Transformations in R² & R³ — Transformations in R² and R³...

SECTION 1

Historical Context & Motivation

People have been moving, flipping, and stretching shapes for thousands of years. Ancient Greek mathematicians studied geometric transformations — ways of changing the position or shape of a figure — long before anyone wrote a matrix. Over time, mathematicians found a powerful shortcut: you can describe every rotation, reflection, projection, and shear using a simple grid of numbers called a matrix. This idea connects geometry (shapes you can see) with algebra (equations you can solve), and it is the backbone of modern computer graphics, robotics, and physics.

~300 BCE

Euclid's Elements

Euclid described rigid motions — translations, rotations, and reflections — as ways to prove two figures are congruent. These ideas form the oldest known theory of geometric transformations.

1827

Möbius & Projective Geometry

August Ferdinand Möbius introduced coordinate-based descriptions of projections, connecting geometry to algebraic formulas for the first time.

1858

Cayley's Matrix Algebra

Arthur Cayley published the first systematic theory of matrices, showing that a 2 × 2 grid of numbers could encode a transformation of the plane.

1960s–1970s

Computer Graphics Revolution

Pioneers like Ivan Sutherland used matrix transformations to rotate and project 3-D objects onto a 2-D screen, launching the era of video games and CGI movies.

The central question this lesson answers is: How can we use matrices to describe every common way of transforming points in 2-D (R²) and 3-D (R³) space? By the end, you will be able to recognize and apply rotations, reflections, projections, and shears using matrix multiplication.

SECTION 2

Core Principles & Definitions

A linear transformation is a rule that takes every point (or vector) in a space and sends it to a new point, while obeying two properties: straight lines stay straight, and the origin stays fixed. Every linear transformation in R² can be written as multiplication by a 2 × 2 matrix, and every linear transformation in R³ can be written as multiplication by a 3 × 3 matrix. The four most important families of transformations are rotations, reflections, projections, and shears.

Rotation

Spins every point around the origin by a fixed angle. Distances from the origin don't change, and angles between vectors are preserved.

Reflection

Flips every point across a line (in R²) or a plane (in R³), like looking in a mirror. Distances are preserved, but orientation reverses.

Projection

Squashes every point onto a line or plane, like a shadow. Dimension is reduced, so information is lost — you can't undo a projection.

Shear

Slides layers of space past each other, like pushing the top of a deck of cards sideways. One axis stays fixed while the other tilts.

✦ KEY TAKEAWAY

Think of a matrix as a recipe for rearranging space. A rotation recipe says "spin everything 45°." A reflection recipe says "flip across this mirror line." A projection recipe says "flatten onto this shadow line." A shear recipe says "slide each row sideways by an amount that depends on its height." The matrix is just the compact instruction card.

SECTION 3

Visual Explanation — Seeing the Four Transformations

The diagram below shows a unit square (with corners at the origin, (1, 0), (0, 1), and (1, 1)) and what happens to it under each of the four transformation types. Watch how the shape changes — and what stays the same — in each case.

The purple dashed outline shows the original unit square. Rotation spins the square around the origin. Reflection flips it across the y-axis. Projection collapses it onto the x-axis. Shear tilts it into a parallelogram.

Notice that rotation and reflection keep the square's area the same — they just move it. Projection crushes the 2-D square into a 1-D line segment, losing all height information. Shear changes the shape from a square to a parallelogram, but the area stays the same because the base and height haven't changed.

SECTION 4

Mathematical Framework — The Matrices

Every linear transformation in R² can be performed by multiplying a point's coordinates by a 2 × 2 matrix. If a point is at position (x, y), the new position (x', y') after the transformation is found by the rule: x' = ax + by and y' = cx + dy, where the matrix has entries a, b, c, d. Below are the key matrices for each transformation type.

Rotation by Angle θ (counterclockwise about the origin)

ROTATION MATRIX (R²)

R(θ) = [ cos θ , −sin θ ; sin θ , cos θ ]

θ is the angle of rotation measured counterclockwise from the positive x-axis. cos θ and sin θ are the cosine and sine of that angle. The matrix preserves lengths and angles.

