RLIPP is used to quantify and compare the importance of subsystems in a visible neural network (VNN), where importance of a subsystem is the Spearman correlation of a linear regression model.
More info here: https://www.nature.com/articles/nmeth.4627
This package implements multiprocessing using joblib (https://joblib.readthedocs.io/en/latest/). It can anywhere from a few hours to weeks to execute the code depending upon the size of the data and the number of available cores.