二項演算と閉包性
mathabstract-algebraundergraduatelecture
抽象代数で最初に確認すべき操作は、二つの元から一つの元を作る操作である。これを二項演算という。
集合 S 上の二項演算とは、写像
\ast:S\times S\to S
のことである。つまり、任意の a,b\in S に対して、結果 a\ast b が一意に定まり、しかも再び S の元になる。
ここでいう写像は、入力に対して出力を一つ定める規則である。S\times S は S の元を二つ並べた組の集合を表す。
この「結果が同じ集合の中に戻る」という条件が閉包性である。したがって、\ast:S\times S\to S と書いた時点で閉包性は二項演算の定義に含まれている。実際の判定では、まず足し算や引き算のような、より広い範囲で意味を持つ規則を考え、それを S の元どうしに使った結果がいつも S に残るかを調べる。
binary operation and closure閉包性へいほうせい
The first operation to check in abstract algebra抽象代数ちゅうしょうだいすう is an operation that takes two elements and produces one element. This is called a binary operation二項演算にこうえんざん.
A binary operation二項演算にこうえんざん on a set S is a map
\ast:S\times S\to S
That is, for any a,b\in S, the result a\ast b is uniquely determined and is again an element of S.
Here a map means a rule that assigns one output to each input. The set S\times S means the set of ordered pairs of elements of S.
The condition that "the result returns to the same set" is closure閉包性へいほうせい. Thus closure is already built into the definition once we write \ast:S\times S\to S. In practice, we often begin with a rule, such as addition or subtraction, that is meaningful on a larger domain and then test whether applying it to elements of S always produces an element of S.
1何故なぜ閉包性へいほうせいが必要ひつようか
閉包性へいほうせいがないと、演算えんざんを S の内部ないぶでいつでも繰くり返かえせない。
たとえば自然数全体しぜんすうぜんたい \mathbb N で引ひき算ざんを考かんがえると、
2-5=-3
は自然数しぜんすうではない。したがって引ひき算ざんは \mathbb N 上じょうの二項演算にこうえんざんではない。
一方いっぽう、整数全体せいすうぜんたい \mathbb Z で引ひき算ざんを考かんがえると、任意にんいの整数せいすう a,b に対たいして a-b は整数せいすうである。したがって引ひき算ざんは \mathbb Z 上じょうの二項演算にこうえんざんである。
1Why closure閉包性へいほうせい is necessary
Without closure閉包性へいほうせい, an operation cannot be repeated inside the same set.
For example, consider subtraction on the natural numbers \mathbb N:
2-5=-3
The result is not a natural number. Therefore subtraction is not a binary operation二項演算にこうえんざん on \mathbb N.
On the other hand, if subtraction is considered on the integers \mathbb Z, then for any integers a,b, the result a-b is an integer. Thus subtraction is a binary operation二項演算にこうえんざん on \mathbb Z.
2結合法則けつごうほうそくと可換性かかんせい
結合法則けつごうほうそくassociative lawは、任意にんいの a,b,c\in S に対たいして、三みっつ以い上うえの元げんを演算えんざんするときに括弧かっこの位置いちが結果けっかに影響えいきょうしないことを表あらわす。
(a\ast b)\ast c=a\ast(b\ast c)
結合法則けつごうほうそくがあると、a\ast b\ast c と書かいても意味いみが曖昧あいまいにならない。
可換性かかんせいcommutativityは、任意にんいの a,b\in S に対たいして、順序じゅんじょを入いれ替かえても結果けっかが変かわらないことを表あらわす。
a\ast b=b\ast a
可換性かかんせいは常つねに仮定かていされるわけではない。行列積ぎょうれつせきや置換ちかんの合成ごうせいは一般いっぱんに可換かかんでない。
data/lecture/math/linear-algebra/meaning-of-matrix-multiplication.lecture.n.md
行列ぎょうれつと置換ちかんは非可換ひかかんな演算えんざんの例れいとして挙あげているだけで、このページの判定はんていには使つかわない。
2Associative law and commutativity
The associative law結合法則けつごうほうそく says that for every a,b,c\in S, the position of parentheses does not affect the result when three or more elements are combined.
