Because MeerKAT data mislabels X and Y (opposing the IEEE convention) one has to apply the anti-diagonal matrix (not simply rotating or flipping the beam) to get the correct E terms out. There has to be an option to support this. Perhaps one can make this arbitrary as apply-correlation-jones [1, 0, 0, 1] by default which can then be changed by the user.
Because MeerKAT data mislabels X and Y (opposing the IEEE convention) one has to apply the anti-diagonal matrix (not simply rotating or flipping the beam) to get the correct E terms out. There has to be an option to support this. Perhaps one can make this arbitrary as apply-correlation-jones [1, 0, 0, 1] by default which can then be changed by the user.