Math 18, Lecture A Fall 2022 Section 6.2 - Orthogonal Sets Definition. A subset {~u1 , ~u2 , . . . , ~up } of Rn is orthogonal if ~ui · ~uj = 0 whenever i 6= j. 1 1 0 1 , 0 , 0 is orthogonal. Example. Show that 0 1 −1 Theorem. If S = {~u1 , ~u2 , . . . , ~up } is an orthogonal set of nonzero vectors in Rn , then S is linearly independent. Definition. Let W be a subspace of Rn . An orthogonal basis for W is a basis for W that is also orthogonal. 1 1 0 Example. Explain why S = 1 , 0 , 0 is an orthogonal basis for R3 . 0 1 −1 Theorem. Let {~u1 , ~u2 , . . . , ~up } be an orthogonal basis for a subspace W of Rn . Then for ~y · ~uj ~y · ~uj each ~y ∈ W , if ~y = c1~u1 + c2~u2 + · · · + cp~up , then cj = or for j = 1, 2, . . . , p. ~uj · ~uj k~uj k2 1 Math 18, Lecture A Fall 2022 Section 6.2 - Orthogonal Sets 1 1 1 0 Example. β = 1 , 0 , 0 is an orthogonal basis for R3 . Find [~y ]β if ~y = 2. 0 1 −1 3 Definition. A subset {~u1 , ~u2 , . . . , ~up } of Rn is orthonormal if it is orthogonal and every vector ~ui in the subset is a unit vector. Let W be a subspace of Rn . An orthonormal basis for W is a basis for W that is also orthonormal. Example. Determine whether the basis for R3 is orthonormal. 1. {~e1 , ~e2 , ~e3 } 1 1 0 1 , 0 , 0 2. 0 1 −1 Definition. An orthogonal matrix is a square matrix U such that U T = U −1 . 2 Math 18, Lecture A Fall 2022 Section 6.2 - Orthogonal Sets Theorem. Let U be any matrix. Then • U has orthonormal columns iff U T U = I. • If U is square, then U is orthogonal iff its columns are orthonormal. • If U is square, then U is orthogonal iff its rows are orthonormal. 0 Example. Determine whether 1 0 √1 2 0 √1 2 √1 2 0 is orthogonal. − √12 Orthogonal Projections Fix a nonzero vector ~u ∈ Rn . Can we decompose a vector ~y ∈ Rn into a sum of a scalar multiple of ~u and a vector orthogonal to ~u? ~y = ŷ + ~z 3 Math 18, Lecture A Fall 2022 Section 6.2 - Orthogonal Sets Definition. Let ~u, ~y ∈ Rn where ~u 6= ~0. Let L be the line through ~0 and ~u. Then • The orthogonal projection of ~y onto L is ŷ = projL ~y = ~y · ~u ~u. ~u · ~u • The component of ~y orthogonal to L is ~z = ~y − projL ~y = ~y − ~y · ~u ~u. ~u · ~u 3 9 Example. Let ~u = and ~y = . −9 5 1. Express ~y as a sum of two orthogonal vectors, one in Span{~u} and the other orthogonal to ~u. 2. Find the distance between ~y and the line L = Span{~u}. 4