Register to reply 
Proving The Continuous Theorem for Sequences 
Share this thread: 
#1
May1007, 11:01 PM

P: 159

1. The problem statement, all variables and given/known data
4.8 Show the following continuous theorem for sequences: if [tex]a_n \rightarrow L[/tex] and f is a real valued function continuous at L, then [tex]bn = f(a_n) \rightarrow f(L)[/tex]. 2. Relevant equations No real relevant equations here. Just good old proof I'm thinking. 3. The attempt at a solution Well, I stared at this for an hour today. I was able to complete the rest of the assignment but this one has me stumped. I realize that [tex]\displaystyle\lim_{n\rightarrow\infty}a_n=L[/tex] and that for a realvalued function to be continuous at L that [tex]\displaystyle\lim_{x\rightarrow x_0}f(x)=f(x_0)=L.[/tex] I don't know what to do from here though. How do I get f(L) from f(x0)=L, and then get f(a_n) from just plain old a_n. This thing makes intuitive sense to me; it's blatantly obvious it's right  proving it has ... well.. proven to be really hard! 


#2
May1007, 11:11 PM

Sci Advisor
HW Helper
P: 2,537

When something seems like it should be obvious, you might want to write out some definitions. Epsilons, deltas, and all that.



#3
May1007, 11:18 PM

P: 159

Well that's the thing; I thought I could just approach it nicely, but the concepts seem to be two different things; it's as if I want to prove a finite limit at infinity. There is a proof I thought was similar to it, in the notes he gave us. However, I fleshed out a few epsilon delta definitions; drew a nice picture so I could see what was going on. My main problem is I have no clue how to get to f(a_n) from a_n, and from f(x0)=L to f(L).



#4
May1107, 12:33 AM

Sci Advisor
HW Helper
P: 2,537

Proving The Continuous Theorem for Sequences
It looks like you might be getting confused by using two different [itex]L[/itex]'s. Typically, the [itex]a_n[/itex] and [itex]b_n[/itex] will not converge to the same value.
Really, this is straightforward stuff... given some [itex]\epsilon[/itex] greater than [itex]0[/itex] can you show that there is some [itex]N[/itex] so that [itex]n>N \Rightarrow f(a_n)f(L) < \epsilon[/itex]? 


#5
May1107, 01:08 AM

P: 159

Hmm, I must not be seeing it...
I get that it's going to follow the basic form of a proof. For all epsilon > 0, there exists an N such that, for all n>N, then f(a_n)f(L)<epsilon. The N is more than likely going to be some integer part of some number plus one. I get how all that works. The problem is... I'm used to an f(x)L, where i know what the L and f(x) are. I have no clue as to what f(L) or f(a_n) are, and I don't see how I can understand what they are from the given information. Sorry; I hate missing obvious stuff =( 


#6
May1107, 01:37 AM

Sci Advisor
HW Helper
P: 2,537

It's a little abstract. Do you know if there's a [itex]\delta[/itex] so that:
[tex]a_nL<\delta \Rightarrow f(a_n)f(L) < \epsilon[/tex] 


#7
May1107, 01:57 AM

P: 159

I'm following you and it looks like I'll be proving a limit at a point, but I'm still stuck with how we can relate a_n to f(a_n), and L to f(L). If I can somehow extract the a_n and the L from the f, I could do it...again, thanks for the help so far and sorry I'm still not quite getting what to do...



Register to reply 
Related Discussions  
Proving a function is continuous  Calculus  11  
Proving asymptotics to sequences  Linear & Abstract Algebra  1  
Proving a function is continuous  Calculus & Beyond Homework  1  
Proving a function is continuous  Calculus & Beyond Homework  2  
Proving limits for converging sequences  Calculus  1 