[daip] IMAGR multiple fields: what's the issue with beam scaling? Also: CALIB with multiply concatenated VLBA datasets.
William Junor
bjunor at lanl.gov
Mon Nov 26 09:50:48 EST 2007
Thanks, Eric.
Well, I'm reassured a little. My concern was that I am losing
information or not CLEANing correctly in Field 2 (where the new science
is). However, given the weakness of the emission here, it's hard to say
if bad things are happening.
On the issue of multiple VLBA datasets concatenated together, I notice
some inconsistencies in various AIPS routines. For instance, where I
have 9 datasets in one concatenated file, CLIP and CALIB only report
information on progress through the first dataset.
I'll persevere.
Bill
> SUSAN BILL WIDER writes:
> > Eric, Amy.
> >
> > I'm running IMAGR with a couple of fields. The first field
> looks about right with a maximum of 1.0 Jy for the beam but the second
> field beam has a maximum of 4.0 Jy. How can this be? Aren't the
> beams meant to have unity maximum? Or is there some additional
> weighting that has to happen? The sum of the gridding weights in both
> cases is the same so what's going on here? (PRTMSG capture appended.)
>
> When there are large numbers of data samples, it is possible for the
> accumulated round-off errors to cause bad beam values, esp in the
> corneres of the image where the correction for the FFT of the
> convolving function is vary large. If you look at these beams you
> will see large positive and negative (1st field has -1.5 Jy as min)
> values in the corners. IMAGR should be able to survive this.
>
> >
> > Another issue: I have concatenated multiple VLBA datasets
> together to improve my sensitivity. The configuration of the arrays
> in most of the datasets is the same but not always. Consequently, the
> antenna numbering scheme for each sub-array changes --- say, if one
> antenna was unanvailable. If I run CALIB on this multi-epoch
> database, CALIB reports that some antennas get flagged because they
> can't be tied across sub-arrays. This shouldn't happen, I think,
> because all we need is that the sub-set of antennas in each effective
> sub-array at each epoch should be linked. I don't remember this sort
> of behavior with VLA multi-epoch data so is there something different
> happening when a non-VLA array is involved?
>
> It depends on how you have concatenated these data. If DBCON has
> called them separate sub-arrays, then there should be no connection
> between the subarrays in any case. The flagging to which you refer is
> due to antennas being disconnected within a subarray - i.e. with no
> way to relate the phase of the antenna to the phase of the reference
> antenna.
>
> Eric Greisen
>
> >
> > Bill
> >
blackhol AIPS (31DEC07) 131 23-NOV-2007
> 05:55:12 Page 1
> > Pops Prior Date Time Task Messages for user
> 131
> >
> > 1 5 22-NOV-2007 21:54:18 IMAGR Task IMAGR (release of
> 31DEC07) begins
> > 1 3 22-NOV-2007 21:55:04 IMAGR UVGET: Using flag table
> version 17 to edit data
> > 1 2 22-NOV-2007 21:59:28 IMAGR Create 3C274 MC
> .IMAGR . 1 (UV) on disk 2 cno 1
> > 1 4 22-NOV-2007 21:59:28 IMAGR Beginning channel 1
> through 64 with 2 IFs
> > 1 3 22-NOV-2007 22:12:58 IMAGR IMACPY: Copied 5327801
> visibilities to be imaged
> > 1 2 22-NOV-2007 22:13:01 IMAGR Create 3C274 MC U
> .IBM001. 2 (MA) on disk 1 cno 126
> > 1 2 22-NOV-2007 22:13:01 IMAGR Create 3C274 MC U
> .IBM002. 2 (MA) on disk 1 cno 127
