ccalapply (Calibration Applicator)

This page provides instruction for using the ccalapply program. The purpose of this software is to apply calibration parameters to Measurement Sets.

Running the program

It can be run with the following command, where “” is a file containing the configuration parameters described in the next section.

$ ccalapply -c

The ccalapply program can either be used in serial mode when it runs on a single input measurement set or in parallel/distributed mode where there are three options (see the distribute and Tiles parset parameters). First, each rank can work with its own measurement set. This may be handy if one has one measurement set per beam or other distribution which leaves the frequency axis intact. The second option involves distribution of data in frequency space internally. Only one measurement set should be given in this case and the appropriate split is done based on the available number of ranks. This is the default behaviour. The third option uses different ranks to process different data tiles in the same dataset and update the data in parallel (with an appropriately patched casacore version), using

Ccalapply.Tiles      = auto
Ccalapply.distribute = false

Configuration Parameters

The following table contains the configuration parameters to be specified in the “” file shown on above command line. Note that each parameter must be prefixed with “Ccalapply.”. For example, the “dataset” parameter becomes “Ccalapply.dataset”.

In addition to the below parameters, those described in Access to calibrator solutions are applicable. Specifically:

  • Ccalapply.calibaccess

  • Ccalapply.calibaccess.parset (if calibaccess is “parset”); or

  • Ccalapply.calibaccess.table (if calibaccess is “table”)








The name of the measurement set to which the calibration parameters will be applied. In the parallel mode with the distribution option is false, usual substitute rules apply.




If true, the noise estimate will be scaled in accordance with the applied calibrator factor to achieve proper weighting.




If true, corresponding visibilities are flagged if the inversion of Mueller matrix fails. Otherwise, an exception is thrown should the matrix inversion fail




If true, the calibration solution corresponding to beam 0 will be applied to all beams




If true, the same calibration solution will be applied to all channels. Use this to speed up the application of selfcal gains.




If true, do linear interpolation of gain solutions




Frequency frame to work in (the frame is converted when the dataset is read). Either lsrk or topo is supported.


unsigned int

max integer

If defined, the chunk size presented at each iteration will be restricted to have at most this number of rows. It doesn’t affect the result, but may give different performance in the case of long spectral axis (think of it as a slice in rows will be taken once and then each spectral slice will be made from this reduced cube for the price of having more iterations).




If the application is executed in parallel mode (i.e. more than one rank is available) and this option is true, the data are split in frequency between all available workers (i.e. ranks with non-zero numbers) which do calibration application. The master rank (always rank 0) writes the result to the measurement set. If this option is false, each rank is expected to deal with its own measurement set and its own reading, calibration application and writing. Note, there is no master in this mode. So the substitution rules should use rank (%r) rather than the worker (%w) number. This option has no effect in the serial mode. Set to false if using the Tiles selection keyword for distribution




Select a subset of data based on tile number. This is only allowed for data stored in tables using tiling. Either specify a 2 elements vector giving the number of tiles and the starting tile (“%w” type substitutions are performed) or the single string “auto”, which will distribute tiles over ranks automatically. The casacore command showtableinfo can be used to check the tiling of a dataset.


Ccalapply.dataset                   =

Ccalapply.calibaccess               = table
Ccalapply.calibaccess.table         =
Ccalapply.distribute                = false
Ccalapply.Tiles                     = auto