Linear Transformations as Matrix-Vector Products - Comprehensive Summary
Linear Transformation: A function T: \(\mathbb{R}^n \to \mathbb{R}^m\) that satisfies: T(u+v) = T(u) + T(v) and T(cu) = cT(u)
Matrix-Vector Product: The result of multiplying a matrix A by a vector x, denoted Ax
Standard Matrix: For any linear transformation T: \(\mathbb{R}^n \to \mathbb{R}^m\), there exists a unique matrix A such that T(x) = Ax
• Every linear transformation can be represented as a matrix-vector product
• The standard matrix is formed by applying T to standard basis vectors
• Matrix multiplication preserves linear combinations
• Composition of transformations corresponds to matrix multiplication
1. Find standard matrix: Apply T to standard basis vectors
2. Verify linearity: Check additivity and homogeneity
3. Compute transformation: Perform matrix-vector multiplication
4. Analyze properties: Determine injectivity, surjectivity, etc.
• Always verify that transformations are linear before using matrix representation
• The standard matrix has columns T(e₁), T(e₂), ..., T(eₙ)
• Matrix-vector products follow standard multiplication rules
• Use geometric intuition for transformations in R² and R³
Standard Matrix: For a linear transformation \(T: \mathbb{R}^n \to \mathbb{R}^m\), the standard matrix A is the \(m \times n\) matrix whose columns are the images of the standard basis vectors: \(A = [T(\vec{e_1}) \, T(\vec{e_2}) \, \ldots \, T(\vec{e_n})]\).
- Identify the standard basis vectors for the domain
- Apply the transformation to each standard basis vector
- Form the matrix with these transformed vectors as columns
- Verify that T(x) = Ax for any vector x
For \(\mathbb{R}^2\), the standard basis vectors are: \(\vec{e_1} = \begin{pmatrix} 1 \\ 0 \end{pmatrix}\) and \(\vec{e_2} = \begin{pmatrix} 0 \\ 1 \end{pmatrix}\).
\(T(\vec{e_1}) = T\begin{pmatrix} 1 \\ 0 \end{pmatrix} = \begin{pmatrix} 2(1) + 3(0) \\ 1 - 0 \end{pmatrix} = \begin{pmatrix} 2 \\ 1 \end{pmatrix}\)
\(T(\vec{e_2}) = T\begin{pmatrix} 0 \\ 1 \end{pmatrix} = \begin{pmatrix} 2(0) + 3(1) \\ 0 - 1 \end{pmatrix} = \begin{pmatrix} 3 \\ -1 \end{pmatrix}\)
The standard matrix A has columns \(T(\vec{e_1})\) and \(T(\vec{e_2})\):
\(A = \begin{pmatrix} 2 & 3 \\ 1 & -1 \end{pmatrix}\)
Check that \(T(\vec{x}) = A\vec{x}\) for any \(\vec{x} = \begin{pmatrix} x_1 \\ x_2 \end{pmatrix}\):
\(A\vec{x} = \begin{pmatrix} 2 & 3 \\ 1 & -1 \end{pmatrix}\begin{pmatrix} x_1 \\ x_2 \end{pmatrix} = \begin{pmatrix} 2x_1 + 3x_2 \\ x_1 - x_2 \end{pmatrix} = T(\vec{x})\). ✓
The standard matrix is \(A = \begin{pmatrix} 2 & 3 \\ 1 & -1 \end{pmatrix}\).
• Standard matrix theorem: Every linear transformation has a unique matrix representation
• Column method: Standard matrix columns are transformations of basis vectors
• Verification: Always check that T(x) = Ax for arbitrary x
Linear Transformation: A function T is linear if it satisfies both: (1) Additivity: \(T(\vec{u} + \vec{v}) = T(\vec{u}) + T(\vec{v})\), and (2) Homogeneity: \(T(c\vec{u}) = cT(\vec{u})\).
Let \(\vec{u} = \begin{pmatrix} 1 \\ 0 \end{pmatrix}\) and \(\vec{v} = \begin{pmatrix} 1 \\ 0 \end{pmatrix}\).
