An outer measure on a commutative ring

We show how to produce a reasonable outer measure on a commutative ring from a given measure on a family of prime ideals of this ring. We provide a few examples and prove several properties of such outer measures.

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Algebra and Discrete Mathematics
Datum:2016
Hauptverfasser: Dudzik, D., Skrzyński, M.
Format: Artikel
Sprache:Englisch
Veröffentlicht: Інститут прикладної математики і механіки НАН України 2016
Online Zugang:https://nasplib.isofts.kiev.ua/handle/123456789/155199
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Zitieren:An outer measure on a commutative ring / D. Dudzik, M. Skrzyński // Algebra and Discrete Mathematics. — 2016. — Vol. 21, № 1. — С. 51-58. — Бібліогр.: 3 назв. — англ.

Institution

Digital Library of Periodicals of National Academy of Sciences of Ukraine
_version_ 1859481424859496448
author Dudzik, D.
Skrzyński, M.
author_facet Dudzik, D.
Skrzyński, M.
citation_txt An outer measure on a commutative ring / D. Dudzik, M. Skrzyński // Algebra and Discrete Mathematics. — 2016. — Vol. 21, № 1. — С. 51-58. — Бібліогр.: 3 назв. — англ.
collection DSpace DC
container_title Algebra and Discrete Mathematics
description We show how to produce a reasonable outer measure on a commutative ring from a given measure on a family of prime ideals of this ring. We provide a few examples and prove several properties of such outer measures.
first_indexed 2025-11-24T12:54:40Z
format Article
fulltext Algebra and Discrete Mathematics RESEARCH ARTICLE Volume 21 (2016). Number 1, pp. 51–58 © Journal “Algebra and Discrete Mathematics” An outer measure on a commutative ring Dariusz Dudzik and Marcin Skrzyński Communicated by V. I. Sushchansky Abstract. We show how to produce a reasonable outer measure on a commutative ring from a given measure on a family of prime ideals of this ring. We provide a few examples and prove several properties of such outer measures. Introduction Throughout the present paper, R is a nonzero commutative ring with identity. We denote by Spec(R) the family of all the prime ideals of R. (Notice that, by definition, every prime ideal is proper). It is well known [1] that topological properties of Spec(R) equipped with the Zariski topology reflect algebraic properties of R. But are there useful relationships between algebraic or geometric properties of R and measures on Spec(R)? This question seems to be quite interesting and not worked out in the specialist literature. The present paper provides some basic remarks concerning the question and, hopefully, is a starting point for further study. In the paper, we will show that an arbitrary measure on a suitable subfamily of Spec(R) induces an outer measure on R with good multi- plicative properties. We will also discuss a few elementary examples of such outer measures. 2010 MSC: 13A15, 28A12. Key words and phrases: outer measure, measure, commutative ring, prime ideal. 