행렬

行列. Matrix.

  • 행렬(行列)은 여럿이 줄지어가는 모습을 뜻하는 말이기도 하다.

정의[편집 | 원본 편집]

수나 식을 직사각형 모양으로 배열한 것. 가로줄을 행(low), 세로줄을 열(column)이라고 한다.

좀 더 수학적으로 정의하면 다음과 같다.

[math]\displaystyle{ R }[/math]이고 모든 [math]\displaystyle{ i=1, 2, \cdots, m }[/math][math]\displaystyle{ j=1, 2, \cdots, n }[/math]에 대하여 [math]\displaystyle{ a_{ij}\in R }[/math]일 때, [math]\displaystyle{ A=\begin{bmatrix} a_{11}&a_{12}&\cdots&a_{1n}\\ a_{21}&a_{22}&\cdots&a_{2n}\\ \vdots&\vdots&\ddots&\vdots\\ a_{m1}&a_{m2}&\cdots&a_{mn} \end{bmatrix} }[/math][math]\displaystyle{ R }[/math] 위의 [math]\displaystyle{ \left(m\times n\right) }[/math] 행렬이라고 부른다.

[math]\displaystyle{ \left(m\times n\right) }[/math] 행렬 전체의 집합은 [math]\displaystyle{ M_{m, n}\left(R\right) }[/math]으로 나타낸다. [math]\displaystyle{ \mathcal M_{m, n}\left(R\right) }[/math][math]\displaystyle{ \mathfrak M_{m, n}\left(R\right) }[/math][math]\displaystyle{ \operatorname{Mat}_{m,n}\left( R \right) }[/math]처럼 나타내기도 한다.

  • 만일 [math]\displaystyle{ m=n }[/math]이면 [math]\displaystyle{ A }[/math][math]\displaystyle{ n }[/math]차 정사각행렬([math]\displaystyle{ n }[/math]‐th square matrix)이라고 한다. [math]\displaystyle{ n }[/math]차 정사각행렬 전체의 집합은 [math]\displaystyle{ M_{n, n}\left(R\right) }[/math] 또는 더 간편하게 [math]\displaystyle{ M_n\left(R\right) }[/math]로 나타낸다.
    • 이때 n을 차수(order)라고 한다.
    • 원소 [math]\displaystyle{ a_{11}, a_{22}, \cdots a_{nn} }[/math]을 포함하는 대각선을 주대각선(principal diagonal)이라고 한다.
  • 만일 [math]\displaystyle{ m=1 }[/math]이면 행벡터(row vector)라고 한다.
  • 만일 [math]\displaystyle{ n=1 }[/math]이면 열벡터(column vector)라고 한다.

행렬의 원소를 강조할 때는 [math]\displaystyle{ A=\left[ a_{ij} \right] = \left[ a_{ij} \right] _{m \times n}=\left( A \right) _{ij} }[/math]로 표기할 수 있다.

구성 성분[편집 | 원본 편집]

[math]\displaystyle{ A=\left(a_{ij}\right) }[/math]에서 [math]\displaystyle{ a_{ij} }[/math]를 행렬 [math]\displaystyle{ A }[/math][math]\displaystyle{ \left(i, j\right) }[/math] 성분([math]\displaystyle{ \left(i, j\right) }[/math]‐th component) 혹은 [math]\displaystyle{ \left(i, j\right) }[/math] 원소([math]\displaystyle{ \left(i, j\right) }[/math]‐th entry)라고 한다.

위 행렬의 [math]\displaystyle{ \left(i, 1\right) }[/math] 성분부터 [math]\displaystyle{ \left(i, n\right) }[/math] 성분까지, 즉 다음 행렬

[math]\displaystyle{ \begin{bmatrix} a_{i1}&a_{i2}&\cdots&a_{in} \end{bmatrix} }[/math]

을 행렬의 [math]\displaystyle{ i }[/math]번째 행벡터라고 하고([math]\displaystyle{ i=1, 2, \cdots, m }[/math]), [math]\displaystyle{ [A]_i }[/math]로 적기도 한다.

