#
FAQ: Probability that gcd(*n*,*m*) = 1?

Here is a frequently asked question:

While playing with programs to determine primes and relative primes, I stumbled over an interesting (at least to me) fact. While the probability of a random number being prime decreases as the range of possible random numbers increases (Prime Number Theorem), the probability of two random numbers being relatively prime is 60.8% Is this something that is either well known by or trivially obvious to prime number gurus?

That there should be such a constant is "obvious", finding its value takes
more work. In 1849 Dirichlet showed that the probability is exactly 6/π^{2}
arguing roughly as follows.

Suppose you pick two random numbers less than *n*, then

- [
*n*/2]^{2}pairs are both divisible by 2. - [
*n*/3]^{2}pairs are both divisible by 3. - [
*n*/5]^{2}pairs are both divisible by 5. - ...

(Here [*x*] is the greatest integer less than or equal to *x*,
usually called the *floor function*.) So the number of relatively prime
pairs less than or equal to *n* is (by the inclusion/exclusion principle):

n^{2}- Σ([n/p]^{2}) + Σ([n/pq]^{2}) - Σ([n/pqr]^{2}) + ...

where the sums are taken over the primes *p, q, r,*... less than *n*.
Letting μ(*x*) be the möbius function this is

Σ(μ(so the desired constant is the limit ask)[n/k]^{2}) (sum over positive integersk)

*n*goes to infinity of this sum divided by

*n*

^{2}, or

Σ(μ(k)/But this series times the sum of the reciprocals of the squares is one, so the sum of this series, the desired limit, is 6/k^{2}) (sum over positive integersk).

^{2}. This number is approximately

60.7927101854026628663276779258365833426152648033479percent. ;-)

Printed from the PrimePages <primes.utm.edu> © Chris Caldwell.