# Find a maximum of complex function

I am trying to find a simple method that does not use the tools of advanced differential calculus to find following maximum, whose existence is justified by the compactness of the close ball $\Delta$ of $\mathbb C$ and continuity of the function $f:z \mapsto |z^3 + 2iz|$ from $\mathbb C$ to $\mathbb C$

$$\large { \displaystyle \max_{z \in {\mathbb C},|z| \leq 1} |z^3 +2i z |}$$ Since : $$(\forall z \in \Delta) \quad f(z) \leq 3$$ is obtained using triangular inequality, we can yet try to find some $z_0 \in {\mathbb C}$ such that $f(z_0)=3$

Does anybody have an idea?

Thanks.

• That's a much better problem statement. It tells us that you have worked quite a bit on the problem, and already know a lot about it. Maybe with my hint you can polish it off. Jun 9, 2012 at 13:04
• @Steve D: I did not want to talk too much for my English (even beginners). @ Paul Slevin,Gerry Myerson, draks : Thank you for your reception and your instructions for using this nice site by which I edited my post. (and excuse my english) Jun 9, 2012 at 13:07
• A very, very simple method is asking Wolfram. Jun 9, 2012 at 13:36

Hint: when do you get equality in the triangle inequality?

• @ Gerry Myerson : When the two terms are directely proportional. I well see this . Jun 9, 2012 at 13:13
• @ Gerry Myereson: Thank's for hint. Using it, i find that $z_0=\frac{\sqrt 2}{2} (1+i)$ gives $f(z_0)=3$ Jun 9, 2012 at 13:27
• Well done! ${}$ Jun 9, 2012 at 23:56

If you've begun a study of complex functions, you may have seen the Maximum Modulus Principle. Since $z^3 + 2iz$ is a polynomial and entire (analytic in the complex plane), the maximum of $|z^3 + 2iz|$ you seek must occur on the boundary of the unit disk. Gerry's Hint then quickly points you in the right direction!

• Ah, drat, you type faster +1
– huon
Jun 9, 2012 at 13:28

Hint: $z^3 + 2iz$ is differentiable on $\mathbb{C}$ (i.e. holomorphic) so you can apply the maximum modulus principle, and deduce that the maximum of $f$ lies on the boundary of $\Delta$, which has a simple parametrisation, so you can use standard techniques from real one-variable calculus to find the maximum.

(A slightly different approach to Gerry Myerson's, much more complicated in this case but also far more general.)

• @ hardmath and dbaupp : Thank you for the idea of maximum module principle.I was looking for a suitable method for beginning students. It turns out that the Pricipe gives the idea already! Thank you! Jun 9, 2012 at 13:37