LQ3 Pointers

Tags: #◇CSMATH2 #◇Linear Algebra #◇Computational Mathematics #◇Matrices #◇Matrix Operations #◇Computer Science #◇Vector Spaces #◇Linear Transformations

---

◇Overview

Linear Combination: A mix of vectors using scaling and addition.

◇Key Concepts

Linear Combination Span Definition Visual Check Basis

◇Details / Theory

◇Part 1: Vector Space Concepts (The Building Blocks)

1. Linear Combination & Span

Linear Combination: A mix of vectors using scaling and addition.
Formula: $v = c_1v_1 + c_2v_2 + \dots + c_nv_n$ Span: The collection of ALL possible vectors you can create from a given set.
Concept: Think of the span as the "territory" covered by the vectors.
Rule: If you have fewer vectors than the dimension of the space (e.g., 2 vectors in $\mathbb{R}^3$), you cannot span the whole space.

2. Linear Independence

Definition: A set of vectors is independent if no vector is "redundant." You cannot build one vector using the others. The Zero Test: To check independence, set a combination to zero:
$c_1v_1 + c_2v_2 + \dots = 0$
If the only solution is that all scalars are zero ($c=0$), they are Linearly Independent.
If you find any non-zero scalars, they are Linearly Dependent. Visual Check: Two vectors are dependent if they are on the same line (one is a multiple of the other).

3. Basis & Dimension

Basis: A "perfect" set of vectors that is efficient (independent) and complete (spans the space).
To be a basis, a set must be: (1) Linearly Independent AND (2) A Spanning Set. Dimension: The number of vectors in a basis.
Rule: Every basis for the same vector space has the exact same number of vectors.
Examples: $\mathbb{R}^2$ has dimension 2; $\mathbb{R}^3$ has dimension 3.

---

◇Part 2: Linear Transformations (The Action)

1. What is a Linear Transformation?

A function $L$ that keeps the grid structure intact. It must follow the "Superposition Principle":

$L(ru + sv) = rL(u) + sL(v)$
Checklist for Linearity:

1. Zero maps to Zero: $L(0)$ must equal 0.
2. No Powers: No $x^2, x^3$, etc.
3. No Constants: No $+5$ or $-1$ added at the end.

2. Kernel (The "Null Space")

Definition: The set of all inputs that get sent to the Zero Vector. How to find it: Set the transformation equation equal to 0 and solve for the variables.
$L(x) = 0$ Significance: The Kernel tells you about uniqueness.

3. Range (The "Image")

Definition: The set of all possible outputs. How to find it: It is the Span of the transformed basis vectors.
Tip: If $L(x,y) = (x+y, x-y, 0)$, separate the variables to find the vectors: $x(1,1,0) + y(1,-1,0)$. The range is the span of those two column vectors.

4. One-to-One vs. Onto

One-to-One (Injective): Every output has a unique input.
The Test: A map is One-to-One if and only if the Kernel contains ONLY the zero vector ($\{0\}$). Onto (Surjective): The map covers the entire target space.
The Test: The dimension of the Range must equal the dimension of the target space.

5. Rank-Nullity Theorem

A fundamental rule connecting the dimensions:

$\text{Dim(Kernel)} + \text{Dim(Range)} = \text{Dim(Domain)}$
Translation: (Dimensions lost) + (Dimensions kept) = (Total dimensions started with). hi, hang in there bro :>

---

◇Part 3: Problem Solving Strategies ("The Algorithms")

A. How to Check for Linear Independence / Find a Basis

1. Matrix Setup: Put your vectors as the columns of a matrix.
2. Row Reduce: Perform Gaussian Elimination to get to Row Echelon Form (REF).
3. Find Pivots: Look for the leading non-zero numbers (usually 1s) in the rows.
4. Conclusion:
   The columns with pivots correspond to the Independent Vectors (your Basis).
   The columns without pivots (free variables) correspond to Dependent Vectors.

B. How to Extend a Set to a Basis

1. Start with your given independent vectors.
2. Add a "Standard Basis" vector (like $(1,0,0)$ or $(0,1,0)$) that is not a linear combination of your current set.
3. Verify: Put them in a matrix and check if the determinant is non-zero, or if you have a full set of pivots.

C. How to Compute Transformations of New Vectors

Problem: Given $L(1,0) = u$ and $L(0,1) = v$, find $L(3, 5)$. Solution:

1. Write the target vector as a combination: $(3, 5) = 3(1,0) + 5(0,1)$.
2. Apply linearity: $L(3, 5) = 3L(1,0) + 5L(0,1)$.
3. Substitute and calculate: $3u + 5v$.

---

◇Part 4: Quick Tips & Common Pitfalls

The "Goldilocks" Rule:
Too few vectors? Can't span.
Too many vectors? Must be dependent.
Just right ($n$ vectors in dimension $n$)? If they are independent, they are automatically a basis. Zero Vector Trap: Any set containing the Zero Vector is automatically Dependent. Rank vs. Basis:
Rank is a number (how many pivots).
Basis is a set of vectors (the columns with pivots). Kernel Dimensions: The Kernel lives in the Domain (input space). The Range lives in the Codomain (output space). Don't mix them up!

