Telescoping product

Calculus Level 2

A ( x ) = x + 1 x + 1 x + 1 A(x)=x+\cfrac { 1 }{ x+\cfrac { 1 }{ x+\cfrac { 1 }{ \ddots } } } With A ( x ) A(x) defined as above, where the continued fraction goes on indefinitely, find the value of the infinite product 1 A ( 1 ) × 1 + 1 1 A ( 1 ) × 1 + 1 1 + 1 1 A ( 1 ) × . \dfrac { 1 }{ A(1) } \times \dfrac { 1+\frac { 1 }{ 1 } }{ A(1) } \times \frac { 1+\frac { 1 }{ 1+\frac { 1 }{ 1 } } }{ A(1) } \times \cdots . If your answer can be expressed as a + b c b , \dfrac { a+\sqrt { b } }{ c\sqrt { b } }, where a , b , c a, b, c are positive integers and b b is square-free, give your answer as 100 a + 10 b + c 100a+10b+c .


Bonus: Can you give a closed formula for y A ( y ) × y + 1 y A ( y ) × y + 1 y + 1 y A ( y ) × \dfrac { y }{ A(y) } \times \dfrac { y+\frac { 1 }{ y } }{ A(y) } \times \dfrac { y+\frac { 1 }{ y+\frac { 1 }{ y } } }{ A(y) } \times \cdots when it converges?


The answer is 152.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

3 solutions

Mark Hennings
Sep 24, 2017

For y > 0 y > 0 , define the sequence G n G_n by the recurrence relation G 0 = 0 , G 1 = 1 , G n + 1 = y G n + G n 1 ( n 1 ) G_0 \; = \; 0\;,\; G_1 \; = \; 1 \;, \; \hspace{2cm} G_{n+1} \; = \; yG_n + G_{n-1} \qquad (n \ge 1) Then it is a simple induction that the sequence of numerators x n x_n given by the recurrence relation x 1 = y , x n + 1 = y + x n 1 ( n 1 ) x_1 \; = \; y \;,\; \hspace{2cm} x_{n+1} \; = \; y + x_n^{-1} \qquad (n \ge 1) is given by the formula x n = G n + 1 G n n 1 x_n \; = \; \frac{G_{n+1}}{G_n} \hspace{2cm} n \ge 1 We can obtain a generalisation of Binet's formula, in that G n = 1 y 2 + 4 [ ( y + y 2 + 4 2 ) n ( y y 2 + 4 2 ) n ] = ϕ ( y ) n ( ϕ ( y ) 1 ) n y 2 + 4 n 0 G_n \; = \; \frac{1}{\sqrt{y^2+4}} \left[\left(\frac{y + \sqrt{y^2+4}}{2}\right)^n - \left(\frac{y - \sqrt{y^2+4}}{2}\right)^n\right] \; = \; \frac{\phi(y)^n - (-\phi(y)^{-1})^n}{\sqrt{y^2+4}} \hspace{2cm} n \ge 0 where ϕ ( y ) = y + y 2 + 4 2 > 1 \phi(y) \; = \; \frac{y + \sqrt{y^2+4}}{2} > 1 From this it is clear that A ( y ) = lim n x n = lim n G n + 1 G n = ϕ ( y ) A(y) \; = \; \lim_{n \to \infty} x_n \; = \; \lim_{n \to \infty}\frac{G_{n+1}}{G_n} \; =\; \phi(y) and that n = 1 x n A ( y ) = lim N n = 1 N x n A ( y ) = lim N G N + 1 G 1 ϕ ( y ) N = ϕ ( y ) y 2 + 4 = y + y 2 + 4 2 y 2 + 4 \prod_{n=1}^\infty \frac{x_n}{A(y)} \;= \; \lim_{N \to \infty}\prod_{n=1}^N \frac{x_n}{A(y)} \; = \; \lim_{N \to \infty}\frac{G_{N+1}}{G_1 \phi(y)^N} \; = \; \frac{\phi(y)}{\sqrt{y^2+4}} \; = \; \frac{y + \sqrt{y^2+4}}{2\sqrt{y^2+4}} so that, in general, a = y a=y , b = y 2 + 4 b = y^2+4 and c = 2 c=2 . In the case y = 1 y=1 , this makes the answer 152 \boxed{152} .

if y < 0 y < 0 , all the x n x_n terms change sign, and hence the limit A A also changes sign (so that A ( y ) = A ( y ) A(-y) = -A(y) ). This means that the sequence of finite products converges to y + y 2 + 4 2 y 2 + 4 \frac{|y| + \sqrt{y^2+4}}{2\sqrt{y^2+4}} for all nonzero y y .

