3

There are at least two ways to define the cosine function:

  1. You can define it with a right triangle in the unit circle and extend the definition to $\mathbb{R}$. (classic definition)

  2. The other definition is the one with the series: $\cos(x) =\sum_0^\infty \frac{(-1)^nx^{2n}}{(2n)!}$

I would like to prove these definitions are equivalent. To prove $1\Rightarrow2$, I showed that $\cos'=-\sin$ and $\sin'=\cos$ and then I used the Taylor formula.

The issue is $2\Rightarrow1$ ; I can't see how to begin. Even proving the $2\pi$ periodicity looks hard.

I need help, please.

  • 1
    You can't show equivalence as they are NOT equivalent: the right triangle thing is the first high school definition, before introducing the trigonometric circle and expanding the definition of the trigonometric functions to all numbers, by means of similarity, radians and stuff. The definition by means of the power series uses the expanded definition. – DonAntonio May 17 '14 at 16:34
  • 1
    @DonAntonio I don't understand. Doesn't he mean the unit circle definition? – Sawarnik May 17 '14 at 16:36
  • 2
    Yes I meant the extended geometric definition of cosine. excuse me. – Adrien Laurent May 17 '14 at 16:38
  • 1
    @Karavel In fact, with the second definition, $\pi$ is not yet defined. Part of the job (and this is not trivial) consists in showing that $\cos$ is periodic; then $\pi$ is defined to be half of the period of $\cos$. You should have a look (for example) at the very first chapter of Rudin's Real and complex analysis – Etienne May 17 '14 at 17:19
  • Thanks a lot. For those who are interested, the book can be read here : http://ruangbacafmipa.staff.ub.ac.id/files/2012/02/Real-and-Complex-Analysis-by-Walter-Rudin.pdf. It's in the zero chapter. – Adrien Laurent May 17 '14 at 19:06
  • @Etienne: In my answer I first define $π$ as twice the first positive root of $\cos$ before proving that it is periodic via proving the geometric nature of $\exp$. =) – user21820 May 20 '14 at 10:30
  • @user21820 That's fine! – Etienne May 20 '14 at 14:05

1 Answers1

3

There actually an elementary way to derive everything from series definitions of the exponential and trigonometric functions:

First obtain the series for $\exp$ on $\mathbb{C}$ about $0$ where $\exp$ is defined as a function such that $\exp' = \exp$ and $\exp(0) = 1$, and prove that the series converges on all of $\mathbb{C}$. Then for any $w \in \mathbb{C}$ find the series for $z \mapsto \exp(z+w)$ about $0$, which will be just $\exp(w)$ times the series for $\exp$, and thus it also converges on all of $\mathbb{C}$ and so $\exp(z+w) = \exp(z) \exp(w)$ for any $z,w \in \mathbb{C}$.

Next prove that the series for $\cos$ converges on $\mathbb{R}$, and likewise the series for $\sin$. Then note that $\exp(ix) = \cos(x) + i \sin(x)$ for any real $x$ because of the series. $\exp(i) \exp(-i) = \exp(0) = 1$ and $\exp(-i) = \cos(-1) + i \sin(-1) = \cos(1) - i \sin(1) = \exp(i)^*$ ("$^*$" denotes conjugation), and so $|\exp(i)| = 1$. Now we easily get $|\exp(ik)| = |\exp(i)^k| = 1^k = 1$ for any integer $k$, and $|\exp(i\frac{k}{m})|^m = |\exp(i\frac{k}{m})^m| = |\exp(ik)| = 1$ implies $|\exp(i\frac{k}{m})| = 1$ for any integers $k,m$. Since $\exp$ is continuous, $|\exp(ix)| = 1$ for any real $x$.

Finally we define $π$ as twice the first positive root of $\cos$, which we know must exist in $(0,2)$ by intermediate value theorem because we have $\cos(x) < 1 - \frac{x^2}{2!} + \frac{x^4}{4!}$ for any $x$ by repeated differentiation and $\cos(0) = 1 > 0$ and $\cos(2) < 1 - \frac{2^2}{2!} + \frac{2^4}{4!} = -\frac{1}{3} < 0$. Similarly we have $\sin(x) > x - \frac{x^3}{3!} > 0$ for any $x \in (0,2)$. These imply $\exp(i\frac{π}{2}) = \cos(\frac{π}{2}) + i \sin(\frac{π}{2}) = 0 + i(1)$, because the unit circle intersects the imaginary axis only at $i,-i$, and $\sin(\frac{π}{2}) > 0$. Hence we get $\exp(i2π) = \exp(i\frac{π}{2})^4 = i^4 = 1$. Marvelous! We're done!

Personally I like this way because it reveals the geometric nature of $\exp$ nicely and completely naturally, but I don't know why none of my teachers ever taught me this way.

user21820
  • 57,693
  • 9
  • 98
  • 256
  • I had never been taught this. In what mathematic field do you learn the cosine definition equivalence? – Adrien Laurent May 20 '14 at 13:57
  • @Karavel: I was taught only the 'circle' definition and very non-rigorous proofs to get from there to the series, but one or two years ago I decided to start from $\exp' = \exp$ and see if I could derive everything else, as I felt that it was enough. So in fact I was never taught the equivalence... – user21820 May 20 '14 at 14:12