3

Is there an operator between Banach spaces with the following properties: $$T:\mathcal{D}(T)\subseteq X\to Y:\text{ injective, dense range, continuously invertible, not closable!}$$ (Note that the ambient spaces should be really Banach spaces - otherwise this is trivial.)

Obviously this operator must automatically satisfy:

  • Operator must be unbounded
  • Range must not be closed
  • Closure of inverse must be not invertible

The question is related to the alternative definition of: Resolvent: Definition

C-star-W-star
  • 16,275
  • The answer in here might help: http://math.stackexchange.com/questions/593857/counterexample-for-the-open-mapping-theorem/658336#658336 – C-star-W-star Jun 03 '14 at 17:49
  • could you please clarify what you are worried about here : http://math.stackexchange.com/questions/818832/matrices-ab-ab-implies-a-commutes-with-b/818838#comment1692436_818838 Hmmm, it deleted the @Freeze; no accounting for machinery. – Will Jagy Jun 03 '14 at 18:15
  • Sure but lets continue it there otherwise it will become messy ;) – C-star-W-star Jun 03 '14 at 18:28

1 Answers1

4

Here's an idea. Let $X=l^{2}$ be the space of sequences $\{ x_{n}\}_{n=1}^{\infty}$ of square-summable sequences of complex numbers.

Let $e_{n}$ the standard basis element which is a sequence with all 0's except for a 1 in the n-th place. Let M be the subspace spanned by finite linear combinations of the vectors $e_{1}+\frac{1}{n}e_{n}$ for $n \ge 2$.

The closure $M^{c}$ of the linear space $M$ contains $e_{1}$ because $\lim_{n} (e_{1}+\frac{1}{n}e_{n})=e_{1}$. $M^{c}$ also contains $e_{n}=n(e_{1}+\frac{1}{n}e_{n})-ne_{1}$ for $n\ge 2$. So the subspace $M$ is dense in $X$ because it contains all of the standard basis elements.

Define $S$ to be the restriction of the left shift to $M$.  $S$ is bounded on $M$, and $\mathcal{N}(S)=\{0\}$ because $e_{1} \notin M$. Let $T$ be the inverse of $S$.

Then $T : \mathcal{D}(T) \rightarrow X$ has domain $\mathcal{D}(T)=S(M)$, and the range of $T$ is $M$ which is dense in $X$. And $S=T^{-1}$ is bounded by $1$, which gives $\|Tx\| \ge \|x\|$ for all $x \in \mathcal{D}(T)$.

However, $T$ is not closable because $(e_{1},0)$ is in the closure of the graph of $S$ and, so, $(0,e_{1})$ is in the closure of the graph of $T$.  Also notice that $\mathcal{D}(T)$ is dense because the range of $M$ under $S$ includes all finite linear combinations of the standard basis elements.

I think this works, but please check my details.

Disintegrating By Parts
  • 87,459
  • 5
  • 65
  • 149
  • I checked the details and they are clearly fine (i.e., it satisfies the criteria in the original question). For example, $TS=I$ (which holds by the definition of $T$ and the fact that $S$ is 1-1) implies that $T$ is 1-1 and maps $\mathcal D(T)$ onto $M$. – user3810316 Dec 17 '19 at 14:08
  • Additionally, $\mathcal D(T)=S(M)=\mathrm{span},{e_n}_{n=1}^\infty$ is dense in $\ell^2$. So both the domain and the range of $T$ are dense (in addition to all other required properties). – user3810316 Jan 03 '20 at 14:27