3.5-exam1-2015-b¶
Exam I, 2015 · Section 3.5: Dimensions of the Four Subspaces Source: exam1_2.pdf, pp. 82, 86-87
Problem¶
2 (32 points) (a) Suppose the matrices \(A\) and \(B\) have the same column space. Give an example where \(A\) and \(B\) have different nullspaces — or say why this is impossible. (b) Again \(A\) and \(B\) have the same column space. Give an example where \(A\) and \(B\) have different ranks \(r\) — or say why this is impossible. (c) CIRCLE True or False: If \(B\) is a square matrix then \(C(B) = C(B^T)\). (d) If the columns of a 5 by 3 matrix \(M\) are linearly independent and \(x\) in \(R^3\) is not the zero vector, then you know that \(Mx\) is ____. (I am looking for an answer that uses independence of columns and \(x \neq 0\).)
Solution
(a) The simplest way to provide an example is to add dependent columns to matrix \(A\). For example, the matrices \([1]\) and \([1 \quad 0]\) have the same column space (both span \(R^1\)) and different nullspaces. (b) Impossible: the rank is the dimension of the column space, so if \(A\) and \(B\) have the same column space, the rank is the same for both matrices. (c) False. Consider \(B = \left[ \begin{smallmatrix} 0 & 1 \\ 0 & 0 \end{smallmatrix} \right]\). Then the column space of \(B\) is spanned by \((1, 0)^T\), while the column space of \(B^T\) is spanned by \((0, 1)^T\). (d) \(Mx\) is a nonzero vector in \(R^5\). If \(Mx = 0\) with \(x \neq 0\), the columns of \(M\) would be linearly dependent; since they are independent, \(Mx \neq 0\).Keywords: column space, nullspace, rank, independence