totalvis: Create an object for visualizing pdp plots for a principal...

In nickseedorff/totalvis: totalvis: A Principal Components Approach to Visualizing Total Effects in Black Box Models

Description

Create an object for visualizing pdp plots for a principal component

Usage

totalvis(
  model,
  X,
  type = "regression",
  pc_num = 1,
  samp_size = 50,
  feature = NULL,
  pin = NULL,
  ice = FALSE
)

Arguments

model	A fitted model object of an appropriate class
X	Design matrix that the object was trained on
type	Outcome type, either 'classification' (binary) or 'regression'
pc_num	Integer of principal component to visualize
samp_size	Number of unique PC values marginalize and plot
feature	Feature to plot a partial dependence plot for
pin	Feature to pin pdp plot to over a range of a principal component
ice	Logical, options to include individual expectation curves

Value

A pdp style plot of the specified principal component

nickseedorff/totalvis documentation built on May 12, 2021, 8:49 p.m.