Math 2040: Matrix Theory and Linear Algebra II Solutions to Assignment 4 5.5 Complex Eigenvalues 5.5.4.

Problem Restatement: Find the eigenvalues and a basis of the eigenspace in C2 of A = 5 2 . 1 3 Final Answer: The complex eigenvalues are = 4 + i and = 4 i. A basis of the eigenspace 1 1 corresponding to = 4 + i is { + i}, and a basis of the eigenspace corresponding 1 0 1 1 to = 4 i is { i}. 1 0 Work: The characteristic polynomial is det(A I ) = det 5 2 = (5 )(3 ) + 1 3 2 = 2 8 + 17. The roots of the characteristic polynomial are the complex eigenvalues 5 (4 + i) 2 = (8 + 64 68)/2 = 4 + i and = 4 i. A I = = 1 3 (4 + i) 1 i 2 z1 . If is a complex eigenvector corresponding to = 4 + i then 1 1 i z2 1 i 2 z1 0 = , so (1 i)z1 2z2 = 0 and z1 +(1 i)z2 = 0. From the 1 1 i z2 0 z1 1+i 1 1 second equation, we get = z2 = z2 ( + i) with z2 a free variable. z2 1 1 0 1 1 Therefore, a basis of the eigenspace corresponding to = 4 + i is { + i}. To get 1 0 a basis of the eigenspace corresponding to = 4 i, we take the complex conjugate, giving 1 1 us { i} (we can do this because, if v is a complex eigenvector corresponding 1 0 to then v is a complex eigenvector corresponding to ). Check: We will test Av = (4 + i)v for v = 1 + 1 1 (4 + i)v = (4 + i)( + 1 Av = 5 2 1 3 5 2 1 3 1 i) = ( 0 1 1 + 5 1 4 0 1 0 i. The left hand side is

1 3 i = + 0 4 4 1 )+( 4 0
1

i and the right hand side is + 1 )i = 1 3 4 + 5 1 i.

The computed left and right hand sides show Av = (4 + i)v holds. Thus, the check is successful, verifying that v is indeed a complex eigenvector corresponding to = 4 + i. Since complex conjugation preserves eigenvectors (that is, v is an eigenvector corresponding to = 4 i), we do not need to check the conjugate eigenvector. 3 3 5.5.8. Problem Restatement: List the eigenvalues of A = . The transformation 3 3 x Ax A is the composition of a rotation and a scaling. Find the angle of rotation , where , and nd the scaling factor r . Final Answer: The eigenvalues are = 3 3 i and = 3 + 3i. The scaling factor is r = 3 + 9 = 2 3, and = /3 since cos = 3/r = 1/2 and sin = 3/r = 3/2. Work: None required. 5.5.16. Problem Restatement: Find an invertible matrix P and a matrix C of the form such that A = P CP 1 , where A = a b b a

5 2 . Use the information found in exercise 5.5.4. 1 3 4 1 and P = 1 4 1 1 and P = .) 1 0 1 1 . (Alternatively, 1 0

Final Answer: With = 1 + i, we get C = with = 1 i, we get C = 4 1 1 4

Work: We refer to the work in the solution to 5.5.4. = 4 + i, so we set C =

4 1 . We 1 4 1 1 use the eigenvector corresponding to = 4 i to construct P . Therefore, P = . 1 0 3 5 Checks: AP = = P C . det(P ) = 1 = 0, so P is invertible. This gives us 4 1 A = P CP 1 , as required.

5.5.22. Problem Restatement: Let A be a complex (or real) n n matrix, and let x Cn be an eigenvector corresponding to an eigenvalue in C. Show that for every nonzero complex scalar , the vector x is an eigenvector of A. Final Answer: Let be any nonzero complex scalar, and let y = x. y = 0 because = 0 and x = 0. Let be the complex eigenvalue corresponding to x. We have Ay = A(x) = Ax = (x) = (x) = y . Therefore, y is an eigenvector of A corresponding to the eigenvalue . Work: None required.

6.1 Inner Product, Length, and Orthogonality 6 6.1.10. Problem Restatement: Find a unit vector in the direction of v = 4 . 3 6/61 Final Answer: A unit vector in the direction of v is u = v/||v || = 4/61 . 3/ 61 Work: ||v || = (6)2 + 42 + (3)2 = 36 + 16 + 9 = 61. 0 4 6.1.14. Problem Restatement: Find the distance between u = 5 and z = 1 . 2 8 Final Answer: The distance between u and z is ||u z || = 2 17. Work: ||u z || = (0 (4))2 + (5 (1))2 + (2 8)2 = 16 + 16 + 36 = 68 = 2 17. 3 1 7 and z = 8 are orthogonal. 6.1.18. Problem Restatement: Determine if y = 4 15 0 7 Final Answer: y and z are not orthogonal, since y z = 0. Work: y z = 3(1) + 7(8) + 4(15) + 0(7) = 1 = 0. 6.1.24. Problem Restatement: Verify the parallelogram law for vectors u and v in Rn : ||u + v ||2 + ||u v ||2 = 2||u||2 + 2||v ||2 . Final Answer: See work below. Work: Using the properties of dot product, we have ||u + v ||2 + ||u v ||2 = (u + v ) (u + v ) + (u v ) (u v ) = u u + u v + v u + v v + u u + u (v ) + (v ) u + (v ) (v ) =uu+uv+vu+vv+uuuvvu+vv = 2(u u) + 2(v v ) = 2||u||2 + 2||v ||2 .

