Skip to content

3.5-exam1-2007-johnson

Quiz 1, 2007 (johnson) · Section 3.5: Dimensions of the Four Subspaces Source: exam1_1.pdf, pp. 180-182

Problem

4 (30 pts.) The following information is known about an \(m \times n\) matrix \(A\):

\[ A \begin{bmatrix} 1 \\ -2 \\ 0 \\ 1 \end{bmatrix} = \begin{bmatrix} 2 \\ 4 \end{bmatrix}, \qquad A \begin{bmatrix} 0 \\ 2 \\ 1 \\ 3 \end{bmatrix} = \begin{bmatrix} 0 \\ 0 \end{bmatrix}, \qquad A \begin{bmatrix} 2 \\ 0 \\ 0 \\ 1 \end{bmatrix} = \begin{bmatrix} 5 \\ 10 \end{bmatrix}, \qquad A \begin{bmatrix} 3 \\ 2 \\ 0 \\ 0 \end{bmatrix} = \begin{bmatrix} 1 \\ 2 \end{bmatrix}. \]

\((\alpha)\) Show that the vectors \(\begin{bmatrix} 1 \\ -2 \\ 0 \\ 1 \end{bmatrix}, \begin{bmatrix} 0 \\ 2 \\ 1 \\ 3 \end{bmatrix}, \begin{bmatrix} 2 \\ 0 \\ 0 \\ 1 \end{bmatrix}, \begin{bmatrix} 3 \\ 2 \\ 0 \\ 0 \end{bmatrix}\) form a basis of \(\mathbb{R}^4\).

\((\beta)\) Give a matrix \(C\) and an invertible matrix \(B\) such that \(A = CB^{-1}\). (You don't have to evaluate \(B^{-1}\) or find \(A\) explicitly. Just say what \(B\) and \(C\) are and use them to reason about \(A\) in the subsequent parts.)

\((\gamma)\) Find a basis for the null space of \(A^T\).

\((\delta)\) What are \(m\), \(n\), and the rank \(r\) of \(A\)?

Solution \((\alpha)\) We are in \(\mathbb{R}^4\), which is four-dimensional, so any four linearly independent vectors form a basis, as shown in class. Thus, we just need to show that these four vectors are linearly independent, which is equivalent to showing that the \(4 \times 4\) matrix whose columns are these vectors has full column rank (null space = \(\{0\}\)). Proceeding by elimination: \[ B = \begin{bmatrix} 1 & 0 & 2 & 3 \\ -2 & 2 & 0 & 2 \\ 0 & 1 & 0 & 0 \\ 1 & 3 & 1 & 0 \end{bmatrix} \to \begin{bmatrix} 1 & 0 & 2 & 3 \\ 0 & 2 & 4 & 8 \\ 0 & 1 & 0 & 0 \\ 0 & 3 & -1 & -3 \end{bmatrix} \to \begin{bmatrix} 1 & 0 & 2 & 3 \\ 0 & 2 & 4 & 8 \\ 0 & 0 & -2 & -4 \\ 0 & 0 & -7 & -15 \end{bmatrix} \to \begin{bmatrix} 1 & 0 & 2 & 3 \\ 0 & 2 & 4 & 8 \\ 0 & 0 & -2 & -4 \\ 0 & 0 & 0 & -1 \end{bmatrix} = U. \] Thus there are four pivots, and hence \(B\) has full column rank as desired. \((\beta)\) The provided equations multiply \(A\) by four vectors to get four vectors, which by definition of matrix multiplication (recall the column picture) can be combined into a single equation where \(A\) is multiplied by a matrix with four columns to yield a matrix with four columns: \[ A \begin{bmatrix} 1 & 0 & 2 & 3 \\ -2 & 2 & 0 & 2 \\ 0 & 1 & 0 & 0 \\ 1 & 3 & 1 & 0 \end{bmatrix} = \begin{bmatrix} 2 & 0 & 5 & 1 \\ 4 & 0 & 10 & 2 \end{bmatrix}. \] Thus if we take \[ C = \begin{bmatrix} 2 & 0 & 5 & 1 \\ 4 & 0 & 10 & 2 \end{bmatrix} \qquad \text{and} \qquad B = \begin{bmatrix} 1 & 0 & 2 & 3 \\ -2 & 2 & 0 & 2 \\ 0 & 1 & 0 & 0 \\ 1 & 3 & 1 & 0 \end{bmatrix}, \] we have \(A = CB^{-1}\). Since \(B\) is precisely the matrix of the basis vectors from part \((\alpha)\), its invertibility follows from above (it is \(4 \times 4\) and has 4 pivots). \((\gamma)\) Since \(A = CB^{-1}\), we have \[ A^T = (B^{-1})^T C^T = (B^T)^{-1} C^T. \] (As in class, because \(B\) is invertible, \(B^T\) is too.) Just as for elimination (multiplying on the left by an invertible elimination matrix), the null space is preserved when \(C^T \to (B^T)^{-1} C^T\). [You need not prove this, because the proof is the same as in class. Recall that if \(C^T x = 0\) then \(A^T x = 0\) from above, and vice versa if we multiply both sides by \(B^T\).] That means we just need to find the null space of \(C^T\) by elimination: \[ \begin{bmatrix} 2 & 4 \\ 0 & 0 \\ 5 & 10 \\ 1 & 2 \end{bmatrix} \to \begin{bmatrix} 2 & 4 \\ 0 & 0 \\ 0 & 0 \\ 0 & 0 \end{bmatrix}, \] in which there is only one free variable, so there is one special solution (the basis of the null space) \[ s_1 = \begin{bmatrix} -2 \\ 1 \end{bmatrix}, \] or any multiple thereof. (You can also find this special solution by inspection, without elimination.) \((\delta)\) Since \(A\) times a 4-vector is a 2-vector, we must have \(m = 2\) and \(n = 4\). Equivalently, from part \((\beta)\) we saw that \(A\) was a \(2 \times 4\) matrix multiplied by a \(4 \times 4\) matrix, giving a \(2 \times 4\) matrix. Moreover, from above the dimension of \(N(A^T)\) is 1, but this must equal \(m - r\), so we obtain \(r = 1\).

Keywords: basis, rank, nullspace, column space, invertible matrix