Skip to content

8.2-final-2000

Final Exam, 2000 · Section 8.2: The Matrix of a Linear Transformation Source: finalexam_1.pdf, p. 49

Problem

6. Suppose \(A\) is a 3 by 3 matrix and you know the three outputs \(y_1 = Ax_1\) and \(y_2 = Ax_2\) and \(y_3 = Ax_3\) from three independent input vectors \(x_1, x_2, x_3\).

(a) (4 points) Find the matrix \(A\) using this hint: Put the vectors \(x_1, x_2, x_3\) into the columns of a matrix \(X\) and multiply \(AX\). Why did I require the \(x\)'s to be independent?

(b) (4 points) Under what condition on \(A\) will the outputs \(y_1, y_2, y_3\) be a basis for \(R^3\)? Explain your answer.

(c) (4 points) If \(x_1, x_2, x_3\) is the input basis and \(y_1, y_2, y_3\) is the output basis, what is the matrix \(M\) that represents this same linear transformation (defined by \(T(x_1) = y_1, T(x_2) = y_2, T(x_3) = y_3\))?

Solution (a) With \(X = [x_1 \ x_2 \ x_3]\) and \(Y = [y_1 \ y_2 \ y_3]\) we have \(AX = Y\), so \(A = YX^{-1}\). The \(x\)'s are required to be independent so that \(X\) is invertible and \(A\) is uniquely determined. (b) The \(y\)'s form a basis of \(R^3\) exactly when \(A\) is invertible: \(\det Y = \det(AX) = (\det A)(\det X)\), and \(\det X \neq 0\), so \(\det Y \neq 0\) iff \(\det A \neq 0\). (c) The matrix is the identity, \(M = I\): the transformation sends the \(i\)-th input basis vector \(x_i\) to the \(i\)-th output basis vector \(y_i\), so its matrix in the input/output bases has column \(i\) equal to the \(i\)-th standard basis vector.

Keywords: linear transformation, matrix of transformation, basis, change of basis