Function reference • maidrr

Package maidrr
`maidrr-package`	maidrr: Model-Agnostic Interpretable Data-driven suRRogate
Datasets included
`mtpl_be`	Belgian motor third party liability insurance portfolio
`mtpl_fr`	French motor third party liability insurance portfolio
Automatic tuning via cross-validation This function automates the tuning process involved in steps 1-3 below in order to find the optimal data segmentation and surrogate GLM.
`autotune()`	Automatic tuning
Helper functions
`lambda_grid()`	Lambda grid
`err_fun()` `poi_dev()` `mse()` `wgt_mse()`	Predefined error functions
`rm_lvls()`	Drop unused factor levels
1) Obtaining model insights
`insights()`	Model insights
Helper functions
`get_vi()`	Calculate variable importance
`plot_vi()`	Plot variable importance
`get_grid()`	Get feature grid
`get_pd()`	Calculate partial dependence
`plot_pd()`	Plot partial dependence
`interaction_pd()`	Pure two-way interaction
`interaction_strength()`	Calculate interaction strength
2) Segmentation of the data
`segmentation()`	Data segmentation
Helper functions
`group_pd()` `group_pd_ckseg()` `group_pd_ckmns()`	Partial dependence grouping
`optimal_ngroups()`	Optimal number of groups
3) Fitting a GLM surrogate
`surrogate()`	Surrogate GLM
4) Explaining a prediction
`explain()`	Explain predictions

Reference