1

help with this excercise,

Find the set $B$ such that its derived set $B´=A$ where,

$$A=\{1,\frac{1}{2},\frac{1}{3},...\}$$

How??

I try but none... :(

NewMathC
  • 107
  • How would you solve this if $A={1}.$? – B. Goddard Aug 27 '17 at 21:50
  • 2
    You have not told us what space you are working in. If you are working in $\mathbb R$ then there is no solution, because the derived set must be closed, and $A$ is not closed in $\mathbb R.$ – bof Aug 28 '17 at 03:35
  • Sometimes when you try to find a set, but fail you eventually have to consider the possibility that such a set does not exist. Then of course when you try to disprove something you may fail, but then you may have more insight in the problem at least... – skyking Aug 28 '17 at 06:38

2 Answers2

1

Such a set does not exist for the usual topology on $\mathbb{R}$, because a derived set containing each $\frac{1}{n}$ must also contain 0:

  1. Suppose every $1/n$ $(n\in \mathbb{N})$ is a limit point of $B$.

  2. Then $B$ contains arbitrarily close approximations to each $1/n$.

  3. Then 0 must also be a limit point of $B$.

  4. Indeed, if $1/n$ gets arbitrarily close to 0, and members of $B$ get arbitrarily close to each $1/n$, then members of $B$ get arbitrarily close to 0.

    (To see this formally, pick $\epsilon > 0$. The ball around 0 of radius $\frac{\epsilon}{2}$ contains some $1/n$. The ball around $1/n$ of radius $\frac{\epsilon}{2}$ contains some $y\in B$. The distance between $y$ and 0 is at most $\epsilon = \frac{\epsilon}{2} + \frac{\epsilon}{2}$, by the triangle inequality. Hence for each ε, we can find $y\in B$ within $\epsilon$ of 0, so 0 is a limit point of $B$.)

  5. Hence if the set of limit points of $B$ contains $\{\frac{1}{n} : n\in \mathbb{N}\}$, it must also contain 0.

  6. There is no set $A$ whose set of limit points includes $\{\frac{1}{n} : n \in \mathbb{N}\}$ and nothing else.


If our goal is to include $\{\frac{1}{n} : n \in \mathbb{N}\}$ and very little else, we can define $H = \{ \frac{1}{n} : n\in \mathbb{N}\}$, so that $H\cup \{0\}$ has zero as its sole limit point. Then define your set to be:

$$B = \{ x + y \mid x, y \in H\} \cup H.$$

Then if we fix $x$ and vary $y\in H$ in the definition of $B$, we can see that each $x\in H$ is a limit point of $B$, and so is 0. It should be straightforward to see whether any other point $x\notin H$ is a limit point.

(Following answer to distantly-related question: Example of a countable compact set in the real numbers)

user326210
  • 17,287
0

HINT: You have not provided enough information to give a unique answer, but one answer could come from studying the integral $\int_0^1{x^m}dx=\left[\frac{x^{m+1}}{m+1} \right]_0^1=\frac{1}{m+1}$ where $m$ is a nonnegative integer, with x being a real number between $0$ and $1$.

If you integrate the function again (being a repeated or successive integral depending on which book on calculus you read), what is the result?

$$\frac{1}{m+1}\int_0^1{x^{m+1}}dx=\frac{1}{m+1}\left[\frac{x^{???}}{???} \right]_0^1=\frac{1}{(m+1)(???)}$$

Therefore if the sets B and A are related by the rules of calculus, then you get an answer. Of course any number of other plausible associations and answers may exist.

  • 1
    It's not clear how this hint would be useful. It rather looks like you didn't understand the question, for example what the definition of derived set is. According to the answer of user326210 there is no solution to this problem. – skyking Aug 28 '17 at 06:36
  • @skyking: I will leave my "incorrect" answer because the general-topology tag was added by B. Goddard (not the OP or Original Poster) after I first added my HINT, to say without defining terms you could make a range of plausible associations to get an answer. In any case it would be much clearer if the "derived" set as I now understand it was called the "limit point set" or even the "limit set". – James Arathoon Aug 28 '17 at 09:34
  • 1
    Try google for "derived set" and you'll see that it's pretty standard notation. For example both wikipedia and wolfam (and other sources) agrees with agrees with the definition. You're not supposed to be required to define terms that are widely accepted. – skyking Aug 28 '17 at 09:40
  • @skyking: That's a shame because there is at least one other term in a different field I want to redefine. – James Arathoon Aug 28 '17 at 09:46
  • Which term in which field do you want to redefine? – skyking Aug 28 '17 at 09:48
  • @skyking: I deleted my comment; I will raise a new question instead. – James Arathoon Aug 28 '17 at 14:08