# Binomial distribution, two-sample test of proportions.
# Example data are from the 6-yr aspirin/heart attack study of 22,000
# physicians (NEJM vol 318(4)).
x=c(104,189)
n=c(11000,11000)
prop.test(x,n,p=NULL,alternative="two.sided",conf.level=.95,correct=TRUE)
# options:  alternative="two.sided" "greater" "less"
# p=NULL sets null hypothesis of p1-p2=0;
# p=c( , ) set null as vector of fixed known proportions.