Hubert Baniecki

The same as in https://github.com/ModelOriented/DALEX/issues/485: change to the `ranger` package to fix CI/tests

`subsample_events()` does not return anything

1

Documentation says that `subsample_events()` returns a `numpy` array. It does not return any object, nor it changes the dataset attribute in place. See the appropriate code: https://github.com/whitews/FlowKit/blob/e049a56973bed9b4aa8455d743f4d2e9f4440399/flowkit/_models/sample.py#L370-L417

It should probably explain what's happening e.g. "specify variables to remove the warning" or "variable_splits overrides variables" https://github.com/ModelOriented/DALEX/blob/142c94d0aab51584a81e31c38d687a897b11cb60/python/dalex/dalex/predict_explanations/_ceteris_paribus/checks.py#L14-L16

- code: https://github.com/NorskRegnesentral/shapr - paper: https://doi.org/10.1016/j.artint.2021.103502 - merge from https://github.com/jakubpw/DALEX

https://arxiv.org/abs/2102.11010

I found out that `varimp()` doesn't work with `blackboost` models. Code to reproduce the error: ```r library(mboost) library(survival) model

changes like this https://github.com/AI4LIFE-GROUP/OpenXAI/commit/9f944dc0ec75265e9091c038dd9658af920d2ebf break the code thanks!

v1.4.0

support categorical features

2

`ranger.unify` and `gbm.unify` don't support categorical features, which is essential when working with `factors` in R.