fit_least_fav_submodel: Least Favorable Submodel

View source: R/utils_ATE.R

fit_least_fav_submodelR Documentation

Least Favorable Submodel

Description

Least Favorable Submodel

Usage

fit_least_fav_submodel(h_aw, data, y, qbar_aw, qbar_1w)

Arguments

h_aw

numeric vector of values for the clever covariate

data

dataframe that includes treatement/exposure

y

Outcome variable used in the fluctuation model

qbar_aw

Initial predictions for Y|A,W - A being observed A

qbar_1w

Initial predictions for Y|A = 1,W - A being deterministically set to 1

Details

Does flucutation of initial outcome estimates using a least favorable submodel in the TMLE update step

Value

A numeric vector of the same length as vals, where the returned values are bounded to machine precision. This is intended to avoid numerical instability issues.


blind-contours/CVtreeMLE documentation built on June 22, 2024, 8:53 p.m.