1
$\begingroup$

" Find the points $z \in \mathbb{C}$ at which the function $g(z) = \cos(\bar{z})$ satisfies the Cauchy-Riemann equations. "

So I've written $g(z) = g(x,y) = \cos(x-iy)=\cos(x)\cosh(y)+i\sin(x)\sinh(y)$. So that $u(x,y)=\cos(x)\cosh(y)$ and $v(x,y)=\sin(x)\sinh(y)$.

From the first C-R equation I get $-\sin(x)\cosh(y) = \sin(x)\cosh(y)$ which is true when $\sin(x)\cosh(y)=0$.

Now here I thought this implied $\sin(x) = 0$ or $\cosh(y) = 0$. However the solutions give only $\sin(x) = 0$, why?

Also, when doing the second equation, you get $\cos(x)\sinh(y) = 0$, but this time the solutions split the case into $\cos(x)=0$ and $\sinh(y)=0$ as I did above.

Why this?

EDIT

Fair enough, I've found this document explaining why it is not zero, as it was said in the comments. Hyperbolic Functions

$\endgroup$
  • 3
    $\begingroup$ For real $y$ we have $\cosh(y) \geq \frac 12e^y$, so it doesn't have real roots. $\endgroup$ – Tim B. Jan 18 '17 at 17:15
  • $\begingroup$ @LeBtz More precisely, $\cosh(y)\ge 1$. $\endgroup$ – Mark Viola Jan 18 '17 at 17:17
  • $\begingroup$ So $\cosh(y) =0$ only for complex arguments? Is it $z = \frac{\pi}{2}i +n\pi$ ? $\endgroup$ – Euler_Salter Jan 18 '17 at 17:19
  • $\begingroup$ Yes for the first part, no for the second. The roots are given by $i(\frac{\pi}{2}+n\pi), ~n\in\mathbb Z$. $\endgroup$ – Tim B. Jan 18 '17 at 17:21
  • $\begingroup$ Oh yeah it makes sense that also the period is complex, thank you $\endgroup$ – Euler_Salter Jan 18 '17 at 17:22
4
$\begingroup$

Now here I thought this implied $\sin(x) = 0$ or $\cosh(y) = 0$. However the solutions give only $\sin(x) = 0$, why?

$$ \cosh x = \frac{e^x + e^{-x}}2$$

So if $\cosh x = 0$, then $e^x + e^{-x} = 0$. But this can't happen because both $e^x$ and $e^{-x}$ are positive, no matter what real value $x$ has. And the sum of two positive numbers can never be zero. Therefore $\cosh x$ can never be zero for real values of $x$.

Also, when doing the second equation, you get $\cos(x)\sinh(y) = 0$, but this time the solutions split the case into $\cos(x)=0$ and $\sinh(y)=0$ as I did above.

Why this?

$$ \sinh x = \frac{e^x - e^{-x}}2$$

So if $\sinh x = 0$, then $e^x - e^{-x} = 0$. And this can happen because the difference of two positive numbers can be zero. And the solution is \begin{align*} e^x - e^{-x} &= 0\\ e^x &= \frac1{e^x}\\ e^{2x} &= 1\\ 2x &= \ln 1\\ x &= 0 \end{align*}

$\endgroup$
  • $\begingroup$ thank you, that's a perfect explanation! $\endgroup$ – Euler_Salter Jan 18 '17 at 17:23
1
$\begingroup$

Hint: $$\cosh^2(y)=1+\sinh^2(y)$$

What happens if $\cosh(y)=0$?

$\endgroup$
  • $\begingroup$ do we have $y = arcsinh(i)$ ? $\endgroup$ – Euler_Salter Jan 18 '17 at 17:26
  • 1
    $\begingroup$ @Euler Yes, but when you write complex numbers in the form $x+iy$ it is always assumed/understood that $y$ is real and hence $\sinh(y)$ is real. $\endgroup$ – N. S. Jan 18 '17 at 18:08
1
$\begingroup$

The $\cosh$ of a real number is never zero.

$\endgroup$

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.