bartRegrDiag

Displays a selection of diagnostic plots for a BART model.

Investigating and visualising Bayesian Additive Regression Tree (BART) (Chipman, H. A., George, E. I., & McCulloch, R. E. 2010) <doi:10.1214/09-AOAS285> model fits. We construct conventional plots to analyze a model’s performance and stability as well as create new tree-based plots to analyze variable importance, interaction, and tree structure. We employ Value Suppressing Uncertainty Palettes (VSUP) to construct heatmaps that display variable importance and interactions jointly using colour scale to represent posterior uncertainty. Our visualisations are designed to work with the most popular BART R packages available, namely 'BART' Rodney Sparapani and Charles Spanbauer and Robert McCulloch 2021 <doi:10.18637/jss.v097.i01>, 'dbarts' (Vincent Dorie 2023) <https://CRAN.R-project.org/package=dbarts>, and 'bartMachine' (Adam Kapelner and Justin Bleich 2016) <doi:10.18637/jss.v070.i04>.

Alan Inglis

bartMan

Create Visualisations for BART Models

Andrew Parnell

Catherine Hurley

Claus Wilke

bartRegrDiag function

<dl><dt>model</dt>
<dd>a model created from either the BART, modelarts, or bartMachine package.</dd>
<dt>response</dt>
<dd>The name of the response for the fit.</dd>
<dt>burnIn</dt>
<dd>Trace plot will only show iterations above selected burn in value.</dd>
<dt>data</dt>
<dd>A dataframe used to build the model.</dd>
<dt>combineFactors</dt>
<dd>Whether or not to combine dummy variables (if present) in display.</dd></dl>

Arguments

bartRegrDiag — bartRegrDiag

<dl>

<dt>model</dt>
<dd>a model created from either the BART, modelarts, or bartMachine package.</dd>


<dt>response</dt>
<dd>The name of the response for the fit.</dd>


<dt>burnIn</dt>
<dd>Trace plot will only show iterations above selected burn in value.</dd>


<dt>data</dt>
<dd>A dataframe used to build the model.</dd>


<dt>combineFactors</dt>
<dd>Whether or not to combine dummy variables (if present) in display.</dd>

</dl>

bartRegrDiag: bartRegrDiag

Description

Usage

Value

Arguments