# Investigating a Sequence

• May 17th 2008, 07:08 AM
jules027
Investigating a Sequence
The sequence is recursively defined by a1 = √(2), an+1 = √(2+an) for n ≥ 1.

The sequence {an} is bounded above. Call this upper bound L.

Show {an} is monotone increasing. Thus {an} has a limit. Why?

Find this limit?

Thank you sooo much for all the help guys!!!
• May 17th 2008, 07:56 AM
flyingsquirrel
Hi
Quote:

Originally Posted by jules027
The sequence is recursively defined by a1 = √(2), an+1 = √(2+an) for n ≥ 1.

The sequence {an} is bounded above. Call this upper bound L.

Show {an} is monotone increasing. Thus {an} has a limit. Why?

Let $\displaystyle f:x\mapsto\sqrt{2+x}$.

$\displaystyle a_{n+1}=\sqrt{2+a_n}=f(a_n)$ hence the sequence is monotone increasing iff $\displaystyle f$ is monotone increasing.
Quote:

Find this limit?
The limit
$\displaystyle l$ of the sequence is such that $\displaystyle f(l)=l$ because $\displaystyle f$ is continuous on $\displaystyle [\sqrt{2},\,L]$.
• May 17th 2008, 08:01 AM
TXGirl
Quote:

Originally Posted by jules027
The sequence {an} is bounded above. Call this upper bound L.

...{an} is monotone increasing. Thus {an} has a limit. Why?

You have by definition that a number L is the limit of a sequence $\displaystyle ({x_n:n\in\mathbb{N}})$ iff $\displaystyle \forall{\epsilon >0} \exists{N}\in\mathbb{N}:\forall{n>N}, {|{x_n}-L|}<\epsilon$

A sequence converges iff it has a limit.

You can see from the definition that every monotone bounded sequence has a limit and thus converges.

So your sequence $\displaystyle (a_n)$ being both monotone and bounded, necessarily converges and thus has a limit.
• May 17th 2008, 08:11 AM
jules027
Quote:

Originally Posted by flyingsquirrel
Hi

Let $\displaystyle f:x\mapsto\sqrt{2+x}$.

$\displaystyle a_{n+1}=\sqrt{2+a_n}=f(a_n)$ hence the sequence is monotone increasing iff $\displaystyle f$ is monotone increasing.
The limit $\displaystyle l$ of the sequence is such that $\displaystyle f(l)=l$ because $\displaystyle f$ is continuous on $\displaystyle [\sqrt{2},\,L]$.

So there would be no computations to reach the limit? It is just the statement,
"The limit $\displaystyle l$ of the sequence is such that $\displaystyle f(l)=l$ because $\displaystyle f$ is continuous on $\displaystyle [\sqrt{2},\,L]$."?
• May 17th 2008, 08:14 AM
Isomorphism
Quote:

Originally Posted by jules027
So there would be no computations to reach the limit? It is just the statement,
"The limit $\displaystyle l$ of the sequence is such that $\displaystyle f(l)=l$ because $\displaystyle f$ is continuous on $\displaystyle [\sqrt{2},\,L]$."?

Actually $\displaystyle f(a_n) = a_{n+1}$. If the limit of the sequence is L, then by letting n tend to infinity we get $\displaystyle f(L) = L$

$\displaystyle \sqrt{2+L} = L \Rightarrow L^2 - L - 2 = 0$
• May 17th 2008, 08:42 AM
flyingsquirrel
Quote:

Originally Posted by jules027

The sequence {an} is bounded above. Call this upper bound L.

Quote:

Originally Posted by Isomorphism
If the limit of the sequence is L ...

These two $\displaystyle L$ are not necessarily the same one.
• May 17th 2008, 03:11 PM
jules027
So how would I get at this limit exactly then? Thanks for the help!
• May 17th 2008, 08:43 PM
Isomorphism
Quote:

Originally Posted by Isomorphism
Actually $\displaystyle f(a_n) = a_{n+1}$. If the limit of the sequence is L, then by letting n tend to infinity we get $\displaystyle f(L) = L$

$\displaystyle \sqrt{2+L} = L \Rightarrow L^2 - L - 2 = 0$

Quote:

Originally Posted by flyingsquirrel
These two $\displaystyle L$ are not necessarily the same one.

flyingsquirrel means generally they need not be the same one. Since what I said could be misleading. You might think it works for any function.

However for $\displaystyle a_{n+1} = \sqrt{2 + a_n}$, we can limit both sides to infinity to get $\displaystyle \lim_{n \to \infty} a_{n+1} =\lim_{n \to \infty} \sqrt{2 + a_n} = \sqrt{2 + \lim_{n \to \infty} a_n}$

Since you have already proved that $\displaystyle (a_n)$ has a limit. Call it l. Its easy to see that $\displaystyle \lim_{n \to \infty} a_n = \lim_{n \to \infty} a_{n+1} = l$. So you can continue like what I did before.
• May 19th 2008, 02:51 PM
jules027
I am still confused about this one. Could somebody try and help? Maybe a step by step procedure of how to get at the answer? I almost have it but I am confused on the steps to get there. Thanks so much!!
• May 19th 2008, 03:02 PM
Mathstud28
Quote:

Originally Posted by Isomorphism
flyingsquirrel means generally they need not be the same one. Since what I said could be misleading. You might think it works for any function.

However for $\displaystyle a_{n+1} = \sqrt{2 + a_n}$, we can limit both sides to infinity to get $\displaystyle \lim_{n \to \infty} a_{n+1} =\lim_{n \to \infty} \sqrt{2 + a_n} = \sqrt{2 + \lim_{n \to \infty} a_n}$

Since you have already proved that $\displaystyle (a_n)$ has a limit. Call it l. Its easy to see that $\displaystyle \lim_{n \to \infty} a_n = \lim_{n \to \infty} a_{n+1} = l$. So you can continue like what I did before.

As Isomorphism showed

$\displaystyle \lim_{n\to\infty}a_{n+1}=\sqrt{2+\lim_{n\to\infty} }\Rightarrow{(\lim_{n\to\infty}a_{n+1})^2=2+\lim_{ n\to\infty}a_{n}}$

and since as $\displaystyle n\to\infty$ $\displaystyle a_n=a_{n+1}$

we say Let $\displaystyle a_n=\lambda$

So we get

$\displaystyle \lambda^2=2+\lambda\Rightarrow{\lambda^2-2-\lambda=0}$

Now just solve for $\displaystyle \lambda$
• May 19th 2008, 05:48 PM
jules027
So the final limit would be from (-1,2)?
• May 19th 2008, 05:51 PM
Mathstud28
Quote:

Originally Posted by jules027
So the final limit would be from (-1,2)?