summary.bdt: Summarization of the results of a call to the bdt function

In Yan2020729/bdt1: Bootstrap Diagnostic Tool

Description

These functions are methods for class bdt, summary.bdt objects

Usage

  ## S3 method for class 'bdt'
summary(object, ...)
  ## S3 method for class 'summary.bdt'
print(x, ...)

Arguments

object	an object of class bdt for summary functions.
x	an object of class summary.bdt for print functions.
...	currently ignored.

Details

print.summary.bdt, called indirectly by entering the command summary(result) (where result has class bdt).

Value

true_Effect	the true ate of simulated bootstrap datasets
gbound	the bounding for propensity scores used defined
gGLM	'GLM' or 'SL' method user defined to estimate ps
bias.aiptw	the Min., 1st Qu., Median, Mean, 3rd Qu., Max value of AIPTW bias ate over M replications
bias.tmle	the Min., 1st Qu., Median, Mean, 3rd Qu., Max value of TMLE bias ate over M replications
sd.aiptw	the average standard error of AIPTW estimatorover M replications
sd.tmle	the average standard error of TMLE over M replications
cov.aiptw	the coverage rate of AIPTW estimator over M replications
cov.tmle	the coverage rate of TMLE over M replications
coverage	coverage rates of TMLE and AIPTW over M replications
sd	influence function based standard errors of TMLE and AIPTW over M replications
bias	a data frame of summaries of three bias over M replications
ps	a data frame of summaries of pooled probabilities over M replications

See Also

bdt

Examples

# Example
  set.seed(1250)
  n <- 1000
  sigma <- matrix(c(2, 1, 1, 1), ncol = 2)
  X <- matrix(rnorm(n*2), ncol = nrow(sigma)) 
  X <- X + matrix(rep(c(.5, 1),each = n), byrow = FALSE, ncol = 2)
  I1 <- rnorm(n,mean = 1, sd = 2)
  I2 <- rnorm(n,mean = 1, sd = 1.9)
  P1 <- rnorm(n,mean = 1, sd = 1.5)
  W <- data.frame(X, I1, I2, P1)
  colnames(W) <- c("W1", "W2", "I1", "I2",  "P1")
  A <- rbinom(n, 1, plogis(0.2 + W[,"W1"] + 0.3*W[,"I1"] + W[,"W1"]*W[,"I1"]
  - 0.2*(W[,"W2"] + W[,"I2"])^2 ))
  Y <-  1 + A + W[,"W1"] + 2*W[,"W2"] + 0.5*(W[,"W1"] + W[,"P1"])^2 + rnorm(n)

  result <- bdt(Y, A, W, outcome_type = "continuous", gGLM = TRUE, gform = "A~W2+W1+I1", M = 60)
  summary(result)

Yan2020729/bdt1 documentation built on March 24, 2021, 8:58 p.m.