- Foundation
- About 18 min
- 5 practice questions
Invertible matrices
Know when a square matrix has an inverse, prove that the inverse is unique and that (AB)⁻¹ = B⁻¹A⁻¹, and find or check the inverse of a 2 × 2 matrix.
If something looks wrong, please tell us.
By the end of this lesson you can
- Define the inverse of a square matrix and check a claimed inverse by multiplication.
- Prove that an inverse is unique and that (AB)⁻¹ = B⁻¹A⁻¹.
- Find the inverse of a 2 × 2 matrix from AX = I, recognise a matrix with no inverse, and use a matrix equation such as A² − 5A + 6I = O to find A⁻¹.
For numbers, dividing by 5 is the same as multiplying by 1/5, the number that gives 1 when multiplied by 5. For matrices there is no division, but some square matrices have a partner that multiplies with them to give the identity I. That partner, the inverse, is what lets us 'divide' and later solve systems of equations in the next chapter.
How matrices are written in these lessons: entries in a row are separated by spaces and rows by semicolons, so [2 −1; 3 0] is the 2 × 2 matrix with first row 2, −1 and second row 3, 0.
Definition
Let A be a square matrix of order m. If there is a square matrix B of the same order with AB = BA = I, then B is the inverse of A, written A⁻¹, and A is invertible. Then A is also the inverse of B. A rectangular matrix has no inverse, because AB and BA can only both be defined and equal when A and B are square of the same order.
B = A⁻¹ ⇔ AB = BA = I. The inverse, if it exists, is unique, and (AB)⁻¹ = B⁻¹A⁻¹.
Two theorems
Uniqueness. Suppose B and C are both inverses of A. Then B = BI = B(AC) = (BA)C = IC = C. Each step uses either the definition or associativity, so an invertible matrix has exactly one inverse.
Inverse of a product. If A and B are invertible of the same order, then (AB)(B⁻¹A⁻¹) = A(BB⁻¹)A⁻¹ = AIA⁻¹ = AA⁻¹ = I, and similarly (B⁻¹A⁻¹)(AB) = B⁻¹(A⁻¹A)B = B⁻¹B = I. So (AB)⁻¹ = B⁻¹A⁻¹: the order reverses, just as for transposes.
Worked examples
Worked example 1 (checking). Is B = [2 −5; −1 3] the inverse of A = [3 5; 1 2]? Working: AB = [3·2 + 5·(−1), 3·(−5) + 5·3; 1·2 + 2·(−1), 1·(−5) + 2·3] = [1 0; 0 1]. BA = [2·3 + (−5)·1, 2·5 + (−5)·2; −1·3 + 3·1, −1·5 + 3·2] = [1 0; 0 1]. Both products are I, so B = A⁻¹.
Worked example 2 (finding). Find the inverse of A = [2 1; 5 3]. Working: let X = [p q; r s] with AX = I. Column 1: 2p + r = 1 and 5p + 3r = 0. From the first, r = 1 − 2p; then 5p + 3 − 6p = 0, so p = 3 and r = −5. Column 2: 2q + s = 0 and 5q + 3s = 1. So s = −2q; then 5q − 6q = 1, giving q = −1 and s = 2. So A⁻¹ = [3 −1; −5 2]. Check XA = [6 − 5, 3 − 3; −10 + 10, −5 + 6] = I ✓.
Worked example 3 (no inverse). Does A = [1 2; 2 4] have an inverse? Working: AX = I gives, for the first column, p + 2r = 1 and 2p + 4r = 0. Doubling the first equation gives 2p + 4r = 2, which contradicts 2p + 4r = 0. So no X exists and A is not invertible. The second row of A is twice the first; in Chapter 4 you will see that this makes its determinant 0.
