For a_{n+1} = 2 - 1/(2a_{n}), a_[1} = 1, show that the sequence is bounded below by 1, that the sequence is increasing, and find its limit.
\[a_{n+1} = 2 - \frac{1}{2a_{n}}\] \(a_{1}=1\)
I have the limit, that's of no issue, but I ran into troubles trying to show it was bounded below by 1 and increasing. I think it's supposed to be done by induction, but still was having issues.
will you work on this tomorrow? (because I'll look tomorrow before noon in my textbook)
Ill be working on it on and off, so if no-one is able to help me before then.
assume \(\large a_k \ge 1\) \(\implies \dfrac{1}{a_k} \le 1 \implies \dfrac{1}{2a_k} \le \dfrac{1}{2} \implies- \dfrac{1}{2a_k}\ge -\dfrac{1}{2} \implies 2-\dfrac{1}{2a_k}\ge 2-\dfrac{1}{2} \ge 1 \)
Is that all that needs to be done to show the first part? O.o
that will do for bounded below part
btw thats only part of induction setup, i did not include base case and all ..
So you didnt need to do a P(k+1) case?
what i did above itself is the P(k+1) case
assuming \(\large a_k \ge 1\), it was shown that \(\large a_{k+1}\ge 1\)
You never needed to have any mention of \(a_{k+1}\)? I would think just using \(a_{k}\) alone would be making the assumption that it's increasing without proving it.....?
i dont need to. but you must mention that the last step is same as \(a_{k+1}\) in your proof
\[\implies \dfrac{1}{a_k} \le 1 \implies \dfrac{1}{2a_k} \le \dfrac{1}{2} \implies- \dfrac{1}{2a_k}\ge -\dfrac{1}{2} \implies \color{Red}{2-\dfrac{1}{2a_k}}\ge 2-\dfrac{1}{2} \ge 1\]
that expression is same as \(\large a_{k+1}\)
Is that always what you're trying to do? You don't ever need an explicit statement for \(a_{k+1}\)?
you're not getting me, what im giving you is just an idea. not the complete proof
any induction proof requires 3 steps: 1) base case 2) induction hypothesis(k) 3) induction step (k+1)
i just gave you 2nd and most of the last steps
you still need to fill in the gaps
I wasn't sure how much of it you were giving me or not giving me, so I guess that's why I was confused. Had no idea where along in the proof you were. But still, am I trying to eventually work my way down to what you had typed up?
are you still not sure ?
Like, for the k+1 step, I assumed you had to start with: \[a_{k+2} = 2 - \frac{ 1 }{ 2a_{k+1} }\] From there, I wasnt sure if I needed to get an explicit expression for \(a_{k+1}\) or \(a_{k+2}\) or \(a_{k}\). I do see what you did, but im confirming if that means im supposed to change the k+1 expression into something with \(a_{k}\) or how im exactly supposed to work it.
In k+1 step, you need to show the given statement works for n=k+1 whenever it works for n=k
you can start with induction hypothesis, massage it a bit and show that the given statement works for n=k+1 too or you can start with k+1 step, and use the induction hypothesis somewhere in between to show that the given statemetn works for n=k+1 too
it doesn't matter how you start
Okay, then is what I have typed up a correct k+1 beginning?
what do you mean by a correct k+1 beginning ?
starting with induction hypothesis is more natural here
I mean is \[a_{k+2} = 2-\frac{ 1 }{ a_{k+1} }\] what the correct k+1 term would be? Like, I know base case, assume true for k, show for k+1, but I felt like I was doing k+1 wrong when doing it on my own. I was trying to manipulate the above and maybe substitute things in, but it was getting me nowhere. As for starting with the induction hypothesis, then that means I'm starting with \[a_{k+1} = 2 - \frac{ 1 }{ 2a_{k} }\] correct? This is where some of my confusion comes in, though. If I start with the induction hypothesis, then \(a_{k+1}\) is right there, it feels like there's no k+1 step to show. If you jump to what I posted first, I have that \(a_{k+2}\) term which I don't think Im allowed to say anything about. It's just weirding me out trying to do induction with a recurrence.
Induction hypothesis : \(\large a_{k+1}\ge 1 \) k+1 step : \(\large a_{k+2}\ge 1\)
induction hypothesis assumes that what you want to prove is true for n = k, so you're right, we need to shift the index
Induction hypothesis : \(\large a_{k+1}\ge 1\) k+1 step : \( a_{k+1}\ge 1 \implies \dfrac{1}{a_{k+1}} \le 1 \implies \dfrac{1}{2a_{k+1}} \le \dfrac{1}{2} \\~\\\implies- \dfrac{1}{2a_{k+1}}\ge -\dfrac{1}{2} \implies 2-\dfrac{1}{2a_{k+1}}\ge 2-\dfrac{1}{2} \ge 1\)
I thought k+1 step used \(a_{k+2}\)? :(
yes, but we don't need to start literally with \(a_{k+2}\) statement.
we can start with induction hypothesis and arrive at \(\large a_{k+2}\) statemetn
Induction hypothesis : \(\large a_{k+1}\ge 1\) k+1 step : \( a_{k+1}\ge 1 \implies \dfrac{1}{a_{k+1}} \le 1 \implies \dfrac{1}{2a_{k+1}} \le \dfrac{1}{2} \\~\\\implies- \dfrac{1}{2a_{k+1}}\ge -\dfrac{1}{2} \implies\color{Red}{ 2-\dfrac{1}{2a_{k+1}}}\ge 2-\dfrac{1}{2} \ge 1\)
that gets replaced with \(\large a_{k+2}\)
Oh....Well, so why is \(a_{k+1}\) equal to 1? you can jump from \(2-\frac{1}{2a_{k+1}}\) to \(2-\frac{1}{2}\)?
