AIPS HELP file for CPASS in 31DEC22
As of Sun Dec 4 9:17:03 2022
CPASS: Task to generate a polynomial 'Bandpass' (BP) table.
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 #
CALSOUR Bandpass calibrator sources.
QUAL -10.0 Calibrator qualifier -1=>all
CALCODE Calibrator code ' '=>all
UVRANGE UV range to select
TIMERANG Time range to select
STOKES Stokes type to select.
SELBAND Bandwidth to select (kHz)
SELFREQ Frequency to select (MHz)
FREQID Freq. ID to select.
BIF 0.0 100.0 Lowest IF number 0=>all
EIF 0.0 100.0 Highest IF number 0=>all
BCHAN 0.0 2048.0 Lowest channel number 0=>all
ECHAN 0.0 2048.0 Highest channel number 0=>all
SUBARRAY 0.0 1000.0 Subarray, 0=>all
ANTENNAS Antennas to select
CLEAN map (optional)
IN2NAME Cleaned map name (name)
IN2CLASS Cleaned map name (class)
IN2SEQ 0.0 9999.0 Cleaned map name (seq. #)
IN2DISK 0.0 9.0 Cleaned map disk unit #
INVERS -1.0 46655.0 CC file version #.
NCOMP # comps to use for model.
1 value per field
FLUX Lowest CC component used.
NMAPS 0.0 4096.0 No. Clean map files
CMETHOD Modeling method:
CMODEL Model type: 'COMP','IMAG'
'SUBI' (see HELP re images)
SMODEL Source model, 1=flux,2=x,3=y
See HELP SMODEL for details.
DOCALIB -1.0 101.0 > 0 calibrate data & weights
> 99 do NOT calibrate weights
GAINUSE CL (or SN) table to apply
DOPOL -1.0 10.0 If >0 correct polarization.
PDVER PD table to apply (DOPOL>0)
BLVER BL table to apply.
FLAGVER Flag table version
DOBAND -1.0 10.0 If >0 apply bandpass cal.
Method used depends on value
of DOBAND (see HELP file).
BPVER Bandpass table version
SOLINT Solution interval (mins)
-1 => do whole time range
REFANT Reference antenna
OUTVERS BP table version to write
SMOOTH Smoothing function. See
HELP SMOOTH for details.
ANTWT Ant. wts (0 => 1.)
BPASSPRM Control information:
1: if > 0 use only the
2: print level.
3: If > 0 do not divide data
by source model
4: If > 0 store phases only
in the BP table.
5: If = 0 divide by 'channel
0' before determining BP.
If > 1 switch off the
channel 0 divide option.
6: min. amp. closure err
7: min. ph. closure err
8: > 0 => scalar average
10: > 1 => normalize ampl.
portion of bandpass by
area under BP function.
Useful for VLBI.
11: > 1, mode to allow
calibration and divide
by channel 0. Useful for
VLBI. Do not use if you
don't understand. See HELP
CPARM Fit parameters:
1: No. of terms.
2: Max. no. iterations.
3: Convergence tolerance.
4: Pre-average interval (sec)
Default: Auto - 5 min.
Cross - 15 min.
5: Fit type (1=Re/Im; 2=A/P)
6: BP table for initial val.
7: Fit phase only (1=> with
ampl.; 2=> without)
8: >0 => Autoscale
9: >0 suppress data weighting
10:>0 => save as coefficients
rather than normal BP table
(no spectral index corr.)
ICHANSEL Array of start and stop chan
numbers, plus a channel
increment and IF to be used
to select channels to sum to
find a 'channel 0'. If all
0, range set to inner 75 percent of
DOSCALE -1.0 2.0 = -1 -> no spectral index
- 0, 1 -> fit spectral index
= 2 -> fit curvature also
SPECINDX Spectral index to correct
SPECURVE Spectral index curvature
'Channel 0' uv-data
IN3NAME Channel 0 uv name (name)
IN3CLASS Channel 0 uv name (class)
IN3SEQ 0.0 9999.0 Channel 0 uv name (seq. #)
IN3DISK 0.0 9.0 Channel 0 uv disk unit #
BADDISK 0.0 9999.0 Disks to avoid for scratch
Task: This task solves for antenna-based complex polynomial bandpasses.