---

◇Part 5: "How To" Step-by-Step Guides

1. How to show a vector $u$ is a Linear Combination of set $S$

Scenario: You are given a vector $u$ and a set of vectors $S = \{v_1, v_2\}$. The question asks: "Is $u$ a linear combination of $S$?" or "Express $u$ as a linear combination of $S$."
Step 1: Write the Equation
Set up the equation $c_1v_1 + c_2v_2 + \dots = u$.
Example: $c_1 \begin{bmatrix}1\\2\end{bmatrix} + c_2 \begin{bmatrix}3\\4\end{bmatrix} = \begin{bmatrix}11\\16\end{bmatrix}$. Step 2: Create the Augmented Matrix
Turn the vectors into columns of a matrix, with the target vector $u$ as the final column on the right side.
$\begin{bmatrix} 1 & 3 & | & 11 \\ 2 & 4 & | & 16 \end{bmatrix}$ Step 3: Row Reduce (REF)
Use Gaussian Elimination to get zeros in the bottom-left corner. Step 4: Check Consistency
Consistent: If you get a valid answer (e.g., $c_2 = 2, c_1 = 5$), then YES, it is a linear combination.
Inconsistent: If you get a row like $\begin{bmatrix} 0 & 0 & | & 5 \end{bmatrix}$ (which means $0 = 5$), then NO, it is impossible.

---

2. How to determine if a Set is Linearly Independent

Scenario: You are given a set of vectors $S = \{v_1, v_2, v_3\}$. Is it linearly independent?
Step 1: Form the Homogeneous Equation
Set the linear combination equal to the zero vector: $c_1v_1 + c_2v_2 + c_3v_3 = 0$. Step 2: Create the Augmented Matrix
Put vectors as columns. The augmented side is all zeros. Step 3: Row Reduce (REF)
Solve the matrix. Step 4: Analyze the Solution
Independent: If the only solution is $c_1=0, c_2=0, c_3=0$ (the trivial solution), then the set is Independent.
Dependent: If there are free variables (non-trivial solutions), the set is Dependent.

---

3. How to find a Basis from a Spanning Set

Scenario: "Find a basis for the subspace spanned by $S = \{v_1, v_2, v_3, v_4\}$."
Step 1: Put Vectors in a Matrix
List the vectors as the columns of a matrix. Step 2: Row Reduce to REF
Reduce until you see the "steps" or pivots. Step 3: Identify Pivot Columns
Locate the columns that have the leading non-zero numbers (pivots). Step 4: Select Original Vectors
The basis consists of the vectors from the original set that correspond to the pivot columns. (Do not use the columns from the reduced matrix itself).

---

4. How to prove $L$ is a Linear Transformation

Scenario: "Let $L(x) = \dots$. Determine if $L$ is a linear transformation."
Step 1: Check the Combined Property
You must prove that $L(ru + sv) = rL(u) + sL(v)$ is true for all vectors $u, v$ and scalars $r, s$. Step 2: Substitute and Expand
Calculate the left side: Plug $(ru + sv)$ into the function.
Calculate the right side: Multiply $r$ by the function $L(u)$ and add $s$ times the function $L(v)$. Step 3: Compare
If Left Side = Right Side, it is a Linear Transformation.
Shortcut: If you see terms like $x^2$, $|x|$, $sin(x)$, or constants ($+1$), it is NOT linear.

---

5. How to find the Kernel of a Linear Transformation

Scenario: "Find the Kernel of $L(x, y) = (x+y, x-2y)$."
Step 1: Set Output to Zero
Write the equation $L(u) = 0_W$ (the zero vector of the target space).
$\begin{bmatrix} x+y \\ x-2y \end{bmatrix} = \begin{bmatrix} 0 \\ 0 \end{bmatrix}$ Step 2: Solve the System
Set up the equations:

1. $x + y = 0$
2. $x - 2y = 0$ Step 3: State the Kernel
   If the only answer is $x=0, y=0$, the Kernel is $\{0\}$ (and $L$ is one-to-one).
   If you get a free variable (like $x = -z$), describe the set of vectors (e.g., $span\{(1, 0, -1)\}$).

---

6. How to find the Range of a Linear Transformation

Scenario: "Find the Range of $L(x, y) = (x+y, x-y, 0)$."
Step 1: Separate the Variables
Take the output formula and pull apart the variables into distinct vectors.
$(x, x, 0) + (y, -y, 0)$
$x(1, 1, 0) + y(1, -1, 0)$ Step 2: Identify the Spanning Vectors
The vectors you factored out are the generators of the range.
$\text{Range } L = \text{Span} \{(1, 1, 0), (1, -1, 0)\}$ Step 3: Check Dimension (Optional)
Count the linearly independent vectors in that span. This is the dimension of the Range (Rank).

◇Common Pitfalls

Checking linear independence by inspection instead of row-reducing — always set up the matrix and reduce to RREF Confusing span (set of all linear combinations) with basis (minimal spanning set that is also linearly independent) Forgetting that dimension = number of vectors in any basis, not the number of vectors given Not setting up the equation $T(\mathbf{x}) = \mathbf{0}$ properly when finding the kernel — write the transformation as a matrix equation first Miscounting rank (pivot columns) vs. nullity (free columns) after row reduction

◇Summary

TL;DR — Key pointers: Span = all linear combinations; linear independence = check via RREF (no free variables); basis = linearly independent spanning set; dimension = basis size. For transformations: kernel = $T(\mathbf{x}) = \mathbf{0}$ (nullity), range = column space of the matrix (rank). Always verify with rank-nullity: $\text{rank} + \text{nullity} = n$.

◇Links

Related: Source / Reference:

◇References & Navigation

• Related Notes: ◇Linear Equations and Matrices