Rainbow graphs and semigroups

We give an algebraic characterization of rainbow graphs. A connected graph Γ is called rainbow if there is a vertex coloring of Γ, which is bijective on the set of neighbors of each vertex of Γ.

Збережено в:
Бібліографічні деталі
Дата:2012
Автори: Protasova, K.D., Provotar, Т.М.
Формат: Стаття
Мова:English
Опубліковано: Видавничий дім "Академперіодика" НАН України 2012
Назва видання:Доповіді НАН України
Теми:
Онлайн доступ:http://dspace.nbuv.gov.ua/handle/123456789/84296
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Цитувати:Rainbow graphs and semigroups / K.D. Protasova, Т.М. Provotar // Доп. НАН України. — 2012. — № 7. — С. 43-47. — Бібліогр.: 6 назв. — англ.

Репозитарії

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id irk-123456789-84296
record_format dspace
spelling irk-123456789-842962015-07-07T03:01:56Z Rainbow graphs and semigroups Protasova, K.D. Provotar, Т.М. Інформатика та кібернетика We give an algebraic characterization of rainbow graphs. A connected graph Γ is called rainbow if there is a vertex coloring of Γ, which is bijective on the set of neighbors of each vertex of Γ. Отримано алгебраїчну характеризацiю веселкових графiв. Зв’язний граф Γ називається веселковим, якщо iснує розфарбування множини вершин Γ, що є бiєктивним на множинi сусiдiв кожної вершини Γ. Получена алгебраическая характеризация радужных графов. Связный граф Γ называется радужным, если существует раскраска множества вершин Γ, биективная на множестве соседей для каждой вершины Γ. 2012 Article Rainbow graphs and semigroups / K.D. Protasova, Т.М. Provotar // Доп. НАН України. — 2012. — № 7. — С. 43-47. — Бібліогр.: 6 назв. — англ. 1025-6415 http://dspace.nbuv.gov.ua/handle/123456789/84296 519.1 en Доповіді НАН України Видавничий дім "Академперіодика" НАН України
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
language English
topic Інформатика та кібернетика
Інформатика та кібернетика
spellingShingle Інформатика та кібернетика
Інформатика та кібернетика
Protasova, K.D.
Provotar, Т.М.
Rainbow graphs and semigroups
Доповіді НАН України
description We give an algebraic characterization of rainbow graphs. A connected graph Γ is called rainbow if there is a vertex coloring of Γ, which is bijective on the set of neighbors of each vertex of Γ.
format Article
author Protasova, K.D.
Provotar, Т.М.
author_facet Protasova, K.D.
Provotar, Т.М.
author_sort Protasova, K.D.
title Rainbow graphs and semigroups
title_short Rainbow graphs and semigroups
title_full Rainbow graphs and semigroups
title_fullStr Rainbow graphs and semigroups
title_full_unstemmed Rainbow graphs and semigroups
title_sort rainbow graphs and semigroups
publisher Видавничий дім "Академперіодика" НАН України
publishDate 2012
topic_facet Інформатика та кібернетика
url http://dspace.nbuv.gov.ua/handle/123456789/84296
citation_txt Rainbow graphs and semigroups / K.D. Protasova, Т.М. Provotar // Доп. НАН України. — 2012. — № 7. — С. 43-47. — Бібліогр.: 6 назв. — англ.
