Form of filters of semisimple modules and direct sums

Some collections of submodules of a module defined by certain conditions are studied. A generalization of the notion of radical (preradical) filter is considered. We study the form of filters of semisimple modules and direct sums.

Gespeichert in:
Bibliographische Detailangaben
Datum:2013
1. Verfasser: Maturin, Yu.
Format: Artikel
Sprache:English
Veröffentlicht: Інститут прикладної математики і механіки НАН України 2013
Schriftenreihe:Algebra and Discrete Mathematics
Online Zugang:http://dspace.nbuv.gov.ua/handle/123456789/152349
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:Form of filters of semisimple modules and direct sums / Yu. Maturin // Algebra and Discrete Mathematics. — 2013. — Vol. 16, № 2. — С. 226–232. — Бібліогр.: 5 назв. — англ.

Institution

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id irk-123456789-152349
record_format dspace
spelling irk-123456789-1523492019-06-11T01:25:20Z Form of filters of semisimple modules and direct sums Maturin, Yu. Some collections of submodules of a module defined by certain conditions are studied. A generalization of the notion of radical (preradical) filter is considered. We study the form of filters of semisimple modules and direct sums. 2013 Article Form of filters of semisimple modules and direct sums / Yu. Maturin // Algebra and Discrete Mathematics. — 2013. — Vol. 16, № 2. — С. 226–232. — Бібліогр.: 5 назв. — англ. 1726-3255 2010 MSC:16D90, 16D10. http://dspace.nbuv.gov.ua/handle/123456789/152349 en Algebra and Discrete Mathematics Інститут прикладної математики і механіки НАН України
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
language English
description Some collections of submodules of a module defined by certain conditions are studied. A generalization of the notion of radical (preradical) filter is considered. We study the form of filters of semisimple modules and direct sums.
format Article
author Maturin, Yu.
spellingShingle Maturin, Yu.
Form of filters of semisimple modules and direct sums
Algebra and Discrete Mathematics
author_facet Maturin, Yu.
author_sort Maturin, Yu.
title Form of filters of semisimple modules and direct sums
title_short Form of filters of semisimple modules and direct sums
title_full Form of filters of semisimple modules and direct sums
title_fullStr Form of filters of semisimple modules and direct sums
title_full_unstemmed Form of filters of semisimple modules and direct sums
title_sort form of filters of semisimple modules and direct sums
publisher Інститут прикладної математики і механіки НАН України
publishDate 2013
url http://dspace.nbuv.gov.ua/handle/123456789/152349
citation_txt Form of filters of semisimple modules and direct sums / Yu. Maturin // Algebra and Discrete Mathematics. — 2013. — Vol. 16, № 2. — С. 226–232. — Бібліогр.: 5 назв. — англ.
series Algebra and Discrete Mathematics
work_keys_str_mv AT maturinyu formoffiltersofsemisimplemodulesanddirectsums
first_indexed 2025-07-13T02:52:57Z
last_indexed 2025-07-13T02:52:57Z
_version_ 1837498559625166848
