Solved Exercises on Basis and Dimension

Master basis, dimension, linear independence, and spanning sets through these 5 detailed exercises.

Basis and Dimension - Comprehensive Summary

Definitions

Basis: A set of vectors that is linearly independent and spans the vector space

Dimension: The number of vectors in any basis for the vector space

Linear Independence: No vector in the set can be written as a linear combination of others

Spanning Set: Every vector in the space can be written as a linear combination of vectors in the set

Core Rules

• Basis = Linearly independent + Spanning

• All bases of a vector space have the same size

• Dimension is well-defined

• In R^n, any n linearly independent vectors form a basis

Methods

1. Check linear independence using determinant or row reduction

2. Verify spanning property by solving for coefficients

3. Use dimension theorem: dim(span) ≤ number of vectors

4. Apply Rank-Nullity theorem when relevant

Key Tips

• Count vectors in proposed basis

• Check against dimension of space

• Use row reduction for efficiency

• Verify both conditions (independence and spanning)

Solution: Exercises 1 to 3
1 Standard Basis
Exercise 1
Prove that the standard basis vectors \(\vec{e_1} = \begin{pmatrix} 1 \\ 0 \\ 0 \end{pmatrix}\), \(\vec{e_2} = \begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix}\), \(\vec{e_3} = \begin{pmatrix} 0 \\ 0 \\ 1 \end{pmatrix}\) form a basis for \(\mathbb{R}^3\).
Definition:

Basis: A set of vectors that is linearly independent and spans the vector space. For a set to be a basis for \(\mathbb{R}^3\), it must contain exactly 3 linearly independent vectors that span \(\mathbb{R}^3\).

Basis Verification Method:
  1. Verify linear independence: Show that the only solution to \(c_1\vec{e_1} + c_2\vec{e_2} + c_3\vec{e_3} = \vec{0}\) is \(c_1 = c_2 = c_3 = 0\)
  2. Verify spanning property: Show that any vector in \(\mathbb{R}^3\) can be written as a linear combination of the basis vectors
  3. Confirm the number of vectors equals the dimension of the space
Linear Independence
Only trivial solution
Spanning
Any \(\vec{v} \in \mathbb{R}^3\) expressible
Dimension
3 vectors in \(\mathbb{R}^3\)
Step 1: Verify linear independence

We need to solve: \(c_1\begin{pmatrix} 1 \\ 0 \\ 0 \end{pmatrix} + c_2\begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix} + c_3\begin{pmatrix} 0 \\ 0 \\ 1 \end{pmatrix} = \begin{pmatrix} 0 \\ 0 \\ 0 \end{pmatrix}\)

This gives us: \(\begin{pmatrix} c_1 \\ c_2 \\ c_3 \end{pmatrix} = \begin{pmatrix} 0 \\ 0 \\ 0 \end{pmatrix}\)

Therefore: \(c_1 = c_2 = c_3 = 0\). The vectors are linearly independent. ✓

Step 2: Verify spanning property

Let \(\vec{v} = \begin{pmatrix} a \\ b \\ c \end{pmatrix}\) be any vector in \(\mathbb{R}^3\).

We can write: \(\vec{v} = a\begin{pmatrix} 1 \\ 0 \\ 0 \end{pmatrix} + b\begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix} + c\begin{pmatrix} 0 \\ 0 \\ 1 \end{pmatrix} = a\vec{e_1} + b\vec{e_2} + c\vec{e_3}\)

Therefore, any vector in \(\mathbb{R}^3\) can be expressed as a linear combination of these vectors. ✓

Step 3: Confirm dimension

The set contains exactly 3 vectors, which equals the dimension of \(\mathbb{R}^3\). ✓

Step 4: Conclusion

Since the set is linearly independent, spans \(\mathbb{R}^3\), and contains the correct number of vectors, it forms a basis for \(\mathbb{R}^3\). ✓

The standard basis vectors form a basis for \(\mathbb{R}^3\).
Final answer:

The standard basis vectors \(\{\vec{e_1}, \vec{e_2}, \vec{e_3}\}\) form a basis for \(\mathbb{R}^3\) because they are linearly independent and span \(\mathbb{R}^3\).