Compute \(T(\vec{u} + \vec{v})\):
\(T\begin{pmatrix} 1 \\ 0 \end{pmatrix} + \begin{pmatrix} 1 \\ 0 \end{pmatrix} = T\begin{pmatrix} 2 \\ 0 \end{pmatrix} = \begin{pmatrix} 2^2 \\ 0 \end{pmatrix} = \begin{pmatrix} 4 \\ 0 \end{pmatrix}\)
\(T(\vec{u}) = T\begin{pmatrix} 1 \\ 0 \end{pmatrix} = \begin{pmatrix} 1^2 \\ 0 \end{pmatrix} = \begin{pmatrix} 1 \\ 0 \end{pmatrix}\)
\(T(\vec{v}) = T\begin{pmatrix} 1 \\ 0 \end{pmatrix} = \begin{pmatrix} 1^2 \\ 0 \end{pmatrix} = \begin{pmatrix} 1 \\ 0 \end{pmatrix}\)
Therefore: \(T(\vec{u}) + T(\vec{v}) = \begin{pmatrix} 1 \\ 0 \end{pmatrix} + \begin{pmatrix} 1 \\ 0 \end{pmatrix} = \begin{pmatrix} 2 \\ 0 \end{pmatrix}\)
We have: \(T(\vec{u} + \vec{v}) = \begin{pmatrix} 4 \\ 0 \end{pmatrix} \neq \begin{pmatrix} 2 \\ 0 \end{pmatrix} = T(\vec{u}) + T(\vec{v})\)
Since additivity fails, T is not linear. ✓
Let \(\vec{u} = \begin{pmatrix} 1 \\ 0 \end{pmatrix}\) and \(c = 2\).
Compute \(T(c\vec{u})\):
\(T(2\begin{pmatrix} 1 \\ 0 \end{pmatrix}) = T\begin{pmatrix} 2 \\ 0 \end{pmatrix} = \begin{pmatrix} 2^2 \\ 0 \end{pmatrix} = \begin{pmatrix} 4 \\ 0 \end{pmatrix}\)
\(cT(\vec{u}) = 2T\begin{pmatrix} 1 \\ 0 \end{pmatrix} = 2\begin{pmatrix} 1^2 \\ 0 \end{pmatrix} = 2\begin{pmatrix} 1 \\ 0 \end{pmatrix} = \begin{pmatrix} 2 \\ 0 \end{pmatrix}\)
We have: \(T(c\vec{u}) = \begin{pmatrix} 4 \\ 0 \end{pmatrix} \neq \begin{pmatrix} 2 \\ 0 \end{pmatrix} = cT(\vec{u})\)
Homogeneity also fails. ✓
T is not a linear transformation because it fails both the additivity and homogeneity properties.
• Additivity test: \(T(\vec{u} + \vec{v}) = T(\vec{u}) + T(\vec{v})\) must hold
• Homogeneity test: \(T(c\vec{u}) = cT(\vec{u})\) must hold
• Counterexample: One violation is sufficient to prove non-linearity
Matrix-Vector Product: The product \(A\vec{x}\) is computed by taking linear combinations of the columns of A using the components of \(\vec{x}\) as weights: \(A\vec{x} = x_1\vec{a_1} + x_2\vec{a_2} + \ldots + x_n\vec{a_n}\).
\(\vec{a_1} = \begin{pmatrix} 1 \\ 0 \\ 5 \end{pmatrix}\), \(\vec{a_2} = \begin{pmatrix} 2 \\ 1 \\ 0 \end{pmatrix}\), \(\vec{a_3} = \begin{pmatrix} 3 \\ 4 \\ 1 \end{pmatrix}\)
\(A\vec{x} = x_1\vec{a_1} + x_2\vec{a_2} + x_3\vec{a_3}\)
\(= 2\begin{pmatrix} 1 \\ 0 \\ 5 \end{pmatrix} + (-1)\begin{pmatrix} 2 \\ 1 \\ 0 \end{pmatrix} + 3\begin{pmatrix} 3 \\ 4 \\ 1 \end{pmatrix}\)
\(= \begin{pmatrix} 2 \\ 0 \\ 10 \end{pmatrix} + \begin{pmatrix} -2 \\ -1 \\ 0 \end{pmatrix} + \begin{pmatrix} 9 \\ 12 \\ 3 \end{pmatrix}\)
\(= \begin{pmatrix} 2 - 2 + 9 \\ 0 - 1 + 12 \\ 10 + 0 + 3 \end{pmatrix} = \begin{pmatrix} 9 \\ 11 \\ 13 \end{pmatrix}\)
Row 1: \((1)(2) + (2)(-1) + (3)(3) = 2 - 2 + 9 = 9\)
Row 2: \((0)(2) + (1)(-1) + (4)(3) = 0 - 1 + 12 = 11\)
Row 3: \((5)(2) + (0)(-1) + (1)(3) = 10 + 0 + 3 = 13\)
Result: \(\begin{pmatrix} 9 \\ 11 \\ 13 \end{pmatrix}\). ✓
The matrix A represents a linear transformation \(T_A: \mathbb{R}^3 \to \mathbb{R}^3\) where \(T_A(\vec{x}) = A\vec{x}\).
This transformation maps the input vector \(\begin{pmatrix} 2 \\ -1 \\ 3 \end{pmatrix}\) to the output vector \(\begin{pmatrix} 9 \\ 11 \\ 13 \end{pmatrix}\).
