Ivan Shishkin, Rye (1878)

Problems/General algebraUnreviewed

From nothing to a group

by pietro.pende·
40
Difficulty scaleÉchelle de difficulté

This score reflects both the level of the required concepts and the difficulty of the solution.Ce score tient compte à la fois du niveau des notions nécessaires et de la difficulté de la résolution.

  1. 1–10First steps / middle schoolPremiers pas / collège
  2. 11–25Beginner / high schoolDébutant / lycée
  3. 26–50Intermediate / undergraduateIntermédiaire / licence
  4. 51–70Advanced / graduateAvancé / master
  5. 71–90Expert / specializedExpert / spécialisé
  6. 91–100Research levelNiveau recherche
These levels are approximate guides.Ces niveaux sont des repères approximatifs.
·
English
EnglishFrançais
This translation may be outdated. Its source text has changed since revision 4712.
Unreviewed. This problem has not been reviewed by trusted users yet.

Let EE be a finite set. Can EE be endowed with a group structure ?

I solved itMark it doneAdd to my listKeep it in your list

Solutions

2
Reveal solutionsAre you sure? Give it a try first.

Solution by Hugues HottFR

Discussions1 useful vote

On peut généraliser à un ensemble non vide quelconque. Pour EE un ensemble non vide, si GG est un groupe qui lui est équipotent, en désignant par φ\varphi une bijection de EE dans GG, alors EE peut être muni d’une structure de groupe pour la loi x∗y=φ−1(φ(x)φ(y))x \ast y = \varphi^{-1}(\varphi(x)\varphi(y)). On peut d’ailleurs vérifier que φ\varphi devient alors naturellement un isomorphisme de groupes. L’exercice se ramène donc à savoir s’il existe ou non un groupe de tout cardinal.

Dans le cas fini, G=Z/nZG = \mathbb Z/n\mathbb Z est un groupe de taille nn pour tout n≥1n \geq 1.

Pour XX un ensemble infini de cardinal κ\kappa, on peut montrer que S(X)′={σ∈S(X)∣supp(σ) est fini}S(X)' = \{\sigma \in S(X) \mid \mathrm{supp}(\sigma) \text{ est fini}\} est un groupe de même cardinal. En effet, pour tout P2(X)P_{2}(X) s’injecte dans S(X)′S(X)' via {x,y}↦(x y)\{x,y\} \mapsto (x~y), d’où ∣S(X)∣′≥∣P2(X)∣=κ|S(X)|' \geq |P_{2}(X)| = \kappa.
Réciproquement, soit σ∈S(X)′\sigma \in S(X)', alors σ\sigma est entièrement déterminée par son support EE et σ∣E\sigma_{\mid E} une bijection de S(E)S(E) sans point fixen d’où une injection de S(X)′S(X)' dans ⋃E⊂P(X)EfiniS(E)\displaystyle \bigcup_{\substack{E \subset P(X) \\ E \text{fini}}} S(E), ce qui donne :
∣S(X)′∣≤∑E⊂P(X)Efini∣S(E)∣=∑n=0∞(∑E∈Pn(X)n!)=κ|S(X)'| \leq \sum_{\substack{E \subset P(X) \\ E \text{fini}}} |S(E)| = \sum_{n=0}^{\infty} \left( \sum_{E \in P_{n}(X)} n! \right) = \kappaFinalement S(X)′S(X)' est de cardinal κ\kappa, et est bien sûr un sous-groupe de S(X)S(X). On aurait pu considérer le groupe libre Z(X)\mathbb{Z}^{(X)}.