Applied rules:

• Linear independence test: Only trivial solution to homogeneous equation

• Spanning verification: Express arbitrary vector as linear combination

• Dimension matching: Number of vectors equals space dimension

2 Non-Standard Basis
Exercise 2
Determine if the vectors \(\vec{v_1} = \begin{pmatrix} 1 \\ 2 \\ 0 \end{pmatrix}\), \(\vec{v_2} = \begin{pmatrix} 0 \\ 1 \\ 1 \end{pmatrix}\), \(\vec{v_3} = \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}\) form a basis for \(\mathbb{R}^3\).
Definition:

Non-Standard Basis: A basis that is not the standard basis. To verify, we must check both linear independence and spanning properties.

Linear Independence
Check det ≠ 0
Determinant
\(\det = 3 \neq 0\)
Conclusion
Forms a basis
Step 1: Set up the matrix

Form a matrix with the vectors as columns: \(A = \begin{pmatrix} 1 & 0 & 1 \\ 2 & 1 & 0 \\ 0 & 1 & 1 \end{pmatrix}\)

Step 2: Calculate the determinant

\(\det(A) = 1\begin{vmatrix} 1 & 0 \\ 1 & 1 \end{vmatrix} - 0\begin{vmatrix} 2 & 0 \\ 0 & 1 \end{vmatrix} + 1\begin{vmatrix} 2 & 1 \\ 0 & 1 \end{vmatrix}\)

\(= 1(1 \cdot 1 - 0 \cdot 1) - 0 + 1(2 \cdot 1 - 1 \cdot 0)\)

\(= 1(1) + 1(2) = 1 + 2 = 3\)

Step 3: Apply the determinant test

Since \(\det(A) = 3 \neq 0\), the vectors are linearly independent. ✓

Step 4: Verify spanning

Since we have 3 linearly independent vectors in \(\mathbb{R}^3\) (which has dimension 3), they automatically span \(\mathbb{R}^3\). ✓

Step 5: Conclusion

Since the vectors are linearly independent and span \(\mathbb{R}^3\), they form a basis for \(\mathbb{R}^3\). ✓

The vectors form a basis for \(\mathbb{R}^3\).
Final answer:

Yes, the vectors \(\{\vec{v_1}, \vec{v_2}, \vec{v_3}\}\) form a basis for \(\mathbb{R}^3\) because they are linearly independent and span the space.

Applied rules:

• Determinant test: Non-zero determinant implies linear independence

• Dimension theorem: n independent vectors in R^n form a basis

• Efficiency shortcut: When dimension matches, independence implies spanning

3 Not a Basis Example
Exercise 3
Show that the vectors \(\vec{u_1} = \begin{pmatrix} 1 \\ 2 \\ 3 \end{pmatrix}\), \(\vec{u_2} = \begin{pmatrix} 2 \\ 4 \\ 6 \end{pmatrix}\), \(\vec{u_3} = \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}\) do NOT form a basis for \(\mathbb{R}^3\).
Definition:

Not a Basis: A set fails to be a basis if it violates either linear independence or spanning property (or both). We'll demonstrate linear dependence.

Linear Dependence
\(\vec{u_2} = 2\vec{u_1}\)
Determinant
\(\det = 0\)
Conclusion
Not a basis
Step 1: Identify the relationship

Notice that \(\vec{u_2} = \begin{pmatrix} 2 \\ 4 \\ 6 \end{pmatrix} = 2\begin{pmatrix} 1 \\ 2 \\ 3 \end{pmatrix} = 2\vec{u_1}\)

Since one vector is a scalar multiple of another, the vectors are linearly dependent. ✓

Step 2: Verify with determinant

Form matrix: \(A = \begin{pmatrix} 1 & 2 & 1 \\ 2 & 4 & 0 \\ 3 & 6 & 1 \end{pmatrix}\)

\(\det(A) = 1\begin{vmatrix} 4 & 0 \\ 6 & 1 \end{vmatrix} - 2\begin{vmatrix} 2 & 0 \\ 3 & 1 \end{vmatrix} + 1\begin{vmatrix} 2 & 4 \\ 3 & 6 \end{vmatrix}\)

\(= 1(4 \cdot 1 - 0 \cdot 6) - 2(2 \cdot 1 - 0 \cdot 3) + 1(2 \cdot 6 - 4 \cdot 3)\)

\(= 1(4) - 2(2) + 1(12 - 12) = 4 - 4 + 0 = 0\)

Since \(\det(A) = 0\), the vectors are linearly dependent. ✓

Step 3: Show explicit dependence relation

We can write: \(2\vec{u_1} - \vec{u_2} + 0\vec{u_3} = \vec{0}\)

This is a non-trivial linear combination equal to zero, confirming linear dependence. ✓

Step 4: Conclusion

Since the vectors are linearly dependent, they cannot form a basis for \(\mathbb{R}^3\). ✓

Step 5: Additional note

These vectors span at most a 2-dimensional subspace of \(\mathbb{R}^3\), not the entire space.

