the FQTOL set to -1, but I have since tried it with 
> it set to 1, and still received the same error message. I also checked my data, and each 
> of them has a FQ table. In addition, while the frequencies for AY152 and AM192 are the 
> same, the frequencies for AA68 and AJ162 are not. 
> 
> AY152 and AM192 
> FREQ         1   4.8851000E+09       1.00  5.0000000E+07    0.00 
> 
> AA68 
> FREQ         1   4.5851000E+09       1.00  3.4375000E+07    0.00 
> 
> AJ162 
> FREQ         1   4.8300630E+09       1.00  4.6386720E+06    0.00 
> 
> Furthermore, I was wondering whether the image made from the resulting DBCON data will 
> have errors if DBCON sent the following messages: 
> 
> HERCUL> DBCON1: Task DBCON  (release of 31DEC10) begins 
> HERCUL> DBCON1: RA does not match, will shift second 
> HERCUL> DBCON1: DEC does not match, will shift second 
> 
> HERCUL> DBCON1: RA and DEC match to within 1 microarcsecond each 
> HERCUL> DBCON1: --I'm not going to shift positions

FQTOL is in kHz so a value of +1 says the frequencies must match to 1E+3 
in Hz.  Yours differ by 3E+8.  The real issue is what do DBCON do.  It 
would be good to see the IMHEADER of the output file after the FREQID 
warning message.  FREQIDs are used to keep data from "different" 
frequencies apart - something you are trying to avoid.  If the output 
has FREQID as a random parameter (the list of parameters starting with 
UU-L, VV-L, WW-L, TIME1...) then DBCON will have written the 2 
frequencies as separate FREQIDs and IMAGR will not image them together. 
  I assume that that is not what you want.  If there is no FREQID 
parameter, then DBCON has done what you want and its "grumble" may be 
ignored.

The phase shifting will take place if the J2000 coordinates of the 
sources in the various files do not match exactly.  It is done well and 
should be okay.  DBCON tells you about phase shifting in the interests 
of "full disclosure".

Eric Greisen