cluster_sim: Calculate expected coverage of several alternative intervals...

In jean997/rccSims: Replicate simulations in "Rank conditional coverage and confidence intervals in high dimensional problems" by Jean Morrison and Noah Simon.

Description

This function can be used to generate simulations in section 3.1

Usage

cluster_sim(beta, Sigma, err.sd = 1, n.samp = 100, n.rep = 100,
  seed = NULL, parallel = FALSE)

Arguments

beta	List of regression coefficients. This should be a least with K elements where K is the number of blocks
Sigma	List of correaltion matrices, one for each block
which.sample	Indices of individuals who are sampled

Value

A list with elements:

COVERAGE

A 5 by p by n array giving the coverage at each simulation. COVERAGE[i, j, k] is either 0 or 1 indicating weather the interval formed using method i for rank j in simulation k covered its respective parameter

WIDTH

A 5 by p by n array giving the width of each interval in each simulation.

simnames

Name of each type of the 5 types of confidence intervals

jean997/rccSims documentation built on May 18, 2019, 11:45 p.m.