TEE.BCa: Compute the bias-corrected accelerated bootstrap confidence...

In TEEReg: Trimmed Elemental Estimation for Linear Models

Description

This function create the bias-corrected accelerated bootstrap confidence intervals based on the trimmed elemental regressions.

Usage

TEE.BCa(formula,data,offset=NULL,p.trimmed=NULL,p.subsample=1,method ="tee",
est.TEE,conf.level,n.boot)

Arguments

formula	define a symbolic description of the model to be fitted.
data	specify the dataset used for performing regression analysis. It must be formatted in data frame.
offset	specify a known component to be included in the linear predictor during fitting. This argument should be either NULL or a numeric vector with length equal to the number of observations.
p.trimmed	define the trimming proportion of elemental subsets. This should be either NULL or a numeric value between 0 and 1. When method = "tee" is specified, a numeric value must be provided.
p.subsample	this is the proportion of subsampling with values between 0 and 1. The default value is 1, meaning that calculations are based on the full data.
method	two options are supported: "ols" stands for ordinary least squares and "tee" stands for trimmed elemental estimation.
est.TEE	this is for trimmed elemental regression estimates.
conf.level	the confidence level.
n.boot	number of bootstrap samples.

Details

For discussions about bootstrap confidence intervals and coverage probabilities of regression parameter estimates using trimmed elemental estimation, see Hall and Mayo (2008).

Value

call	call to the function.
ci	bias-corrected accelerated bootstrap confidence interval estimates for the regression parameters.

Author(s)

Wei Jiang and Matthew S. Mayo

References

M. Hall and M. S. Mayo. Bootstrap confidence intervals and coverage probabilities of regression parameter estimates using trimmed elemental estimation. Journal of Modern Applied Statistical Methods, 7: 514-525, 2008.

Examples

data(telephone)
fit <- TEE(formula=Y~X,data=telephone,p.trimmed=0.5,p.subsample=0.5,method="tee")
TEE.BCa(formula=Y~X,data=telephone,est.TEE=fit$coefficients,p.trimmed=0.5,p.subsample=0.5,
method="tee",conf.level=0.05,n.boot=20)

Example output

$call
TEE.BCa(formula = Y ~ X, data = telephone, p.trimmed = 0.5, p.subsample = 0.5, 
    method = "tee", est.TEE = fit$coefficients, conf.level = 0.05, 
    n.boot = 20)

$ci
            estimates(TEE) Lower limit Upper limit
(Intercept)     -59.838079  -71.485059  -50.786948
X                 1.236983    1.077209    1.435576

TEEReg documentation built on May 2, 2019, 3:38 p.m.