Let \(f:\mathbb{C} \rightarrow \mathbb{C}\) be continuous and assume that \(f(z)=f(2z)\) for all \(z \in \mathbb{C}\). Show that \(f\) is constant.
If you're looking for a proof here, I *think* you can try this by contradiction. Assume \(f(z)\) is non-constant, perhaps by saying it's monotonically increasing or decreasing for some interval. Is it possible then for \(f(z)=f(2z)\) to be true? Just an idea, I don't know if it'd take you anywhere.
I should take that back. I'm not even sure monotonicity applies to complex-valued functions...
Yeah I'm not too sure about that either
If you had the fact that \(f\) was holomorphic, I might be able to do something, but I don't see anything (right now) if it's just continuous.
I think I have an idea. Since it's continuous everywhere, it's continuous at 0. So let \(\{z_i\}_{i=1}^\infty\) be a sequence of complex numbers approaching 0 along a straight line so that \(z_i=2z_{i+1}\). So since \(f\) is continuous, \[\lim_{n\to\infty}f(z_n)=f(0).\]But from the way we defined our sequence, \(\lim_{n\to\infty}f(z_n)=f(z_1)\). So \(f(z_1)=f(0)\). Now, notice that we can choose \(z_1\) arbitrarily in \(\mathbb{C}\) and define the same sequence, and it will always approach \(0\). But that means that \(f(0)=f(z)\) for all \(z\in\mathbb{C}\), so the function must be constant.
how do you say that \(\lim_{n\rightarrow \infty} f(z_n)=f(z_1)\)
That's from the condition that \(f(z)=f(2z)\). The way we defined the sequence, we know that \(f(z_i)=f(z_j)\) for all \(i,j\in\mathbb{Z}^+\). So if we say \(f(z_1)=L\), then we could rewrite the limit as \(\lim_{n\to\infty}L\). But \(L\) is a constant, so the limit is clearly equal to \(L=f(z_1)=f(0)\).
Gosh I feel really dumb but I'm not sure I quite understand :( Maybe because I have trouble understand how complex sequences work? Honestly our prof spent like 30 mins on that... can we imagine that they work similarly as in real-valued functions? Also, if you constructed this particular sequences, how does it imply it's true in all cases?
Convergent sequences in the complex numbers are slightly different from sequence in the real numbers, but the particular sequence I constructed is almost exactly like a sequence in the real numbers since every point in the sequence lies on the same line. As for continuous functions, I'm using the definition: A function \(f\) is continuous at a point \(c\) if for any sequence \(\{x_1\}_{n=1}^\infty\) that converges to \(c\), then the sequence \(\{f(x_n)\}_{n=1}^\infty\) converges to \(f(c)\) (so the limit as \(n\to\infty\) of \(f(x_n)\) will be \(f(c)\). It's true in all cases, since if I choose ANY \(z_1\in\mathbb{C}\), then define the sequence \(z_{i+1}=z_i/2\), the sequence \(\{z_i\}_{i=1}^\infty\) will approach the origin. Since the function is continuous, we use the above definition to see that \(f(z_i)\) approaches \(f(0)\) as \(i\to\infty\). Finally, from the condition that \(f(z)=f(2z)\), or equivalently \(f(z)=f(z/2)\), we notice that \(f(z_i)\) is constant for all \(i\in\mathbb{Z}^+\). So the sequence must approach that same constant. But we already know it approaches \(f(0)\), so \(f(0)\) must be that constant. So \(f(0)=f(z_1)\) for all \(z_1\in\mathbb{C}\), so \(f\) must be constant.
ohhh oh my I had to read this MANY times but it has finally sunk in! I get it now. Thank you sooooo much!!!
You're welcome. Out of curiosity, what is this for? Complex analysis? Real analysis? Or something different?
It is for complex analysis
Nice. You'll be surprised at how much nicer things work in the complex plane rather than the real line.
Hehe good to know
Join our real-time social learning platform and learn together with your friends!