TRICEP_glm_boot: Bootstrap TRICEP/BICEP for general composite tests using a...

In hoangtt1989/BICEP: Bi-linear algorithm for composite empirical likelihood

Description

Bootstrap TRICEP/BICEP for general composite tests using a proximal operator

Usage

TRICEP_glm_boot(X, y = NULL, prox_fun = prox_inequality,
  prox_fun_arg = list(test_idx = 1, test_val = 0), family = c("gaussian",
  "binomial", "poisson", "mean"), ..., beta_init = NULL, B = 200,
  prob = 0.95, parallel = T, seed = 123)

Arguments

X	a design matrix with observations in rows and variables in columns.
y	a vector of the responses. Default is NULL in case family = "mean" (test a vector mean).
prox_fun	a proximal operator. This should be an R object.
prox_fun_arg	a list of arguments passed to prox_fun.
family	either the GLM family or "mean" for a vector mean.
...	additional arguments passed to TRICEP_glm.
beta_init	an initial guess for the beta vector.
B	the number of boostrap replications.
prob	the coverage probability.
parallel	a boolean indicating whether each replicate should be parallelized. A backend for foreach must be registered if this is T.
seed	an integer for setting the seed.

Value

A list with the bootstrapped test statistic, the p value, and the replicates.

hoangtt1989/BICEP documentation built on May 28, 2019, 3:37 p.m.