A 2 × 2 matrix can describe a transformation of the whole plane. Multiply the matrix by a point's position vector and you get the image of that point. In AI HL you need the matrices for rotations about the origin, reflections in lines through the origin, stretches and enlargements. You also need to combine them in the right order, add a translation vector, and use the determinant as the factor by which areas change.
The topic sits in the additional higher level part of Topic 3, geometry and trigonometry, and it relies on the matrix skills from Topic 1. AI SL students do not study it, and every AI paper requires a GDC, so the multiplication is quick. The marks go to setting matrices up correctly and describing transformations precisely. If you want steady help through the HL content, see our IB maths AI HL tuition page. Matrix operations, eigenvalues and transition matrices are covered in our separate guide to matrices, eigenvalues and Markov chains, so this post stays with geometry.
The one idea that builds every matrix
Look at where the unit vectors go. The first column of the matrix is the image of (1, 0), and the second column is the image of (0, 1).
Take a rotation of 90° anticlockwise about the origin. (1, 0) goes to (0, 1) and (0, 1) goes to (−1, 0). So the matrix has columns (0, 1) and (−1, 0):
R = [[0, −1], [1, 0]] (written row by row).
With this idea, you never need to memorise a matrix you can rebuild in ten seconds. Here are the standard forms:
| Transformation | Matrix (row by row) | Determinant |
|---|---|---|
| Rotation θ anticlockwise about O | [[cos θ, −sin θ], [sin θ, cos θ]] | 1 |
| Rotation θ clockwise about O | [[cos θ, sin θ], [−sin θ, cos θ]] | 1 |
| Reflection in y = (tan θ)x | [[cos 2θ, sin 2θ], [sin 2θ, −cos 2θ]] | −1 |
| Reflection in the x-axis | [[1, 0], [0, −1]] | −1 |
| Reflection in y = x | [[0, 1], [1, 0]] | −1 |
| Stretch parallel to the x-axis, factor k | [[k, 0], [0, 1]] | k |
| Stretch parallel to the y-axis, factor k | [[1, 0], [0, k]] | k |
| Enlargement, centre O, scale factor k | [[k, 0], [0, k]] | k² |
A stretch parallel to the x-axis keeps the y-axis fixed and multiplies every x-coordinate by k.
Worked example 1: the order of composition matters
A shape is rotated 90° anticlockwise about O, and the result is then reflected in the line y = x.
R = [[0, −1], [1, 0]] and F = [[0, 1], [1, 0]].
The matrix nearest the point acts first, so the combined matrix is F × R, not R × F.
F × R = [[0, 1], [1, 0]] × [[0, −1], [1, 0]] = [[1, 0], [0, −1]].
This is a single reflection in the x-axis.
Now reverse the order: reflect first, then rotate.
R × F = [[−1, 0], [0, 1]], which is a reflection in the y-axis.
Same two transformations, different single result. In exams, write "first A, then B, so the combined matrix is BA" as a line of working.
Worked example 2: the determinant as an area scale factor
Triangle P has vertices (1, 1), (4, 1) and (1, 3). It is transformed by M = [[2, 1], [−1, 3]].
(a) Find the vertices of the image.
Put the vertices as columns and multiply:
M × [[1, 4, 1], [1, 1, 3]] = [[3, 9, 5], [2, −1, 8]].
The image has vertices (3, 2), (9, −1) and (5, 8).
(b) Find the area of the image.
Area of P = ½ × base 3 × height 2 = 3 units².
det M = 2 × 3 − 1 × (−1) = 7.
Area of image = |det M| × 3 = 21 units².
Check: the shoelace formula on the three image vertices also gives 21. A negative determinant would mean the shape has been flipped as well, so always use the absolute value for area.
Worked example 3: reflection in a line that is not an axis
Find the matrix for a reflection in the line y = 0.5x, and the image of (5, 0).
The line makes angle θ with the x-axis, where tan θ = 0.5. You need cos 2θ and sin 2θ:
cos 2θ = (1 − tan²θ) ÷ (1 + tan²θ) = 0.75 ÷ 1.25 = 0.6.
sin 2θ = 2 tan θ ÷ (1 + tan²θ) = 1 ÷ 1.25 = 0.8.
Matrix: [[0.6, 0.8], [0.8, −0.6]].
Image of (5, 0): (0.6 × 5, 0.8 × 5) = (3, 4).
Check: the midpoint of (5, 0) and (3, 4) is (4, 2), which lies on y = 0.5x. The point and its image are both 5 units from O. On the GDC you can work in degrees here instead: θ = 26.565…°, so 2θ = 53.13°, and cos 53.13° = 0.6.
Worked example 4: adding a translation, then working backwards
AI HL also uses transformations of the form T(x) = Ax + b, where b is a translation vector.