The resulting BP tables are applied using AIPS adverbs DOBAND and
BPVER, in the same manner as standard bandpasses determined by
BPASS. Solutions can be made for cross-power or auto-correlation
bandpasses. In fitting data correlated at the VLBA, the
antenna-based fringe rotation applied at the correlator is
accounted for in the fit.
Model images made with both values of IMAGR's DO3DIMAG
option are handled correctly, as are multi-scale images. Set
NMAPS = NFIELD * NGAUSS.
Fitting of standard bandpasses may require a fair number
of terms (CPARM(1)) and may therefore take a long time to
compute. CPASS is sometimes used on data that have had the
time-averaged bandpass function corrected. It then can use a
few terms only with weaker sources to find modest time variable
portions of the bandpass shape.
Unless you normalize the output bandpass with BPASSPRM(5) and/or
BPASSPRM(10), it is important that SETJY or GETJY have been used on
the calibrator source in advance. This is particularly true if you
choose to use the spectral index parameters. Spectral index
parameters are used for "known" sources including 3C286, 3C48,
3C147, 3C138, 3C123, 3C196, and 3C295. They are also fit to the
fluxes in the SU table for "unknown" sources. To avoid the use of
spectral index, set DOSCALE < 0.
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.
CALSOUR....List of sources for which bandpass response
functions are to be determined.
All ' ' = all sources; a "-" before a source name.
means all except ANY source named. If the data
file is a single source file no source name need
QUAL.......Only sources with a source qualifier number in the
SU table matching QUAL will be used if QUAL is not
CALCODE....Calibrators may be selected on the basis of the
' ' => any calibrator code selected
'* ' => any non blank code (cal. only)
'-CAL' => blank codes only (no calibrators)
anything else = calibrator code to select.
NB: The CALCODE test is applied in addition to the
other tests, i.e. CALSOUR and QUAL, in the
selection of sources for which to determine
UVRANGE....Range (min, max) of projected baselines to include
0,0 => all baselines (units: klamda)
TIMERANG...Time range of the data to be selected. In order:
Start day, hour, min. sec,
end day, hour, min. sec. Days relative to ref.
STOKES.....The desired Stokes type of the output data:
'RR','LL',' '. ' '=>all.
SELBAND....Bandwidth of data to be selected. If more than
one IF is present SELBAND is the width of the
first IF required. Units = kHz, 0=> all
SELFREQ....Frequency of data to be selected. If more than
one IF is present SELFREQ is the frequency of the
first IF required. Units = MHz, 0=> all
FREQID.....Frequency identifier to select (you may determine
which is applicable from the OPTYPE='SCAN' listing
produced by LISTR). If either SELBAND or SELFREQ
are set their values overide that of FREQID.
However, setting SELBAND and SELFREQ may result in
an ambiguity, in which case the task will request
that you use FREQID.
BIF........First IF to select. 0=>all.
EIF........Highest IF to select. 0=>all higher than BIF
BCHAN......First channel to use in fit. 0=>all.
ECHAN......Highest channel to use in fit. All channels are read in
and the solution applies to all channels. The outermost
channels may be excluded from the fitting process in this
SUBARRAY...Subarray number to select. 0=>all.
ANTENNAS...A list of the antennas for which bandpasses are
to be determined..
If any number is negative then all antennas listed
are NOT to be used and all others are.
The following specify a CLEAN model to be used if a single
source was specified in SOURCES:
IN2NAME....Cleaned map name (name). Standard defaults.
Note: a CLEAN image for only a single-source may
be given although it may be in a multi-source file.
If the source table contains a flux, then that flux will
be used to scale the components model to obtain the
stated total flux. This is needed since initial Cleans
may not obtain the full flux even though they represent
all the essentials of the source structure.
IN2CLASS...Cleaned map name (class). Standard defaults.
IN2SEQ.....Cleaned map name (seq. #). 0 -> highest.
IN2DISK....Disk drive # of cleaned map. 0 => any.
