plot.gg_beta_uvarpro: Plot a 'gg_beta_uvarpro' object

In ggRandomForests: Visually Exploring Random Forests

Plot a gg_beta_uvarpro object

Description

Horizontal bar chart of the mean absolute lasso coefficient \mathrm{mean}(|\hat{\beta}|) per variable from an unsupervised varPro fit, sorted descending so the eye lands on the top variable first. Bars are filled blue above the selection cutoff, gray otherwise, with a dashed red line at the cutoff.

Usage

## S3 method for class 'gg_beta_uvarpro'
plot(x, ...)

Arguments

x	A gg_beta_uvarpro object from gg_beta_uvarpro().
...	Not currently used.

Value

A ggplot object.

Reading the chart

Each bar is the average magnitude of a per-region lasso coefficient for that variable, computed by varPro::get.beta.entropy() over the unsupervised entropy regions of a varPro::uvarpro() fit. There is no response: the score measures how strongly a variable is reconstructed by the others within released regions, i.e. an unsupervised importance / redundancy signal rather than a predictive one. As with gg_beta_varpro(), the numeric scale carries the predictors' units, so bar lengths are comparable within a data set but not blindly across variables on very different scales.

See Also

gg_beta_uvarpro(), gg_beta_varpro(), gg_udependent().

Examples

if (requireNamespace("varPro", quietly = TRUE)) {
  set.seed(1)
  o <- varPro::uvarpro(mtcars, ntree = 50)
  plot(gg_beta_uvarpro(o))
}

ggRandomForests documentation built on Aug. 4, 2026, 5:09 p.m.