Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Simple Differentiability and Continuity Question

  1. Nov 16, 2011 #1
    1. The problem statement, all variables and given/known data

    If f(x) = 3 for x < 0 and f(x) = 2x for x ≥ 0, is f(x) differentiable at x = 0? State and justify why/why not.

    2. Relevant equations

    3. The attempt at a solution

    Obviously, since f(x) is not continuous and the limit doesn't exist as x[itex]\rightarrow[/itex]0, the function shouldn't be differentiable at that point. But I can't justify this because I don't really understand why not. Simply put, I get that the limit doesn't exist at 0 and, hence, the function is discontinuous there, but I don't see why that invalidates the idea that the gradient of the line f(x) = 2x at x = 0 is 2. Further, as far as I can tell (and I have a strong feeling I'm going wrong here somewhere), f'(x) = 0 for x < 0 and f'(x) = 2 for x ≥ 0, which seems to suggest that the derivative of f(x) for all values of x ≥ 0 is 2. What am I missing here?

    As always, any help would be much appreciated, thanks.
  2. jcsd
  3. Nov 16, 2011 #2


    User Avatar
    Science Advisor

    You say, twice, that the limit does not exist at 0. What limit? I think you mean just [itex]lim_{x\to 0} f(x)[/itex] but the derivative, at x= 0, is defined as
    [tex]\lim_{h\to 0}\frac{f(h)- f(0)}{h}= \lim{h\to 0} \frac{f(h)}{h}[/tex]
    if the numerator does not go to 0, then that limit cannot exist.
  4. Nov 17, 2011 #3
    You're right; I was considering the problem in too narrow a scope and only considering the limit as x[itex]\rightarrow[/itex]0. So let's see: the limit as x[itex]\rightarrow[/itex]0 of f(x) does not exist, therefore, the limit as h[itex]\rightarrow[/itex]0 of f(0 + h) - f(0) does not exist either. The limit as h[itex]\rightarrow[/itex]0 of h is obviously just 0, and, since in [tex]\lim_{h\to 0}\frac{f(0 + h) - f(0)}{h}[/tex] the denominator approaches 0 as expected, but the numerator does not exist, then the limit of the quotient and hence the derivative does not exist. Is that right?
  5. Nov 17, 2011 #4


    User Avatar
    Science Advisor

    For this particular function, yes, the limit of the numerator does not exist. But the reason why a function has to be continuous to be differentiable is more general. Consider another example: f(x)= 2x for x not equal to 0, f(0)= 1. Now [itex]\lim_{h\to 0}f(h)= 0[/itex] but [itex]\lim_{h\to 0} f(h)- f(0)= -1[/itex]. The limit of the numerator does exist but is not 0 so [itex]\lim_{h\to 0} (f(h)- f(0))/h[/itex] does not exist.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook