Lecture 7 - Linear Combinations and Vector Equations

Learning Objectives

Linear Combinations

Definition. Given vectors \( \bbm v_1, \bbm v_2, \ldots, \bbm v_p \in \mathbb R^n \), a linear combination of these vectors is any vector of the form \( c_1 \bbm v_1 + c_2 \bbm v_2 + \cdots + c_p \bbm v_p, \) where \( c_1, c_2, \ldots, c_p \) are scalars.

Example 1. If \( \bbm u = \vecthree 0 {-3} 5 \) and \( \bbm v = \vecthree 1 4 7 \), then \( 2\bbm u + 3\bbm v = \vecthree 3 6 {31} \) is a linear combination of \( \bbm u \) and \( \bbm v \).

Example 2. If \( \bbm p = \vectwo 1 0 \), \( \bbm q = \vectwo {-2} {-2} \), and \( \bbm r = \vectwo {-4} 8 \), then \( -1\bbm p + 0\bbm q + \frac 1 2 \bbm r = \vectwo {-3} 4 \) is a linear combination of \( \bbm p \), \( \bbm q \), and \( \bbm r \).

Note that any or all of the scalars in the linear combination are allowed to be zero. This means that the zero vector \( \bbm 0 \) is always a linear combination of any given collection of vectors.

Vector Equations

Definition. A vector equation has the form \( x_1 \bbm v_1 + x_2 \bbm v_2 + \cdots + x_p \bbm v_p = \bbm b \), where \( \bbm v_1, \bbm v_2, \ldots, \bbm v_p\) and \( \bbm b\) are vectors in \( \mathbb R^n \). A solution of this vector equation is a value for each variable \( x_1, x_2, \ldots, x_p \) that makes the left-hand side equal the right-hand side.

Example 3. Solve the vector equation \[ x_1 \vecthree 2 {-1} 0 + x_2 \vecthree 1 0 {-3} = \vecthree 4 {-1} {-6}. \]

Working out the vector operations on the left-hand sides gives the equation \( \vecthree {2x_1 + x_2} {-x_1} {-3x_2} = \vecthree 4 {-1} {-6} \), which can be thought of as a system of three linear equations. Solving this system in the normal way gives the solution \( x_1 = 1 \) and \( x_2 = 2 \). \( \Box \)

Example 3 illustrates that we can set up an augmented matrix to solve a vector equation \( x_1 \bbm v_1 + x_2 \bbm v_2 + \cdots + x_p \bbm v_p = \bbm b \). The columns of this augmented matrix are \( \bbm v_1, \bbm v_2, \ldots, \bbm v_p \) and \( \bbm b \).

Equivalent Questions

Consider again the vector equation in Example 3. We have two equivalent questions that arise given this vector equation:

In general, given any vector equation \( x_1 \bbm v_1 + x_2 \bbm v_2 + \cdots + x_p \bbm v_p = \bbm b \), the two equivalent questions are:

If the vector \( \bbm b \) is a linear combination of \( \bbm v_1, \bbm v_2, \ldots, \bbm v_p \), then we can use the scalars from that linear combination as values for the \( x\)'s to solve our equation. If the equation is consistent (meaning that it has at least one solution), then we can use the \( x\)-values from that solution to build the linear combination.

Example 4. Let \( \bbm v_1 = \vecthree 1 {-2} {-5} \), \( \bbm v_2 = \vecthree 2 5 6 \), and \( \bbm b = \vecthree 7 4 {-3} \). Is \( \bbm b\) a linear combination of \( \bbm v_1 \) and \( \bbm v_2 \)?

To solve this problem, we first consider the equivalent question: Does the equation \( x_1 \bbm v_1 + x_2 \bbm v_2 = \bbm b \) have a solution? As we saw in Example 3, we can construct an augmented matrix for this question whose columns are \( \bbm v_1 \), \( \bbm v_2 \), and \( \bbm b \): \[ \begin{bmatrix} 1 & 2 & 7 \\ -2 & 5 & 4 \\ -5 & 6 & -3 \end{bmatrix} \longrightarrow \begin{bmatrix} 1 & 0 & 3 \\ 0 & 1 & 2 \\ 0 & 0 & 0 \end{bmatrix} \]

The row-reduced augmented matrix shows us that, since there is no pivot in the last column, the vector equation is consistent. This tells us that, yes, \( \bbm b \) is a linear combination of \( \bbm v_1 \) and \( \bbm v_2 \). \( \Box \)

Note that, in our solution to Example 4, we actually had more information: we discovered exactly how to construct \( \bbm b \) as a linear combination of \( \bbm v_1 \) and \( \bbm v_2 \). However, the question did not ask us for that information. This illustrates an important idea: Always answer the question you are being asked.

Example 5. Let \( \bbm w_1 = \vecthree 1 {-2} 0 \), \( \bbm w_2 = \vecthree 0 1 1 \), \( \bbm w_3 = \vecthree 2 {-7} {-3} \), and \( \bbm b = \vecthree {-1} 0 3 \). Is \( \bbm b \) a linear combination of \( \bbm w_1 \), \( \bbm w_2 \), and \( \bbm w_3 \)?

To answer this question, we consider whether the equation \( x_1 \bbm w_1 + x_2 \bbm w_2 + x_3 \bbm w_3 = \bbm b \) is consistent. To solve this equation, we set up the corresponding augmented matrix and row-reduce: \[ \begin{bmatrix} 1 & 0 & 2 & -1 \\ -2 & 1 & -7 & 0 \\ 0 & 1 & -3 & 3 \end{bmatrix} \longrightarrow \begin{bmatrix} 1 & 0 & 2 & 0 \\ 0 & 1 & -3 & 0 \\ 0 & 0 & 0 & 1 \end{bmatrix} \]

Since there is a pivot in the last column of our augmented matrix, the vector equation has no solutions. This means that, no, \( \bbm b \) is not a linear combination of \( \bbm w_1 \), \( \bbm w_2 \), and \( \bbm w_3 \). \( \Box \)

Writing Good Solutions

The previous examples illustrate a good format for your solutions that you should emulate in your own writing. Specifically, make sure to include the following elements in your written solutions:

  1. Introduction. Rewrite the given question as an equivalent question about a vector equation.
  2. Row-Reduction. Construct and row-reduce the augmented matrix corresponding to the vector equation.
  3. Conclusion. Based on your row-reduced matrix, draw a conclusion about the vector equation, and use that to answer the original question.

Many students, in their initial attempts at these kinds of solutions, will omit items (1) and (3) above and only show a row-reduction process. It is very important to explain what that matrix represents, and how the result of the row-reduction provides information about the question being considered. Look back over the examples you have seen and adapt your writing style to make sure that you include these essential elements.

« Lecture 6 Back to Top Lecture 8 »