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

Pointwise Limits and Uniform Convergence Help

  1. Apr 23, 2010 #1
    Hi all
    Came across this question which i have attempted to answer but am sure i have gone wrong somewhere, any help would be appreciated...

    Suppose fn(x):R-->R, where fn(x)=x+(1/n) (n belongs to N). Find the pointwise limits of the sequences (fn), (1/n fn) and (fn2). In each case determine whether the convergence is uniform.

    Now i managed to get all three as being uniformly convergent and am sure i went wrong somewhere. I am still struggling to understand uniform convergence completely. The pointwise limits i computed were x, 0, x2 respectively...sorry i havnt put my working up struggling to use latex.
  2. jcsd
  3. Apr 23, 2010 #2
    Your limits are right. (Let [tex]f[/tex] be the limit of [tex]f_n[/tex]). The first sequence is uniformly convergent: given [tex]\epsilon >0[/tex], choose N to be the lowest integer that is greater than:[tex]\frac{1}{\epsilon}[/tex].

    Then, for all [tex]n>N[/tex]:

    For the second one, note that [tex]sup_x|f_n|[/tex] does not go to 0. This tells you it is not uniformly convergent. To prove this, suppose we are given [tex]\epsilon >0[/tex]. For any choice of N, let [tex]x>\epsilon .N-\frac{1}{N}[/tex].

    [tex]|f-f_N|=|0-\frac{x}{N}-\frac{1}{N^2}|=|\frac{x}{N}+\frac{1}{N^2}|>\frac{\epsilon .N-\frac{1}{N}}{N}+\frac{1}{N^2}=\epsilon[/tex].

    The third one is very similar. I'm afraid all this might be a bit bewildering if you don't yet understand uniform convergence though! Keep thinking about it and eventually it will click.

    PS - I may have made some mistakes here as I did it very quickly, so be sure to check everything!
    Last edited: Apr 23, 2010
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook