I am reading the following question:

Let $X$ be an infinite dimensional Banach space. Prove that every Hamel basis of X is uncountable.

And I am wondering why

$$X=\bigcup_{n\in \mathbb N}X_n$$

Since the right hand side is just an union of some sets and the left hand side is the complete space. Shouldn't it be:

$$X=lin(\bigcup_{n\in \mathbb N}X_n)$$ ?

  • $\begingroup$ Note that the $X_n$'s form an increasing chain of subsets, so at each stage $\bigcup_{n\le N}X_n=X_N$. That's why we don't need $\operatorname{span}(\bigcup_{n\in\Bbb N}X_n)$, since any vector is spanned by finitely many basis vectors, which is in some $X_n$. $\endgroup$ – Mario Carneiro Feb 5 '15 at 11:53
  • $\begingroup$ But I still don't understand it. $\cup_{n\in \mathbb N}X_n$ is a collection of independent vectors. Not the whole space. :/ $\endgroup$ – Duke Feb 5 '15 at 11:56
  • 2
    $\begingroup$ Oh, now I see the issue. $X_n=\operatorname{span}(v_1,\dots,v_n)$, not $X_n=\{v_1,\dots,v_n\}$. The author appears to be using square brackets to denote this. $\endgroup$ – Mario Carneiro Feb 5 '15 at 11:59
  • $\begingroup$ Thanks a lot. That was stupid from me. I only know the notation <...>. $\endgroup$ – Duke Feb 5 '15 at 12:46

The notation $[v_1,\ldots,v_n]$ refers to the vector space spanned by $B=\{v_1,\ldots,v_n\}$ and not $B$ itself. If $x\in X$ then by definition of a Hamel basis, $x=\sum_{j=1}^ka_jv_j$ for some $k\ge1$, $a_j\in\mathbb{K}$. This implies $x\in X_k\subseteq\bigcup_{n\in\mathbb{N}}X_n$.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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