mcEst: Compute the Monte Carlo estimate

In podTockom/tstmle01: Targeted Maximum Likelihood Estimation of Causal Effects of Interventions on a Single Binary Time Series

Usage

mcEst(fit, start = 1, node = "W", t, Anode, intervention = NULL,
  lag = 0, MC = 100, returnMC = FALSE, returnMC_full = FALSE,
  clevCov = FALSE, set = NULL, full = FALSE, update = FALSE)

Arguments

fit	fit object obtained by initEst.
start	Start generating Monte Carlo estimates from this point earliest time point is 1. For the clever covariate, this is also our s.
node	Start generating Monte Carlo estimates from O_i node W, A or Y.
t	Outcome time point of interest. It must be greater than the intervention node A.
Anode	Intervention node.
intervention	Specify \item lagThis is an user impossed dependency lag necessary for the calculation of the clever covariate in the targeting step. It refers to past. Default is 0, meaning that it starts right after the node in question. Also, note that lag should be equal or smaller than start. \item MCHow many Monte Carlo samples should be generated. \item returnMCIf TRUE, returns all MC draws up until outcome time. \item returnMC_fullIf TRUE, returns all full MC draws. \item clevCovIf TRUE, this MC is used for the calculation of the clever covariate. Instead of observed data, it used intervened \item setSet the s node to either 1 or 0. Used for the clever covariate calculation. \item fullTRUE if full time-series should be used. \item updateTRUE, use updated fits to get Monte Carlo draws. An object of class tstmle01. estimate Mean of the outcome at time t under specified intervention, or no intervention. outcome Outcome at time t for each MC interation. intervention Intervention specified. MC How many Monte Carlo samples should be generated. Anode Intervention node as a function of O_i. MCdata If returnMC is TRUE, returns a data.frame with MC time-series. set Returns what the s node was set to. This function computes the Monte Carlo estimate of the expected value of the parameter of interest under specified intervention and under the estimated mechanism.

podTockom/tstmle01 documentation built on May 14, 2019, 2:03 a.m.