check_bart_error_assumptions: Check BART Error Assumptions

In bartMachine: Bayesian Additive Regression Trees

Check BART Error Assumptions

Description

Diagnostic tools to assess whether the errors of the BART model for regression are normally distributed and homoskedastic, as assumed by the model. This function generates a normal quantile plot of the residuals with a Shapiro-Wilks p-value as well as a residual plot.

Usage

check_bart_error_assumptions(bart_machine, hetero_plot = "yhats")

Arguments

bart_machine	An object of class “bartMachine”.
hetero_plot	If “yhats”, the residuals are plotted against the fitted values of the response. If “ys”, the residuals are plotted against the actual values of the response.

Value

None.

Author(s)

Adam Kapelner and Justin Bleich

See Also

plot_convergence_diagnostics

Examples

## Not run: 
#generate Friedman data
set.seed(11)
n  = 300 
p = 5
X = data.frame(matrix(runif(n * p), ncol = p))
y = 10 * sin(pi* X[ ,1] * X[,2]) +20 * (X[,3] -.5)^2 + 10 * X[ ,4] + 5 * X[,5] + rnorm(n)

##build BART regression model
bart_machine = bartMachine(X, y)

#check error diagnostics
check_bart_error_assumptions(bart_machine)

## End(Not run)

bartMachine documentation built on July 9, 2023, 5:59 p.m.