Can $\prod_{n=0}^{\infty}\left(1+ \frac{1}{2^n}\right)$ be expressed in terms of the Pell constant?

101 Views Asked by At

In a comment to this question here, I made, based on my "defective" memory, the claim that

\begin{equation} \prod_{n=0}^{\infty}\left(1+ \frac{1}{2^n}\right)=\frac{2}{1-P}, \end{equation} where $P$ is the Pell constant defined as:

\begin{equation} P=1-\prod_{n=0}^\infty\left(1-\frac1{2^{2n+1}}\right) \end{equation}

and, luckily, got corrected by user mrtaurho. However, although my underlying assumption was false, the result still appears to be correct (by a numerical check).

As I don't believe in coincidences but find this relation interesting, I thought I'd ask whether there is anybody here who could prove or disprove this relation?

Thank you very much in advance for your interest.

1

There are 1 best solutions below

2
On

This is pretty easy to see, if you ignore matters of convergence (but everything converges absolutely anyways so it's fine):

We just want to prove that $$\prod_{m=0}^\infty\left(1+\frac1{2^m}\right)\prod_{n=0}^\infty\left(1-\frac1{2^{2n+1}}\right)=2.$$

which is just

$$\prod_{m=1}^\infty\left(1+\frac1{2^m}\right)\prod_{n=0}^\infty\left(1-\frac1{2^{2n+1}}\right)=1.$$

We match up terms -- a single $1-\frac1 {2^{2n+1}}$ term gets paired with all $m$ such that $2n+1|m$. In other words, we arrange the product to:

$$\prod_{n=0}^\infty\left(\left(1-\frac1{2^{2n+1}}\right)\prod_{m=0}^\infty\left(1+\frac1{2^{2^m(2n+1)}}\right)\right)=1.$$

Since every factor in the infinite product on the left is just $1$, the two sides match like we wanted.