fulltext Algebra and Discrete Mathematics RESEARCH ARTICLE Volume 16 (2013). Number 2. pp. 226 – 232 c© Journal “Algebra and Discrete Mathematics” Form of filters of semisimple modules and direct sums Yuriy Maturin Communicated by A. I. Kashu Abstract. Some collections of submodules of a module defined by certain conditions are studied. A generalization of the notion of radical (preradical) filter is considered. We study the form of filters of semisimple modules and direct sums. All rings are considered to be associative with unit 16=0 and all modules are left and unitary. Let R be a ring. Put (N : f)M = {x ∈ M |f(x) ∈ N} , End(M)N = {f ∈ End(M)|f(M) ⊆ N} . Let E be some non-empty collection of submodules of a left R-module M . We consider the following conditions: (1) L ∈ E,L ≤ N ≤ M ⇒ N ∈ E; (2) L ∈ E, f ∈ End(M) ⇒ (L : f)M ∈ E; (3) N,L ∈ E ⇒ N ∩ L ∈ E; (4) N ∈ E,N ∈ Gen(M), L ≤ N ≤ M ∧ ∧ ∀g ∈ End(M)N : (L : g)M ∈ E ⇒ L ∈ E; (5) N,L ∈ E,N ∈ Gen(M) ⇒ N ∩ L ∈ E. 2010 MSC: 16D90, 16D10. Key words and phrases: ring, module, filter. Yu. Maturin 227 Consider a generalization of the notion of radical (preradical) filter (see [3, 4]). A non-empty collection E of submodules of a left R-module M satis- fying ((1)), ((2)), ((3)) is called a preradical filter of M (see [4]). A non-empty collection E of submodules of a left R-module M satis- fying ((1)), ((2)), ((4)) is called a radical filter of M (see [4]). It is easy to see that for every radical filter of M ((5)) is held. A preradical (radical) filter E of a left R-module M is said to be trivial if either E = {L|L ≤M} or E = {M}. Let M be a semisimple left R-module with a unique homogeneous component and let M = ⊕ i∈I Mi, where Mi is simple for each i ∈ I. If N = ⊕ i∈J Ni, where Ni is simple for each i ∈ J and M ∼= N , then Card(I) = Card(J). Put Cards(M) := Card(I). Let M be a semisimple R-module with a unique homogeneous compo- nent. If Cards(M) is infinite, then we set Ep(M) := {L|L ≤M,Cards(M/L) < p}, where p is an infinite cardinal number. Theorem 1. Let M be a semisimple R-module with a unique homoge- neous component. If Cards(M) is infinite, then every non-trivial radical [preradical] filter of M is of the form Ep(M) for some infinite cardinal number p ≤ Cards(M). Proof. Let M be a semisimple R-module with a unique homogeneous component, Cards(M) = ∞, and E a non-trivial radical [preradical] filter of M . Put q := Cards(M). It is obvious that for each L ∈ E there exists H ≤ M such that M = L⊕H. Hence CardsH ≤ q. We claim that CardsH 6= q. Indeed, suppose, contrary to our claim, that CardsH = q. Since M is a semisimple R-module with a unique ho- mogeneous component, for some set I we have that M = ⊕ i∈I Mi, where Mi 228Form of filters of semisimple modules and direct sums is simple for each i ∈ I and for every i, j ∈ I there exists an isomorphism fij : Mi → Mj . Hence CardI = Cards(M) = q. Taking into account that Cards(M) is infinite, by (2.1) [5, p. 417], q + q = q. Consider a set X such that CardX = q and X∩I = ∅. Since q+q = q, there exists a bijection w : X ∪ I → I. Put Y := w(X), Z := w(I). Therefore, I = Y ∪Z, Y ∩Z = ∅, q = CardI = CardY = CardZ. Now we obtain M = A⊕B, where A = ⊕ i∈Y Mi,B = ⊕ i∈Z Mi. Since H ≤ M , there exists an isomorphism u : H → ⊕ i∈T Mi for some T ⊆ I (see Proposition 9.4 [1]). It is clear that CardsH = CardT = q. Whence q = CardY = CardZ = CardT . Let g : Y → T, c : Z → T be bijections. Consider the following maps: G : A → H,C : B → H, where G( ∑ i∈Y mi) = u−1( ∑ i∈Y fi,g(i)(mi)), (mi ∈ Mi(i ∈ I), Card{i ∈ Y |mi 6= 0} < ∞), C( ∑ i∈Z mi) = u−1( ∑ i∈Z fi,c(i)(mi)), (mi ∈ Mi(i ∈ I), Card{i ∈ Z|mi 6= 0} < ∞). It is easily seen that these maps are isomorphisms. Let