Scribd Logo
Upload

Welcome to Scribd!

  • 4 views

    Module 11 Assignment Solutions

    Uploaded by

    wes.s.devlin

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Module 11 Assignment Solutions

    Uploaded by

    wes.s.devlin
    4 views3 pages

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    4 views3 pages

    Module 11 Assignment Solutions

    Uploaded by

    wes.s.devlin

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 3

    Module 11 Assignment Solutions 1

    Matrix Theory

    ***Please do not share or re-distribute these solutions!***

    1. To find the eigenvalues of A, we need to compute

    4−λ 4 1
    3
    χA (t) = (−1) det(A − λI ) = (−1) −1 −λ 0 .
    0 0 2−λ

    Applying co-factor expansion along the bottom row, we get

    4−λ 4
    χA (t) = −(2−λ) = −(2−λ) [(4 − λ)(−λ) − 4(−1)] = (λ−2)(λ2 −4λ+4) = (λ−2)3 .
    −1 −λ

    Thus the only eigenvalue of A is 2.


    Next, we need to find an eigenvector associated with eigenvalue 2. To do so, we need to
    find the null space of A − 2I . We begin by finding the row reduced echelon form of A − 2I :
             
    2 4 1 1 2 12 1 2 21 1 2 12 1 2 0
     E1/2 [1]   E1 [2,1]   E2 [2]   E−1/2 [1,2] 
    −1 −2 0 −−−→ −1 −2 0  −−−→ 0 0 12  −−→ 0 0 1  −−−−−→ 0 0 1 .
     
    0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

    Applying the rest of Gauss-Jordan elimination, we delete the last row, add a row of zeros
    to the second row, and change the missing pivot to −1 to get
     
    1 2 0
    0 −1 0 .
     
    0 0 1

    This gives us the eigenvector u = (2, −1, 0)T associated wtih eigenvalue −1.
    Next, we find a basis for C3×1 containing u by creating the matrix [u | I ] and reducing to
    row reduced echelon form:
         
    1 1
    2 1 0 0 1 2
    0 0 1 2
    0 0
     E1/2 [1]   E1 [2,1]  1
    −1 0 1 0 −−−→ −1 0 1 0 −−−→ 0 1 0
     
    2
    0 0 0 1 0 0 0 1 0 0 0 1
       
    1 21 0 0 1 0 −1 0
    E2 [2]   E−1/2 [1,2] 
    −−→ 0 1 2 0 −−−−−→ 0 1 2 0 .

    0 0 0 1 0 0 0 1
    2

    Since the first, second, and fourth columns are pivot columns, the corresponding
    columns of [u | I ] form a basis for C3×1 : {u1 , u2 , u3 } = {(2, −1, 0)T , (1, 0, 0)T , (0, 0, 1)T }.
    We can now use the Gram-Schmidt algorithm to generate an orthogonal basis.

    v1 = (2, −1, 0)T


             
    1 2 1 2 1/5
      hu2 , v1 i     2    
    v2 = 0 − −1 = 0 − −1 = 2/5
    hv1 , v1 i 5
    0 0 0 0 0
           
    0 2 1 0
      hu3 , v1 i   hu3 , v2 i    
    v3 = 0 − −1 − 0 = 0
    hv1 , v1 i hv2 , v2 i
    1 0 0 1
    Normalizing, we create the unitary matrix
     
    √2 √1 0
     √51 √2
    5
    P= − 5 0 .

    5
    0 0 1
    Then  
    2 5 √2
    5
    ∗ √1 
    P AP = 0 2 =U
     
    5
    0 0 2
    which is already upper triangular, so we are done! Our Schur triangularization is
     2   
    √ √1 0 2 5 √2
     5 5 5
    P = − √15 √25 0 and U = 0 2 √15  .
      

    0 0 1 0 0 2

    2. (a) We know by that for any n × n matrices A1 , A2 , tr(A1 A2 ) = tr(A2 A1 ) (see page 132 in
    the textbook). Letting A1 = AB and A2 = C , observe that applying this property of
    trace yields
    tr(ABC ) = tr((AB)C ) = tr(C (AB)) = tr(C AB)
    as desired.
    (b) Let A ∈ Cn×n have Schur Triangularization U = P ∗ AP where P ∈ Cn×n is unitary
    matrix and U ∈ Cn×n is upper triangular. By part (a),

    tr(U) = tr(P ∗ AP) = tr(PP ∗ A) = tr(I A) = tr(A).

    Since the elements on the main diagonal of U are the eigenvalues λ1 , . . . , λn of A,


    we can conclude
    n
    X
    tr(A) = tr(U) = λj .
    j=1
    3

    Note. This argument can also be made using the Jordan Normal Form.
    (c) Let A ∈ Cn×n have Schur Triangularization U = P ∗ AP where P ∈ Cn×n is unitary
    matrix and U ∈ Cn×n is upper triangular. Because det(AB) = det(A)det(B) by
    Theorem 3.30 (6),

    det(U) = det(P ∗ AP) = det(P ∗ )det(A)det(P).

    Since P is unitary, P ∗ = P −1 and by Theorem 3.30 (7), det(P ∗ ) = (det(P))−1 . That is,
    det(P ∗ )det(P) = 1. Thus, det(U) = det(A).
    Additionally, since the elements on the main diagonal of U are the eigenvalues
    λ1 , . . . , λn of A and the determinant of a triangular matrix is the product of the
    diagonal entries by Theorem 3.30 (5), we can conclude
    n
    Y
    det(A) = det(U) = λj .
    j=1

    Note. This argument can also be made using the Jordan Normal Form.

    3. Since null(T − 2I ) = 1 and null(T − 3I ) = 2, 2 has geometric multiplcity 1 and 3 has


    geometric multiplicity 2. This tells us there is one Jordan block associated with
    eigenvalue 2 and two Jordan blocks associated with eigenvalue 3.
    Since χT (t) = (t − 2)2 (t − 3)3 , we know that T has eigenvalues 2 and 3, where the
    algebraic multiplicity of 2 is 2 and the algebraic multiplicity of 3 is 3. This tells us the sum
    of the sizes of the Jordan blocks associated with eigenvalue 2 is 2 (i.e., the Jordan block
    associated with eigenvalue 2 is J(2, 2)) and the sum of the sizes of the Jordan blocks
    associated with eigenvalue 3 is 3. The only possible way to write 3 as a sum of two
    positive integers is 1 + 2, so the two Jordan blocks associated with eigenvalue 3 are
    J(3, 1) and J(3, 2). Overall, this tells us the Jordan normal form of T (up to permutation of
    the Jordan blocks) is
     
    2 1 0 0 0
    0 2 0 0 0
     
    diag{J(2, 2), J(3, 1), J(3, 2)} = 0 0 3 0 0 .
     
     
    0 0 0 3 1
    0 0 0 0 3

    You might also like