Optimal Transport of Classifiers to Fairness: Appendix A Derivations and Proofs