Linear Independence & Dependence Advanced Revision Notes

Contextual Definition: Let $V$ be a vector space defined over a field $F$ . Consider a finite subset of vectors $S = \{v_1, v_2, \dots, v_n\}$ in $V$ .
The Homogeneous Linear Combination Equation: The foundational check for independence involves analyzing the solutions to the equation: $c_1v_1 + c_2v_2 + \dots + c_nv_n = 0$ . In this equation:
- $c_i \in F$ represent scalars.
- $0$ represents the zero vector within the vector space $V$ .
Linear Independence: A set $S$ is classified as linearly independent if and only if the trivial solution is the unique solution to the homogeneous equation. This requires that $c_1 = c_2 = \dots = c_n = 0$ . No single vector within the set can be expressed as a linear combination of the remaining vectors.
Linear Dependence: A set $S$ is classified as linearly dependent if there exists at least one non-trivial solution to the homogeneous equation. This means there are scalars $c_1, c_2, \dots, c_n$ , not all of which are zero, such that the equation holds true.
Infinite Sets: In the context of infinite-dimensional spaces, such as polynomial spaces or the space of continuous functions on an interval $C[a, b]$ , an infinite set of vectors is considered linearly independent if and only if every finite subset of that set is linearly independent.

Conceptual Visualization: Vectors should be viewed as spatial directions. The span of a set of vectors represents the entire subspace that can be reached by traveling solely along those directions.
Expanding the Space (Independence): Linearly independent vectors are additive in terms of dimensionality:
- 1 Vector: A single non-zero vector ( $v_1 \neq 0$ ) defines a 1D line.
- 2 Vectors: Two vectors ( $v_1, v_2$ ) define a 2D plane, provided that $v_2$ does not lie on the line defined by $v_1$ .
- 3 Vectors: Three vectors ( $v_1, v_2, v_3$ ) define a 3D volume, provided that $v_3$ does not lie within the plane defined by $v_1$ and $v_2$ .
Collapsing Volume (Dependence): Linearly dependent sets contain redundant directional information. In the context of $\mathbb{R}^n$ , if a set of $n$ vectors is dependent, the $n$ -dimensional volume (parallelepiped) they would otherwise form "collapses" to exactly zero. They are constrained within a lower-dimensional subspace.

The Pigeonhole Principle of Vector Spaces: In any vector space with dimension $n$ , any subset that contains more than $n$ vectors ( $m > n$ ) is strictly linearly dependent.
Subsets and Supersets Rule:
- Any subset of a linearly independent set is also linearly independent.
- Any superset of a linearly dependent set is also linearly dependent.
The Zero Vector Rule: Any set that contains the zero vector ( $0$ ) is unconditionally and instantly linearly dependent.

Matrix Construction: When vectors in $\mathbb{R}^n$ are arranged as the columns (or rows) of an $n \times n$ matrix $A$ , several tests apply:
The Determinant Test:
- If $\text{det}(A) \neq 0$ , the vectors are linearly independent.
- If $\text{det}(A) = 0$ , the spatial volume has collapsed, indicating the vectors are linearly dependent.
The Eigenvalue Test: Because the determinant is the product of the eigenvalues ( $\text{det}(A) = \lambda_1 \times \lambda_2 \times \dots \times \lambda_n$ ), a set is linearly dependent if and only if at least one eigenvalue is exactly zero.
The Rank Test: For an $m \times n$ matrix, the number of linearly independent rows or columns is equal to the rank of the matrix. Linear dependence exists if $\text{Rank}(A) < \min(m, n)$ .

Definition of the Wronskian: When analyzing function spaces (common in differential equations), the Wronskian $W(x)$ is utilized.
Independence Condition: If $W(x) \neq 0$ at even a single point within the specified interval, the functions are linearly independent.
The High-Yield Trap: Caution must be exercised if $W(x) = 0$ everywhere. This result does not universally guarantee that the functions are linearly dependent.
The Exception: If the functions are specifically known to be solutions to an $n$ -th order linear homogeneous Ordinary Differential Equation (ODE), then $W(x) = 0$ strictly implies linear dependence.

The Proportionality Check: Determine if one vector is a direct scalar multiple of another. For example, if $v_1 = (1, -2, 3)$ and $v_2 = (-2, 4, -6)$ , then $v_2 = -2v_1$ , confirming dependence.
The Hidden Sum Check: Check if adding or subtracting two vectors results in a third vector. For example, if $v_1 = (1, 0, 1)$ , $v_2 = (0, 1, 1)$ , and $v_3 = (1, 1, 2)$ , then $v_1 + v_2 = v_3$ , confirming dependence.
The Zero Vector Check: If the set contains a zero vector or if a matrix contains a row or column of all zeros, it is instantly dependent.
Dimension Mismatch: If you are presented with 4 vectors in $\mathbb{R}^3$ , they are linearly dependent by definition without any need for calculation.

Problem Scenario: Find a scalar parameter $\lambda$ that makes the set $S = \{(1, 1, 2), (1, 2, 1), (2, 3, \lambda)\}$ linearly dependent.
Standard Categorical Method (Slow): This involves setting the determinant of the matrix to zero:
- $\text{det} \begin{pmatrix} 1 & 1 & 2 \\ 1 & 2 & 1 \\ 2 & 3 & \lambda \end{pmatrix} = 0$
- Solving this equation leads to $\lambda = 3$ .
Rapid Visual Observation Method (Fast): Let $v_1 = (1, 1, 2)$ , $v_2 = (1, 2, 1)$ , and $v_3 = (2, 3, \lambda)$ . Compare the components of the first two dimensions:
- x-components: $1 + 1 = 2$ (matches the x-component of $v_3$ ).
- y-components: $1 + 2 = 3$ (matches the y-component of $v_3$ ).
Conclusion: For the set to be dependent, $v_3$ must exist in the span of $v_1$ and $v_2$ . Given the patterns in the first two dimensions, the most likely relationship is $v_3 = v_1 + v_2$ . Applying this to the z-component: $2 + 1 = \lambda$ , therefore $\lambda = 3$ . This pattern recognition is a crucial time-saving skill for competitive exams like GATE or CSIR NET.