Linear Independence Problems with solutions

These linear independence exercises span every core topic in undergraduate linear algebra: the definition of a linearly independent set and the trivial solution, testing sets of vectors in \(\mathbb{R}^n\) by row reduction, using the determinant for square systems, finding dependence relations and expressing vectors as linear combinations, connecting independence to rank, nullity, and the null space, applying the Wronskian to sets of functions, and proving results in abstract vector spaces. Problems are ordered from straightforward definition checks to multi-step proofs, giving a complete path from first encounter to exam readiness.

Definition of Linear Independence and Dependence

A set \(\{v_1, v_2, \ldots, v_k\}\) is linearly independent if the vector equation \(c_1 v_1 + c_2 v_2 + \cdots + c_k v_k = \mathbf{0}\) has only the trivial solution \(c_1 = c_2 = \cdots = c_k = 0\). These problems train students to apply this definition directly, identify dependence by inspection, and write explicit dependence relations.

Problem 1: Applying the Definition in \(\mathbb{R}^2\)

Easy

Consider the vectors \(u = \begin{pmatrix} 2 \\ 5 \end{pmatrix}\) and \(v = \begin{pmatrix} 6 \\ 15 \end{pmatrix}\) in \(\mathbb{R}^2\).

  1. Write out the vector equation \(c_1 u + c_2 v = \mathbf{0}\) as a system of two scalar equations and solve it.
  2. State whether \(\{u, v\}\) is linearly independent or linearly dependent, and justify your answer using the definition.
Hint
Notice that each component of \(v\) is exactly 3 times the corresponding component of \(u\). Ask yourself whether a nontrivial solution to the scalar system exists.
View Solution
Solution to question 1:

The equation \(c_1 u + c_2 v = \mathbf{0}\) yields the system:

\[ 2c_1 + 6c_2 = 0, \quad 5c_1 + 15c_2 = 0. \]

Both equations reduce to \(c_1 = -3c_2\). Since \(c_2\) is a free variable, there are infinitely many solutions, for instance \(c_1 = 3\), \(c_2 = -1\).

Solution to question 2:

Because the system has a nontrivial solution (not all coefficients zero), the set \(\{u, v\}\) is linearly dependent. Explicitly, \(3u – v = \mathbf{0}\), so \(v = 3u\): one vector is a scalar multiple of the other.

Problem 2: Independence by Inspection in \(\mathbb{R}^3\)

Easy

Without performing any row reduction, determine whether each set below is linearly independent or linearly dependent. Justify each answer with a brief reason referencing a standard theorem or observation.

  1. \(\left\{ \begin{pmatrix}1\\0\\0\end{pmatrix},\ \begin{pmatrix}0\\1\\0\end{pmatrix},\ \begin{pmatrix}0\\0\\1\end{pmatrix},\ \begin{pmatrix}2\\-1\\3\end{pmatrix} \right\}\) in \(\mathbb{R}^3\).
  2. \(\left\{ \begin{pmatrix}4\\-2\\6\end{pmatrix},\ \begin{pmatrix}-6\\3\\-9\end{pmatrix} \right\}\) in \(\mathbb{R}^3\).
  3. \(\left\{ \begin{pmatrix}1\\2\\3\end{pmatrix},\ \mathbf{0},\ \begin{pmatrix}0\\1\\4\end{pmatrix} \right\}\) in \(\mathbb{R}^3\).
Hint
Recall three quick facts: (a) any set containing the zero vector is dependent; (b) any set of more than \(n\) vectors in \(\mathbb{R}^n\) is dependent; (c) two vectors are dependent if and only if one is a scalar multiple of the other.
View Solution
Solution to question 1:

Linearly dependent. There are 4 vectors in \(\mathbb{R}^3\). Since the number of vectors exceeds the dimension of the space (\(4 > 3\)), the set must be linearly dependent.

Solution to question 2:

Linearly dependent. Observe that \(-6/4 = 3/(-2) = -9/6 = -3/2\), so the second vector equals \(-\tfrac{3}{2}\) times the first. Two vectors are dependent if and only if one is a scalar multiple of the other.

Solution to question 3:

Linearly dependent. Any set containing the zero vector is linearly dependent, because \(0 \cdot v_1 + 1 \cdot \mathbf{0} + 0 \cdot v_3 = \mathbf{0}\) is a nontrivial combination equal to zero.

Problem 3: Writing an Explicit Dependence Relation

Medium

Let \(v_1 = \begin{pmatrix}1\\2\\-1\end{pmatrix}\), \(v_2 = \begin{pmatrix}3\\0\\1\end{pmatrix}\), and \(v_3 = \begin{pmatrix}5\\4\\-1\end{pmatrix}\).

  1. Show that \(\{v_1, v_2, v_3\}\) is linearly dependent by finding a nontrivial solution to \(c_1 v_1 + c_2 v_2 + c_3 v_3 = \mathbf{0}\).
  2. Express \(v_3\) as a linear combination of \(v_1\) and \(v_2\).
Hint
Form the augmented matrix \([v_1 \mid v_2 \mid v_3 \mid \mathbf{0}]\) and row-reduce. A free variable in the solution confirms dependence and directly gives the relation coefficients.
View Solution
Solution to question 1:
\[ \begin{pmatrix}1 & 3 & 5 & 0\\2 & 0 & 4 & 0\\-1 & 1 & -1 & 0\end{pmatrix} \xrightarrow{R_2-2R_1,\ R_3+R_1} \begin{pmatrix}1 & 3 & 5 & 0\\0 & -6 & -6 & 0\\0 & 4 & 4 & 0\end{pmatrix} \xrightarrow{\frac{2}{3}R_2+R_3\to R_3} \begin{pmatrix}1 & 3 & 5 & 0\\0 & -6 & -6 & 0\\0 & 0 & 0 & 0\end{pmatrix}. \]

Column 3 is free (\(c_3 = t\)). From row 2: \(-6c_2 = 6t \Rightarrow c_2 = -t\). From row 1: \(c_1 = -3c_2 – 5c_3 = 3t – 5t = -2t\). Taking \(t = 1\) gives the nontrivial solution \((c_1,c_2,c_3) = (-2,-1,1)\), so \(-2v_1 – v_2 + v_3 = \mathbf{0}\).

Solution to question 2:
\[ v_3 = 2v_1 + v_2 = 2\begin{pmatrix}1\\2\\-1\end{pmatrix} + \begin{pmatrix}3\\0\\1\end{pmatrix} = \begin{pmatrix}5\\4\\-1\end{pmatrix}. \checkmark \]

Testing Linear Independence by Row Reduction

Row reduction is the most systematic method for testing independence: form the matrix \(A = [v_1 \mid v_2 \mid \cdots \mid v_k]\) and reduce to echelon form. The vectors are linearly independent if and only if every column contains a pivot position, equivalently if the homogeneous system \(Ax = \mathbf{0}\) has only the trivial solution.

Problem 4: Row Reduction Test in \(\mathbb{R}^3\)

Easy

Determine whether the set \(\{v_1, v_2, v_3\}\) is linearly independent, where

\[ v_1 = \begin{pmatrix}1\\0\\2\end{pmatrix},\quad v_2 = \begin{pmatrix}3\\1\\-1\end{pmatrix},\quad v_3 = \begin{pmatrix}0\\2\\5\end{pmatrix}. \]

  1. Set up the matrix \(A = [v_1 \mid v_2 \mid v_3]\) and reduce it to row echelon form.
  2. Count the pivot positions and state your conclusion.
Hint
If after row reduction you obtain three pivot positions (one per column), the columns are linearly independent. If any column lacks a pivot, the set is dependent.
View Solution
Solution to question 1:
\[ A = \begin{pmatrix}1 & 3 & 0\\0 & 1 & 2\\2 & -1 & 5\end{pmatrix} \xrightarrow{R_3 – 2R_1} \begin{pmatrix}1 & 3 & 0\\0 & 1 & 2\\0 & -7 & 5\end{pmatrix} \xrightarrow{R_3 + 7R_2} \begin{pmatrix}1 & 3 & 0\\0 & 1 & 2\\0 & 0 & 19\end{pmatrix}. \]
Solution to question 2:

There are 3 pivot positions, one in each column. The homogeneous system \(Ax = \mathbf{0}\) has only the trivial solution, so \(\{v_1, v_2, v_3\}\) is linearly independent.

