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 subspace of $\ell_1^4$ (i.e. $\mathbb{R}^4$ equipped with the $\ell^1$ norm). Can one always extend a linear operator $l:X\rightarrow \ell_1^4$ to $L:\ell_1^4\rightarrow \ell_1^4$ such that $L$ has the same operator norm as $l$?


share|cite|improve this question

closed as too localized by quanta, t.b., Qiaochu Yuan May 14 '11 at 2:57

This question is unlikely to help any future visitors; it is only relevant to a small geographic area, a specific moment in time, or an extraordinarily narrow situation that is not generally applicable to the worldwide audience of the internet. For help making this question more broadly applicable, visit the help center.If this question can be reworded to fit the rules in the help center, please edit the question.

@Arturo, what Bill said is for two spaces, $X$ and $Y$. Here $X$ and $Y$ are the same space. The question really is: what if $X$ is not \ell{\infty}^n$? – user4730 Dec 14 '10 at 5:32
Dear Arturo, what did you say a couple of minutes ago? – user4730 Dec 14 '10 at 6:05
@user4730: Something only half right. I said we can assume $||l||=1$, which is true enough (just scale), and something false. – Arturo Magidin Dec 14 '10 at 6:12
@user1736: No. The Hahn-Banach theorem applies only to linear forms (functionals), not to operators. – Florian Mar 14 '11 at 18:51
@Willie: Can't we just close this question? It is both completely uninteresting and unmotivated and pops up every now and again on MO and here e.g. in this version. The OP got an answer by Fedja in the comments on MO in one version of this question, which he didn't understand then deleted the question etc. I'm tired of this. – t.b. May 13 '11 at 22:42

If $X$ is a closed subspace of $\ell_1^4$, then it is true if the latter has the metric extension property. If you are working over the reals, then Google :"binary intersection property Nachbin". I think Grothendieck proved that if a space has the metric extension property, then it is either finite dimensional or non-separable. I hope this helps!.

share|cite|improve this answer
I am wondering if $\ell_1^4$ has the metric extension property. – user4730 Dec 15 '10 at 12:33

The answer is (trivially) affirmative if you suppose that $X$ be one of the following subspaces:

$$\{(x^1, 0, 0, 0) \mid x^1 \in \mathbb{R}\}\quad \{(x^1, x^2, 0, 0) \mid x^i\in \mathbb{R}\},\quad \{(x^1, x^2, x^3, 0) \mid x^i \in \mathbb{R}\}.$$

Then the operator $l \colon X \to \ell^1_4$ can be represented as one of the following matrices

$$\begin{bmatrix} l^1_1 \\ l^2_1 \\ l^3_1 \\ l^4_1 \end{bmatrix},\quad \begin{bmatrix} l^1_1& l^1_2 \\ l^2_1 & l^2_2\\ l^3_1 &l^3_2\\ l^4_1 & l^4_2 \end{bmatrix}, \quad \begin{bmatrix} l^1_1 & l^1_2 & l^1_3 \\ l^2_1 & l^2_2 & l^2_3 \\ l^3_1 & l^3_2 & l^3_3 \\ l^4_1 & l^4_2 & l^4_3\end{bmatrix}$$

meaning that

$$l(x^1, x^2, x^3, x^4)= \begin{bmatrix}l^i_j\end{bmatrix}\begin{bmatrix} x^1 \\ x^2 \\ x^3 \\ x^4 \end{bmatrix}.$$

Now the operator norm of $l$ is the usual $\ell^1$ norm of the matrix $\begin{bmatrix} l^i_j \end{bmatrix}$, that is

$$\lVert l \rVert = \max_j \sum_{i=1}^4 \lvert l^i_j \rvert$$

where $j$ ranges over $1..\dim X$. To extend $l$ preserving norm it will be enough to fill up the corresponding matrix with zeroes:

$$\begin{bmatrix} l^1_1 & 0 & 0 &0 \\ l^2_1 &0 &0 & 0 \\ l^3_1 &0 & 0 & 0\\ l^4_1 &0 &0 &0 \end{bmatrix},\quad \begin{bmatrix} l^1_1& l^1_2 & 0 & 0\\ l^2_1 & l^2_2 & 0 & 0\\ l^3_1 &l^3_2 & 0&0 \\ l^4_1 & l^4_2 &0 & 0\end{bmatrix}, \quad \begin{bmatrix} l^1_1 & l^1_2 & l^1_3 & 0 \\ l^2_1 & l^2_2 & l^2_3 & 0\\ l^3_1 & l^3_2 & l^3_3 & 0\\ l^4_1 & l^4_2 & l^4_3 & 0\end{bmatrix}$$

I don't know if this can be of some help. If the norm were $\ell^2$ then we could isometrically map every subspace $X$ into one of the preceding four and be done with it. With $\ell^1$ can we do the same thing...?

share|cite|improve this answer

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