Matrices — Types, Operations & Determinants

Matrices are rectangular arrays of numbers arranged in rows and columns. To get moving fast, four things carry most of the weight: the size of a matrix, the common types, which operations are defined, and what the determinant tells you when the matrix is square. A matrix can organize data, but in early linear algebra it also represents a rule that transforms vectors. You mostly need to know how size controls the rules.

The Key Quantities and Their Notation

Size is written as rows by columns. For example,

is a $2 \times 3$ matrix: $2$ rows and $3$ columns. Size is not just a label; it controls what the matrix can do and which operations make sense.

The determinant is a single number attached to a square matrix. For a $2 \times 2$ matrix

it is

Here $a, b, c, d$ are the four entries; the determinant pairs the main-diagonal product $ad$ against the off-diagonal product $bc$. It is not defined for non-square matrices.

Why the Determinant Means Something

The determinant is not just a number you compute; it reports something structural. The beginner-level reading is:

• If $\det(A) \ne 0$, the matrix is invertible.
• If $\det(A) = 0$, the matrix is not invertible.

Geometrically, for a $2 \times 2$ matrix, $|\det(A)|$ is the factor by which areas are scaled, and the sign tells you whether orientation is preserved or reversed. That is the intuition behind $ad - bc$: it measures how the transformation stretches signed area.

The Types You Will Meet

Most introductory problems use a small set of types.

• Row and column matrices: a row matrix has one row (such as $1 \times 3$); a column matrix has one column (such as $3 \times 1$).
• Square matrices: equal rows and columns, such as $2 \times 2$ or $3 \times 3$. Determinants and inverses are defined only for these.
• Diagonal matrices: square, with zeros everywhere except possibly the main diagonal, where the important values concentrate.
• Identity matrix: the matrix version of the number $1$. For $2 \times 2$, $I = \begin{bmatrix} 1 & 0 \\ 0 & 1 \end{bmatrix}$, and multiplying by $I$ leaves a compatible matrix unchanged.
• Zero matrix: all entries $0$; it acts like the additive zero for matrices of the same size.

Which Operations Are Defined

Addition and subtraction work only when the matrices have exactly the same size, done entry by entry. If the sizes differ, the operation is not defined.

Scalar multiplication multiplies every entry by a number:

Matrix multiplication follows a different rule. If $A$ is $m \times n$ and $B$ is $n \times p$, then $AB$ is an $m \times p$ matrix. The inner dimensions must match:

is defined, but

is not defined when $n \ne r$. Order matters too: even when both products exist, $AB$ and $BA$ are usually different.

Transpose swaps rows and columns, turning a $2 \times 3$ matrix into a $3 \times 2$ matrix, which changes how it lines up in multiplication.

Worked Example: Reading a Matrix Through Its Determinant

Take

This is square, so its determinant is defined. Apply $ad - bc$:

Because $\det(A) = 5 \ne 0$, the matrix is invertible. This single example ties the ideas together:

• The matrix is $2 \times 2$, so it is square.
• Square means a determinant is defined.
• A nonzero determinant means an inverse exists.
• As a transformation of the plane, it scales signed area by $5$.

Try It and Check Yourself

Pick a small $2 \times 2$ matrix and answer four questions: what is its size, is it square, what is its determinant, and does it have an inverse? Predict every answer before you reach for a calculator, then use the calculator only to confirm. That turns the tool into a check rather than a substitute for understanding.

Calculation Traps

One frequent slip is trying to add matrices of different sizes; another is multiplying without checking the inner dimensions first. Students also assume $AB = BA$, which is usually false. With determinants, the main errors are applying them to non-square matrices and misremembering the $2 \times 2$ formula as $ad + bc$ instead of $ad - bc$.

Matrices appear wherever relationships between many quantities must be organized at once: systems of equations, linear transformations, computer graphics, data analysis, engineering models, and numerical computing. The details change by field, but the same rules about size, multiplication, and invertibility keep mattering.