Calibration solutions for QuickLook images

[hsuyeep]

QuickLook images are currently imaged in python on a subband independent of the ones going to the pipeline. We would like to calibrate these images periodically, say 1/day. The solutions can be generated by:

• A periodic run of the Matlab calibration pipeline on e.g. octave on the atv host
• A short run on the atv subband using the calim. pipeline.
• Periodic transfer of calibration solutions routinely being generated by the calibration pipeline on the other subbands, with the assumption that the same solutions can roughly be applied.

Nevertheless, we might need a socket connection to atv, or perhaps an easier task is to scp the files from whereever they are generated onto the host running the atv.

[hsuyeep]

Added ability to specify pipeline ip address and port from which to read the gain solutions. New commandline: ./atv.py --antpos=/usr/local/share/aartfaac/antennasets/lba_outer.dat --freq=54000000 --output=/data/atv --snapshot=/var/www/ --calip=ais001 --calport=4500

If the pipeline process is not running, we read the calibration solutions from a fixed location on disk. This is currently hardcoded in the code. The parsing was checked by writing another program which reads a file for the text output expected to be generated by a pipeline, and then waits on a specific port, sending it to atv when requested by the process. Vote to include this program into the test suite.

TODO:

• Store the calibration vectors read into folder specified with --caldir option.
• Periodically update the calibration vectors (currently being done only once) based on time elapsed.

[hsuyeep]

Moved this issue to the aartfaac-tools repo. Closing issue here.