vivid

Variable importance, interaction measures and partial dependence plots are important summaries in the interpretation of statistical and machine learning models. In our R package vivid (variable importance and variable interaction displays) we create new visualisation techniques for exploring these model summaries. We construct heatmap and graph-based displays showing variable importance and interaction jointly, which are carefully designed to highlight important aspects of the fit. We also construct a new matrix-type layout showing all single and bivariate partial dependence plots, and an alternative layout based on graph Eulerians focusing on key subsets. Our new visualisations are model-agnostic and are applicable to regression and classification supervised learning settings. They enhance interpretation even in situations where the number of variables is large and the interaction structure complex.

Installation

The zenplots package (which is used within vivid) requires the graph package from BioConductor. To install the graph and zenplots packages use:

if (!requireNamespace("graph", quietly = TRUE)){
  install.packages("BiocManager")
  BiocManager::install("graph")
}
install.packages("zenplots")

You can install the released version of vivid from CRAN with:

install.packages("vivid")

And the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("AlanInglis/vivid")

You can then load the package with:

library(vivid)