Re: [Numpy-discussion] Possible Deprecation of np.ediff1d