Using a matrix equation to find A⁻¹
- Start from the equation e.g. A² − 5A + 6I = O
- Move the I term A² − 5A = −6I
- Factor out A A(A − 5I) = −6I, i.e. A(5I − A) = 6I
- Read off the inverse A⁻¹ = (1/6)(5I − A)
Worked example 4. In the previous lesson, A = [2 1; 0 3] satisfied A² − 5A + 6I = O. Working: A(5I − A) = 6I, so A⁻¹ = (1/6)(5I − A) = (1/6)[5 − 2, 0 − 1; 0 − 0, 5 − 3] = (1/6)[3 −1; 0 2]. Check: A[3 −1; 0 2] = [6 + 0, −2 + 2; 0, 6] = 6I, so dividing by 6 gives I ✓.
Worked example 5 (product rule). With A = [3 5; 1 2] and B = [2 1; 5 3] from above, A⁻¹ = [2 −5; −1 3] and B⁻¹ = [3 −1; −5 2]. Then (AB)⁻¹ = B⁻¹A⁻¹ = [3·2 + (−1)(−1), 3·(−5) + (−1)·3; −5·2 + 2·(−1), −5·(−5) + 2·3] = [7 −18; −12 31]. (Check: AB = [31 18; 12 7], and [31 18; 12 7][7 −18; −12 31] = [217 − 216, −558 + 558; 84 − 84, −216 + 217] = I ✓.)
| Matrix | Inverse | Check |
|---|---|---|
| [3 5; 1 2] | [2 −5; −1 3] | product is I both ways |
| [2 1; 5 3] | [3 −1; −5 2] | product is I both ways |
| [2 1; 0 3] | (1/6)[3 −1; 0 2] | from A² − 5A + 6I = O |
| [1 2; 2 4] | none | equations inconsistent |
Key takeaways: only square matrices can have inverses, and not all do. An inverse is unique; check it by multiplying both ways. (AB)⁻¹ = B⁻¹A⁻¹. A 2 × 2 inverse can be found by solving AX = I, and a polynomial equation in A can be rearranged to give A⁻¹.
Check your understanding
-
1. Which matrix is the inverse of [1 2; 0 1]?
- [1 −2; 0 1]
- [1 0; −2 1]
- [−1 2; 0 −1]
- [1 2; 0 −1]
Show answer
Answer: A. [1 2; 0 1][1 −2; 0 1] = [1, −2 + 2; 0, 1] = I, and the product in the other order is also I. So the inverse is [1 −2; 0 1].
-
2. If A and B are invertible matrices of the same order, which expression equals (AB)⁻¹?
- A⁻¹B⁻¹
- (BA)⁻¹
- B⁻¹A⁻¹
- AB
Show answer
Answer: C. (AB)(B⁻¹A⁻¹) = A(BB⁻¹)A⁻¹ = AA⁻¹ = I, and similarly the other way. So (AB)⁻¹ = B⁻¹A⁻¹, with the order reversed.
-
3. A square matrix A satisfies A² − 3A + 2I = O. Which expression equals A⁻¹?
- 3I − A
- (A − 3I)/2
- (3I + A)/2
- (3I − A)/2
Show answer
Answer: D. A² − 3A = −2I, so A(3I − A) = 2I. Dividing by 2: A × (3I − A)/2 = I. Hence A⁻¹ = (3I − A)/2.
-
4. What is the inverse of A = [4 3; 1 1]?
- [1 3; 1 4]
- [−1 3; 1 −4]
- [1 −3; −1 4]
- [4 −3; −1 1]
Show answer
Answer: C. Check: [4 3; 1 1][1 −3; −1 4] = [4 − 3, −12 + 12; 1 − 1, −3 + 4] = I, and [1 −3; −1 4][4 3; 1 1] = [4 − 3, 3 − 3; −4 + 4, −3 + 4] = I. So the inverse is [1 −3; −1 4].
-
5. Assertion (A): A 2 × 3 matrix has no inverse.
Reason (R): If AB = BA = I, then A and B must be square matrices of the same order.- Both A and R are true, and R is the correct explanation of A
- Both A and R are true, but R is not the correct explanation of A
- A is true but R is false
- A is false but R is true
Show answer
Answer: A. For a 2 × 3 matrix A, AB = I needs B of order 3 × 2 and gives a 2 × 2 product, while BA would be 3 × 3, so they cannot both equal the same I. That is exactly R, and it explains A.
Sign in to keep track of the lessons you finish.