gnB_norm: ACP Estimation by Polynomial Approximation under Normality...

In koohyun-kwon/OptACI: Methods to Construct Optimal Average Coverage Confidence Intervals

Description

This is function is a version of gnB function where higher moments of the true mean vector are estimated assuming the true mean vector is normally distributed.

Usage

gnB_norm(
  chi,
  lam,
  xvec,
  Jn,
  Jn2,
  orcind = c("default", "m2", "theta"),
  m2,
  th_vec
)

Arguments

chi	a scalar half-length value χ.
lam	a vector of shrinkage factors λ.
xvec	normalized observed outcome vector, corresponding to z_i in the paper.
Jn	the order of polynomial approximation to be used for series estimation.
Jn2	the total order of polynomial approximation.
orcind	oracle specification to be used; when orcind = "m2", the value of true second moment of the true mean vector is provided in m2; when orcind = "th", the entire true mean vector is provided; the values other than "default" are used for simulations.
m2	the value of the true second moment of the true mean vector; used when orcind = "m2".
th_vec	the true mean vector; used when orcind = "th".

Details

Jn terms are estimated in the same way as gnB, while Jn2 - Jn terms are estimated under the normality assumption.

Value

vector of estimated average coverage probability values, corresponding to each value of λ in lam.

Examples

th_vec <- stats::rnorm(50)
xvec <- stats::rnorm(50, th_vec)
gnB_norm(1, c(0.3, 0.4), xvec, 2, 4)
gnB_norm(1, c(0.3, 0.4), xvec, 0, 4)
m2 <- mean(th_vec^2)
gnB_norm(1, c(0.3, 0.4), xvec, 2, 4, orcind = "m2", m2)
gnB_norm(1, c(0.3, 0.4), xvec, 2, 4, orcind = "th", NULL, th_vec)

koohyun-kwon/OptACI documentation built on Oct. 6, 2020, 8:09 a.m.