Posted by: courtneyobrien | September 30, 2008

## Hafner-Sarnak-McCurley

This past weekend, I learned that my friend’s dad was the Sarnak in the Hafner-Sarnak-McCurley constant (http://mathworld.wolfram.com/Hafner-Sarnak-McCurleyConstant.html).  I had never heard of this constant before this weekend, but I thought you would all enjoy learning about it.

The Hafner-Sarnak-McCurley constant represents the probability that the determinants of two randomly chosen square integer matrices will be relatively prime. The probability depends on the matrix size, n, in accordance with the formula

$D(n)=\prod_{k=1}^\infty\left(1-\left(1-\prod_{j=1}^n(1-p_k^{-j})\right)^2\right)$,

where $p_k$ is the kth prime number. The limit of this expression as n approaches infinity is roughly 0.3532363719.

Professor Peter Sarnak has been called “one of the most influential mathematicians in the world.” (http://www.math.ufl.edu/dept_news_events/ramanujan/2007-8/). He received his undergraduate education at the University of Witwatersrand in South Africa and was awarded a Ph.D. from Stanford University in 1980.  He is a Professor at the Institute for Advanced Study in Princeton, NJ, and his current research focuses on the theory of zeta functions and automorphic forms with applications to number theory, combinatorics, and mathematical physics. Professor Sarnak was the recipient of The American Mathematical Society’s Levi L. Conant Prize in 2003 and the Frank Nelson Cole Prize in 2005.  In 2002, he was named a Member of the National Academy of Sciences and a Fellow of The Royal Society of London.

1. Thanks for telling us about this! I hadn’t seen the formula before. These sorts of formulas are very interesting. As we’ll see when we get to probability, it’s actually a very tricky thing to “choose a random integer.” It’s easy to describe, mathematically, what it means to choose a random element from a finite set. But how do you do it from an infinite set? It would seem that a random integer would have zero probability of being positive and less than $10$, since there are infinitely many larger alternatives. But the same argument would apply to being positive and less than $10^{10}$ or $10^{10^{10}}$! So maybe it is larger than zero? Ok, suppose the probability is larger than zero for it to be positive and less than $10$. Call the probability $p$. Then it’s got to have the same probability of being between 10 and 20, or of being between 20 and 30. But there are infinitely many such segments, each with probability $p$ — this adds up to way more than 1. And you can’t have something of probability more than 1! So doesn’t this mean every integer has probability zero of being chosen? How can you pick anything?!
You can see a derivation of the formula for $D(1)$ here.