Estimating Pi

Recently I’ve been working on some jackknife and bootstrapping problems.  While working on those projects I figured it would be a fun distraction to take the process and estimate pi.  I’m sure this problem has been tackled countless times but I have never bothered to try it using a Monte Carlo approach.  Here is the code that can be used to estimate pi.  The idea is to generate a 2 by 2 box and then draw a circle inside of it.  Then start randomly selecting points inside the box.  This example uses 10,000 iterations using a sample size of 500.  An interesting follow-up exercise would be to change the sample size, mean, and standard deviation to and see what it does to the jackknife variance.

nsims <- 10000; size <- 500; alpha <- .05; theta <- function(x){sd(x)} pi.estimate <- function(nsims,size,alpha){ out <-; call <-; pi.sim <- matrix(NA, nrow=nsims); for(i in 1:nsims){ x <- runif(size,-1,1); y <- runif(size,-1,1); radius <- -1; x.y <- cbind(x,y); d.origin <- sqrt(x^2+y^2); x.y.d <- cbind(x.y,d.origin); est.pi <- 4*length(x.y.d[,3][x.y.d[,3]<1])/ length(x.y.d[,3]) pi.sim[i] <- est.pi; } pi.mean <- mean(pi.sim); jk <- jackknife(c(pi.sim), theta); <- jk$*qnorm(1-alpha/2); pi.mean <- pi.mean; <- jk$; <-; <-; <-; return(list(pi.mean=pi.mean,,,, call=call)); } pi.estimate(nsims,size,alpha);  
Leave a comment


  1. Carl Witthoft

     /  December 12, 2012

    Seems like a lot of SLOC — by comparison, look at a recent post

    and even that can be tightened up with some vectorization.

  1. True Significance of a T Statistic | Statistical Research

Leave a Reply

Your email address will not be published. Required fields are marked *

4 − two =

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

%d bloggers like this: