6
$\begingroup$

What is the value of this product? $$\prod_{n=1}^{\infty}\left(1+\frac{1}{2^n}\right)$$

Me and my friend came across this product. Is the product till infinity equal to $1$?

If no, what is the answer?

$\endgroup$
14
$\begingroup$

Is the product till infinity equal to $1$?

Certainly not! All the individual terms are greater than $1$. So if you multiply them together, you will always be increasing and cannot get back to $1$.

If no, what is the answer?

The product in question is $$ \prod_{n=1}^\infty (1 + x^n) $$ where $x = \frac12$. This product equals $$ \sum_{n=0}^\infty q(n) x^n $$ where $q(n)$ is the number of partitions of $n$ into distinct parts (each part $\ge 1$), and also equals $$ \prod_{n=1}^\infty \frac{1}{1 - x^{2n-1}} = \frac{\Phi(x^2)}{\Phi(x)} $$ (see Wikipedia), where here $\Phi$ is the Euler function, not to be confused with Euler's totient function. So your product is equal to $$ \boxed{\frac{\Phi(1/4)}{\Phi(1/2)} = 2.38423\ldots}. $$ I don't expect this can be simplified.

$\endgroup$
  • $\begingroup$ Can you please see his updated question and try to answer the same way? I am quite interested. $\endgroup$ – Win Vineeth Sep 13 '16 at 7:57
  • $\begingroup$ @WinVineeth I assume you're referring to the different product that came up in the comments, which was the problem the OP got this from? That is $2$, and it is answered here. $\endgroup$ – 6005 Sep 13 '16 at 8:00
  • 1
    $\begingroup$ Don't delete this answer. It's great and correctly answers the original question. If the OP wants to ask something different, they can simply open a new question. $\endgroup$ – Wood Sep 13 '16 at 8:05
3
$\begingroup$

In the link you provided, the product seems to be $$\prod_{n=0}^{\infty} \left (1+\frac{1}{2^{2^n}}\right)$$

Note that $$(1-\frac{1}{2})\prod_{n=0}^{\infty} (1+\frac{1}{2^{2^n}})=\lim_{n \to \infty} 1-\frac{1}{2^{2^{n+1}}}=1$$ From the fact that $$(1+\frac{1}{2^{2^k}})(1-\frac{1}{2^{2^k}})=1-\frac{1}{2^{2^{k+1}}}$$

$\endgroup$
3
$\begingroup$

$$\prod_{n\geq 1}\left(1+\frac{1}{2^n}\right)=\exp\sum_{n\geq 1}\sum_{m\geq 1}\frac{(-1)^{m+1}}{m 2^{mn}}=\exp\sum_{m\geq 1}\frac{(-1)^{m+1}}{m(2^m-1)} $$ where the last series is (rapidly) convergent by Leibniz' test. It follows that the original product is between $\exp\left(\frac{121}{140}\right)$ and $\exp\left(\frac{3779}{4340}\right)$. By truncating the shown series at $m=30$ we get $$\prod_{n\geq 1}\left(1+\frac{1}{2^n}\right)\approx 2.384231029.$$

$\endgroup$
1
$\begingroup$

Another way to obtain the same answer of 6005. If we take the log of this product and the Taylor expansion of log we get $$\sum_{n\geq1}\sum_{k\geq1}\frac{\left(-1\right)^{k+1}}{k2^{mk}}=\sum_{k\geq1}\frac{\left(-1\right)^{k}}{k\left(1-2^{k}\right)}=-\sum_{k\geq1}\frac{\left(-1\right)^{k}}{k2^{k}\left(1-\frac{1}{2^{k}}\right)} $$ $$=-\sum_{k\geq1}\frac{1}{k4^{k}\left(1-\frac{1}{4^{k}}\right)}+\sum_{k\geq1}\frac{1}{k2^{k}\left(1-\frac{1}{2^{k}}\right)} $$ and since we have $$\log\left(\Phi\left(q\right)\right)=-\sum_{k\geq1}\frac{q^{n}}{k\left(1-q^{n}\right)} $$ where $\Phi\left(q\right) $ is the Euler's function we have $$\sum_{n\geq1}\sum_{k\geq1}\frac{\left(-1\right)^{k+1}}{k2^{mk}}=\log\left(\frac{\Phi\left(1/4\right)}{\Phi\left(1/2\right)}\right). $$ As wrote by 6005, probably there is no simplification for this result. The result can be written also as a q-Pochhammer symbol $$\prod_{n\geq1}\left(1+\frac{1}{2^{n}}\right)=\left(-1;\frac{1}{2}\right)_{\infty}.$$

$\endgroup$
1
$\begingroup$

This is not a full answer. There are already full answers posted.

I can prove that the sequence $x_n=\prod_{i=1}^n\left(1+\frac{1}{2^i}\right)$ converges, is strictly increasing and less than $e$, so $\prod_{i=1}^{+\infty}\left(1+\frac{1}{2^i}\right)\le e$. Notice that $x_n>0$, $\forall n$, so $x_{n+1}=x_n\left(1+\frac{1}{2^n}\right)>x_n$, $\forall n$. $$x_1=1.5\le x_n=e^{\ln\prod_{i=1}^{n}\left(1+\frac{1}{2^i}\right)}=$$

$\ln\left(1+\frac{1}{n}\right)<\frac{1}{n}$, $\forall n\ge 1$, $n\in\mathbb Z$ is true. A more general inequality $e^x> x+1$, $\forall x\in\mathbb R$, $x\neq 0$ is also true.

$$=e^{\sum_{i=1}^n \ln\left(1+\frac{1}{2^i}\right)}<e^{\sum_{i=1}^n \frac{1}{2^i}}<e^{\sum_{i=1}^{+\infty}\frac{1}{2^i}}=e$$

WolframAlpha says that the infinite product is $\approx 2.384$.

$\endgroup$

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.