What Are The Types Of Matrices?
The types of matrices are the named categories a matrix falls into based on its shape, its entries, or how it behaves under an operation. A matrix is a rectangular array of numbers arranged in rows and columns, and its order is written as (rows × columns). Naming a matrix's type is a shortcut, because the name tells you at a glance what the matrix is allowed to do.
Mathematicians sort matrices by asking four questions:
By shape: how many rows and columns does it have? (row, column, square, rectangular)
By entries: where do the zeros and the equal values sit? (diagonal, scalar, identity, zero, triangular)
By transpose: what happens when its rows and columns are swapped? (symmetric, skew-symmetric)
By multiplication and inverse: how does the matrix act on others? (singular, non-singular, orthogonal, idempotent, nilpotent, involutory)
Every type below is one answer to one of those four questions. Some matrices answer several at once, and that overlap is where the mix-ups start. To classify any matrix, all you need are its elements and its order.
What Are The Types Of Matrices Based On Shape?
Shape is the first thing you notice about a matrix, and it splits the whole family into four groups by comparing the number of rows with the number of columns.
Row Matrix
A row matrix has exactly one row and any number of columns, so its order is 1 × n. It lies flat, like a single line of numbers. (This one has no dedicated Bhanzu page yet, so it appears here as plain text.)
$$\begin{bmatrix} 3 & -1 & 5 & 0 \end{bmatrix}$$
Column Matrix
A column matrix has exactly one column and any number of rows, so its order is m × 1. It stands upright, a single stack of numbers.
$$\begin{bmatrix} 2 \ 7 \ 4 \end{bmatrix}$$
Square Matrix
A square matrix has the same number of rows and columns, order n × n. Squareness matters, because determinants, inverses, and most of the special types further down this page only exist for square matrices.
$$\begin{bmatrix} 1 & 2 \ 3 & 4 \end{bmatrix}$$
Rectangular Matrix
A rectangular matrix has a different number of rows and columns, order m × n with m ≠ n. Every matrix that is not square is rectangular.
$$\begin{bmatrix} 1 & 2 & 3 \ 4 & 5 & 6 \end{bmatrix}$$
What Are The Types Of Matrices Based On Their Entries?
The next group looks inside the grid at where the zeros sit and which entries repeat. This is where the three most-confused types live, so watch the pattern of the main diagonal, the line of entries running from the top-left corner to the bottom-right.
Diagonal Matrix
A diagonal matrix is a square matrix whose entries off the main diagonal are all zero. The diagonal entries themselves can be anything, including zero.
$$\begin{bmatrix} 5 & 0 \ 0 & 3 \end{bmatrix}$$
Scalar Matrix
A scalar matrix is a diagonal matrix whose diagonal entries are all the same number. It is the tighter case: diagonal, plus every diagonal entry equal. (No dedicated Bhanzu page yet, so it is named here as plain text.)
$$\begin{bmatrix} 7 & 0 \ 0 & 7 \end{bmatrix}$$
Identity Or Unit Matrix
An identity matrix, also called a unit matrix, is a scalar matrix whose repeated value is exactly 1. Written $I$, it is the matrix version of the number one: multiplying by it changes nothing, so $AI = IA = A$.
$$I = \begin{bmatrix} 1 & 0 \ 0 & 1 \end{bmatrix}$$
Zero Or Null Matrix
A zero matrix, also called a null matrix and written $O$, has every entry equal to zero. It can be any order, square or rectangular, and it acts like the number zero: $A + O = A$. (No dedicated Bhanzu page yet.)
$$O = \begin{bmatrix} 0 & 0 \ 0 & 0 \end{bmatrix}$$
Upper And Lower Triangular Matrix
A triangular matrix is a square matrix with zeros on one side of the main diagonal. An upper triangular matrix has all zeros below the diagonal; a lower triangular matrix has all zeros above it.
$$\text{Upper: } \begin{bmatrix} 2 & 4 & 1 \ 0 & 3 & 5 \ 0 & 0 & 6 \end{bmatrix} \qquad \text{Lower: } \begin{bmatrix} 2 & 0 & 0 \ 4 & 3 & 0 \ 1 & 5 & 6 \end{bmatrix}$$
Which Types Of Matrices Come From The Transpose?
