Proving convergence of recursive sequence.

Click For Summary
The discussion focuses on proving the convergence of the recursive sequence defined by x_n = (1/2)(x_{n-1} + c/x_{n-1}) for c > 0. Participants suggest using the theorem that states a monotone bounded sequence converges, prompting exploration of whether the sequence is increasing or decreasing and identifying bounds. Examples illustrate that the sequence may not consistently behave as expected, as shown with specific values for c and x_0. Additionally, the use of graphical tools like cobweb plots is recommended to gain intuition about the sequence's behavior. Overall, the conversation emphasizes the need for a rigorous proof of convergence while exploring various approaches.
Darth Frodo
Messages
211
Reaction score
1

Homework Statement



Prove for c>0 the sequence {x_n} = \frac{1}{2}(x_{n-1} + \frac{c}{x_{n-1}}) converges.


The Attempt at a Solution



This is proving difficult, I have never dealt with recursive sequences before. Any help would be appreciated. Thanks.
 
Physics news on Phys.org
So, I have an idea, but I'm not sure if it proves convergence.

If x_{n} = g(x_{n-1}) then all I have to do it let x = g(x).EDIT: Upon further thought, it does not as convergence => x_{n} = g(x_{n-1}). It is not a iff situation. Back at square one.
 
Last edited:
You have the following handy theorem:

If ##(x_n)_n## is a monotone increasing/decreasing sequence and if the ##(x_n)_n## are bounded, then the ##(x_n)_n## converge.

I would start by trying to apply this one. Can you show the sequence is increasing/decreasing? Can you find a lower/upper bound?
 
Darth Frodo said:
So, I have an idea, but I'm not sure if it proves convergence.

If x_{n} = g(x_{n-1}) then all I have to do it let x = g(x).

This is a nice remark, because if the sequence converges, you can find out the limit point this way. Indeed, you have ##x_n = g(x_{n-1})##. Thus

x = \lim_{n\rightarrow +\infty} x_n = \lim_{n\rightarrow +\infty} g(x_{n-1}) = g(x)

So can you figure out what the limit point ##x## is? (of course you still have to prove it actually converges!)
 
So this is what I have so far.

x_0 = r r is a non-zero real number

x_{n+1} = x_n(\frac{1}{2} + \frac{c}{2x_n^2})Case 1: ##r^2 < c~ \Rightarrow## sequence is increasing

Case 2: ##r^2 > c ~\Rightarrow## sequence is decreasing
 
Last edited:
Darth Frodo said:
So this is what I have so far.

x_0 = r r is a non-zero real number

x_{n+1} = x_n(\frac{1}{2} + \frac{c}{2x_n^2})


Case 1: ##r^2 < c~ \Rightarrow## sequence is increasing

Case 2: ##r^2 > c ~\Rightarrow## sequence is decreasing

Not quite:

Suppose c = 4 and x_0 = 1, then x_1 = 5/2 and x_2 = 89/40, so neither increasing nor decreasing!

You're on the right track, though.
 
Darth Frodo said:

Homework Statement



Prove for c>0 the sequence {x_n} = \frac{1}{2}(x_{n-1} + \frac{c}{x_{n-1}}) converges.


The Attempt at a Solution



This is proving difficult, I have never dealt with recursive sequences before. Any help would be appreciated. Thanks.

To see what is happening, look at the graph of
y = f(x), \text{ where } f(x) = \frac{1}{2} \left( x + \frac{c}{x} \right)
Look up material about "cobweb plots"; see, eg., http://en.wikipedia.org/wiki/Cobweb_plot . This will give you some intuition about what is happening; it is not yet a 'proof', but may help get you started.
 

Similar threads

Replies
4
Views
2K
Replies
1
Views
1K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 13 ·
Replies
13
Views
2K
  • · Replies 13 ·
Replies
13
Views
2K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 17 ·
Replies
17
Views
1K
  • · Replies 4 ·
Replies
4
Views
2K