mrPdPlotBootstrap: Bootstrap Partial Dependence Plots

In nfj1380/mrIML: Multi-Response (Multivariate) Interpretable Machine Learning

Bootstrap Partial Dependence Plots

Description

This function extracts and plots the bootrapped partial dependence functions calculated by mrBootstrap() for each response variable.

Usage

mrPdPlotBootstrap(
  mrIML_obj,
  mrBootstrap_obj,
  vi_obj = NULL,
  target,
  global_top_var = 2
)

Arguments

mrIML_obj	A list object returned by mrIMLpredicts().
mrBootstrap_obj	A list object returned by mrBootstrap().
vi_obj	A list object returned by mrVip(). If vi_obj is not provided, then it is created inside mrPD_bootstrap by running mrVip().
target	The target variable for generating plots.
global_top_var	The number of top variables to consider (default: 2).

Value

A list with two elements:

• ⁠[[1]]⁠: A data frame of the partial dependence grid for each response model, predictor variable, and bootstrap.

• ⁠[[2]]⁠: A list of partial dependence plots for each predictor variable in the target response model.

Examples

library(tidymodels)

data <- MRFcov::Bird.parasites
Y <- data %>%
  select(-scale.prop.zos) %>%
  dplyr::select(order(everything()))
X <- data %>%
  select(scale.prop.zos)

model_rf <- rand_forest(
  trees = 50, # 50 trees are set for brevity. Aim to start with 1000
  mode = "classification",
  mtry = tune(),
  min_n = tune()
) %>%
  set_engine("randomForest")

mrIML_rf <- mrIMLpredicts(
  X = X,
  Y = Y,
  X1 = Y,
  Model = model_rf,
  prop = 0.7,
  k = 2,
  racing = FALSE
)

mrIML_rf_boot <- mrIML_rf %>%
  mrBootstrap(num_bootstrap = 5)

mrIML_rf_PD <- mrPdPlotBootstrap(
  mrIML_rf,
  mrIML_rf_boot,
  target = "Plas",
  global_top_var = 4
)

head(mrIML_rf_PD[[1]])
mrIML_rf_PD[[2]]

nfj1380/mrIML documentation built on June 2, 2025, 1:03 a.m.