Zum Inhalt springen

Adjunkte

aus Wikipedia, der freien Enzyklopädie

Die Adjunkte, klassische Adjungierte (nicht zu verwechseln mit der echten adjungierten Matrix) oder komplementäre Matrix einer Matrix ist ein Begriff aus dem mathematischen Teilgebiet der linearen Algebra. Man bezeichnet damit die Transponierte der Kofaktormatrix, also die Transponierte jener Matrix, deren Einträge die vorzeichenbehafteten Minoren (Unterdeterminanten) sind.

Mit Hilfe der Adjunkten kann man die Inverse einer regulären quadratischen Matrix berechnen.

Definition

Die Adjunkte <math>\operatorname{adj}(A)</math> einer quadratischen Matrix <math>A \in K^{n \times n}</math> mit Einträgen aus einem Körper (oder allgemeiner aus einem kommutativen Ring) <math>K</math> ist definiert als

<math>\operatorname{adj}(A) = \operatorname{Cof}(A)^\mathsf{T} = \tilde A^\mathsf{T} =

\begin{pmatrix} \tilde a_{11} & \tilde a_{12} & \cdots & \tilde a_{1n}\\ \tilde a_{21} & \tilde a_{22} & & \tilde a_{2n}\\ \vdots & & \ddots & \vdots\\ \tilde a_{n1} & \tilde a_{n2} & \cdots & \tilde a_{nn}\end{pmatrix}^\mathsf{T} = \begin{pmatrix} \tilde a_{11} & \tilde a_{21} & \cdots & \tilde a_{n1}\\ \tilde a_{12} & \tilde a_{22} & & \tilde a_{n2}\\ \vdots & & \ddots & \vdots\\ \tilde a_{1n} & \tilde a_{2n} & \cdots & \tilde a_{nn}\end{pmatrix}</math>. Es ist hierbei zu beachten, dass an der Stelle <math>(j,i)</math> der Kofaktor <math>\tilde a_{ij}</math> steht. Die Kofaktoren <math>\tilde a_{ij}</math> berechnen sich zu

<math>\tilde a_{ij} = (-1)^{i+j}\cdot M_{ij} = (-1)^{i+j}\cdot \det \begin{pmatrix}

a_{1,1} & \cdots & a_{1,j-1} & a_{1,j+1} & \cdots & a_{1,n} \\ \vdots & \ddots & \vdots & \vdots & \ddots & \vdots \\ a_{i-1,1} & \cdots & a_{i-1,j-1} & a_{i-1,j+1} & \cdots & a_{i-1,n} \\ a_{i+1,1} & \cdots & a_{i+1,j-1} & a_{i+1,j+1} & \cdots & a_{i+1,n} \\ \vdots & \ddots & \vdots & \vdots & \ddots & \vdots \\ a_{n,1} & \cdots & a_{n,j-1} & a_{n,j+1} & \cdots & a_{n,n}\end{pmatrix}</math>.

Die Minoren <math>M_{ij}</math> sind also die Werte der Unterdeterminanten der Matrix <math>A</math>, die durch Streichen der <math>i</math>-ten Zeile und der <math>j</math>-ten Spalte entstehen.

Da die Adjunkte in heutigen Lehrbüchern selten auftaucht und in älteren Werken die Notation nicht immer eindeutig ist, ist Vorsicht geboten. Oft wird dieselbe Notation für die Adjunkte und die Adjungierte (also bei reellen Matrizen deren Transponierte, bei komplexen Matrizen deren konjugiert-transponierte) verwendet.

Beispiele

(2 × 2)-Matrix

Eine beliebige <math>2 \times 2</math>-Matrix hat die Form

<math>A = \begin{pmatrix} Vorlage:A & Vorlage:Bibel/Link\\ Vorlage:C & Vorlage:D \end{pmatrix}</math>

Die Adjunkte zu dieser Matrix ist

<math>\operatorname{adj} (A) = \begin{pmatrix} \,\,\,Vorlage:D & \!\!Vorlage:-c\\ Vorlage:-b & Vorlage:A \end{pmatrix} ^\mathsf{T} = \begin{pmatrix} \,\,\,Vorlage:D & \!\!Vorlage:-b\\ Vorlage:-c & Vorlage:A \end{pmatrix}</math>

