cmdSuperLearner.onebin: cmdSuperLearner for a specific number of bins

In tedwestling/ctsCausal: Methods for Causal Inference with Continuous Exposures

cmdSuperLearner for a specific number of bins

Description

This function estimates the conditional mixed density using a given number of bins b.

Usage

cmdSuperLearner.onebin(A, W, b, SL.library, verbose, validRows)

Arguments

A	n x 1 numeric vector of exposure values.
W	n x p data.frame of covariate values to condition upon.
b	Integer number of bins >= 2.
SL.library	Library to use for bin-specific probabilities.
verbose	Logical indicating whether to print progress reports to the command line.
validRows	List of rows in each CV fold.

Value

Returns a named list with the following elements:

bins	List of length b containing the sets used for each bin.
bin.fits	List of length b containing the estimated SuperLearner objects for each bin.
a.ecdf	Empirical CDF of the exposure.
SL.bin.probs	SuperLearner conditional probabilities of being in each bin.
SL.densities	SuperLearner conditional standardized mixed density correspondint to each bin.
cv.library.densities	Cross-validated library conditional standardized mixed density corresponding to each bin.
library.densities	Library conditional standardized mixed density corresponding to each bin fit on the full data.
alg.names	Algorithm names.

Examples

# Define parameters
n <- 300
W <- data.frame(matrix(rnorm(3 * n), ncol = 3))
Z <- rbinom(n, size = 1, prob = 1/(1 + exp(2-W[,1] + W[,2])))
A <- (1-Z) * rnorm(n, mean = W[,2] - W[,3], sd = abs(1 + W[,1]))
validRows <- cmdCVFolds(n = n, cvControl = list(V = 10, shuffle=TRUE, validRows = NULL))
bin.fit <- cmdSuperLearner.onebin(A, W, b = 2, SL.library = c("SL.mean", "SL.glm", "SL.gam", "SL.earth"), verbose=TRUE, validRows = validRows)

tedwestling/ctsCausal documentation built on Dec. 7, 2022, 3:33 p.m.