Problem 5: Four Vectors in \(\mathbb{R}^3\) (Reduction and Dependence Relation)

Medium

Consider the vectors in \(\mathbb{R}^3\):

\[ v_1 = \begin{pmatrix}1\\2\\1\end{pmatrix},\quad v_2 = \begin{pmatrix}0\\1\\3\end{pmatrix},\quad v_3 = \begin{pmatrix}2\\0\\-4\end{pmatrix},\quad v_4 = \begin{pmatrix}3\\3\\0\end{pmatrix}. \]

  1. Without calculation, explain why \(\{v_1, v_2, v_3, v_4\}\) must be linearly dependent.
  2. Use row reduction on \([v_1 \mid v_2 \mid v_3 \mid v_4]\) to find a nontrivial dependence relation among these four vectors.
Hint
For part (1), compare the number of vectors with the dimension of the ambient space. For part (2), row-reduce and identify the free variable, then back-substitute.
View Solution
Solution to question 1:

There are 4 vectors in \(\mathbb{R}^3\), and \(4 > 3 = \dim(\mathbb{R}^3)\). Any set containing more vectors than the dimension of the ambient space is automatically linearly dependent.

Solution to question 2:
\[ \begin{pmatrix}1&0&2&3\\2&1&0&3\\1&3&-4&0\end{pmatrix} \xrightarrow{R_2-2R_1,\ R_3-R_1} \begin{pmatrix}1&0&2&3\\0&1&-4&-3\\0&3&-6&-3\end{pmatrix} \xrightarrow{R_3-3R_2} \begin{pmatrix}1&0&2&3\\0&1&-4&-3\\0&0&6&6\end{pmatrix}. \]

Set \(c_4 = 1\) (free after noting the system has a non-trivial kernel). From row 3: \(6c_3 + 6 = 0 \Rightarrow c_3 = -1\). Row 2: \(c_2 – 4(-1) – 3 = 0 \Rightarrow c_2 = -1\). Row 1: \(c_1 + 2(-1) + 3 = 0 \Rightarrow c_1 = -1\). The dependence relation is:

\[ -v_1 – v_2 – v_3 + v_4 = \mathbf{0}, \quad \text{i.e.,} \quad v_4 = v_1 + v_2 + v_3. \]

Problem 6: Vectors in \(\mathbb{R}^4\) with a Real Parameter

Hard

Let \(h \in \mathbb{R}\) and consider

\[ v_1 = \begin{pmatrix}1\\0\\0\\h\end{pmatrix},\quad v_2 = \begin{pmatrix}0\\1\\0\\2\end{pmatrix},\quad v_3 = \begin{pmatrix}0\\0\\1\\-1\end{pmatrix},\quad v_4 = \begin{pmatrix}2\\-1\\h\\0\end{pmatrix}. \]

  1. Row-reduce the matrix \([v_1 \mid v_2 \mid v_3 \mid v_4]\) and determine for which values of \(h\) the set is linearly independent.
  2. For the value(s) of \(h\) where the set is linearly dependent, write an explicit dependence relation.
Hint
Form the \(4 \times 4\) matrix with these columns. Use the first three rows to eliminate entries in row 4 involving \(h\). The fourth row becomes a scalar multiple of \((0,0,0, 2-3h)\); dependence occurs when this scalar equals zero.
View Solution
Solution to question 1:

Writing \(A = [v_1|v_2|v_3|v_4]\):

\[ A = \begin{pmatrix}1&0&0&2\\0&1&0&-1\\0&0&1&h\\h&2&-1&0\end{pmatrix}. \]

Apply \(R_4 \leftarrow R_4 – hR_1 – 2R_2 + R_3\):

\[ R_4 \rightarrow \begin{pmatrix}h-h & 2-2 & -1+1 & 0 – 2h + 2 – h\end{pmatrix} = \begin{pmatrix}0 & 0 & 0 & 2-3h\end{pmatrix}. \]

The fourth pivot is \(2 – 3h\). The set is linearly independent when \(2 – 3h \neq 0\), i.e., \(\boxed{h \neq \tfrac{2}{3}}\).

Solution to question 2:

When \(h = \tfrac{2}{3}\), row 4 vanishes and \(c_4\) is free. Set \(c_4 = 1\). Back-substitution gives \(c_3 = -h = -\tfrac{2}{3}\), \(c_2 = 1\), \(c_1 = -2\). The dependence relation is:

\[ -2v_1 + v_2 – \tfrac{2}{3}v_3 + v_4 = \mathbf{0}. \]

Determinant Test for Linear Independence

When a set consists of exactly \(n\) vectors in \(\mathbb{R}^n\), the independence question reduces to a single scalar: the vectors are linearly independent if and only if \(\det[v_1 \mid \cdots \mid v_n] \neq 0\). This section develops fluency with computing and interpreting determinants in this context, including parametric problems where the critical values of a parameter must be found.

Problem 7: \(3 \times 3\) Determinant Test

Easy

Let \(w_1 = (2,\, 1,\, 0)\), \(w_2 = (0,\, 3,\, -1)\), and \(w_3 = (1,\, 0,\, 4)\).

