Counterfactual marginal effects from fitted distributional parameters

Usage

marginal_effects(
  object,
  newdata,
  variable,
  values = NULL,
  parameter = "mu",
  reference = NULL,
  se.fit = FALSE,
  level = 0.95,
  vcov_method = "analytical",
  ...
)

Arguments

object: A fitted gamlss.longitudinal object.
newdata: Data used as the counterfactual baseline.
variable: Single variable to vary.
values: Values to assign to variable. Defaults to observed factor levels for factors/characters or the 25th, 50th, and 75th percentiles for numeric variables.
parameter: Distributional parameter to summarize, usually "mu".
reference: Optional reference value. Defaults to the first value.
se.fit: Logical; when TRUE and parameter = "mu", attach approximate delta-method standard errors for response-scale averages.
level: Confidence level used when se.fit = TRUE.
vcov_method: Variance-covariance method passed to vcov.gamlss.longitudinal().
...: Additional arguments passed to predict.gamlss.longitudinal().

Value

A data frame with average fitted parameter values and contrasts.