6.2 Orthogonal Sets 1 0 5 6.2.2. Problem Restatement: Determine if { 2 , 1 , 2 } is an orthogonal set of 1 2 1 vectors. Final Answer: Yes, this is an orthogonal set of vectors. 0 1 Work: 2 1 = 1(0) + (2)1 + 1(2) = 0 2 + 2 = 0, 1 2 5 1 2 2 = 1(5) + (2)(2) + 1(1) = 5 + 4 + 1 = 0, and 1 1 0 5 1 2 = 0(5) + 1(2) + 2(1) = 0 2 + 2 = 0. 2 1 5 4 3 4 1 3 6.2.6. Problem Restatement: Determine if { 0 , 3 , 5 } is an orthogonal set of 3 8 1 vectors. Final Answer: The second and third vectors in the set are not orthogonal, so the set is not an orthogonal set of vectors. 4 3 1 3 Work: 3 5 = (4)3 + 1(3) + (3)5 + 8(1) = 12 + 3 15 8 32 = 0. 8 1
3 6.2.10. Problem Restatement: Show that {u1 , u2 , u3 } is an orthogonal basis express of R and 3 2 1 x as a linear combinations of the us where u1 = 3 , u2 = 2 , u3 = 1 , and 0 1 4 5 x = 3 . 1

Final Answer: ui uj = 0 for i = j , all i = 1, 2, 3 and j = 1, 2, 3, as shown in the work below, so {u1 , u2 , u3 } is a linearly independent set of three nonzero orthogonal vectors in R3 .
4

Therefore, {u1 , u2 , u3 } is an orthogonal basis of R3 . Also, in the work below we establish 1 x= 4 u +3 u2 + 1 u. 3 1 3 3 Work: u1 u2 = 3(2)+(3)(2)+0(1) = 0, u1 u3 = 3(1)+(3)1+0(4) = 0, and u2 u3 = 2(1)+ 3)(3)+0(1) 3)+(1)(1) xu1 xu2 xu3 2(1) + (1)4 = 0. x = u u1 + u u2 + u u3 = 3(5)+( u1 + 2(5)+2( u2 + 32 +(3)2 +02 22 +22 +(1)2 1 u1 2 u2 3 u3
1(5)+1(3)+4(1) u3 12 +12 +42

24 u 18 1

+3 u + 9 2

6 u 18 3

=4 u +1 u +1 u. 3 1 3 2 3 3

6.2.14. Problem Restatement: Let y =

2 7 and u = . Write y as the sum of two orthog6 1 onal vectors, one in Span{u} and one orthogonal to u. + 4 5
28 5

Final Answer: y =

14 5 2 5

Span{u} and

4 5

28 5

is orthogonal to u.
y u u uu 14 5 2 5

Work: y = y + z where the orthogonal projection onto Span{u} is y = 14 2 2 u= 5 and the vector orthogonal to u is z = y y = 2 5 6 5 6.3 Orthogonal Projections

14+6 u 49+1

4 5 28 5

20 u 50

1 2 1 1 2 , u2 = 1 , u3 = 1 , u4 = 1 , and 6.3.2. Problem Restatement: u1 = 1 1 2 1 1 1 1 2 4 5 4 v= 3 . Assuming {u1 , u2 , u3 , u4 } is an orthogonal basis of R , write v as the sum of 3 two vectors, one in Span{u1 } and one in Span{u2 , u3 , u4 }. 2 2 4 in Span{u1 }, and y = v x = 1 in Final Answer: v = x + y where x = 2u1 = 2 5 2 1 Span{u2 , u3 , u4 }. Work: Span{u1 } and Span{u2 , u3 , u4 } are orthogonal complements of each other, so if x is the orthogonal projection of v onto Span{u1 } then y = v x is the orthogonal projection 3)+1(3) u1 v of v onto Span{u2 , u3 , u4 }. u = 1(4)+2(5)+1( = 14 = 2. Therefore, x = 2u and 12 +22 +12 +12 7 1 u1 y = v 2u.

 6 4 6.3.6. Problem Restatement: y = 4 , u1 = 1 1 1 an orthogonal set, and then nd the orthogonal

0 , and u2 = 1 . Verify that {u1 , u2 } is 1 projection of y onto Span{u1 , u2 }.

