Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $X$ be a Banach space and $A\subset X$ be bounded. Suppose that, for any $\varepsilon>0$, $\exists\; F_\varepsilon$ a subspace of $X$ of finite dimension such that $$\text{dist}(x,F_\varepsilon)\leq\varepsilon,\quad\forall x\in A.$$

Then prove that $A$ is relatively compact in $X$, i.e. its closure is compact in $X$.

Can anybody give me an hint?

share|cite|improve this question
up vote 2 down vote accepted

I would reason as follows. Fix $\varepsilon>0$ and accordingly the finite dimensional subspace $F_\varepsilon$ granted by the hypotheses of the problem. Then $F_\varepsilon$ is closed. In particular $\forall x\in A$ exists $v_x\in F_\varepsilon$ such that $\|x-v_x\|=\text{dist}(x,F_\varepsilon)\leq \varepsilon$. Since $A$ is bounded, the set of all such $v_x$ is bounded, because $\|v_x\|\leq \|x-v_x\|+\|x\|\leq \varepsilon+M$. Then it istotally bounded because bounded in a finite dimensional space. Suppose $v_p$ is an $\varepsilon$ net in this set. Then we may choose, for any $p$, $x_p$ as (one of) the element(s) of $A$ such that $\|x_p-v_p\|=\text{dist}(x_p,F_\varepsilon).$ Notice that $x_p$ exists by construction of the set of the $v_x's$. Then the $\{x_p\}$ are a $3\varepsilon$ net in $A$ because if we choose $x\in A$, let $v_p$ be such that $\|v_x-v_p\|\leq \varepsilon$ by definition of $\varepsilon$ net. Then $\|x-x_p\|\leq \|x-v_x\|+\|v_x-v_p\|+\|v_p-x_p\|\leq 3\varepsilon.$

Hope this is correct.

share|cite|improve this answer

Fix $\varepsilon>0$. Let $B$ be a ball containing $A$. The set $G_\varepsilon=F_\varepsilon\cap B$ is non-empty bounded subset of finite dimensional subspace, so you can find here finite $\varepsilon$-net $M_\varepsilon$. It is easy to prove that $M_\varepsilon$ is a $2\varepsilon$-net for $A$. The rest is clear.

share|cite|improve this answer
what if $F_\varepsilon\cap A$ is empty? – guido giuliani Aug 24 '12 at 15:25
Well, I think we can salvage this proof – Norbert Aug 24 '12 at 16:26

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.