転置行列
正方行列\(A\in M_{n}\left( \mathbb{R} \right) \)を任意に選びます。つまり、\begin{equation*}A=\left( a_{ij}\right) =\begin{pmatrix}
a_{11} & a_{12} & \cdots & a_{1n} \\
a_{21} & a_{22} & \cdots & a_{2n} \\
\cdots & \cdots & \cdots & \cdots \\
a_{n1} & a_{n2} & \cdots & a_{nn}\end{pmatrix}\end{equation*}です。行列\(A\)の\(ij\)要素\(a_{ij}\)と\(ji\)要素\(a_{ji}\)を入れ替えることで得られる行列を、\begin{equation*}A^{t}=\left( a_{ji}\right) =\begin{pmatrix}
a_{11} & a_{21} & \cdots & a_{n1} \\
a_{12} & a_{22} & \cdots & a_{n2} \\
\cdots & \cdots & \cdots & \cdots \\
a_{1n} & a_{2n} & \cdots & a_{nn}\end{pmatrix}\end{equation*}で表記し、これを\(A\)の転置(transpose)や転置行列(transposed matrix)などと呼びます。明らかに、\begin{equation*}A^{t}\in M_{n}\left( \mathbb{R} \right)
\end{equation*}です。
a_{11} & a_{12} \\
a_{21} & a_{22}\end{pmatrix}\end{equation*}の転置行列は、\begin{equation*}
A^{t}=\begin{pmatrix}
a_{11} & a_{21} \\
a_{12} & a_{22}\end{pmatrix}\end{equation*}となります。具体例を挙げると、\begin{eqnarray*}
\begin{pmatrix}
4 & -5 \\
-1 & -2\end{pmatrix}^{t} &=&\begin{pmatrix}
4 & -1 \\
-5 & -2\end{pmatrix}
\\
\begin{pmatrix}
1 & 0 \\
0 & 1\end{pmatrix}^{t} &=&\begin{pmatrix}
1 & 0 \\
0 & 1\end{pmatrix}\end{eqnarray*}などが成り立ちます。
a_{11} & a_{12} & a_{13} \\
a_{21} & a_{22} & a_{23} \\
a_{31} & a_{32} & a_{33}\end{pmatrix}\end{equation*}の転置行列は、\begin{equation*}
A^{t}=\begin{pmatrix}
a_{11} & a_{21} & a_{31} \\
a_{12} & a_{22} & a_{32} \\
a_{13} & a_{23} & a_{33}\end{pmatrix}\end{equation*}となります。具体例を挙げると、\begin{eqnarray*}
\begin{pmatrix}
2 & 3 & 4 \\
5 & 6 & 7 \\
8 & 9 & 1\end{pmatrix}^{t} &=&\begin{pmatrix}
2 & 5 & 8 \\
3 & 6 & 9 \\
4 & 7 & 1\end{pmatrix}
\\
\begin{pmatrix}
1 & 0 & 0 \\
0 & 1 & 0 \\
0 & 0 & 1\end{pmatrix}^{t} &=&\begin{pmatrix}
1 & 0 & 0 \\
0 & 1 & 0 \\
0 & 0 & 1\end{pmatrix}\end{eqnarray*}などが成り立ちます。
転置行列の行列式
正方行列\(A\in M_{n}\left( \mathbb{R} \right) \)を任意に選んだ上で、その転置行列\(A^{t}\in M_{n}\left( \mathbb{R} \right) \)をとります。このとき、両者の行列式の値の間には、\begin{equation*}\left\vert A^{t}\right\vert =\left\vert A\right\vert
\end{equation*}という関係が成り立ちます。つまり、正方行列とその転置の行列式の値は一致するということです。
\end{equation*}という関係が成り立つ。
a_{11} & a_{12} \\
a_{21} & a_{22}\end{pmatrix}\end{equation*}の転置行列は、\begin{equation*}
A^{t}=\begin{pmatrix}
a_{11} & a_{21} \\
a_{12} & a_{22}\end{pmatrix}\end{equation*}です。サラスの公式を用いてこれらの行列式の値を求めると、\begin{eqnarray*}
\left\vert A\right\vert &=&\begin{vmatrix}
a_{11} & a_{12} \\
a_{21} & a_{22}\end{vmatrix}=a_{11}a_{22}-a_{12}a_{21} \\