> > 1 2 22-NOV-2007 22:13:01 IMAGR Create 3C274 MC U
> .ICL001. 2 (MA) on disk 1 cno 128
> > 1 2 22-NOV-2007 22:13:01 IMAGR Create 3C274 MC U
> .ICL002. 2 (MA) on disk 1 cno 129
> > 1 2 22-NOV-2007 22:13:03 IMAGR QINIT: did a GET of
> 5120 Kwords, OFF -344181861
> > 1 2 22-NOV-2007 22:13:03 IMAGR GRDFLT: X and Y
> convolution type = SPHEROIDAL
> > 1 2 22-NOV-2007 22:13:03 IMAGR GRDFLT: X and Y parms =
> 3.0000 1.0000
> > 1 2 22-NOV-2007 22:13:03 IMAGR GRDFLT: convolution
> function sampled every 1/100 of a cell
> > 1 2 22-NOV-2007 22:13:03 IMAGR GRDMEM: Ave 128
> Channels; 3.245212E+08 to 3.403962E+08 Hz
> > 1 4 22-NOV-2007 22:43:30 IMAGR Field 1 Sum of
> gridding weights = 1.34944E+02
> > 1 4 22-NOV-2007 22:43:31 IMAGR Field 1 Beam min =
> -1.5 Jy, max = 1.0 Jy
> > 1 4 22-NOV-2007 22:43:31 IMAGR Field 1 fit FWHM=
> 70.414 x 41.497 Milliarcsec, PA= -25.9
> > 1 4 22-NOV-2007 23:03:02 IMAGR Field 2 Sum of
> gridding weights = 1.35039E+02
> > 1 4 22-NOV-2007 23:03:04 IMAGR Field 2 Beam min =
> -1.4 Jy, max = 4.0 Jy
> > 1 4 22-NOV-2007 23:03:04 IMAGR Field 2 fit FWHM=
> 70.435 x 41.507 Milliarcsec, PA= -25.9
> > 1 4 22-NOV-2007 23:03:05 IMAGR CLBHIS: minimum
> component 0.500 of current peak
> > 1 4 23-NOV-2007 00:17:16 IMAGR Field 1 min = -236.7
> MilliJy,max = 1.2 Jy
> > 1 4 23-NOV-2007 01:33:28 IMAGR Field 2 min = -127.4
> MilliJy,max = 37.6 MilliJy
> > 1 2 23-NOV-2007 01:33:29 IMAGR BGC Clean: using 1019
> cell beam + residuals > 343.39 MilliJy
> > 1 2 23-NOV-2007 01:33:31 IMAGR 226 Residual map
> points loaded
> > 1 4 23-NOV-2007 01:33:31 IMAGR Field 1 min
> algorithm flux= 884.381 MilliJy iter= 4
> > 1 3 23-NOV-2007 01:33:31 IMAGR Field 1 Clean flux
> density= 414.179 MilliJy 4 comps
> > 1 3 23-NOV-2007 01:33:31 IMAGR Total Cleaned flux
> density = 414.179 MilliJy 4 comps
> > 1 4 23-NOV-2007 01:33:31 IMAGR SETDO3: imaging done
> with multiple tangent planes
> > 1 2 23-NOV-2007 01:33:31 IMAGR VISDFT: Begin DFT
> component subtraction
> > 1 2 23-NOV-2007 01:33:31 IMAGR VISDFT: Model
> components of type Point
> > 1 3 23-NOV-2007 01:33:31 IMAGR I Polarization model
> processed
> > 1 2 23-NOV-2007 02:18:36 IMAGR BGC Clean: using 1019
> cell beam + residuals > 343.39 MilliJy
> > 1 2 23-NOV-2007 02:18:37 IMAGR 0 Residual map
> points loaded
> > 1 4 23-NOV-2007 03:33:07 IMAGR Field 1 min = -192.5
> MilliJy,max = 797.2 MilliJy
> > 1 4 23-NOV-2007 04:49:13 IMAGR Field 2 min = -79.8
> MilliJy,max = 28.9 MilliJy
> > 1 2 23-NOV-2007 04:49:16 IMAGR BGC Clean: using 1019
> cell beam + residuals > 228.67 MilliJy
> > 1 2 23-NOV-2007 04:49:17 IMAGR 392 Residual map
> points loaded
> > 1 4 23-NOV-2007 04:49:17 IMAGR Field 1 min
> algorithm flux= 459.853 MilliJy iter= 12
> > 1 3 23-NOV-2007 04:49:17 IMAGR Field 1 Clean flux
> density= 894.323 MilliJy 12 comps
> > 1 3 23-NOV-2007 04:49:17 IMAGR Total Cleaned flux
> density = 894.323 MilliJy 12 comps
> > 1 2 23-NOV-2007 04:49:17 IMAGR VISDFT: Begin DFT
> component subtraction
> > 1 3 23-NOV-2007 04:49:17 IMAGR I Polarization model
> processed
> > _______________________________________________
> > Daip mailing list
> > Daip at listmgr.cv.nrao.edu
> > http://listmgr.cv.nrao.edu/mailman/listinfo/daip
>
--
ISR-2, Space and Remote Sensing,
MS D436, Los Alamos National Laboratory,
P.O. Box 1663, Los Alamos, NM 87545
Phone: (505) 667 0788, FAX: (505) 664 0362
Cell: (505) 699 7692
More information about the Daip
mailing list