aligned spin priors

This is not super-high-priority, but at some point, it would be good to

• Add aligned spin priors. One prior corresponds some funny distribution of dimensionless spin (a), which is whatever it needs to be in order to get the same distribution of chi_eff from a mis-aligned isotropic distribution with fixed a. There is an analytic form in Lange et al. (Note: make sure to check that Lange et al. functions reproduce plots from Grant and Colm's previous work.)

@grant.meadors @colm.talbot