ciccr (version 0.3.0)

AAA_DML: Average Adjusted Association

Description

Averages the log odds ratio using prospective or retrospective high-dimensional logistic regression

Usage

AAA_DML(y, t, x, type = "pro", k = 10)

Value

An S3 object of type "ciccr". The object has the following elements.

est

a scalar estimate

se

standard error

Arguments

y

n-dimensional vector of binary outcomes

t

n-dimensional vector of binary treatments

x

n by d matrix of covariates

type

'pro' if the average is based on prospective regression; 'retro' if it is based on prospective regression (default = 'pro')

k

number of folds in k-fold partition (default = 10)

References

Jun, S.J. and Lee, S. (2020). Causal Inference under Outcome-Based Sampling with Monotonicity Assumptions. https://arxiv.org/abs/2004.08318.

Examples

# use the ACS dataset included in the package
  y = ciccr::ACS$topincome
  t = ciccr::ACS$baplus
  age = ciccr::ACS$age
  x = splines::bs(age, df=6) # b-splines for age

  results = AAA_DML(y, t, x, 'pro', k=2)