Reflection Across the x-axis, y-axis, or y = x

REFLECTION MATRICES (R²)

Ref_x = [ 1, 0 ; 0, −1 ] Ref_y = [ −1, 0 ; 0, 1 ] Ref_{y=x} = [ 0, 1 ; 1, 0 ]

Ref_x flips over the x-axis (negates y). Ref_y flips over the y-axis (negates x). Ref_y=x swaps x and y coordinates.

Projection onto the x-axis or y-axis

PROJECTION MATRICES (R²)

Proj_x = [ 1, 0 ; 0, 0 ] Proj_y = [ 0, 0 ; 0, 1 ]

Proj_x drops the y-coordinate to zero, flattening everything onto the x-axis. Proj_y drops the x-coordinate to zero.

Horizontal and Vertical Shear

SHEAR MATRICES (R²)

Shear_h(k) = [ 1, k ; 0, 1 ] Shear_v(k) = [ 1, 0 ; k, 1 ]

k is the shear factor. In a horizontal shear, the x-coordinate shifts by k × y, so higher rows slide further right (when k > 0). In a vertical shear, the y-coordinate shifts by k × x.

🧊 Extending to R³

In three dimensions, these same ideas apply with 3 × 3 matrices. For example, a rotation about the z-axis in R³ uses the matrix [ cos θ, −sin θ, 0 ; sin θ, cos θ, 0 ; 0, 0, 1 ]. The extra row and column keep the z-coordinate unchanged. Reflections, projections, and shears extend to R³ in the same way — just add a third row and column.

SECTION 5

Detailed Breakdown — What Each Transformation Preserves

One of the best ways to tell transformations apart is to ask: what does this transformation keep the same? Some preserve distances, some preserve areas, and some preserve neither. The table below summarizes these properties.

Comparison of properties preserved by each transformation type

Transformation	Preserves Length?	Preserves Area?	Preserves Angles?	Invertible?
Rotation	Yes	Yes	Yes	Yes (rotate back)
Reflection	Yes	Yes	Yes	Yes (reflect again)
Projection	No	No (area → 0)	No	No (info lost)
Shear	No	Yes	No	Yes (shear back)

The determinant of the matrix tells you how the area changes. A determinant of 1 means area is preserved. A determinant of 0 means the shape collapses — and the transformation cannot be undone.

A key insight is the connection between the determinant and invertibility. If the determinant of a matrix is zero, the transformation squashes everything down to a lower dimension. Once that information is lost, you can never get it back — there is no "undo" matrix. Rotations, reflections, and shears all have nonzero determinants, so they are invertible.

SECTION 6

Worked Example — Rotating a Point 90° in R²

Let's rotate the point (3, 1) counterclockwise by 90° about the origin. We'll build the rotation matrix, multiply, and check that the result makes geometric sense.

Rotate (3, 1) by 90° Counterclockwise

Step 1 — Write the Rotation MatrixFor θ = 90°, we have cos 90° = 0 and sin 90° = 1. Plug these into the rotation matrix formula: R(90°) = [ 0, −1 ; 1, 0 ].

Step 2 — Set Up the MultiplicationWe multiply the matrix by the column vector [3 ; 1]: [x' ; y'] = [ 0, −1 ; 1, 0 ] × [3 ; 1].

Step 3 — Compute x'x' = (0)(3) + (−1)(1) = 0 − 1 = −1.

x' = −1

Step 4 — Compute y'y' = (1)(3) + (0)(1) = 3 + 0 = 3.

y' = 3

Step 5 — Interpret the ResultThe new point is (−1, 3). This makes sense: rotating 90° counterclockwise swaps the coordinates and negates the new x-coordinate, sending (3, 1) to (−1, 3). The distance from the origin is √(9 + 1) = √10 before and √(1 + 9) = √10 after — distance is preserved, as expected for a rotation.

Rotated point: (−1, 3)

SECTION 7

Comparing the Transformations — Strengths & Limitations

Each transformation type has strengths (things it's great at) and limitations (things it can't do). Understanding these helps you choose the right transformation for a given application. For instance, if you need to preserve the shape of an object while changing its orientation, you'd use a rotation or reflection. If you need to cast a 3-D model's shadow, you'd use a projection.

