tlmixture: Targeted learning for exposure mixtures

In ck37/tlmixture: Data-adaptive Creation Of Exposure (Treatment) Mixtures Using Targeted Learning

Description

This is our main function.

Usage

tlmixture(
  data,
  outcome,
  exposures,
  quantiles_mixtures = 3L,
  quantiles_exposures = 4L,
  folds_cvtmle = 2L,
  folds_sl = 2L,
  estimator_outcome = c("SL.mean", "SL.glmnet"),
  estimator_propensity = estimator_outcome,
  cluster_exposures = FALSE,
  mixture_fn = mixture_glm,
  refit_mixtures = TRUE,
  verbose = FALSE
)

Arguments

data	Data frame with outcome, exposure, and adjustment variables.
outcome	Name of the outcome variable.
exposures	A vector of exposure names, or (not yet supported) a list where each element is a vector of pre-clustered exposures.
quantiles_mixtures	Number of quantiles to use for discretizing mixture (default 3 - low, medium, high).
quantiles_exposures	Number of quantiles to use for discretizing continuous exposures (default 4).
folds_cvtmle	Number of CV-TMLE folds (default 2).
folds_sl	Number of SL folds during outcome and propensity estimation.
estimator_outcome	SuperLearner library for outcome estimation.
estimator_propensity	SuperLearner library for propensity estimation.
cluster_exposures	Whether to automatically cluster a vector of exposures into sub-groups (default FALSE; TRUE not yet supported).
mixture_fn	Current options: mixture_glm, mixture_pls, or mixture_sl
refit_mixtures	After CV-TMEL, refit mixture functions to full dataset.
verbose	If TRUE, display more detailed info during execution.

ck37/tlmixture documentation built on Dec. 19, 2021, 5:13 p.m.