(3 × 3)-Matrix

Eine beliebige <math>3 \times 3</math>-Matrix hat die Form

<math>A = \begin{pmatrix}a & b & c\\ d & e & f \\ g & h & i\end{pmatrix}</math>

Die Adjunkte zu dieser Matrix ist

<math>\begin{align}

\operatorname{adj} (A) & = \begin{pmatrix} \quad\det\begin{pmatrix}e & f\\ h & i\end{pmatrix} & - \det\begin{pmatrix}d & f\\ g & i\end{pmatrix} & \quad\det\begin{pmatrix}d & e\\ g & h\end{pmatrix} \\ - \det\begin{pmatrix}b & c\\ h & i\end{pmatrix} & \quad\det\begin{pmatrix}a & c\\ g & i\end{pmatrix} & - \det\begin{pmatrix}a & b\\ g & h\end{pmatrix} \\ \quad\det\begin{pmatrix}b & c\\ e & f\end{pmatrix} & - \det\begin{pmatrix}a & c\\ d & f\end{pmatrix} & \quad\det\begin{pmatrix}a & b\\ d & e\end{pmatrix} \end{pmatrix}^\mathsf{T}\\[.7em] & = \begin{pmatrix} ei - fh & fg - di & dh - eg \\ ch - bi & ai - cg & bg - ah \\ bf - ce & cd - af & ae - bd \end{pmatrix}^\mathsf{T}\\[.7em] & = \begin{pmatrix} ei - fh & ch - bi & bf - ce \\ fg - di & ai - cg & cd - af \\ dh - eg & bg - ah & ae - bd \end{pmatrix} \end{align}</math>

Eigenschaften

Nachfolgende Beziehungen gelten für alle Matrizen aus <math>K^{n\times n}</math>

<math>\operatorname{adj}(E) = E</math>, wobei <math>E</math> eine Einheitsmatrix ist.
<math>\operatorname{adj}(0) = 0</math> für <math>n>1</math>, wobei 0 die Nullmatrix ist. Für <math>1\times 1</math>-Matrizen <math>A=[a_{11}]</math> gilt jedoch immer, auch für die Nullmatrix: <math>\operatorname{adj}([a_{11}]) = [1]</math>.
<math>\operatorname{adj}(AB) = \operatorname{adj}(B) \cdot \operatorname{adj}(A)</math>
<math>\operatorname{adj}(A^\mathsf{T}) = \operatorname{adj}(A)^\mathsf{T}</math>
<math>A \cdot \operatorname{adj}(A) = \operatorname{adj}(A) \cdot A = \det(A) \cdot E</math>
<math>\operatorname{adj}(\lambda A)=\lambda^{n-1}\operatorname{adj}(A)</math> wobei <math>\lambda\in K</math>
<math>\det(\operatorname{adj}(A))=(\det A)^{n-1}</math>
<math>\operatorname{adj}(\operatorname{adj}(A))=(\det A)^{n-2}A</math>, insbesondere für <math>2\times2</math>-Matrizen gilt <math>\operatorname{adj}(\operatorname{adj}(A))=A</math>

Für invertierbare Matrizen gilt zusätzlich

<math>(\operatorname{adj}(A))^{-1}=\frac1{\det(A)}A=\operatorname{adj}(A^{-1})</math>

Berechnung der Inversen einer Matrix

Die einzelnen Spalten der Inversen einer Matrix <math>A</math> werden jeweils von der Lösung des Gleichungssystems <math>Ax = e_j</math> mit dem <math>j</math>-ten Einheitsvektor auf der rechten Seite gebildet. Berechnet man diese mit der Cramerschen Regel, so erhält man die Formel

<math>A^{-1} = \frac{1}{\det(A)} \,\operatorname{adj} (A).</math>

Hieraus erhält man sofort eine Formel für die Inverse einer regulären <math>2 \times 2</math>-Matrix:

<math>A^{-1} = \frac{1}{\det(A)} \,\operatorname{adj} (A) = \frac{1}{ad - bc} \begin{pmatrix} \,\,\,Vorlage:D & \!\!Vorlage:-b\\ Vorlage:-c & Vorlage:A \end{pmatrix}. </math>

Literatur

  • Siegfried Bosch: Lineare Algebra. 4., überarbeitete Auflage. Springer, Berlin u. a. 2008, ISBN 978-3-540-76437-3.