또 행렬의 [math]\displaystyle{ \left(1, j\right) }[/math] 성분부터 [math]\displaystyle{ \left(m, j\right) }[/math] 성분까지, 즉 다음 행렬

[math]\displaystyle{ \begin{bmatrix} a_{1j}\\ a_{2j}\\ \vdots\\ a_{mj} \end{bmatrix} }[/math]

을 행렬의 [math]\displaystyle{ j }[/math]번째 열벡터라고 하고([math]\displaystyle{ j=1, 2, \cdots, n }[/math]), [math]\displaystyle{ [A]^j }[/math]로 적기도 한다.

위 행벡터, 열벡터를 이용해서 행렬을 다음과 같이 나타내기도 한다.

[math]\displaystyle{ A =\begin{bmatrix} - &[A]_1& -\\ - &[A]_2& -\\ \vdots&\vdots&\vdots\\ - &[A]_m& - \end{bmatrix} =\begin{bmatrix} {|} & {|} & {\cdots} & {|} \\ {[A]^{1}} & {[A]^{2}} & {\cdots} & {[A]^{n}} \\ {|} & {|} & {\cdots} & {|} \end{bmatrix} }[/math]

행렬의 성질[편집 | 원본 편집]

연산[편집 | 원본 편집]

덧셈과 뺄셈[편집 | 원본 편집]

행렬의 덧셈과 뺄셈은 행렬의 크기가 같을 때만 정의된다. 행렬의 덧셈·뺄셈 얘기가 나올 때 ‘크기가 맞을 때’라고 하면 이 조건을 의미하는 것이다.

행렬 [math]\displaystyle{ A, B }[/math][math]\displaystyle{ \left(m\times n\right) }[/math]행렬일 때, 행렬의 덧셈과 뺄셈을 다음과 같이 성분별로(componentwisely) 정의한다. 즉,

  • (덧셈) [math]\displaystyle{ A+B=\left(a_{ij}+b_{ij}\right) }[/math]
  • (뺄셈) [math]\displaystyle{ A-B=\left(a_{ij}-b_{ij}\right) }[/math]

예를 들어, [math]\displaystyle{ R=\mathbb{R} }[/math]이고 [math]\displaystyle{ A,B\in M_2(R) }[/math]에 대해 [math]\displaystyle{ A=\begin{bmatrix} 9 & -12\\ -5 & 3 \end{bmatrix},B=\begin{bmatrix} -2 & 7\\ 0 & 1 \end{bmatrix} }[/math]이면

[math]\displaystyle{ \begin{align} A+B&=\begin{bmatrix} 9 & -12\\ -5 & 3 \end{bmatrix}+\begin{bmatrix} -2 & 7\\ 0 & 1 \end{bmatrix}\\ &=\begin{bmatrix} 9+(-2) & -12+7\\ -5+0 & 3+1 \end{bmatrix}\\ &=\begin{bmatrix} 7 & -5\\ -5 & 4 \end{bmatrix} \end{align} }[/math]

이며, [math]\displaystyle{ R=\mathbb{Z}_3 }[/math]이고 [math]\displaystyle{ A,B\in M_2(R) }[/math]에 대해 [math]\displaystyle{ A=\begin{bmatrix} 1 & 0\\ 2 & 1 \end{bmatrix},B=\begin{bmatrix} 2 & 2\\ 0 & 1 \end{bmatrix} }[/math]이면

[math]\displaystyle{ \begin{align} A-B&=\begin{bmatrix} 1 & 0\\ 2 & 1 \end{bmatrix}-\begin{bmatrix} 2 & 2\\ 0 & 1 \end{bmatrix}\\ &=\begin{bmatrix} 1-2 & 0-2\\ 2-0 & 1-1 \end{bmatrix}\\ &=\begin{bmatrix} -1 & -2\\ 2 & 0 \end{bmatrix}\\ &=\begin{bmatrix} 2 & 1\\ 2 & 0 \end{bmatrix} \end{align} }[/math]

[math]\displaystyle{ R }[/math]의 덧셈의 결합법칙과 교환법칙이 성립하기 때문에, 행렬의 덧셈에 대해 결합법칙과 교환법칙이 성립함은 자명하다.

[math]\displaystyle{ R }[/math]상수곱[편집 | 원본 편집]

행렬의 집합 위에서는 [math]\displaystyle{ R }[/math]상수곱이라는 작용(action)이 정의된다.