Let A = [[3, 1], [2, 1]] and b = (−2, 5). A point Q is mapped to (9, 12). Find Q.
Subtract the translation first: AQ = (9, 12) − (−2, 5) = (11, 7).
det A = 3 − 2 = 1, so A⁻¹ = [[1, −1], [−2, 3]].
Q = A⁻¹ × (11, 7) = (11 − 7, −22 + 21) = (4, −1).
Check forwards: A × (4, −1) = (11, 7), and adding b gives (9, 12). ✓
The common error is multiplying by the inverse before removing the translation. Undo the steps in reverse order, like taking off shoes before socks.
Mistakes our tutors see most often
- Composing in reading order. "A then B" is BA.
- Clockwise and anticlockwise swapped. Check with (1, 0): a 90° anticlockwise turn must send it to (0, 1).
- Using θ instead of 2θ in the reflection matrix.
- Area scaled by det without the modulus, giving a negative area.
- Confusing stretch direction. "Parallel to the x-axis" changes x-coordinates.
- Vague descriptions. "A rotation" earns little. "A rotation of 36.9° anticlockwise about the origin" earns the mark.
Exam technique
- Describe fully: type, angle and direction for rotations, the line for reflections, factor and direction for stretches, centre and factor for enlargements.
- Use the unit-vector check on any matrix you write down. It takes seconds and catches sign errors.
- Use the GDC for products and inverses, but write the matrix product you are computing so method marks are available. Our guide to entering matrices on your GDC covers the menus.
- Identify a mystery matrix by testing (1, 0) and (0, 1). Then check the determinant: 1 suggests a rotation, −1 a reflection.
- Repeated transformations become matrix powers: applying M five times is M⁵.
These matrix transformations are different from the graph transformations of functions, such as y = f(x − 2) or y = 3f(x). AA students study those in graph transformations in AA. Do not mix the two notations in an answer.
Self-check
- Show that applying the rotation of 60° anticlockwise about O six times gives the identity.
- Describe the transformation [[0.8, −0.6], [0.6, 0.8]].
- Find the image of (2, 5) under [[1, 0], [0, 3]], and the area scale factor.
- A stretch parallel to the x-axis with factor 2 is followed by a reflection in the x-axis. Find the single matrix and its determinant.
- Find the image of (4, 2) after a rotation of 30° anticlockwise about O followed by an enlargement, centre O, scale factor 2.
Answers: 1) Six turns of 60° make 360°, and the GDC confirms M⁶ = I. 2) A rotation of 36.9° anticlockwise about O, since cos θ = 0.8 and sin θ = 0.6. 3) (2, 15), factor 3. 4) [[2, 0], [0, −1]], determinant −2. 5) (4√3 − 2, 4 + 2√3) ≈ (4.93, 7.46).
Where this leads
A rotation matrix behaves like multiplying by a complex number of modulus 1, a link explored in our post on vectors and complex numbers at AI HL. Matrices then return in eigenvectors and coupled differential equations. Our sibling guide to calculus topics on the AI HL course shows where those systems come from, and the complete AI HL syllabus puts every topic in context.
How Maths Bodhi helps
In the free demo, the tutor gives a student two matrices to identify and one composition to set up. This shows quickly whether the issue is the geometry, the order or the description. Tutors are matched to AI HL, and practice is past-paper style with the student's written method checked each session. For families comparing support across AA and AI, see our IB maths tuition for every course.
Home tuition is available in Gurgaon (Gurugram). Students elsewhere in India and abroad learn with us online. You can book a free demo class.
Written and checked by the Maths Bodhi team. Facts checked against official IB sources, November 2026.
Frequently asked questions
- Are matrix transformations in IB maths AI HL?
- Yes. In the current course, AI HL includes 2 × 2 matrices for rotations, reflections, stretches and enlargements, compositions of these, translations added as a vector, and the determinant as an area scale factor. AI SL does not include them.
- How do I find the matrix for a transformation?
- Find the images of (1, 0) and (0, 1). The first image is the first column of the matrix and the second image is the second column.
- What does the determinant of a transformation matrix tell you?
- Its absolute value is the area scale factor. A negative determinant means the orientation is reversed, which happens with a reflection.
- If I do transformation A then B, which matrix product do I use?
- Use BA. The matrix next to the point acts first, so the first transformation goes on the right.
- What is the matrix for a reflection in the line y = (tan θ)x?
- It is [[cos 2θ, sin 2θ], [sin 2θ, −cos 2θ]] written row by row. Note the 2θ, which is a common source of lost marks.
- Can I use my GDC for matrix transformations?
- Yes, every AI paper requires a GDC, so you can multiply matrices and find inverses on it. You still need to write the product you are working out and describe the transformation in words.
