MHB Simplify cos(a)cos(2a)cos(3a)....cos(999a) if a=(2pi)/1999

  • Thread starter Thread starter anemone
  • Start date Start date
  • Tags Tags
    Simplify
Click For Summary
The discussion focuses on simplifying the product cos(a)cos(2a)cos(3a)...cos(999a) where a = (2π)/1999. Initial attempts involve pairing terms and applying Chebyshev's formula for multiple angle cosines, leading to a polynomial equation whose roots are related to the cosine values. The product of these roots is determined to be 1/2^(2n), allowing the conclusion that the product equals ±1/2^n. Further analysis reveals the need to consider the sign of the product based on the distribution of positive and negative cosine values. The final result indicates that the product can be expressed as ±(-1)^(n(n+1)/2)/2^n, reflecting the parity of n.
anemone
Gold Member
MHB
POTW Director
Messages
3,851
Reaction score
115
Simplify cos(a)cos(2a)cos(3a)...cos(999a) if a=(2pi)/1999

I don't see any way to approach this problem but my attempt is to group cos(a).cos(999a), then cos(2a).cos(998a), and so on.
Then I get 500 pairs where the first two pairs are 1/2[cos(100a)+cos(998a)] and 1/2[cos(100a)+cos(996a)], and etc.
Next, I see that cos(a)cos(2a)cos(3a)...cos(999a)= 1/2^500[[cos(100a)+cos(998a)][cos(100a)+cos(996a)][cos(100a)+cos(994a)][...].
Now, I'm stuck. I don't know how to proceed.
Any help, please?
 
Mathematics news on Phys.org
anemone said:
Simplify cos(a)cos(2a)cos(3a)...cos(999a) if a=(2pi)/1999

I don't see any way to approach this problem but my attempt is to group cos(a).cos(999a), then cos(2a).cos(998a), and so on.
Then I get 500 pairs where the first two pairs are 1/2[cos(100a)+cos(998a)] and 1/2[cos(100a)+cos(996a)], and etc.
Next, I see that cos(a)cos(2a)cos(3a)...cos(999a)= 1/2^500[[cos(100a)+cos(998a)][cos(100a)+cos(996a)][cos(100a)+cos(994a)][...].
Now, I'm stuck. I don't know how to proceed.
Any help, please?
(This problem cropped up in the dying days of the Math Help Forum. I'll have to reconstruct an outline of the solution.)

Here's how to compute the product $\displaystyle \prod_{k=1}^n \cos\tfrac{2k\pi}{2n+1} = \cos a\cos2a\cdots\cos na$, where $a = \tfrac{2\pi}{2n+1}.$

To start with, notice that the numbers $\theta = \tfrac{2k\pi}{2n+1}\ (0\leq k\leq 2n)$ are the solutions of the equation $\cos(2n+1)\theta = 1.$ Chebyshev's formula for multiple angle cosines says that $\cos(2n+1)\theta$ is a polynomial of degree $2n+1$ in $\cos\theta$, of the form $$\cos(2n+1)\theta = 2^{2n}\cos^{2n+1}\theta + \ldots \pm(2n+1)\cos\theta.$$ Put $x = \cos\theta$ to see that numbers $\cos\tfrac{2k\pi}{2n+1}\ (0\leq k\leq 2n)$ are the solutions of the equation $$2^{2n}x^{2n+1} + \ldots \pm(2n+1)x - 1 = 0.\quad(*)$$ One of these solutions (corresponding to $k=0$) is $x=1$. The other roots come in pairs, because of the fact that $\cos\varphi = \cos(2\pi-\varphi).$ Thus the roots of $(*)$ are $x=1$ (occurring once), together with the numbers $\cos\tfrac{2k\pi}{2n+1}\ (1\leq k\leq n)$ (occurring twice each). But the product of the roots of $(*)$ is $1/2^{2n}.$ It follows that $\displaystyle \prod_{k=1}^n \cos\tfrac{2k\pi}{2n+1} = \pm\frac1{2^n}.$

The final part of the calculation is to decide whether to take the plus or the minus sign in that formula. Notice that $\cos\theta$ is positive if $0\leq\theta<\pi/2$, and negative if $\pi/2<\theta\leq\pi$. In the above product, roughly the first half of the terms will be positive and the second half will be negative. For small values of n, you can check that the sign is given as in this table: $$\begin{array}{c|c|c|c|c|c|c|c}n&1&2&3&4&5&6&7\\ \hline \text{sign}&-&-&+&+&-&-&+ \end{array}$$ I'll leave you to work out what it should be for $n=999.$
 
Thanks, Opalg!

You approach is so easy to follow and I like it very much.
But it's also going uphill to the point where I don't understand why the following is true.:confused:

But the product of the roots of $(*)$ is $1/2^{2n}.$ It follows that $\displaystyle \prod_{k=1}^n \cos\tfrac{2k\pi}{2n+1} = \pm\frac1{2^n}.$

Could you please elaborate on that one?
Thanks again.
 
anemone said:
I don't understand why the following is true.:confused:

Opalg said:
But the product of the roots of $(*)$ is $1/2^{2n}.$ It follows that $\displaystyle \prod_{k=1}^n \cos\tfrac{2k\pi}{2n+1} = \pm\frac1{2^n}.$
Could you please elaborate on that one?
The roots of $(*)$ are $\cos\tfrac{2k\pi}{2n+1}\ (1\leq k\leq n)$ (occurring twice each), together with 1. So their product is $\displaystyle \Bigl(\prod_{k=1}^n \cos\tfrac{2k\pi}{2n+1}\Bigr)^{\!\!\!2}.$ But the product of the roots of a polynomial of odd degree is the negative of the constant term divided by the coefficient of the leading term. Therefore $\displaystyle \Bigl(\prod_{k=1}^n \cos\tfrac{2k\pi}{2n+1}\Bigr)^{\!\!\!2} = \frac1{2^{2n}}.$ Now you just have to take the square root of both sides.
 
