Bourbaki

Proposition 4

Tag 02SN, permanent. Cite it as [Bourbaki, Tag 02SN].

Théories spectrales, chapter III, § 4. Perturbations dans les espaces de Banach

Let E and F be Banach spaces, let u0:E→Fu_0: E\rightarrow F be a Fredholm mapping and v0v_0 a quasi-inverse of u0u_0. There exist an open neighbourhood U of u0u_0 in L(E;F)\mathscr{L}(E; F) and an analytic mapping φ:U→L(F;E)\varphi : U\rightarrow \mathscr{L}(F; E) such that

(i) One has φ(u0)=v0\varphi (u_0) =v_0;

(ii) For every uu in U, the mapping φ(u)\varphi (u) is a quasi-inverse of uu, and in particular uu is a Fredholm mapping;

(iii) For every uu in U, one has ind(uu) == ind(u0u_0).

By prop. 2 of III, p. 42, (ii), there exist topological direct sum decompositions E=E1⊕E2E = E_1\oplus E_2 and F=F1⊕F2F = F_1\oplus F_2, and there exists α0∈I(E1;F1)\alpha_0\in \mathscr{I}(E_1; F_1), such that E2E_2 and F2F_2 are finite-dimensional and that u0u_0 is represented by the matrix (α0000)\begin{pmatrix} \alpha_0 & 0 \\ 0 & 0 \end{pmatrix} relative to these decompositions.

Let U be the set of elements uu of L(E;F)\mathscr{L}(E; F) such that, in the matrix representation (αβγδ)\begin{pmatrix} \alpha & \beta \\ \gamma & \delta \end{pmatrix} of uu with respect to these decompositions, one has α∈I(E1;F1)\alpha \in \mathscr{I}(E_1; F_1). Since I(E1;F1)\mathscr{I}(E_1; F_1) is open in L(E1;F1)\mathscr{L}(E_1; F_1) (prop. 3 of III, p. 57), U is an open neighbourhood of u0u_0 in L(E;F)\mathscr{L}(E; F). For u=(αβγδ)u=\begin{pmatrix} \alpha & \beta \\ \gamma & \delta \end{pmatrix} in U, put

φ(u)=v0+(α−1−α0−1000)(1) \varphi (u) =v_0+\begin{pmatrix} \alpha^{-1}-\alpha_0^{-1} & 0 \\ 0 & 0 \end{pmatrix} \tag{1}

One has φ(u0)=v0\varphi (u_0) =v_0 and the mapping φ\varphi is analytic (loc. cit.). Modulo continuous linear mappings of finite rank, one has the congruences

u≡(α000),v0≡(α0−1000),φ(u)≡(α−1000) u\equiv \begin{pmatrix} \alpha & 0 \\ 0 & 0 \end{pmatrix},v_0\equiv \begin{pmatrix} \alpha_0^{-1} & 0 \\ 0 & 0 \end{pmatrix},\varphi (u)\equiv \begin{pmatrix} \alpha^{-1} & 0 \\ 0 & 0 \end{pmatrix}

Consequently, φ(u)\varphi (u) is a quasi-inverse of uu. Every element uu of U defines by restriction an isomorphism of E1E_1 onto a topological supplement of F2F_2 in F. By prop. 3 of III, p. 44, one therefore has

ind(uu) == codimF(u(E1))−_F(u(E_1))- codimE(E1)_E(E_1)

= dim(F2_2)−- dim(E2_2) == ind(u0u_0),

which concludes the proof.

Bourbaki sets the proof directly after the statement, and the corpus does not separate them, so what is above is the statement and its proof together.

Cited by

Nothing.

Cites

Nothing.

Machine translation of the English, by gpt-5.4, against glossary version 34. Not checked by a person.