SuperLearner-stacked AIPW (not yet ported).

The python version stacks RF + ridge + OLS/logit + mean (and optionally xgboost) via cross-validated convex weights. The R port would require pulling in SuperLearner or hand-rolling the stacked-cross-fit construction.

Usage

morie_otis_aipw_superlearner(...)

Arguments

...

Arguments mirroring morie_otis_aipw_ate().

Value

Stops with a NotYetPorted message; for the time being, call morie_otis_aipw_ate() with the default cross-fit OLS+logit stack.

Examples

if (FALSE) { # \dontrun{
  morie_otis_aipw_superlearner(df, treatment = "d", outcome = "y",
                                covariates = "x")
} # }