3.5-final-2005-strang-b¶
Final Exam, 2005 (strang) · Section 3.5: Dimensions of the Four Subspaces Source: finalexam_1.pdf, p. 130
Problem¶
9 (10 pts.) (a) Describe all vectors that are orthogonal to the nullspace of this singular matrix \(A\). You can do this without computing the nullspace.
\[
A = \begin{bmatrix} 1 & 3 & 7 \\ 2 & 2 & 6 \\ 2 & 1 & 4 \end{bmatrix}.
\]
(b) If you apply Gram-Schmidt to the columns of this \(A\), what orthonormal vectors do you get?
(c) Find a "reduced" \(LU\) factorization of \(A\), with only 2 columns in \(L\) and 2 rows in \(U\). Can you write \(A\) as the sum of two rank 1 matrices?
Solution
(a) The vectors orthogonal to the nullspace of \(A\) are the rows of \(A\). Since we know that the matrix \(A\) is singular and it is clearly not rank one, it follows that the rank of \(A\) is two. The first two rows are independent and therefore the orthogonal complement of the nullspace of \(A\) is spanned by the two vectors \[ \begin{bmatrix} 1 \\ 3 \\ 7 \end{bmatrix} \quad \text{and} \quad \begin{bmatrix} 2 \\ 2 \\ 6 \end{bmatrix}. \] (b) We get the vectors \[ \frac{1}{3} \begin{bmatrix} 1 \\ 2 \\ 2 \end{bmatrix}, \quad \frac{1}{\sqrt{5}} \begin{bmatrix} 2 \\ 0 \\ -1 \end{bmatrix} \quad \text{and} \quad \begin{bmatrix} 0 \\ 0 \\ 0 \end{bmatrix}. \] (c) The "reduced" \(LU\) decomposition, from ignoring the zero row in \(U\), is \[ A = \begin{bmatrix} 1 & 0 \\ 2 & 1 \\ 2 & \frac{5}{4} \end{bmatrix} \begin{bmatrix} 1 & 3 & 7 \\ 0 & -4 & -8 \end{bmatrix}. \] Here are the details: Starting elimination we find \[ \begin{bmatrix} 1 & 0 & 0 \\ -2 & 1 & 0 \\ -2 & 0 & 1 \end{bmatrix} \begin{bmatrix} 1 & 3 & 7 \\ 2 & 2 & 6 \\ 2 & 1 & 4 \end{bmatrix} = \begin{bmatrix} 1 & 3 & 7 \\ 0 & -4 & -8 \\ 0 & -5 & -10 \end{bmatrix}, \] and proceeding further we find \[ \begin{bmatrix} 1 & 0 & 0 \\ 0 & 1 & 0 \\ 0 & -\frac{5}{4} & 1 \end{bmatrix} \begin{bmatrix} 1 & 3 & 7 \\ 0 & -4 & -8 \\ 0 & -5 & -10 \end{bmatrix} = \begin{bmatrix} 1 & 3 & 7 \\ 0 & -4 & -8 \\ 0 & 0 & 0 \end{bmatrix}. \] Collecting all the information together we obtain \[ \begin{bmatrix} 1 & 3 & 7 \\ 2 & 2 & 6 \\ 2 & 1 & 4 \end{bmatrix} = \begin{bmatrix} 1 & 0 & 0 \\ 2 & 1 & 0 \\ 2 & 0 & 1 \end{bmatrix} \begin{bmatrix} 1 & 0 & 0 \\ 0 & 1 & 0 \\ 0 & \frac{5}{4} & 1 \end{bmatrix} \begin{bmatrix} 1 & 3 & 7 \\ 0 & -4 & -8 \\ 0 & 0 & 0 \end{bmatrix}, \] and multiplying the first two matrices on the right-hand side we deduce that \[ \begin{bmatrix} 1 & 3 & 7 \\ 2 & 2 & 6 \\ 2 & 1 & 4 \end{bmatrix} = \begin{bmatrix} 1 & 0 & 0 \\ 2 & 1 & 0 \\ 2 & \frac{5}{4} & 1 \end{bmatrix} \begin{bmatrix} 1 & 3 & 7 \\ 0 & -4 & -8 \\ 0 & 0 & 0 \end{bmatrix}. \] Since the last row of the last matrix is all zero, we conclude that \[ \begin{bmatrix} 1 & 3 & 7 \\ 2 & 2 & 6 \\ 2 & 1 & 4 \end{bmatrix} = \begin{bmatrix} 1 & 0 \\ 2 & 1 \\ 2 & \frac{5}{4} \end{bmatrix} \begin{bmatrix} 1 & 3 & 7 \\ 0 & -4 & -8 \end{bmatrix}. \] This is the "reduced" \(LU\) factorization of \(A\). Multiplying columns of \(L\) by rows of \(U\), this is \[ A = \begin{bmatrix} 1 \\ 0 \\ 0 \end{bmatrix} \begin{bmatrix} 1 & 3 & 7 \end{bmatrix} + \begin{bmatrix} 0 \\ 1 \\ \frac{5}{4} \end{bmatrix} \begin{bmatrix} 0 & -4 & -8 \end{bmatrix}. \]Keywords: row space, Gram-Schmidt, reduced LU, rank one