Generalise comp mass conv (Resolve #436)

**Created a new merge request for this @colm.talbot since my old branch was froma very old version of master and rebasing was proving to be challenging but this new MR is from a branch off the recent master with the changes you kindly pointed me to in the previous MR. Thanks!! **

This MR resolves issue #436 (closed) by including the prior ratio weights as well as the Jacobian term for the function to resample posterior samples with priors defined in Chirp Mass / Mass Ratio to samples with Uniform priors in Component Masses

Here is two example plots showing the function converting both Uniform in mc/q and square power law in mc/q to flat in m1/m2 with the blue the unconverted samples, orange the converted samples, and green being samples drawn from the flat in m1/m2 priorset. (legend got cut off my pics)