estimate_Q {tmleLite}R Documentation

Initial Estimation of Q portion of the Likelihood

Description

An internal function called by the tmle function to obtain an initial estimate of theQ portion of the likelihood based on user-supplied values for Q(A,W) and counterfactual outcomes Q(1,W),Q(0,W), or a user-supplied regression formula, or based on a model data-adaptively selected by the DSA algorithm. If DSA is not available and Q is not user-supplied, estimation is based on a main terms regression using glm.

Usage

estimate_Q(Q, DSAargs, Y, A, W, Delta, family, wts, id)

Arguments

Q a 3-column matrix (Q(A,W), Q(1,W), Q(0,W)), a user-specified regression formula, or NULL. If Q is NULL DSA is used to select a model. If DSA is unavailable estimation is based on a main terms regression using glm
DSAargs an optional list specifying values for all arguments to the DSA algorithm
Y continuous or binary outcome variable
A binary treatment indicator, 1 - treatment, 0 - control
W vector, matrix, or dataframe containing baseline covariates
Delta indicator of missing outcome or treatment assignment. 1 - observed, 0 - missing
family family specification for regressions, generally ‘gaussian’ for continuous oucomes, ‘binomial’ for binary outcomes
wts optional weights on observations. Defaults to unweighted
id id identifying repeated measures

Value

Q a matrix with one row per observation, 3 columns containing the initial estimate of Q(A,W) and predicted values for counterfactual outcomes Q(1,W), Q(0,W)
coef coefficients for each term in working model used for initial estimation of Q
type type of estimation procedure: NULL-user supplied,‘glm’-generalized linear regression using user-specified or main terms model, ‘DSA’-model selected by DSA

Author(s)

Susan Gruber

See Also

tmle, estimate_g, DSA


[Package tmleLite version 1.0-1 Index]