1
$\begingroup$

Determinants are used here as scaling factors for the areas of the polygons generated by connecting the individual roots of $z$.

  1. The roots of complex numbers are given by the formula $$z_k=\sqrt[n]{r}\Big(\cos\frac{\theta+2k\pi}{n}+i\sin\frac{\theta+2k\pi}{n}\Big)$$
  2. Two consecutive roots ($k=0$ and $k=1$) can be represented as $$\begin{bmatrix}\sqrt[n]{r}(\cos\frac{\theta}{n})&\sqrt[n]{r}(\cos\frac{\theta+2\pi}{n}) \\ \sqrt[n]{r}(\sin\frac{\theta}{n})&\sqrt[n]{r}(\sin\frac{\theta+2\pi}{n})\end{bmatrix}$$
  3. And the determinant of that, after some algebraic manipulation, is given by $$\text{det}=\sqrt[n]{r^2}\Big(\sin\frac{2\pi}{n}\Big)$$
  4. After excluding the cases where $\displaystyle\sin\frac{2\pi}{n}$ is $0$, I arrived with a final formula of $$r=\Big(\frac{\text{det}}{\sin\frac{2\pi}{n}}\Big)^{\frac{n}{2}}$$

However, the values generated by this formula are incorrect. Here is an example with $n=4$, $\text{det}=3$; the resultant area is $162$ instead of $6$).

The area $a$ for a polygon with $n=4$ and $r=1$ is $4(\frac{r^2}{2})=4(\frac{1}{2})=2$. The area $A$ for the same polygon is then expected to be $a\times\text{det}=2\times3=6$. Instead I get the following: $$r=\Big(\frac{\text{det}}{\sin\frac{2\pi}{n}}\Big)^{\frac{n}{2}}=\Big(\frac{3}{\sin\frac{2\pi}{4}}\Big)^{\frac{4}{2}}=9$$ $$A=4(\frac{r^2}{2})=4(\frac{9^2}{2})=162\neq6$$

At which step did I make an error? Are my assumptions correct in the first place?

Edit: My biggest concern right now is whether complex numbers are allowed to be represented as in step 2.

  • 0
    for what $r$ resultant area is 72 or 6 (correct)?2017-01-08
  • 0
    Do you know area formula for $n$ points in a plane?2017-01-08
  • 0
    @MyGlasses For $r$ given by the latest formula after plugging in $n=4$ and $\text{det}=3$.2017-01-08
  • 0
    First step is $r$ instead of $\sqrt[r]{r}$, the roots are on a circle with radius $r$ not $\sqrt[r]{r}$.2017-01-08
  • 0
    @MyGlasses This doesn't seem to work for different cases. I tried $n=6$, $\text{det}=\frac{1}{3}$ and got an area of $1$ instead of $\frac{\sqrt{3}}{2}$.2017-01-08

1 Answers 1

1

Let me write some words, from your statements. Maybe it works:

Suppose all vertices of a regular polygon are on a circle with radius $r$.

(1). The n-th roots of 1 as a complex number are given by the formula $$z_k=\cos\frac{\theta+2k\pi}{n}+i\sin\frac{\theta+2k\pi}{n}$$

(2). All of these roots lie on circle $|z|=r$, so following points are the vertices of regular polygon $$w_k=r\Big(\cos\frac{\theta+2k\pi}{n}+i\sin\frac{\theta+2k\pi}{n}\Big)$$

(3). Every two consecutive points (points $k$ and $k+1$) with origin as triangle third vertex, swap an area $A_k$ equal to $$A_k=\frac12\text{det}\begin{bmatrix}r\cos\frac{\theta+2k\pi}{n}&r\cos\frac{\theta+2(k+1)\pi}{n}\\r\sin\frac{\theta+2k\pi}{n}&r\sin\frac{\theta+2(k+1)\pi}{n}\end{bmatrix}$$

(4). And the determinant of that, after some algebraic manipulation, gives us $$A_k=\frac12r^2\sin\frac{2\pi}{n}$$

(5). After adding all areas, the total area of polygon is $$A=\sum_{k=0}^{n-1}A_k=\frac{nr^2}{2}\sin\frac{2\pi}{n}$$

(6). If our process were done in true way, We must have $$\lim_{n\to\infty}A=\pi r^2$$ is the area o circle with radius $r$, because as $n\to\infty$, the polygon will convert to circle.

For example with $n=4$, the area of square is $$A=\frac{4r^2}{2}\sin\frac{2\pi}{4}=2r^2$$

With $n=6$, the area is $$A=\frac{6r^2}{2}\sin\frac{2\pi}{6}=\frac{3\sqrt{3}}{2}r^2$$

enter image description here

  • 0
    Thank you, very insightful answer, though there are still two questions I have. Your answer gives the area of the polygon in terms of the radius (or vice versa) - is it still possible to find a formula for the radius in terms of the determinant (which was my initial approach)? Is there a formal, algebraical way to prove step 6?2017-01-08