52 An outer measure on a commutative ring By “measure” we mean a “non-negative σ-additive measure”. We denote by 2X the power set of a set X. We define |X| = { the cardinality of X, if X is finite, +∞, otherwise. By R× we denote the set of invertible elements of R. Notice that ℘ ∩ R× = ∅ whenever ℘ ∈ Spec(R). We define Max(R) to be the family of all the maximal ideals of R. One can prove that Max(R) ⊆ Spec(R) and ⋃ Max(R) = R \ R×. We refer to [1] for more information about commutative rings and to [2] for elements of measure theory. 1. Construction We will use the definition of outer measure taken from [2]. Definition 1. We say that µ∗ : 2X −→ [0, +∞] is an outer measure on a set X, if the following conditions are satisfied: (1) µ∗(A) 6 ∞∑ n=1 µ∗(Bn) for every sequence {Bn}∞ n=1 of subsets of X and every A ⊆ ∞⋃ n=1 Bn, (2) µ∗(∅) = 0. Let P ⊆ Spec(R) be such that ⋃ P = R\R×, and let M be a σ-algebra of subsets of P. For a set A ⊆ R we define Ω(A) = { S ∈ M : ⋃ S ⊇ A \ R× } . Proposition 1. Suppose that µ : M −→ [0, +∞] is a measure. Then the function µ∗ : 2R −→ [0, +∞] defined by µ∗(A) = inf S∈Ω(A) µ(S) is an outer measure on R. (This outer measure will be referred to as the outer measure induced by µ). Proof. It is obvious that µ∗(∅) = 0. Let {Bn}∞ n=1 be a sequence of subsets of R and let ε be an arbitrary positive real number. Observe that ∀ n ∈ N \ {0} ∃ Sn ∈ Ω(Bn) : µ(Sn) 6 µ∗(Bn) + ε 2n . D. Dudzik, M. Skrzyński 53 If A ⊆ ∞⋃ n=1 Bn, then ∞⋃ n=1 Sn ∈ Ω(A) and hence µ∗(A) 6 ∞∑ n=1 µ(Sn) 6 ε + ∞∑ n=1 µ∗(Bn). Since ε is arbitrary, the above inequalities yield µ∗(A) 6 ∞∑ n=1 µ∗(Bn). The outer measure induced by a measure on a family of prime ideals is a slight modification of a well known measure-theoretical construction. In the next section we give examples that illustrate and motivate this modification. 2. Examples We denote by (a) the principal ideal generated by an element a ∈ R. Consider a further example of a “covering by prime ideals”. Example 1. We assume that R is a unique factorization domain and define Pirr(R) = {(0)} ∪ {(a) : a ∈ R, a is irreducible}. Observe that Pirr(R) ⊆ Spec(R) and ⋃ Pirr(R) = R \ R×. Moreover, if n ∈ N \ {0, 1} and R = C[x1, . . . , xn], then Pirr(R) ∩ Max(R) = ∅. Recall that for every ideal I of the ring of integers there exists exactly one m ∈ N ∪ {0} such that I = (m). Notice also that Max(Z) = {(p) : p ∈ P}, where P stands for the set of prime numbers. Proposition 2. Let µ∗ : 2Z −→ [0, +∞] be the outer measure in- duced by the counting measure on Max(Z), and let A ⊆ Z be such that A \ {−1, 1} 6= ∅. Then (i) µ∗({−1, 1}) = 0, (ii) µ∗(A) = 1 if and only if ∃ d ∈ N \ {0, 1} ∀ k ∈ A \ {−1, 1} : d | k (in particular, µ∗(A) = 1 whenever A is a singleton or a proper ideal of Z), (iii) µ∗(A) 6 |A|. Moreover, in the case where A ∩ {−1, 1} = ∅ and A is a finite set, µ∗(A) = |A| if and only if the elements of A are pairwise relatively prime. 