Since A is a matrix, the transformation \(T_A\) is linear, satisfying:
- Additivity: \(T_A(\vec{u} + \vec{v}) = T_A(\vec{u}) + T_A(\vec{v})\)
- Homogeneity: \(T_A(c\vec{u}) = cT_A(\vec{u})\)
The matrix-vector product is \(A\vec{x} = \begin{pmatrix} 9 \\ 11 \\ 13 \end{pmatrix}\). This represents the linear transformation of \(\vec{x}\) by matrix A.
• Column interpretation: Matrix-vector product is linear combination of columns
• Row-by-column: Alternative method using dot products
• Linear transformation: Every matrix represents a linear transformation
Composition of Transformations: The composition \((T_2 \circ T_1)(\vec{x}) = T_2(T_1(\vec{x}))\) is the transformation obtained by applying \(T_1\) first, then \(T_2\). The standard matrix of the composition is the product of the individual standard matrices.
Apply \(T_1\) to standard basis vectors:
\(T_1\begin{pmatrix} 1 \\ 0 \end{pmatrix} = \begin{pmatrix} 1 \\ 2 \end{pmatrix}\), \(T_1\begin{pmatrix} 0 \\ 1 \end{pmatrix} = \begin{pmatrix} 1 \\ -1 \end{pmatrix}\)
So \(A_1 = \begin{pmatrix} 1 & 1 \\ 2 & -1 \end{pmatrix}\)
Apply \(T_2\) to standard basis vectors:
\(T_2\begin{pmatrix} 1 \\ 0 \end{pmatrix} = \begin{pmatrix} 3 \\ 1 \end{pmatrix}\), \(T_2\begin{pmatrix} 0 \\ 1 \end{pmatrix} = \begin{pmatrix} 0 \\ 2 \end{pmatrix}\)
So \(A_2 = \begin{pmatrix} 3 & 0 \\ 1 & 2 \end{pmatrix}\)
The standard matrix for \(T_2 \circ T_1\) is \(A_2A_1\):
\(A_2A_1 = \begin{pmatrix} 3 & 0 \\ 1 & 2 \end{pmatrix}\begin{pmatrix} 1 & 1 \\ 2 & -1 \end{pmatrix}\)
Row 1, Col 1: \((3)(1) + (0)(2) = 3\)
Row 1, Col 2: \((3)(1) + (0)(-1) = 3\)
Row 2, Col 1: \((1)(1) + (2)(2) = 5\)
Row 2, Col 2: \((1)(1) + (2)(-1) = -1\)
Therefore: \(A_2A_1 = \begin{pmatrix} 3 & 3 \\ 5 & -1 \end{pmatrix}\)
Let's verify with \(\vec{x} = \begin{pmatrix} 1 \\ 1 \end{pmatrix}\):
\(T_1\begin{pmatrix} 1 \\ 1 \end{pmatrix} = \begin{pmatrix} 1+1 \\ 2-1 \end{pmatrix} = \begin{pmatrix} 2 \\ 1 \end{pmatrix}\)
\(T_2\begin{pmatrix} 2 \\ 1 \end{pmatrix} = \begin{pmatrix} 3(2) \\ 2 + 2(1) \end{pmatrix} = \begin{pmatrix} 6 \\ 4 \end{pmatrix}\)
\(A_2A_1\begin{pmatrix} 1 \\ 1 \end{pmatrix} = \begin{pmatrix} 3 & 3 \\ 5 & -1 \end{pmatrix}\begin{pmatrix} 1 \\ 1 \end{pmatrix} = \begin{pmatrix} 3+3 \\ 5-1 \end{pmatrix} = \begin{pmatrix} 6 \\ 4 \end{pmatrix}\)
Results match! ✓
The standard matrix for the composition \(T_2 \circ T_1\) is \(\begin{pmatrix} 3 & 3 \\ 5 & -1 \end{pmatrix}\).
• Composition property: \((T_2 \circ T_1)(\vec{x}) = T_2(T_1(\vec{x}))\)
• Matrix multiplication: Standard matrix of composition is product of individual matrices
• Order matters: \(T_2 \circ T_1\) corresponds to \(A_2A_1\), not \(A_1A_2\)
Kernel (Null Space): \(\ker(T) = \{\vec{x} \in \mathbb{R}^n \mid T(\vec{x}) = \vec{0}\}\). This is the set of all vectors that T maps to the zero vector.
Range (Image): \(\text{range}(T) = \{T(\vec{x}) \mid \vec{x} \in \mathbb{R}^n\}\). This is the set of all possible outputs of T.
