Transformations using Matrices

• Linear transformation in the plane maps a point with position vector $\begin{pmatrix}x \\ y\end{pmatrix}$ to a new point $\begin{pmatrix}x' \\ y'\end{pmatrix}$ by multiplication with a $2 \times 2$ matrix $M$. This works because the transformation acts consistently on all vectors through the rule $\begin{pmatrix}x' \\ y'\end{pmatrix}=M\begin{pmatrix}x \\ y\end{pmatrix}$, so every point is determined by the same matrix. It applies to transformations centred at the origin such as rotations, reflections, enlargements, and stretches.

• Object and image are the original point or shape and the transformed result. This language matters because exam questions often ask you to move from object to image or to reverse the process from image back to object. Being precise about which coordinates are before and after the transformation prevents direction errors.

• Column vectors are used because a matrix transformation acts on vectors, not directly on ordered pairs written horizontally. Writing a point as $\begin{pmatrix}x \\ y\end{pmatrix}$ allows the matrix to combine the $x$- and $y$-components according to its entries. This gives the transformed coordinates through standard matrix multiplication.

• Basis-vector viewpoint is central: the first column of the matrix is the image of $\begin{pmatrix}1 \\ 0\end{pmatrix}$ and the second column is the image of $\begin{pmatrix}0 \\ 1\end{pmatrix}$. This is why many standard transformation matrices can be built by seeing where these unit vectors move. It is also why the columns encode the geometry of the whole transformation.

• Transforming a shape can be done by transforming each vertex separately, or by placing all vertex vectors into a single vertex matrix and multiplying once. Straight lines remain straight under linear transformations because matrix multiplication preserves linear combinations. However, the order of vertices may reverse if orientation changes, which is linked to the sign of the determinant.

• Matrix multiplication gives the new coordinates because each transformed vector is a linear combination of the transformed basis vectors. For $M=\begin{pmatrix}a & b \\ c & d\end{pmatrix}$, the image of $\begin{pmatrix}x \\ y\end{pmatrix}$ is $\begin{pmatrix}x' \\ y'\end{pmatrix}=\begin{pmatrix}a & b \\ c & d\end{pmatrix}\begin{pmatrix}x \\ y\end{pmatrix}=\begin{pmatrix}ax+by \\ cx+dy\end{pmatrix}.$ This shows exactly how the entries of the matrix control the mixing and scaling of coordinates.

• The determinant of a $2 \times 2$ matrix, $\det(M)=ad-bc,$ measures how areas change under the transformation. The magnitude $|\det(M)|$ is the area scale factor, so if $|\det(M)|=k$, every region has its area multiplied by $k$. This gives a powerful link between algebra and geometry.

• The sign of the determinant tells you whether orientation is preserved or reversed. A positive determinant keeps clockwise and anticlockwise order the same, while a negative determinant reverses the sense of the vertices. This is especially useful when distinguishing reflections from rotations, since reflections reverse orientation but rotations do not.

• Invertibility depends on the determinant because a transformation can be undone only if no information has been collapsed. If $\det(M) \neq 0$, the inverse exists and each image point comes from exactly one original point. If $\det(M)=0$, the transformation squashes the plane into a line or a point, so reversal is impossible.

• Standard transformation matrices arise from geometric rules about how unit directions move. Reflections swap or negate coordinate directions, stretches scale one direction more than the other, and rotations preserve lengths while changing direction. Understanding these geometric effects is more reliable than memorising isolated matrices.

Transforming a single point

• Write the point as a column vector and multiply by the matrix on the left. For $M=\begin{pmatrix}a & b \\ c & d\end{pmatrix},\quad \mathbf{p}=\begin{pmatrix}x \\ y\end{pmatrix},$ compute $M\mathbf{p}$ carefully entry by entry. This method is the default whenever you are asked for the image of a point under a given transformation.
• Interpret the result as coordinates of the image point. If the product is $\begin{pmatrix}u \\ v\end{pmatrix}$, then the transformed point is $(u,v)$. This small final conversion is simple, but many students lose clarity by leaving answers as vectors when coordinates are expected.

Transforming a set of vertices

• Transform each vertex separately if the shape has only a few points. This is reliable because each point follows the same matrix rule, and then the image shape is formed by joining corresponding image vertices in order. It is especially helpful when you want to track orientation or identify which vertex came from which original point.
• Use a vertex matrix when there are many points: place each vertex as a column, then multiply once by the transformation matrix. This is efficient because the resulting columns are the transformed vertices in matching order. It reduces repeated arithmetic and makes pattern checking easier.

Reversing a transformation

• Use the inverse matrix if you know the image and need the original object. If $M\mathbf{p}=\mathbf{p'}$, then $\mathbf{p}=M^{-1}\mathbf{p'}.$ This works only when $\det(M) \neq 0$, so checking invertibility is part of the method.
• Find the inverse of a $2 \times 2$ matrix using $M^{-1}=\frac{1}{ad-bc}\begin{pmatrix}d & -b \\ -c & a\end{pmatrix},$ where $M=\begin{pmatrix}a & b \\ c & d\end{pmatrix}$. The determinant in the denominator rescales the adjugate matrix so that multiplying by $M$ returns the identity. This is the standard algebraic tool when the inverse is not obvious from geometry.

Combining transformations

• Multiply matrices in the correct order. If a point is transformed first by $P$ and then by $Q$, the single matrix is $M=QP.$ The first transformation appears on the right because it acts on the vector first, and this reversed order is one of the most examined ideas in the topic.
• Reverse both order and direction when inverting a combination. If $M=QP$, then $M^{-1}=(QP)^{-1}=P^{-1}Q^{-1}.$ This reflects the logical idea that to undo two actions, you must undo the second one first and then undo the first.

