Is the Sequence Defined by x_n+1 = x_n/2 + 1 Bounded Above by 2?

  • Thread starter Thread starter cragar
  • Start date Start date
  • Tags Tags
    Proof Sequence
Join the discussion
Registration is free. Start your own thread to ask a follow-up.
8 replies · 2K views
cragar
Messages
2,546
Reaction score
3

Homework Statement


Show that the sequence [itex](x_1,x_2,x_3,...)[/itex]
defined by; Let [itex]x_1=1[/itex] for each [itex]n \in \mathbb{N}[/itex]
[itex]x_{n+1}= \frac{x_n}{2}+1[/itex]
[itex]x_2=\frac{3}{2}[/itex]
Show that this sequence is bounded above by 2; that is prove that [itex]x_n\leq2[/itex] for all [itex]n\in\mathbb{N}[/itex]

The Attempt at a Solution


This seems weird to me because it doesn't seem like it would be bounded above by 2, I could find some x that was bigger that 2. unless I don't understand the question.
 
Physics news on Phys.org
The statement that it's bounded from above by 2 means precisely that all terms in the sequence are ≤2. If you can find a term that's >2, then you have proved that it's not bounded from above by 2.

I suggest that you try to determine what the consequences would be if there's an integer N≥2 such that ##x_N>2##.
 
Last edited:
okay I should have looked at it closer, it does look like it will stay below 2. so should I try to manipulate it into a form that will work and then use induction. thanks for the replies by the way.
 
cragar said:

Homework Statement


Show that the sequence [itex](x_1,x_2,x_3,...)[/itex]
defined by; Let [itex]x_1=1[/itex] for each [itex]n \in \mathbb{N}[/itex]
[itex]x_{n+1}= \frac{x_n}{2}+1[/itex]
[itex]x_2=\frac{3}{2}[/itex]
Show that this sequence is bounded above by 2; that is prove that [itex]x_n\leq2[/itex] for all [itex]n\in\mathbb{N}[/itex]

The Attempt at a Solution


This seems weird to me because it doesn't seem like it would be bounded above by 2, I could find some x that was bigger that 2. unless I don't understand the question.

You can arrive at a nice contradiction by "going backward".

Suppose there was some k (k > 1) such that [itex]x_k = 2 + \epsilon[/itex], where [itex]\epsilon \geq 0[/itex].

Calculate [itex]x_{k-1}, x_{k-2}[/itex] (observe that they're increasing as you're going backward) and derive the general term [itex]x_{k-r}[/itex]. Prove this by induction (really easy).

Now set [itex]r = k-1[/itex]. What do you get?
 
Even more easily, using the same approach I detailed you can prove the "forward" case by induction. If you show that [itex]x_k < 2 \Rightarrow x_{k+1} < 2[/itex], you're done.
 
can I do this. [itex]x_1=1[/itex] so 1 is less than 2 so the base case is confirmed.
now [itex]2(x_{n+1}-1)=x_n\leq2[/itex]
then say that [itex]2(x_{n+1}-1)\leq2[/itex]
the divide both sides by 2 and then add the 1 so now we have
[itex]x_{n+1}\leq2[/itex]
so does this work, kinda seems recursive
 
cragar said:
can I do this. [itex]x_1=1[/itex] so 1 is less than 2 so the base case is confirmed.
now [itex]2(x_{n+1}-1)=x_n\leq2[/itex]
then say that [itex]2(x_{n+1}-1)\leq2[/itex]
the divide both sides by 2 and then add the 1 so now we have
[itex]x_{n+1}\leq2[/itex]
so does this work, kinda seems recursive
This is a good way to do it. Since you have found one solution, I don't mind telling you another. This is what I had in mind before:

For all n such that ##x_n>2##, we have ##\frac{x_{n-1}}{2}+1<2##, which implies that ##x_{n-1}>2##. So if one term is >2, every one that came before it is too. This contradicts ##x_1\leq 2##.
 
interesting idea fredrik, thanks to everyone who made a post.