n, r : M → M are maps such that n(a + b) = G(a), (a ∈ A, b ∈ B) and r(a + b) = C(b), (a ∈ A, b ∈ B). It is clear that n, r : M → M are endomorphisms. Since L ∩ H = 0 and G,C are isomorphisms, (L : n)M = B and (L : r)M = A. As L ∈ E , by ((2)), we get B ∈ E and A ∈ E. By ((3)) or ((5)) , 0 = A ∩ B ∈ E. Consequently, E is trivial. This contradicts our assumption. Hence CardsH < q. The natural isomorphism H ∼= M/L implies that Cards(M/L) < q. Now we consider the set Ω of all cardinal numbers v such that v ≤ q∀L ∈ E : Cards(M/L) < v. Yu. Maturin 229 Ω 6= ∅, because q ∈ Ω. By [2, p. 82] , there exists the least element p belonging to Ω. Thus ∀L ∈ E : Cards(M/L) < p. It means that E ⊆ Ep(M). Let L ∈ Ep(M). Whence Cards(M/L) < p. We claim that there exists D ∈ E such that Cards(M/L) ≤ Cards(M/D). Conversely, suppose that ∀D ∈ E : Cards(M/D) < Cards(M/L). But Cards(M/L) < p ≤ q. Hence Cards(M/L) ∈ Ω. Since p is the least element belonging to Ω, p ≤ Cards(M/L), contrary to Cards(M/L) < p. Now we have that there exists D ∈ E such that Cards(M/L) ≤ Cards(M/D). It is easily seen that for L,D there exist H,K ≤ M such that M = L⊕H,M = D⊕K. Since M/L ∼= H,M/D ∼= K, Cards(H) ≤ Cards(K). Since H ≤ M and K ≤ M , there exist isomorphisms u : H → ⊕ i∈T Mi for some T ⊆ I and w : K → ⊕ i∈S Mi for some S ⊆ I. Therefore CardT ≤ CardS. From this we have that there exists an injective map γ : T → S. Consider the following map: ψ : ⊕ i∈T Mi → ⊕ i∈S Mi, where ψ( ∑ i∈T mi) = ∑ i∈Y fi,γ(i)(mi), (mi ∈ Mi(i ∈ I), Card{i ∈ T |mi 6= 0} < ∞). It is obvious that ψ is a monomorphism. Now consider the following map: η : M → M, where η(l + h) = w−1ψu(h), (l ∈ L, h ∈ H). It is clear that η ∈ End(M). Since D ∩ K = 0 and im η ⊆ K, for every l ∈ K,h ∈ H: η(l + h) ∈ D ⇔ w−1ψu(h) ∈ D ⇔ w−1ψu(h) = 0. Since u,w are isomorphisms and ψ is monomorphism, for every h ∈ H: w−1ψu(h) = 0 ⇔ h = 0. From the above it follows that (D : η)M = L. Since E is a radical [preradical] filter of M and D ∈ E, (D : η)M = L shows that L ∈ E, by ((2)). It means that Ep(M) ⊆ E. But E ⊆ Ep(M). Hence E = Ep(M) 230Form of filters of semisimple modules and direct sums Theorem 2. If M is a left R-module such that M = M1 ⊕M2 ⊕ . . .⊕Mn, where Mi = TrM (Mi) for each i ∈ {1, 2, . . . , n} and ∀S : S ≤ M ⇒ S ∈ Gen(M), then every radical [preradical] filter E of M is of the form E = {J1 + J2 + . . .+ Jn|Ji ∈ Ei(i ∈ {1, 2, . . . , n})}, where Ei is a radical [preradical] filter of Mi for each i ∈ {1, 2, . . . , n}. Proof. Let E be a radical [preradical] filter of M and M = M1 ⊕M2 ⊕ . . .⊕Mn, where Mi = TrM (Mi) for each i ∈ {1, 2, . . . , n}. Put Ei := {fi(K)|K ∈E} for each i ∈ {1, 2, . . . , n}, where fi : M → M,fi(m1 +m2 + . . .+mn) = mi, (m1 ∈ M1,m2 ∈ M2, . . . ,mn ∈ Mn) for each i ∈ {1, 2, . . . , n}. (1) Let L ∈ Ei, L ≤ N ≤ Mi. Hence there exists P ∈ E such that L = fi(P ). Since L ≤ N , P ≤ f−1 i (N). By (1), f−1 i (N) ∈ E, because P ∈ E. Therefore N = fi(f −1 i (N)) ∈ Ei. (2) Let L ∈ Ei, f ∈ End(Mi). Hence there exists P ∈ E such that L = fi(P ). Consider F : M → M, where F : m1+m2+. . .+mi+. . .