Throughout the following, we base everything on the matrix $\mathbf{A} =
\begin{bmatrix}
a_{11} & a_{12} & \cdots & a_{1n} \\
a_{21} & a_{22} & \cdots & a_{2n} \\
\vdots & \vdots & \ddots & \vdots \\
a_{m1} & a_{m2} & \cdots & a_{mn}
\end{bmatrix}$and explicitly denote the dimensions. If the definition is valid for an arbitrary [[Field]], we denote it by $F$, otherwise we explicitly use e.g. $\mathbb{R}$ or $\mathbb{C}$.
>[!note]
>- For derivatives involving matrices, see **[[Matrix Calculus]]**.
>- For the [[Group|group]] structure resulting from invertible matrices, see **[[General Linear Group and Matrix Groups]]**.
>[!brainwaves] In pure Mathematics
>Matrices represent [[Linear Maps and Operators]]
---
- [[Determinant of a Square Matrix]]
- [[Kernel or Nullspace of a linear Map]]
- [[Rank of a Matrix]]
- [[Trace of a Square Matrix]]
- [[Definiteness]]
---
#### Special Maps of a Matrix $\mathbf{A}$
| **Special Matrices** | **Definition** | **Properties** | **Remarks** |
| -------------------- | -------------------------------------------- | ---------------------------------------------------------------------- | ------------------------------------------------ |
| **Transpose** | $[\mathbf{A}^{T}]_{ij}=[\mathbf{A}]_{ij}$ | $(\mathbf{A+B})^{T}=\mathbf{A}^{T}+\mathbf{B}^{T}$ | |
| | | $(\mathbf{AB})^{T}=\mathbf{B}^{T}\mathbf{A}^{T}$ | |
| | | $(\mathbf{A}^{T})^{-1}=(\mathbf{A}^{-1})^{T}$ | |
| | | $\text{det}(\mathbf{A}^{T})=\text{det}(\mathbf{A})$ | |
| **Inverse** | $\mathbf{A}\cdot \mathbf{A}^{-1}=\mathbf{I}$ | $(\mathbf{A \cdot B})^{-1}=\mathbf{B}^{-1}\cdot \mathbf{A}^{-1}$ | Only defined for square matrices |
| | | $(c\mathbf{A})^{-1}=c^{-1}\mathbf{A}^{-1}$ | |
| | | $\det(\mathbf{A}^{-1})=(\det(\mathbf{A}))^{-1}$ | |
| **Projector** | $\mathbf{A}\cdot \mathbf{A}=\mathbf{A}$ | othogonal, if $\langle \mathbf{x}-\mathbf{Ax}, \mathbf{Ax} \rangle =0$ | Orthogonal, if projector matrix symmetric |
| | | | Only defined for square matrices |
---
#### Special Matrix Structures
- **Symmetric** $\mathbf{A}^{T}=\mathbf{A}$
- Sum and difference of symmetric matrices is symmetric, product not
- If $\mathbf{A}$ symmetric and $n \in \mathbb{N}$, $\mathbf{A}^{n}$ is symmetric
- If $\mathbf{A}^{-1}$ exists, it is symmetric if and only if $\mathbf{A}$ is symmetric
- **Skew-Symmetric** $\mathbf{A}^\top=-\mathbf{A}$
- Sum and scalar multiples of skew symmetric matrices are skew symmetric
- trace, determinant $0$
-
- **Hermitian** $\mathbf{A}=\mathbf{A}^{H}=\overline{\mathbf{A}^{T}}$
- Complex conjugate of every element
- $\langle \mathbf{w}, \mathbf{Av}\rangle=\langle \mathbf{v}, \mathbf{Aw}\rangle$ for any pair of vector
- For $\mathbf{v}\in \mathbb{C}^{n}$ it holds that $\langle \mathbf{v}, \mathbf{Av}\rangle\ \in \mathbb{R}$
- $(\mathbf{AB})^{H}=\mathbf{B}^{H}\mathbf{A}^{H}$
- **Unitary** $\mathbf{U}^{H}\mathbf{U}=\mathbf{I}$
- $U^{H}=U^{-1}$
- Eigenvectors form orthogonal basis of $\mathbb{C}^{n}$
- $|\det(\mathbf{U})|=1$
- Product of unitary matrices is unitary
- Preserve angles between vectors and their lengths
- **Orthogonal** $\mathbf{A}^{T}\mathbf{A}=\mathbf{A}\mathbf{A}^{T}=\mathbf{I}$
- $\mathbf{A}^{T}=\mathbf{A}^{-1}$
- Determinant is either $1$ or $-1$
- Real analog to unitary
- **Upper Hessenberg**
- Upper triangular matrix with additional element below diagonal