Condidence Intervals for Single Index

Description

Computes 95% bootstrap confidence intervals for the single index.

Usage

1
ci.index(y, U, V, B=100, use.parallel=TRUE, ...)

Arguments

y

vector, response

U

matrix, whose columns are covariates for factor one

V

matrix, whose columns are covariates for factor two

B

number of bootstap samples to take

use.parallel

if TRUE, exploits multiple cores by using foreach, doSNOW, etc

...

other arguments for siRSM, e.g., interaction.only=TRUE, trial=10

Value

A data.frame specifying the lower 2.5% and upper 97.5% confidence limits, the mean, and the standard error for each coordinate of the index.

Author(s)

Mu Zhu