A Matrix Is a Transformation

By Victor Arce

A matrix can look like a lifeless box of numbers, but the clearest way to understand one is geometric: a 2×2 matrix is a transformation of space. It picks up the entire plane — every point, every vector — and moves it to a new place, while keeping grid lines straight, evenly spaced, and the origin fixed. That property is what we call linear.

Step through it below. Watch the grid and the basis vectors î and ĵ morph as the matrix is applied, then see any vector ride along for free.

The columns tell you everything

A linear transformation is completely pinned down by where it sends the two basis vectors î = (1, 0) and ĵ = (0, 1). And those landing spots are exactly the columns of the matrix:

M = [[2, 1], [1, 2]] sends î → (2, 1) (first column) and ĵ → (1, 2) (second column).

Once you know where î and ĵ go, every other vector is determined, because a vector is just a combination of them. v = (1, 1) = 1·î + 1·ĵ, so after the transform it becomes 1·(2, 1) + 1·(1, 2) = (3, 3). The components didn't change — only the basis they ride on did. That is matrix-vector multiplication:

M · v = (2·1 + 1·1, 1·1 + 2·1) = (3, 3).

The determinant is an area

The last step shows the unit square — the little box spanned by î and ĵ — becoming a parallelogram. Its area is the determinant:

det M = 2·2 − 1·1 = 3.

So this matrix triples every area. A determinant of 1 preserves area, a determinant between 0 and 1 shrinks it, and a negative determinant flips orientation (a mirror). A determinant of 0 collapses space onto a line — the transformation can't be undone.

This single idea — track î and ĵ, everything else follows — is the engine behind rotating and scaling in graphics, solving systems of equations, and transforming data in machine learning. Matrix multiplication is just doing one such transformation after another.

Back to Matrix Math