54 An outer measure on a commutative ring Proof. Since {−1, 1} = Z ×, we have ∅ ∈ Ω({−1, 1}). Equality (i) follows. By the above characterization of Max(Z) and the definition of counting measure, µ∗(A) = 1 if and only if A \ {−1, 1} ⊆ (p1) for a prime number p1. The latter condition means precisely that ∃ p1 ∈ P ∀ k ∈ A \ {−1, 1} : p1 | k. Finally, if d ∈ N \ {0, 1}, k ∈ Z and d | k, then k is divisible by every prime factor of d. Property (ii) follows. Property (iii) is an immediate consequence of the definition of outer measure and the fact that µ∗({k}) 6 1 for all k ∈ Z. Assume that A ∩ {−1, 1} = ∅ and A is a finite set. Let us define ℓ = |A|. Observe that µ∗(A) 6= |A| if and only if ∃ S ∈ Ω(A) : |S| 6 ℓ − 1. Since the cardinality of A is greater than the cardinality of S, the latter condition holds true if and only if ∃ s, t ∈ A ∃ p2 ∈ P : { s 6= t, s, t ∈ (p2), and this means precisely that there exist two distinct elements of A which are not relatively prime. Let n ∈ N \ {0}. Consider a σ-algebra N of subsets of Cn, a measure λ : N −→ [0, +∞], and the map Φ : Cn ∋ z 7−→ {f ∈ C[x1, . . . , xn] : f(z) = 0} ∈ Max(C[x1, . . . , xn]). The family M = {S ⊆ Max(C[x1, . . . , xn]) : Φ−1(S) ∈ N} is a σ-algebra of subsets of Max(C[x1, . . . , xn]). The function η : M ∋ S 7→ λ(Φ−1(S)) ∈ [0, +∞] is a measure. Let us define U = C[x1, . . . , xn]×. (Obviously, U = C \ {0}). Proposition 3. If η∗ : 2C[x1,...,xn] −→ [0, +∞] is the outer measure induced by η and A ⊆ C[x1, . . . , xn] is such that A \ U 6= {0}, then η∗(A) = inf{λ(Z) : Z ∈ N, Z ∩ f−1(0) 6= ∅ for every f ∈ A \ C}. Proof. If A ⊆ U , then {Z ∈ N : Z ∩ f−1(0) 6= ∅ for every f ∈ A \ C} = N, and hence inf{λ(Z) : Z ∈ N, Z ∩ f−1(0) 6= ∅ for every f ∈ A \ C} = 0 = η∗(A). D. Dudzik, M. Skrzyński 55 By Hilbert’s Nullstellensatz, the map Φ is bijective. Consequently, M = {Φ(Z) : Z ∈ N}. Suppose that A \ C 6= ∅. Then for any Z ∈ N the following equivalences hold true: Φ(Z) ∈ Ω(A) ⇐⇒ (∀ f ∈ A \ U ∃ ℘ ∈ Φ(Z) : f ∈ ℘) ⇐⇒ (∀ f ∈ A \ C ∃ z ∈ Z : f(z) = 0) ⇐⇒ ( ∀ f ∈ A \ C : Z ∩ f−1(0) 6= ∅ ) . (The second equivalence holds because 0 belongs to every ideal). Therefore, η∗(A) = inf S∈Ω(A) η(S) = = inf{λ(Φ−1(Φ(Z))) : Z ∈ N, Z ∩ f−1(0) 6= ∅ for every f ∈ A \ C}, which completes the proof. Example 2. Let η∗ : 2C[x,y] −→ [0, +∞] be the outer measure induced by the counting measure on Max(C[x, y]). Consider the set E = {f, g, h, k} ⊂ C[x, y], where f(x, y) = x2 − y + 1, g(x, y) = y2, h(x, y) = xy − 1, k(x, y) = xy + 1. Since f−1(0) ∩ g−1(0) ∩ h−1(0) ∩ k−1(0) = ∅ and f−1(0) ∩ g−1(0) 6= ∅, we have η∗(E) ∈ {2, 3}. Observe that {f, g}, {f, h} and {f, k} are the only two-element subsets of E which have a common zero. Consequently, no three-element subset of E has a common zero. It follows, therefore, that η∗(E) = 3. Notice that in the example above, if I is a proper ideal of C[x, y], then I ⊆ ℘ for an ideal ℘ ∈ Max(C[x, y]) and hence η∗(I) = 1. Let K be a nonempty compact subset of Rn and let C(K,R) stand for the ring of all the continuous functions f : K −→ R. Recall that C(K,R)× = {f ∈ C(K,R) : f(x) 6= 0 for all x ∈ K}. The map Ψ : K ∋ x 7−→ {f ∈ C(K,R) : f(x) = 0} ∈ Max(C(K,R)) is well known to be a bijection [3]. Consequently, if B is a σ-algebra of subsets of K and ξ : B −→ [0, +∞] is a measure, then M = {Ψ(Z) : Z ∈ B} is a σ-algebra of subsets of Max(C(K,R)) and η : M ∋ S 7→ ξ(Ψ−1(S)) ∈ [0, +∞] is a measure. The obvious counterpart of Proposition 3 remains true. 56 An outer measure on a commutative ring Example 3. Let η∗ : 2C(K,R) −→ [0, +∞] be the outer measure induced by η. We will denote by W the set of all the polynomial functions f : K −→ R. Since ∀ x ∈ K ∃ f ∈ W : f−1(0) = {x}, we have η∗(W ) = η∗(C(K,R)) = ξ(K). Now, suppose that K is the Euclidean closed unit ball and ξ is the n-dimensional Lebesgue measure. If E stands for the set of all the radially symmetric functions belonging to C(K,R) and L is the straight line segment that joins the origin to a boundary point of K, then ∀ f ∈ E \ C(K,R)× : L ∩ f−1(0) 6= ∅. Consequently, η∗(E) = ξ(L) = 0 whenever n > 2. It is easy to see that if n = 1, then η∗(E) = 1. 3. General properties In the theorem below (it is the main result of the paper) we use the notations and assumptions of Proposition 1. For n ∈ N \ {0} and A1, . . . , An ⊆ R we define A1 . . . An = {a1 . . . an : a1 ∈ A1, . . . , an ∈ An}. Moreover, if A ⊆ R, then An = {an : a ∈ A} and A•n = A . . . A ︸ ︷︷ ︸ n . Theorem 1. Let A, B ⊆ R and let C be a nonempty subset of R×. Then (i) µ∗(R×) = 0, (ii) µ∗(A) = µ∗(A \ R×), (iii) µ∗({0}) = min{µ∗(E) : E ⊆ R, E \ R× 6= ∅}, (iv) ∀ n ∈ N \ {0} : µ∗(An) = µ∗(A•n) = µ∗(A), (v) µ∗(AC) = µ∗(A), (vi) µ∗(AB) > max{µ∗(A), µ∗(B)} whenever A ∩ R× 6= ∅ and B ∩ R× 6= ∅, (vii) µ∗(AB) 6 µ∗(A) + µ∗(B), (viii) µ∗(AB) = µ∗(A) whenever A ∩ R× = ∅ and B ∩ R× 6= ∅, (ix) µ∗(AB) = min{µ∗(A), µ∗(B)} whenever A ∩ R× = ∅ and B ∩ R× = ∅. Proof. Properties (i) and (ii) are obvious. Property (iii) follows from the facts that 0 /∈ R× and 0 belongs to every ideal of R. Fix a positive integer n. Let a1, . . . , an ∈ R. The product a1 . . . an is not invertible if and only if there exists an index i ∈ {1, . . . , n} such that D. Dudzik, M. Skrzyński 57 ai is not invertible. Similarly, a1 . . . an ∈ ℘ for an ideal ℘ ∈ Spec(R) if and only if there exists an index i ∈ {1, . . . , n} such that ai ∈ ℘. Therefore, Ω(An) = Ω(A•n) = Ω(A). Property (iv) follows. Let a ∈ R and c ∈ R×. Observe that ac /∈ R× if and only if a /∈ R×. Moreover, ∀ ℘ ∈ Spec(R) : ac ∈ ℘ ⇔ a ∈ ℘. Consequently, Ω(AC) = Ω(A). Suppose that C1 = A ∩ R× 6= ∅ and C2 = B ∩ R× 6= ∅. Since AC2 ∪BC1 ⊆ AB, we have max{µ∗(AC2), µ∗(BC1)} 6 µ∗(AB). Property (v) yields µ∗(AC2) = µ∗(A) and µ∗(BC1) = µ∗(B). This completes the proof of (vi). Let S ∈ Ω(A) and T ∈ Ω(B). Suppose that ab /∈ R× for some a ∈ A and b ∈ B. Then a /∈ R× or b /∈ R×. By the definition of ideal, we get therefore ab ∈ ⋃ S ∪ ⋃ T . Consequently, S ∪ T ∈ Ω(AB) and hence µ∗(AB) 6 µ(S) + µ(T ). Since S and T are arbitrarily chosen, it follows that µ∗(AB) 6 µ∗(A) + µ∗(B). Assume that A ∩ R× = ∅ and C2 = B ∩ R× 6= ∅. Then, by the defi- nition of ideal, Ω(A) ⊆ Ω(AB) which implies that µ∗(AB) 6 µ∗(A). On the other hand, by (v), we have µ∗(A) = µ∗(AC2) 6 µ∗(AB). Therefore, µ∗(AB) = µ∗(A). Finally, assume that A ∩ R× = ∅ and B ∩ R× = ∅. Then µ∗(AB) 6 min{µ∗(A), µ∗(B)} (cf. the proof of property (viii)). Suppose now that µ∗(AB) < min{µ∗(A), µ∗(B)}. Then ∃ U ∈ Ω(AB) : { µ∗( ⋃ U) < µ∗(A), µ∗( ⋃ U) < µ∗(B). (Notice that µ∗( ⋃ U) 6 µ(U)). Consequently, µ∗(A \ ⋃ U) > µ∗(A) − µ∗(A ∩ ⋃ U) > µ∗(A) − µ∗( ⋃ U) > 0 and, in the same way, µ∗(B \ ⋃ U) > 0. Since AB ∩ R× = ∅ and therefore AB ⊆ ⋃ U , we get ∃ a ∈ A ∃ b ∈ B ∃ ℘ ∈ U ⊆ Spec(R) : { ab ∈ ℘, a /∈ ℘, b /∈ ℘, a contradiction. Property (ix) follows. 