With the awesome explanation, there's no reason to say that I don't understand!:D
BIG thanks to you, Opalg!
Also, big thanks to this site!:cool:
 
Here's what I tried:

From $\displaystyle(2n+1-(2k-1))a=(2n+1-(2k-1))\frac{2\pi}{2n+1}=2\pi-(2k-1)\frac{2\pi}{2n+1}=2\pi-(2k-1)a$, it follows that $\sin{(2n+1-(2k-1))a}=\sin(-(2k-1)a)=-\sin{(2k-1)a}$. So we simply have, $\displaystyle\frac{\sin{(2n+1-(2k-1))a}}{{\sin(2k-1)a}}=-1$(*).The identity $\sin(2\alpha)=2\sin(\alpha)\cos(\alpha)$, or $\displaystyle \cos(\alpha)=\frac{1}{2}\frac{\sin(2\alpha)}{\sin(\alpha)}$ gives:\[\prod_{1\leq k\leq n}\cos(ka)=\prod_{1\leq k\leq n}\frac{1}{2}\frac{\sin(2ka)}{\sin(ka)}=\frac{1}{2^n}\prod_{1\leq k\leq n}\frac{\sin(2ka)}{\sin(ka)}=\frac{\sin(2a)}{\sin(1a)}\cdot\frac{\sin(4a)}{\sin(2a)}\cdot\frac{\sin(6a)}{\sin(3a)}\cdot\frac{\sin(8a)}{\sin(4a)}\cdots\frac{\sin(2na)}{\sin(na)},\]where the RHS product $A_n$ (without $\displaystyle\frac{1}{2^n}$) after cancellation is something like (for $n=6$)$\displaystyle A_6=\frac{\sin(2a)}{\sin(1a)}\cdot \frac{\sin(4a)}{\sin(2a)}\cdot\frac{\sin(6a)}{\sin(3a)}\cdot\frac{\sin(8a)}{\sin(4a)}\cdot\frac{\sin(10a)}{\sin(5a)}\cdot\frac{\sin(12a)}{\sin(6a)}= \frac{1}{\sin(1a)} \cdot\frac{1}{1}\cdot\frac{1}{\sin(3a)}\cdot\frac{\sin(8a)}{1}\cdot\frac{\sin(10a)}{\sin(5a)} \cdot\frac{\sin(12a)}{1}=\frac{\sin(8a)}{\sin(1a)}\cdot\frac{\sin(10a)}{\sin(3a)} \frac{\sin(12a)}{\sin(5a)}$.Rearanging the numerator, we get: \[A_n=\displaystyle \frac{\sin(2na)}{\sin(1a)}\cdot\frac{\sin((2n-2)a)}{\sin(3a)}\cdot\frac{\sin((2n-4)a)}{\sin(5a)}\cdots=\frac{\sin{(2n+1-1)a}}{{sin1a}}\cdot\frac{\sin{(2n+1-3)a}}{{\sin3a}}\cdots\frac{\sin{(2n+1-(2t-1))a}}{{\sin(2t-1)a}}=\prod_{1\leq k\leq t}\frac{\sin{(2n+1-(2k-1))a}}{{\sin(2k-1)a}},\] where (by inspection) $t=\lfloor (n+1)/2\rfloor$.But then from (*), we immediately find that $A_n=\prod_{1\leq k\leq t}(-1)=(-1)^t$. Finally, $\displaystyle\prod_{1\leq k\leq n}\cos(ka)=\frac{1}{2^n}A_n=\frac{1}{2^n}(-1)^t$.Note: Since $\frac{n(n+1)}{2}$ and $t$ have the same parity, the answer may be written as: $\displaystyle\prod_{1\leq k\leq n}\cos(ka)=\frac{(-1)^{\frac{n(n+1)}{2}}}{2^n}$.

መለሰ
 
Thread 'Erroneously  finding discrepancy in transpose rule'
Obviously, there is something elementary I am missing here. To form the transpose of a matrix, one exchanges rows and columns, so the transpose of a scalar, considered as (or isomorphic to) a one-entry matrix, should stay the same, including if the scalar is a complex number. On the other hand, in the isomorphism between the complex plane and the real plane, a complex number a+bi corresponds to a matrix in the real plane; taking the transpose we get which then corresponds to a-bi...

Similar threads

Replies
8
Views
2K
  • · Replies 1 ·
Replies
1
Views
4K
  • · Replies 10 ·
Replies
10
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
Replies
6
Views
3K
  • · Replies 5 ·
Replies
5
Views
4K
  • · Replies 4 ·
Replies
4
Views
15K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 21 ·
Replies
21
Views
2K