est_ipcw: Estimate Inverse Probability of Censoring Weights
In txshift: Efficient Estimation of the Causal Effects of Stochastic Interventions

Description Usage Arguments Details Value

Estimate Inverse Probability of Censoring Weights

1	est_ipcw(V, Delta, fit_type = c("sl", "glm"), sl_learners = NULL)

`V`	A `numeric` vector, `matrix`, `data.frame` or similar object giving the observed values of the covariates known to potentially inform the censoring mechanism.
`Delta`	A `numeric` vector giving observed values of the indicator function corresponding to the censoring mechanism.
`fit_type`	A `character` indicating whether to perform the fit using GLMs or a Super Learner. If use of Super Learner is desired, then the argument `sl_learners` must be provided.
`sl_learners`	An `Lrnr_sl` object, a Super Learner instantiated externally using sl3.

Compute inverse probability of censoring weights for the two-phase sampling mechanism. These inverse weights are based on the probability of appearing in the second-phase sample based on variables measured on all participants.

A list containing a numeric vector corresponding to the inverse probability of censoring weights required for computing an IPCW-TMLE and numeric vector of the estimated missingness mechanism. Formally, the former is nothing more than term class as computed using standard logistic regression.

txshift documentation built on Oct. 23, 2020, 8:27 p.m.