Skip to contents

quantile accumulated local effects (ALE)

Source: R/QALE.R
QALE.Rd

Computes the quantile ALEs of a SPQR class object. The function plots the ALE main effects across tau using line plots for a single covariate, and the ALE interaction effects across tau using contour plots for two covariates .

Usage

QALE(
  object,
  var.index,
  tau = seq(0.1, 0.9, 0.1),
  n.bins = 40,
  ci.level = 0,
  getAll = FALSE,
  pred.fun = NULL
)

Arguments

object

An object of class SPQR.

var.index

a numeric scalar or length-two vector of indices of the covariates for which the ALEs will be calculated. When length(var.index) = 1, the function computes the main effect for X[,var.index]. When length(var.index) = 2, the function computes the interaction effect between X[,var.index[1]] and X[,var.index[2]].

tau

The quantiles of interest.

n.bins

the maximum number of intervals into which the covariate range is divided when calculating the ALEs. The actual number of intervals depends on the number of unique values in X[,var.index]. When length(var.index) = 2, n.bins is applied to both covariates.

ci.level

The credible level for computing the pointwise credible intervals for ALE when length(var.index) = 1. The default is 0 indicating no credible intervals should be computed.

getAll

If TRUE and length(var.index) = 1, extracts all posterior samples of ALE.

pred.fun

A function that will be used instead of predict.SPQR() for computing predicted quantiles given covariates. This can be useful when the user wants to compare the QALE calculated using SPQR to that using other quantile regression models, or maybe that using the true model in a simulation study.

Value

x

If length(var.index) = 1, a (n.bins+1)-length vector specifying the ordered predictor values at which the ALE plot function is calculated. These are the break points for the n.bins intervals into which the predictor range is divided, plus the lower boundary of the first interval and the upper boundary of the last interval. If length(var.index) = 2, a list of two such vectors, the first containing the X[,var.index[1]] values and the second containing the X[,var.index[2]] values at which the ALE plot function is calculated.

ALE

If length(var.index) = 1, a (n.bins+1) by length(tau) matrix of predicted ALE values for every combination of x and tau. If length(var.index) = 2, a 3-dimensional array of predicted ALE values. Each slice corresponds to a quantile. Within each slice, the rows correspond to X[,var.index[1]] and the columns correspond to X[,var.index[2]].

Examples

# \donttest{
set.seed(919)
n <- 200
X <- runif(n,0,2)
Y <- rnorm(n,X^2,0.3+X/2)
control <- list(iter = 200, warmup = 150, thin = 1)
fit <- SPQR(X=X, Y=Y, n.knots=12, n.hidden=3, method="MCMC",
            control=control, normalize=TRUE, verbose = FALSE)

## compute quantile ALE main effect of X at tau = 0.2,0.5,0.8
ale <- QALE(fit, var.index=1, tau=c(0.2,0.5,0.8))
# }