Design & prototype an automated solution for AA0.5 station beam calibration

MCCS sky modelling and calibration tools have been incorporated into the new ska-low-mccs-calibration python module, which is available on gitlab and documented in the SKA developer portal . The main branch of the package contains all of the functionality developed for this feature. This includes: Conversion between MCCS data products and SDP data models. Automatic generation of sky models and model visibilities. Automatic calculation and application of calibration solutions. Access to new SDP solvers with improved polarisation calibration capabilities. Generation of QA metrics. Unit testing and documentation. Many of the functions in the module are based on existing station calibration scripts and have been updated to meet SKA software standards. They have also been adapted for more efficient processing in both runtime and memory usage. The interfaces to the main autonomous functions use SDP data models to minimise the coupling with intricacies of MCCS data products and the DAQ. The calibrate_mccs_visibility function can automatically form a sky model based on the contents of a Visibility Xarray Dataset. It returns calibration solutions, calibrated visibilities and QA metrics, and it is intended that this function will be called in an autonomous fashion from a MCCS script or pipeline. In the course of the PI, processing and design discussions were held with R.Subrahmanyan , Drew.Devereux , S.Asayama and M.Serylak , and a DP system demo was given. Plots and QA metrics presented in the demo were for the most part taken from Ravi's original notebooks, and some of these are returned by the calibrate_mccs_visibility function after calibration. Other metrics that indicate the state of the data and the sky model are also returned, however use of the package with more data of varying quality is needed to fine tune what is returned. Further details on the architecture of MCCS station calibration can be found here .