INVERS.....CC file version #. 0=> highest numbered version
NCOMP......Number of Clean components to use for the model, one
value per field. If all values are zero, then all
components in all fields are used. If any value is not
zero, then abs(NCOMP(i)) (or fewer depending on FLUX and
negativity) components are used for field i, even if
NCOMP(i) is zero. If any of the NCOMP is less than 0,
then components are only used in each field i up to
abs(NCOMP(i)), FLUX, or the first negative whichever
comes first. If abs(NCOMP(i)) is greater than the number
of components in field i, the actual number is used. For
NCOMP = -1,0
says to use one component from field one unless it is
negative or < FLUX and no components from any other
field. This would usually not be desirable.
NCOMP = -1000000
says to use all components from each field up to the
first negative in that field.
NCOMP = -200 100 23 0 300 5
says to use no more than 200 components from field 1, 100
from field 2, 23 from field 3, 300 from field 5, 5 from
field 6 and none from any other field. Fewer are used if
a negative is encountered or the components go below
FLUX.......Only components > FLUX in absolute value are used in the
NMAPS......Number of image files to use for model. For multi-scale
models, set NMAPS = NFIELD * NGAUSS to include the Clean
components of the extended resolutions. If more than one
file is to be used, the NAME, CLASS, DISK and SEQ of the
subsequent image files will be the same as the first file
except that the LAST 3 or 4 characters of the CLASS will
be an increasing sequence above that in IN2CLASS. Thus,
if INCLASS='ICL005', classes 'ICL005' through 'ICLnnn'
or 'ICnnnn', where nnn = 5 + NMAPS - 1 will be used. Old
names (in which the 4'th character is not a number) are
also supported: the last two characters are '01' through
'E7' for fields 2 through 512. In old names, the highest
field number allowed is 512; in new names it is 4096.
CMETHOD....This determines the method used to compute the
model visibility values.
'DFT' uses the direct Fourier transform, this
method is the most accurate.
'GRID' does a gridded-FFT interpolation model
' ' allows the program to use the fastest
NOTE: when using a model derived from data with
different uv sampling it is best to use 'DFT'
CMODEL.....This indicates the type of input model; 'COMP' means that
the input model consists of Clean components, 'IMAG'
indicates that the input model consists of images.
'SUBI' means that the model consists of a sub-image of
the original IMAGR output. If CMODEL is ' ' Clean
components will be used if present and the image if not.
SUBI should work for sub-images made with DO3DIM true and
sib-images of the central facet made with DO3DIM false,
but probably will not work well for shifted facets with
DO3DIM false. Use BLANK rather than SUBIM in such cases.
CALIB will set a scaling factor to correct image units
from JY/BEAM to JY/PIXEL for image models. If the source
table contains a flux, then that flux will be used to
scale the components model to obtain the stated total
flux. This is needed since initial Cleans may not obtain
the full flux even though they represent all the
essentials of the source structure.
SMODEL.....A single component model to be used instead of a
CLEAN components model; if abs (SMODEL) > 0 then
use of this model is requested.
SMODEL(1) = flux density (Jy)
SMODEL(2) = X offset in sky (arcsec)
SMODEL(3) = Y offset in sky (arcsec)
SMODEL(4) = Model type:
0 => point model
1 => elliptical Gaussian and
SMODEL(5) = major axis size (arcsec)
SMODEL(6) = minor axis size (arcsec)
SMODEL(7) = P. A. of major axis (degrees)
2 => uniform sphere and
SMODEL(5) = radius (arcsec)
DOCALIB....If true (>0), calibrate the data using information in the
specified Cal (CL) table for multi-source or SN table for
single-source data. Also calibrate the weights unless
DOCALIB > 99 (use this for old non-physical weights).
The calibration is applied prior to the bandpass
GAINUSE....version number of the CL table to apply to
multisource files or the SN table for single
source files. 0 => highest.
DOPOL......If > 0 then correct data for instrumental polarization as
represented in the AN or PD table. This correction is
only useful if PCAL has been run or feed polarization
parameters have been otherwise obtained. See HELP DOPOL
for available correction modes: 1 is normal, 2 and 3 are
for VLBI. 1-3 use a PD table if available; 6, 7, 8 are
the same but use the AN (continuum solution) even if a PD
table is present.
