Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.Sign up
ENH: Add the first Wasserstein and the Cramér-von Mises statistical distances #7563
Hi @CharlesMasson, looks good from a quick browse. Some questions/thoughts:
Thanks @rgommers for your feedback. I addressed your comments.
This distinction is highlighted in Székely's paper (at the beginning of the section 2): http://personal.bgsu.edu/~mrizzo/energy/Szekely-E-statistics.pdf
As shown in that paper, in one dimension, the non-distribution-free Cramér-von Mises distance is equivalent to the energy distance (https://en.wikipedia.org/wiki/Energy_distance).
For those reasons and to avoid any ambiguities, I renamed the function to
Thanks for the ping @irabinovitch, working on it now. And just pushed a rebase.
Note to other reviewers, this was already discussed on the mailing list: https://mail.python.org/pipermail/scipy-dev/2017-July/022005.html. Not many responses, but no objections so OK to add.