  1. Form the \(3 \times 3\) matrix whose columns are \(w_1, w_2, w_3\) and compute its determinant by cofactor expansion along the first row.
  2. Conclude whether \(\{w_1, w_2, w_3\}\) is linearly independent or dependent.
Hint
The cofactor expansion along row 1 is \(\det(A) = a_{11}C_{11} + a_{12}C_{12} + a_{13}C_{13}\), where each cofactor involves a \(2\times 2\) determinant and a sign \((-1)^{i+j}\).
View Solution
Solution to question 1:
\[ A = \begin{pmatrix}2&0&1\\1&3&0\\0&-1&4\end{pmatrix}. \]
\[ \det(A) = 2\begin{vmatrix}3&0\\-1&4\end{vmatrix} – 0\begin{vmatrix}1&0\\0&4\end{vmatrix} + 1\begin{vmatrix}1&3\\0&-1\end{vmatrix} = 2(12-0) – 0 + 1(-1-0) = 24 – 1 = 23. \]
Solution to question 2:

Since \(\det(A) = 23 \neq 0\), the matrix is invertible and the set \(\{w_1, w_2, w_3\}\) is linearly independent.

Problem 8: Finding the Parameter That Forces Dependence

Medium

For a real parameter \(k\), define the matrix

\[ A(k) = \begin{pmatrix}k & 1 & 0\\2 & k & 1\\0 & 1 & k\end{pmatrix}. \]

  1. Compute \(\det(A(k))\) as a polynomial in \(k\).
  2. Find all values of \(k\) for which the columns of \(A(k)\) are linearly dependent.
Hint
Expand the determinant along the first row. You should obtain a cubic polynomial in \(k\). Factor it carefully — try \(k=0\) as a root first.
View Solution
Solution to question 1:
\[ \det(A(k)) = k(k^2 – 1) – 1(2k – 0) + 0 = k^3 – k – 2k = k^3 – 3k. \]

Factoring: \(\det(A(k)) = k(k^2 – 3)\).

Solution to question 2:

The columns are linearly dependent exactly when \(\det(A(k)) = 0\):

\[ k(k^2 – 3) = 0 \implies k = 0,\quad k = \sqrt{3},\quad k = -\sqrt{3}. \]

Problem 9: Upper-Triangular Matrices and Independence — A Proof

Hard

Let \(M\) be an \(n \times n\) upper-triangular matrix, meaning \(m_{ij} = 0\) for all \(i > j\).

  1. Prove that if all diagonal entries of \(M\) are nonzero, then the column vectors of \(M\) are linearly independent.
  2. Give a \(2\times 2\) example showing that the conclusion may fail if even one diagonal entry is zero.
Hint
For part (1), recall that the determinant of an upper-triangular matrix equals the product of its diagonal entries. Then apply the determinant test for independence.
View Solution
Solution to question 1:

For any triangular matrix, \(\det(M) = m_{11} m_{22} \cdots m_{nn}\). If all diagonal entries are nonzero, then \(\det(M) \neq 0\). A square matrix with nonzero determinant is invertible; by the determinant test for independence, its columns are therefore linearly independent. \(\square\)

Solution to question 2:

Let \(M = \begin{pmatrix}0 & 1\\ 0 & 1\end{pmatrix}\). This is upper-triangular with \(m_{11} = 0\). Its determinant is \(0 \cdot 1 – 1 \cdot 0 = 0\), so its columns \(\begin{pmatrix}0\\0\end{pmatrix}\) and \(\begin{pmatrix}1\\1\end{pmatrix}\) are linearly dependent (the first column is the zero vector).

Linear Independence, Rank, and the Null Space

Linear independence of the columns of a matrix \(A\) is equivalent to \(\operatorname{null}(A) = \{\mathbf{0}\}\), which occurs precisely when \(\operatorname{rank}(A)\) equals the number of columns. The Rank-Nullity Theorem, \(\operatorname{rank}(A) + \operatorname{nullity}(A) = n\), ties these ideas together and is the engine behind many theoretical and computational results.

Problem 10: Rank and Column Independence

Easy

Let \(A = \begin{pmatrix}1 & 2 & 0\\ 0 & 1 & -1\\ 2 & 5 & -1\end{pmatrix}\).

