Slices in the Unit Ball of a Uniform Algebra
نویسندگان
چکیده
We show that every nonvoid relatively weakly open subset, in particular every slice, of the unit ball of an infinite-dimensional uniform algebra has diameter 2. It is an important task in Banach space theory to determine the extreme point structure of the unit ball for various examples of Banach spaces. The most common way to describe “corners” of convex sets is by looking for extreme points, exposed points, denting points and strongly exposed points. Every strongly exposed point is both denting and exposed and every denting or exposed point is extreme. In this note A denotes an infinite-dimensional uniform algebra, i.e., an infinite-dimensional closed subalgebra of some C(K)-space which separates the points of K and contains the constant functions. In [1] Beneker and Wiegerinck demonstrated the non-existence of strongly exposed points in BA, the closed unit ball of A. Here we shall prove a stronger result by more elementary means. A corollary of our result is that the set of denting points is, in fact, also empty. Recall that we may assume that K is the Silov boundary of A. It is a fundamental result in the theory of uniform algebras that then the set of strong boundary points is dense in K; cf. [4, p. 48 and p. 78]. (A point x ∈ K is a strong boundary point if for every neighbourhood V of x and every δ > 0 there is some f ∈ A such that f(x) = ‖f‖ = 1 and |f | ≤ δ off V .) We now turn to our first result, which gives a quantitative statement of the non-dentability of BA. Theorem 1. Every slice of the unit ball of an infinite-dimensional uniform algebra A has diameter 2. Proof. Take an arbitrary slice S = {a ∈ BA: re l(a) ≥ 1−ε}, where ‖l‖ = 1. We will produce two functions in S having distance nearly 2. Let 0 < δ ≤ ε/11. We first pick some f ∈ BA such that re l(f) ≥ 1− δ. The functional l can be represented by a regular Borel measure μ on K with ‖μ‖ = 1, i.e., l(a) = ∫ K a dμ for all a ∈ A. Let ∅ 6= V0 ⊂ K be an open set with |μ|(V0) ≤ δ; such a set exists since K is infinite. Fix a strong boundary point x0 ∈ V0. Using the definition of a strong boundary point, inductively construct functions g1, g2, . . . ∈ A and nonvoid open subsets Date: February 1, 2008. 1991 Mathematics Subject Classification. Primary: 46B20.
منابع مشابه
SOME REMARKS ON WEAKLY INVERTIBLE FUNCTIONS IN THE UNIT BALL AND POLYDISK
We will present an approach to deal with a problem of existence of (not) weakly invertible functions in various spaces of analytic functions in the unit ball and polydisk based on estimates for integral operators acting between functional classes of different dimensions.
متن کاملSurjective Real-Linear Uniform Isometries Between Complex Function Algebras
In this paper, we first give a description of a surjective unit-preserving real-linear uniform isometry $ T : A longrightarrow B$, where $ A $ and $ B $ are complex function spaces on compact Hausdorff spaces $ X $ and $ Y $, respectively, whenever ${rm ER}left (A, Xright ) = {rm Ch}left (A, Xright )$ and ${rm ER}left (B, Yright ) = {rm Ch}left (B, Yright )$. Next, we give a description of $ T...
متن کاملBall Versus Distance Convexity of Metric Spaces
We consider two different notions of convexity of metric spaces, namely (strict/uniform) ball convexity and (strict/uniform) distance convexity. Our main theorem states that (strict/uniform) distance convexity is preserved under a fairly general product construction, whereas we provide an example which shows that the same does not hold for (strict/uniform) ball convexity, not even when consider...
متن کاملParabolic starlike mappings of the unit ball $B^n$
Let $f$ be a locally univalent function on the unit disk $U$. We consider the normalized extensions of $f$ to the Euclidean unit ball $B^nsubseteqmathbb{C}^n$ given by $$Phi_{n,gamma}(f)(z)=left(f(z_1),(f'(z_1))^gammahat{z}right),$$ where $gammain[0,1/2]$, $z=(z_1,hat{z})in B^n$ and $$Psi_{n,beta}(f)(z)=left(f(z_1),(frac{f(z_1)}{z_1})^betahat{z}right),$$ in which $betain[0,1]$, $f(z_1)neq 0$ a...
متن کاملAreas of Projections of Analytic Sets
If V is a pure 1-dimensional analytic subvariety of C", then the area of V is the sum of the areas, counting multiplicity, of the projections of V onto the n coordinate lines [5, 6]. If V is a pure 1-dimensional variety in the unit ball in C" which passes through the origin, then [5] the area of V is at least ~. Together these theorems imply that a 1variety through 0 in the unit ball has the pr...
متن کامل