A problem about calculating sample entropy

[Q-Qing]

I think there is a problem in calculating B of sample entropy. Variable xmi and xmi should be the same, which both contain N-m templates of length m. As stated in the original paper, sample entropy only considers the first N - m vectors of length m, ensuring that, xm(i) and xm+1(i) were defined.

[liuchaojie85]

I think 'B = np.sum([np.sum(np.abs(xmii-xmj).max(axis=1) <= r)-1 for xmii in xmi])' is revised ‘B = np.sum([np.sum(np.abs(xmii-xmj).max(axis=1) <= r)-1 for xmii in xmj])’