Notes on Homework 10/Practice Exam 2 Due: Never 1. (a) A subspace S of a vector space V is a nonempty subset of V that is closed under addition and scalar multiplication. (b) A linear transformation T : V → W is a function from vector space V to vector space W such that, for any v1 , v2 ∈ V, T (v1 + v2 ) = T (v1 ) + T (v2 ) and, for any α in the field, T (α v1 ) = α T (v1 ). (c) An eigenvalue λ is a constant associated to a matrix A such that there is a vector v with Av = λv. (d) The image of a linear transformation T : V → W is {w ∈ W |T (v) = w for some v ∈ V } (e) A linear transformation is injective if T (v1 ) = T (v2 ) necesssarily means that v1 = v2 . Alternatively, T is injective if for all w ∈ W, there is no more than one v ∈ V such that T (v) = w. (f) A set of vectors B is a spanning set for vector space V if v is in the span of B for any v ∈ V. (g) A basis of a vector space is just a linearly independent spanning set. 2. Nonempty, closed under addition, and closed under scalar multiplication. 3. Simply put, T (v1 + v2 ) = T (v1 ) + T (v2 ) and T (α v) = α T (v). 4. Here are eight (plus a couple extras): • A is not invertible. • A has 0 as an eigenvalue. • RREF ( A) is not an identity matrix. • The nullspace of A contains more than just the zero vector. • The columns of A are linearly dependent. • The columns of A do not form a basis for the domain. • The rank of A is less than the number of columns of A. • The nullity of A is positive. • det( A) = 0. • Ax = b does not have a unique solution for every choice of b. 5. We need to show that three properties hold (as described in #2): Nonempty: The polynomial p( z) = 0 is certainly in S since p(0) = 0. X Closed under addition: Let g( z), h( z) ∈ S. Then ( g + h)( z) is certainly a polynomial, and ( g + h)(0) = g(0) + h(0) = 0 + 0 = 0, by definition of function addition. So ( g + h)( z) ∈ S. X Professor Dan Bates Colorado State University M369 Linear Algebra Fall 2008 Closed under scalar multiplication: Let g( z) ∈ S, α ∈ F. (α g)( z) is certainly a polynomial, and (α g)(0) = α g(0) = α · 0 = 0. So (α g)( z) ∈ S. X 6. 7. (a) We need to show that two properties hold (as desribed in #3): x1 + y1 y1 x1 Let x = x2 , y = y2 ∈ R3 . Then T ( x + y) = T x2 + y2 x + y3 y3 x3 3 ( x1 + y1 ) − ( x2 + y2 ) x1 − x2 y1 − y2 = = + = T ( x) + T ( y).X 2( x3 + y3 ) + ( x2 + y2 ) 2x3 + x2 2y3 + y2 α x1 x1 − x2 α x1 − α x2 Also, let α ∈ F. Then T (α x) = T α x2 = =α = 2x3 + x2 2α x3 + α x2 α x3 α T ( x).X 1 −1 0 . (b) By inspection, [T ]ε←ε = 0 1 2 (c) A linear transformation is injective if and only if the kernel (a.k.a. nullspace) is trivial, 1 0 2 i.e., just the zero vector. Row-reducing our matrix, we get to . The kernel is 0 1 2 nontrivial,so the is not injective. Alternatively, you could notice transformation linear 1 0 0 . that both 1 and 0 map to 0 1 0 −2 1−x 5 (a) p A ( x) = det = (1 − x)2 − 25 = x2 − 2x − 24 = ( x − 6)( x + 4). 5 1−x (b) The eigenvalues of A are the roots of p A ( x): x = −4, 6. 5 5 5 5 . (c) For λ = −4, we have ε A (−4) = null( A − λ I ) = null( A + 4I ) = null 1 1 −1 Row-reducing, we get to , so ε A (−4) = span . 0 0 1 −5 5 For λ = 6, we have ε A (6) = null( A − λ I ) = null ( A − 6I ) = null . 5 −5 1 1 −1 . Row-reducing, we get to , so ε A (6) = span 1 0 0 (d) Yes, A is diagonalizable. We know this because, for each eigenvalue, the algebraic multiplicity and the geometric multiplicity are the same. (e) Yes, A is invertible, since 0 is not an eigenvalue. 8. (a) Professor Dan Bates Colorado State University M369 Linear Algebra Fall 2008 (b) Let v1 = 1 2 , v2 = 3 0 . Then [id ]E ←B = ([id(v1 )]E [id (v2 )]E ) 1 3 = 2 0 so that [id ]B←E = ([id]E ←B )−1 1 0 2 = 1 1 . 3 −6 (c) Then [ f ]B←B = [id]B←E [ f ]E ←E [id ]E ←B 1 1 5 0 1 3 2 = 1 1 5 1 2 0 3 −6 7 15 = 25 −23 . 2 9. 2 (a) True. f is injective if and only if the nullity is 0. However, the rank-nullity theorem tells us that rank( f )+nullity( f ) = dim (W ) (the domain). However, since image( f ) ⊂ V, rank( f ) =dim(im( f )) ≤dimV <dimW, so nullity( f ) > 0. 1 0 1 (b) False. As a simple counterexample, take , , ∈ R2 . The first two are 0 1 1 linearly independent, but the three together certainly are not! (c) True. Since the first n vectors span V, any v ∈ V can be written as a linear combination of these n vectors. This linear combination can be extended to all n + 1 vectors simply by using the coefficient of 0 for vn+1 . Thus, any vector v ∈ V can be written as a linear combination of all n + 1 vectors, i.e., they span V. (d) False. We had a homework problem about this. Recall the example of two lines in R3 . That example is closed under scalar multiplication but definitely not under addi 0 1 tion. In particular, take the spans of 0 and 1 in R3 as U and W, respectively. 0 0 1 0 1 0 1 0 , 1 ∈ U ∪ W, but 0 + 1 = 1 6∈ U ∪ W. 0 0 0 0 0 Professor Dan Bates Colorado State University M369 Linear Algebra Fall 2008