In some discussions, it was suggested to explore a process where we image and self-calibrate using data from long baselines in order to filter out the difficult to deconvolve diffuse emission. Once that is completed, attempt to do a deep multi-scale clean.
Although the basics of this process were already enabled, what was not available was a way of disabling either masking or self-calibration tasks throughout the self-cal and imaging loop. This attempts to bring some of this in.
I am not entirely comfortable with the current approach, which exposes some extra CLI options where rounds can be specified as ones to skip (for self-cal) or rounds to create beam masks for (as for masking, which has always been opt-in). Perhaps the better approach is to add this go/no go options to the Options classes. But that is sort of in conflict with my eorry of putting too much flow control into what should be pure options.
In some discussions, it was suggested to explore a process where we image and self-calibrate using data from long baselines in order to filter out the difficult to deconvolve diffuse emission. Once that is completed, attempt to do a deep multi-scale clean.
Although the basics of this process were already enabled, what was not available was a way of disabling either masking or self-calibration tasks throughout the self-cal and imaging loop. This attempts to bring some of this in.
I am not entirely comfortable with the current approach, which exposes some extra CLI options where rounds can be specified as ones to skip (for self-cal) or rounds to create beam masks for (as for masking, which has always been opt-in). Perhaps the better approach is to add this go/no go options to the
Options
classes. But that is sort of in conflict with my eorry of putting too much flow control into what should be pure options.