# Mathematical Methods of Physics/Matrices

We have already, in the previous chapter, introduced the concept of matrices as representations for linear transformations. Here, we will deal with them more thoroughly.

## Definition

Let $F$  be a field and let $M=\{1,2,\ldots ,m\}$ ,$N=\{1,2,\ldots ,n\}$ . An n×m matrix is a function $A:N\times M\to F$ .

We denote $A(i,j)=a_{ij}$ . Thus, the matrix $A$  can be written as the array of numbers $A={\begin{pmatrix}a_{11}&a_{12}&a_{13}&\ldots &a_{1m}\\a_{21}&a_{22}&a_{23}&\ldots &a_{2m}\\a_{31}&a_{32}&a_{33}&\ldots &a_{3m}\\\vdots &\vdots &\vdots &\ddots &\vdots \\a_{n1}&a_{n2}&a_{n3}&\ldots &a_{nm}\\\end{pmatrix}}$

Consider the set of all n×m matrices defined on a field $F$ . Let us define scalar product $cA$  to be the matrix $B$  whose elements are given by $b_{ij}=ca_{ij}$ . Also let addition of two matrices $A+B$  be the matrix $C$  whose elements are given by $c_{ij}=a_{ij}+b_{ij}$

With these definitions, we can see that the set of all n×m matrices on $F$  form a vector space over $F$

## Linear Transformations

Let $U,V$  be vector spaces over the field $F$ . Consider the set of all linear transformations $T:U\to V$ .

Define addition of transformations as $(T_{1}+T_{2})\mathbf {u} =T_{1}\mathbf {u} +T_{2}\mathbf {u}$  and scalar product as $(cT)\mathbf {u} =c(T\mathbf {u} )$ . Thus, the set of all linear transformations from $U$  to $V$  is a vector space. This space is denoted as $L(U,V)$ .

Observe that $L(U,V)$  is an $mn$  dimensional vector space

## Operations on Matrices

### Determinant

The determinant of a matrix is defined iteratively (a determinant can be defined only if the matrix is square).

If $A$  is a matrix, its determinant is denoted as $|A|$

We define, $\left|{\begin{pmatrix}a_{11}&a_{12}\\a_{21}&a_{22}\\\end{pmatrix}}\right|=a_{11}a_{22}-a_{21}a_{12}$

For $n=3$ , we define $\left|{\begin{pmatrix}a_{11}&a_{12}&a_{13}\\a_{21}&a_{22}&a_{23}\\a_{31}&a_{32}&a_{33}\\\end{pmatrix}}\right|=a_{11}\left|{\begin{pmatrix}a_{22}&a_{23}\\a_{32}&a_{33}\\\end{pmatrix}}\right|-a_{12}\left|{\begin{pmatrix}a_{21}&a_{23}\\a_{31}&a_{33}\\\end{pmatrix}}\right|+a_{13}\left|{\begin{pmatrix}a_{21}&a_{22}\\a_{31}&a_{32}\\\end{pmatrix}}\right|$

We thus define the determinant for any square matrix

### Trace

Let $A$  be an n×n (square) matrix with elements $a_{ij}$

The trace of $A$  is defined as the sum of its diagonal elements, that is,

$tr(A)=\sum _{i=1}^{n}a_{ii}$

This is conventionally denoted as $tr(A)=\sum _{i,j=1}^{n}a_{ij}\delta _{ij}$ , where $\delta _{ij}$ , called the Kronecker delta is a symbol which you will encounter constantly in this book. It is defined as

$\delta _{ij}=\left\{{\begin{matrix}1,&{\mbox{if }}i=j\\0,&{\mbox{if }}i\neq j\end{matrix}}\right.$

The Kronecker delta itself denotes the members of an n×n matrix called the n×n unit matrix, denoted as $I$

### Transpose

Let $A$  be an m×n matrix, with elements $a_{ij}$ . The n×m matrix $A^{T}$  with elements $a_{ij}^{T}$  is called the transpose of $A$  when $a_{ij}^{T}=a_{ji}$

## Matrix Product

Let $A$  be an m×n matrix and let $B$  be an n×p matrix.

We define the product of $A,B$  to be the m×p matrix $C$  whose elements are given by

$c_{ij}=\sum _{k=1}^{n}a_{ik}b_{kj}$  and we write $C=AB$

### Properties

(i) Product of matrices is not commutative. Indeed, for two matrices $A,B$ , the product $BA$  need not be well-defined even though $AB$  can be defined as above.
(ii) For any matrix n×n $A$  we have $AI=IA=A$ , where $I$  is the n×n unit matrix.