I tried calculating the optimal coupling from the optimal dual potentials in the branch optimalcoupling following Prop. 6 of [SFVTP19], but it doesn't seem correct (as shown by the tests). I don't actually need the coupling now, so I'll look into this more later.
I don't think there's a problem with the dual potentials, as the tests show I recover the same answer for optimal transport as the optimization problem does.
I tried calculating the optimal coupling from the optimal dual potentials in the branch
optimalcoupling
following Prop. 6 of [SFVTP19], but it doesn't seem correct (as shown by the tests). I don't actually need the coupling now, so I'll look into this more later.I don't think there's a problem with the dual potentials, as the tests show I recover the same answer for optimal transport as the optimization problem does.