Crazy Cosine Calculus!

Calculus Level 3

Let I m = 0 2 π cos ( x ) cos ( 2 x ) cos ( m x ) d x I_m = \displaystyle \int_0^{2\pi} \cos(x) \cos(2x) \dots \cos(mx) dx . What is the sum of all integers 100 m 110 100 \leq m \leq 110 such that I m 0 I_m \neq 0 ?


The answer is 522.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Aakarshit Uppal
Jan 31, 2015

This can be solved using the property

f ( 2 a x ) = f ( x ) f(2a - x) = f(x) 0 2 a f ( x ) d x = 2 0 a f ( x ) d x \Rightarrow \int_{0}^{2a}f(x)dx = 2\int_{0}^{a}f(x)dx

Let f ( x ) = cos ( x ) cos ( 2 x ) cos ( m x ) f(x) = \cos(x) \cos(2x) \dots\cos(mx) f ( 2 π x ) = cos ( 2 π x ) cos ( 4 π 2 x ) cos ( 2 m π m x ) \Rightarrow f(2\pi - x) = \cos(2\pi - x) \cos(4\pi - 2x) \dots\cos(2m\pi - mx) f ( 2 π x ) = cos ( x ) cos ( 2 x ) cos ( m x ) \Rightarrow f(2\pi - x) = \cos(x) \cos(2x) \dots\cos(mx)

(Since cos ( 2 n π ± θ ) = cos ( θ ) \cos(2n\pi \pm \theta) = \cos(\theta) where n n is an integer )

f ( 2 π x ) = f ( x ) \Rightarrow f(2\pi - x) = f(x)

Thus,

I m = 2 0 π cos ( x ) cos ( 2 x ) cos ( m x ) d x . . . ( i ) I_m = 2\int_{0}^{\pi}\cos(x) \cos(2x) \dots\cos(mx) dx ...(i)

Now, we know that 0 a f ( x ) d x = 0 a f ( a x ) d x \int_{0}^{a}f(x)dx = \int_{0}^{a} f(a-x)dx

Therefore

I m = 2 0 π cos ( π x ) cos ( 2 π 2 x ) cos ( m π m x ) d x I_m = 2\int_{0}^{\pi}\cos(\pi - x) \cos(2\pi - 2x) \dots\cos(m\pi - mx) dx

Thus I m = + 2 0 π cos ( x ) cos ( 2 x ) cos ( m x ) d x = I m I_m = +2\int_{0}^{\pi}\cos(x) \cos(2x) \dots\cos(mx) dx = I_m

when m = 4 n + 3 m = 4n+3 or m = 4 n ; n I n t e g e r s m = 4n ; n \in Integers

And

I m = 2 0 π cos ( x ) cos ( 2 x ) cos ( m x ) d x = I m I_m = -2\int_{0}^{\pi}\cos(x) \cos(2x) \dots\cos(mx) dx = -I_m 2 I m = 0 I m = 0 \Rightarrow 2 I_m = 0 \Rightarrow I_m = 0

when m = 4 n + 1 m = 4n+1 or m = 4 n + 2 ; n I n t e g e r s m = 4n+2 ; n \in Integers

Thus, I m 0 I_m \neq 0 for

m = 100 , 103 , 104 , 107 , 108 m = 100 , 103 , 104 , 107 , 108

Thus,

A n s w e r = 100 + 103 + 104 + 107 + 108 = 522 Answer = 100+103+104+107+108 = \boxed{\color{#20A900}{522}}

I think you only showed that I_m is zero if m=4n+1 or 4n+2, but you did not prove that it is not zero otherwise. Or do I miss something here?

György Gehér - 2 years, 9 months ago

Nicely done!

ibrahim abdullah - 5 years, 4 months ago
Michael Tong
Jan 24, 2014

If we use the identity c o s z = e i z + e i z 2 \large{cos z = \frac{e^{iz} + e^{-iz}}{2}} , we can rewrite the integral as 0 2 π k = 1 m e i k x + e i k x 2 d x = 1 2 m a k = ± 1 0 2 π e i ( a 1 + 2 a 2 + 3 a 3 + + m a m ) x d x \displaystyle \int_0^{2\pi} \displaystyle \prod_{k=1}^m \frac{e^{ikx} + e^{-ikx}}{2} dx = \frac{1}{2^m} \displaystyle \sum_{a_k = \pm 1} \displaystyle \int_0^{2\pi} e^{i(a_1 + 2a_2 + 3a_3 + \dots + ma_m)x} dx .

