Skip to contents

Function to calculate AIPTW

Usage

calc_aiptw(
  a,
  A,
  A_name,
  W,
  Y,
  I_Y,
  d_pred,
  mean_dZ,
  outcome_model,
  treatment_model,
  missingness_model,
  ps_trunc_level,
  idx_sub
)

Arguments

a

value of treatment to set for all observations

A

vector of observed treatments

A_name

name of treatment variable

W

dataframe of covariates

Y

vector of observed outcome variable

I_Y

indicator that Y is missing (I_Y = 1 if missing, I_Y = 0 if observed)

d_pred

vector of treatment decisions (d_pred = 1 if recommend treatment under rule)

mean_dZ

proportion recommended treatment under rule

outcome_model

outcome Nuisance model

treatment_model

treatment Nuisance model

missingness_model

missingness Nuisance model

ps_trunc_level

threshold to truncate propensity score

idx_sub

indices of observations that are recommended treatment (d_pred == 1)

Value

(1) a - value of treatment (2) aiptw - vector aiptw estimators (3) plug_in_est - vector of plug-in estimators (4) augmentation - vector of augmentation terms (5) var_aug - variance of augmentation term