Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

This conjecture is false. A nice counter example is given by David Speyer in the comments.

Yesterday I imagined up a nice little conjecture on real trigonometric polynomials. Much to my surprise, I couldn't find any such thing on googling the net. I don't know what are the implications or importance (if any) of this conjecture, if it were correct. I'd like to know whether the conjecture is true, and also a proof. In case its already treated in the literature, please give a pointer to a source.


Let $f_N(x)$ be a real trigonometric polynomial of degree $N$. Let $$P = \{x_p|x_p \in (0,2\pi) \land (f^'_N(x_p) = 0 \lor f^''_N(x_p) = 0)\}$$ Let $$B = \{(x_i,x_j)| x_i,x_j\in P \wedge x_i\ne x_j\}$$ Then $$\min\limits_{(x_i,x_j)\in B} |x_i-x_j| \ge \frac{\pi}{2N}$$

Note : $f^'_N(x)$,$f^''_N(x)$ are respectively, the first and second derivatives of $f_N(x)$

PS : Please let me know if you find this theorem interesting or useful in any part of Mathematics.

EDIT 1 : Adding a short summary of the statement in words as requested by Chris Taylor to avoid any possible confusion to the reader while parsing the symbols.

Let $P$ be the set of all points where either the first derivative or second derivative of a given real trigonometric polynomial of degree $N$ vanish. The conjecture is that the shortest of the distances between any two distinct points in $P$ is at least $\frac{\pi}{2N}$. I hope this summary in words is accurate. If not please let me know the mistakes.

share|cite|improve this question
One does not call it a theorem unless it had been proved. – Sasha May 23 '12 at 11:39
@Sasha : theorem changed to conjecture. – Rajesh Dachiraju May 23 '12 at 11:48
Am I confused? This seems very false. Consider $(4 - \epsilon) \cos \theta - \cos (2 \theta)$. This is a trig. polynomial of degree $2$ with critical points at $0$, $\pi$ and $\pm \cos^{-1} (1-\epsilon/4)$. By choosing $\epsilon$ sufficiently small (and positive) I can make the three critical points near $0$ as close as I want. – David Speyer May 23 '12 at 13:31
@David Speyer : Thank you very much for your nice counter example. I am really sorry for the goof up and for wasting your time. I'd like to delete this question after a while with your permission. – Rajesh Dachiraju May 24 '12 at 2:09
@David: if Rajesh decides to not delete his question, you may want to post that comment as an answer. – Willie Wong May 24 '12 at 9:03
up vote 4 down vote accepted

Adding this answer by popular request:

The degree $2$ trignometric polynomial $(4-\epsilon) \cos \theta - \cos (2 \theta)$ is a counter-example. It has critical points at $0$, $\pi$ and $\pm \cos^{-1} (1-\epsilon/4)$; by choosing $\epsilon$ small enough, I can make the zeroes near $0$ as close as I like.

Found by thinking first "there is usually no obstacle making roots of a polynomial close together" and then by thinking "well, where are the roots of $\frac{d}{d \theta} (a \cos (2 \theta) + b \cos (\theta))$?"

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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