The next two types are defined by what happens when its transpose is taken, that is, when the rows and columns are swapped and $A^{T}$ is compared with $A$.
Symmetric Matrix
A symmetric matrix equals its own transpose, $A^{T} = A$. The grid is a mirror image across the main diagonal, so the entry in row $i$, column $j$ matches the entry in row $j$, column $i$.
$$\begin{bmatrix} 1 & 2 & 3 \ 2 & 5 & 4 \ 3 & 4 & 6 \end{bmatrix}$$
Skew-Symmetric Matrix
A skew-symmetric matrix equals the negative of its transpose, $A^{T} = -A$. Flipping across the diagonal flips every sign, which forces every diagonal entry to be zero, since a number can only equal its own negative if it is zero.
$$\begin{bmatrix} 0 & 2 & -3 \ -2 & 0 & 4 \ 3 & -4 & 0 \end{bmatrix}$$
Which Types Of Matrices Come From Multiplication And Inverses?
The last group is defined by how a square matrix behaves when it is multiplied, or whether it can be undone at all. The gatekeeper here is the determinant, a single number computed from a square matrix that decides whether an inverse exists.
Singular And Non-Singular Matrix
A singular matrix is a square matrix whose determinant is zero, and it has no inverse. A non-singular matrix has a non-zero determinant, so it can be inverted; for that reason it is also called an invertible matrix.
$$\text{Singular: } \begin{bmatrix} 2 & 4 \ 1 & 2 \end{bmatrix} ; (\det = 0) \qquad \text{Non-singular: } \begin{bmatrix} 1 & 2 \ 3 & 4 \end{bmatrix} ; (\det = -2)$$
Orthogonal Matrix
An orthogonal matrix is a square matrix whose transpose is also its inverse, so $A^{T}A = I$. These matrices rotate and reflect without stretching, which is why graphics and robotics rely on them.
$$\begin{bmatrix} 0 & 1 \ -1 & 0 \end{bmatrix}$$
Idempotent Matrix
An idempotent matrix stays the same when multiplied by itself, $A^{2} = A$. Projection matrices, which flatten points onto a line or plane, are the classic example. (No dedicated Bhanzu page yet.)
$$\begin{bmatrix} 1 & 0 \ 0 & 0 \end{bmatrix}$$
Nilpotent Matrix
A nilpotent matrix becomes the zero matrix when raised to some power, $A^{k} = O$. The smallest such $k$ is its index. (No dedicated Bhanzu page yet, so it is named here as plain text.)
$$\begin{bmatrix} 0 & 1 \ 0 & 0 \end{bmatrix} \quad \text{since} \quad \begin{bmatrix} 0 & 1 \ 0 & 0 \end{bmatrix}^{2} = \begin{bmatrix} 0 & 0 \ 0 & 0 \end{bmatrix}$$
Involutory Matrix
An involutory matrix is its own inverse, $A^{2} = I$. Applying it twice returns you to the start, which is exactly how a reflection behaves. (No dedicated Bhanzu page yet.)
$$\begin{bmatrix} 1 & 0 \ 0 & -1 \end{bmatrix}$$
Matrices with complex entries add a few more names, such as the Hermitian matrix, which equals its own conjugate transpose. For matrices of real numbers, the types above cover almost everything you will meet in school and early college.
Types Of Matrices: The Full Summary Table
Here is the whole taxonomy on one screen. Read the condition column first, then check the example against it.
Table: Every common type of matrix, its defining condition, and a worked example.
