1
$\begingroup$

The following is found using a combination of: (a) a polygon with an infinite number of sides is a circle, (b) the perimeter of that polygon is the circumference of the circle that it becomes (of course), (c) the sine theorem, and (d) the ratio between the circumference of a circle and the diameter is $\pi$. $$\pi=\lim_{n\to\infty} n\sin \left(\frac{180}{n}\right)^o$$ I have been trying to rewrite the above expression without $sin$ in it explicitly. So, I researched other expressions for $sin$, and found this: $$\sin (x) = x-\frac{x^3}{3!}+\frac{x^5}{5!}-\frac{x^7}{7!}+\frac{x^9}{9!}-\frac{x^{11}}{11!}+\cdots$$ Attempting to rewrite this Taylor series in Sigma notation yielded this: $$\sin(x)=\sum_{n=1}^{\infty} \frac{x^{2n-1}}{(2n-1)!}(-1)^{n-1}$$ The issue for me is that I have an angle measure in degrees, but $x$ above is in radians. I would convert the degrees to radians and be done with it just like that, except that I feel that solving for $\pi$ with $\pi$ is like defining a word with itself, and is 'cheating'.

I have been trying to find a way to make this conversion without $\pi$ but to no avail. I also tried using trig identities, but that has just led me in circles.

Is there another way to find the sine of $x^o$ without involving $\pi$ at all?

An even more explicit and blunt form of my question: What is another way to express $sin(x^o)$ without using $\pi$?

To be clear, I am not looking for anything that involves approximations or infinite series that cannot be wholly expressed in a finite amount of space (ex. Taylor series above is okay because it can be expressed in sigma notation).

$\endgroup$
  • 1
    $\begingroup$ How would finding the sine of an angle $x$, where $x$ is measured in degrees, help you to write an exact expression for $\pi$? $\endgroup$ – David K Feb 19 '16 at 19:17
  • $\begingroup$ @DavidK - edited. Please tell me of I should elaborate $\endgroup$ – Daniel Feb 19 '16 at 20:19
  • $\begingroup$ Sines and cosines of rational numbers of degrees can be computed as roots of polynomials, without involving $\pi$ at all. For example, $\sin(1 \text{ degree})$ is a root of a polynomial of degree $48$ that's too big to fit in a comment. $\endgroup$ – Robert Israel Feb 19 '16 at 20:32
  • 1
    $\begingroup$ If you want $\sin(180/n \text{ degrees})$ for $n$ a power of $2$, just repeatedly use the half-angle formulas. $\endgroup$ – Robert Israel Feb 19 '16 at 20:41
  • $\begingroup$ (1)You cannot convert from radians to degrees or vice-versa without using $\pi$ because $180^o= \pi$ radians. (2) A circle is not really an infinite-sided polygon. (3) To prove $\pi=\lim_{n\to \infty} n \sin (180/n)^o$ you need a def'n of $\pi$ to begin. Define $\pi$ as the length of a semi-circle of radius $1$. Then you need to define what "length of a semi-circle" means: It's the limit of the perimeters of polygons that uniformly approximate it This implies $\pi=\lim n\sin (180/n)^o.$ (4)The power series for $\sin x $ is in radians. $\endgroup$ – DanielWainfleet Feb 19 '16 at 21:45
1
$\begingroup$

As noted in a comment, there are equations you can write for the trigonometric functions of rational portions of a right angle (and therefore for the sine of any rational number of degrees) without using $\pi$ or any trigonometric functions. To actually use these equations may prove somewhat cumbersome, however.

Instead of attempting a formula for an arbitrary integer or rational number of degrees, let me just address the desire to evaluate $$ \newcommand{\Sin}{\mathop{\mathrm{Sin}}} \newcommand{\Cos}{\mathop{\mathrm{Cos}}} \lim_{n\to\infty} n \Sin\left(\frac{180}{n}\right) $$ where $\Sin$ is the sine function that takes its parameter in degrees, for example, $\Sin(90) = 1$.

It is sufficient to examine the following limit for integer values of $k$: $$ \lim_{k\to\infty} 2^k \Sin\left(\frac{180}{2^k}\right). $$ The quantity $\Sin\left(\frac{180}{2^k}\right)$ is easy to compute (at least, it is easy compared to such values as $\Sin(1)$). Let $\Cos(x)$ be the cosine function for $x$ measured in degrees; then apply the half-angle formula for cosines of angles in the interval from $0$ to $180$ degrees, inclusive: $$ \Cos\left(\frac{x}{2}\right) = \sqrt{\frac{1 + \Cos x}{2}}. $$

If $x = \dfrac{180}{2^m}$, then $\dfrac x2 = \dfrac{180}{2^{m+1}}$ and the half-angle formula just says that $$ \Cos\left(\frac{180}{2^{m+1}}\right) = \sqrt{\frac{1 + \Cos \left(\frac{180}{2^m}\right)}{2}}. $$

You can find the value for any $k$ by starting at $m=1$ and applying the half-angle formula repeatedly for $m=2,3,\ldots, k$: \begin{align} \Cos\left(\frac{180}{2}\right) &= \Cos(90) = 0,\\ \Cos\left(\frac{180}{4}\right) &= \sqrt{\frac{1 + \Cos \left(\frac{180}{2}\right)}{2}} = \sqrt{\frac{1 + 0}{2}} = \frac{1}{\sqrt 2}, \\ \Cos\left(\frac{180}{8}\right) &= \sqrt{\frac{1 + \Cos \left(\frac{180}{4}\right)}{2}} = \sqrt{\frac{1 + \frac{1}{\sqrt 2}}{2}} = \sqrt{\frac{\sqrt 2 + 1}{2\sqrt 2}}, \\ \Cos\left(\frac{180}{16}\right) &= \sqrt{\frac{1 + \Cos \left(\frac{180}{8}\right)}{2}} = \sqrt{\frac{1 + \sqrt{\frac{\sqrt 2 + 1}{2\sqrt 2}}}{2}} = \sqrt{\frac{\sqrt{2\sqrt 2} + \sqrt{\sqrt 2 + 1}}{2\sqrt{2\sqrt 2}}}, \\ \end{align} and so forth to obtain $\Cos\left(\frac{180}{2^k}\right)$ for $k$ as large as desired. Then obtain the sine by $$ \Sin\left(\frac{180}{2^k}\right) = \sqrt{1 - \left(\Cos\left(\frac{180}{2^k}\right)\right)^2}. $$

$\endgroup$
  • 1
    $\begingroup$ How would I calculate $Cos(\frac{180}{2^k})$ with a degree measure, then? Wouldn't that involve yet another Taylor series that takes radians and not degrees? (From looking at this page: people.math.sc.edu/girardi/m142/handouts/…) $\endgroup$ – Daniel Feb 19 '16 at 21:15
  • $\begingroup$ No, you do not need to use Taylor series. I have shown all the calculations explicitly for $k = 1,2,3,4$. Just continue like that for higher powers of $2$. You do have to take a square root each time. $\endgroup$ – David K Feb 19 '16 at 21:28
  • 1
    $\begingroup$ Do I not need to somehow calculate $Cos(\frac{180}{2^k})$, though? How would I calculate that? $\endgroup$ – Daniel Feb 19 '16 at 21:35
  • $\begingroup$ I have given step-by-step instructions. First compute $\newcommand{\Cos}{\mathop{\mathrm{Cos}}}\Cos(180/2)$, then $\Cos(180/4)$, then $\Cos(180/8)$, and so forth using the half-angle formula exactly as shown. The denominator doubles each time, so if you do this enough times the denominator will be $2^k$. $\endgroup$ – David K Feb 19 '16 at 21:44
  • $\begingroup$ How do we know that these formulas use degrees and not radians? $\endgroup$ – Daniel Feb 19 '16 at 22:24
0
$\begingroup$

In relation to David K's answer, I also note one small problem:

$$\sin(\frac x2)=\pm\sqrt{\frac{1-\cos(x)}2}$$

Obviously, $\sin(\frac x2)$ being negative is fully possible.

However, we know that for $0\le x\le\pi$, $\sin(x)>0$, in other words, it is positive.

On the interval from $-\frac{\pi}2\le x\le\frac{\pi}2$, $\cos(x)>0$.

So, we may start with $\cos(1)\approx0.540302306$.

We have $$\cos(0.5)\approx0.877581562$$

$$\cos(0.25)\approx0.968912422$$

$$\cos(0.125)\approx0.992197667$$

etc. (these values were obtained through the half angle formula)

Similarly, we have: (using half angle formula)

$$\sin(1)\approx0.841470985$$

$$\sin(0.5)\approx0.479425539$$

$$\sin(0.25)\approx0.247403959$$

$$\sin(0.125)\approx0.124674733$$

etc.

Now, knowing our trigonometric identities, one can use the sum of angles formula to approximate the desired value to desired accuracy.

$\endgroup$

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.