savage_dickey: Calculate Bayes Factor using Savage-Dickey density ratio.
In LudvigOlsen/LRO.utilities: Utility functions and addins

Description Usage Arguments Details Value Author(s) Examples

View source: R/savage_dickey.R

Creates plot of prior and posterior distributions and calculates the Bayes factor at the point of interest (Q).

savage_dickey(
  post,
  prior,
  Q,
  xlab = "Parameter values",
  ylab = "Density",
  plot = TRUE,
  print_plot = FALSE
)

`post`	Samples from posterior distribution (Numeric)
`prior`	Samples from prior distribution (Numeric)
`Q`	Point on x-axis for calculating Bayes factor.
`xlab`	Label for x-axis
`ylab`	Label for y-axis
`plot`	Create density plot (Logical)
`print_plot`	(Logical)

Bayes factors are calculated using the polspline package.

BF10 <- dlogspline(Q, logspline(post)) / dlogspline(Q,logspline(prior))

BF01 <- dlogspline(Q, logspline(prior)) / dlogspline(Q,logspline(post))

List with ggplot2 object (optional), BF10 and BF01, Maximum a posteriori and priori at point Q

Benjamin Hugh Zachariae

Ludvig Renbo Olsen

# Sampling from two different gaussian distributions
prior <- rnorm(1000, mean=0, sd=1)
posterior <- rnorm(1000, mean=2, sd=2)

# Calculating BF and generating plot of the given distributions
s_d <- savage_dickey(posterior, prior, Q = 0, plot = TRUE)