post_int: Posterior intervals

In ggdistribute: A 'ggplot2' Extension for Plotting Unimodal Distributions

Description

Returns cutoff points from a posterior distribution

Usage

post_int(x, mid = c("median", "mean", "mode"), int = c("hdi", "ci"),
  widths = c(0.5, 0.95), adj = 1.5, rope = NULL, warn = FALSE)

Arguments

x	Vector of numeric values. Typically a posterior sample.
mid	Central tendency estimator. Defaults to "median". Other options include "mean" and "mode".
int	interval type, either "hdi" or "ci"
widths	interval widths
adj	Bandwidth adjustment used only with the "mode" estimator. See dmode.
rope	Region of practical equivalence. Check how much of the distribution is within rope value.
warn	Turn off warning for flat intervals found (multiple possible values)

Value

data.table

Examples

x <- rpois(5000, 15)
ints <- post_int(x, warn = FALSE)
hist(x, br=50)
abline(v=ints$c, col="cyan")
abline(v=ints[, c("l.wide", "r.wide")], col="magenta")

post_int(x, "median", warn = FALSE)
post_int(x, "mean", warn = FALSE)
post_int(x, "mode", adj=2, rope = c(14, 16), warn = FALSE)

ggdistribute documentation built on May 2, 2019, 10:25 a.m.