PDVER......PD table to apply if PCAL was run with SPECTRAL true and
0 < DOPOL < 6. <= 0 => highest.
BLVER......Version number of the baseline based calibration
(BL) table to appply. <0 => apply no BL table,
0 => highest.
FLAGVER....specifies the version of the flagging table to be
applied. 0 => highest numbered table.
<0 => no flagging to be applied.
DOBAND.....If true (>0) then correct the data for the shape of the
antenna bandpasses using the BP table specified by BPVER.
The correction has five modes:
(a) if DOBAND=1 all entries for an antenna in the table
are averaged together before correcting the data.
(b) if DOBAND=2 the entry nearest in time (including
solution weights) is used to correct the data.
(c) if DOBAND=3 the table entries are interpolated in
time (using solution weights) and the data are then
(d) if DOBAND=4 the entry nearest in time (ignoring
solution weights) is used to correct the data.
(e) if DOBAND=5 the table entries are interpolated in
time (ignoring solution weights) and the data are then
BPVER......Specifies the version of the BP table to be applied.
<0 => no bandpass correction done.
** NOTE - AIPS does not now have a way to "concatanate"
** BP tables. Thus, if one determines a BP table having
** applied a BP table, one has no way to get that net
** calibration applied to the data. Nonetheless, it may
** be of interest, for example, to use DOBAND=1 for
** example and then to see what is left as a
** time-variable BP, examining it with BPLOT.
SOLINT.....the interval over which to average the data
before solving for the bandpasses (MINS).
(0 => scan, -1 => whole timerange)
REFANT.....the antenna to use as a reference in the
least squares solution.
OUTVERS....the version of the BP table to fill. (0 => 1)
SMOOTH.....Specifies the type of spectral smoothing to be applied to
a uv database . The default is not to apply any smoothing.
The elements of SMOOTH are as follows:
SMOOTH(1) = type of smoothing to apply: 0 => no smoothing
To smooth before applying bandpass calibration
1 => Hanning, 2 => Gaussian, 3 => Boxcar, 4 => Sinc
To smooth after applying bandpass calibration
5 => Hanning, 6 => Gaussian, 7 => Boxcar, 8 => Sinc
SMOOTH(2) = the "diameter" of the function, i.e. width
between first nulls of Hanning triangle and sinc
function, FWHM of Gaussian, width of Boxcar. Defaults
(if < 0.1) are 4, 2, 2 and 3 channels for SMOOTH(1) =
1 - 4 and 5 - 8, resp.
SMOOTH(3) = the diameter over which the convolving
function has value - in channels. Defaults: 1,3,1,4
times SMOOTH(2) used when input SMOOTH(3) < net
Note that any use of SMOOTH in CPASS will require that
the same SMOOTH values be used when accessing any data to
be calibrated by the BP table produced by CPASS.
ANTWT......Antenna weights for up to 30 antennas. (0=>1.0)
(1) > 0 => fill BP table with autocorrelation
data only, ignoring the phases.
(2) print level
(3) > 0 => do not divide data by a model of the
source. The default is that the data will
be divided by a source model, the model
used will be (in order of decreasing
(1) the flux density specified in
the SU table
(2) a point source model (SMODEL)
(3) CLEAN components,
If BPASSPRM(5) = 0 the data will not be
divided by a source model.
(4) > 0 => store only the phase of the complex
bandpass function in the BP table. You may
want to do this if, for instance, your data
have been normalized by the autocorrelation
data as is possible at the VLA.
(5) = 0 => divide the line data by 'channel 0',
this is a useful operation if you wish to
determine your antenna based bandpasses from
uncalibrated data. Channel 0 is a 'VLAism'
for the vector averaged center 75 percent of the
observing band. If IN3NAME is specified
CPASS will use an external channel 0, else
it will generate one internally.
Note that this is on a record by record basis and,
if you have good phase stability, it might be
better to do the normalization on the SOLINT
average instead; see CPARM(8).