+mn 7→ f(mi), (m1 ∈ M1, . . . ,mn ∈ Mn). Thus F ∈ End(M). We claim that fi((P : F )M ) ≤ (L : f)Mi . Indeed, let xi ∈ fi((P : F )M ). We have that xi ∈ Mi. Thus there exists x ∈ (P : F )M such that fi(x) = xi. Hence f(xi) = F (x) ∈ P . It is clear that f(xi) ∈ Mi. Therefore f(xi) = fi(f(xi)) ∈ fi(P ) = L. Whence xi ∈ (L : f)Mi . We obtain fi((P : F )M ) ≤ (L : f)Mi . Since P ∈ E and F ∈ End(M), (P : F )M ∈ E, by (2). (P : F )M ∈ E implies fi((P : F )M ) ∈ Ei. Since fi((P : F )M ) ≤ (L : f)Mi , (1) implies (L : f)Mi ∈ Ei. (3) Let L,N ∈ Ei. Hence there exist P, T ∈ E such that L = fi(P ) and N = fi(T ). By (3) (for the preradical filter E), P ∩ T ∈ E. Therefore fi(P ∩ T ) ∈ Ei. Since fi(P ∩ T ) ⊆ fi(P ) ∩ fi(T ) = L∩N and fi(P ∩ T ) ∈ Ei, we obtain L ∩N ∈ Ei, by (1). (4) Let N ∈ Ei, N ∈ Gen(Mi), L ≤ N ≤ Mi ∧ ∀g ∈ End(Mi)N : (L : g)Mi ∈ Ei. Hence N = fi(T ) for some T ∈ E. Since T ⊆ f−1 i (N), f−1 i (N) ∈ E, by (1). And f−1 i (N) ∈ Gen(M). L ≤ N implies f−1 i (L) ≤ f−1 i (N). Yu. Maturin 231 Let G be an arbitrary element of End(M) f−1 i (N). By Proposition 8.16 [1] , Ms = TrM (Ms) is a fully invariant submodule of M for each s ∈ {1, 2, . . . , n}. Hence G(Ms) ⊆ Ms for each s ∈ {1, 2, . . . , n}. Consider g : Mi → Mi,m 7→ G(m), (m ∈ Mi). Since ∀g ∈ End(Mi)N : (L : g)Mi ∈ Ei, there exists Yg ∈ Ei such that g(Yg) ≤ L. Since G(Ms) ⊆ Ms for each s ∈ {1, 2, . . . , n}, G(f−1 i (Yg)) = G(M1 ⊕ . . .⊕ Mi−1 ⊕ Yg ⊕Mi+1 ⊕ . . .⊕Mn) ⊆ ⊆ M1 ⊕ . . .⊕Mi−1⊕ G(Yg) ⊕Mi+1 ⊕ . . .⊕Mn = = M1 ⊕ . . .⊕Mi−1⊕ g(Yg) ⊕Mi+1 ⊕ . . .⊕Mn ⊆ ⊆ M1 ⊕ . . .⊕Mi−1⊕ L⊕Mi+1 ⊕ . . .⊕Mn = f−1 i (L). Hence f−1 i (Yg) ⊆ (f−1 i (L) : G)M . Since Yg ∈ Ei, there exists P ∈ E such that Yg = fi(P ). Thus P ⊆ f−1 i (Yg). Hence P ⊆ (f−1 i (L) : G)M &P ∈ E. By (1), (f−1 i (L) : G)M ∈ E. Since f−1 i (N) ∈ E, f−1 i (N) ∈ Gen(M), f−1 i (L) ≤ f−1 i (N) ≤ M and ∀G ∈ End(M) f−1 i (N) : (f−1 i (L) : G)M ∈ E, obtain f−1 i (L) ∈ E. Therefore L = fi(f −1 i (L)) ∈ Ei. Let J ∈ E. Put Ji := fi(J), (i ∈ {1, 2, . . . , n}). By Proposition 8.20 [1], TrJ(M) = TrJ(M1 ⊕ M2 ⊕ . . . ⊕ Mn) = n∑ i=1 TrJ(Mi). Since J ≤ M , TrJ (Mi) ≤ TrM (Mi) = Mi for any i ∈ {1, 2, . . . , n}, by Proposition 8.16 [1]. Hence TrJ(M) = n ⊕ i=1 TrJ(Mi), because M = M1 ⊕M2 ⊕ . . .⊕Mn. Since J ∈ Gen(M), TrJ(M) = J , by Proposition 8.12 [1]. Whence J = n ⊕ i=1 TrJ(Mi)&∀i ∈ {1, 2, . . . , n} : TrJ(Mi) ≤ Mi. Therefore TrJ(Mi) = Ji for any i ∈ {1, 2, . . . , n}. Thus J = J1 + J2 + . . .+ Jn, where J1 ∈ E1, J2 ∈ E2, . . . , Jn ∈ En. Let Pi ∈ Ei for each i ∈ {1, 2, . . . , n}. Hence there exists Hi ∈ E such that Pi = fi(Hi). Thus Hi ⊆ f−1 i (Pi). By (1), f−1 i (Pi) ∈ E. f−1 i (Pi) ∈ Gen(M) for any i ∈ {1, 2, . . . , n}. By ((3)) or ((5)), f−1 1 (P1) ∩ f−1 2 (P2) ∩ . . .∩f−1 n (Pn) ∈ E. Since f−1 i (Pi) = M1+. . .+Mi−1+Pi+Mi+1 +. . .+Mn for any i ∈ {1, 2, . . . , n}, P1 + P2 + . . .+ Pn ∈ E. 232Form of filters of semisimple modules and direct sums References [1] F.W. Anderson, K.R. Fuller, Rings and categories of modules. Springer, Berlin- Heidelberg-New York, 1973. [2] F. Hausdorff, Set theory. Second edition. Chelsea Publishing Company, New York, 2005. [3] A.I. Kashu, Radicals and torsions in modules. Stiintca, Chisinau, 1983. (in Russian). [4] Yu.P. Maturin, Preradicals and submodules. Algebra and Discrete Mathematics, Vol.10 (2010), No.1, pp. 88-96. [5] W. Sierpinski, Cardinal and ordinal numbers. Second edition. PWN, Warsaw, 1965. Contact information Yuriy Maturin Institute of Physics, Mathematics and Computer Science,Drohobych Ivan Franko State Pedagog- ical University, Stryjska, 3, Drohobych 82100, Lviv Region, Ukraine E-Mail: yuriy_maturin@hotmail.com URL: www.drohobych.net/ddpu/ Received by the editors: 16.10.2012 and in final form 14.08.2013.