add_weighting: Treatment Effect Estimation Using Propensity Scores

In ygeunkim/propensityml: Propensity score weighting using machine learning methods

Description

estimates treatment effect based on ps estimation (e.g. inverse probability treatment weighting)

Usage

add_weighting(
  data,
  treatment,
  trt_indicator = 1,
  object = NULL,
  formula = NULL,
  method = c("logit", "rf", "cart", "SVM"),
  mc_col = NULL,
  sc_col = NULL,
  parallel = FALSE,
  ...
)

Arguments

data	A data frame to be used
treatment	Treatment variable name
trt_indicator	Value that indicates the unit is treated
object	A propmod object if already fitted.
formula	If not, write a formula to be fitted. Remember that you don't have to worry about group variable. .SD do exclude by.
method	Estimating methods "logit" - ps_glm "rf" - ps_rf "cart" - ps_cart "SVM" - ps_svm
mc_col	Indicator column name for MC simulation if exists
sc_col	Indicator column name for various scenarios if exists
parallel	parallelize some operation
...	Additional arguments of fitting functions

Details

This functions add columns by

\frac{trt_i}{\hat{e}_i} - \frac{1- trt_i}{1 - \hat{e}_i}

and

trt_i - (1 - trt_i) \frac{\hat{e}_i}{1 - \hat{e}_i}

References

Lee, B. K., Lessler, J., & Stuart, E. A. (2010). Improving propensity score weighting using machine learning. Statistics in Medicine. Statistics in Medicine, 29(3), 337-346.

Pirracchio, R., Petersen, M. L., & Laan, M. van der. (2015). Improving Propensity Score Estimators’ Robustness to Model Misspecification Using Super Learner. American Journal of Epidemiology, 181(2), 108–119. https://doi.org/10.1093/aje/kwu253

See Also

add_propensity

ygeunkim/propensityml documentation built on Jan. 1, 2021, 1:44 p.m.