Skip to contents

Quantile-Based Variable Importance for SPQR / SPQRX

Source: R/SPQRX.R
eval_explain_VI.Rd

Computes variable importance across quantile levels using Quantile Accumulated Local Effects (QALE). Importance is defined as the standard deviation of the ALE function across feature intervals for each quantile level.

Usage

eval_explain_VI(
  model,
  x,
  tau = seq(0.1, 0.9, 0.1),
  var.indexs = c(1, 2),
  type = "QF"
)

Arguments

model

A fitted model object returned by fit_spqrx().

x

Matrix or data frame of covariates used for evaluation.

tau

Numeric vector of quantile levels in (0,1).

var.indexs

Integer vector specifying the indices of variables for which importance is computed.

type

Provides the output of model to be explain in this case either Quantile function or XI parameter for extremes

Value

A matrix of variable importance values. Rows correspond to variables and columns correspond to quantile levels.