Biostat CDA - Codes for Chapter 01

  1. Binomial distribution: normal approximation 
     ## not very good approximation
 pi=0.1; n=30
 plot(0:n, dbinom(0:n,n,pi), type="h", lwd=2, ylab="Density")
 curve(dnorm(x,n*pi,sqrt(n*pi*(1-pi))), col=2, lty=2, add=TRUE)
 ## very good approximation
 pi=0.5; n=30
 plot(0:n, dbinom(0:n,n,pi), type="h", lwd=2, ylab="Density")
 curve(dnorm(x,n*pi,sqrt(n*pi*(1-pi))), col=2, lty=2, add=TRUE)
  2. Poisson distribution: normal approximation 
     mu = 16; rng = mu+4*sqrt(mu)
 plot(0:rng, dpois(0:rng,mu), type="h", lwd=2, ylab="Density")
 curve(dnorm(x,mu,sqrt(mu)), col=2, lty=2, add=TRUE)