  1. Find the row echelon form of \(A\) and determine \(\operatorname{rank}(A)\).
  2. Are the columns of \(A\) linearly independent? Explain using the rank and the Rank-Nullity Theorem.
Hint
The columns of an \(m \times n\) matrix are linearly independent if and only if \(\operatorname{rank}(A) = n\), i.e., every column is a pivot column and \(\operatorname{nullity}(A) = 0\).
View Solution
Solution to question 1:
\[ \begin{pmatrix}1&2&0\\0&1&-1\\2&5&-1\end{pmatrix} \xrightarrow{R_3-2R_1} \begin{pmatrix}1&2&0\\0&1&-1\\0&1&-1\end{pmatrix} \xrightarrow{R_3-R_2} \begin{pmatrix}1&2&0\\0&1&-1\\0&0&0\end{pmatrix}. \]

There are two pivot positions, so \(\operatorname{rank}(A) = 2\).

Solution to question 2:

\(A\) has 3 columns but \(\operatorname{rank}(A) = 2 < 3\). By the Rank-Nullity Theorem, \(\operatorname{nullity}(A) = 3 - 2 = 1 > 0\), meaning \(Ax = \mathbf{0}\) has a nontrivial solution. The columns are therefore linearly dependent.

Problem 11: Finding a Basis for the Null Space

Medium

Let \(B = \begin{pmatrix}1&-1&2&0\\2&-2&4&1\\0&0&0&1\end{pmatrix}\).

  1. Find \(\operatorname{rank}(B)\) and \(\operatorname{nullity}(B)\) using the Rank-Nullity Theorem.
  2. Find a basis for \(\operatorname{null}(B)\) and verify that the basis vectors are linearly independent.
Hint
Row-reduce \(B\) to reduced row echelon form. Identify the pivot columns and the free variables. Each free variable produces one basis vector for the null space.
View Solution
Solution to question 1:
\[ B \xrightarrow{R_2 – 2R_1} \begin{pmatrix}1&-1&2&0\\0&0&0&1\\0&0&0&1\end{pmatrix} \xrightarrow{R_3-R_2} \begin{pmatrix}1&-1&2&0\\0&0&0&1\\0&0&0&0\end{pmatrix}. \]

Pivot columns: 1 and 4. So \(\operatorname{rank}(B) = 2\) and \(\operatorname{nullity}(B) = 4 – 2 = 2\).

Solution to question 2:

Free variables: \(x_2 = s\) and \(x_3 = t\). From row 2: \(x_4 = 0\). From row 1: \(x_1 = x_2 – 2x_3 = s – 2t\). The general solution is

\[ x = s\begin{pmatrix}1\\1\\0\\0\end{pmatrix} + t\begin{pmatrix}-2\\0\\1\\0\end{pmatrix}. \]

A basis for \(\operatorname{null}(B)\) is \(\left\{n_1 = (1,1,0,0)^T,\, n_2 = (-2,0,1,0)^T\right\}\). These are independent because the system \(c_1 n_1 + c_2 n_2 = \mathbf{0}\) gives, from component 2: \(c_1 = 0\), and from component 3: \(c_2 = 0\).

Problem 12: Rank-Nullity Reasoning Without Explicit Computation

Hard

Suppose \(A\) is a \(7 \times 6\) matrix with \(\operatorname{rank}(A) = 4\).

  1. Determine \(\operatorname{nullity}(A)\), \(\operatorname{nullity}(A^T)\), and the dimensions of the four fundamental subspaces of \(A\).
  2. Are the columns of \(A\) linearly independent? Are the rows of \(A\) linearly independent? Justify both answers.
Hint
Apply the Rank-Nullity Theorem to \(A\) (6 columns) and to \(A^T\) (7 columns). Recall that \(\operatorname{rank}(A^T) = \operatorname{rank}(A)\). Columns are independent iff rank equals the number of columns.
View Solution
Solution to question 1:

By the Rank-Nullity Theorem applied to \(A\) (with 6 columns): \(\operatorname{nullity}(A) = 6 – 4 = 2\). \(A^T\) is \(6 \times 7\) with \(\operatorname{rank}(A^T) = 4\), so \(\operatorname{nullity}(A^T) = 7 – 4 = 3\). The four fundamental subspaces have dimensions:

SubspaceDimension
Column space of \(A\) (in \(\mathbb{R}^7\))4
Null space of \(A\) (in \(\mathbb{R}^6\))2
Row space of \(A\) (in \(\mathbb{R}^6\))4
Left null space of \(A\) (in \(\mathbb{R}^7\))3

Solution to question 2:

Columns: \(A\) has 6 columns but \(\operatorname{rank}(A) = 4 < 6\). The columns are not linearly independent (\(\operatorname{nullity}(A) = 2 > 0\)).