(a\ast b)\ast c=a\ast(b\ast c)
When the associative law結合法則けつごうほうそく holds, writing a\ast b\ast c is not ambiguous.
Commutativity可換性かかんせい says that for every a,b\in S, changing the order順序じゅんじょ does not change the result.
a\ast b=b\ast a
Commutativity is not always assumed. Matrix multiplication and composition of permutations are generally not commutative.
data/lecture/math/linear-algebra/meaning-of-matrix-multiplication.lecture.n.md
Matrices and permutations are mentioned only as examples of noncommutative operations; they are not used in the tests on this page.
3単位元たんいげんと逆元ぎゃくげん
単位元たんいげんidentity elementとは、すべての a\in S に対たいして演算えんざんしても相手あいてを変かえない e\in S である。
e\ast a=a\ast e=a
e を単位元たんいげんとする。a\in S の逆元ぎゃくげんinverse elementとは、a と左右さゆうどちらの順じゅんで演算えんざんしても単位元たんいげんへ戻もどす a^{-1}\in S である。
a\ast a^{-1}=a^{-1}\ast a=e
両側りょうがわ単位元たんいげんは存在そんざいすれば一意いちいである。実際じっさい、e,f がともに両側りょうがわ単位元たんいげんなら、e=e\ast f=f となる。
単位元たんいげんと逆元ぎゃくげんは、方程式ほうていしきを解とくために重要じゅうようである。a\ast x=b の左ひだりから a^{-1} を演算えんざんし、括弧かっこを付つけ替かえて x=a^{-1}\ast b と導みちびくには、逆元ぎゃくげんに加くわえて結合法則けつごうほうそくと単位元たんいげんを仮定かていする。すると、
a^{-1}\ast(a\ast x)=(a^{-1}\ast a)\ast x=e\ast x=x
だから、a\ast x=b の両辺りょうへんの左ひだりから a^{-1} を演算えんざんして x=a^{-1}\ast b を得える。逆ぎゃくにこの x を代入だいにゅうすれば a\ast(a^{-1}\ast b)=(a\ast a^{-1})\ast b=e\ast b=b なので、これは実際じっさいに解かいである。
さらに、結合法則けつごうほうそくと両側りょうがわ単位元たんいげんがあるとき、ある元げんの両側りょうがわ逆元ぎゃくげんも存在そんざいすれば一意いちいである。u,v がともに a の両側りょうがわ逆元ぎゃくげんなら、
u=u\ast e=u\ast(a\ast v)=(u\ast a)\ast v=e\ast v=v
となる。ここで結合法則けつごうほうそくを使つかったことが重要じゅうようである。
3Identity element and inverse element逆元ぎゃくげん
An identity element単位元たんいげん is an element e\in S that leaves every a\in S unchanged on both sides.
e\ast a=a\ast e=a
Let e be an identity element. An inverse element逆元ぎゃくげん of a\in S is an element a^{-1}\in S that brings a back to the identity in either order.
a\ast a^{-1}=a^{-1}\ast a=e
A two-sided identity, if it exists, is unique. Indeed, if both e and f are two-sided identities, then e=e\ast f=f.
Identity elements and inverses are important for solving equations. To derive x=a^{-1}\ast b from a\ast x=b by operating with a^{-1} on the left and reassociating, assume associativity and an identity in addition to the inverse. Then
a^{-1}\ast(a\ast x)=(a^{-1}\ast a)\ast x=e\ast x=x.
Thus applying a^{-1} on the left of both sides gives x=a^{-1}\ast b. Conversely, substituting this value gives a\ast(a^{-1}\ast b)=(a\ast a^{-1})\ast b=e\ast b=b, so it really is a solution.
Moreover, when associativity and a two-sided identity hold, a two-sided inverse of an element is unique if it exists. If u and v are both two-sided inverses of a, then
u=u\ast e=u\ast(a\ast v)=(u\ast a)\ast v=e\ast v=v.
The use of associativity in this argument is essential.
