AIPS HELP file for UJOIN in 31DEC24
As of Wed Nov 6 6:44:04 2024
UJOIN: Converts IFs to additional spectral channels
INPUTS
INNAME Input UV file name (name)
INCLASS Input UV file name (class)
INSEQ 0.0 9999.0 Input UV file name (seq. #)
INDISK 0.0 9.0 Input UV file disk unit #
OUTNAME Output UV file name (name)
OUTCLASS Output UV file name (class)
OUTSEQ -1.0 9999.0 Output UV file name (seq. #)
OUTDISK 0.0 9.0 Output UV file disk unit #.
CHANSEL 0.0 9999.0 Begin, end input channels,
begin output channel each IF
OPCODE 'DIFF' to output the vis
difference in the overlap
DOWEIGHT -1.0 1.0 0 => delete spectrum if any
IF or channel is flagged
< -0.5 => keep data even if
one of the IFs is flagged
> 0.5 => delete channel if
one of the IFs is flagged
HELP SECTION
UJOIN
Task: This task is intended to convert two IFs into one longer
spectrum, averaging the overlapped channels. It can also write a
difference data set to let you test how nearly equal the two
overlapped regions actually are.
NOTE: This task does not apply flag tables or calibration. All
such must be applied before this task is run - use SPLIT,
FUDGE, or some other appropriate task.
Adverbs:
INNAME.....Input UV file name (name). Standard defaults.
INCLASS....Input UV file name (class). Standard defaults.
INSEQ......Input UV file name (seq. #). 0 => highest.
INDISK.....Disk drive # of input UV file. 0 => any.
OUTNAME....Output UV file name (name). Standard defaults.
OUTCLASS...Output UV file name (class). Standard defaults.
OUTSEQ.....Output UV file name (seq. #). 0 => highest unique
OUTDISK....Disk drive # of output UV file. 0 => highest with
space for the file.
CHANSEL....3 entries each for up to 10 IFs:
CHANSEL(1)= 1st input channel to use from IF 1
CHANSEL(2)= last input channel to use from IF 1
CHANSEL(3)= output channel corresponding to 1st input
channel used from IF 1
CHANSEL(4)= 1st input channel to use from IF 2 etc.
OPCODE.....'DIFF' odd # IFs added, even numbered IFs subtracted;
else all IFs added with weights and then divided by sum of
weights in each output channel.
DOWEIGHT...= 0 => delete any spectrum in which one or more
channels/IFs are flagged. This means that weights will
depend on channel only because of overlap.
< -0.5 => weight the data by the sum of the incoming
weights (weight = +1 or -1 for DIFF), so that an
overlapped channel having one of the IFs flagged will be
indicated only by a lower weight. (This is not a good
idea on OPCODE 'DIFF')
> 0.5 => delete any channel in which one or more IFs is
flagged.
EXPLAIN SECTION
UJOIN: Task to average/concatenate IFs to form a single-IF
output spectrum
Documentor: Michael Rupen
Related tasks: AVSPC, MCUBE, SPECR, UVGLU, VBGLU, WTSUM
One often observes with IFs tuned to slightly different but
overlapping frequencies, to obtain a broader bandwidth or higher
spectral resolution than possible with a single IF. UJOIN takes such
data and averages or concatenates the data from the IFs to create a
single-IF output data set with a single continuous spectrum.
CHANSEL:
You must specify explicitly the correspondence between input and
output channels for each IF, using CHANSEL. Consider a 2-IF
experiment with 5 channels per IF, set up as follows:
IF 1 Channels 1 2 3 4 5
IF 2 Channels 1 2 3 4 5
We want a 7-channel output spectrum like so:
IF 1 Channels 1 2 3 4 5 6 7
For this one would set
CHANSEL = 1 5 1 1 5 3
UJOIN would produce a single-IF data set with channels 1-2 taken
directly from IF 1, channels 3-5 formed from the (weighted) average
of the corresponding channels in each IF, and channels 6-7 copied
directly from IF 2.
As a more realistic example, consider a 2-IF experiment with 15
channels per IF, where the first 3 and last 2 channels are
significantly degraded by a roll-off in the bandpass. Suppose the
IFs overlap by 7 channels. In this case we have
IF 1 Channels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
IF 2 Channels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
and we want to toss channels 1-3 of IF 1 and 14-15 of IF 2. In this
case we would set
CHANSEL = 4 15 12 1 13 1
OPCODE
By default, UJOIN takes the weighted sum of the input IFs on a
channel-by-channel basis. With OPCODE='DIFF' the odd IFs are
summed and the even IFs are subtracted; if there are an even number
of IFs the resulting spectrum should be zero (apart from noise) on
each baseline in the overlap region. This provides a useful check
on the consistency of the input IFs. The actual data weights
(other than flagging) are ignored in DIFF mode.
DOWEIGHT
A slight complication results if the uv-coverage of the IFs is
different, as it usually is: what should be done if a given
baseline has no data for one channel or one IF? UJOIN offers three
options, specified by DOWEIGHT:
DOWEIGHT = 0 flag a baseline/Stokes in EVERY channel, if that
baseline/Stokes is flagged in *any* of the channels of *any* IF.
This insures that the uv-coverage is identical in every channel
-- an excellent idea for delicate spectral line work, to avoid
interpreting a change in the beam as a change in the line
emission.
DOWEIGHT <= -0.5 weight the data by the sum of the incoming
weights. In this case a baseline in an overlapped channel which
is flagged in one IF will wind up with a lower weight, but will
still appear in the output data set.
DOWEIGHT >= 0,5 flag a baseline in a given channel if that
baseline/channel pair is flagged in *any* input IF.
Comments and Suggestions
UJOIN should usually be run after amplitude, phase, and bandpass
calibration, and before continuum subtration. Care should be taken
with the bandpass calibration in particular to ensure that the
spectrum produced by UJOIN doesn't create false emission/absorption
-- one useful check is to run UJOIN on your phase calibrator, and
confirm (with POSSM) that the resulting spectrum is indeed flat, or
at least linear, with frequency.
Unless you are very confident that you know what you're doing, set
DOWEIGHT=0 to insure that the uv-coverage of the UJOIN'd file is
the same in every channel. See van Gorkom and Ekers 1994 (lecture
18 of _Synthesis Imaging in Radio Astronomy_, ASP Conference Series
vol. 6, eds. Perley, Schwab, & Bridle) for an expanation and
examples of the subtle errors that might otherwise result.