Williamson's Variable Importance Measure (WVIM)

Base class generalizazing refit-based variable importance measures. Default corresponds to leaving out each feature iters_refit times, which corresponds to LOCO (Leave One Covariate Out).

Super class

xplainfi::FeatureImportanceMethod -> WVIM

Public fields

direction

(character(1)) Either "leave-out" or "leave-in".

Methods

Public methods

• WVIM$new()

• WVIM$compute()

• WVIM$clone()

Inherited methods

• xplainfi::FeatureImportanceMethod$importance()
• xplainfi::FeatureImportanceMethod$obs_loss()
• xplainfi::FeatureImportanceMethod$print()
• xplainfi::FeatureImportanceMethod$reset()
• xplainfi::FeatureImportanceMethod$scores()

---

Method new()

Creates a new instance of this R6 class.

Usage

WVIM$new(
  task,
  learner,
  measure,
  resampling = NULL,
  features = NULL,
  groups = NULL,
  direction = c("leave-out", "leave-in"),
  label = "Williamson's Variable Importance Measure (WVIM)",
  iters_refit = 1L
)

Arguments

task, learner, measure, resampling, features, groups

Passed to FeatureImportanceMethod for construction.

direction

(character(1)) Either "leave-out" or "leave-in".

label

(character(1)) Method label.

iters_refit

(integer(1)) Number of refit iterations per resampling iteration.

---

Method compute()

Computes leave-out or leave-in feature importance. wvim_design_matrix(task$feature_names, "leave-out") corresponds to LOCO.

Usage

WVIM$compute(store_backends = TRUE)

Arguments

store_backends

(logical(1)) Passed to mlr3::resample to store backends in resample result. Required for some measures, but may increase memory footprint.

---

Method clone()

The objects of this class are cloneable with this method.

Usage

WVIM$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.