PoSI-package: Valid Post-Selection Inference for Linear LS Regression

In PoSI: Valid Post-Selection Inference for Linear LS Regression

Description

In linear LS regression, calculate for a given regressor matrix the multiplier K of coefficient standard errors such that the confidence intervals [b - K*SE(b), b + K*SE(b)] have a guaranteed coverage probability for all coefficient estimates b in any submodels after performing arbitrary model selection.

Details

Package:	PoSI
Type:	Package
Version:	1.1
Date:	2020-10-24
License:	GPL-3

Author(s)

Andreas Buja and Kai Zhang

Maintainers: Andreas Buja <[e-mail]>, Kai Zhang <[e-mail]> and Wan Zhang <[e-mail]>

References

“Valid Post-Selection Inference,” Berk, R., Brown, L., Buja, A., Zhang, K., Zhao, L., The Annals of Statistics, 41 (2), 802–837~(2013).

See Also

lm, link{model.matrix}

Examples

data(Boston, package="MASS")
summary(PoSI(Boston[,-14]))

PoSI documentation built on Jan. 13, 2021, 5:10 a.m.