Ingest measurement sets from casda

[tjgalvin]

To date flint has been processing uncalibrated measurement sets in order to bandpass calibrate, apply solutions, image, and self-calibrate.

It would be useful to add a component to flint to start imaging using measurement sets from casda, which have already been imaged and self-calibrated. This would involve:

• properly digesting the askap pipeline naming scheme
• ensuring the appropriate fixms preprocessing has been applied
• renaming the measurement sets to flint standard.

Item 1 might not be a large issue so long as the SBID, field name and beam information are available somewhere. The SBID is already a pain we have to rely on a directory containing the MSs to be named with the SBID.

Otherwise provided an appropriate imaging strategy things should be ok.

[tjgalvin]

I have started this in the casda branch

[tjgalvin]

Have the framework to:

• regex apart the MS files names
• copy and preprocess the MSs from the askap pipeline form on CASDA and transform them to flint name / wsclean compatible
• CLI to test it out

Need to get some data and look at it now. Provided that works can mock up a quick and easy imaging + co-add pipeline.

[tjgalvin]

I have pulled into the prefect continuum_pipeline flow a step to distinguish between ASKAP pipeline produced measurement sets obtained through CASDA, which I refer to as casda throughout the code and do not require a bandpass to be applied, and the raw ASKAP measurement sets that flint was developed against, which do required a bandpass to be applied.

In the typical continuum pipeline there is the capability to perform rounds of further self-calibration, but this is not necessary and can essentially be disabled by setting the CLI argument to 0.

Since this is part of the normal continuum pipeline there is the capability to use a template file to specify imaging options that can be changed throughout rounds of subsequent self-calibration and / or imaging.

Peeling through potatopeel by @sunmish is also part to the flow provided an appropriate container is available. At the moment the list of sources peeling is performed against is small and limited to only those stored within the reference table packaged within flint.

Primitive archiving options are available, though the corresponding options (which can be placed in the imaging strategy file) and not properly connected to the actual tasks - I am a lazy pirate.

Some more testing to be done, but I believe it seems to be working as expected.

If there are any more bespoke imaging options / modes requested that deviate away from the slim approach here it might be better to consider a different specialised workflow.

[tjgalvin]

This has been addressed in #114.