Marco Brezzi
Sep 23, 2017

Define the truncated version of A ( x ) A(x) as

A k ( x ) = x + 1 x + x + 1 x k x’s A_k(x)=\underbrace{x+\dfrac{1}{x+\dfrac{\ddots}{x+\frac{1}{x}}}}_{k\text{ x's}}

The product we have to find is

P ( 1 ) = k = 1 A k ( 1 ) A ( 1 ) P(1)=\displaystyle\prod_{k=1}^{\infty}\dfrac{A_k(1)}{A(1)}

Claim:

A k ( 1 ) = F k + 1 F k A_k(1)=\dfrac{F_{k+1}}{F_k}

Where F k F_k is the k k -th Fibonacci number

Proof:

Base case: k = 1 k=1

A k ( 1 ) = 1 = 1 1 = F 2 F 1 A_k(1)=1=\dfrac{1}{1}=\dfrac{F_2}{F_1}

Induction step: k k + 1 k\Longrightarrow k+1

Since

A k + 1 ( 1 ) = 1 + 1 A k ( 1 ) A_{k+1}(1)=1+\dfrac{1}{A_k(1)}

It follows that

A k + 1 ( 1 ) = 1 + 1 F k + 1 F k = 1 + F k F k + 1 = F k + 1 + F k F k + 1 = F k + 2 F k + 1 c c c c \begin{aligned} A_{k+1}(1)&=1+\dfrac{1}{\frac{F_{k+1}}{F_k}}\\ &=1+\dfrac{F_k}{F_{k+1}}\\ &=\dfrac{F_{k+1}+F_k}{F_{k+1}}\\ &=\dfrac{F_{k+2}}{F_{k+1}}\phantom{cccc}\square \end{aligned}

Which also gives the value of A ( 1 ) A(1)

A ( 1 ) = lim k A k ( 1 ) = lim k F k + 1 F k = ϕ A(1)=\displaystyle\lim_{k\rightarrow\infty}A_k(1)=\displaystyle\lim_{k\rightarrow\infty}\dfrac{F_{k+1}}{F_k}=\phi

Considering a finite version of the product

P n ( 1 ) = k = 1 n A k ( 1 ) ϕ = 1 ϕ n ( F 2 F 1 F 3 F 2 F n + 1 F n ) = F n + 1 ϕ n P_n(1)=\displaystyle\prod_{k=1}^n \dfrac{A_k(1)}{\phi}=\dfrac{1}{\phi^n}\left(\dfrac{F_2}{F_1}\cdot\dfrac{F_3}{F_2}\cdot\cdot\cdot\dfrac{F_{n+1}}{F_n}\right)=\dfrac{F_{n+1}}{\phi^n}

Hence

P ( 1 ) = lim n P n ( 1 ) = lim n F n + 1 ϕ n = lim n 1 5 ( ϕ n + 1 ( ϕ ) ( n + 1 ) ) ϕ n c c c c c c c c c c Binet’s Formula = ϕ 5 = 1 + 5 2 5 = a + b c b \begin{aligned} P(1)&=\displaystyle\lim_{n\rightarrow\infty}P_n(1)\\ &=\displaystyle\lim_{n\rightarrow\infty}\dfrac{F_{n+1}}{\phi^n}\\ &=\displaystyle\lim_{n\rightarrow\infty}\dfrac{\frac{1}{\sqrt{5}}(\phi^{n+1}-(-\phi)^{-(n+1)})}{\phi^n}\phantom{cccccccccc}\color{#3D99F6}\text{Binet's Formula}\\ &=\dfrac{\phi}{\sqrt{5}}=\dfrac{1+\sqrt{5}}{2\sqrt{5}}=\dfrac{a+\sqrt{b}}{c\sqrt{b}} \end{aligned}

