Previous Calibration: gaintable, gainfield, interp and spwmap

NRAO Home > CASA > CASA Cookbook and User Reference Manual

4.4.1.4 Previous Calibration: gaintable, gainfield, interp and spwmap

Calibration tables that have already been determined can also be applied before solving for the new table:

gaintable    =         ’’   #   Prior gain calibration table(s) to apply
gainfield    =         ’’   #   Field selection on prior gaintable(s)
interp       =         ’’   #   Interpolation mode (in time) for prior gaintable(s)
spwmap       =         []   #   Spectral window mapping for each gaintable (see help)

This is controlled by the gaintable parameter, which takes a string or list of strings giving one or more calibration tables to pre-apply. For example,

gaintable = [’ngc5921.bcal’,’ngc5921.gcal’]

specifies two tables, in this case bandpass and gain calibration tables respectively.

The other parameters key off gaintable, taking single values or lists, with an entry for each table in gaintable. The order is given by that in gaintable.

The gainfield parameter specifies which fields from the respective gaintable to use to apply. This is a list, with each entry a string or list of strings. The default ’’ for an entry means to use all in that table. For example,

gaintable = [’ngc5921.bcal’,’ngc5921.gcal’]
gainfield = [ ’1331+305’, [’1331+305’,’1445+099’] ]

or using indices

gainfield = [ ’0’, [’0’,’1’] ]

to specify the field ’1331+305’ from the table ’ngc5921.bcal’ and fields ’1331+305’ and ’1445+099’ from the second table ’ngc5921.gcal’. We could also have wildcarded the selection, e.g.

gainfield = [ ’0’, ’*’ ]

taking all fields from the second table. And of course we could have used the default

gainfield = [ ’0’, ’’ ]

or even

gainfield = [ ’0’ ]

which is to take all.

The interp parameter chooses the interpolation scheme to be used when pre-applying the solution in the tables. This interpolation is (currently) only in time. The choices are currently ’nearest’, ’linear’, and ’aipslin’:

’nearest’ just picks the entry nearest in time to the visibility in question;
’linear’ interpolation calibrates each datum with calibration phases and amplitudes linearly interpolated from neighboring time values. In the case of phase, this mode will assume that phase never jumps more than 180^∘ between neighboring points, and so undersampled cycle-slips will not be corrected for. Solutions will not be extrapolated arbitrarily in time for data before the first solution or after the last solution; such data will be calibrated using ’nearest’ to avoid unreasonable extrapolations.
’aipslin’ emulates the classic AIPS interpolation mode with linearly interpolated amplitudes and phases derived from interpolation of the complex calibration values. While this method avoids having to track cycle slips (which is unstable for solutions with very low SNR), it will yield a phase interpolation which becomes increasingly non-linear as the spanned phase difference increases. The non-linearity mimics the behavior of interp=’nearest’ as the spanned phase difference approaches 180^∘ (the phase of the interpolated complex calibration value initially changes very slowly, then rapidly jumps to the second value at the midpoint of the interval).

If the uncalibrated phase is changing rapidly, a ’nearest’ interpolation is not desirable. Usually, interp=’linear’ is the best choice. For example,

interp = [ ’nearest’, ’linear’ ]

uses nearest “interpolation” on the first table, and linear on the second.

The spwmap parameter sets the spectral window combinations to form for the gaintable(s). This is a list, or a list of lists, of integers giving the spw IDs to map. There is one list for each table in gaintable, with an entry for each ID in the MS. For example,

spwmap=[0,0,1,1] # apply from spw=0 to 0,1 and 1 to 2,3

for an MS with spw=0,1,2,3. For multiple gaintable, use lists of lists, e.g.

spwmap=[ [0,0,1,1], [0,1,0,1] ] # 2nd table spw=0 to 0,2 and 1 to 1,3

[next] [prev] [prev-tail] [front] [up]

More information about CASA may be found at the CASA web page

This code is available under the terms of the GNU General Public Lincense