0
$\begingroup$

Suppose we have a vector space of 2x2 matrices with a basis { E11, E12, E21, E22] with Eij = a 2x2 matrix with a 1 in the i-th row and the j-th column.

In this lineair space, the transformation T : R2x2 -> R2x2 is defined by T(A) = AT

Define the matrix of T in respect to the given basis.

Now I'm struggling with that. T transposes a matrix so do I need to express it in terms of Eij or...? And if not what's the point of having that basis so excplicitely written down?

I also need the eigenvalues and eigenvectors of this transformation but without calculating. So even if I got the matrix, I wouldn't know how to extract eigenvalues unless it would be a lower triangular matrix.

A solution on a similar question was

\begin{bmatrix}1&0&0&0\\0&0&1&0\\0&1&0&0\\0&0&0&1\end{bmatrix}

Yet this is a 4x4 matrix so how could it be applied to 2x2 matrices? And more important. How do I extract the eigenvalues of this by reasoning.

$\endgroup$
  • $\begingroup$ Yes, you need to express $T$ in the basis $E_{i,j}$. $\endgroup$ – xbh Sep 5 '18 at 17:21
  • $\begingroup$ For eigenvalues, try definition, since you clearly knew what happen when you apply $T$ to matrices. $\endgroup$ – xbh Sep 5 '18 at 17:23
  • 1
    $\begingroup$ P.S. We recommend users to use MathJax to write math expressions. Here is the tutorial. $\endgroup$ – xbh Sep 5 '18 at 17:28
1
$\begingroup$

To write down the matrix of a linear transformation with respect to a given basis you have to see how the transformation acts on the base of the domain and write down the result as a linear combination of the basis of the image. Words! Here's a similar question where a gave better insights on how to solve this kind of problems.

Now let's see how does it work in your case: let us evaluate how the map acts on our base, remember that we have $$T: U\equiv \mathbb R^{2\times 2}\longrightarrow \mathbb R^{2\times2}\equiv V$$ and for both $U$ and $V$ we have the same base, the standard base $E_{ij}$. Now $$T\left(E_{11}\right) = E_{11}^T = E_{11} = \overbrace{\color{red}{1}E_{11}+ \color{blue}{0}E_{12}+\color{orange}{0}E_{21}+\color{green}{0}E_{22}}^{\text{Linear combination of base of }V}\\[5pt] T\left(E_{12}\right) = E_{12}^T = E_{21} =\color{red}{0}E_{11}+ \color{blue}{0}E_{12}+\color{orange}{1}E_{21}+\color{green}{0}E_{22}\\[5pt] T\left(E_{21}\right) = E_{21}^T = E_{12}=\color{red}{0}E_{11}+ \color{blue}{1}E_{12}+\color{orange}{0}E_{21}+\color{green}{0}E_{22}\\[5pt] T\left(E_{22}\right) = E_{22}^T = E_{22}=\color{red}{0}E_{11}+ \color{blue}{0}E_{12}+\color{orange}{0}E_{21}+\color{green}{1}E_{22} $$ remember that all that transposition does is to invert the indices $(a_{ij})^T = a_{ji}$. As you can see I wrote down the linear combination in the base of the image for every element mapped with $T$. The matrix is now to be written down as $$\begin{align}[T]=&\left(\begin{matrix}\color{red}{1}&\color{red}{0}&\color{red}{0}&\color{red}{0}\\ \color{blue}{0}& \color{blue}{0}& \color{blue}{1}& \color{blue}{0}\\ \color{orange}{0}&\color{orange}{1}&\color{orange}{0}&\color{orange}{0} \\ \color{green}{0}&\color{green}{0}&\color{green}{0}&\color{green}{1}\end{matrix}\right)\end{align}$$

So this is the matrix of the transformation! You could be disoriented by the fact that, by definition $$T(A) = [T]v$$ how is it possible that, if $v$ is a matrix, we have to do a $4\times4$ matrix times a $2\times 2$ matrix? That is because there's a little problem: $v$ is not a matrix! $v$ is a vector that contains the coefficients of the linear combination in a given base for a given element of $\mathbb R^{2\times 2}$. For example, if a wanted to evaluate $T$ on the identity matrix I should first evaluate the linear combination in the basis for the identity matrix which is $$v_{I} = (1,0,0,1) = 1E_{11}+0E_{12}+0E_{21}+1E_{22} \\[10pt] \implies T(I) = [T]v_{I} = \left(\begin{matrix}1&0&0&0\\0&0&1&0\\0&1&0&0\\0&0&0&1\end{matrix}\right)\left(\begin{matrix}1\\0\\0\\1\end{matrix}\right) = \left(\begin{matrix}1&0\\0&1\end{matrix}\right)$$ as expected.

The eigenvectors to this transformations are clearly the matrices such that $$T(A) = A^T = \lambda A$$ which, as a trivial case, with $\lambda=1$ are symmetric matrices. Note that $T^2=I$ because $(A^T)^T = A$ so if $$T(A) = A^T = \lambda A$$ and $$T(A^T) = T(\lambda A) = \lambda T(A) = \lambda^2 A = I $$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.