Skip to content

5.2-exam2-2015

Exam II, 2015 · Section 5.2: Computing and Using Determinants Source: exam2_2.pdf, pp. 98, 101

Problem

3. (34 points)

(a) If \(q_1, q_2, q_3\) are orthonormal vectors in \(R^3\), what are the possible determinants of this matrix \(A\) with columns \(2q_1\) and \(3q_2\) and \(5q_3\)? Why?

\[ A = \left[ \begin{array}{ccc} 2q_1 & 3q_2 & 5q_3 \end{array} \right] \]

(b) For a matrix \(A\), suppose the cofactor \(C_{11}\) of the first entry \(a_{11}\) is zero. What information does that give about \(A^{-1}\)? Can this inverse exist?

(c) Find the 3 eigenvalues of this matrix \(A\) and find all of its eigenvectors. Why is the diagonalization \(S^{-1}AS = \Lambda\) not possible?

\[ A = \left[ \begin{array}{ccc} 2 & 1 & 0 \\ 0 & 2 & 0 \\ 0 & 0 & 2 \end{array} \right] \]
Solution (a) The determinant of the matrix \(Q = [q_1 \ q_2 \ q_3]\) has to be 1 or \(-1\). This is because \(Q^TQ = I\), which means that \(\det Q^T \cdot \det Q = 1\), that is, \((\det Q)^2 = 1\). When we multiply a column by a number, the determinant is multiplied by the same number. Thus, the determinant of \(A\) is either 30 or \(-30\). (b) The cofactor \(C_{11}\) being zero does not give us enough information to decide whether the inverse exists or not. For example, in the matrix \(\begin{bmatrix} 1 & 0 \\ 0 & 0 \end{bmatrix}\) this cofactor is zero and the inverse does not exist, and in the matrix \(\begin{bmatrix} 1 & 1 \\ 1 & 0 \end{bmatrix}\) this cofactor is zero and the inverse exists. If this inverse exists, then we know that the entry \((1, 1)\) in this inverse is zero. (c) The eigenvalues of this matrix are \((2, 2, 2)\). But the rank of \(A - 2I\) is 1. That means you can only find two independent eigenvectors. When the number of independent eigenvectors is smaller than the size of the matrix, then the diagonalization is not possible because you cannot build the square matrix of eigenvectors \(S\).

Keywords: determinant, orthonormal vectors, cofactors, eigenvalues, diagonalization