Type | Condition | Example |
|---|---|---|
Row matrix | One row (order 1 × n) | $\begin{bmatrix} 3 & -1 & 5 \end{bmatrix}$ |
Column matrix | One column (order m × 1) | $\begin{bmatrix} 2 \ 7 \end{bmatrix}$ |
Square matrix | Rows = columns (n × n) | $\begin{bmatrix} 1 & 2 \ 3 & 4 \end{bmatrix}$ |
Rectangular matrix | Rows ≠ columns (m × n) | $\begin{bmatrix} 1 & 2 & 3 \ 4 & 5 & 6 \end{bmatrix}$ |
Diagonal matrix | Square; off-diagonal entries all 0 | $\begin{bmatrix} 5 & 0 \ 0 & 3 \end{bmatrix}$ |
Scalar matrix | Diagonal; diagonal entries all equal | $\begin{bmatrix} 7 & 0 \ 0 & 7 \end{bmatrix}$ |
Identity (unit) matrix | Scalar; diagonal entries all 1 | $\begin{bmatrix} 1 & 0 \ 0 & 1 \end{bmatrix}$ |
Zero (null) matrix | Every entry 0 | $\begin{bmatrix} 0 & 0 \ 0 & 0 \end{bmatrix}$ |
Upper triangular | Square; zeros below the diagonal | $\begin{bmatrix} 2 & 4 \ 0 & 3 \end{bmatrix}$ |
Lower triangular | Square; zeros above the diagonal | $\begin{bmatrix} 2 & 0 \ 4 & 3 \end{bmatrix}$ |
Symmetric matrix | $A^{T} = A$ | $\begin{bmatrix} 1 & 2 \ 2 & 5 \end{bmatrix}$ |
Skew-symmetric matrix | $A^{T} = -A$; zero diagonal | $\begin{bmatrix} 0 & 2 \ -2 & 0 \end{bmatrix}$ |
Singular matrix | Square; $\det = 0$ | $\begin{bmatrix} 2 & 4 \ 1 & 2 \end{bmatrix}$ |
Non-singular matrix | Square; $\det \neq 0$ | $\begin{bmatrix} 1 & 2 \ 3 & 4 \end{bmatrix}$ |
Orthogonal matrix | $A^{T}A = I$ | $\begin{bmatrix} 0 & 1 \ -1 & 0 \end{bmatrix}$ |
Idempotent matrix | $A^{2} = A$ | $\begin{bmatrix} 1 & 0 \ 0 & 0 \end{bmatrix}$ |
Nilpotent matrix | $A^{k} = O$ for some $k$ | $\begin{bmatrix} 0 & 1 \ 0 & 0 \end{bmatrix}$ |
Involutory matrix | $A^{2} = I$ | $\begin{bmatrix} 1 & 0 \ 0 & -1 \end{bmatrix}$ |
Why Do We Sort Matrices Into Types?
Classifying matrices is not busywork. A type is a promise about structure, and that promise tells you what is allowed and saves you work before a single calculation begins.
It decides what operations are even legal. Only square matrices have a determinant or an inverse, and only same-order matrices can be added. The type screens out impossible steps in advance.
It shortcuts the arithmetic. The determinant of a triangular matrix is just the product of its diagonal. The inverse of a diagonal matrix is found by flipping each diagonal entry. Spot the type, skip the long method.
It names the building blocks. The zero matrix and the identity matrix play the roles that 0 and 1 play for ordinary numbers, which is what lets matrix algebra behave like arithmetic at all.
So the type is the label on the box. Once you can read it, half of every matrix problem is answered before you start, because you already know which tools fit and which do not.
Who Discovered And Named The Types Of Matrices?
Grids of numbers are ancient, but the idea of treating them as objects with their own algebra, and giving the special ones names, is surprisingly recent and came largely from two friends in Victorian England.
Two more names shaped how we talk about matrix types today:
James Joseph Sylvester (1814–1897, England) coined the word "matrix" in 1850, from the Latin for "womb," and worked closely with his friend Arthur Cayley.
Olga Taussky-Todd (1906–1995, Austria and the United States) helped turn scattered facts about matrices into a unified matrix theory in the twentieth century, and once described herself as a torchbearer for the whole field.
Where Are The Types Of Matrices Used In The Real World?
Each type earns its name because it does a specific job somewhere outside the classroom.
Computer graphics and games: orthogonal matrices rotate and reflect characters and cameras without distorting them, and the identity matrix is the "do nothing" starting point every transformation builds on.
Data science and machine learning: symmetric matrices show up as covariance matrices, and diagonal matrices are the goal of diagonalization, the step behind Principal Component Analysis that finds the directions data varies most.
Networks and social graphs: the adjacency matrix of an undirected network is symmetric, because a mutual connection points both ways.