100 a + 10 b + c = 152 \Longrightarrow 100a+10b+c=\boxed{152}

Kevin Tong
Oct 8, 2017

First, let's try to write a closed form for A ( x ) A(x) . First we notice A ( x ) = x + 1 A ( x ) ( A ( x ) ) 2 = x A ( x ) + 1 ( A ( x ) ) 2 x A ( x ) 1 = 0 A(x)=x+\frac{1}{A(x)} \\ (A(x))^2=xA(x)+1 \\ (A(x))^2-xA(x)-1=0 By the quadratic formula, we get A ( x ) = x ± x 2 + 4 2 A(x)=\frac{x \pm \sqrt{x^2+4}}{2} Since in the original equation, we are adding a number to 1 1 , we can conclude that the ± \pm should be + + . Using our new closed form of A ( x ) A(x) , we get that A ( 1 ) = 1 + 1 2 + 4 2 = 1 + 5 2 A(1)=\frac{1+\sqrt{1^2+4}}{2}=\frac{1+\sqrt{5}}{2} Now, looking at the infinite product, we notice that the reciprocal of A ( 1 ) A(1) is multiplied an infinite number of times, so we can take it out and try to evaluate the numerator. Looking at the first few terms, we notice, whenever we go up to the other term, we are simply taking the reciprocal of the previous term and adding 1 1 . In other words a n = 1 a n 1 + 1 = a n 1 + 1 a n 1 a_n=\frac{1}{a_{n-1}}+1 = \frac{a_{n-1}+1}{a_{n-1}} Since the terms can always be written in fractional form, we can look at the numerator and denominator separately, so a n 1 + 1 a n 1 = p n 1 q n 1 + q n 1 q n 1 p n 1 q n 1 = p n 1 + q n 1 p n 1 where p is the numerator and q is the denominator \frac{a_{n-1}+1}{a_{n-1}}=\frac{\frac{p_{n-1}}{q_{n-1}}+\frac{q_{n-1}}{q_{n-1}}}{\frac{p_{n-1}}{q_{n-1}}}=\frac{p_{n-1}+q_{n-1}}{p_{n-1}} \textrm{ where p is the numerator and q is the denominator}\\ Notice, the numerator now forms the Fibonacci Sequence. Also, notice, the numerator of the first term and denominator of the second term cancel out when we multiply them, so this product is simply the infinith Fibonacci number. We can evaluate this as a limit, namely lim n 1 5 ( 1 + 5 2 ) n + 1 Which is an approximation of the nth Fibonacci Number \lim_{n \to \infty} \frac{1}{\sqrt{5}}( \frac{1+\sqrt{5}}{2})^{n+1} \textrm{ Which is an approximation of the nth Fibonacci Number} Now putting this into the overall product, we get lim n ( 2 1 + 5 ) n 1 5 ( 1 + 5 2 ) n + 1 = 1 5 ( 1 + 5 2 ) = 1 + 5 2 5 a = 1 , b = 5 , c = 2 , so 100 a + 10 b + c = 100 + 50 + 2 = 152 \lim_{n \to \infty} (\frac{2}{1+\sqrt{5}})^n \cdot \frac{1}{\sqrt{5}}(\frac{1+\sqrt{5}}{2})^{n+1}=\frac{1}{\sqrt{5}} \cdot (\frac{1+\sqrt{5}}{2})=\frac{1+\sqrt{5}}{2\sqrt{5}} \implies a=1, b=5, c=2, \textrm{ so } 100a+10b+c=100+50+2=\boxed{152}

Bonus: From what we did above, we can simply get that the closed formula for y A ( y ) × y + 1 y A ( y ) × = lim n ( 1 A ( y ) ) n × 1 5 ( 1 + 5 2 ) n + 1 = lim n ( 1 + 5 2 5 ) ( 1 + 5 y + y 2 + 4 ) n \frac{y}{A(y)} \times \frac{y+\frac{1}{y}}{A(y)} \times \dots = \lim_{n \to \infty} (\frac{1}{A(y)})^n \times \frac{1}{\sqrt{5}}(\frac{1+\sqrt{5}}{2})^{n+1}=\boxed{\lim_{n \to \infty}(\frac{1+\sqrt{5}}{2\sqrt{5}}) \cdot (\frac{1+\sqrt{5}}{|y|+\sqrt{y^2+4}})^n}

Your general y y formula is wrong. If y > 1 |y| > 1 , your result says that the limit is 0 0 , which is not the case. See my solution.

Mark Hennings - 3 years, 8 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...