11
$\begingroup$

Problem B6 on the 2016 Putnam exam is to calculate:

$$\sum\limits_{k=1}^\infty \left( \frac{(-1)^{k-1}}{k} \sum\limits_{n=0}^\infty \frac{1}{k 2^n + 1} \right)$$

The direct approach

Sum[(-1)^{k-1}/k Sum[1/(k 2^n + 1),{n,0,\[Infinity]}], {k,1, \[Infinity]}]

does not resolve to the analytic solution. However, the numerical value, computed by N[%] gives the correct value: $1$.

How can we compute the analytic solution to this summation?

$\endgroup$
1
  • 1
    $\begingroup$ A few analytical solutions here for problem B6. $\endgroup$
    – Syed
    Oct 11, 2022 at 9:20

1 Answer 1

16
$\begingroup$

One way using MellinTransform and InverseMellinTransform:

func = (-1)^(k - 1)/k*1/(A*k 2^n + 1)(* where A = 1*)

InverseMellinTransform[Sum[Sum[MellinTransform[func, A, s] // 
PowerExpand, {k, 1, Infinity}], {n, 0, Infinity}], s, A] /. A -> 1

(*\[Pi] InverseMellinTransform[Csc[\[Pi] s] Zeta[1 + s], s, 1]*)(*Can't compute ! Weakness !*)

Using:

$$\zeta (1+s)=\sum _{k=1}^{\infty } \frac{1}{k^{1+s}}$$

Then:

InverseMellinTransform[\[Pi] Csc[\[Pi] s] 1/k^(1 + s), s, A] /. A -> 1 // FullSimplify
(*1/(k + k^2)*)
Sum[%, {k, 1, Infinity}]

(* 1 *)

Addition for Table of InverseMellinTransform: $$\mathcal{M}_s^{-1}[\pi \csc (\pi s) \zeta (1+s)](A)=\gamma +\psi ^{(0)}\left(1+\frac{1}{A}\right)=H_{\frac{1}{A}}$$ for: $0<\Re(s)<1$

$$\mathcal{M}_s^{-1}[\pi \csc (\pi s) \zeta (2-s)](A)=\frac{\gamma }{A}+\frac{\psi ^{(0)}(1+A)}{A}=\frac{H_A}{A}$$

for: $0<\Re(s)<1$

$\endgroup$
1
  • 2
    $\begingroup$ Mariusz Iwaniuk: Very nice... I'm guessing from a real mathematician who knows Mathematica rather than from a computer scientist. ($\checkmark$) $\endgroup$ Oct 11, 2022 at 12:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.