Let $A = \{(\sqrt{5} + 1)/2\}^{5}$ and let $\alpha,\beta$ be positive reals such that $\alpha\beta = \pi^{2}/5$. Then it is known that $$\left\{A + R^{5}(e^{-2\alpha})\right\}\left\{A + R^{5}(e^{-2\beta})\right\} = 5\sqrt{5}A$$ where $$R(q) = \cfrac{q^{1/5}}{1 + \cfrac{q}{1 + \cfrac{q^{2}}{1 + \cfrac{q^{3}}{1 + \cdots}}}}$$ is the Rogers-Ramanujan continued fraction. A proof of the above result is available on my blog http://paramanands.blogspot.com/2013/09/values-of-rogers-ramanujan-continued-fraction-part-2.html

Using the above I am trying to evaluate $R(e^{-2\pi/5})$ using the already known value of $R(e^{-2\pi})$ given by $$R(e^{-2\pi}) = \sqrt{\frac{5 + \sqrt{5}}{2}} - \frac{\sqrt{5} + 1}{2}$$

My calculation yields the value $$R(e^{-2\pi/5}) = \frac{\sqrt{5} + 1}{2}\left(3\sqrt{85 - 38\sqrt{5}} - \frac{17\sqrt{5} - 37}{2}\right)^{1/5}$$ and I have checked using calculator that both the values of $R(e^{-2\pi})$ and $R(e^{-2\pi/5})$ shown above satisfy the $\alpha,\beta$ relation mentioned above.

My problem is that the value of $R(e^{-2\pi/5})$ does not look like a unit. On the other hand the value $R(e^{-2\pi})$ is a unit and there is a general theorem that all values of $R(q)$ are units for $q = e^{-\pi\sqrt{n}}$ where $n$ is positive rational. Probably I am making some calculation mistakes or may be I am not able to figure out how the expression for $R(e^{-2\pi/5})$ is a unit. Unfortunately I have not been able to find any online reference to check this value of $R(e^{-2\pi/5})$ and I am stuck with these crazy radical manipulations. Please help me out.

  • $\begingroup$ Could you please define unit in this context? $\endgroup$
    – Daniel R
    Sep 15 '13 at 19:15
  • $\begingroup$ To Daniel: Roughly an algebraic integer $\alpha$ is said to be a unit if there is another algebraic integer $\beta$ such that $\alpha\beta = 1$. $\endgroup$
    – Paramanand Singh
    Sep 16 '13 at 3:30

It turns out that if I move the factor $(\sqrt{5} + 1)/2$ inside the 5th root things look much better and then $$R(e^{-2\pi/5}) = \left(3\sqrt{\frac{5 + \sqrt{5}}{2}} - \frac{9 + \sqrt{5}}{2}\right)^{1/5}$$

From this form it is easy to see the unitness. Clearly we have $$\left(3\sqrt{\frac{5 + \sqrt{5}}{2}} - \frac{9 + \sqrt{5}}{2}\right)\left(3\sqrt{\frac{5 + \sqrt{5}}{2}} + \frac{9 + \sqrt{5}}{2}\right) = 1$$

It was silly of me not to have thought of merging the factor $(\sqrt{5} + 1)/2$ inside the 5th root.


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.