Distributive rings of continuous functions and F-spaces

Bezout module Bezout ring distributive lattice E-cozero set EF-space F-space rings of continuous functions with values in topological rings or fields set of zeros

Mathematics Subject Classification ID

Topological and ordered rings and modules (16W80) Rings and algebras of continuous, differentiable or analytic functions (46E25) Normed modules and Banach modules, topological modules (if not placed in 13-XX or 16-XX) (46H25) Functional analysis over fields other than (mathbb{R}) or (mathbb{C}) or the quaternions; non-Archimedean functional analysis (46S10)

Review text

The author continues his study of rings of continuous functions with values in topological rings or fields. Let E be a separable topological associative ring with identity \(1\neq 0\) and X a topological space. We denote by \(C(X,E)\) the set of continuous functions of X into E, which forms an associative ring with respect to pointwise operations. An element f in \(C(X,E)\) is called bounded if \(f(X)\) is relatively compact in E. \(C^*(X,E)\) denotes the set of all bounded functions in \(C(X,E)\). A subset S of X is said to be \(C^*_ E\)-embedded in X if every element in \(C^*(S,E)\) is extended to an element in \(C(X,E)\). Given any element f in \(C(X,E)\), we denote by \(Z(f)\) the set of zeros of f. A subset in X is called an E-cozero set if it is the complement in X of some Z(f). X is, by definition, an EF-space if every E-cozero subset of X is \(C^*_ E\)-embedded in X. Two subsets \(A_ 1\) and \(A_ 2\) are said to be E- separated if we have \(f(A_ 1)=\{0\}\), \(f(A_ 2)=\{1\}\) for some f in \(C(X,E).\) The main result, Theorem 1, states fourteen mutually equivalent properties, a sample being given below. Theorem 1: For any continuous field (=a nondiscrete locally compact field) E and any topological space X, the following are equivalent: (1) \(C(X,E)\) is a Bezout ring, i.e. every finitely generated ideal of \(C(X,E)\) is principal; (2) the lattice of ideals of \(C(X,E)\) is distributive; (3) every ideal of \(C(X,E)\) is flat; (9) if an E-cozero subset of X is partitioned into two open sets \(A_ 1\) and \(A_ 2\), then these are E-separated; (11) X is an EF-space. Theorem 2 deals with general topological rings. For such a ring E the set \(E^ X\) of all E-valued functions on X forms a right \(C(X,E)\)-module under pointwise operations. The module \(E^ X\) is called distributive if the lattice of submodules of \(E^ X\) is distributive. It is called a Bezout module if every finitely generated submodule is cyclic. Theorem 2. (a) If every right (or left) invertible element in E is invertible, then the right \(C(X,E)\)-module \(E^ X\) is a Bezout module if and only if X is discrete and E is a right Bezout ring. (b) The \(C(X,E)\)-module \(E^ X\) is distributive if and only if X is discrete and the ring E is right distributive.

Recommendations

Cites work

Cited in

(19)

This page was built for publication: Distributive rings of continuous functions and F-spaces

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q797101)