BSDesparseLasso: Bootstrap-calibrated Desparsified Lasso

In WaverlyWei/debiased.subgroup: Sharp Inference on Selected Subgroup in Observational Studies

Bootstrap-calibrated Desparsified Lasso

Description

This method first constructs the debiased estimator of β via the desparsified Lasso procedure. Then it calculates the calibration term \hat{b}_{max} =(1-n^{r-0.5})(\hat{β}_{max}-\hat{β}_{j,lasso}). Through B bootstrap iterations, it recalibrates the bootstrap statistic T_b. The bias-reduced estimate is computed as: \hat{b}_{max}-\frac{1}{B}∑_{b=1}^BT_b.

Usage

BSDesparseLasso(y, x, r = NULL, G = NULL, B = NULL, alpha = 0.95, fold = 3)

Arguments

y	response
x	design matrix
r	tuning parameter
G	subgroup indicator
B	bootstrap iterations
alpha	level of CI

Value

LowerBound	lower confidence bound
UpperBound	upper confidence bound
betaMax	bias-reduced maximum beta estimate
betaEst	debiased beta estimate for each subgroup
op	optimal tuning

WaverlyWei/debiased.subgroup documentation built on June 27, 2022, 7:22 p.m.