The vectors do not form a basis for \(\mathbb{R}^3\).
Final answer:

The vectors do not form a basis for \(\mathbb{R}^3\) because they are linearly dependent (specifically, \(\vec{u_2} = 2\vec{u_1}\)).

Applied rules:

• Dependence detection: One vector is scalar multiple of another

• Determinant test: Zero determinant confirms dependence

• Explicit relation: Find non-trivial combination equal to zero

Basis and Dimension Properties
\(\text{dim}(V) = \text{number of vectors in any basis of } V\)
Dimension Definition
Standard Basis
\(\mathbb{R}^n \text{ has dimension } n\)
Standard basis has n vectors
Independence-Spanning
\text{Basis} \Leftrightarrow \text{Lin. Indep.} \land \text{Spans}\)
Basis iff both conditions satisfied
Dimension Theorem
\text{dim}(\text{col}(A)) + \text{dim}(\text{null}(A)) = n\)
Rank-nullity theorem
Uniqueness of Dimension: All bases of a vector space have the same number of vectors
Maximal Independence: A basis is a maximal linearly independent set
Minimal Spanning: A basis is a minimal spanning set
Solution: Exercises 4 to 5
4 Subspace Basis
Exercise 4
Find a basis and determine the dimension of the subspace \(W = \text{span}\left\{\begin{pmatrix} 1 \\ 2 \\ 1 \end{pmatrix}, \begin{pmatrix} 2 \\ 4 \\ 2 \end{pmatrix}, \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}, \begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix}\right\}\).
Definition:

Spanning Set Reduction: To find a basis for the span of a set of vectors, we need to eliminate dependent vectors while preserving the span. The remaining vectors form a basis.

Matrix Form
Row reduce
Pivot Columns
Columns 1, 3
Basis
\(\left\{\begin{pmatrix} 1 \\ 2 \\ 1 \end{pmatrix}, \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}\right\}\)
Step 1: Form matrix with vectors as rows

Let \(A = \begin{pmatrix} 1 & 2 & 1 \\ 2 & 4 & 2 \\ 1 & 0 & 1 \\ 0 & 1 & 0 \end{pmatrix}\)

Step 2: Row reduce to echelon form

\(\begin{pmatrix} 1 & 2 & 1 \\ 2 & 4 & 2 \\ 1 & 0 & 1 \\ 0 & 1 & 0 \end{pmatrix} \xrightarrow{\text{row ops}} \begin{pmatrix} 1 & 2 & 1 \\ 0 & 0 & 0 \\ 0 & -2 & 0 \\ 0 & 1 & 0 \end{pmatrix} \xrightarrow{\text{more ops}} \begin{pmatrix} 1 & 2 & 1 \\ 0 & 1 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{pmatrix}\)

Step 3: Identify pivot columns

The pivot columns in the row-reduced form are columns 1 and 2.

This corresponds to the first and fourth original vectors.

Step 4: Select basis vectors

A basis for W is \(\left\{\begin{pmatrix} 1 \\ 2 \\ 1 \end{pmatrix}, \begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix}\right\}\).

Wait, let me recalculate more carefully:

Looking at the original vectors: \(\vec{v_1} = \begin{pmatrix} 1 \\ 2 \\ 1 \end{pmatrix}\), \(\vec{v_2} = \begin{pmatrix} 2 \\ 4 \\ 2 \end{pmatrix}\), \(\vec{v_3} = \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}\), \(\vec{v_4} = \begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix}\)

Note: \(\vec{v_2} = 2\vec{v_1}\), so \(\vec{v_2}\) is dependent on \(\vec{v_1}\).

Now check if \(\vec{v_1}\) and \(\vec{v_3}\) are independent: \(\begin{pmatrix} 1 & 1 \\ 2 & 0 \\ 1 & 1 \end{pmatrix}\) has rank 2.

