next up previous contents index
Nächste Seite: Rang einer Matrix Aufwärts: Matrizen Vorherige Seite: Lineare Abbildungen und Koordinatenvektoren   Inhalt   Index

Matrizen und Basiswechsel

Statt zur Beschreibung von linearen Abbildungen kann man Matrizen auch zur Beschreibung eines Basiswechsels benutzen.
Seien $ {\mathcal B}= (v_1, \ldots, v_n)$ und $ {\mathcal B}' = (v'_1, \ldots, v'_n)$ zwei Basen des Vektorraums $ V$ . Sei $ v \in V$ . Dann ist $ v = \sum_{j=1}^{n} c_j v_j = \sum_{j=1}^{n} c'_j v'_j$ . Wie berechnen sich die Koordinaten $ c'_1, \ldots, c'_n$ bezüglich $ {\mathcal B}'$ aus denen bezüglich $ {\mathcal B}$ ?

Wir drücken zunächst die Elemente von $ {\mathcal B}'$ als Linearkombinationen bezüglich der Basis $ {\mathcal B}$ aus:

$\displaystyle v'_j = \sum_{i = 1}^{n} s_{ij} v_i.$

Die auftretenden Koeffizienten fassen wir in der Matrix

$\displaystyle S_{{\mathcal B}, {\mathcal B}'} = (s_{ij})_{{i=1, \ldots, n} \atop {j = 1, \ldots, n}}$

zusammen.

Der $ j$ -te Spaltenvektor $ s_j^{\downarrow} = \left( \begin{array}{ccc}
s_{1j} \\ \vdots \\ s_{nj}
\end{array} \right)$ von $ S_{{\mathcal B}, {\mathcal B}'}$ ist also der Koordinatenvektor von $ v'_j$ bezüglich der Basis $ {\mathcal B}$ .

Genauso drücken wir $ v_k$ als Linearkombination bezüglich der Basis $ {\mathcal B}'$ aus:

$\displaystyle v_k = \sum_{l=1}^{n} t_{lk} v'_l,$

und definieren analog $ S_{{\mathcal B}', {\mathcal B}} = (t_{lk})_{{l = 1, \ldots, n} \atop {k =1, \ldots, n}}$ . Dann erhalten wir für alle $ k = 1, \ldots, n$ :

$\displaystyle v_k = \sum_{j=1}^{n} t_{jk} v'_j = \sum_{j=1}^{n} t_{jk} (\sum_{i=1}^{n} s_{ij} v_i) = \sum_{i=1}^{n}
(\sum_{j=1}^{n} s_{ij} t_{jk}) v_i.$

Aus dieser Gleichung folgt $ \sum_{j=1}^{n} s_{ij} t_{jk} = \delta_{ik}$ für alle $ i = 1, \ldots, n$ (wobei $ \delta_{ik}$ das Kronecker-Symbol bezeichnet; vgl. Beispiel 1, Seite [*]), denn die Koordinatendarstellung von $ v_k$ bezüglich der Basis $ {\mathcal B}= (v_1, \ldots, v_n)$ ist nun einmal

$\displaystyle v_k = 0 \cdot v_1 + 0 \cdot v_2 + \cdots + 0 \cdot v_{k-1} + 1 \cdot v_k + 0 \cdot v_{k+1} + \cdots +
0 \cdot v_n.$

Das bedeutet aber $ S_{{\mathcal B}, {\mathcal B}'} \cdot S_{{\mathcal B}', {\mathcal B}} = E_n$ und analog gilt $ S_{B', B} \cdot S_{{\mathcal B}, {\mathcal B}'} = E_n$ . $ S_{{\mathcal B}, {\mathcal B}'}$ ist also eine invertierbare Matrix und $ S_{{\mathcal B}', {\mathcal B}} = S_{{\mathcal B}, {\mathcal B}'}^{-1}$ . Damit haben wir gezeigt:





Satz 10.27   Sind $ {\mathcal B}$ und $ {\mathcal B}'$ Basen des Vektorraums $ V$ , so ist die Matrix $ S_{{\mathcal B}, {\mathcal B}'}$ , die als Spalten die Koordinatenvektoren der Basiselemente aus $ {\mathcal B}'$ bezüglich der Basis $ {\mathcal B}$ besitzt, invertierbar und es gilt $ S_{{\mathcal B}, {\mathcal B}'}^{-1} = S_{{\mathcal B}', {\mathcal B}}$ . $ S_{{\mathcal B}, {\mathcal B}'}$ heißt Basiswechselmatrix.



Wir können nun die eingangs gestellte Frage beantworten. Es ist $ v = \sum_{j=1}^{n} c_j v_j = \sum_{j = 1}^{n} c_j \sum_{l=1}^{n} t_{lj} v'_l = \sum_{l=1}^{n}
(\sum_{j=1}^{n} t_{lj} c_j) v'_l$ .