Solution by pietro.pende

Discussions0 useful votes

First of all, if EE is empty, then EE cannot be endowed with a group structure for lack of a neutral element.
Consider E≠∅E \neq \varnothing and let n∈N∗n \in \mathbb{N}^{*} denote the cardinality of EE. Since the group (Z/nZ,+)(\mathbb{Z}/n\mathbb{Z},+) has the same cardinality as EE, there exists a bijection from EE to Z/nZ\mathbb{Z}/n\mathbb{Z}, which we shall denote by φ\varphi.
We then consider the map τ:E×E⟶E\tau : E \times E \longrightarrow E defined as follows:
∀(x,y)∈E2,τ(x,y)=φ−1(φ(x)+φ(y))  .\forall(x,y) \in E^2, \quad \tau(x,y) = \varphi^{-1} \left( \varphi(x) + \varphi(y)\right) \;.We shall then verify that (E,τ)(E,\tau) is a group.
Let 0‾\overline{0} denote the neutral element of (Z/nZ,+)(\Z/n\Z,+), and set
e:=φ−1(0‾)∈E.e:=\varphi^{-1}(\overline{0})\in E.

Let us first prove associativity. Let (x,y,z)∈E3(x,y,z)\in E^3. By definition of τ\tau, we have
φ(τ(x,y))=φ(x)+φ(y).\varphi(\tau(x,y))=\varphi(x)+\varphi(y). Therefore
φ(τ(τ(x,y),z))=φ(τ(x,y))+φ(z)=(φ(x)+φ(y))+φ(z)=φ(x)+(φ(y)+φ(z))=φ(x)+φ(τ(y,z))=φ(τ(x,τ(y,z))).\begin{aligned} \varphi\bigl(\tau(\tau(x,y),z)\bigr) &=\varphi(\tau(x,y))+\varphi(z)\\ &=(\varphi(x)+\varphi(y))+\varphi(z)\\ &=\varphi(x)+(\varphi(y)+\varphi(z))\\ &=\varphi(x)+\varphi(\tau(y,z))\\ &=\varphi\bigl(\tau(x,\tau(y,z))\bigr). \end{aligned} As φ\varphi is injective, it follows that
τ(τ(x,y),z)=τ(x,τ(y,z)).\tau(\tau(x,y),z)=\tau(x,\tau(y,z)). Thus τ\tau is associative.

Let us verify the existence of a neutral element. For every x∈Ex\in E,
φ(τ(e,x))=φ(e)+φ(x)=0‾+φ(x)=φ(x),\varphi(\tau(e,x)) = \varphi(e)+\varphi(x) = \overline{0}+\varphi(x) = \varphi(x), hence, by injectivity of φ\varphi,
τ(e,x)=x.\tau(e,x)=x. Similarly,
φ(τ(x,e))=φ(x)+φ(e)=φ(x)+0‾=φ(x),\varphi(\tau(x,e)) = \varphi(x)+\varphi(e) = \varphi(x)+\overline{0} = \varphi(x), whence
τ(x,e)=x.\tau(x,e)=x. Thus ee is a two-sided neutral element for τ\tau.

Finally, let us verify the existence of an inverse for every element. Let x∈Ex\in E. In the group (Z/nZ,+)(\Z/n\Z,+), the element φ(x)\varphi(x) admits the additive inverse −φ(x)-\varphi(x). Set
y:=φ−1(−φ(x))∈E.y:=\varphi^{-1}(-\varphi(x))\in E. Then
φ(τ(x,y))=φ(x)+φ(y)=φ(x)−φ(x)=0‾=φ(e),\varphi(\tau(x,y)) = \varphi(x)+\varphi(y) = \varphi(x)-\varphi(x) = \overline{0} = \varphi(e), so, by injectivity of φ\varphi,
τ(x,y)=e.\tau(x,y)=e. Similarly,
φ(τ(y,x))=φ(y)+φ(x)=−φ(x)+φ(x)=0‾=φ(e),\varphi(\tau(y,x)) = \varphi(y)+\varphi(x) = -\varphi(x)+\varphi(x) = \overline{0} = \varphi(e), whence
τ(y,x)=e.\tau(y,x)=e. Thus every element of EE admits a two-sided inverse for τ\tau.

The three axioms are therefore satisfied; consequently, (E,τ)(E,\tau) is a group (and it is even abelian).

Report

For an unclear, ambiguous, or possibly incorrect statement, please use the Discussion tab on the right. Report content that needs moderator intervention, such as dangerous, clearly non-mathematical, or plagiarized content.