series Доповіді НАН України
work_keys_str_mv AT protasovakd rainbowgraphsandsemigroups
AT provotartm rainbowgraphsandsemigroups
first_indexed 2025-07-06T11:17:11Z
last_indexed 2025-07-06T11:17:11Z
_version_ 1836896106071457792
fulltext УДК 519.1 © 2012 K.D. Protasova, Т.М. Provotar Rainbow graphs and semigroups (Presented by Corresponding Member of the NAS of Ukraine S. I. Lyashko) We give an algebraic characterization of rainbow graphs. A connected graph Γ is called rainbow if there is a vertex coloring of Γ, which is bijective on the set of neighbors of each vertex of Γ. A rainbow graph [3] is a connected graph Γ with the set of vertices V (Γ) and the set of edges E(Γ) that can be vertex-colored χ : V (Γ) −→ κ so that every color x ∈ κ is represented once, and only once, among the neighbors N(v) = {u ∈ V (Γ): {u, v} ∈ E(Γ)} of each vertex v ∈ V (Γ). For applications of rainbow graphs, see [1]. If one removes the edge-matching of the monochrome edges of a rainbow graph, one gets a kaleidoscopical graph [2, Chapter 6]. Let κ be a cardinal. A rainbow semigroup RS(κ) is a semigroup in the alphabet κ determined by the relations xxx = x, xyx = x for all x, y ∈ κ. We identify RS(κ) with the set of all non- empty words in κ with no factors xxx, xyx. For x ∈ κ, a rainbow group RG(κ, x) is a subset of RS(κ) containing x and all words of the form xwx, w ∈ RS(κ). The word xx is the identity of RG(κ, x), x−1 = x, and (xwx)−1 = xxw̃xx, where w̃ is the word w written in the reverse order. Theorem 1. For any cardinal κ and each x ∈ κ, the following statements hold: (i) the idempotents of RS(κ) are only yz, where y, z ∈ κ; (ii) RG(κ, x) is a free product of the cyclic group 〈x〉 of order 2 and the family of infinite cyclic groups {〈xabx〉 : a, b ∈ κ, a 6= x, b 6= x}; (iii) RS(κ) is a sandwich product RS(κ) = L(x)×RG(κ, x)×R(x), where L(x) = {yx : y ∈ ∈ κ}, R(x) = {xy : y ∈ κ}, and the multiplication (l1, w1, r1)(l2, w2, r2) = (l1, w1r1l2w2, r2). Let κ be a cardinal, x ∈ κ. An equivalence ∼ on RS(κ) is called a rainbow equivalence if, for any w1, w2 ∈ RS(κ), we have • w1 ∼ w2 =⇒ l(w1) = l(w2), where l(w) is the first letter of w; • w1 ∼ w2 =⇒ yw1 = yw2 for each y ∈ κ; • l(w) = y =⇒ w and yw are not equivalent; • w ∼ wxx for each w ∈ KS(κ). Each rainbow equivalence ∼ on RS(κ) determines the rainbow graph Γ(κ, k) as follows. The set V (Γ) of vertices of Γ is a factor-set RS(κ)/ ∼= {[w] : w ∈ RS(κ)}, where [w] is the class of equivalence ∼ containing w. By definition, {u, v} ∈ E(Γ) if and only if u 6= v, and there exists w ∈ u such that yw ∈ v. Then the mapping χ : V (Γ) −→ κ defined by χ([w]) = l(w) does not depend on the choice of w and determines a rainbow coloring of Γ. In turn, every rainbow equivalence ∼ on RS(κ) is uniquely determined by the subgroup Sx = [xx] ⋂ RG(κ, x) of RG(κ, x) because w1 ∼ w2 ⇐⇒ l(w1) = l(w2) ∧ xw1xx ∼ xw2xx ⇐⇒ (xw1xx) −1(xw2xx) ∈ Sx. ISSN 1025-6415 Доповiдi Нацiональної академiї наук України, 2012, №7 43 We say that two rainbow graphs Γ1, Γ2 with rainbow colorings χ1 : V (Γ1) −→ κ, χ2 : V (Γ2) −→ κ are rainbow isomorphic if there exists a bijection f : V (Γ1) −→ V (Γ2) such that • ∀u, v ∈ V (Γ1) : {u, v} ∈ E(Γ1) ⇐⇒ {f(u), f(v)} ∈ E(Γ2); • ∀u ∈ V (Γ1) : χ1(u) = χ2(f(u)). Now, we are ready to characterize all rainbow graphs up to rainbow isomorphisms. Let Γ be a rainbow graph with rainbow coloring χ : V (Γ) −→ κ. We define a transitive action of RS(κ) on the set V (Γ) as follows. Let v ∈ V (Γ), x ∈ κ. Pick u ∈ N(v) such that χ(u) = x and put x(v) = u. Then we extend the action onto KS(κ) inductively. If w = RS(κ), w = xw′, x ∈ κ, we put w(v) = χ(w′(v)). Given any v1, v2 ∈ V (Γ), the sequence of colors of the vertices on a path from v1 to v2 determines a word w ∈ RS(κ) such that w(v1) = v2, so RS(κ) acts on V (Γ) transitively. Clearly, the group RG(κ, x) acts transitively on the set of vertices of color x. We fix v ∈ V (Γ) with χ(v) = x, determine a rainbow equivalence ∼ on RS(κ) by the rule w ∼ w′ ⇐⇒ w(v) = w′(v), and note that the graphs Γ and Γ(κ,∼) are rainbow isomorphic via the bijection f : V (Γ) −→ → KS(κ)/ ∼, f(u) = {w ∈ KS(κ) : w(v) = u}. Thus, we get the following statement. Theorem 2. For every rainbow graph Γ with rainbow coloring χ : V (Γ) −→ κ, there exists a rainbow equivalence ∼ on RS(κ) such that Γ and Γ(κ,∼) are rainbow isomorphic. Every rainbow equivalence on RS(κ) is uniquely determined by some subgroup of RG(κ, x). Let Γ(V,E) be a connected graph with the set of vertices V , and let the set of edges E, d be the path metric on V , B(v, r) = {u ∈ V : d(v, u) 6 r}, v ∈ V , r ∈ ω = {0, 1, . . .}. A graph Γ(V,E) is called kaleidoscopical [6] if there exists a coloring (a surjective mapping) χ : V −→ κ, κ is a cardinal such that the restriction χ | B(v, 1): B(v, 1) −→ κ is a bijection on each unit ball B(v, 1), v ∈ V . For kaleidoscopical graphs, see also [2, Chapter 6] and [5]. Let G be a group, and let X be a transitive G-space with the action G×X −→ X, (g, x) 7−→ → gx. A subset A of X, |A| = κ is said to be a kaleidoscopical configuration [4] if there exists a coloring χ : X −→ κ such that, for each g ∈ G, the restriction χ | gA : gA −→ κ is a bijection. We note that kaleidoscopical graphs and kaleidoscopical configurations can be considered as partial cases of kaleidoscopical hypergraphs defined in [2, p.5]. Recall that a hypergraph is a pair (X,F), where X is a set, F is a family of subsets of X. A hypergraph (X,F) is said to be kaleidoscopical if there exists a coloring χ : X −→ κ such that, for each F ∈ F, the restriction χ | F : F −→ κ is a bijection. Clearly, a graph Γ(V,E) is kaleidoscopical if and only if the hypergraph (V, {B(v, 1): v ∈ V }) is kaleidoscopical. A subset A of a G-space X is kaleidoscopical if and only if the hypergraph (X, {g(A) : g ∈ G}) is kaleidoscopical. We say that two hypergraphs (X1,F1), (X2,F2) with kaleidoscopical colorings χ1 : X1 −→ κ, χ2 : X2 −→ κ are kaleidoscopically isomorphic if there is a bijection f : X1 −→ X2 such that • ∀A ⊆ X1 : A ∈ F1 ⇐⇒ f(A) ∈ F2; • ∀x ∈ X1 : χ1(x) = χ2(f(x)). We describe an algebraic construction which gives all kaleidoscopical graphs up to isomor- phisms. The kaleidoscopical semigroup KS(κ) is a semigroup in the alphabet κ determined by the relations xx = x, xyx = x for all x, y ∈ κ. For our purposes, it is convenient to identify KS(κ) with the