(6) If > 0, then the values of any amplitude
closure errors whose abs. percentage value
exceeds BPASSPRM(6) will be printed.
(7) If > 0, then the values of any phase closure
errors whose value exceeds BPASSPRM(7)
degrees will be printed.
(8) If > 0, then the amplitudes will be scalar
averaged before determining the solutions.
(10) If > 0 then will normalize the amplitude
portion of the bandpass function by the area
under the curve (to force the area to = 1.0).
This is useful for VLBI data where a useful
'channel 0' is sometimes difficult to obtain.
(11) If > 0 will then allow the user to apply
calibration and divide by channel 0. This is
disallowed in normal useage, however VLBI
users need this because they need to apply
delay and fringe rate solutions to the data
before dividing by channel 0 (delay is
CPARM..... Parameters relevant to the least-squares fit:
(1) No. of terms to use in the polynomial
expansion. 0 -> 20
(2) Maximum number of iterations allowed. 0 -> 50
(3) Convergence limit. 0 -> 0.001
***Convergence is a real issue, you may need to set
CPARM(2) and CPARM(3) fairly high.***********
Failure to converge will be reported but the
solution will be saved anyway.
(4) Pre-average interval (seconds).
CPASS pre-averages the uv-data over sub-intervals
of each solution interval to improve the SNR before
the bandpass fit. For VLBA data the pre-average
interval should not be very long compared to
the time-scale over which the fringe-rate is
changing significantly at each antenna.
For low-SNR VLBA data, however, it is more important
to pre-average over a longer interval when solving
for cross-power bandpass solutions.
For VLA data the pre-average interval can be set to
a large value so that the entire solution interval
is averaged before the fit.
Default: Autocorrelation bandpass - 5 min.
Cross-correlation bandpass - 15 min.
The data are not averaged beyond scan boundaries no
matter what the pre-averaging value is.
(5) Fit type: 0 -> 2
1: Chebyshev expansion of (Re, Im).
2: Chebyshev expansion of (Amp, Phs).
(6) BP table to use for initial estimates of
the polynomial coefficients. Must be of
compatible type to the present fit.
(7) Fit phase only ? (1 => fit for phase only
using initial amplitude solutions; 2 =>
fit for phase without reference to amplitude
(8) If >= 0 then auto-scale the data before the
fit to unit amplitude and zero mean phase.
See also BPASSPRM(5). With data of good phase
stability, the scan average will produce a much
less noisy normalization than the record-by-record
divide by channel 0. ***** You must use one (or
both) of these normalizations; the fitting routine
in CPASS does not work on un-normalized data. *****
(9) The fitting is done using data weights. If
CPARM(9) <= 0, the pre-average is used to find an
rms for each baseline and channel and the data
weight is set based on this rms tempered some by
the input weight. If CPARM(9) > 0, the pre-average
rms is ignored. If CPARM(9) > 1.5, the data
weights are set to one everywhere (except flagged
data of course).
(10) > 0 => Save the results as the Chebyshev
polynomial coefficients, otherwise convert to a
normal Real/imaginary BP table. When applying BP
tables with interpolations, whatever is stored
in the BP table is interpolated. This option
allows you to interpolate coefficients rather than
ICHANSEL.. Array of start, stop, and increment channel numbers plus
an IF used for channel selection in the averaging to
compute a channel 0. Up to 20 sets if channels/IF may be
entered. The first having ICHANSEL(2,i) <= 0 terminates
the list. ICHANSEL(4,i) is the IF number, with <= 0
meaning all IFs. If an IF has no ICHANSEL set for it,
then the inner 75 percent of that IF is used. Note that these
are absolute channel numbers; they are not relative to
DOSCALE....Controls whether spectral index corrections are made.
DOSCALE < 0 => do no spectral index correction.
DOSCALE >= 0 => do spectral index correction, using
SPECINDX and SPECURVE for the first calibrator or
using known parameters, or fitting fluxes in the SU
DOSCALE >= 2 => fit spectral index and curvature for
unknown sources. Do this only for well-determined
fluxes over a wide frequency range.