Recognising standard transformations

• Reflections, rotations, stretches, and enlargements may all be represented by $2 \times 2$ matrices, but they have different geometric signatures. Reflections reverse orientation, rotations preserve lengths and orientation, stretches change scale differently in different directions, and enlargements scale equally in both coordinate directions. Distinguishing these quickly is essential for describing a matrix transformation fully.
• | Feature | Reflection | Rotation | Stretch | Enlargement | | --- | --- | --- | --- | --- | | Lengths preserved? | Yes | Yes | Not usually | Not unless scale factor is $1$ | | Angles preserved? | Yes | Yes | Not usually | Yes | | Orientation preserved? | No | Yes | Usually yes if scale factors positive | Yes if factor positive | | Determinant | Negative for standard reflections | Positive, usually $1$ | Product of axis scale factors | $k^2$ for scale factor $k$ | | Typical effect | Mirror image | Turn about origin | One-directional scaling | Uniform scaling | This comparison helps you interpret a matrix from its form and determinant, rather than relying only on memory.

Reflection versus rotation when matrices look similar

• Some matrices can appear deceptive, especially those containing zeros and negative ones. For example, a matrix may look like a sign change or a swap of coordinates, so the safest method is to see where $\begin{pmatrix}1 \\ 0\end{pmatrix}$ and $\begin{pmatrix}0 \\ 1\end{pmatrix}$ go. Their images reveal whether the transformation is a mirror action or a turn.

Stretch versus enlargement

• An enlargement scales equally in all directions from the origin, so its matrix is $kI=\begin{pmatrix}k & 0 \\ 0 & k\end{pmatrix}$. A stretch scales one axis direction differently from the other, such as $\begin{pmatrix}a & 0 \\ 0 & 1\end{pmatrix}$ or $\begin{pmatrix}1 & 0 \\ 0 & b\end{pmatrix}$. This distinction matters because a combined stretch only becomes an enlargement when the scale factors are equal.

Single transformation versus combined transformation

• A single matrix may describe one standard transformation or the net effect of several transformations. If the question asks for a description, decide whether the matrix directly matches a standard form or whether you must interpret its action on basis vectors. In exam settings, this distinction prevents you from overcomplicating simple matrices or oversimplifying composite ones.

• Always state a transformation fully by giving the type and all required details. For example, a rotation needs angle, direction, and centre; a reflection needs the mirror line; an enlargement needs scale factor and centre; and a stretch needs scale factor and direction. Examiners often withhold marks when one of these descriptors is missing even if the general idea is correct.

• Check order carefully in composite transformations because the rightmost matrix acts first. A reliable habit is to write a vector on the far right and then read the operations from right to left. This avoids the common mistake of multiplying in the order the transformations are spoken rather than the order matrices must act.

• Use the determinant as a fast diagnostic tool. If $|\det(M)|=1$, area is preserved; if $\det(M)<0$, orientation is reversed; if $\det(M)=0$, the transformation is not invertible. These quick checks help you judge whether a proposed description is even possible.

• Test basis vectors when identifying an unknown matrix. Compute the images of $\begin{pmatrix}1 \\ 0\end{pmatrix}$ and $\begin{pmatrix}0 \\ 1\end{pmatrix}$ or read them directly from the columns. This is often faster and less error-prone than trying to guess from the matrix entries alone.

• Be careful with modulus in area questions. Since area scale factor is $|\det(M)|$, algebraic equations involving scale factor may lead to more than one solution when a parameter is involved. Ignoring the absolute value can lose valid answers or produce the wrong sign restriction.

• Confusing a rotation of $180^\circ$ with a reflection is a frequent error because both can involve negative coordinates. A $180^\circ$ rotation sends every vector to its opposite and preserves orientation, while a reflection flips across a line and reverses orientation. Checking the determinant sign immediately separates these cases.

• Using the wrong multiplication order is one of the most damaging mistakes in composite transformations. Since matrix multiplication is not commutative, $PQ$ and $QP$ usually represent different transformations. Even if both products exist, only one matches the stated order of actions.

• Assuming every matrix has an inverse is incorrect. If the determinant is zero, the transformation collapses dimension and cannot be undone uniquely. Students often try to use an inverse formula automatically without first checking this condition.

• Describing stretches informally as moving left, right, up, or down misses the geometric meaning. A stretch should be described as parallel to the $x$-axis or parallel to the $y$-axis, because it scales displacement in one coordinate direction. Precise wording reflects precise understanding.

• Forgetting that area scale factor uses the absolute value of the determinant leads to sign errors. A negative determinant does not mean negative area; it means reversed orientation. The area itself is always scaled by a non-negative factor.

• Transformations using matrices connect algebra, geometry, and vector spaces. The same matrix that changes coordinates on a graph also acts as a linear map in abstract linear algebra. This makes the topic a foundation for later work on eigenvalues, vector spaces, and change of basis.

• Determinants link local coordinate action to global geometric effect. In two dimensions they measure area scaling, and in higher dimensions they measure volume scaling. This idea reappears in calculus, differential equations, computer graphics, and physics.

• Rotation matrices generalise naturally through trigonometry. The standard formula $\begin{pmatrix}\cos\theta & -\sin\theta \\ \sin\theta & \cos\theta\end{pmatrix}$ shows how geometry and circular functions interact, making this topic an important bridge between pure algebra and trigonometric modelling. It also explains why exact angle values and sign conventions matter.

• Composite and inverse transformations model real systems where actions occur in sequence and then must be undone. This perspective appears in robotics, animation, image processing, and coordinate changes in mechanics. Learning to reason about order and reversal in matrices has strong transfer value beyond exam questions.