search

mad-lab-fau / BioPsyKit

A Python package for the analysis of biopsychological data.
MIT License
42 stars 8 forks source link

[JOSS Review] Paper #16

Closed espenhgn closed 3 years ago

espenhgn commented 3 years ago

Thanks for the submission of BioPsyKit to JOSS! This Python package appears to be a useful software contribution to the biopsychology research field that aims to combine main tools and methods into one. At large, the package wraps around existing analysis and visualization tools like scipy, pandas, scikit-learn, seaborn, matplotlib etc. and implements fairly standard protocols encountered in the field, ranging from handling questionnaire data, biomarker data and time-continuous electrophysiology data like ECG and EEG. The software implementation appears sound and appears well tested in the form of many unit tests, but I did have various problems running the different example notebooks (#14 ) and did find the installation cumbersome due to difficulties in resolving dependencies (#13 ) which may be a dealbreaker for some.

Otherwise, I will keep it short here as the other reviewer already raised a few valid points (#12 ). In particular, I agree that a comparison with existing/alternative tools (also outside the Python realm) with overlapping functionality would be most useful. Does BioPsyKit for instance incorporate new features not available in its alternatives, beyond its somewhat uniform user high(er) level interface and data structures?

espenhgn commented 3 years ago

Thanks for the updated version of the manuscript. I have nothing further to add.