A perfect welcome to 2016, Part 2

Let S = ω 480 S=\sum\omega^{480}

where the sum is taken over all primitive 201 6 th 2016^\text{th} roots of unity ω \omega . Enter S 8 \dfrac{S}{8} as your answer.


Last year's problem


The answer is 6.

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

Otto Bretscher
Jan 1, 2016

Aareyan has written a fine solution. For the sake of variety, let's look at another approach:

Let c m ( n ) c_m(n) be the sum of the n n th powers of the primitive m m th roots of unity. There is a useful formula, due to Otto Hölder: c m ( n ) = ϕ ( m ) μ ( r ) ϕ ( r ) c_m(n)=\frac{\phi(m)\mu(r)}{\phi(r)} where r = m gcd ( m , n ) r=\frac{m}{\gcd(m,n)} . To prove this formula, first observe that both sides are multiplicative functions in m m . The verification for a prime power m = p k m=p^k is a bit tedious but straightforward.

In our case we have n = 480 , m = 2016 n=480, m=2016 and d = 21 d=21 , so that S = ϕ ( 2016 ) μ ( 21 ) ϕ ( 21 ) = 48 S=\frac{\phi(2016)\mu(21)}{\phi(21)}=48 . The answer we seek is 6 \boxed{6} ... a perfect number... a perfect year.

Aareyan Manzoor
Jan 1, 2016

nice problem. let c m ( n ) c_m(n) be the sum of the nth power of the mth roots of unity.if you dont know about them see this awesome wiki .

from the wiki we see that c m ( n ) = { 0 , m ∤ n m , m n c_m(n)=\begin{cases}0 &&,m\not\mid n\\m&&,m\mid n\end{cases} . let r m ( n ) r_m(n)

be the nth powers of the primitive mth roots of unity.

note that every nth root of unity is a dth primitive root of unity for d dividing n.

for example,lets look at the 4th roots of unity: 1 , 1 , i , i 1,-1,i,-i . the i , i i,-i are primitive 4th roots of unity, 1 -1 is a primitive 2nd root of unity and finally 1 1 is a primitive 1st root of unity.assuming this elucidated the earlier statement, lets move on.

we can say that the sum of the nth power of the mth roots of unity are the sum of the primitive dth roots of unity such that d n d|n . or c m ( n ) = d n ( r d ( n ) ) c_m(n)=\sum_{d|n}(r_d(n)) .

written in Dirichlet convolution form this is c = r u c=r*u . convoluting both sides with μ \mu we get c μ = r I = r c*\mu=r*I=r or r m ( n ) = d m c d ( n ) μ ( m d ) r_m(n)=\sum_{d|m}c_d(n)\mu(\dfrac{m}{d}) .

we remember c d ( n ) c_d(n) "dissapears" for d ∤ n d\not\mid n .so basically we want d|n and d|m. from this we want a common divisor of n,m or we want d g c d ( m , n ) d|gcd(m,n) .

so the summation turns into r m ( n ) = d g c d ( m , n ) d μ ( m d ) r_m(n)=\sum_{d|gcd(m,n)}d\mu(\dfrac{m}{d}) .

put m,n=2016,480. their greatest common is 96 = 2 5 3 96=2^5*3 .

remember that μ \mu dissapears over non-squarefree integers. 2016 = 2 5 3 2 7 2016=2^5*3^2*7 . for 2015 d \dfrac{2015}{d} to be square free, power of 2 atleast 4 and power of 3 atleast 1. so the possible d are 2 4 3 = 48 2^4*3=48 and 2 5 3 = 96 2^5*3=96 .

put this in to get 48 + 96 = 48 -48+96=\boxed{48} . divide by 8 to get 6 \boxed{6} .

Yes, a perfect solution with a perfect answer, 6 (+1) It's gonna be a perfect year!

A few small things:

"we can say that the sum of the nth power of the mth roots of unity are the sum of the dth roots of unity such that " should be something like ... "is the sum of the nth powers of the dth primitive roots of unity..."

"let r m ( n ) r_m(n) be the nth powers of the primitive mth roots of unity." should be "...be the sum of the nth powers of the dth primitive..."

Otto Bretscher - 5 years, 5 months ago

Log in to reply

now i get why divide by 8.. because 6 is the smallest perfect number.

Aareyan Manzoor - 5 years, 5 months ago

Log in to reply

You got it....same thing in "Perfect Welcome to 2016, part 1"

Otto Bretscher - 5 years, 5 months ago

I'm impressed how quickly you understand these new topics... you really have an amazing talent (and enthusiasm) for maths, Comrade Aareyan!

Otto Bretscher - 5 years, 5 months ago

Log in to reply

thanks a lot. i really am able to learn this because of practice from you awesome problems! thank you again.

Aareyan Manzoor - 5 years, 5 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...