In bayesball/TeachBayes: Teaching Bayesian Inference

knitr::opts_chunk$set(fig.width=6, 
                      fig.height=4, fig.path='Figs/',
                      echo=TRUE, warning=FALSE, message=FALSE)

The TeachBayes package has several functions to facilitate working with a discrete prior for two proportions.

library(TeachBayes)

Uniform Prior

Start with a uniform prior on (p1, p2), where each proportion takes on values .05, .15, ..., .95.

prior <- testing_prior(.05, .95, 10,
                       uniform=TRUE)

Construct a graph of this distribution.

draw_two_p(prior)

This finds the probability distribution of the difference in proportions p1 - p2, and plots the distribution.

(diff_dist <- two_p_summarize(prior))
prob_plot(diff_dist)

Collect some data from two binomial samples.

y1n1 <- c(10, 20)
y2n2 <- c(8, 24)

Update (find posterior):

post <- two_p_update(prior, y1n1, y2n2)

Graph and summarize:

draw_two_p(post)

prob_plot(two_p_summarize(post))

Testing Prior

prior <- testing_prior(.05, .95, 10, pequal=0.5)

Construct a graph of this distribution and summarize.

draw_two_p(prior)

prob_plot(two_p_summarize(prior))

Collect some data from two binomial samples.

y1n1 <- c(10, 20)
y2n2 <- c(8, 24)

Update (find posterior):

post <- two_p_update(prior, y1n1, y2n2)

Graph and summarize:

draw_two_p(post)

prob_plot(two_p_summarize(post))

bayesball/TeachBayes documentation built on Jan. 5, 2020, 1:47 a.m.