1 prior <- c(working =0.99,broken =0.01)
 2 likelihood <- rbind(working = c(good=0.99,bad=0.01),broken =c(good=0.6,bad=0.4))
 3 data <- c("bad","bad","bad","bad")
 4 
 5 bayes <- function(prior, likelihood, data)
 6 {
 7     posterior <- matrix(0, nrow=length(data), ncol=length(prior))
 8     dimnames(posterior) <- list(data, names(prior))
 9 
10     initial_prior = prior
11     for(i in 1:length(data))
12     {
13         posterior[i, ] <- 
14             prior*likelihood[ , data[i]]/
15             sum(prior * likelihood[ , data[i]])
16 
17         prior <- posterior[i , ]
18     }
19 
20     return(rbind(initial_prior,posterior))
21 }
22 
23 
24 bayes(prior,likelihood,data)
25 matplot(bayes(prior,likelihood,data),t=\'b\',lty =1,pch =20,col =c(3,2))