shuffled parameters for some waveforms and when using `--ignore_parameters`
With a new (not yet on master) time-domain waveform and parallel bilby json files, we've noticed cases where pesummary comparison pages had parameters shuffled: "ra is taking the value of the maximum frequency, maximum frequency is taking the values of phase, phase is taking the value of pn_tidal_order (or another parameter set to zero) and pn_tidal_order is taking the value of ra, completing the circle"...
@charlie.hoy has already debugged this so far that the problem can be avoided by dropping the use of --ignore_parameters (which we had used for the calibration parameters) and will dig deeper to see what's happening, probably some indexing that doesn't get updated correctly after the "ignore" step. Thanks a lot Charlie!