Pipeline: Calibrating MWAX data

Owned by Samuel James McSweeney
Last updated: Nov 04, 2021

This page outlines both existing and planned pipelines for producing calibration solutions (with pulsar applications in mind).

Case 1: In-beam calibration

Overview

Green = Intermediate data products

Blue = processing steps

Steps

  1. Offline Fine PFB: The documentation for the offline fine PFB is found here. The FINEPFB filter should be used (as of this writing, 4 Nov 2021, this is the only available option).
  2. Offline Correlator: The documentation for the offline correlator is found here. On Garrawarla, there are two versions available, which differ by the integration time set by the underlying xGPU dependency. The offline_correlator/master module uses a "base" integration time of 0.1 secs, while the offline_correlator/millisecond module uses 0.02 secs. For the purposes of generating calibration solutions, the master version should be used. This step can also be achieved using the legacy pipeline, documented here.
  3. RTS: Currently, the only documentation is found on the legacy page.

Case 2: Separate MWAX calibration observation

Overview

Green = Intermediate data products

Blue = processing steps

As of this writing, is is not yet confirmed whether the various steps in this pipeline are able to be glued together in this particular way. In particular, it is believed that the RTS can read and process UV Fits files, but this may require some special configuration settings of the RTS that we are currently unaware of. This information will be updated as it becomes available.

Steps

  1. Birli: A link to documentation can be found here. Currently still in development.
  2. RTS: Currently, the only documentation is found on the legacy page.

Verifying solutions

The legacy page describes one set of tools for visualising the RTS solutions, and these can be applied equally well to the RTS solutions produced by any of the methods above. However, VCSBeam (>= v2.18) now contains tools for visualising the solutions in an alternative, complementary way.

  1. Use mwa_plot_calibration to generate an ASCII file containing the calibration solution in a way that can be readily plotted.
  2. Use plot_calibration.py to generate the plots themselves.

Example usage

$ mwa_plot_calibration -m /PATH/TO/1240826896_metafits_ppds.fits -c /PATH/TO/1240827912.metafits -C /PATH/TO/RTS/OUTPUT -N 24 -R NONE -U 0,0 -X > solution.txt
$ /PATH/TO/plot_calibration.py solution.txt

Example plots