MATH 304 Linear Algebra Lecture 20: The Gram-Schmidt process (continued). Eigenvalues and eigenvectors.

Transcription

1 MATH 304 Linear Algebra Lecture 20: The Gram-Schmidt process (continued). Eigenvalues and eigenvectors.

2 Orthogonal sets Let V be a vector space with an inner product. Definition. Nonzero vectors v 1,v 2,...,v k V form an orthogonal set if they are orthogonal to each other: v i,v j = 0 for i j. If, in addition, all vectors are of unit norm, v i = 1, then v 1,v 2,...,v k is called an orthonormal set. Theorem Any orthogonal set is linearly independent.

3 Orthogonal projection Theorem Let V be a finite-dimensional inner product space and V 0 be a subspace of V. Then any vector x V is uniquely represented as x = p + o, where p V 0 and o V 0. The component p is the orthogonal projection of the vector x onto the subspace V 0. The distance from x to the subspace V 0 is o. If v 1,v 2,...,v n is an orthogonal basis for V 0 then p = x,v 1 v 1,v 1 v 1 + x,v 2 v 2,v 2 v x,v n v n,v n v n.

4 o x p V 0

5 The Gram-Schmidt orthogonalization process Let V be a vector space with an inner product. Suppose x 1,x 2,...,x n is a basis for V. Let v 1 = x 1, v 2 = x 2 x 2,v 1 v 1,v 1 v 1, v 3 = x 3 x 3,v 1 v 1,v 1 v 1 x 3,v 2 v 2,v 2 v 2,... v n = x n x n,v 1 v 1,v 1 v 1 x n,v n 1 v n 1,v n 1 v n 1. Then v 1,v 2,...,v n is an orthogonal basis for V.

6 v 3 x 3 p 3 Span(v 1,v 2 ) = Span(x 1,x 2 )

7 Any basis x 1,x 2,...,x n Orthogonal basis v 1,v 2,...,v n Properties of the Gram-Schmidt process: v k = x k (α 1 x α k 1 x k 1 ), 1 k n; the span of v 1,...,v k is the same as the span of x 1,...,x k ; v k is orthogonal to x 1,...,x k 1 ; v k = x k p k, where p k is the orthogonal projection of the vector x k on the subspace spanned by x 1,...,x k 1 ; v k is the distance from x k to the subspace spanned by x 1,...,x k 1.

8 Normalization Let V be a vector space with an inner product. Suppose v 1,v 2,...,v n is an orthogonal basis for V. Let w 1 = v 1 v 1, w 2 = v 2 v 2,..., w n = v n v n. Then w 1,w 2,...,w n is an orthonormal basis for V. Theorem Any finite-dimensional vector space with an inner product has an orthonormal basis. Remark. An infinite-dimensional vector space with an inner product may or may not have an orthonormal basis.

9 Orthogonalization / Normalization An alternative form of the Gram-Schmidt process combines orthogonalization with normalization. Suppose x 1,x 2,...,x n is a basis for an inner product space V. Let v 1 = x 1, w 1 = v 1 v 1, v 2 = x 2 x 2,w 1 w 1, w 2 = v 2 v 2, v 3 = x 3 x 3,w 1 w 1 x 3,w 2 w 2, w 3 = v 3 v 3,... v n = x n x n,w 1 w 1 x n,w n 1 w n 1, w n = v n v n. Then w 1,w 2,...,w n is an orthonormal basis for V.

10 Problem. Let Π be the plane in R 3 spanned by vectors x 1 = (1, 2, 2) and x 2 = ( 1, 0, 2). (i) Find an orthonormal basis for Π. (ii) Extend it to an orthonormal basis for R 3. x 1,x 2 is a basis for the plane Π. We can extend it to a basis for R 3 by adding one vector from the standard basis. For instance, vectors x 1, x 2, and x 3 = (0, 0, 1) form a basis for R 3 because = = 2 0.

11 Using the Gram-Schmidt process, we orthogonalize the basis x 1 = (1, 2, 2), x 2 = ( 1, 0, 2), x 3 = (0, 0, 1): v 1 = x 1 = (1, 2, 2), v 2 = x 2 x 2,v 1 v 1,v 1 v 1 = ( 1, 0, 2) 3 (1, 2, 2) 9 = ( 4/3, 2/3, 4/3), v 3 = x 3 x 3,v 1 v 1,v 1 v 1 x 3,v 2 v 2,v 2 v 2 = (0, 0, 1) 2 4/3 (1, 2, 2) ( 4/3, 2/3, 4/3) 9 4 = (2/9, 2/9, 1/9).

12 Now v 1 = (1, 2, 2), v 2 = ( 4/3, 2/3, 4/3), v 3 = (2/9, 2/9, 1/9) is an orthogonal basis for R 3 while v 1,v 2 is an orthogonal basis for Π. It remains to normalize these vectors. v 1,v 1 = 9 = v 1 = 3 v 2,v 2 = 4 = v 2 = 2 v 3,v 3 = 1/9 = v 3 = 1/3 w 1 = v 1 / v 1 = (1/3, 2/3, 2/3) = 1 3 (1, 2, 2), w 2 = v 2 / v 2 = ( 2/3, 1/3, 2/3) = 1 3 ( 2, 1, 2), w 3 = v 3 / v 3 = (2/3, 2/3, 1/3) = 1 3 (2, 2, 1). w 1,w 2 is an orthonormal basis for Π. w 1,w 2,w 3 is an orthonormal basis for R 3.

