Cdboot_multi: Fast computation of internal double bootstrap calculations

In perccal: Implementing Double Bootstrap Linear Regression Confidence Intervals Using the 'perc-cal' Method

Description

This is the workhorse function of the package, speeding up computations within double bootstrap routine.

Usage

Cdboot_multi(xxyy, lgridlo, lgridhi, B, B2, G)

Arguments

xxyy	(n by p+1) matrix for X (design matrix) and response vector y.
lgridlo	Lower quantile values of double bootstrap distribution to obtain.
lgridhi	Upper quantile values of double bootstrap distribution to obtain.
B	Number of 1st stage bootstrap samples.
B2	Number of 2nd stage double bootstrap samples.
G	Calculate quantile-based empirical coverage at this many grid points

Value

theta_hat_boot first-level bootstrap estimates of all slope coefficients

theta_qtl_lgrid_lo (p+1 by B by G by 1) matrix for lower quantiles at all grid points for all predictors over all bootstrap samples.

theta_qtl_lgrid_hi (p+1 by B by G by 1) matrix for upper quantiles at all grid points for all predictors over all bootstrap samples.

perccal documentation built on May 2, 2019, 1:06 p.m.