Multiscale sample entropy: Docstring has dead link

[JohannesWiesner]

Hi! I would love to use your multiscale_entropy function :) Just as a sanity check: The link in your docstring leads to a dead end. I guess this link also refers to the reference in sample_entropy? Therefore you are using the implementation from Costa et al.? Which again should mean that your function gives the same output as neurokit2.entropy_multiscale? :) I would prefer using pyEntropy over neurokit2 since the latter seems to downgrade my pandas version and your package seems to be more lightweight :)

[nikdon]

Hey, unfortunately yes, the link is dead and most probably there is no way to restore what was there. There is a PR #4 which has not progressed for a long time. And it brought up a discussion on the way to calculate MSE. In the PR mentioned Costa's implementation. I'm encouraging you to take a look there to find out differences

[JohannesWiesner]

Yes, there seem to be a lot of packages out there that implement multiscale sample entropy and refer to Costa et al. Would be interesting to cross-check if they give the same (or at least similar) results. For example, I also discovered the pacakge EntropyHub yesterday which also seems to have implmemented it.

[nikdon]

Added a reference to the web archive in #28