Following on #41, flagging in can hurt the ACA data much more than the main array or VLA data as any given solution is a much larger fraction of the overall number of solutions. To remedy this, explicitly set the number of antennas flagged threshold for the ACA to 0.
This works on the test dataset where this issue became apparent, and I am now running some additional tests to verify the impacts.
Following on #41, flagging in can hurt the ACA data much more than the main array or VLA data as any given solution is a much larger fraction of the overall number of solutions. To remedy this, explicitly set the number of antennas flagged threshold for the ACA to 0.
This works on the test dataset where this issue became apparent, and I am now running some additional tests to verify the impacts.