How to Understand Matrices

The direct answer: a matrix is a rectangular grid of numbers, described by its rows and columns, that you can add, scale, and multiply by rules that differ from ordinary arithmetic. Matrices encode systems of equations, transformations, and data, and they are the central object of linear algebra. This guide covers the operations and the ideas of determinant and inverse.

Matrices at a Glance

The Basics

A matrix with m rows and n columns is called m by n. The entry in row i, column j is written a_ij. A 2 by 3 matrix looks like:

[ 1 2 3 ] [ 4 5 6 ]

A vector is just a matrix with one column. That is why matrices and vectors are taught together. The size matters for every operation that follows.

Addition and Scalar Multiplication

You can add two matrices only when they share the same dimensions. Add entry by entry:

[ 1 2 ] + [ 3 4 ] = [ 4 6 ] [ 5 6 ] [ 7 8 ] [ 12 14 ]

To multiply a matrix by a number c, multiply every entry by c:

3 times [ 1 2 ] = [ 3 6 ] [ 3 4 ] [ 9 12 ]

Scalar multiplication is the easy one. Students rarely err here, which makes it a good warm up before the operation that trips everyone up.

Matrix Multiplication

Multiplication is the operation students most often get wrong, because it is not entrywise. To find entry (i, j) of the product AB, take row i of A and column j of B, multiply matching pairs, and add them.

The 2 by 2 pattern

For 2 by 2 matrices:

[ a b ] [ e f ] = [ a e + b g a f + b h ] [ c d ] [ g h ] [ c e + d g c f + d h ]

Example:

[ 1 2 ] [ 3 0 ] = [ 13 + 21 10 + 22 ] = [ 5 4 ] [ 0 1 ] [ 1 2 ] [ 03 + 11 00 + 12 ] [ 1 2 ]

Why order matters

Matrix multiplication is generally not commutative: AB is usually not equal to BA. Always keep the order. The dimensions tell you whether the product is even defined, a point the Khan Academy linear algebra track drills with diagrams Khan Academy.

The Identity and the Inverse

The identity matrix I has 1s on the main diagonal and 0s elsewhere. For 2 by 2:

I = [ 1 0 ] [ 0 1 ]

It behaves like the number 1: A times I = A.

A square matrix A has an inverse A inverse only when its determinant is not zero. The inverse undoes the matrix: A times A inverse = I. If the determinant is zero, the matrix is singular and has no inverse, which in a system of equations means the rows are dependent.

The 2 by 2 inverse formula

For a 2 by 2 matrix [ a b ; c d ], the determinant is a d minus b c. The inverse, when a d minus b c is not zero, is:

(1 divided by (a d minus b c)) times [ d minus b ] [ minus c a ]

Example: A = [ 2 1 ; 1 1 ]. Determinant = 21 minus 11 = 1. Inverse = [ 1 minus 1 ; minus 1 2 ]. Check: A times A inverse = [ 21+1(-1) 2(-1)+12 ; 11+1(-1) 1(-1)+12 ] = [ 1 0 ; 0 1 ] = I.

Why Matrices Matter

A system of linear equations becomes Ax = b, where A holds coefficients. Solving the system is finding x, often via row reduction or an inverse. Beyond algebra, matrices represent rotations and scalings in graphics, transitions in probability, and the structure of datasets in machine learning. They are the language behind most large scale computation, which is why linear algebra treats them as the main object rather than a side topic.

Solving a System With the Inverse

The inverse is not just a definition to memorize. It solves Ax = b directly.

The method

If A is invertible, multiply both sides of Ax = b on the left by A inverse:

A inverse times A times x = A inverse times b x = A inverse times b

For the 2 by 2 from earlier, A = [ 2 1 ; 1 1 ] with inverse [ 1 minus 1 ; minus 1 2 ], solve Ax = [ 3 ; 2 ]:

x = [ 1 minus 1 ; minus 1 2 ] times [ 3 ; 2 ] = [ 13 + minus 12 ; minus 13 + 22 ] = [ 1 ; 1 ].

Check: 21 + 11 = 3 and 11 + 11 = 2. Correct.

When this fails

If the determinant is zero, no inverse exists and this route closes. The system then either has no solution or infinitely many, and you fall back to row reduction to tell which. That is the practical reason the determinant is the first thing you compute before reaching for the inverse formula.

A 3 by 3 Determinant in Brief

The 2 by 2 formula generalizes, and the pattern explains why size matters. For a 3 by 3 matrix, the determinant expands along the top row by minors:

det = a(ei minus fh) minus b(di minus fg) plus c(dh minus eg)

The plus, minus, plus pattern follows the checkerboard of signs. Each term drops one row and one column to leave a 2 by 2 you already know how to compute. The calculation is longer but the idea is identical: a single number that tells you whether the matrix is invertible.

What a nonzero determinant means geometrically

For a 2 by 2 matrix, the absolute value of the determinant equals the area of the parallelogram spanned by its column vectors. For 3 by 3 it equals a volume. A determinant of zero means the columns collapse onto a line or plane, which is the geometric picture behind "no inverse." This view connects algebra to the transformations section: a matrix with determinant 1 preserves area, a fact used constantly in computer graphics.

Common Misconceptions

• Multiplying matrices entrywise instead of row by column.
• Assuming AB = BA.
• Adding matrices of different sizes.
• Forgetting that only square matrices can have inverses, and only when the determinant is nonzero.
• Dropping the 1 over determinant factor when writing an inverse.
• Thinking division exists for matrices. It is replaced by multiplying by an inverse, and only when that inverse exists.
• Believing a zero determinant is a calculation error. It can be a real signal that the system has no unique solution.

Frequently Asked Questions

Can you multiply any two matrices?

Only when the number of columns in the first equals the number of rows in the second. An m by n matrix times an n by p matrix gives an m by p matrix.

What does a determinant of zero mean?

The matrix has no inverse and its rows (or columns) are linearly dependent. In a system Ax = b, it signals either no solution or infinitely many.

Is matrix multiplication like regular multiplication?

It shares some rules, like associativity, but not commutativity, and division is replaced by multiplying by an inverse.

Do I need matrices for calculus?

Yes, in multivariable calculus and differential equations. They also appear in data and computer science.

How do I know if a matrix is invertible?

Compute its determinant. If it is nonzero, the inverse exists.

How are matrices related to vectors?

A vector is a matrix with one column, so the same multiplication and transformation ideas apply. See vectors for the building blocks.

Sources

• Khan Academy, matrix transformations
• Math is Fun, matrices