hdpbci: Bootstrap confidence interval of the qth quantile

In GRousselet/rogme: Robust Graphical Methods For Group Comparisons

Bootstrap confidence interval of the qth quantile

Description

Compute a percentile bootstrap confidence interval for the qth quantile via the Harrell–Davis estimator. Appears to be best method when there are tied values.

Usage

hdpbci(x, q = 0.5, alpha = 0.05, nboot = 2000, nv = 0)

Arguments

x	A vector of continuous observations.
q	A quantile between 0 and 1 - default = 0.5, meaning a confidence interval for the median is computed.
alpha	Alpha level - default = 0.05, such that a 95 interval is computed.
nboot	Number of bootstrap samples - default = 2000.
nv	Null value when computing a p-value.

Value

A list with variables: q estimate, ci, n, p.value

Note

Adaptation of qcipb from Rallfun-v33.txt. See https://github.com/nicebread/WRS/

See Also

hd

GRousselet/rogme documentation built on Nov. 12, 2022, 4:38 a.m.