Furthermore, note that 0 2 π e i n x d x = 2 π \int_0^{2\pi} e^{inx} dx = 2\pi if n = 0 n = 0 , otherwise it is equal to 0 0 as n n ranges in the integers. Recall that e i x = cos x + i sin x e^{ix} = \cos x + i \sin x , so the real and imaginary parts of the graph of e i n x e^{inx} from 0 0 to 2 π 2\pi is the cosine and sine graph, respectively, repeated n n times, so the integral is zero. Unless n = 0 n = 0 , in which case it is the horizontal line y = 1 y = 1 , which has an area of 2 π 2\pi .

Since our integral ranges over all 2 m 2^m m-tuples of ( a 1 , a 2 , , a m ) (a_1, a_2, \dots, a_m) with a k = ± 1 a_k = \pm 1 , it suffices to show that, for a given m m , there exists such a sequence { a k } \{a_k\} such that a 1 + 2 a 2 + + m a m = 0 a_1 + 2a_2 + \dots + ma_m = 0 .

So, 0 = a 1 + 2 a 2 + + m a m 1 + 2 + + m = m ( m + 1 ) 2 ( m o d 2 ) 0 = a_1 + 2a_2 + \dots + ma_m \equiv 1 + 2 + \dots + m = \frac{m(m+1)}{2} \pmod 2 , thus 4 m ( m + 1 ) 4 | m(m+1) and m 0 , 3 ( m o d 4 ) m \equiv 0, 3 \pmod 4 .

Consider m 0 ( m o d 4 ) m \equiv 0 \pmod 4 . Then we find a solution in ( 1 2 3 + 4 ) + ( 5 6 7 + 8 ) + + ( ( m 3 ) ( m 2 ) ( m 1 ) + m ) (1 - 2 - 3 + 4) + (5 - 6 - 7 + 8) + \dots + ((m-3) - (m-2) - (m-1) + m)

Consider m 3 ( m o d 4 ) m \equiv 3 \pmod 4 . Then we find a solution in ( 1 + 2 3 ) + ( 4 5 6 + 7 ) + + ( ( m 3 ) ( m 2 ) ( m 1 ) + m ) (1+2-3) + (4-5-6+7) + \dots + ((m-3)-(m-2)-(m-1)+m) .

Thus, I m 0 I_m \neq 0 iff m 0 , 3 ( m o d 4 ) m \equiv 0, 3 \pmod 4 . This gives the solution set of { 100 , 103 , 104 , 107 , 108 } \{100, 103, 104, 107, 108\} , with sum 522 \boxed{522} . \blacksquare

Great problem!

Peter Byers - 7 years, 4 months ago

Log in to reply

yes indeed, nice problem. i sorta forgot about the cos(z) = (e^(iz) + e^(-iz))/2 =P

but the problem beatifully combines number theory, calculus and complex numbers along with clever thinking and a couple tricks here and there =)

but im guessing this is one of the "easier" putnam problems? (or is it one of the harder ones? @Michael Tong) <-- this reminds me, Brilliant should add a tagging system similar to facebook and forums so that one can tag another as "@Username".

Muhammad Shariq - 7 years, 4 months ago

Log in to reply

This was from the first year of putnam, but it was near the end (recall that they give 12 questions; 6 in each batch, I believe this was the 2nd to last problem so that means it's on the harder side)

Michael Tong - 7 years, 4 months ago

Looks like your wishes came true.

Cole Coupland - 7 years, 2 months ago

I used the same argument but without Euler's formula:

(*) cos ( x ) cos ( y ) = 1 2 ( cos ( x y ) + cos ( x + y ) , x , y R f m ( x ) : = k = 1 m cos ( k x ) = ( ) 1 2 m 1 a k = ± 1 cos ( ( 1 a 1 + + m a m ) x ) \begin{aligned} \text{(*)}\quad\cos(x)\cos(y)&=\frac{1}{2}(\cos(x-y)+\cos(x+y),&&x,\:y\in\mathbb{R}\\\\ \Rightarrow\quad f_m(x)&:=\prod_{k=1}^m\cos(kx)\underset{\text(*)}{=} \frac{1}{2^{m-1}}\sum_{a_k=\pm 1}\cos\bigl((1a_1+\ldots+ma_m)x\bigr) \end{aligned}

The rest works exactly the same!

Carsten Meyer - 2 years, 3 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...