Per-variable lasso-beta importance from an unsupervised varPro fit

Tidy wrapper around varPro::get.beta.entropy() for a uvarpro object. Where gg_beta_varpro() refines the supervised release-rule contrast, gg_beta_uvarpro() does the unsupervised analogue: uvarpro() builds entropy regions with no response, and get.beta.entropy() fits a cross-validated lasso within each region to ask how strongly every other variable explains the released variable. Averaging the absolute lasso coefficients per variable gives one number per variable: an unsupervised, lasso-flavoured importance.

Usage

gg_beta_uvarpro(object, ..., cutoff = NULL, beta_fit = NULL)

Arguments

object: A uvarpro object from varPro::uvarpro().
...: Forwarded to varPro::get.beta.entropy() when beta_fit = NULL (e.g. pre.filter, second.stage, use.cv). Ignored, with a warning, when beta_fit is supplied.
cutoff: Selection threshold on beta_mean. NULL (default) uses mean(beta_mean); a scalar sets it explicitly. Variables at or above the cutoff are flagged selected.
beta_fit: Optional pre-computed varPro::get.beta.entropy() matrix for object. When supplied, must be a numeric matrix with column names (the variables); ... is then ignored.

Value

A gg_beta_uvarpro object (a data.frame), one row per variable, most-important first, with columns:

variable: factor; levels reversed so the most-important variable lands at the top after coord_flip() (the gg_vimp convention).
beta_mean: mean(|lasso beta|) over the released regions (colMeans(beta, na.rm = TRUE)).
n_released: number of regions contributing a non-NA coefficient for the variable.
selected: logical; beta_mean >= cutoff.

The provenance attribute records source, family ("unsupv"), cutoff, n_var, n_released_regions, and precomputed.

Details

get.beta.entropy(o) returns a (released-variable x variable) numeric matrix of absolute lasso coefficients. The column mean (na.rm = TRUE) is the per-variable importance reported here, matching the canonical sort(colMeans(beta, na.rm = TRUE), decreasing = TRUE) idiom in the varPro::uvarpro() help ("iowa housing - illustrates lasso importance").

Because get.beta.entropy() is expensive (a cross-validated glmnet per region), the beta_fit argument accepts a pre-computed matrix so you can iterate on the cutoff without re-fitting. The pairing mirrors the beta_fit argument of gg_beta_varpro().

Examples

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

# }