ale_profile: ALE profile

Description

Internal function used by light_profile to calculate ALE profiles.

Usage

ale_profile(
  x,
  v,
  breaks = NULL,
  n_bins = 11,
  cut_type = c("equal", "quantile"),
  counts = TRUE,
  counts_weighted = FALSE,
  pred = NULL,
  evaluate_at = NULL,
  indices = NULL,
  n_max = 1000,
  seed = NULL,
  two_sided = FALSE,
  calibrate = TRUE,
  ...
)

Value

A tibble containing results.

Arguments

x: An object of class flashlight.
v: The variable to be profiled.
breaks: Cut breaks for a numeric v. Only used if no evaluate_at is specified.
n_bins: Maxmium number of unique values to evaluate for numeric v. Only used if no evaluate_at is specified.
cut_type: For the default "equal", bins of equal width are created for v by pretty. Choose "quantile" to create quantile bins.
counts: Should counts be added?
counts_weighted: If counts is TRUE: Should counts be weighted by the case weights? If TRUE, the sum of w is returned by group.
pred: Optional vector with predictions.
evaluate_at: Vector with values of v used to evaluate the profile. Only relevant for type = "partial dependence".
indices: A vector of row numbers to consider.
n_max: Maximum number of ICE profiles to calculate within interval (not within data).
seed: Integer random seed passed to light_ice.
two_sided: Standard ALE profiles are calculated via left derivatives. Set to TRUE if two-sided derivatives should be calculated. Only works for continuous v. More specifically: Usually, local effects at value x are calculated using points between x-e and x. Set ale_two_sided = TRUE to use points between x-e/2 and x+e/2.
calibrate: Should values be calibrated based on average preditions? Default is TRUE.
...: Other arguments passed to this function (currently unused).