characterization of maximal ideals of the algebra of continuous functions on a compact set


Let X be a compactPlanetmathPlanetmath topological spaceMathworldPlanetmath and let C(X) be the algebra of continuousMathworldPlanetmathPlanetmath real-valued functions on this space. In this entry, we shall examine the maximal idealsPlanetmathPlanetmath of this algebra.

Theorem 1.

Let X be a compact topological space and I be an ideal of C(X). Then either I=C(x) or there exists a point pX such that f(p)=0 for all fI.

Proof.

Assume that, for every point pX, there exists a continuous function fI such that f(p)0. Then, by continuity, there must exist an open set U containing p so that f(q)0 for all qU. Thus, we may assign to each point pX a continuous function fI and an open set U of X such that f(q)0 for all qU. Since this collectionMathworldPlanetmath of open sets covers X, which is compact, there must exists a finite subcover which also covers X. Call this subcover U1,,Un and the corresponding functions f1,fn. Consider the function g defined as g(x)=(f1(x))2++(fn(x))2. Since I is an ideal, gI. For every point pX, there exists an integer i between 1 and n such that fi(p)0. This implies that g(p)0. Since g is a continuous function on a compact set, it must attain a minimum. By construction of g, the value of g at its minimum cannot be negative; by what we just showed, it cannot equal zero either. Hence being bounded from below by a positive number, g has a continuous inversePlanetmathPlanetmathPlanetmath. But, if an ideal contains an invertible element, it must be the whole algebra. Hence, we conclude that either there exists a point px such that f(p)=0 for all fI or I=C(x). ∎

Theorem 2.

Let X be a compact Hausdorff topological space. Then an ideal is maximal if and only if it is the ideal of all points which go zero at a given point.

Proof.

By the previous theorem, every non-trivial ideal must be a subset of an ideal of functions which vanish at a given point. Hence, it only remains to prove that ideals of functions vanishing at a point is maxiamal.

Let p be a point of X. Assume that the ideal of functions vanishing at p is properly contained in ideal I. Then there must exist a function fI such that f(p)0 (otherwise, the inclusion would not be proper). Since f is continuous, there will exist an open neighborhood U of p such that f(x)0 when xU. By Urysohn’s theorem, there exists a continuous function h:X such that f(p)=0 and f(x)=0 for all xXU. Since I was assumed to contain all functions vanishing at p, we must have fI. Hence, the function g defined by g(x)=(f(x))2+(h(x))2 must also lie in I. By construction, g(g)>0 when xU and when g(x)XU. Because X is compact, g must attain a minimum somewhere, hence is bounded from below by a positive number. Thus g has a continuous inverse, so I=C(X), hence the ideal of functions vanishing at p is maximal. ∎

Title characterization of maximal ideals of the algebra of continuous functions on a compact set
Canonical name CharacterizationOfMaximalIdealsOfTheAlgebraOfContinuousFunctionsOnACompactSet
Date of creation 2013-03-22 17:45:08
Last modified on 2013-03-22 17:45:08
Owner rspuzio (6075)
Last modified by rspuzio (6075)
Numerical id 9
Author rspuzio (6075)
Entry type Theorem
Classification msc 46L05
Classification msc 46J20
Classification msc 46J10
Classification msc 16W80