2x2 matrix determinant

We have already seen “determinants” in the context of linear systems. This is actually a quantity that we can compute for any square matrice. Its value can provide us some important information about the matrix. In particular, just by look at the determinant of a matrix, we can know if the matrix is invertible.

Determinant: 2x2 cases

As we saw earlier, a matrix with a zero row or zero column will not be invertible. Similarly, a matrix that has column that is a multiple or another column will also not be invertible. In general, the “invertibility” can be determined by computing the “matrix determinant”. For a $2 \times 2$ matrix

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

its determinant is a number, denoted $\det A$, is given by the formula

\[\det A = \det \begin{bmatrix} a & b \\ c & d \end{bmatrix} = ad - bc.\]

From a geometric point of view, we can see that the determinant tells us if the two column vectors are collinear or zero. That is, the determinant is 0 exactly when

  • One of the column vector is zero;
  • Both columns are zero;
  • The two column vectors have the same direction; or
  • The two column vectors have the exact opposite directions.

Exercise. It is worth thinking about the geometric meaning of $|\det A|$ as well. Prove that $|\det A|$ is the area of the parallelogram spanned by the two columns of $A$.

Exercise. What does the sign of the determinant tell us?

Exercise. Prove that for a lower triangular or upper triangular matrix, i.e., matrices of the form

\[\begin{aligned} & \begin{bmatrix} a_{11} & a_{12} \\ 0 & a_{22} \end{bmatrix} &&\text{or} & \begin{bmatrix} a_{11} & 0 \\ a_{21} & a_{22} \end{bmatrix}, \end{aligned}\]

their determinants are simply the product of their diagonal entries.

From an algebraic point of view, determinants tell us if a matrix is invertible. For a $2 \times 2$ matrix $A$,

\[\det A \ne 0 \quad\Leftrightarrow\quad A^{-1} \text{ exists}.\]

Exercise. Can you see why this criterion works?

For $2 \times 2$ matrices, the following properties hold:

\[\begin{aligned} \det(A^\top) &= \det(A) & \det(AB) &= \det(A) \det(B). \end{aligned}\]

Consequently, if $A$ is invertible, then

\[\det(A^{-1}) = \frac{1}{\det(A)}.\]

Determinant and row/column operations

From the formula

\[\det \begin{bmatrix} a & b \\ c & d \end{bmatrix} = ad - bc\]

we can see that

  • Scaling one row/column by a factor will scale the determinant by the same factor;
  • Permuting the rows/column will alter the sign of the determinant;
  • Adding a multiple of a row/column to another row/column will not change the determinant.

These properties suggest that we can also compute the determinant of a matrix by performing elementary row/column operations and reducing the matrix into a simple form (although this seems unnecessary for the $2 \times 2$ case). We will soon explore this idea for matrices of larger sizes

Inverse formula: 2x2 cases

Indeed, when $\det A \ne 0$, we have a formula for the inverse:

\[\begin{bmatrix} a & b \\ c & d \end{bmatrix}^{-1} = \frac{1}{ad - bc} \begin{bmatrix} d & -b \\ -c & a \end{bmatrix},\]

if $ad - bc \ne 0$.

Exercise. Determine if the following matrices are invertible. If yes, compute their inverses.

\[\begin{aligned} A &= \begin{bmatrix} 1 & 2 \\ 3 & 6 \end{bmatrix} & B &= \begin{bmatrix} 1 & 2 \\ 3 & 7 \end{bmatrix} & C &= \begin{bmatrix} 1 & 2 \\ 3 & 8 \end{bmatrix} \\ D &= \begin{bmatrix} 0 & 1 \\ 0 & 0 \end{bmatrix} & E &= \begin{bmatrix} 3 & -4 \\ 4 & 3 \end{bmatrix} & R &= \begin{bmatrix} \cos \theta & -\sin \theta \\ \sin \theta & \cos \theta \end{bmatrix} \end{aligned}\]