Transformation	Best Used For	Key Limitation
Rotation	Changing direction without distortion — spinning a game character, pointing a satellite, or modeling planetary orbits.	Cannot change the size or shape of an object. Also, rotating around a point other than the origin requires extra steps (translate, rotate, translate back).
Reflection	Creating mirror images — symmetry detection, flipping textures, and physics simulations of bouncing light.	Reverses orientation (a clockwise shape becomes counterclockwise), which can cause issues in 3-D models where surface normals matter.
Projection	Displaying 3-D scenes on a 2-D screen, computing shadows, and reducing dimensions in data science.	Irreversible — once you project, the lost dimension cannot be recovered. Applying it twice gives the same result as applying it once (it's idempotent).
Shear	Simulating italic text, wind effects on structures, or creating oblique drawing projections.	Distorts angles and lengths, making shapes look "slanted." Not suitable when you need to preserve the exact shape.

✦ KEY TAKEAWAY

Choosing a transformation is like choosing a tool from a toolbox. A rotation is a wrench that turns things without bending them. A reflection is a mirror. A projection is a flashlight casting a shadow. A shear is a push on one side of a stack. Knowing which tool to reach for depends on what you want to keep and what you're willing to change.

SECTION 8

Connection to Advanced Theory

The four transformations you've learned are building blocks for much bigger ideas. In advanced linear algebra and beyond, mathematicians combine and generalize these transformations in powerful ways. Here's a peek at where these concepts lead.

What You Learned	Where It Leads
Rotation matrices in R² and R³	Rotation groups (SO(2), SO(3)) used in physics to describe symmetries of particles and galaxies.
Reflections preserving length	Orthogonal matrices and the full orthogonal group O(n), used in signal processing and quantum mechanics.
Projections onto lines/planes	Orthogonal projection and least-squares regression — the foundation of data fitting and machine learning.
Shear matrices with det = 1	The special linear group SL(n) and volume-preserving flows in fluid dynamics.
Combining transformations (matrix products)	Eigenvalues and eigenvectors — directions that don't change during a transformation — central to all of advanced linear algebra.

One exciting idea is composition: you can combine two transformations by multiplying their matrices together. For example, reflecting across the x-axis and then rotating by 90° is the same as multiplying R(90°) × Ref_x. This lets you build complex animations and simulations from simple pieces.

SECTION 9

Practice Problems

PROBLEM 1 — CONCEPTUAL

A transformation takes the unit square and turns it into a parallelogram of the same area, but the angles have changed. Which type of transformation is this — rotation, reflection, projection, or shear? Explain your reasoning.

PROBLEM 2 — BASIC CALCULATION

Reflect the point (4, −2) across the y-axis using the appropriate reflection matrix. What is the new point?

PROBLEM 3 — INTERMEDIATE

Apply the horizontal shear matrix with k = 2 to the point (1, 3). Then verify that the transformation preserves area by computing the determinant of the shear matrix.

PROBLEM 4 — APPLIED

A video game designer wants to rotate a spaceship sprite by 180° about the origin. The nose of the spaceship is at point (5, 2). Where does the nose end up after the rotation? Could the designer achieve the same result with two reflections? If so, which two?

PROBLEM 5 — CRITICAL THINKING

If you project the point (a, b) onto the x-axis and then project the result onto the y-axis, what point do you get? Is the order of the two projections important? Explain why projections are not invertible by considering what happens to two different starting points.

SUMMARY

Lesson Summary

Linear transformations in R² and R³ can be described by matrix multiplication. A rotation spins points around the origin by an angle θ using the matrix [cos θ, −sin θ; sin θ, cos θ], preserving both lengths and angles. A reflection flips points across a line (or plane in R³), preserving lengths and angles but reversing orientation — its determinant is −1. A projection collapses points onto a lower-dimensional subspace (like a shadow), losing information so the determinant is 0 and the transformation is not invertible. A shear slides layers of space past each other, preserving area (determinant = 1) but distorting angles.

The determinant of the transformation matrix tells you the area scale factor: |det| = 1 means area is preserved, det = 0 means the shape collapses. Transformations can be composed by multiplying their matrices, building complex motions from simple ones. These ideas extend to R³ using 3 × 3 matrices and are the foundation of computer graphics, robotics, data science, and physics.