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!

Calculous proof question

  1. Feb 27, 2010 #1
    1. The problem statement, all variables and given/known data
    Let f be a function from R to R such that [tex] f'(x)>f(x), f(0)=1 [/tex] for every x. prove that for every [tex] x>0 f(x)>e^x [/tex]

    3. The attempt at a solution
    From the given information we know that [tex] f'(0) \geq 0 [/tex] and so [tex] f'(x) \geq 0 [/tex]

    define [tex] h(x) = f(x)-e^x [/tex]
    so h(0) =0

    [tex] h'(x)=f'(x) -e^x = lim_{h ->0 } \frac {f(x+h)-f(x)}{h} -e^x= lim_{h ->0 } \frac {f(x+h)-f(x) -he^x}{h}= lim_{h ->0 } \frac {f(x+h)-f(x) -0}{h} = f'(x) \geq 0 [/tex]

    Then h(x) is montonous rising and so is never smaller then zero. thus [tex] f(x) \geq e^x [/tex]

    Is this correct?

    Last edited: Feb 27, 2010
  2. jcsd
  3. Feb 27, 2010 #2
    Well you seemed to have defined f(0) = 0, so that h(0) = -1...
  4. Feb 27, 2010 #3
    oops. I made a typo, that was supposed to be f(0)=1. Fixed in the original post.
  5. Feb 27, 2010 #4
    OK cool. Even though you have the right idea, the second to last equality in your main derivation cannot be justified, since you are taking the limit of some term in the numerator of a quotient which is a serious violation of the limit laws (basically you need to know all limits in the the numerator and denominator exist before you can perform such operations).

    My advice would be to consider g(x) = f(x)/e^x instead and show g is increasing. At no point should you need to resort to the definition of the derivative.
  6. Feb 27, 2010 #5
    That seems much simpler:
    Just to make sure i got it right:
    h(x) = [tex] \frac {f(x)}{e^x}, h(0)=1 [/tex]
    [tex] h'(x)= \frac {f'(x)e^x -f(x)e^x}{e^(2x)} \geq 0 [/tex] because f'(x)>f(x)
  7. Feb 27, 2010 #6
    That's fine. Notice the last inequality is in fact strict, which only helps you I guess.
  8. Feb 27, 2010 #7
    Thanks alot. Much apreciated
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook