Skip to contents

Variable Importance Summary

Source: R/importance.r
importance.Rd

Extract variable importance from a fitted RLT model. When variance estimation was enabled via var.mode, standard deviations, Z-scores, and significance codes are also reported. Negative variance estimates yield NA for SD, Z, and significance.

Usage

importance(object, ...)

Arguments

object

A fitted RLT object from RLT.

...

Additional arguments (unused).

Value

A data.frame with columns:

  • Variable: variable name

  • VI: variable importance

  • SD: standard deviation of VI (NA if not estimated or negative variance)

  • Z: Z-score (VI / SD, NA if SD is NA)

  • Sig: significance code ("" if not estimated or negative variance)

Significance codes: *** |Z| >= 2.58, ** |Z| >= 1.96, * |Z| >= 1.64.

Examples

if (FALSE) { # \dontrun{
fit <- RLT(x, y, model = "classification", importance = TRUE, var.mode = TRUE)
importance(fit)
} # }