4具体例ぐたいれい:演算表えんざんひょうで見みる
集合しゅうごう S=\{0,1,2\} に、3 で割わった余あまりの足たし算ざんを入いれる。
a\ast b を、a+b を 3 で割わった余あまりと定さだめる。
このとき演算表えんざんひょうは次つぎのようになる。行ぎょうを左ひだりの元げん、列れつを右みぎの元げんとする。
| \ast | 0 | 1 | 2 |
| 0 | 0 | 1 | 2 |
| 1 | 1 | 2 | 0 |
| 2 | 2 | 0 | 1 |
全すべての結果けっかが S に入はいっているので閉包性へいほうせいがある。0 は単位元たんいげんであり、0 の逆元ぎゃくげんは 0、1 の逆元ぎゃくげんは 2、2 の逆元ぎゃくげんは 1 である。また、(a\ast b)\ast c と a\ast(b\ast c) はどちらも a+b+c を 3 で割わった余あまりになるので等ひとしく、結合法則けつごうほうそくも成立せいりつする。したがって、この例れいは次つぎの講義こうぎで定義ていぎする群ぐんになる。
4Concrete example: reading an operation table
Put addition modulo 3 on the set S=\{0,1,2\}: define a\ast b to be the remainder of a+b upon division by 3.
The operation table is as follows, with the left operand indexing the row and the right operand indexing the column.
| \ast | 0 | 1 | 2 |
| 0 | 0 | 1 | 2 |
| 1 | 1 | 2 | 0 |
| 2 | 2 | 0 | 1 |
Every result lies in S, so closure holds. The element 0 is the identity; the inverse of 0 is 0, the inverse of 1 is 2, and the inverse of 2 is 1. Both (a\ast b)\ast c and a\ast(b\ast c) are the remainder of a+b+c upon division by 3, so they are equal and the operation is associative. Thus this example is a group群ぐん, as defined in the next lecture.
6判定はんてい補足ほそく:演算えんざんの性質せいしつは別々べつべつに確認かくにんする
集合しゅうごう S と規則きそく \ast が与あたえられたとき、まず a,b\in S\Rightarrow a\ast b\in S を確認かくにんして、\ast:S\times S\to S が二項演算にこうえんざんbinary operationとして定さだまっているかを見みる。これは閉包性へいほうせいclosureだけの確認かくにんであり、結合法則けつごうほうそく、可換性かかんせい、単位元たんいげん、逆元ぎゃくげんはまだ保証ほしょうしない。
たとえば \mathbb Z 上じょうの引ひき算ざんsubtraction a\ast b=a-b は閉包性へいほうせいclosureを持もつが、
(5-3)-1=1,\qquad 5-(3-1)=3
なので結合法則けつごうほうそくを満みたさない。したがって、閉包性へいほうせいclosureがあることと群ぐんに近ちかい性質せいしつを持もつことは別問題べつもんだいである。
6Test supplement: check properties of an operation演算えんざん separately
Given a set S and a rule \ast, first check a,b\in S\Rightarrow a\ast b\in S to see whether \ast:S\times S\to S is a binary operation二項演算にこうえんざん. This checks only closure閉包性へいほうせい; it does not yet guarantee associativity, commutativity, an identity element単位元たんいげん, or inverse elements.
For example, subtraction a\ast b=a-b on \mathbb Z has closure, but
(5-3)-1=1,\qquad 5-(3-1)=3
so it does not satisfy associativity. Thus having closure and having group-like properties are separate questions.
7まとめ
二項演算にこうえんざんは、集合しゅうごうの二ふたつの元げんから同おなじ集合しゅうごうの元げんを作つくる写像しゃぞうである。閉包性へいほうせい、結合法則けつごうほうそく、可換性かかんせい、単位元たんいげん、逆元ぎゃくげんは、群ぐん・環かん・体たいを定義ていぎするための基本きほん語彙ごいである。
7Summary
A binary operation二項演算にこうえんざん is a map that takes two elements of a set and produces an element of the same set. Closure閉包性へいほうせい, the associative law結合法則けつごうほうそく, commutativity可換性かかんせい, an identity element単位元たんいげん, and inverse elements逆元ぎゃくげん are the basic vocabulary used to define groups群ぐん, rings環かん, and fields体たい.