Jensen-Shannon Divergence (not distance)

[rkjulian]

First, thank you for this terrific package!

Maybe this is intentional, but the value returned for the Jensen-Shannon "distance" (coded as 'd') is the divergence, not the distance metric. Taking the square root produces the distance according to the several sources referenced on the Wikipedia page for the JSD: https://en.wikipedia.org/wiki/Jensen-Shannon_divergence. I checked the Python function scipy.spatial.distance.jensenshannon(), and it returns the square root of the value produced by this package.

Maybe some clarification could be added to the documentation if you thought it was needed.

[HajkD]

Hi @rkjulian

Apologies for the late reply, I am resurfacing slowly from my job transition from Tuebingen to Dundee.

You are absolutely correct and I think this needs to be clarified.

I will add this to the documentation to make this point crystal clear.

Thank you so much for reaching out to me.

With very best wishes,
Hajk