\left\vert A^{t}\right\vert &=&\begin{vmatrix}
a_{11} & a_{21} \\
a_{12} & a_{22}\end{vmatrix}=a_{11}a_{22}-a_{21}a_{12}
\end{eqnarray*}となるため、\begin{equation*}
\left\vert A\right\vert =\left\vert A^{t}\right\vert
\end{equation*}が成立しています。この結果は先の命題の主張と整合的です。
a_{11} & a_{12} & a_{13} \\
a_{21} & a_{22} & a_{23} \\
a_{31} & a_{32} & a_{33}\end{pmatrix}\end{equation*}の転置行列は、\begin{equation*}
A^{t}=\begin{pmatrix}
a_{11} & a_{21} & a_{31} \\
a_{12} & a_{22} & a_{32} \\
a_{13} & a_{23} & a_{33}\end{pmatrix}\end{equation*}です。サラスの公式を用いてこれらの行列式の値を求めると、\begin{eqnarray*}
\left\vert A\right\vert &=&\begin{vmatrix}
a_{11} & a_{12} & a_{13} \\
a_{21} & a_{22} & a_{23} \\
a_{31} & a_{32} & a_{33}\end{vmatrix}=a_{11}a_{22}a_{33}+a_{12}a_{23}a_{31}+a_{13}a_{21}a_{32}-a_{11}a_{23}a_{32}-a_{12}a_{21}a_{33}-a_{13}a_{22}a_{31}
\\
\left\vert A^{t}\right\vert &=&\begin{vmatrix}
a_{11} & a_{21} & a_{31} \\
a_{12} & a_{22} & a_{32} \\
a_{13} & a_{23} & a_{33}\end{vmatrix}=a_{11}a_{22}a_{33}+a_{21}a_{32}a_{13}+a_{31}a_{12}a_{23}-a_{11}a_{32}a_{23}-a_{21}a_{12}a_{33}-a_{31}a_{22}a_{13}
\end{eqnarray*}となるため、\begin{equation*}
\left\vert A\right\vert =\left\vert A^{t}\right\vert
\end{equation*}が成立しています。この結果は先の命題の主張と整合的です。
後ほど示すように、一般に、行列式の2つの行を入れ替えた場合、その前後において行列式の値は符号だけが変化します。つまり、\begin{equation}
\begin{vmatrix}
a_{11} & a_{12} & \cdots & a_{1n} \\
\cdots & \cdots & \cdots & \cdots \\
a_{i1} & a_{i2} & \cdots & a_{in} \\
\cdots & \cdots & \cdots & \cdots \\
a_{j1} & a_{j2} & \cdots & a_{jn} \\
\cdots & \cdots & \cdots & \cdots \\
a_{n1} & a_{n2} & \cdots & a_{nn}\end{vmatrix}=-\begin{vmatrix}
a_{11} & a_{12} & \cdots & a_{1n} \\
\cdots & \cdots & \cdots & \cdots \\
a_{j1} & a_{j2} & \cdots & a_{jn} \\
\cdots & \cdots & \cdots & \cdots \\
a_{i1} & a_{i2} & \cdots & a_{in} \\
\cdots & \cdots & \cdots & \cdots \\
a_{n1} & a_{n2} & \cdots & a_{nn}\end{vmatrix}
\quad \cdots (1)
\end{equation}という関係が成り立つということです。先の命題より、正方行列とその転置行列の行列式の値は一致するため、\(\left( 1\right) \)の左辺の行列式に関して、\begin{equation}\begin{vmatrix}
a_{11} & \cdots & a_{i1} & \cdots & a_{j1} & \cdots & a_{n1} \\
a_{12} & \cdots & a_{i2} & \cdots & a_{j2} & \cdots & a_{n2} \\
\cdots & \cdots & \cdots & \cdots & \cdots & \cdots & \cdots \\
a_{1n} & \cdots & a_{in} & \cdots & a_{jn} & \cdots & a_{nn}\end{vmatrix}=\begin{vmatrix}
a_{11} & a_{12} & \cdots & a_{1n} \\
