mid.terms: Extract Terms from MID Models

Description

mid.terms() extracts term labels from a fitted MID model or derivative objects. Its primary strength is the ability to filter terms based on their type (main effects vs. interactions) or their associated variable names.

Usage

mid.terms(
  object,
  main.effects = TRUE,
  interactions = TRUE,
  require = NULL,
  remove = NULL,
  ...
)

Value

mid.terms() returns a character vector of the selected term labels.

Arguments

object: a "mid" object or another object that contains model terms. Can be a "mid.importance", "mid.conditional", or "mid.breakdown" object.
main.effects: logical. If FALSE, the main effect terms are excluded.
interactions: logical. If FALSE, the interactions terms are excluded.
require: a character vector of variable names. Only terms related to at least one of these variables are returned.
remove: a character vector of variable names. Terms related to any of these variables are excluded.
...: aliases are supported for convenience: "me" for main.effects and "ie" for interactions.

Details

A "term" in a MID model refers to either a main effect (e.g., "Wind") or an interaction effect (e.g., "Wind:Temp"). This function provides a flexible way to select a subset of these terms, which is useful for plotting, summarizing, or other downstream analyses.

Examples

Run this code

data(airquality, package = "datasets")
mid <- interpret(Ozone ~ .^2, airquality, lambda = 1)

# Get only main effect terms
mid.terms(mid, interactions = FALSE)

# Get terms related to "Wind" or "Temp"
mid.terms(mid, require = c("Wind", "Temp"))

# Get terms related to "Wind" or "Temp", but exclude any with "Day"
mid.terms(mid, require = c("Wind", "Temp"), remove = "Day")

# Get the predicted contributions of only the terms associated with "Wind"
terms_wind <- mid.terms(mid, require = "Wind")
predict(mid, airquality[1:3,], terms = terms_wind, type = "terms")