How to use Wasserstein distance in OptimalTransport.jl

Hi all,

I was hoping to compare probability distributions using Wasserstein distance. As a starting point, I was hoping to compare two discrete probability distributions, say [.25.,25.,25.,25] and [0,.5,.2,.3]. I found this thread, but it referenced what appears to be an abandoned package.

Thanks!

Evidently, I was looking at the wrong package. Here is a simple example:

using ExactOptimalTransport
wasserstein(Categorical([.3,.7]), Categorical(.5,.5))
1 Like