MATH 323.501 Exam 2 Solutions November 12, 2013 1. For each statement below, write down whether it is true or false. (a) The rank of a matrix B plus the dimension of the null space of B is equal to the number of rows of B. (b) If L : V → W is a linear transformation, then the kernel of L is trivial if and only if L is surjective. (c) Suppose V1 and V2 are subspaces of R4 with dim(V1 ) = 2 and dim(V2 ) = 3. Then dim(V1 ∩ V2 ) is either 1 or 2. (d) Suppose that v1 , v2 , v3 , v4 are linearly dependent vectors in Rn . Then we have Span(v1 , v2 , v3 , v4 ) = Span(v2 , v3 , v4 ). (e) If λ is an eigenvalue of a matrix A, then A2 − λA is singular. Solutions: (a) False : The rank plus the nullity is equal to the number of columns. (b) False : The kernel is trivial if and only if L is injective. (c) True : Since V1 ∩ V2 ⊂ V1 , and dim(V1 ) = 2, we know that dim(V1 ∩ V2 ) is either 0, 1, or 2. We need to rule out that the dimension is 0. Let {v1 , v2 } be a basis of V1 and let {w1 , w2 , w3 } be a basis of V2 . There is a non-trivial linear dependcy among these 5 vectors. Use this linear dependency to show that there is a non-zero vector in V1 ∩ V2 . (d) False : We know that at least one of v1 , . . . , v4 is redundant when taking the span, but we cannot say that it is definitly v1 . (e) True : Since A2 − λA = A(A − λI), it follows that det(A2 − λA) = det(A) det(A − λI). Since λ is an eigenvalue, we have det(A − λI) = 0, and so det(A2 − λA) = 0. 2. Let L : R2 → R2 be defined by x1 x 2 − x1 L = . x2 2 − x2 Is L a linear transformation? Justify your answer. Solution: It is not a linear transformation . The reason is that it fails to respect both addition and scalar multiplication. We only need to exhibit a counterexample to one of these, so we compute 1 3 0 L 3 =L = 1 3 −1 and 1 0 0 3L =3 = . 1 1 3 1 Since these two values are not the same, we see that L does not respect scalar multiplication. Answers may vary of course, and it is also possible to provide a counterexample for addition. 3. Let A be the matrix 1 1 −2 0 A = 3 5 4 2 , 0 1 5 1 and let V be the column space of A. (a) Are the columns of A linearly independent or linearly dependent? Justify your answer. (b) Find a basis for V . Show your work. (c) What is the rank of A? Explain. Solution: (a) The columns are linearly dependent . The reason is that each column represents a vector in R3 . Thus we are considering 4 vectors in this 3-dimensional space, so they must be linearly dependent. (b) We perform row operations on A 1 0 0 and find that its reduced row echelon form is 1 −2 0 1 5 1 . 0 0 0 The leading 1’s appear in the first two columns, so the first two columns of A will n o 1 1 3 , 5 make a basis for the column space of A. Thus is a basis. 0 1 (c) The rank of A is 2 . We know that the rank of a matrix is the same as the dimension of its column space (or the dimension of its row space, or the number of non-zero rows when it is put in row echelon form). 4. 1 Let A = {( 11 ) , ( −1 )} ⊆ R2 . We can take it as given that A is a basis for R2 . Suppose that B is another basis of R2 with change of basis matrix 2 5 P = . 1 3 That is, for every v ∈ R2 , we have [v]A = P [v]B . (a) Find a matrix Q such that for every v ∈ R2 , [v]B = Q [v]A . (b) Suppose that L : R2 → R2 is a linear transformation and that with respect to the 1 −2 B basis A we have [L]A A = −3 4 . Find [L]B . (c) What are the vectors in B with respect to the standard coordinates on R2 ? Solution: (a) If we take the equation [v]A = P [v]B given and multiply through on the left by P −1 , we see that P −1 [v]A = [v]B . Thus Q = P −1 , which we calculate as 3 −5 Q = −1 . 2 2 (b) The matrix P is the change of basis matrix from basis B into basis A. Thus B P = [I]A B . Likewise, Q = [I]A . Thus A A [L]BB = [I]BA [L]A A [I]B = Q [L]A P. We calculate and find [L]BB = 3 −5 1 −2 2 5 10 12 = . −1 2 −3 4 1 3 −4 −5 (c) If we let V = [I]SA be the transition matrix from A to the standard basis and let U = [I]SB be the transition matrix from B to the standard basis, then A S −1 P = [I]A U. B = [I]S [I]B = V Multiplying through by V on the left we see that, U = V P, and the columns of U will be the vectors in B with respect to the standard basis. Now 1 V = ( 11 −1 ), since those are the coordinates of the basis vectors in A with respect to the standard coordinates. Therefore, 1 1 2 5 3 8 U= = , 1 −1 1 3 1 2 and so with respect to the standard basis B = {( 31 ) , ( 82 )} . 5. Suppose V is a vector space and that {u, v, w} is a basis for V . Prove that {u + 2v + 3w, 4v + 5w, 6w} is also a basis for V . Solution: Since {u, v, w} contains 3 elements and is a basis for V , the dimension of V is 3. Since dim(V ) = 3, any other set of 3 linearly independent vectors in V will also be a basis. Thus it suffices to show that u + 2v + 3w, 4v + 5w, and 6w are linearly independent. Suppose that we have constants a, b, c ∈ R such that a(u + 2v + 3w) + b(4v + 5w) + c(6w) = 0. Then by manipulating the left-hand side and gathering terms we have au + (2a + 4b)v + (3a + 5b + 6c)w = 0. Since u, v, w are linearly independent, this implies that a = 0, 2a + 4b = 0, 3a + 5b + 6c = 0. As a = 0, the second equation implies that b = 0 also. Then using the third equation, a = b = 0 implies that c = 0. Therefore, u + 2v + 3w, 4v + 5w, and 6w are linearly independent. 3 6. Consider the matrix 1 0 −2 A = 2 0 −4 . −1 0 2 (a) Find the eigenvalues of A. (b) For each eigenvalue λ of A, find vectors that span the eigenspace corresponding to λ. Solution: (a) We calculate 1−λ 0 −2 −λ −4 = (1 − λ)((−λ)(2 − λ) − 2(−λ)) = −λ2 (λ − 3). det 2 −1 0 2−λ Therefore the eigenvalues are λ = 0, 3 . (b) λ = 0: To calculate the eigenspace E0 for λ = 0, we compute the null space of A itself. The reduced row echelon form of A is 1 0 −2 0 0 0 , 0 0 0 and so the defining equation of the nullspace is x1 = 2x3 . From this we see that n o 2 0 0 , 1 E0 = Span . 1 0 λ = 3: To calculate the eigenspace E3 for λ = 3, −2 0 2 −3 A − 3I = −1 0 we compute the null sapce of −2 −4 , −1 whose reduced row echelon form is 1 0 1 0 1 2 . 0 0 0 Thus the defining equations of E3 are x1 = −x3 and x2 = −2x3 , so E3 = Span 4 n −1 o −2 1 .