행렬 [math]\displaystyle{ A, B }[/math][math]\displaystyle{ \left(m\times n\right) }[/math]행렬이고 [math]\displaystyle{ r\in R }[/math]일 때, 행렬의 [math]\displaystyle{ R }[/math]상수곱을 다음과 같이 성분별로(componentwisely) 정의한다. 즉,

  • ([math]\displaystyle{ R }[/math]상수곱) [math]\displaystyle{ rA=\left(ra_{ij}\right) }[/math]

행렬의 덧셈과 뺄셈에 대한 위 상수곱의 분배법칙이 성립함은 자명하다(환 [math]\displaystyle{ R }[/math]의 덧셈에 대한 환 [math]\displaystyle{ R }[/math]의 곱셈의 교환법칙이 성립하기 때문). 또, 위 상수곱은 환 [math]\displaystyle{ R }[/math]의 곱셈과도 compatible하다(환 [math]\displaystyle{ R }[/math]의 결합법칙 때문).

몇 가지만 더 확인하면 앞서 정의한 덧셈과 [math]\displaystyle{ R }[/math]상수곱에 대하여 [math]\displaystyle{ \left(m\times n\right) }[/math]행렬 전체의 집합 [math]\displaystyle{ M_{\left(m, n\right)}\left(R\right) }[/math][math]\displaystyle{ R }[/math]가군이 되고, 특별히 [math]\displaystyle{ R }[/math]이면 [math]\displaystyle{ R }[/math]벡터공간이 됨을 금방 확인할 수 있다.

곱셈[편집 | 원본 편집]

행렬의 곱셈은 앞 행렬의 행의 개수와 뒤 행렬의 열의 개수가 같을 때만 정의된다. 행렬의 곱셈 얘기가 나올 때 ‘크기가 맞을 때’라고 하면 이 조건을 의미하는 것이다.

행렬 [math]\displaystyle{ A }[/math]의 행 개수와 행렬 [math]\displaystyle{ B }[/math]의 열 개수가 같을 때, 두 행렬의 곱셈은 [math]\displaystyle{ A }[/math]의 각 행벡터와 [math]\displaystyle{ B }[/math]의 각 열벡터의 내적(inner product)을 원소로 가지는 행렬로 정의된다.

즉, 행렬 [math]\displaystyle{ A }[/math][math]\displaystyle{ \left(m\times n\right) }[/math]행렬이고 [math]\displaystyle{ B }[/math][math]\displaystyle{ \left(n\times r\right) }[/math]행렬일 때, 두 행렬의 곱 [math]\displaystyle{ C=AB }[/math][math]\displaystyle{ \left(c_{ij}\right)=\left(\left[A\right] _i\cdot\left[B\right] ^j\right) }[/math][math]\displaystyle{ \left(m\times r\right) }[/math]행렬이다.

행렬의 곱셈을 이렇게 일견 부자연스럽게 정의하는 데는 다 이유가 있다. 이는 선형대수학기본 정리, 즉 행렬과 선형사상은 같은 것이라는 명제를 배우고 나면 금방 이해할 수 있다.

가령 행렬[math]\displaystyle{ A,B }[/math]를 각 각 함수[math]\displaystyle{ L_1(\mathbf{x})=A\mathbf{x}, L_2(\mathbf{x})=B\mathbf{x} }[/math]로서 생각했을 때 우리가 위의 곱셈을 따른다면 두 함수의 합성은 [math]\displaystyle{ L_1 \circ L_2(\mathbf{x})=AB\mathbf{x} }[/math] 로서 나타나게 된다. 즉 함수의 합성을 곱의 계산으로 할 있게 되는 것이다!

행렬의 곱셈은 결합법칙은 만족하나 교환법칙은 성립하지 않는다. 이는 함수의 합성이 그렇기 때문이다. 한편 덧셈과 뺄셈에 대한 곱셈의 분배법칙은 성립한다.

앞서 정의한 덧셈과 [math]\displaystyle{ R }[/math]상수곱 그리고 지금 정의한 곱셈에 대하여 [math]\displaystyle{ n }[/math]차 정사각행렬 전체의 집합 [math]\displaystyle{ M_n\left(R\right) }[/math][math]\displaystyle{ R }[/math]대수(결합대수)가 됨을 보일 수 있다.