13 Problem. Find the distance from the point y = (0, 0, 0, 1) to the subspace Π R 4 spanned by vectors x 1 = (1, 1, 1, 1), x 2 = (1, 1, 3, 1), and x 3 = ( 3, 7, 1, 3). Let us apply the Gram-Schmidt process to vectors x 1,x 2,x 3,y. We should obtain an orthogonal system v 1,v 2,v 3,v 4. The desired distance will be v 4.

14 x 1 = (1, 1, 1, 1), x 2 = (1, 1, 3, 1), x 3 = ( 3, 7, 1, 3), y = (0, 0, 0, 1). v 1 = x 1 = (1, 1, 1, 1), v 2 = x 2 x 2,v 1 v 1,v 1 v 1 = (1, 1, 3, 1) 4 (1, 1, 1, 1) 4 = (0, 2, 2, 0), v 3 = x 3 x 3,v 1 v 1,v 1 v 1 x 3,v 2 v 2,v 2 v 2 = ( 3, 7, 1, 3) (1, 1, 1, 1) (0, 2, 2, 0) 4 8 = (0, 0, 0, 0).

15 The Gram-Schmidt process can be used to check linear independence of vectors! The vector x 3 is a linear combination of x 1 and x 2. Π is a plane, not a 3-dimensional subspace. We should orthogonalize vectors x 1,x 2,y. v 4 = y y,v 1 v 1,v 1 v 1 y,v 2 v 2,v 2 v 2 = (0, 0, 0, 1) 1 4 (1, 1, 1, 1) 0 (0, 2, 2, 0) 8 = (1/4, 1/4, 1/4, 3/4). ( 1 v 4 = 4, 1 4, 1 4, 3 ) 1 12 = (1, 1, 1, 3) = =

16 Problem. Find the distance from the point z = (0, 0, 1, 0) to the plane Π that passes through the point x 0 = (1, 0, 0, 0) and is parallel to the vectors v 1 = (1, 1, 1, 1) and v 2 = (0, 2, 2, 0). The plane Π is not a subspace of R 4 as it does not pass through the origin. Let Π 0 = Span(v 1,v 2 ). Then Π = Π 0 + x 0. Hence the distance from the point z to the plane Π is the same as the distance from the point z x 0 to the plane Π 0. We shall apply the Gram-Schmidt process to vectors v 1,v 2,z x 0. This will yield an orthogonal system w 1,w 2,w 3. The desired distance will be w 3.

17 v 1 = (1, 1, 1, 1), v 2 = (0, 2, 2, 0), z x 0 = ( 1, 0, 1, 0). w 1 = v 1 = (1, 1, 1, 1), w 2 = v 2 v 2,w 1 w 1,w 1 w 1 = v 2 = (0, 2, 2, 0) as v 2 v 1. w 3 = (z x 0 ) z x 0,w 1 w 1,w 1 w 1 z x 0,w 2 w 2 w 2,w 2 = ( 1, 0, 1, 0) 0 4 (1, 1, 1, 1) 2 (0, 2, 2, 0) 8 = ( 1, 1/2, 1/2, 0). ( w 3 = 1, 1 2, 1 ) 2, 0 1 = ( 2, 1, 1, 0) = = 3 2.

18 Linear transformations of R 2 Any linear mapping f : R 2 R 2 is represented as multiplication of a 2-dimensional column vector by a 2 2 matrix: f (x) = Ax or f ( x y ) = ( a b c d ) ( ) x. y Linear transformations corresponding to particular matrices can have various geometric properties.

19 A = ( ) Rotation by 90 o

20 ( ) A = Rotation by 45 o

21 A = ( ) Reflection in the vertical axis

22 A = ( ) Reflection in the line x y = 0

23 A = ( ) 1 1/2 0 1 Horizontal shear

24 ( ) 1/2 0 A = 0 1/2 Scaling

25 ( ) 3 0 A = 0 1/3 Squeeze

26 A = ( ) Vertical projection on the horizontal axis

27 A = ( ) Horizontal projection on the line x + y = 0

28 A = ( ) Identity

29 Eigenvalues and eigenvectors of a matrix Definition. Let A be an n n matrix. A number λ R is called an eigenvalue of the matrix A if Av = λv for a nonzero column vector v R n. The vector v is called an eigenvector of A belonging to (or associated with) the eigenvalue λ. Remarks. Alternative notation: eigenvalue = characteristic value, eigenvector = characteristic vector. The zero vector is never considered an eigenvector.

30 ( ) 2 0 Example. A =. 0 3 ( ) ( ) ( ) ( ) = = 2, ( ) ( ) ( ) ( ) = = Hence (1, 0) is an eigenvector of A belonging to the eigenvalue 2, while (0, 2) is an eigenvector of A belonging to the eigenvalue 3.