Sequences and continuous functions

Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
1 reply · 2K views
the_kid
Messages
114
Reaction score
0

Homework Statement


a) Let {s[itex]_{n}[/itex]} and {t[itex]_{n}[/itex]} be two sequences converging to s and t. Suppose that s[itex]_{n}[/itex] < (1+[itex]\frac{1}{n}[/itex])t[itex]_{n}[/itex]

Show that s [itex]\leq[/itex]t.

b) Let f, g be continuous functions in the interval [a, b]. If f(x)>g(x) for all x[itex]\in[/itex][a, b], then show that there exists a positive real z>1 such that f(x)[itex]\geq[/itex]zg(x) for all x[itex]\in[/itex][a, b].

Homework Equations


The Attempt at a Solution



Ok, so I've already done part a. I'm trying to figure out part b. I think my ideas are on the right track, but I'm looking for some help fleshing them out a bit more.

Argue by contradiction. Suppose there does not exist z such that f(x)[itex]\geq[/itex]zg(x). I'm not sure exactly what this implies in terms of deriving a contradiction; some general guidance would be appreciated it. Also, I'm using the definition that the notion that f(x) is continuous at c is equivalent to the following:

For every sequence {s[itex]_{n}[/itex]} in the domain of f converging to c, one has

lim, n-->[itex]\infty[/itex] f(s[itex]_{n}[/itex])=f(c). I somehow want to construct an analogous argument to that in part a, but I'm not sure how the z is similar to the (1+[itex]\frac{1}){n}) term. Can anyone help with this?[/itex]
 
Physics news on Phys.org
I think this can go easier, define h:

[tex]f(x)-g(x)=h(x)[/tex]

because f,g are continuous in [a,b], h is continuous in [a,b] but then h(x) has a minimum >0 in [a,b] call it c thus:

[tex]f \geq g+c = g \left( 1+\frac{c}{g(x)} \right)[/tex]

This holds for all x in [a,b] thus:

[tex]f \geq g+c = g \left( 1+\frac{c}{ ||g(x)||_{\infty} } \right) = zg(x)[/tex]with [tex]||g(x)||_{\infty} =sup_{x \in [a,b]} g(x)[/tex] and [tex]z= 1+\frac{c}{ ||g(x)||_{\infty} }[/tex]Something like this should work ( I think the part with the inf term should be replaced by something else).
 
Last edited: