v0.2

Fixes a bug pointed out by Isobel Romero-Shaw. The issue was confusion betweeen classic det-frame/source-frame masses conversion. Basically, the redshift shouldn't have been used anywhere except while extracting the eccentricity at a specified frequency.

To reiterate, the general algorithm is:

1. Use the source frame quantities (initial separation, initial eccentricity, masses) to get the e vs f curve. Interpolate this curve.
2. If you want to extract e at fixed (det frame) f22, evaluate the interpolant at f22 * (1 + redshift)
3. f instead you want e at fixed M * f22, then evaluate the interpolant at M_times_f22 / Mtot_src

Thanks to Isobel for pointing this out!

Full Changelog: v0.1...v0.2