submarg: Return subgroup margins

In anniejw6/testingtools: Helper functions to run and analyze experiments

Description

Calculates subgroup margins by running on subset of data. Doesn't matter for OLS but does matter, e.g., for logit.

Usage

submarg(
  mod,
  subgrp_var,
  treat_var = "treatment",
  mod_df = mod$data,
  weight = mod$prior.weights,
  ...
)

Arguments

mod	glm
subgrp_var	Subgroup variable, character. The subgroup variable must be a categorical variable
treat_var	Treatment variable, character
mod_df	Dataframe used to estimate model, defaults to mod$data
weight	Weight parameter to pass to 'modmarg::marg', defaults to mod$prior.weights
...	Other parameters passed to modmarg::marg

Examples

library(modmarg)
data(margex)
mod <- glm(outcome ~ treatment * yc + distance, data = margex, family = 'binomial')
submarg(mod, 'yc', 'treatment')

anniejw6/testingtools documentation built on Jan. 26, 2021, 6:41 a.m.