I come cross a problem when I read a book of complex analysis:

If three complex number $a,b,c$ satisfy the relation of


Prove that: these numbers must be three vertices of an equilateral triangle on the complex plane.

if $a,b,c$ are real numbers, we have $a=b=c$. but I’m not sure how to prove it with complex number. The hint I got is:

Calculate $((b-a)w+(b-c))*((b-a)w^2+(b-c))$, where $w$ is nonreal cube root of unity.

  • 1
    $\begingroup$ What do you mean by “it’s obvious if $a,b,c$ are real numbers”? I’m not sure how the question can be made sense of in that case $\endgroup$ – User May 15 '18 at 23:06
  • $\begingroup$ If they're all real numbers, then we have $a=b=c$, which are vertices of a degenerate equilateral triangle? $\endgroup$ – G Tony Jacobs May 15 '18 at 23:08
  • $\begingroup$ @GTonyJacobs, yes - a degenerate, equilateral triangle on the real axis to be precise. $\endgroup$ – dezdichado May 15 '18 at 23:09
  • $\begingroup$ Of course it's on the real axis; we said they're all real numbers. It's still not clear that's what OP meant. $\endgroup$ – G Tony Jacobs May 15 '18 at 23:10
  • $\begingroup$ By obvious I mean, if a,b,c are real, we can show that (a-b)^2+(a-c)^2+(b-c)^2=0,from which we know a=b=c. Sorry for the confusion. Question rephrased.@User $\endgroup$ – Sizhe May 15 '18 at 23:11

What is the geometric meaning of

$((b-a)\omega + (c-a))((b-a)\omega^2 + (c-a))$

If we subtract $a$ from each point we translate them such that $a$ is on the origin.

$(b-a)\omega$ rotates $(b-a)$ 120 degrees clockwise $(b-a)\omega^2$rotates $(b-a)$ 120 degrees counter-clockwise.

If $a,b,c$ form an equilateral triangle, one of these will be exactly the negative of $(c-a)$

In which case one of the factors of $((b-a)\omega + (c-a))((b-a)\omega^2 + (c-a))$ equals $0.$

enter image description here

  • $\begingroup$ very intuitive explanation ! $\endgroup$ – G Cab May 15 '18 at 23:46
  • $\begingroup$ Fantastic, thank you! $\endgroup$ – Sizhe May 15 '18 at 23:58

The hint is one way to write it. Another is $$ (x + y \omega + z \omega^2)(x + y \omega^2 + z \omega) = x^2 + y^2 + z^2 - yz-zx-xy $$

  • $\begingroup$ Thanks for your reply. Would you mind to explain more? $\endgroup$ – Sizhe May 15 '18 at 23:15
  • $\begingroup$ @Sizhe one of the factors must be zero. In either case, you need to relate the pairwise differences $x-y, y-z, z-x$ and show $|x-y| = |y-z| = |z-x|.$ I suppose this is why they gave the hint in that style. $\endgroup$ – Will Jagy May 15 '18 at 23:19
  • $\begingroup$ @Sizhe now that I think about it, the requirement really is that $y-z$ is equal to one of these four: $(x-y)\omega, (x-y)\omega^2, -(x-y)\omega, -(x-y)\omega^2.$ You need to draw some pictures, I think two out of four are suitable. $\endgroup$ – Will Jagy May 15 '18 at 23:23
  • $\begingroup$ Thank you, now I think I get it.@WillJagy $\endgroup$ – Sizhe May 15 '18 at 23:59

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.