Rows: The row space has dimension 4, but \(A\) has 7 rows. Since \(4 < 7\), the rows are also not linearly independent.

Linear Independence of Functions and the Wronskian

In function spaces, linear independence is tested using the Wronskian determinant. For differentiable functions \(f_1, \ldots, f_k\) on an interval \(I\), if \(W[f_1,\ldots,f_k](x_0) \neq 0\) at some point \(x_0 \in I\), the functions are linearly independent on \(I\). These exercises also reinforce how the definition alone suffices for direct proofs of dependence.

Problem 13: Wronskian of Two Exponential Functions

Easy

Consider \(f(x) = e^{2x}\) and \(g(x) = e^{-x}\) on \(\mathbb{R}\).

  1. Compute the Wronskian \(W[f, g](x)\).
  2. Use the result to conclude whether \(\{f, g\}\) is linearly independent on \(\mathbb{R}\).
Hint
The Wronskian of two functions is \(W[f,g] = f\,g’ – f’\,g\). Compute each derivative and substitute.
View Solution
Solution to question 1:
\[ W[f,g](x) = \begin{vmatrix}e^{2x} & e^{-x}\\ 2e^{2x} & -e^{-x}\end{vmatrix} = e^{2x}(-e^{-x}) – e^{-x}(2e^{2x}) = -e^x – 2e^x = -3e^x. \]
Solution to question 2:

Since \(W[f,g](x) = -3e^x \neq 0\) for every \(x \in \mathbb{R}\), the functions \(f\) and \(g\) are linearly independent on \(\mathbb{R}\).

Problem 14: Three Trigonometric Functions (Wronskian and Direct Check)

Medium

Let \(f_1(x) = 1\), \(f_2(x) = \cos^2 x\), \(f_3(x) = \sin^2 x\) on \(\mathbb{R}\).

  1. Compute \(W[f_1, f_2, f_3](x)\) and state what you can conclude from the result.
  2. Show directly from the definition that \(\{f_1, f_2, f_3\}\) is linearly dependent, and write an explicit dependence relation.
Hint
For part (2), recall the Pythagorean identity \(\cos^2 x + \sin^2 x = 1\). This gives a nontrivial relation among the three functions valid for all \(x\).
View Solution
Solution to question 1:

Compute derivatives: \(f_1′ = 0\), \(f_2′ = -\sin(2x)\), \(f_3′ = \sin(2x)\), \(f_2^{\prime\prime} = -2\cos(2x)\), \(f_3{\prime\prime} = 2\cos(2x)\). Expanding along the first column:

\[ W = \begin{vmatrix}1 & \cos^2 x & \sin^2 x\\ 0 & -\sin(2x) & \sin(2x)\\ 0 & -2\cos(2x) & 2\cos(2x)\end{vmatrix} = 1\cdot\bigl[(-\sin 2x)(2\cos 2x)-(\sin 2x)(-2\cos 2x)\bigr] = 0. \]

The Wronskian is identically zero. For a general set of functions this is inconclusive, but dependence is confirmed directly below.

Solution to question 2:

The Pythagorean identity states \(\cos^2 x + \sin^2 x = 1\) for all \(x \in \mathbb{R}\), so:

\[ 1 \cdot f_1(x) – 1 \cdot f_2(x) – 1 \cdot f_3(x) = 1 – \cos^2 x – \sin^2 x = 0. \]

The coefficients \((c_1,c_2,c_3)=(1,-1,-1)\) are not all zero, so \(\{f_1,f_2,f_3\}\) is linearly dependent.

Problem 15: Independence of Exponentials with Distinct Exponents via the Vandermonde Determinant

Hard

Let \(\lambda_1, \lambda_2, \lambda_3\) be three distinct real numbers and set \(f_i(x) = e^{\lambda_i x}\), \(i=1,2,3\).

