My textbook said that

Suppose that $\{f_1,f_2,\dots\}$ is a linearly independent sequence in a Hilbert space $V$. There is a vector in $V$ which is not a finite linear combination of the $f_j$.

I am wondering if I have a Hamel basis $B$ for $V$. Then $B$ is linearly independent and by the above assertion, there is a vector $v\in V$ such that $v$ cannot be expressed as a finite linear combination, which contradicts with the fact that $B$ is a basis.

The proof on the textbook is

The Gram-Schmidt process produces an orthonormal sequence $\{e_1,e_2,\dots\}$ with $span\{e_1,e_2,\dots,e_n\}=span\{f_1,f_2,\dots,f_n\}$ for any $n$. Write down any sum $\sum_1^\infty c_je_j$ where infinitely many of $c_j$ are nonzero and $\sum_1^\infty |c_j|^2<\infty$. The sum then converges to a vector $g$. Then $g$ is not in the span of $\{e_1,e_2,\dots,e_n\}$ for any $n$, and thus not in the space of $\{f_1,f_2,\dots,f_n\}$ for any $n$.


As a vector space, it must have a Hamel basis, and the theorem says that such a Hamel basis cannot be denumerable, so it must be uncountable.

  • $\begingroup$ So an infinite-dimensional Hilbert space must have an uncountable Hamel basis? $\endgroup$ – SHBaoS Feb 4 '18 at 4:58
  • $\begingroup$ Yes, this is correct. $\endgroup$ – user284331 Feb 4 '18 at 4:58

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.