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 $H$ a Hilbert space and $(a_n)_{n \in \mathbb{N}}$ a complete orthonormal system, i.e. a basis. Do you know a example of a orthonormal system $(b_n)_{n\in \mathbb{N}}$ satisfying $\infty > \sum_{k \in \mathbb{N}} \lVert a_k-b_k \lVert \geq 1$, which is not complete?

share|cite|improve this question

Sketch of the proof

0)Since $\sum_{k\in\mathbb{N}}\Vert a_k-b_k\Vert<+\infty$, then $\sum_{k\in\mathbb{N}}\Vert a_k-b_k\Vert^2<+\infty$. And we reduce to the case considered in the "A Hilbert Space problem book" by P.R. Halmos. This problem no 7 in Russian edition.

1) Assume that there exist $b_0\neq 0$, orthogonal to all $b_k$. Then for all $N$ the system ${b_0,\ldots,b_N}$ is linearly independent.

2) Consider $n$ such that $\sum_{k>n}\Vert a_k-b_k\Vert^2<1$.

3) Projections $\{c_0,\ldots,c_n\}$ of $b_0,\ldots,b_n$ on $\mathrm{span}\{a_1,\ldots,a_n\}$ are linearly dependent, i.e. $\sum_{k=0}^n \alpha_k c_k=0$ for some coefficients $\{\alpha_0,\ldots,\alpha_n\}$, that are not zero simultaneously.

4) Now prove that vector $d=\sum_{k=0}^n\alpha_k b_k$ is orthogonal to $a_1,\ldots,a_n$ and then prove that $$ \Vert d\Vert^2\leq\sum\limits_{k>n}\Vert d\Vert^2\Vert a_k-b_k\Vert^2. $$

6) This implies $d=0$. Hence $\{b_0,\ldots,b_n\}$ is linearly dependent. Contradiction, so for all $b_0$ orthogonal to $b_1,\ldots,b_n$ we have $b_0=0$. This means completeness of $b_1,\ldots,b_n$.

share|cite|improve this answer

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.