Ist also $ x = (c_1, \ldots, c_n)^t$ der Koordinatenvektor von $ v$ bezüglich $ {\mathcal B}$ , so ist $ x' = S_{{\mathcal B}', {\mathcal B}} x$ der Koordinatenvektor von $ v$ bezüglich $ {\mathcal B}'$ . Daher gilt auch $ x = S_{{\mathcal B}', {\mathcal B}}^{-1} x' = S_{{\mathcal B}, {\mathcal B}'} x'$ .





Satz 10.28   (Basiswechsel) Um die Koordinaten von $ v$ bezüglich $ {\mathcal B}'$ aus denen bezüglich $ {\mathcal B}$ zu berechnen, multipliziert man den Koordinatenvektor bezüglich $ {\mathcal B}$ von links mit derjenigen Matrix, die als Spalten die Koordinaten der Basisvektoren aus $ {\mathcal B}$ bezüglich $ {\mathcal B}'$ hat:

$\displaystyle x' = S_{{\mathcal B}', {\mathcal B}} x \ .$



Beispiel: Sei $ V = \mathbb{R}^2, \, {\mathcal B}\, = \, (e_1, e_2), \, {\mathcal B}' \, = \, ...
...t{2}} (e_1 + e_2), \frac{\displaystyle 1}{\displaystyle \sqrt{2}}
(- e_1 + e_2)$ . $ {\mathcal B}'$ entsteht aus $ {\mathcal B}$ durch Drehung um 45 Grad.
Dann ist also $ S_{{\mathcal B}, {\mathcal B}'} \, = \, \frac{\displaystyle 1}{\displaystyle \...
...ystyle \sqrt{2}} \left( \begin{array}{ccc}
1 & 1\\
-1 & 1
\end{array} \right) $ . Machen Sie bitte die Probe!
Sei $ v = \, \left (\begin{array}{c}1\\ 2\end{array} \right )$ . Dies ist dann auch der Koordinatenvektor bezüglich $ {\mathcal B}$ . Dann hat $ v$ bezüglich der Basis $ {\mathcal B}'$ den Koordinatenvektor $ x' \, = \, \frac{\displaystyle 1}{\displaystyle \sqrt{2}} \left( \begin{array}...
...yle 1}{\displaystyle \sqrt{2}} \left (\begin{array}{c}3\\ 1\end{array} \right )$ , das heißt

$\displaystyle v \, = \, \frac{\displaystyle 1}{\displaystyle \sqrt{2}} \cdot 3 ...
...} \cdot 1 \cdot (\frac{\displaystyle 1}{\displaystyle \sqrt{2}} (- e_1 + e_2)).$

Aufgabe: Berechnen Sie bitte die Matrizen der folgenden Basiswechsel im $ \mathbb{R}^3$ . Dabei ist
$ {\mathcal B}\, = \, (e_1, e_2, e_3)$ , $ {\mathcal B}' \, = \, (\left( \begin{array}{ccc}
\cos(\varphi) \\ \sin(\varphi...
...\ 0
\end{array} \right), \left (\begin{array}{c}0\\ 0\\ 1 \end{array} \right ))$ , $ {\mathcal B}'' = (e_1, e_3, e_2)$ .

Nun können wir leicht berechnen, wie sich die Darstellungsmatrizen einer linearen Abbildung bei Basiswechsel verhalten: Sei $ \alpha : V \rightarrow W$ eine lineare Abbildung, $ {\mathcal B}, {\mathcal B}'$ seien Basen in $ V$ mit den Basiswechselmatrizen $ S_{{\mathcal B}, {\mathcal B}'}, \, S_{{\mathcal B}', {\mathcal B}}$ wie oben. $ {\mathcal C}, {\mathcal C}'$ seien Basen in $ W$ . $ S_{{\mathcal C}, {\mathcal C}'}$ und $ S_{{\mathcal C}', {\mathcal C}}$ seien entsprechend gebildet.

Aufgabe: Beantworten Sie bitte die Frage: Was enthalten $ S_{{\mathcal B}, {\mathcal B}'}$ bzw. $ S_{{\mathcal B}', {\mathcal B}}$ als Spalten? Was enthalten $ S_{{\mathcal C}, {\mathcal C}'}$ und $ S_{{\mathcal C}', {\mathcal C}}$ als Spalten? Sei $ A_{\alpha}^{{\mathcal B}, {\mathcal C}}$ die Darstellungsmatrix von $ \alpha$ bezüglich der Basen $ {\mathcal B}$ und $ {\mathcal C}$ . Was sind die Spalten von $ A^{{\mathcal B}, {\mathcal C}}_{\alpha}$ ?





