Raphael Vallat

I agree, I have been wanting to revamp all the notebooks at some point in the near future, so this is definitely something we can add at that point. That...

Hi @remrama, Both are actually included in the [requirements.txt](https://github.com/raphaelvallat/yasa/blob/master/requirements.txt) file. However I see now that `lightgbm` is not included in the [setup.py](https://github.com/raphaelvallat/yasa/blob/master/setup.py) file — we should absolutely fix this. I...

Hi @remrama, Thanks for the detailed issue. I don't see any reasons why not to implement your proposed solution, and I agree that it may improve flexibility and at the...

Thanks for opening the issue @jckrumm! For this function, I've essentially translated into Python pre-existing R and Matlab implementations (see [here](https://github.com/selcukorkmaz/MVN/blob/master/R/mvn.R) and [here](https://www.mathworks.com/matlabcentral/fileexchange/17931-hzmvntest)), which I believe are already quite optimized...

Hi @jckrumm, Thanks for your detailed response! > Are you suggesting I should instead copy your code and modify to enforce the input type in the intermediate computations? Absolutely. The...

I'm closing this issue for now, but please feel free to reopen. Thanks

Thanks @avm19! Ok I had a closer look into the source code, and I think we can get better performance (and similar results) by replacing https://github.com/raphaelvallat/pingouin/blob/adf07187bae972a14a47b3c7c85826d4b7a7b67c/pingouin/multivariate.py#L81 with ```python Y_diag =...

Hi @TanjaS91, Thanks for opening the issue. This is a normal behavior since the conversion from r d is an approximation and not an exact science. Furthermore, instead of using...

@TanjaS91, that's right — thank you for catching this mistake! Going back to my previous example: ```python xy = np.hstack((x, y)) xy_bool = np.repeat([0, 1], 100) r_biserial = pearsonr(xy_bool, xy)[0]...

Hi @dmarcos97, Thanks for this. It is quite worrying that all of the implementation return a different value. In AntroPy, the sample entropy is calculated using a custom Numba-accelerated function...