Check if \(\vec{v_4}\) is independent of \(\{\vec{v_1}, \vec{v_3}\}\): Form \(\begin{pmatrix} 1 & 1 & 0 \\ 2 & 0 & 1 \\ 1 & 1 & 0 \end{pmatrix}\). The third column is independent of the first two.

Actually, let's reconsider: \(\vec{v_4} = \begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix}\) is independent of \(\{\vec{v_1}, \vec{v_3}\}\).

Step 5: Final verification

Check if \(\{\vec{v_1}, \vec{v_3}, \vec{v_4}\}\) are linearly independent:

\(c_1\begin{pmatrix} 1 \\ 2 \\ 1 \end{pmatrix} + c_2\begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix} + c_3\begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix} = \begin{pmatrix} 0 \\ 0 \\ 0 \end{pmatrix}\)

This gives: \(c_1 + c_2 = 0\), \(2c_1 + c_3 = 0\), \(c_1 + c_2 = 0\)

From first and third: \(c_1 + c_2 = 0\)

From second: \(c_3 = -2c_1\)

Let \(c_2 = -c_1\), then all equations satisfied only when \(c_1 = c_2 = c_3 = 0\)

Actually, wait: if \(c_1 = 1\), \(c_2 = -1\), \(c_3 = -2\), then \(2(1) + (-2) = 0\), so this is not independent!

Let me be more systematic: \(\begin{pmatrix} 1 & 1 & 0 & | & 0 \\ 2 & 0 & 1 & | & 0 \\ 1 & 1 & 0 & | & 0 \end{pmatrix}\)

Row reducing: \(\begin{pmatrix} 1 & 1 & 0 & | & 0 \\ 0 & -2 & 1 & | & 0 \\ 0 & 0 & 0 & | & 0 \end{pmatrix}\)

This gives: \(c_1 + c_2 = 0\) and \(-2c_2 + c_3 = 0\), so \(c_1 = -c_2\) and \(c_3 = 2c_2\)

So \(\vec{v_4} = 2\vec{v_3} - 2\vec{v_1}\) (checking: \(2\begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix} - 2\begin{pmatrix} 1 \\ 2 \\ 1 \end{pmatrix} = \begin{pmatrix} 0 \\ -4 \\ 0 \end{pmatrix} \neq \vec{v_4}\))

Actually: \(\vec{v_4} = \frac{1}{2}\vec{v_1} - \frac{1}{2}\vec{v_3}\) (checking: \(\frac{1}{2}\begin{pmatrix} 1 \\ 2 \\ 1 \end{pmatrix} - \frac{1}{2}\begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix} = \begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix} = \vec{v_4}\))

So \(\vec{v_4}\) is dependent on \(\{\vec{v_1}, \vec{v_3}\}\).

Step 6: Conclusion

A basis for W is \(\left\{\begin{pmatrix} 1 \\ 2 \\ 1 \end{pmatrix}, \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}\right\}\).

Since \(\vec{v_2} = 2\vec{v_1}\) and \(\vec{v_4} = \frac{1}{2}\vec{v_1} - \frac{1}{2}\vec{v_3}\), only \(\vec{v_1}\) and \(\vec{v_3}\) are needed.

Basis: \(\left\{\begin{pmatrix} 1 \\ 2 \\ 1 \end{pmatrix}, \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}\right\}\), Dimension: 2
Final answer:

A basis for W is \(\left\{\begin{pmatrix} 1 \\ 2 \\ 1 \end{pmatrix}, \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}\right\}\) and \(\text{dim}(W) = 2\).

Applied rules:

• Row reduction method: Efficiently identify independent vectors

• Span preservation: Removing dependent vectors doesn't change the span

• Dimension counting: Number of basis vectors equals dimension

5 Polynomial Space
Exercise 5
Find a basis and determine the dimension of the subspace \(P_2 = \{p(x) \in \mathcal{P}_2 \mid p(1) = 0\}\), where \(\mathcal{P}_2\) is the space of polynomials of degree ≤ 2.
Definition:

Polynomial Subspace: The space of quadratic polynomials that vanish at x = 1. We need to find polynomials of the form \(p(x) = ax^2 + bx + c\) where \(p(1) = a + b + c = 0\).

