3
$\begingroup$

We have $$\sum_{m=0}^\infty \prod_{n=1}^m\frac{n}{3n-1}=\frac{3}{2}+\frac{\ln(\sqrt[3]{2}-1)}{2^{7/3}}+\frac{\sqrt{3}}{2^{4/3}}\arctan\frac{\sqrt{3}}{2\sqrt[3]{2}-1}\tag1$$ $$\sum_{m=0}^\infty \prod_{n=1}^m\frac{n}{3n-2}=\frac{3}{2}-\frac{\ln(\sqrt[3]{2}-1)}{2^{5/3}}+\frac{\sqrt{3}}{2^{2/3}}\arctan\frac{\sqrt{3}}{2\sqrt[3]{2}-1}\tag2$$ with the first discussed in this post. (Update) Courtesy of the answers below, we also have, $$\sum_{m=0}^\infty \prod_{n=1}^m\frac{n}{4n-1}=\frac{4}{3}+\frac{\ln\big((4^{1/4}-3^{1/4})\sqrt{2+\sqrt{3}}\big)}{3^{5/4}\sqrt{2}}+\frac{\sqrt{2}}{3^{5/4}}\arctan\big(3^{1/4}\sqrt{2+\sqrt{3}}\big)\tag3$$ $$\sum_{m=0}^\infty \prod_{n=1}^m\frac{n}{4n-3}=\frac{4}{3}-\frac{\ln\big((4^{1/4}-3^{1/4})\sqrt{2+\sqrt{3}}\big)}{3^{3/4}\sqrt{2}}+\frac{\sqrt{2}}{3^{3/4}}\arctan\big(3^{1/4}\sqrt{2+\sqrt{3}}\big)\tag4$$Walpha gives a closed-form for $(3)$, but uses two log functions and two arctans. The two answers below show it can be simplified further similar to $(1),(2)$.

$\endgroup$
  • $\begingroup$ I calculated it further, but my result differs numerically from the RHS of your equation 3. $\endgroup$ – John Bentin Nov 12 '16 at 10:20
  • $\begingroup$ @JohnBentin: I've double-checked $(3)$, and both sides do equate to $1.4639\dots$ Perhaps you mistyped something? $\endgroup$ – Tito Piezas III Nov 12 '16 at 10:51
  • $\begingroup$ Indeed I did slip up. See my corrected version below. $\endgroup$ – John Bentin Nov 12 '16 at 17:54
1
$\begingroup$

Yes. The logarithms have the same denominator. So factoring this out gives a simple difference of logs: namely the simple log of a quotient, which itself can be simplified to $$2\ln\left(\frac12+\frac{\surd3}2-\sqrt{2\surd3}\right)$$(if my calculation is right). A similar method will simplify the inverse cotangents to a single inverse cotangent, by use of the formula $$\alpha+\beta=\operatorname{arcot}\frac{\cot\alpha\cot\beta-1}{\cot\alpha+\cot\beta}$$(exactly or to within $\pm\pi$).

(Revised) According to my (corrected) calculation, Wolfram's expression for formula 3 can be written as $$\frac43+\frac4{3a}\arctan \frac{2a+a^3}4-\frac2{3a}\ln\frac{2+2a+a^2}4,$$where $a:=\sqrt{2\surd3}$.

$\endgroup$
  • $\begingroup$ Your version evaluates to $1.2449\dots$ while the value of the series is $1.4639\dots$. Kindly re-check the expression. $\endgroup$ – Tito Piezas III Nov 12 '16 at 10:56
  • $\begingroup$ Thank you @TitoPiezasIII. The corrected version does now evaluate to $1.463939186$ on my calculator. $\endgroup$ – John Bentin Nov 12 '16 at 17:51
1
$\begingroup$

What you need is $\enspace\displaystyle \arctan x+\arctan y=\arctan\frac{x+y}{1-xy}\enspace$ because the values in front of the two $\arctan$ are equal of the results of wolframalpha. Also the values in front of the two $\ln$ are the same.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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