\cdots & \cdots & \cdots & \cdots \\
a_{i1} & a_{i2} & \cdots & a_{in} \\
\cdots & \cdots & \cdots & \cdots \\
a_{j1} & a_{j2} & \cdots & a_{jn} \\
\cdots & \cdots & \cdots & \cdots \\
a_{n1} & a_{n2} & \cdots & a_{nn}\end{vmatrix}
\quad \cdots (2)
\end{equation}が成り立ち、\(\left( 1\right) \)の右辺の行列式に関して、\begin{equation}\begin{vmatrix}
a_{11} & \cdots & a_{j1} & \cdots & a_{i1} & \cdots & a_{n1} \\
a_{12} & \cdots & a_{j2} & \cdots & a_{i2} & \cdots & a_{n2} \\
\cdots & \cdots & \cdots & \cdots & \cdots & \cdots & \cdots \\
a_{1n} & \cdots & a_{jn} & \cdots & a_{in} & \cdots & a_{nn}\end{vmatrix}=\begin{vmatrix}
a_{11} & a_{12} & \cdots & a_{1n} \\
\cdots & \cdots & \cdots & \cdots \\
a_{j1} & a_{j2} & \cdots & a_{jn} \\
\cdots & \cdots & \cdots & \cdots \\
a_{i1} & a_{i2} & \cdots & a_{in} \\
\cdots & \cdots & \cdots & \cdots \\
a_{n1} & a_{n2} & \cdots & a_{nn}\end{vmatrix}
\quad \cdots (3)
\end{equation}が成り立ちます。\(\left(1\right) ,\left( 2\right) ,\left( 3\right) \)より、\begin{equation}\begin{vmatrix}
a_{11} & \cdots & a_{i1} & \cdots & a_{j1} & \cdots & a_{n1} \\
a_{12} & \cdots & a_{i2} & \cdots & a_{j2} & \cdots & a_{n2} \\
\cdots & \cdots & \cdots & \cdots & \cdots & \cdots & \cdots \\
a_{1n} & \cdots & a_{in} & \cdots & a_{jn} & \cdots & a_{nn}\end{vmatrix}=-\begin{vmatrix}
a_{11} & \cdots & a_{j1} & \cdots & a_{i1} & \cdots & a_{n1} \\
a_{12} & \cdots & a_{j2} & \cdots & a_{i2} & \cdots & a_{n2} \\
\cdots & \cdots & \cdots & \cdots & \cdots & \cdots & \cdots \\
a_{1n} & \cdots & a_{jn} & \cdots & a_{in} & \cdots & a_{nn}\end{vmatrix}
\quad \cdots (4)
\end{equation}を得ますが、これは、行列式の2つの列を入れ替えた場合、その前後において行列式の値は符号だけが変化するという主張です。つまり、先の命題を踏まえると、行列式の行に関する関係式\(\left( 1\right) \)から、行列式の列に関する同様の関係式\(\left( 4\right) \)を導くことができるということです。このような関係は一般的に成り立ちます。つまり、行列式の行に関する何らかの関係式が導かれたとき、先の命題より、行列式の列に関する同様の関係式を導くことができます。
行と列の関係を逆にした場合にも同様の議論が成立します。つまり、行列式の列に関する何らかの関係式が導かれたとき、先の命題より、行列式の行に関する同様の関係式を導くことができます。
演習問題
A=\begin{pmatrix}
3 & 4 \\
-1 & 2\end{pmatrix}\end{equation*}の行列式\(\left\vert A\right\vert \)の値を求めてください。また、転置行列\(A^{t}\)を明らかにした上で行列式\(\left\vert A^{t}\right\vert \)の値を求め、それが\(\left\vert A\right\vert \)と一致することを確認してください。
x & x+1 \\
-x+1 & -3x\end{pmatrix}\end{equation*}の行列式\(\left\vert A\right\vert \)の値を求めてください。また、転置行列\(A^{t}\)を明らかにした上で行列式\(\left\vert A^{t}\right\vert \)の値を求め、それが\(\left\vert A\right\vert \)と一致することを確認してください。
プレミアム会員専用コンテンツです
【ログイン】【会員登録】