Consider the following function, where below $r$ is a fixed real number, and $q$ is a positive integer

$$f(q) = \sum_{\substack{k = 1 \\ (k, q) = 1}}^{q} e^{\frac{2 \pi i k r}{q}}$$ I wished to determine the asymptotic behavior of this function as $q \rightarrow \infty$, for fixed $r$. When $r$ is an integer, I note that the above sum becomes the Ramanujan sum $c_q(r)$ which is $O(1)$ (since $|c_q(r)| \leq \sigma(r)$)

However for general $r$, it's not at all clear to me how to concretely bound this function, beyond the naive bound that $$f(q) \leq \phi(q)$$ I have a heuristic to think that $$f(q) = O(\sqrt{\phi(q)})$$ since we can think of $f(q)$ as a sum of $\phi(q)$ "random" exponential terms, which would have an expectation value of $O(\sqrt{\phi(q)})$. However, I am not sure how to make this argument rigorous.

  • $\begingroup$ When $r=q$ the sum equals the trivial bound $\phi(q)$, and when $r$ is a real number very close to $q$ then the sum will still be very close to $\phi(q)$. Also, $|c_q(r)| \le \sigma(r)$ doesn't imply $c_q(r) = O(1)$ (or, perhaps, it needs to be made clearer which parameter(s) the $O$-constant may depend on). $\endgroup$ – Greg Martin Jul 31 '20 at 17:00
  • $\begingroup$ Sorry, in the question above I was thinking of $r$ to be a fixed real number, and the asymptotic behavior as $q \rightarrow \infty$. In this context, if $r$ is an integer, then $|c_q(r)| \leq \sigma(r)$ does imply $c_q(r) = O(1)$ since then $\sigma(r)$ is a constant. I have updated the question to try to clarify this, if it helps. $\endgroup$ – chaiKaram Jul 31 '20 at 17:15

If $r$ is not an integer, we can rewrite the sum using geometric series: \begin{align*} \sum_{\substack{0\le k<q \\ (k, q) = 1}} e^{2 \pi i k r/q} &= \sum_{k=0}^{q-1} e^{2 \pi i k r/q} \sum_{d\mid(k,q)} \mu(d) \\ &= \sum_{d\mid q} \mu(d) \sum_{j=0}^{q/d-1} e^{2 \pi i j r d/q} \\ &= \sum_{d\mid q} \mu(d) \frac{e^{2 \pi i r}-1}{e^{2 \pi i r d/q}-1} \\ &= (e^{2 \pi i r}-1) \sum_{d\mid q} \frac{\mu(d)}{e^{2 \pi i r d/q}-1} = (e^{2 \pi i r}-1) \sum_{d\mid q} \frac{\mu(q/d)}{e^{2 \pi i r/d}-1}. \end{align*} When $d\ge r$, say, we have $e^{2 \pi i r/d}-1 = 2 \pi i \frac rd + O_r(\frac1{d^2}) = 2 \pi i \frac rd (1+O_r(\frac1d))$. Therefore* \begin{align*} \sum_{\substack{0\le k<q \\ (k, q) = 1}} e^{2 \pi i k r/q} &= (e^{2 \pi i r}-1) \sum_{d\mid q} \mu(\tfrac qd) \frac d{2 \pi i r} (1+O_r(\tfrac1d)) \\ &= \frac{e^{2 \pi i r}-1}{2\pi ir} \sum_{d\mid q} \big( d \mu(\tfrac qd) + O_r(1) \big) \\ &= \frac{e^{2 \pi i r}-1}{2\pi ir} \phi(q) + O_r(\tau(q)), \end{align*} where $\tau(q)$ is the number of divisors of $q$. So, surprisingly, the asymptotic behavior really is like $\phi(q)$. (Numerical experiments confirm this asymptotic formula.)

*I cheated with the terms $d<r$, but I think that this just gives a $O_r(1)$ error.

  • $\begingroup$ As far as I can tell, the argument up to the third line ($= \sum_{d | q} \mu(d) \sum \ldots$ should hold regardless of whether or not $r$ is an integer. But this leads to concerning conclusions. As a concrete example, let $q$ be prime, and let $r$ be a multiple of $q$. Then we would expect that the sum should evaluate to $\phi(q)$, but the third line from your argument reduces to $\frac{e^{2\pi i r} - 1}{e^{2 \pi i r / q} - 1} - 1$ which reduces to 0/0 - 1. Also could you please explain how to go from line 1-2 of your argument (changing order of summation)? $\endgroup$ – chaiKaram Jul 31 '20 at 20:17
  • $\begingroup$ Agreed, it is very surprising from the $\sum_{(k,q)=1}$ formula, but obvious once put in the Möbius inversion form. I think we should add it to en.wikipedia.org/wiki/… $\endgroup$ – reuns Jul 31 '20 at 21:04
  • $\begingroup$ @reuns Sorry could you please elaborate on this point? $\endgroup$ – chaiKaram Jul 31 '20 at 21:19
  • $\begingroup$ @chaiKaram One cannot use this geometric series formula when the common ratio equals $1$ (of course, in that case then the sum is just the number of terms). So you've correctiy identified the place where the proof depends upon $r$ being irrational, and where a different calculation would be used to recover the usual information on the Ramanujan sum. $\endgroup$ – Greg Martin Aug 1 '20 at 4:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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