demo/Chapter.2.3.R

In bayesball/LearnBayes: Learning Bayesian Inference

####################################
# Section 2.3 Using a Discrete Prior
####################################

 library(LearnBayes)

 p = seq(0.05, 0.95, by = 0.1)
 prior = c(1, 5.2, 8, 7.2, 4.6, 2.1, 0.7, 0.1, 0, 0)
 prior = prior/sum(prior)
 plot(p, prior, type = "h", ylab="Prior Probability")

S=readline(prompt="Type  <Return>   to continue : ")

 data = c(11, 16)
 post = pdisc(p, prior, data)
 round(cbind(p, prior, post),2)

 library(lattice)
 PRIOR=data.frame("prior",p,prior)
 POST=data.frame("posterior",p,post)
 names(PRIOR)=c("Type","P","Probability")
 names(POST)=c("Type","P","Probability")
 data=rbind(PRIOR,POST)

if (.Platform$OS.type == "unix") x11() else windows()
 xyplot(Probability~P|Type,data=data,layout=c(1,2),
        type="h",lwd=3,col="black")