1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Differentiable statements

  1. Feb 20, 2012 #1
    1. The problem statement, all variables and given/known data
    can someone help me to see whether the following statements choose or false

    a) if lim(X->infinite)f(x) exists and is finite and lim(X->infinite)f'(x)=b then b=0
    i think it is right but i don't know how to prove it

    b)if lim(X->infinite)f(x) exists and is finite then lim(X->infinite)f'(x)=0
    i have no idea guesses right

    c)lim(X->infinite)f'(x)=0 then lim(X->infinite)f(x) exists
    counterexample: f(x)=lnx

    d))lim(X->infinite)f'(x)=0 then lim(X->infinite)f(x)/x=0
    no idea as well
  2. jcsd
  3. Feb 20, 2012 #2


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    For (a), suppose b > 0. Then f'(x) > b/2 for all sufficiently large x. What does this imply about f(x) for sufficiently large x?

    For (b), think about a function that has jump discontinuities at arbitrarily large values of x. Then f'(x) will not exist at the jumps, so lim(X->infinite)f'(x) won't exist, either. Can you come up with an example like this that still satisfies "lim(X->infinite)f(x) exists and is finite"?

    For (c), your counterexample is fine.

    For (d), given epsilon > 0, you have |f'(x)| < epsilon for all sufficiently large x. Pick a specific such x, call it x0. Then for all x beyond that point, f(x) must be bounded between two lines, of slope +/- epsilon. Work out the equations for those lines, and see what this implies about f(x)/x for x > x0.
  4. Feb 20, 2012 #3
    sorry ,for a) i still can understand how to prove it , can u give me more details then cfor b) my example is y=[x] the integer part of x
    and for d) i think that for all sufficiently large x , take x0, for all x beyond that point then then
    -epsilon<f'(x)<+epsilon , then -epsilonx<f(x)>epsilonx then ,f(x)/x for all x0 bounded by epsilon that is 0 does it make sense?
  5. Feb 20, 2012 #4


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Your counterexample for (b) doesn't work, because [x] goes to infinity as x goes to infinity. You have to make the jumps smaller and smaller as x goes to infinity.

    I didn't understand your argument for (d). Can you write it out with a bit more detail?

    For (a), suppose b > 0. Then for sufficiently large x, f'(x) > b/2. More precisely, there exists an x0 such that f'(x) > b/2 for all x > x0. Therefore, for all x > x0, you have f(x) > (b/2)x + f(x0). [Why?] This means that f(x) -> infinity as x -> infinity. [Why?]

    The b < 0 case is almost identical.
  6. Feb 21, 2012 #5
    for b)i can create a function s.t. in some interval it is a straight line ,but when x is large enougn then f(x)=1 100<x<10000 f(x)=1/2 10000<x<!00000 f(x)=1/3 100000<x<1000000000.....

    for d) we have -epsilon<f'(x)<epsilon for x >x0 , then -epsilonx-f(x0)<f(X)<epsilonx+f(x0) since then the slope of f(X) between -epsilon and epsilon.then diveided then inequality on both side by x assume x >0 then -epsilon-f(x0)/x<f(X)<epsilon+f(X0)/x then as x->infinity then f(X0)/x=0 so we get f(X)/x convergent to 0 done
  7. Feb 21, 2012 #6


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Yes, that works. Or, if you wanted a function that is easier to express, you could use something like f(x) = 1/[x].

    Actually it would be [itex]-\epsilon x + f(x_0) < f(x) < \epsilon x + f(x_0)[/itex] (plus, not minus, in the first expression).

    The rest looks OK.

    You should probably provide some justification for how you transformed

    [tex]-\epsilon < f'(x) < \epsilon[/tex]


    [tex]-\epsilon x + f(x_0) < f(x) < \epsilon x + f(x_0)[/tex]

    i.e. Did you use the fundamental theorem of calculus? And if so, what happens if [itex]f'[/itex] is not integrable?
  8. Feb 21, 2012 #7

    no, but anyway thank you for helping me to solve the problem
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook