Determinants and Invertibility

When a matrix acts on a region of space, it can stretch it, squish it, rotate it, or flatten it entirely. The determinant is the number that captures exactly how much area (or volume) changes — and whether orientation flips. It's also the gatekeeper of invertibility: a matrix can be undone if and only if its determinant is nonzero.

---

The 2×2 Determinant

Take a matrix

$$A = \begin{bmatrix} a & b \\ c & d \end{bmatrix}$$

Its determinant is:

$$\det(A) = ad - bc$$

To see why this formula makes geometric sense, think of the two columns $\mathbf{u} = \begin{bmatrix} a \\ c \end{bmatrix}$ and $\mathbf{v} = \begin{bmatrix} b \\ d \end{bmatrix}$ as vectors in the plane. They span a parallelogram. The area of that parallelogram is exactly $|ad - bc|$.

The signed version $ad - bc$ carries orientation: positive means the pair $(\mathbf{u}, \mathbf{v})$ has the same "handedness" as the standard basis, negative means they've been flipped.

Drag the two vectors and watch the parallelogram and determinant value update live. Notice that when the vectors become parallel, the parallelogram collapses and $\det \to 0$.

---

What It Means When det = 0

When $\det(A) = 0$, the two columns lie along the same line — they are linearly dependent. The transformation $A$ squashes the entire plane onto a line (or a point). This is catastrophic for one reason: information is permanently lost.

If $A\mathbf{x}_1 = A\mathbf{x}_2$ for two distinct $\mathbf{x}_1 \neq \mathbf{x}_2$, you can't tell them apart after the transformation. No matter how hard you try, there's no way to reconstruct which input produced a given output.

Formally, $\ker(A) \neq \{\mathbf{0}\}$ — there are nonzero vectors that get mapped to zero.

---

Sign of the Determinant

• $\det(A) > 0$: the transformation preserves orientation. A counterclockwise loop stays counterclockwise.
• $\det(A) < 0$: the transformation reverses orientation — like a reflection. A counterclockwise loop becomes clockwise.
• $\det(A) = 0$: the transformation collapses space to a lower dimension.

Pure rotations always have $\det = 1$ (they preserve area and orientation). Reflections have $\det = -1$. A uniform scaling by factor $k$ gives $\det = k^2$ for 2D.

---

det(AB) = det(A) · det(B)

When you compose two linear maps, their scaling factors multiply:

$$\det(AB) = \det(A) \cdot \det(B)$$

Intuitively: if $A$ doubles area and $B$ triples it, then $AB$ multiplies area by six. This multiplicativity is what makes determinants so powerful — they turn composition (a complex operation) into ordinary multiplication.

A direct consequence: $\det(A^{-1}) = 1/\det(A)$ (since $\det(A) \cdot \det(A^{-1}) = \det(I) = 1$).

---

Invertibility: det(A) ≠ 0

A square matrix $A$ is invertible if and only if $\det(A) \neq 0$.

The logic flows directly from geometry: if $\det(A) \neq 0$, the transformation preserves dimension — no information is lost, so there's a unique input for every output, meaning we can reverse it. If $\det(A) = 0$, distinct inputs collide and no reversal is possible.

Use the sliders to see how rotation alone keeps $\det = 1$, while a squish factor $s < 1$ shrinks the determinant — and at $s = 0$ the square collapses to a line, making the transformation irreversible.

---

The 2×2 Inverse Formula

For an invertible 2×2 matrix, the inverse has a clean closed form:

$$A^{-1} = \frac{1}{\det(A)} \begin{bmatrix} d & -b \\ -c & a \end{bmatrix}$$

You can verify this directly: $A \cdot A^{-1} = I$. The $1/\det(A)$ factor undoes the scaling, while swapping $a \leftrightarrow d$ and negating $b, c$ reverses the rotation and shear.

Geometrically: the matrix $\begin{bmatrix} d & -b \\ -c & a \end{bmatrix}$ maps each column of $A$ to the corresponding standard basis vector (it "undoes" the parallelogram), and dividing by $\det(A)$ corrects the scale.

Watch as $A$ transforms the grid, then $A^{-1}$ brings it back exactly.

---

The 3×3 Determinant

For a $3 \times 3$ matrix, the determinant is computed by cofactor expansion along the first row:

$$\det \begin{bmatrix} a_{11} & a_{12} & a_{13} \\ a_{21} & a_{22} & a_{23} \\ a_{31} & a_{32} & a_{33} \end{bmatrix} = a_{11} \det \begin{bmatrix} a_{22} & a_{23} \\ a_{32} & a_{33} \end{bmatrix} - a_{12} \det \begin{bmatrix} a_{21} & a_{23} \\ a_{31} & a_{33} \end{bmatrix} + a_{13} \det \begin{bmatrix} a_{21} & a_{22} \\ a_{31} & a_{32} \end{bmatrix}$$

Each $2 \times 2$ minor is the determinant of the submatrix obtained by deleting that element's row and column. The alternating signs ($+, -, +$) come from the checkerboard pattern of cofactors.

The 3D geometric interpretation: $|\det(A)|$ equals the volume of the parallelepiped spanned by the three column vectors, and the sign captures orientation in 3D (right-handed vs left-handed).

---

Summary

Concept	Key Fact
$\det(A)$	Signed area/volume scaling factor
$\det = 0$	Columns dependent; transformation collapses dimension
$\det > 0$	Orientation preserved
$\det < 0$	Orientation flipped
$\det(AB)$	$= \det(A) \cdot \det(B)$
Invertibility	$A$ invertible $\iff \det(A) \neq 0$
2×2 inverse	$\frac{1}{\det(A)}\begin{bmatrix}d&-b\\-c&a\end{bmatrix}$

The determinant is one of the most information-dense numbers in linear algebra. Next, we'll use these ideas to understand eigenvalues — another place where the determinant plays a starring role.