Apply T to standard basis vectors:
\(T\begin{pmatrix} 1 \\ 0 \\ 0 \end{pmatrix} = \begin{pmatrix} 1 \\ 3 \end{pmatrix}\), \(T\begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix} = \begin{pmatrix} 2 \\ -1 \end{pmatrix}\), \(T\begin{pmatrix} 0 \\ 0 \\ 1 \end{pmatrix} = \begin{pmatrix} -1 \\ 4 \end{pmatrix}\)
So \(A = \begin{pmatrix} 1 & 2 & -1 \\ 3 & -1 & 4 \end{pmatrix}\)
Solve \(A\vec{x} = \vec{0}\): \(\begin{pmatrix} 1 & 2 & -1 \\ 3 & -1 & 4 \end{pmatrix}\begin{pmatrix} x_1 \\ x_2 \\ x_3 \end{pmatrix} = \begin{pmatrix} 0 \\ 0 \end{pmatrix}\)
This gives the system: \(x_1 + 2x_2 - x_3 = 0\) and \(3x_1 - x_2 + 4x_3 = 0\)
Row reduce the augmented matrix: \(\begin{pmatrix} 1 & 2 & -1 & | & 0 \\ 3 & -1 & 4 & | & 0 \end{pmatrix}\)
\(R_2 \leftarrow R_2 - 3R_1\): \(\begin{pmatrix} 1 & 2 & -1 & | & 0 \\ 0 & -7 & 7 & | & 0 \end{pmatrix}\)
\(R_2 \leftarrow -\frac{1}{7}R_2\): \(\begin{pmatrix} 1 & 2 & -1 & | & 0 \\ 0 & 1 & -1 & | & 0 \end{pmatrix}\)
\(R_1 \leftarrow R_1 - 2R_2\): \(\begin{pmatrix} 1 & 0 & 1 & | & 0 \\ 0 & 1 & -1 & | & 0 \end{pmatrix}\)
So: \(x_1 + x_3 = 0\) and \(x_2 - x_3 = 0\), which gives \(x_1 = -x_3\) and \(x_2 = x_3\)
Let \(x_3 = t\), then: \(\vec{x} = \begin{pmatrix} -t \\ t \\ t \end{pmatrix} = t\begin{pmatrix} -1 \\ 1 \\ 1 \end{pmatrix}\)
Therefore: \(\ker(T) = \text{span}\left\{\begin{pmatrix} -1 \\ 1 \\ 1 \end{pmatrix}\right\}\)
The range of T is the column space of A. Since A is \(2 \times 3\), the range is a subspace of \(\mathbb{R}^2\).
From the row reduction above, the first two columns of A are linearly independent (they form a basis for the column space).
So \(\text{range}(T) = \text{span}\left\{\begin{pmatrix} 1 \\ 3 \end{pmatrix}, \begin{pmatrix} 2 \\ -1 \end{pmatrix}\right\}\).
Since these two vectors are linearly independent in \(\mathbb{R}^2\), they span all of \(\mathbb{R}^2\).
Therefore: \(\text{range}(T) = \mathbb{R}^2\).
\(\dim(\ker(T)) + \dim(\text{range}(T)) = 1 + 2 = 3 = \dim(\text{domain})\). ✓
The kernel of T is \(\text{span}\left\{\begin{pmatrix} -1 \\ 1 \\ 1 \end{pmatrix}\right\}\) and the range of T is \(\mathbb{R}^2\).
• Kernel computation: Solve homogeneous system \(A\vec{x} = \vec{0}\)
• Range computation: Find column space of the standard matrix
• Rank-Nullity theorem: Verify that dimensions add to domain dimension
Linear Transformation: A function that preserves vector addition and scalar multiplication, always representable as matrix-vector multiplication.
Standard Matrix: The unique matrix that represents a linear transformation with respect to standard bases.
Kernel: The null space of the transformation, representing all vectors mapped to zero.
Range: The column space of the transformation, representing all possible outputs.
- For standard matrix: Apply transformation to standard basis vectors
- For kernel: Solve homogeneous system Ax = 0
- For range: Find column space of standard matrix
- For composition: Multiply corresponding matrices in reverse order
• Every linear transformation has a unique matrix representation
• Composition of transformations corresponds to matrix multiplication
• The kernel of T equals the null space of its standard matrix
• The range of T equals the column space of its standard matrix
• Injective ⟺ kernel contains only zero vector
Geometric Interpretation: Linear transformations preserve straight lines and ratios of distances, mapping the unit square to a parallelogram.
- Original unit square: vertices (0,0), (1,0), (1,1), (0,1)
- Transformed vertices: (0,0), (2,1), (3,0), (1,-1)
- Linear transformations map parallelograms to parallelograms
- The transformation matrix is \(\begin{pmatrix} 2 & 1 \\ 1 & -1 \end{pmatrix}\)