set of all non-empty words in κ with no factors xx, xyx, where x, y ∈ κ. 44 ISSN 1025-6415 Reports of the National Academy of Sciences of Ukraine, 2012, №7 For every x ∈ κ, the set KG(κ, x) of all words from KS(κ) with the first and the last letter x is a subgroup (with the identity x) of the semigroup KS(κ). To obtain the inverse element to the word w ∈ KG(κ, x), it suffices to write w in the inverse order. The group KG(κ, x) is called the kaleidoscopical group in the alphabet κ with the identity x. For finite cardinals κ, the following theorem is proved in [2, pp. 64–66]: but corresponding arguments work for arbitrary κ. Theorem 3. For any cardinal κ, the following statements hold: (i) idempotents of the semigroup KS(κ) are the only words x, xy, where x, y ∈ κ, x 6= y, (ii) the kaleidoscopical group KG(k, x) is a free group with the set of free generators {xyzx : y, z ∈ κ \ {x}, y 6= z}, (iii) the kaleidoscopical semigroup KS(κ) is isomorphic to the sandwich product L(x) × × KG(κ, x) × R(x) with the multiplication (l1, g1, r1)(l2, g2, r2) = (l1, g1r1l2g2, r2), where L(x) = {yx : y ∈ κ}, R(x) = {xy : y ∈ κ}. We fix x ∈ κ, denote the first letter of the word w ∈ KS(κ) by æ(w), and say that an equivalence ∼ on KS(κ) is kaleidoscopical if, for all w,w′ ∈ KS(κ) and y ∈ κ, w ∼ w′ =⇒ æ(w) = æ(w′) ∧ yw = yw′, w ∼ w′ ⇐⇒ wx ∼ w′x. Let [w] be the class of equivalence ∼ containing w ∈ KS(κ). We put Sx = [x] ⋂ KG(κ, x), observe that Sx is a subgroup of KG(κ, x), and show that ∼ is uniquely determined by Sx: w ∼ w′ ⇐⇒ æ(w) = æ(w′) ∧ xwx ∼ xw′x ⇐⇒ (xwx)−1(xw′x) ∈ Sx. We see also that any subgroup of KG(κ, x) can be taken as Sx to determine a kaleidoscopical equivalence on KS(κ). A kaleidoscopical equivalence ∼ determines a graph Γ(κ,∼) with the set of vertices KS(κ)/ ∼ and the set of edges E defined by the rule: (u, v) ∈ E ⇐⇒ ∃w ∈ u∃ y ∈ κ : æ(w) 6= y ∧ yw ∈ v. A coloring χ : KS(κ)/ ∼−→ κ defined by χ([w]) = æ(w) shows that Γ(κ,∼) is kaleidoscopical. Now let Γ(V,E) be a kaleidoscopical graph with kaleidoscopical coloring χ : V −→ κ. We define a transitive action of the semigroup KS(κ) on the set V as follows. Let v ∈ V , x ∈ κ. Pick u ∈ B(v, 1) such that χ(u) = x and put x(v) = u. Then we extend the action onto KS(κ) inductively. If w = KS(κ), w = xw′, w′ ∈ KS(κ), x ∈ κ, we put w(v) = x(w′(v)). Given any v1, v2 ∈ V , the sequence of colors of the vertices on a path from v1 to v2 determines a word w ∈ KS(κ) such that w(v1) = v2, so KS(κ) acts on V transitively. Clearly, the group KG(κ, x) acts transitively on the set χ−1(x) of vertices of color x. ISSN 1025-6415 Доповiдi Нацiональної академiї наук України, 2012, №7 45 We fix v ∈ V with χ(v) = x, determine a kaleidoscopical equivalence ∼ on KS(κ) by the rule w ∼ w′ ⇐⇒ w(v) = w′(v), and note that the graphs Γ(V,E) and Γ(κ,∼) are kaleidoscopically isomorphic via the bijection f : V −→ KS(κ)/ ∼, f(u) = {w ∈ KS(κ) : w(v) = u}. All above considerations are focused in the following theorem. Theorem 4. For every kaleidoscopical graph Γ(V,E) with kaleidoscopical coloring χ : V −→ → κ, there exists a kaleidoscopical equivalence ∼ on the semigroup KS(κ) such that Γ(V,E) is kaleidoscopically isomorphic