Satz 10.29   Die Darstellungsmatrix $ A^{{\mathcal B}', {\mathcal C}'}_\alpha$ von $ \alpha$ berechnet sich aus $ A^{{\mathcal B}, {\mathcal C}}_\alpha$ durch

$\displaystyle A^{{\mathcal B}',
{\mathcal C}'}_\alpha = S^{-1}_{{\mathcal C}, {...
...l C}} A^{{\mathcal B}, {\mathcal C}}_{\alpha} S_{{\mathcal B}, {\mathcal B}'}.
$



Beweis: Sei $ v \in V$ beliebig. Sei $ x$ bzw. $ x'$ der Koordinatenvektor von $ v$ bezüglich $ {\mathcal B}$ bzw. $ {\mathcal B}'$ . Analog sei $ y$ bzw. $ y'$ der Koordinatenvektor von $ \alpha (v)$ bezüglich $ {\mathcal C}$ bzw. $ {\mathcal C}'$ . Nach Satz 10.25 ist also $ y \, = \, A_{\alpha}^{{\mathcal B}, {\mathcal C}} x$ und $ y' \, = \, A_{\alpha}^{{\mathcal B}', {\mathcal C}'} x'$ und nach Satz 10.28 ist $ x \, = \, S_{{\mathcal B}, {\mathcal B}'} x'$ und $ y' \, = \, S_{{\mathcal C}', {\mathcal C}} y$ . Daraus folgt: $ y' \, = \, S_{{\mathcal C}', {\mathcal C}} y \, = \, S_{{\mathcal C}', {\mathc...
... C}} A_{\alpha}^{{\mathcal B}, {\mathcal C}}
S_{{\mathcal B}, {\mathcal B}'} x'$ , d. h. $ A_{\alpha}^{{\mathcal B}', {\mathcal C}'} x' \, = \, S_{{\mathcal C}', {\mathcal C}} A_{\alpha}^{{\mathcal B}, {\mathcal C}} S_{{\mathcal B}, {\mathcal B}'} x'$ .
Durchläuft $ v$ ganz $ V$ , so durchläuft $ x'$ ganz $ K^n$ . Nach Korollar 10.26 b) ist daher $ A_{\alpha}^{{\mathcal B}', {\mathcal C}'} = S_{{\mathcal C}', {\mathcal C}} A_{\alpha}^{{\mathcal B}, {\mathcal C}} S_{{\mathcal B}, {\mathcal B}'}$ . Es ist $ S_{{\mathcal C}', {\mathcal C}} \, = \,
S_{{\mathcal C}, {\mathcal C}'}^{-1}$ nach Satz 10.27.$ \Box$

Beispiel: Sei $ V = W = \mathbb{R}^2, \, {\mathcal B}\, = \, {\mathcal C}\, = \, (e_1, e_2), \...
...mathcal C}' = (\frac{1}{\sqrt{2}} (e_1 + e_2),
\frac{1}{\sqrt{2}} (-e_1 + e_2))$ . Sei $ \alpha : \mathbb{R}^2 \rightarrow \mathbb{R}^2$ definiert durch $ \alpha(e_1) = e_2$ und $ \alpha(e_2) = e_1$ . Geometrisch gesprochen ist $ \alpha$ die Spiegelung an der Hauptdiagonalen.
Es ist $ A_{\alpha}^{{\mathcal B}}\, ( = A_{\alpha}^{{\mathcal B}, {\mathcal B}}) = \left( \begin{array}{cc}
0 & 1 \\
1 & 0
\end{array} \right)$ . Wir berechnen $ A_{\alpha}^{{\mathcal B}'} (= A_{\alpha}^{{\mathcal B}', {\mathcal B}'})$ mit Hilfe von Satz 10.29. Es ist $ S_{{\mathcal B}, {\mathcal B}'} \, = \,
\frac{1}{\sqrt{2}} \left( \begin{array}{rr}
1 & -1 \\
1 & 1
\end{array} \right)$ und $ S_{{\mathcal B}', {\mathcal B}} = S_{{\mathcal B}, {\mathcal B}'}^{-1} = \frac{1}{\sqrt{2}} \left( \begin{array}{cc}
1 & 1\\
-1 & 1
\end{array} \right)$ . Also ist $ A_{\alpha}^{{\mathcal B}'} = S_{{\mathcal B}', {\mathcal B}} A_{\alpha}^{{\mat...
...} \right) \, = \, \left( \begin{array}{rr}
1 & 0 \\
0 & -1
\end{array} \right)$ . An dieser Darstellungsmatrix wird sofort klar, dass $ \alpha$ eine Spiegelung ist.


next up previous contents index
Nächste Seite: Rang einer Matrix Aufwärts: Matrizen Vorherige Seite: Lineare Abbildungen und Koordinatenvektoren   Inhalt   Index
friedric 2004-08-18