Constraint
\(c = -a - b\)
General Form
\(p(x) = a(x^2-1) + b(x-1)\)
Basis
\(\{x^2-1, x-1\}\)
Step 1: Express the constraint

For \(p(x) = ax^2 + bx + c\) to satisfy \(p(1) = 0\):

\(p(1) = a(1)^2 + b(1) + c = a + b + c = 0\)

Therefore: \(c = -a - b\)

Step 2: Rewrite the general form

Any polynomial in the subspace can be written as:

\(p(x) = ax^2 + bx + (-a - b) = ax^2 + bx - a - b\)

\(= a(x^2 - 1) + b(x - 1)\)

Step 3: Identify potential basis vectors

From the expression \(p(x) = a(x^2 - 1) + b(x - 1)\), we see that any polynomial in the subspace is a linear combination of \((x^2 - 1)\) and \((x - 1)\).

So our candidate basis is \(\{x^2 - 1, x - 1\}\).

Step 4: Verify linear independence

We need to show that the only solution to \(c_1(x^2 - 1) + c_2(x - 1) = 0\) is \(c_1 = c_2 = 0\).

\(c_1x^2 - c_1 + c_2x - c_2 = 0\)

\(c_1x^2 + c_2x + (-c_1 - c_2) = 0\)

For this to be the zero polynomial: \(c_1 = 0\), \(c_2 = 0\), and \(-c_1 - c_2 = 0\).

All conditions are satisfied only when \(c_1 = c_2 = 0\). ✓

Step 5: Verify spanning

By construction, any polynomial \(p(x)\) in the subspace with \(p(1) = 0\) can be written as \(p(x) = a(x^2 - 1) + b(x - 1)\), which is a linear combination of our basis vectors. ✓

Step 6: Conclusion

The set \(\{x^2 - 1, x - 1\}\) is linearly independent and spans the subspace, so it forms a basis.

Basis: \(\{x^2 - 1, x - 1\}\), Dimension: 2
Final answer:

A basis for the subspace is \(\{x^2 - 1, x - 1\}\) and the dimension is 2.

Applied rules:

• Constraint substitution: Use the condition to eliminate variables

• Factorization approach: Write polynomials in factored form when possible

• Polynomial independence: Coefficients of different powers must be zero

Basis and Dimension: Advanced Concepts
\(\text{basis} \Leftrightarrow \text{linearly independent} \land \text{spans the space}\)
Basis Characterization
Key definitions:

Basis: A set of vectors that is linearly independent and spans the vector space. It provides a coordinate system for the space.

Dimension: The number of vectors in any basis of a vector space. This is well-defined and unique for each space.

Finite-Dimensional: A vector space is finite-dimensional if it has a finite basis.

Complete methodology:
  1. For linear independence: Show that the only solution to the homogeneous equation is the trivial solution
  2. For spanning: Show that any vector in the space can be written as a linear combination of the proposed basis vectors
  3. For dimension: Count the number of vectors in a basis
  4. Verification: Ensure the number of vectors matches the expected dimension
Tip 1: In R^n, if you have n vectors, check the determinant - non-zero means basis.
Tip 2: For subspaces, use row reduction to identify independent vectors.
Tip 3: If you have too many vectors, some are necessarily dependent.
Tip 4: Use the rank-nullity theorem to verify dimensions.
Common errors: Forgetting to check both independence and spanning, miscounting dimensions, not considering constraints in subspaces.
Applications: Coordinate systems, linear transformations, differential equations, Fourier series, data compression.
Properties to know by heart:

• All bases of a vector space have the same number of elements

• In R^n, any n linearly independent vectors form a basis

• A basis is both a minimal spanning set and a maximal independent set

• Dimension of a subspace ≤ dimension of the parent space

• \(\text{dim}(\text{col}(A)) + \text{dim}(\text{null}(A)) = \text{number of columns of } A\)

Basis Visualization: Different Bases for R²
Exercise 6: Basis Comparison
Visualize how different bases for R² provide different coordinate systems. Show the standard basis and an alternative basis \(\left\{\begin{pmatrix} 2 \\ 1 \end{pmatrix}, \begin{pmatrix} 1 \\ 2 \end{pmatrix}\right\}\).

Geometric Interpretation: Different bases create different "grid" systems for the same space. Any vector can be represented uniquely in terms of any basis.