SPECINDX...The fit bandpasses are corrected for a calibrator
spectral index. The calibrator flux is assumed to be
frequency^SPECINDX so the bandpass is multiplied by
(freq/f0) ** SPECINDX when SPECINDX is not zero.
If SPECINDX = 0 and the calibrator source is one of
3C286, 3C48, 3C147, 3C138, 3C123, 3C196, or 3C295 (the
standard "known" ones) SPECINDX and SPECURVE will be set
to the known (Perley 20xx) values from SETJY.
If SPECINDX=0 and the source is not one of the known
ones, then a spectral index (with curvature if DOSCALE=2)
is fit to the fluxes in the source table.
SPECURVE...If SPECINDX is not zero, a curvature term may be added to
the above correction as
A = log(f/f0) * SPECINDX + log(f/f0)^2 * SPECURVE(1)
+ log(f/f0)^3 * SPECURV(2) + log(f/f0)^3 * SPECURV(2)
and then multiply by 10^(-A/2.) where the 2 is because
the BP tables are voltages rather than powers. f0 is
taken to be 1.0 GHz by convention and all logs are base
The following specify a uv-file which can be used as an external
'channel 0'. If all the IN3* adverbs are blank, then the task
will determine channel 0 from the line data itself, unless
BPASSPRM(5) = 1, in which case no channel 0 division will be done.
IN3NAME....Channel 0 uv name (name). Standard defaults.
IN3CLASS...Channel 0 uv name (class). Standard defaults.
IN3SEQ.....Channel 0 uv name (seq. #). 0 -> highest.
IN3DISK....Disk drive # of Channel 0 uv 0 => any.
BADDISK....A list of disks on which scratch files are not to
be placed. This will not affect the output file.
CPASS: Task to determine the instrumental bandpass response
Documentor: A. J. Kemball
Related Programs: BPASS, CALIB, POSSM, LISTR, SPLIT, TVFLG
This purpose of this task is to determine the antenna-based
instrumental bandpass response, modeled as a complex polynomial
expansion, using either autocorrelation or cross-correlation
calibrator data. This task is closely related in function to BPASS,
which in contrast determines the bandpass response on a channel by
channel basis. CPASS offers an alternative bandpass calibration method
which may be useful in some cases. In fitting the polynomial bandpass
response function, CPASS correctly models the antenna-based fringe
rotation applied to both autocorrelation and cross-correlation data by
the VLBA correlator, and is therefore useful for determining bandpass
solutions for high-frequency VLBA data. It has also proved useful in
deriving bandpass solutions for narrow-band VLBA data.
MODES OF OPERATION
CPASS has two major modes of operation:
i) AUTOCORRELATION BANDPASSES
In this mode CPASS uses only the autocorrelation data in
determining the bandpass response function for each antenna.
The phase response of the resulting solutions is zero. This mode
is selected by setting BPASSPRM(1) = 1.
ii) CROSS-CORRELATION BANDPASSES
In this mode CPASS factorises the baseline-based cross-power
data on the specified continuum calibrator sources into
antenna-based complex bandpass response functions using a
non-linear least squares method. This is analogous to
self-calibration except the solution at each antenna is a
parametrized bandpass response. The phase response of the
reference antenna is zero in this case and this mode is
selected by setting BPASSPRM(1) = 0. In solving for cross-
power bandpasses a phase-only solution may be performed with
an existing AC bandpass solution used to fix the amplitude
The effect of calibrator source resolution can be minimized by
dividing the uv-data by a pre-determined source model or by the
so-called "channel-zero". These options are selected using BPASSPRM(3)
and BPASSPRM(5) respectively.
The polynomial bandpass solutions are written to the standard BP
tables used by BPASS and are applied using adverb DOBAND in the
standard manner. Multiple solutions can be appended to the same BP
table by running CPASS several times with the same output BP table
specified by OUTVERS.
The uv-data can be selected, calibrated, edited and smoothed prior to
the bandpass solutions using the standard AIPS adverbs for this
purpose (i.e. DOCALIB, GAINUSE, FLAGVER, BIF, EIF, SMOOTH etc.).