58 An outer measure on a commutative ring We will conclude the paper with an example illustrating the behavior of µ∗(AB) in the case where A and B both contain invertible elements. Example 4. Let µ∗ : 2Z −→ [0, +∞] be the outer measure induced by the counting measure on Max(Z). If A = {1, 2, 3}, B1 = {1, 2, 3, 5}, B2 = {1, 2, 5, 7} and B3 = {1, 5, 7, 11}, then µ∗(A) = 2, µ∗(B1) = µ∗(B2) = µ∗(B3) = 3, µ∗(AB1) = 3, µ∗(AB2) = 4 and µ∗(AB3) = 5. References [1] M. F. Atiyah, I. G. MacDonald, Introduction to Commutative Algebra, Westview Press, 1994. [2] H. Federer, Geometric Measure Theory, Springer, 1969. [3] W. Rudin, Functional Analysis, McGraw-Hill, 1991. Contact information D. Dudzik Institute of Mathematics, Pedagogical University of Cracow, ul. Podchora̧żych 2, 30-084 Kraków, Poland E-Mail(s): dariusz.dudzik@gmail.com M. Skrzyński Institute of Mathematics, Cracow University of Technology, ul. Warszawska 24, 31-155 Kraków, Poland E-Mail(s): pfskrzyn@cyf-kr.edu.pl Received by the editors: 15.10.2015 and in final form 05.01.2016.
id nasplib_isofts_kiev_ua-123456789-155199
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
issn 1726-3255
language English
last_indexed 2025-11-24T12:54:40Z
publishDate 2016
publisher Інститут прикладної математики і механіки НАН України
record_format dspace
spelling Dudzik, D.
Skrzyński, M.
2019-06-16T10:48:33Z
2019-06-16T10:48:33Z
2016
An outer measure on a commutative ring / D. Dudzik, M. Skrzyński // Algebra and Discrete Mathematics. — 2016. — Vol. 21, № 1. — С. 51-58. — Бібліогр.: 3 назв. — англ.
1726-3255
2010 MSC:13A15, 28A12.
https://nasplib.isofts.kiev.ua/handle/123456789/155199
We show how to produce a reasonable outer measure on a commutative ring from a given measure on a family of prime ideals of this ring. We provide a few examples and prove several properties of such outer measures.
en
Інститут прикладної математики і механіки НАН України
Algebra and Discrete Mathematics
An outer measure on a commutative ring
Article
published earlier
spellingShingle An outer measure on a commutative ring
Dudzik, D.
Skrzyński, M.
title An outer measure on a commutative ring
title_full An outer measure on a commutative ring
title_fullStr An outer measure on a commutative ring
title_full_unstemmed An outer measure on a commutative ring
title_short An outer measure on a commutative ring
title_sort outer measure on a commutative ring
url https://nasplib.isofts.kiev.ua/handle/123456789/155199
work_keys_str_mv AT dudzikd anoutermeasureonacommutativering
AT skrzynskim anoutermeasureonacommutativering
AT dudzikd outermeasureonacommutativering
AT skrzynskim outermeasureonacommutativering