  • Standard basis: \(\left\{\begin{pmatrix} 1 \\ 0 \end{pmatrix}, \begin{pmatrix} 0 \\ 1 \end{pmatrix}\right\}\)
  • Alternative basis: \(\left\{\begin{pmatrix} 2 \\ 1 \end{pmatrix}, \begin{pmatrix} 1 \\ 2 \end{pmatrix}\right\}\)
  • Same space, different coordinate systems

Questions & Answers

Question: I'm confused about why we need both linear independence AND spanning to form a basis. Why isn't one enough?

Answer: Both conditions are essential for different reasons:

Linear Independence: This ensures minimality. If vectors are linearly dependent, one vector can be expressed as a combination of others, making it redundant. A basis should have no redundant vectors.

Spanning: This ensures completeness. If a set doesn't span the space, there are vectors in the space that cannot be expressed using the basis vectors. A basis should be able to reach every vector in the space.

Here's what happens if you only have one condition:

  • Independent but not spanning: Too few vectors (e.g., \(\{(1,0)\}\) in \(\mathbb{R}^2\))
  • Spanning but not independent: Too many vectors (e.g., \(\{(1,0), (0,1), (1,1)\}\) in \(\mathbb{R}^2\))
  • Both conditions: Just the right number of vectors (e.g., \(\{(1,0), (0,1)\}\) in \(\mathbb{R}^2\))

A basis is the "perfect" set - minimal but complete.

Question: How do we know that all bases of a vector space have the same number of vectors? That seems like it could be different.

Answer: This is a fundamental theorem in linear algebra called the "Dimension Theorem." Here's the intuition:

Steinitz Exchange Lemma: If you have a linearly independent set of size \(m\) and a spanning set of size \(n\), then \(m \leq n\).

Application to bases: If \(B_1\) and \(B_2\) are both bases with sizes \(m\) and \(n\) respectively:

  • Since \(B_1\) is linearly independent and \(B_2\) spans, we have \(m \leq n\)
  • Since \(B_2\) is linearly independent and \(B_1\) spans, we have \(n \leq m\)
  • Therefore \(m = n\)

This theorem guarantees that dimension is well-defined and doesn't depend on which basis you choose. It's one of the most important results in linear algebra.

Question: How do I practically check if a set of vectors spans a space?

Answer: There are several practical methods:

Method 1 - General Vector Approach: Show that an arbitrary vector in the space can be written as a linear combination of your vectors. For example, for \(\mathbb{R}^3\), show that \((a,b,c) = c_1\vec{v_1} + c_2\vec{v_2} + c_3\vec{v_3}\) has a solution for any \((a,b,c)\).

Method 2 - Known Basis Approach: Show that you can express a known basis in terms of your vectors. If you can write the standard basis vectors using your set, then your set spans the space.

Method 3 - Dimension Approach: If you have \(n\) linearly independent vectors in an \(n\)-dimensional space, they automatically span the space.

Method 4 - Row Reduction: Form a matrix with your vectors as rows/columns and row reduce. If the rank equals the dimension of the space, the vectors span it.

Method 3 is often the most efficient when applicable.

Question: How are bases used in computer graphics and 3D modeling?

Answer: Bases are fundamental in computer graphics:

Local Coordinate Systems:

  • Objects have their own local basis (object space)
  • Camera has its own basis (view space)
  • World coordinates use a global basis

Transformations:

  • Change of basis matrices convert between coordinate systems
  • Rotation matrices define new orthonormal bases
  • Scaling transforms modify basis vector lengths

Surface Properties:

  • Surface normal, tangent, and bitangent form a local basis at each point
  • Used for lighting calculations and texture mapping

Understanding bases helps in implementing efficient transformations and maintaining geometric relationships in 3D scenes.

Question: How do bases relate to principal component analysis (PCA) in machine learning?

Answer: PCA fundamentally constructs a new basis:

Eigenvector Basis: PCA finds the principal components, which are the eigenvectors of the covariance matrix. These eigenvectors form an orthonormal basis for the data space.

Dimensionality Reduction:

  • The top k eigenvectors form a k-dimensional subspace
  • Data is projected onto this lower-dimensional basis
  • The new basis captures maximum variance in fewer dimensions

Coordinate Transformation:

  • Original data is re-expressed in terms of the principal component basis
  • Each data point gets new coordinates in the PCA basis
  • This is a change of basis operation

PCA essentially finds the optimal basis for representing data with minimal information loss.