Estimates the Average Marginal Effect (AME) or Marginal Effect at the Mean (MEM) for all paths in a model. This is especially useful for comparing coefficients across different families (e.g., Gaussian vs. Binomial) as it converts them to a common scale (expected change in the response).
Usage
marginal_effects(
fit,
at = NULL,
prob = 0.95,
samples = 1000,
multinomial_probabilities = FALSE
)Arguments
- fit
A because fit object.
- at
Character or list. If NULL (default), calculates Average Marginal Effects (AME). If "mean", calculates Marginal Effects at the Mean (MEM).
- prob
Numeric; probability mass for the credible interval (default 0.95).
- samples
Integer. Number of posterior samples to use (default 1000 for precision and consistency across plots).
- multinomial_probabilities
Logical. If TRUE, returns granular probability shifts for each category of multinomial (unordered) responses instead of a single expected value shift. Default FALSE.