This preview shows page 1. Sign up to view the full content.
Unformatted text preview: implies diagonalizable
35 kshum The converse is false If a matrix A is similar to a diagonal matrix, then A may
have repeated eigenvalues. Eigenvalue = 2, 2,–3. Diagonalizable matrix may have repeated eigenvalues
36 kshum Facts
Distinct eigenvalues Eigenvalues not all distinct 37 diagonalizable may or may not
be diagonalizable kshum Calculation of inverse required in diagonalization May take a lot of time
to compute when
the size of matrix is large 38 kshum Symmetric matrix Definition: A matrix M is symmetric if MT = M.
Examples: Line of symmetry 39 kshum Diagonalization of symmetric matrix Transpose instead
of inverse 40 kshum Orthogonal matrix A matrix M is called orthogonal if
Each column has norm 1
MT I M Dot product = 1
41 kshum Orthogonal matrix A matrix M is called orthogonal if
Any two distinct columns are orthogonal Dot product = 0
42 kshum Orthogonal matrix If then M is also orthogonal Each row has norm 1
Any two distinct rows are orthogonal M
Dot product = 0
43 MT
kshum Theorem Two eigenvectors of a symmetric matrix are orthogonal if
the corresponding eigenvalues are different.
Let A be a symmetric matrix, AT = A.
Let u is an eigenvector with eigenvalue λ, A u = λ u
Let v is an eigenvector with eigenvalue µ. A v = µ v
Suppose λ not equal to µ. (a scalar) 44 kshum Corollary
1.
2. 3. For an n× n symmetric A with n distinct eigenvalues, we
can find n orthogonal eigenvectors.
Form a matrix P with the n eigenvectors as columns.
Normalize each column so that the norm of each column
is equal to one.
Then PT P = I and PTAP is a diagonal matrix, whose
diagonal entries consist of the eigenvalues of A. Example:
PT 45 A P kshum Principal axes theorem: all symmetric matrices are
orthogonally diagonalizable
A stronger statement is possible
1. For an n× n symmetric A with n distinct eigenvalues, we
can find n orthogonal eigenvectors. 2. Form a matrix P with the n eigenvectors as columns.
Normalize each column so that the norm of each column
is equal to one.
Then PT P = I and PTAP is a diagonal matrix, whose
diagonal entries consist of the eigenvalues of A. 3. The conclusion in the previous page holds even
without the hypothesis of distinct eigenvalues.
46 Proof omitted
kshum Facts
A has distinct eigenvalues Eigenvalues not all distinct Symmetric A,
eigenvalues may or
may not be distinct
47 P– 1AP = D may or may not
be diagonalizable PTAP = D
P TP = I
kshum Application to conic sections Is x2 – 4xy +2y2 = 1 a ellipse, or a hyperbola? Rewrite using symmetric matrix Find the characteristic polynomial Solve for the eigenvalues 48 kshum Application to conic sections
Diagonalize
Change coordinates Hyperbola
49 kshum x2 – 4xy +2y2 = 1
15 10 y 5 0 5 10 15
15 10 5 0 5 10 x
50 kshum 15 2x2 + 2xy + y2 = 1
Rewrite using symmetric matrix Compute the characteristic polynomial Find the eigenvalues 51 kshum 2x2 + 2xy + y2 = 1
Diagonalize Columns of P are eigenvectors,
normalized to norm 1. Change of variables 52 kshum 2x2 + 2xy + y2 = 1
v
0.8
0.6
0.4 y 0.2
0
0.2
0.4
0.6
0.8
1 0.5 0 0.5 1 u x
53 kshum Final 5 May, Wed
UC Gym
9:3011:30
Coverage: Kreyszig chapter 1.1~1.5, 2.1, 2.2, 2.4, 2.7~2.9, 4.0~4.5,
7.1~7.9, 8.1~8.5
Lecture notes, tutorial notes, assignments Closebook exam
You may bring one single sheet of handwritten note 54 A4 size, doublesided kshum...
View
Full
Document
This document was uploaded on 09/16/2013.
 Spring '13

Click to edit the document details