Python_packages.md

I have tested all packages myself on MacOS.

General

1. black: Neat code formatter - black my_file.py. There is a jupyter notebooks alternative - nb_black with as well. Enable it with %load_ext nb_black in your notebook and disable it with %load_ext lab_black.
2. pre-commit A tool which performs formatting and makes checks such as pep8. Used in CI.
3. dask My favorite parallel computing lib in Python.

Viz

1. matplotlib-venn I use it to make Venn diagrams.

ML

1. pycaret **so impressed** An omnipotent ML tool, including data preprocess, automl part (selection of best model), great result overview and option to make ML pipeline from selected best model. Just check this example notebook.

2. shap Package to compute explainability Shapley values. You can read this paper in Nature on how to go deeper into the case of Tree models.

3 shap-select Not tried: implements a heuristic for fast feature selection for tabular regression and classification models.