Author |
Topic: EVALUATE THIS INFINITE SUM (Read 540 times) |
|
pcbouhid
Uberpuzzler
Gender:
Posts: 647
|
|
EVALUATE THIS INFINITE SUM
« on: Nov 20th, 2005, 4:55am » |
Quote Modify
|
Evaluate this infinite sum: k=inf 1 SUM ------------------- k=0 (3k+1)(3k+2)(3k+3) No computers allowed!
|
« Last Edit: Nov 20th, 2005, 4:58am by pcbouhid » |
IP Logged |
Don´t follow me, I´m lost too.
|
|
|
Eigenray
wu::riddles Moderator Uberpuzzler
Gender:
Posts: 1948
|
|
Re: EVALUATE THIS INFINITE SUM
« Reply #1 on: Nov 20th, 2005, 6:03am » |
Quote Modify
|
Fun. First rewrite it as 2S = [sum] 1/(3k+1) - 2/(3k+2) + 1/(3k+3) Now let L(x) = x + x2/2 + x3/3 + ... = -log(1-x). Ignoring convergence, 2(w-w')S = L(w) - L(w') - w'L(w) + wL(w') = (L(w)+L(w'))*(w-w')/2 + (L(w)-L(w'))*3/2, where w = e2 pi i/3, w'=w2 are cube roots of 1. Evaluating L(w)+L(w') = -log[(1-w)(1-w')] = -log|1-w|2 = -log 3, L(w)-L(w') = log[(1-w')/(1-w)] = log (-w') = i pi/3, it follows 2i sqrt(3) S = -(log 3) i sqrt(3)/2 + i pi/2, so S = [ pi/sqrt(3) - log 3 ]/4
|
|
IP Logged |
|
|
|
Barukh
Uberpuzzler
Gender:
Posts: 2276
|
|
Re: EVALUATE THIS INFINITE SUM
« Reply #2 on: Nov 20th, 2005, 10:11am » |
Quote Modify
|
on Nov 20th, 2005, 6:03am, Eigenray wrote: Fun2 ! Is it possible in a similar manner to evaluate infinite sums of n-degree polynomials?
|
|
IP Logged |
|
|
|
|