  1. Evaluate \(W[f_1, f_2, f_3]\) at \(x = 0\) and identify the resulting determinant as a Vandermonde determinant.
  2. Use the Vandermonde formula \(\det V(\lambda_1,\lambda_2,\lambda_3) = (\lambda_2-\lambda_1)(\lambda_3-\lambda_1)(\lambda_3-\lambda_2)\) to prove that \(\{e^{\lambda_1 x}, e^{\lambda_2 x}, e^{\lambda_3 x}\}\) is linearly independent when the \(\lambda_i\) are distinct.
Hint
Note that \(f_i^{(k)}(0) = \lambda_i^k\). At \(x = 0\) the Wronskian matrix has entries \(\lambda_i^{j-1}\) in row \(j\), which is precisely the structure of a Vandermonde matrix.
View Solution
Solution to question 1:

Since \(f_i^{(k)}(x) = \lambda_i^k e^{\lambda_i x}\), at \(x=0\) we have \(f_i^{(k)}(0) = \lambda_i^k\). The Wronskian matrix at \(x=0\) is therefore:

\[ W[f_1,f_2,f_3](0) = \begin{vmatrix}1 & 1 & 1\\ \lambda_1 & \lambda_2 & \lambda_3\\ \lambda_1^2 & \lambda_2^2 & \lambda_3^2\end{vmatrix}, \]

which is the Vandermonde determinant \(V(\lambda_1,\lambda_2,\lambda_3)\).

Solution to question 2:

By the Vandermonde formula:

\[ \det V(\lambda_1,\lambda_2,\lambda_3) = (\lambda_2-\lambda_1)(\lambda_3-\lambda_1)(\lambda_3-\lambda_2). \]

Since the \(\lambda_i\) are distinct, each factor is nonzero, so \(\det V \neq 0\). Therefore \(W[f_1,f_2,f_3](0) \neq 0\), and \(\{e^{\lambda_1 x}, e^{\lambda_2 x}, e^{\lambda_3 x}\}\) is linearly independent on \(\mathbb{R}\). \(\square\)

Linear Independence in Abstract Vector Spaces

The definition of linear independence applies verbatim in polynomial spaces, matrix spaces, and arbitrary vector spaces: the only linear combination of the elements equal to the zero element is the trivial one. These exercises build proof-writing skills and connect independence to bases, spanning sets, and dimension.

Problem 16: Independence in the Polynomial Space \(\mathcal{P}_2\)

Easy

In the vector space \(\mathcal{P}_2\) of polynomials of degree at most 2, consider \(p_1 = 1 + x\), \(p_2 = 1 – x\), \(p_3 = x^2\).

  1. Write \(c_1 p_1 + c_2 p_2 + c_3 p_3 = 0\) (zero polynomial) and equate coefficients of \(1\), \(x\), and \(x^2\) to zero to solve for \(c_1, c_2, c_3\).
  2. Is \(\{p_1, p_2, p_3\}\) a basis for \(\mathcal{P}_2\)? Justify your answer.
Hint
Expand \(c_1(1+x)+c_2(1-x)+c_3 x^2\) and collect the constant, linear, and quadratic terms separately. Set each to zero.
View Solution
Solution to question 1:
\[ c_1(1+x)+c_2(1-x)+c_3 x^2 = (c_1+c_2)+(c_1-c_2)x+c_3 x^2 = 0. \]

Equating coefficients: \(c_1+c_2=0\), \(c_1-c_2=0\), \(c_3=0\). The first two equations give \(c_1=c_2\) and \(c_1=-c_2\), so \(c_1=c_2=0\). Combined with \(c_3=0\), the only solution is trivial.

Solution to question 2:

\(\{p_1,p_2,p_3\}\) is linearly independent. Since \(\dim(\mathcal{P}_2)=3\) and the set contains 3 independent vectors, it also spans \(\mathcal{P}_2\) and is therefore a basis.

Problem 17: Independence in the Space of \(2\times 2\) Matrices

Medium

In \(M_{2\times 2}(\mathbb{R})\), the space of all \(2\times 2\) real matrices, consider

\[ A_1=\begin{pmatrix}1&0\\0&0\end{pmatrix},\quad A_2=\begin{pmatrix}0&1\\1&0\end{pmatrix},\quad A_3=\begin{pmatrix}0&0\\0&1\end{pmatrix},\quad A_4=\begin{pmatrix}1&0\\0&1\end{pmatrix}. \]

  1. Prove that \(\{A_1, A_2, A_3\}\) is linearly independent.
  2. Determine whether \(\{A_1, A_2, A_3, A_4\}\) is linearly independent. If not, express \(A_4\) as a linear combination of the others.
Hint
For a linear combination of matrices to equal the zero matrix, each entry of the combination must be zero. Use this entry-by-entry condition to determine the coefficients.
View Solution
Solution to question 1:

Suppose \(c_1 A_1+c_2 A_2+c_3 A_3 = \mathbf{0}_{2\times2}\). Equating entries:

\[ \begin{pmatrix}c_1 & c_2\\ c_2 & c_3\end{pmatrix}=\begin{pmatrix}0&0\\0&0\end{pmatrix} \implies c_1=0,\quad c_2=0,\quad c_3=0. \]

Only the trivial solution exists, so \(\{A_1,A_2,A_3\}\) is linearly independent.

