3
$\begingroup$

This is my first question. It appeared while solving a research problem in cryptography. I am computer science student, so I apologize for lack of mathematical rigor in this question. Thanks for any help.

Consider the Riemann zeta function at s = +1. It diverges, but the expression for the function is $\zeta(1) = \lim_{n \rightarrow \infty} \sum_{i = 1}^{n} \frac{1}{i}$ , the truncated sum of which are the $n$-th harmonic number, $\mathcal{H}(n)$.

The question is, how about the expression $\zeta(1) = \lim_{n \rightarrow \infty} \prod_{\textrm{primes } p_i \leq n} \frac{1}{1-p_i^{-1}}$. is the value of the truncated product $\mathcal{H}(n)$ too?

My simulations for large values of $n$ tells me that it is some function of $\log n$ (for example comparing the ratio of the function for $n$ and $n^2$ and $n^3$ etc) How do we prove this?

In summary, what is the value of $\prod_{\textrm{primes } p_i \leq n} \frac{1}{1-p_i^{-1}}$? Thanks

$\endgroup$

3 Answers 3

11
$\begingroup$

Formula (8) on this page gives the result

$$\prod_{p \le n} \frac1{1-p^{-1}} = e^\gamma \log n \,(1 + o(1)).$$

$\endgroup$
4
$\begingroup$

Note that, asymptotically, $\mathcal{H}(n)\simeq \ln(n) + \gamma + \frac{1}{2n} + O(n^{-2})$. In other words, both expressions diverge like $\ln(n)$ but not exactly in the same way.

$\endgroup$
3
$\begingroup$

Check out

J. Barkley Rosser and Lowell Schoenfeld Approximate formulas for some functions of prime numbers, especially Theorem 7.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.