Skip to contents

Simple Bootstrap Confidence Intervals

Source: R/bootCI.R
bootCI.Rd

Convenience wrapper for calculating bootstrap confidence intervals for univariate and bivariate statistics.

Usage

bootCI(
  x,
  y = NULL,
  FUN,
  conf.level = 0.95,
  sides = c("two.sided", "left", "right"),
  R = 999,
  ...
)

Arguments

x

a (non-empty) numeric vector of data values.

y

NULL (default) or a vector with compatible dimensions to x, when a bivariate statistic is used.

FUN

the function to be used.

conf.level

confidence level of the interval.

sides

a character string specifying the side of the confidence interval, must be one of "two.sided" (default), "left" or "right". You can specify just the initial letter. "left" would be analogue to a hypothesis of "greater" in a t.test.

R

number of bootstrap replicates, a single positive whole number.

...

further arguments. The bootstrap options are taken out first, as in the other interval functions of the package: type, the interval type passed to boot::boot.ci(), one of "bca" (default), "perc", "basic", "norm" or "stud", and parallel and ncpus, passed to boot::boot(). Everything else is passed to FUN.

Value

A named numeric vector with three elements:

est

the estimate calculated by FUN.

lci

lower confidence interval bound.

uci

upper confidence interval bound.

Details

type, parallel and ncpus therefore cannot reach FUN through the dots. A statistic that has an argument of one of these names - the type of quantile(), say - is wrapped: FUN = function(z) quantile(z, 0.9, type = 6).

"stud" needs a variance estimate for every replicate, which a general FUN does not deliver; boot::boot.ci() then returns no such interval and bootCI() stops with a message saying so.

Examples


set.seed(1984)
bootCI(mtcars$mpg, FUN=mean, na.rm=TRUE)
#>      est      lci      uci 
#> 20.09062 18.12561 22.31055 
bootCI(mtcars$mpg, FUN=mean, trim=0.1, na.rm=TRUE, type="basic")
#>      est      lci      uci 
#> 19.69615 17.46923 21.78846 

# bootCI(mtcars$mpg, FUN=DescToolsX::skewX, na.rm=TRUE, type="basic")

# bootCI(Pizza$operator, Pizza$area, FUN=cramerV)

spearman <- function(x,y) cor(x, y, method="spearman", use="p")
bootCI(mtcars$mpg, mtcars$hp, FUN=spearman)
#>        est        lci        uci 
#> -0.8946646 -0.9592579 -0.7950177