i have just added 2 to both sides of the preceding inequality
Oh, you didnt actually substitute \(a_{k+1}\)
ahh i see now why its bothering you so much ;p
yes no substitution, we have just started with induction hypothesis inequality and manipulated it
And the induction hypothesis is only \(a_{k+1} \ge1\)?
yup
but an induction proof requires a basis step also, i doubt you never worked induction proofs before hmm
Alright. So we never really need to deal with the actual "\(a_{k+2}\)", we just make use of what it's equal to. And no, Ive done induction before, just the recursion stuff, coming up with \(a_{k+2}\) stuff, that weirded me out. It's like, you can ask me to show 1 + 2 + ... + n = n(n+1)/2 and that doesnt bug me. Base case, assume true for p(k), do p(k+1) step. But even though I was plugging in k+1, I could break apart "n's" and separate whatever I wanted. "\(a_{k+2}\)" I can't factor or break apart as a statement alone, so I didnt know what I was supposed to do with it, it seemed like some crazy obstruction, lol.
well he actually did show a_(k+2) >=1
remember that a_(k+2)=2-1/[2*a_(k+1)]
maybe lets try and work it by starting with \(\large a_{k+2}\) also
\[a_{k+1}\ge 1 \implies \dfrac{1}{a_{k+1}} \le 1 \implies \dfrac{1}{2a_{k+1}} \le \dfrac{1}{2} \\~\\\implies- \dfrac{1}{2a_{k+1}}\ge -\dfrac{1}{2} \implies\color{Red}{ 2-\dfrac{1}{2a_{k+1}}}\ge 2-\dfrac{1}{2} \ge 1 \] --- \[\text{ We had }a_{n+1}=2-\frac{1}{2a_n} \\ \text{ so } a_{n+2}=2-\frac{2}{a_{n+1}}\]
Yeah, that would be cool.
oops my fraction thing didn't work out very well
that one 2 is suppose to be on bottom
Yeah, I realize that now, naya (or however you prefer to be called), I just was still getting tripped up on whether or not I need the \(a_{k+2}\) part or if I needed an explicit expression of \(a_{k+2}\)= ? or \(a_{k+1}\)= ? So I suppose I needed to see how the form of a solution step looked so I wasnt messing with things trying to get ___ = ___. I'm getting a better idea of what you're trying to do, though.
assumption : \(\large a_{k+1}\ge 1\) \(\large a_{k+2} = 2 - \dfrac{1}{2a_{k+1}} \ge 2 - \dfrac{1}{2\cdot 1} \) nope that doesn't look smooth.
Which is probably why Im here banging my head on the table :D It looks like dead end brain cramp-ville o_o
what is unsmooth about that?
i knw that feeling :)
\[a_{k+1} \ge 1 \\ \frac{1}{a_{k+1}} \le \frac{1}{1}=1 \\ \frac{1}{a_{k+1}} \le 1 \\ \frac{1}{2a_{k+1}} \le \frac{1}{2} \\ -\frac{1}{2a_{k+1}} \ge \frac{-1}{2} \\ a_{k+2}=2-\frac{1}{2a_{k+1}} \ge 2-\frac{1}{2}\] It is a little more ugly than just working from inductive hypothesis
that second proof ends abruptly in two steps and leaves so much for the reader to think about the transition to second step
you also have to include reasoning that was already included above from the inductive hypothesis
Yeah, working from the induction step makes much more sense. I wouldnt know what to do with the 2nd proof. And that's what I was trying to work with the whole time, trying substitutions and such. But I guess I didnt know you could manipulate the original induction hypothesis. It seemed like the pattern was you start with the k+1 step and somewhere in the process of manipulating it you throw in the induction hypothesis in order to do some sort of substitution. So I guess I wouldve never known to try it this way.
Alrighty, so after all that, I now would have to show increasing, haha. Is that kind of going to work in the same way, trying to start with the induction hypothesis and build your way up to the k+1 case?
to show the sequence is increasing, we prove \(\large \forall n\ge 1, ~ a_{n+1}\gt a_n \)
Yeah, you typed it out before me xD Out of curiosity, could I also show: \[(a_{n+1})^{2} > (a_{n})^{2}\]? I dont mean to say this is the correct thing to do, Im just curious if youre allowed to do this?
sure, if thats easy.. because when x an y are positive : \[\large x^2 \gt y^2 \implies x \gt y\]
but we can mimic the same steps for bounded below proof here
start with induction hypothesis, manipulate it and try to arrive at : \( a_{k+1}\gt a_k \implies a_{k+2} \gt a_{k+1}\)
Okay, another question then. If \[a_{k+1} = 2 - \frac{ 1 }{ 2a_{k} }\]we cant ever say: \[a_{k} = 2-\frac{ 1 }{ 2_{k-1} }\]can we?
Oops, forgot the a in there. \[a_{k} = 2 - \frac{ 1 }{ 2a_{k-1} }\]
why not ? you're just shifting the index. so we should be okay as long as \(a_{k-1}\) dont go below the first term
Ah, okies :) Well, Im not sure what the use of it would be in this problem. But using the same idea as before, I was able to figure it out, got it to work out ^_^ Okay, thanks a lot and thanks for being patient :3
np :) finding the limit of sequence should be pretty straightforward i guess
Yeah, I got that no problem, lol. That was the easy part. See ya later :P
Join our real-time social learning platform and learn together with your friends!