universalStep: universalStep

In benkeser/drinf: Computes doubly-robust estimators and doubly-robust confidence intervals

universalStep

Description

Take one step in the practical construction of the universally least favorable submodel.

Usage

universalStep(L2, Qn, gn, Qnr.gnr, PnDQ2, PnDQ1, PnDg1, PnDg0, normPnD,
  dx = 0.005, tolg, tolQ, ...)

Arguments

Qn	A list of current estimates of Q2n and Q1n
gn	A list of current estimates of g0n and g1n
Qnr.gnr	A list of current estimates of reduced dim. regressions
normPnD	A numeric giving the norm of PnD
dx	A numeric step size
tolg	A numeric indicating the truncation level for conditional treatment probabilities.
tolQ	A numeric indicating the truncation level for transformed outcome regressions.
PnD	A matrix of dimension 3x1 that is the empirical mean of the relevant portions of the influence function (one for original EIF, one for extra G terms, one for extra Q terms).

benkeser/drinf documentation built on June 10, 2025, 11:25 p.m.