Final Answer: u1 u2 = (4)0 + (1)1 + 1(1) = 0, so {u1 , u2 } is an orthogonal set. The 3 3 u1 + 5 u . Since 2 u1 + 5 u = y , we orthogonal projection of y onto Span{u1 , u2 } is y = 2 2 2 2 2 get y = y ; that is y Span{u1 , u2 }. Work: y =
y u1 y u1 u +u u2 u1 u1 1 2 u2

6(4)+4(1)+1(1) u + 6(0)+4(1)+1(1) u2 (4)2 +(1)2 +12 1 02 +12 +12

3 1 6.3.12. Problem Restatement: Find the closest point to y = 1 in the subspace W spanned 13 1 4 2 1 by v1 = 1 and v2 = 0 . 2 3 Final Answer: the closest point to y = in the subspace W spanned by v1 = and v2 is 1 5 y = vy1vv11 v1 + vy2vv22 v2 = 3v1 + v2 = 3 . 9 1 1 Check: If z = y y = 4 1 , then it is easy to see z is orthogonal to v1 = and v2 . 1 Work: v1 v2 = (1)(4) + (2)(1) + (1)(0) + (2)(3) = 0, so v1 v2 . y v1 = (3)(1) + (1)(2) + (1)(1) + (13)(2) = 30. y v2 = (3)(4) + (1)(1) + (1)(0) + (13)(3) = 26. v1 v1 = (1)2 + (2)2 + (1)2 + (2)2 = 10. v2 v2 = (4)2 + (1)2 + (0)2 + (3)2 = 26. y v1 = 3, and vy2vv22 = 1. v1 v1

5 = 27 u +2 u2 = 3 u +5 u. 18 1 2 1 2 2

6.3.14. Problem Restatement: to vectors of the form Find the best approximation z by 2 2 5 4 0 2 c1 v1 + c2 v2 , where z = 0 , v1 = 1 , and v2 = 4 1 3 2 Final Answer: {c1 v1 + c2 v2 | c1 , c2 R} = Span{v1 , v2 }. Therefore, the best approximation 2 0 1 1 v1 + 0v2 = 2 to z by vectors of the form c1 v1 + c2 v2 is z = vz1vv11 v1 + vz2vv22 v2 = 2 1 . 3 1 4 Check: If u = z z = 1 , then it is easy to see uv1 and uv2 .
2 1 2

Work: v1 v2 = (2)(5) + (0)(2) + (1)(4) + (3)(2) = 0, so v1 v2 . z v1 = (2)(2) + (4)(0) + (0)(1) + (1)(3) = 7. z v2 = (2)(5) + (4)(2) + (0)(4) + (1)(2) = 0. v1 v1 = (2)(2) + (0)(0) + (1)(1) + (3)(3) = 14. v2 v2 = (5)(5) + (2)(2) + (4)(4) + (2)(2) = 34. 1 c1 = vv11vz1 = 2 , and c2 = vy2vv22 = 0. 6.7 Inner Product Spaces 6.7.4. Problem Restatement: Let P2 have inner product dened by evaluation at 1, 0, and 1 (that is, < p(t), q (t) >= p(1)q (1) + p(0)q (0) + p(1)q (1)). Compute < p(t), q (t) > where p(t) = 3t t2 , and q (t) = 3 + 2t2 . Final Answer: < p(t), q (t) >= (3(1) (1)2 )(3 + 2(1)2 ) + (3(0) (0)2 )(3 + 2(0)2 ) + (3(1) (1)2 )(3 + 2(1)2 ) = (4)(5) + (0)(3) + (2)(5) = 10. Work: None required. 6.7.6. Problem Restatement: Compute ||p|| and ||q ||, for p and q in 6.7.4. Final Answer: ||p|| = 2 5, and ||q || = 59. Work: ||p||2 =< p(t), p(t) >= (4)2 +02 +22 = 20, and ||q ||2 =< q (t), q (t) >= 52 +32 +52 = 59.

6.7.8. Problem Restatement: Find the orthogonal projection of q onto Span{p}, for p and q in 6.7.4. Final Answer: Find the orthogonal projection of q onto Span{p} is <q,p> 1 2 q = <p,p> p = 10 (3t t2 ) = 3 t+ 2 t. 20 2 Work: None required. 6.7.10. Problem Restatement: Let R3 have inner product given by evaluation at 3, 1, 1, and 3. Let p0 (t) = 1, p1 (t) = t, and q (t) = t2 5 (found in the back of text). Find the best approximation to t3 by polynomials in Span{p0 , p1 , q }. Final Answer: According to the information given in 4.7.9, {p0 , p1 , q } is an orthogonal set. Therefore, the best approximation to t3 by polynomials in Span{p0 , p1 , q } is 3 ,q> <t3 ,p0 > <t3 ,p1 > 0 t3 = <p p0 + <p p1 + <t q=0 p + 164 p + 64 q = 41 p = 41 t. <q,q> 4 0 20 1 5 1 5 0 ,p0 > 1 ,p1 > Work: < p0 , p0 >= 12 + 12 + 12 + 12 = 4. < p1 , p1 >= (3)2 + (1)2 + (1)2 + (3)2 = 20. < q, q >= (4)2 + (4)2 + (4)2 + (4)2 = 64. < t3 , p0 >= (3)3 (1) + (1)3 (1) + (1)3 (1) + (3)3 (1) = 0. < t3 , p1 >= (3)3 (3) + (1)3 (1) + (1)3 (1) + (3)3 (3) = 164. < t3 , q >= (3)3 (4) + (1)3 (4) + (1)3 (4) + (3)3 (4) = 0.