Christopher Schölzel

Hey, thank you very much for looking into this! :smile: And also thanks for the hint at Lévy motion in your NeuroKit2 comment. I am always happy to have data...

As workaround one could use a CI script which generates releases that include the content of the relevant submodules as asset. However, due to #1235 , this currently also does...

On a different note, could something like this be a solution for the licensing issues? Or would that mean too much configuration effort in the backend?  I think not...

I finally came back to this: Retracing my steps from https://github.com/neuropsychology/NeuroKit/issues/206, there is indeed a change required: We need to apply the square root only after we summed _all_ of...

Other To-dos: - [ ] Use Lévy motion as another unit test. - [x] Take a closer look at the actual values of the unit tests that still work. Did...

I looked for another paper reporting the Hurst parameter obtained by DFA from the Lorenz system and found [Wallot et al. (2023)](https://onlinelibrary.wiley.com/doi/pdfdirect/10.1111/tops.12688). The authors actually report values that are closer...

I compared the PhysioNet implementation against nolds. Fortunately, the PhysioNet binary outputs all data points on the log-log-plot, which allows a very fine-grained comparison of the two implementations. After correcting...

I added the regression test. Open TODOs: - [x] Fix expected DFA value for Lorenz in unit tests and examples and change reference from González-Salas et al. (2016) to Wallot...

Thanks a lot for the hint. Much appreciated. :heart:

Update for myself: Unfortunately, I noticed that the error I had in my formula was due to the same error existing in Hardstone et al. (2012), which I based my...