No label defined (Q4662277)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 2149762
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 2149762 |
Statements
30 March 2005
0 references
0.897266685962677
0 references
0.8848899006843567
0 references
0.8840593695640564
0 references
The paper is a delightfully told story about the existence of a continuous linear analytic extension of Whitney jets. The problem goes back to the well-known Borel theorem (1895) according to which for every family \((c_n)_{n\in\mathbb N_0}\) of complex numbers, there is a function \(f\in\mathcal C^\infty(\mathbb R)\) such that \(\text{D}^nf(0)=c_n\) for every \(n\in\mathbb N_0.\) Its first improvement is due to Bernstein (1913) and states that the function \(f\) of the Borel theorem can be chosen to be analytic outside \(0\). Shortly afterwards Ritt (1916) showed that for any \(\varepsilon\in(0,\pi/2)\), there is such a function \(f\) having a holomorphic extension on the sector \(\{\rho e^{i\theta}:\;\rho>0, -\varepsilon<\theta<\pi+\varepsilon\}\). A fundamental improvement of the above results is due to Whitney (1934) who solved the problem for any closed subset \(F\) of \(\mathbb R^n\): A jet \(\varphi=(\varphi_{\alpha})_{\alpha\in\mathbb N^n_0}\) with \(\varphi_{\alpha}\in \mathcal C(F),\;\forall\;\alpha\in\mathbb N_0^n\), comes from a \(\mathcal C^\infty\) function \(f\) on \(\mathbb R^n\) if and only if it is a Whitney jet, i.e., for every compact subset \(K\) of \(F\), \(m\in\mathbb N_0\) and \(\alpha\in\mathbb N_0^n\) verifying \(|\alpha|\leq m\), \[ \sup_{\substack{ 0<|x-y|\leq t\\ x,y\in K}} \biggl\{\varphi_{\alpha}(y)-\sum_{|\beta|\leq m-|\alpha|}\frac{\varphi_{\alpha+\beta}(x)}{\beta!}(y-x)^\beta\biggr\}/|y-x|^{m-|\alpha|}\to 0\;\text{ if }\;t\to0^+. \] Moreover in such a case, one may suppose that \(f\) is holomorphic on \(\{z\in\mathbb C^n: x\in\mathbb R^n\setminus F\), \(|y|<d(x,F)\}\), and hence analytic on \(\mathbb R^n\setminus F\). Let \(\mathcal E(F)\) denote the vector space of all Whitney jets on \(F\). It can be endowed in a natural way with a Fréchet topology. Then the first part of Whitney's theorem can be restated as follows. The continuous linear restriction map \[ R: \mathcal C^\infty(\mathbb R^n)\to \mathcal E(F);\qquad f\rightarrowtail(\text{D}^\alpha f_{|F})_{\alpha\in\mathbb N_0^n} \] is surjective. Now one can ask the following question: when does there exist a continuous linear extension map from \(\mathcal E(F)\) into \(\mathcal C^\infty(\mathbb R^n)?\) It is well-known that there is no such a map if \(F\) is a singleton. Mityagin (1961) showed that there is a continuous linear extension map from \(\mathcal E([-1,1])\) into \(\mathcal C^\infty(\mathbb R)\). This result opened the hunt for positive or negative examples as well as for necessary or sufficient conditions for the existence of such an extension map. Here the author mentions results of Seeley, Stein, Bierstone, alone or with Milman or Schwarz, Tidten, Pawłucki-Pleśniak and Pleśniak alone, Goncharov and Frerick. He fears he has forgotten to mention many other names \dots and I agree with him. In fact, it is not the existence problem which focuses most the author's attention in this survey. He is more interested in the last part of Whitney's theorem dealing with analyticity. Let us mention here one of the important contributions to this question given by Schmets and Valdivia (1997): Let \(K\) be a non void compact subset of \(\mathbb R^n\). If there is a continuous linear extension map from \(\mathcal E(K)\) into \(\mathcal C^\infty(\mathbb R^n)\), then there also is such a map with images analytic on \(\mathbb R^n\setminus K\). The same result has also been obtained by different methods by Langenbruch (1999). Frerick and Vogt (2002) characterized proper closed subsets of \(\mathbb R^n\) for which the above Schmets-Valdivia result still holds. Some stronger results have been recently obtained in the case of \(n=1\) by Frerick in cooperation with Brück or Boonen. Let us remark that more details about the existence of a continuous linear extension map from \(\mathcal E(K)\) into \(\mathcal C^\infty(\mathbb R^n)\) can be found in the reviewer's survey article [``Recent Progress in Multivariate Markov Inequality'', Approximation Theory: in memory of A. K. Varma, New York, NY: Marcel Dekker. Pure Appl. Math., Marcel Dekker. 212, 449--464. New York (1998; Zbl 0984.41007)]. For the entire collection see [Zbl 1054.00007].
0 references