특수한 행렬[편집 | 원본 편집]

아래 내용은 모든 행렬에서 정의된다.

  • 영행렬(zero matrix) [math]\displaystyle{ O }[/math]
    모든 원소가 0인 행렬로, 행렬의 덧셈의 항등원이다.
  • 전치행렬(transpose of a matrix) [math]\displaystyle{ A^T }[/math]
    원래 행렬의 행과 열을 뒤바꾼 행렬이다. 원래 행렬을 [math]\displaystyle{ A }[/math]에 대해 전치행렬을 [math]\displaystyle{ A^T }[/math]와 같이 표기한다. 자세한 내용은 전치행렬을 참조.

아래 내용은 정사각행렬에서만 정의된다.

  • 단위행렬(identity matrix) [math]\displaystyle{ I }[/math]
    주대각선(행 번호와 열 번호가 같은 원소들)이 1이고 나머지는 0인 행렬이다. 이는 정사각행렬의 곱셈의 항등원이 됨을 확인할 수 있다. 고등학교에서는 [math]\displaystyle{ E }[/math]로도 많이 표기한다.
  • 역행렬(inverse matrix) [math]\displaystyle{ A^{-1} }[/math]
    어떤 행렬의 곱셈에 대한 역원을 말한다. 자세한 내용은 역행렬을 참조.
    역행렬은 아무 때나 존재하는 것이 아니며, 역행렬이 존재하는 행렬을 가역행렬(invertible matrix)이라 한다. 단위행렬은 가역이고, 단위행렬의 역행렬은 자기 자신이다. 선형대수학에서 어떤 행렬이 가역인 것과 동치인 조건을 엄청 많이 배울 수 있다.
  • 멱영행렬(nilpotent matrix)
    여러 번 거듭제곱하면 영행렬이 되는 행렬이다. 어떤 행렬이 멱영행렬이면 행렬의 차수가 [math]\displaystyle{ n }[/math]일 때 [math]\displaystyle{ n }[/math]번 이하로 거듭제곱하면 영행렬이 됨을 증명할 수 있다. 단위행렬은 멱영행렬이 아니고, 따라서 멱영행렬은 당연히 가역이 아니다.
  • 대각행렬(diagonal matrix)
    주대각선 외의 원소가 0인 행렬이다. 주대각선의 원소가 [math]\displaystyle{ a_1,a_2,\cdots a_n }[/math]일 때, 대각행렬을 [math]\displaystyle{ \operatorname{diag}\left(a_1,a_2,\cdots a_n\right)= \begin{pmatrix} a_1 & 0 & \cdots & 0 \\ 0 & a_2 & \cdots & 0 \\ \vdots & \vdots & \ddots & \vdots \\ 0 & 0 & \cdots & a_n \end{pmatrix} }[/math]로 표기한다. 자세한 내용은 대각행렬을 참조.
  • 대칭행렬(symmetric matrix)
    [math]\displaystyle{ A = A^T }[/math]인 행렬이다. 모든 대각행렬은 당연히 대칭행렬이다. 모든 대칭행렬은 직교기저로써 대각화할 수 있다.
  • 직교행렬(orthogonal matrix)
    [math]\displaystyle{ A A^T = A^T A = I }[/math]인 행렬이다. 직교행렬은 점곱을 보존하고, 나아가 길이와 거리도 보존한다.

아래 내용은 복소수체 [math]\displaystyle{ \mathbb{C} }[/math] 위의 행렬에 대해서만 정의된다.

  • 켤레전치행렬(conjugate transpose matrix) [math]\displaystyle{ A^\ast }[/math]
  • 정규행렬(normal matrix)
    [math]\displaystyle{ A A^\ast = A^\ast A }[/math]인 정사각행렬이다. 모든 정규행렬은 직교기저로써 대각화할 수 있다.
  • 에르미트행렬(Hermitian matrix)
    [math]\displaystyle{ A = A^\ast }[/math]인 정사각행렬이다. 모든 에르미트행렬은 정규행렬이다.
  • 유니터리행렬(unitary matrix)
    [math]\displaystyle{ A A^\ast = A^\ast A = I }[/math]인 정사각행렬이다. 유니터리행렬은 안곱을 보존하고, 나아가 길이와 거리도 보존한다. 모든 유니터리행렬은 정규행렬이다.

각주