to Γ(κ,∼). Every kaleidoscopical equivalence ∼ on KS(κ) is uni- quely determined by some subgroup of the group KG(κ, x). Every group G can be considered as a G-space with the left regular action (g, x) 7−→ yx. Let A be a kaleidoscopical configuration in G. By [4, Corollary 1.3], A is complemented, i. e. there exists a subset B of G such that the multiplication A×B −→ G, (a, b) 7−→ ab is bijective. Let A be a system of generators of a group G such that A = A−1 and e ∈ A, e is the identity of G. We consider the Cayley graph Cay(G,A) with the set of vertices G and the set of edges E defined by the rule: (g, h) ∈ E ⇐⇒ g−1h ∈ A, g 6= h. Clearly, Cay(G,A) is connected. Assume that Cay(G,A) is kaleidoscopical with kaleidoscopi- cal coloring χ : G −→ |A|. Since B(g, 1) = gA and χ is bijective on each ball B(g, 1), we see that A is a kaleidoscopical configuration. On the other hand, if A is a kaleidoscopical confi- guration in G with kaleidoscopical coloring χ : G −→ A, then χ is bijective on each set gA. So, Cay(G,A) is kaleidoscopical. Thus, we get the following theorem. Theorem 5. Let G be a group, and let A be a system of generators of G such that A = A−1 and e ∈ A. Then A is a kaleidoscopical configuration if and only if Cay(G,A) is kaleidoscopical. We conclude the paper with two open questions. Question 1. How can one detect whether a kaleidoscopical hypergraph is kaleidoscopically isomorphic to a hypergraph of unit balls of some kaleidoscopical graph? Question 2. How can one detect whether a kaleidoscopical hypergraph is kaleidoscopically isomorphic to a hypergraph determined by a kaleidoscopical configuration in a G-space? 1. Lazebnik F., Woldar A. J. General properties of some families of graphs defined by systems of equations // J. Graph Theory. – 2001. – 38. – P. 65–86. 2. Protasov I., Banakh T. Ball structures and colorings of groups and graphs. – Lviv: VNTL, 2003. – 148 p. 3. Woldar A. J. Rainbow graphs // Codes and Designs, edited by K.T. Arasu and A. Seress. – Berlin: W. de Gruyter, 2002. – P. 313–322. 4. Banakh T., Petrenko O., Protasov I., Slobodianuk S. Kaleidoscopical configurations in G-space // Electron. J. Combinatorics. – 2012. – 19. – 16 p. 5. Protasov I.V., Protasova K.D. Kaleidoscopical graphs and Hamming codes // Voronoi’s Impact on Modern Science. – Kiev: Institute Math. of the NAS of Ukraine, 2008. – P. 240–245. 6. Protasova K.D. Kaleidoscopical graphs // Mat. Studii. – 2002. – 18. – P. 3–9. Received 23.12.2011Taras Schevchenko Kyiv National University 46 ISSN 1025-6415 Reports of the National Academy of Sciences of Ukraine, 2012, №7 K.Д. Протасова, T.M. Провотар Веселковi графи i напiвгрупи Отримано алгебраїчну характеризацiю веселкових графiв. Зв’язний граф Γ називається ве- селковим, якщо iснує розфарбування множини вершин Γ, що є бiєктивним на множинi сусiдiв кожної вершини Γ. K.Д. Протасова, T.M. Провотар Радужные графы и полугруппы Получена алгебраическая характеризация радужных графов. Связный граф Γ называется радужным, если существует раскраска множества вершин Γ, биективная на множестве соседей для каждой вершины Γ. ISSN 1025-6415 Доповiдi Нацiональної академiї наук України, 2012, №7 47