Solution to question 2:

\(\{A_1,A_2,A_3,A_4\}\) is linearly dependent since \(\dim(M_{2\times2})=4\) and we can verify directly that \(A_4 = A_1 + A_3\):

\[ A_1+A_3=\begin{pmatrix}1&0\\0&0\end{pmatrix}+\begin{pmatrix}0&0\\0&1\end{pmatrix}=\begin{pmatrix}1&0\\0&1\end{pmatrix}=A_4. \]

The dependence relation is \(A_1 + A_3 – A_4 = \mathbf{0}\).

Problem 18: Independence Is Preserved by Injective Linear Maps

Hard

Let \(T: V \to W\) be a linear transformation with \(\ker(T) = \{\mathbf{0}\}\).

  1. Prove that if \(\{v_1, \ldots, v_k\}\) is linearly independent in \(V\), then \(\{T(v_1), \ldots, T(v_k)\}\) is linearly independent in \(W\).
  2. Show by example that the conclusion can fail when \(\ker(T) \neq \{\mathbf{0}\}\).
Hint
For part (1): suppose a linear combination of the images equals zero in \(W\). Use linearity to pull this back to \(V\), invoke \(\ker(T)=\{\mathbf{0}\}\), then use independence of the \(v_i\).
View Solution
Solution to question 1:

Suppose \(c_1 T(v_1)+\cdots+c_k T(v_k)=\mathbf{0}_W\). By linearity of \(T\):

\[ T\!\left(\sum_{i=1}^k c_i v_i\right) = \mathbf{0}_W. \]

So \(\sum c_i v_i \in \ker(T) = \{\mathbf{0}_V\}\), giving \(c_1 v_1+\cdots+c_k v_k=\mathbf{0}_V\). Since the \(v_i\) are linearly independent in \(V\), all \(c_i=0\). Therefore \(\{T(v_1),\ldots,T(v_k)\}\) is linearly independent. \(\square\)

Solution to question 2:

Let \(T:\mathbb{R}^2\to\mathbb{R}\) be the projection \(T(x,y)=x\). Then \(\ker(T)=\{(0,y)\}\neq\{\mathbf{0}\}\). The set \(\{v_1,v_2\}=\{(1,0),(1,1)\}\) is linearly independent in \(\mathbb{R}^2\), but \(T(v_1)=1=T(v_2)\), so \(\{T(v_1),T(v_2)\}=\{1,1\}\) is a repeated singleton — linearly dependent in \(\mathbb{R}\).

Problem 19: Mutually Orthogonal Nonzero Vectors Are Linearly Independent

Hard

Let \(\{v_1, v_2, \ldots, v_k\}\) be nonzero vectors in an inner product space with \(\langle v_i, v_j\rangle = 0\) for all \(i \neq j\).

  1. Prove that \(\{v_1, \ldots, v_k\}\) is linearly independent.
  2. Explain why the standard basis \(\{e_1, \ldots, e_n\}\) in \(\mathbb{R}^n\) with the dot product is an immediate corollary of this result.
Hint
Begin with \(\sum c_i v_i = \mathbf{0}\). Take the inner product of both sides with a fixed \(v_j\). Orthogonality kills all terms except the \(j\)-th; then use \(\|v_j\|^2 > 0\).
View Solution
Solution to question 1:

Suppose \(c_1 v_1+\cdots+c_k v_k = \mathbf{0}\). Fix any index \(j\) and take the inner product of both sides with \(v_j\):

\[ \left\langle \sum_{i=1}^k c_i v_i,\; v_j\right\rangle = \langle\mathbf{0}, v_j\rangle = 0. \]

By linearity and orthogonality (\(\langle v_i,v_j\rangle=0\) for \(i\neq j\)):

\[ c_j\langle v_j, v_j\rangle = c_j\|v_j\|^2 = 0. \]

Since \(v_j \neq \mathbf{0}\), we have \(\|v_j\|^2 > 0\), so \(c_j = 0\). Since \(j\) was arbitrary, all coefficients are zero and the set is linearly independent. \(\square\)

Solution to question 2:

The standard basis vectors satisfy \(e_i \cdot e_j = \delta_{ij}\), so they are mutually orthogonal and nonzero. The theorem guarantees their linear independence as an immediate special case — no row reduction needed.