Engineering and physics: the stiffness matrix of a structure and the inertia tensor of a rotating body are symmetric by physical law, and triangular matrices power the LU decomposition that solves large systems of equations fast.
Cryptography: invertible (non-singular) matrices encode messages that only the matching inverse can decode.
One taxonomy of grids quietly runs graphics engines, learning machines, social networks, and bridges. Naming the type is how each field grabs exactly the tool it needs.
What Are The Most Common Types Of Matrices Mistakes?
These four mix-ups account for most lost marks on matrix classification. Each is a case of reading the pattern too quickly.
Confusing diagonal, scalar, and identity matrices.
Where it slips in:
A student calls any matrix with zeros off the diagonal an identity, or labels every diagonal matrix a scalar.
Don't do this:
Do not treat the three names as interchangeable. They are three different levels of strictness.
The correct way:
Read the three as nested levels of strictness: diagonal (off-diagonal entries all zero) then scalar (diagonal entries all equal) then identity (that equal value fixed at 1). Every identity is scalar and every scalar is diagonal, but not the reverse.
Mixing up row and column matrices.
Where it slips in:
A student judges by how many entries there are rather than by the shape, and writes a row where a column was needed.
Don't do this:
Do not go by the count of numbers. Go by the order.
The correct way:
A row matrix has exactly one row and lies flat, order 1 × n. A column matrix has exactly one column and stands upright, order m × 1. The order settles it every time.
Confusing symmetric with skew-symmetric.
Where it slips in:
A student checks $A^{T} = A$ for both, or forgets that a skew-symmetric matrix must have a zero diagonal.
Don't do this:
Do not ignore the sign or the diagonal.
The correct way:
Symmetric means $A^{T} = A$, a clean mirror across the diagonal. Skew-symmetric means $A^{T} = -A$, which flips every sign and forces each diagonal entry to be zero.
Assuming every square matrix has an inverse.
Where it slips in:
A student sees a square matrix and starts inverting it without checking anything first.
Don't do this:
Do not assume square means invertible.
The correct way:
Compute the determinant first. If it is zero the matrix is singular and has no inverse; if it is non-zero the matrix is non-singular and can be inverted.
Practice Problems On Types Of Matrices
Name every type each matrix belongs to. Answers follow each problem.
Classify $\begin{bmatrix} 0 & 0 & 0 \end{bmatrix}$.
(Answer: row matrix and zero matrix, order 1 × 3.)Is $\begin{bmatrix} 4 & 0 \ 0 & 4 \end{bmatrix}$ diagonal, scalar, or identity?
(Answer: diagonal and scalar, since the diagonal entries are equal; not identity, because they are not 1.)Classify $\begin{bmatrix} 0 & -5 \ 5 & 0 \end{bmatrix}$.
(Answer: square and skew-symmetric, since $A^{T} = -A$ and the diagonal is zero.)Is $\begin{bmatrix} 2 & 6 \ 1 & 3 \end{bmatrix}$ singular or non-singular?
(Answer: singular, because $\det = (2)(3) - (6)(1) = 0$.)What type is $\begin{bmatrix} 3 & 1 & 0 \ 0 & 5 & 2 \ 0 & 0 & 7 \end{bmatrix}$, and what is its determinant?
(Answer: upper triangular; $\det = 3 \times 5 \times 7 = 105$, the product of the diagonal.)Show that $\begin{bmatrix} 1 & 0 \ 0 & -1 \end{bmatrix}$ is involutory.
(Answer: $A^{2} = \begin{bmatrix} 1 & 0 \ 0 & 1 \end{bmatrix} = I$, so it is its own inverse.)
Where Should You Go Next After Types Of Matrices?
Knowing the types is the map. The next steps are the roads that connect them, and each one deepens a name you just met.
Order Of A Matrix. The rows-by-columns idea that every shape-based type is built on.
Transpose Of A Matrix. The operation that defines the symmetric and skew-symmetric types.
Matrices And Determinants. The determinant that separates singular from non-singular matrices, and much more.
If your child is building these foundations, a live Bhanzu trainer teaches the types of matrices starting from what each one is for, not just what it looks like, in the Bhanzu algebra program.
Was this article helpful?
Your feedback helps us write better content