Determinants

A rigorous treatment of determinants as signed volume scaling, invertibility tests, and algebraic invariants.

A linear transformation moves an entire grid while preserving lines and the origin. A square source grid maps to a skewed image grid. Basis vectors and one eigen-direction are highlighted, while a signed area tile represents the determinant’s scale factor.

Precise definition

The determinant assigns a scalar $\det A$ to a square matrix. It is multilinear in the rows or columns, changes sign when two are swapped, and equals 1 on the identity. Geometrically, $|\det A|$ is the factor by which the associated linear map scales $n$-dimensional volume; the sign records orientation.

Notation and mathematical language

For $2\times2$, $\det\begin{pmatrix}a&b\\c&d\end{pmatrix}=ad-bc$. Row replacement leaves the determinant unchanged, swapping rows negates it, and scaling one row by $k$ scales the determinant by $k$. Also $\det(AB)=\det A\det B$ and $\det A^T=\det A$.

Conceptual picture

A zero determinant means the columns collapse dimension and are linearly dependent. Equivalent statements for a square matrix include invertibility, full rank, a pivot in every row and column, and only the trivial solution to $A\mathbf x=0$.

Fully worked example

Interpretation and application

Determinants appear in change of variables, orientation, eigenvalue products, and uniqueness tests. A large determinant does not necessarily mean a matrix is numerically well-conditioned because rescaling units changes it.