2
$\begingroup$

How are powers of complex numbers defined? Suppose I have some number $z\in\mathbb{C}$. It makes sense that there are $n$ unique solutions to $w^n=z$. Where we define this power in terms of complex multiplication.

How is this extended to $z^w$ where $w\in \mathbb{C}$. I assume some sort of limiting process is required from natural to rational to real to complex? Is there a good resource for finding where this is defined?

Ted Shifrin
128k8 gold badges114 silver badges175 bronze badges
asked Jul 8, 2022 at 2:39
$\endgroup$
5
  • $\begingroup$ Here is a way using complex components. There is a formula at the end. $\endgroup$ Commented Jul 8, 2022 at 2:45
  • $\begingroup$ $w^{a+bi}=e^{(a+bi) \ln(w)},ドル where $\ln(w) = \ln|w| + i\arg(w)$ (taking the principal branch of the logarithm) $\endgroup$ Commented Jul 8, 2022 at 2:45
  • $\begingroup$ You can either use the binomial theorem, simplifying powers of $i$ or convert the complex numbers to polar coordinates which is really convenient for multiplication problems. $\endgroup$ Commented Jul 8, 2022 at 2:57
  • $\begingroup$ Any analysis book. For $n \in \mathbb N$ defining it as $b\cdot b\cdot...\cdot b$ is just as acceptable for complex numbers as it is for real numbers. But for $b^x; x\not \in \mathbb Z$ we need a different definition for it just as we needed a different definition for the reals. $\endgroup$ Commented Jul 9, 2022 at 5:27
  • $\begingroup$ Very carefully. No seriously, you use the complex exponential and logarithmic functions. $\endgroup$ Commented Jul 9, 2022 at 5:27

2 Answers 2

2
$\begingroup$

It's defined using Euler's Formula:

$$e^{i\theta} = \cos\theta + i \sin\theta$$

Or, adding a real part to the exponent:

$$e^{x+iy} = e^xe^{iy} = e^x(\cos y + i \sin y)$$

So if you have a complex number expressed in polar coordinates $z = r(\cos \theta + i \sin \theta)$, then:

$$\log z = \log r + i\theta$$

And you can then calculate $z^w = e^{w \log z}$ just like you can for real numbers.

The tricky part with complex logarithms is that because the trig functions are periodic, the choice of $\theta$ isn't unique.

$$\log z = \log r + i(\theta + 2\pi n), n \in \mathbb{Z}$$

So, do you use an angle in the interval $[0, 2\pi)$, $[-\pi, \pi)$, or some other 2ドル\pi$-wide interval? This choice, called a "branch cut", is a matter of context or convenience.

answered Jul 9, 2022 at 21:30
$\endgroup$
0
$\begingroup$

Fix a branch of natural log, the most common one being $\log : B = \mathbb{C} \setminus (-\infty, 0] \to \mathbb{C}$ that inverts $\exp : A = (-\infty, \infty) \times (-\pi, \pi) \to B$. Then for $z \in B$ and $w \in \mathbb{C}$, define $z^w = \exp(w \log(z))$. For details see sections 3 and 4 of https://mtaylor.web.unc.edu/notes/complex-analysis-course/

answered Jul 9, 2022 at 5:14
$\endgroup$

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.