If ω = e 2 0 1 5 2 i π . Then
k = 1 ∑ 2 0 1 4 1 + ω k + ω 2 k 1 = ?
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.
If u , v are complex numbers of modulus 1 such that 1 + u + u 2 = 1 + v + v 2 , then ( u − v ) ( u + v + 1 ) = 0 , so either u = v or u + v + 1 = 0 , and this means that either u = v or else { u , v } = { e 3 2 π i , e − 3 2 π i } . Thus if N ≥ 1 is an integer such that 2 N + 1 is not divisible by 3 , and if we define ω = e 2 N + 1 2 π i , then none of ω k (for 0 ≤ k ≤ 2 N ) are equal to e ± 3 2 π i , and hence we deduce that 1 + ω k + ω 2 k , for 0 ≤ k ≤ 2 N are distinct complex numbers. We shall find the polynomial (of degree 2 N + 1 ) whose roots are 1 + ω k + ω 2 k for 0 ≤ k ≤ 2 N .
If X = 1 + ω k + ω 2 k , then ( ω k + 2 1 ) 2 = X − 4 3 , and hence ω k = − 2 1 ± X − 4 3 , and hence 0 = 1 − ( − 2 1 ± X − 4 3 ) 2 N + 1 = 1 + ( 2 1 ∓ X − 4 3 ) 2 N + 1 = 1 + j = 0 ∑ 2 N + 1 ( j 2 N + 1 ) ( 2 1 ) 2 N + 1 − j ( ∓ X − 4 3 ) j = 1 + j = 0 ∑ N ( 2 j 2 N + 1 ) ( 2 1 ) 2 ( N − j ) + 1 ( X − 4 3 ) j ∓ X − 4 3 j = 0 ∑ N ( 2 j + 1 2 N + 1 ) ( 2 1 ) 2 ( N − j ) ( X − 4 3 ) j and hence f ( X ) = ( X − 4 3 ) [ j = 0 ∑ N ( 2 j + 1 2 N + 1 ) ( 2 1 ) 2 ( N − j ) ( X − 4 3 ) j ] 2 − [ 1 + j = 0 ∑ N ( 2 j 2 N + 1 ) ( 2 1 ) 2 ( N − j ) + 1 ( X − 4 3 ) j ] 2 = 0 Now f ( X ) is a polynomial of degree 2 N + 1 , and it has 1 + ω k + ω 2 k as a root for any 0 ≤ k ≤ 2 N , and hence it follows that this is the complete list of its roots. Thus we deduce that j = 0 ∑ 2 N 1 + ω k + ω 2 k 1 = − f ( 0 ) f ′ ( 0 ) and hence j = 1 ∑ 2 N 1 + ω k + ω 2 k 1 = − f ( 0 ) f ′ ( 0 ) − 3 1 Now standard combinatorial tricks tell us that f ( 0 ) = − 4 3 [ j = 0 ∑ N ( 2 j + 1 2 N + 1 ) ( 2 1 ) 2 ( N − j ) ( − 4 3 ) j ] 2 − [ 1 + j = 0 ∑ N ( 2 j 2 N + 1 ) ( 2 1 ) 2 ( N − j ) + 1 ( − 4 3 ) j ] 2 = − 2 − 2 cos ( 3 ( 2 N + 1 ) π ) and, similarly, f ′ ( 0 ) = 3 1 ( 1 + 2 N ) ( 3 − 2 3 sin ( 3 2 N π ) ) With N = 1 0 0 7 , we deduce that the final answer is − f ( 0 ) f ′ ( 0 ) − 3 1 = 1 3 4 3
Problem Loading...
Note Loading...
Set Loading...
This is actually not too bad: let n = 2 0 1 5 , and find g such that 3 g ≡ 1 ( m o d n ) . Now k = 1 ∑ n − 1 1 + ω k + ω 2 k 1 = k = 1 ∑ n − 1 1 − ω 3 k 1 − ω k = k = 1 ∑ n − 1 1 − ω 3 k 1 − ω 3 g k = k = 1 ∑ n − 1 ( 1 + ω 3 k + ω 6 k + ⋯ + ω 3 ( g − 1 ) k ) = − g + k = 0 ∑ n − 1 ( 1 + ω 3 k + ω 6 k + ⋯ + ω 3 ( g − 1 ) k ) = n − g + k = 0 ∑ n − 1 ( ω 3 k + ω 6 k + ⋯ + ω 3 ( g − 1 ) k ) = n − g , because for ζ any n th root of unity, ∑ k = 0 n − 1 ζ k = { n 0 if ζ = 1 otherwise. (This is a standard fact: multiplying the sum by ζ permutes the terms, so ζ S = S , so S = 0 unless ζ = 1 . )
For n = 2 0 1 5 , we get g = 6 7 2 , and the answer is n − g = 1 3 4 3 .
Edit: I thought this problem was familiar. I solved it five years ago , in pretty much the same way!