Comments on some specific adverbs follow:
SOLINT defines the time interval to use in determining each bandpass
solution. The default (SOLINT = 0) determines a bandpass solution for
each scan defined in the NX table. Setting SOLINT = -1 will determine
only one bandpass solution for each antenna for the entire run.
REFANT defines the antenna to be used as the reference in the
least squares solution. A useful choice is an antenna with good
SNR, at the center of the array that was present for a significant
fraction of the observing run.
The version of the BP table to fill. The default (OUTVERS = 0)
will cause a new table to be generated. Solutions can be appended
to an existing BP table but BPASS and CPASS solutions cannot be
stored in the same table.
Specifies the type of spectral smoothing that can be applied
to the data before the bandpasses are determined. See
HELP SMOOTH for more details.
** WARNING ** If SMOOTH is specified then all channels of
the database must be present in a single file, it will not
work properly if for instance you have split your data into
4 files each of 16 channels. The rest of CPASS will work
with no problem in that mode but the SMOOTH option will
result in corrupted data.
See the HELP section for a full description of the BPASSPRM
PARAMETERS CONTROLLING THE LEAST SQUARES FIT
The least squares fit is controlled using parameters specified in
adverb CPARM. This task uses a NETLIB routine by S.G. Nash (Siam
J. Numer. Anal. 21 (1984) which uses a truncated Newton algorithm.
CPASS can be somewhat slow when solving for a large number of free
parameters. The convergence can be monitoring by setting the print
level using BPASSPRM(2)=1, and CPASS re-started with different
least squares parameters to improve this if necessary.
A description of the least squares parameters is given below:
CPARM(1)...Number of terms to use in the polynomial expansion for
the bandpass response function at each antenna. Reasonable
values to use here are in the range 10-30. This number
cannot exceed the number of frequency channels in each IF.
CPARM(2)...Maximum number of iterations allowed. The least squares
solution will terminate after this number of function
CPARM(3)...Convergence limit. This is the fractional convergence
CPARM(4)...Pre-average interval (seconds). CPASS pre-averages the
uv-data over sub-intervals of each solution interval to
improve the SNR before the bandpass fit. For VLBA data
the pre-average interval should not be very long compared
to the time-scale over which the fringe-rate is changing
significantly at each antenna. For low-SNR VLBA data,
however, it is more important to pre-average over a longer
interval when solving for cross-power bandpass solutions.
For VLA data the pre-average interval can be set to
a large value so that the entire solution interval is
averaged before the fit.
Default: Autocorrelation bandpass - 5 min.
Cross-correlation bandpass - 15 min.
CPARM(5)...Fit type. The task can solve for bandpass functions
where either (Re,Im) or (Amp,Phs) are each expanded
separately as Chebyshev series. The (Amp,Phs) solution
is recommended, and the former may only be necessary
if residual delays have not been removed before the
Default: 2 (amplitude and phase).
CPARM(6)...In solving for a cross-power bandpass, a pre-existing
AC bandpass solution can be used either as a starting point
for the fit or to fix the amplitude response of the
bandpass solution. This BP table must have been fit
using the same fit type (CPARM(5)) as is presently
being used. Again, A&P is recommended.
Default: 0 (do not read a BP table for initial values)
CPARM(7)...This parameter determines whether phase alone is fitted.
CPARM(7) = 1 will fit for phase only using the amplitude
response specified in CPARM(6). CPARM(7) = 2 will fit
for phase alone without reference to the amplitude
CPARM(8)...Auto-normalization. This parameter enables auto-scaling
of the data to unit mean amplitude and zero mean phase
after each pre-average interval. This is necessary to
correct for calibrator resolution not previously removed.
Default: 1 (auto-scaling enabled).
CPARM(10) Option to save coefficients in the BP table (used to be the
only choice). > 0 save coefficients, <= 0 save an
evaluated real and imaginary bandpass of the "normal" sort.
All calibration routines plus POSSM can handle either. The
time smoothing and averaging is done over whatever numbers
are in the BP table, and it is more intuitive to average
normal bandpasses rather than Chebyshev polynomials.
These parameters will be consolidated and revised as further
experience is gained with the algorithm.