Wasserstein distance for histograms

[ignace]

Hello,
How can one compute the Wasserstein distance between two histograms.
To be clear, what I mean with a histogram is, e.g., the following h :
using StatsBase
x = randn(10^4)
h = fit(Histogram, x)
Looking forward to any suggestion.

[rafael.guerra]

@ignace

Looking forward to any suggestion

Including calling Python’s wasserstein_distance()?

[Dan]

The trick was to get the right synonym for Wasserstein distance. In this case earth mover distance. A little google gives:

https://github.com/mirkobunse/EarthMoversDistance.jl

and following the instructions on the README I was able to calculate a distance.
Thanks mirkobunse (Mirko Bunse).

Also, this thread might be relevant:

Computing discrete Wasserstein (EMD) distance in Julia?

Is there any library in Julia for computing discrete Wasserstein (EMD) given two discrete distributions? Python seems to have a tool for that http://pot.readthedocs.io/en/stable/index.html, but if a pure Julia solution is available, it will be better than using PyCall.