I need some help with the following problem:

Let $0\neq z\in\mathbb{C}$, prove $\lvert\frac{z}{\lvert z\rvert}-1\lvert\leq\rvert Arg(z)\rvert$

where we take the argument to be in $\left(-\frac{\pi}{2},\frac{\pi}{2}\right]$.

What I did:

I say that it is clear that it sufficient to prove that if $z'$ is on the unit circle then $|z'-1|\leq|Arg(\alpha z)|$ where $0\neq\alpha\in\mathbb{R}$.

Since if $0\neq\alpha\in\mathbb{R}$ then $Arg(\alpha z)=Arg(z)$ we need to prove that $|z'-1|\leq|Arg(z)|$.

Now, let $a,b\in\mathbb{R}$s.t $z'=a+bi$ then $\sqrt{a^{2}+b^{2}}=1$ and we need to prove that $\sqrt{(a-1)^{2}+b^{2}}\leq|Arg(z')|$.

This is where I am stuck, can someone please provide a hint ? am I even on the right track ?


$z = r e^{i\theta}$. Then you want to prove that $$\left \vert e^{i \theta} - 1\right \vert \leq \theta$$ $$e^{i \theta} - 1 = \cos(\theta) + i \sin(\theta) - 1 = -2\sin^2(\theta/2) + 2i \sin(\theta/2) \cos(\theta/2) = 2i \sin(\theta/2) e^{i \theta/2}$$

Hence, $$\vert e^{i \theta} - 1 \vert = 2 \left \vert \sin(\theta/2) \right \vert \leq 2 \times \left \vert \theta/2 \right \vert= \left \vert \theta \right \vert$$

  • $\begingroup$ Thank you for your answer (and for leaving me some of the work!). but as I understand it $\theta$ is not unique in this representation. so I'm a bit confused...everything I write in the LHS can ba changed from $\theta$ to $\theta+2\pi k$ , is it OK ? $\endgroup$ – Belgi Oct 30 '12 at 22:06
  • $\begingroup$ @Belgi It actually doesn't matter since $\vert \sin(\phi) \vert \leq \vert \phi \vert$ for all $\phi$ $\endgroup$ – user17762 Oct 30 '12 at 22:09
  • $\begingroup$ Maybe the $\theta /2$ in the exponent in the last step in the big equation should have a minus sign ? $\endgroup$ – Belgi Oct 30 '12 at 22:18
  • $\begingroup$ @Belgi: I think it is fine, since $$-2\sin^2(\theta/2) + 2i \sin(\theta/2) \cos(\theta/2) = 2i^2\sin^2(\theta/2) + 2i \sin(\theta/2) \cos(\theta/2)$$ $\endgroup$ – user17762 Oct 30 '12 at 22:22
  • $\begingroup$ Thank you very much for your help! I do have to ask this - it seems that the proof is OK for every $\theta$ in $arg(z)$ but the LHS is fixed and the LHS can be as large as I want...so there must be something that I don't undetstand... $\endgroup$ – Belgi Oct 30 '12 at 22:29

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.