Linear Algebra – Week 1 Vector Space Definition 1.1 — Vector Space A vector space V over a field F is a set equipped with two operations, addition and scalar multiplication: (a) Addition + : V × V → V (b) Scalar multiplication · : F × V → V that satisfies the following eight axioms for all u, v, w ∈ V and all a, b ∈ F : VS1. u + (v + w) = (u + v) + w VS2. u + v = v + u VS3. There exists an element 0 ∈ V such that u + 0 = u for all u ∈ V VS4. For each u ∈ V , there exists an element −u ∈ V such that u + (−u) = 0 VS5. a(bv) = (ab)v VS6. 1v = v, where 1 is the multiplicative identity in F VS7. a(u + v) = au + av VS8. (a + b)v = av + bv Definition 1.2 — Vector and Scalar If V is a vector space over F , then the elements of V are called vectors, the elements of F are called scalars. Additionally, 0 ∈ V is called the zero vector, and −u is called the inverse element of u in Definition 1.1. Example. a) V = F [x] is a vector space over F . Where addition and scalar multiplication are the regular operations. b) V = F n are vector spaces over F , where n ∈ N. For u = (x1 , x2 , . . . , xn ), v = (y1 , y2 , . . . , yn ) ∈ V, a ∈ F . Where u + v = (x1 + y1 , x2 + y2 , . . . , xn + yn ), av = (ax1 , ax2 , . . . , axn ). 1 Linear Algebra – Week 1 2 c) V = R>0 is a vector space over F = R For x, y ∈ V, α ∈ F : Addition is x ⊕ y = xy. Scalar multiplication is α ⊙ x = xα = eα log x . The zero vector 0 is 1 ∈ V . Checking the eight axioms are satisfied for the above three examples is left as exercise. We now prove some elementary theorems and corollaries about the vectors: Theorem 1.1 — Cancellation Law for Vector Addition Let v, v1 , v2 ∈ V . If v + v1 = v + v2 , then v1 = v2 . Proof. Let w be a inverse of v. w + v + v1 = w + v + v2 =⇒ 0 + v1 = 0 + v2 =⇒ v1 = v2 ■ Corollary 1.2 — Uniqueness of the Zero Vector The zero vector 0 (also called the origin) in VS3 is unique. Proof. Suppose there are two origins 01 , 02 . We have 01 + 02 = 02 + 01 =⇒ 01 = 02 (VS1) (VS3) ■ Linear Algebra – Week 1 3 Corollary 1.3 — Uniqueness of the Inverse Element The inverse element of v is unique, for all v ∈ V . Proof. Suppose there are two inverses w1 , w2 of v. We have 0 = v + w1 and 0 = v + w2 . Thus v + w1 = v + w2 , then by Theorem 1.1 we have w1 = w2 Remark. The unique inverse of v is −v := (−1) · v We can perform the operation: v + (−v) = v + (−1) · v = 1 · v + (−1) · v = (1 + (−1)) · v = (0 · v) =0 By definition (VS6) (VS8) By field operation See below To show that 0v = 0, we calculate (1 + 0) · v in two ways: (1 + 0)v = 1v = v = v + 0 (1 + 0)v = 1v + 0v = v + 0v And this gives us v + 0 = v + 0v, then by Theorem 1.1 we have 0v = 0. ■ Linear Algebra – Week 1 4 Definition 1.3 — Linear Combination Let V be a vector space and S ⊂ V . v ∈ V is a linear combination of vectors in S if v can be expressed as: v = α 1 v 1 + . . . + α n vn for some α ∈ F, vi ∈ S. Example. a) If V = F 3 , S = {v1 = (1, 1, 0), v2 = (0, 1, 1)}. Then v = (2, 5, 3) = 2v1 + 3v2 is a linear combination of vectors in S, but w = (2, 6, 3) is not. b) If V = F [x], S = {x + 1, x2 + 3x, x2 + 5} Then x2 is a linear combination of vectors in S, since we can solve x2 = α1 (x + 1) + α2 (x2 + 3x) + α(x2 + 5) for α1 , α2 , α3 . We get , α2 = 58 , α3 = 83 . α1 = −15 8 x3 is NOT a linear combination of vectors in S. If x3 = α1 (x + 1) + α2 (x2 + 3x) + α3 (x2 + 5) then x3 − α1 (x + 1) − α2 (x2 + 3x) − α(x2 + 5) = 0 should have infinitely roots in F . But a polynomial equation of degree of 3 can only have 3 solutions. Linear Algebra – Week 1 5 Span Definition 1.4 — Span Let V be a vector space over F , and let S be a subset of V . The subspace spanned by S over F is denoted by: spanF (S) := {α1 v1 + . . . + αn vn | vi ∈ S, αi ∈ F } spanF (S) can be interperted as "the set of all linear combinations of vectors in S over F ". Example. a) Let V = R2 , F = R, S = {v1 = (1, 2)}. Then spanF (S) = {αv1 | α ∈ R} = {(α, 2α) | α ∈ R} y spanF (S) x b) Let V = R3 , F = R, S = {v1 = (1, 2, 3), v2 = (0, 1, 1)} Then the span is: spanR {v1 , v2 } = {α1 v1 + α2 v2 | α1 , α2 ∈ R} = {(α1 , 2α1 + α2 , 3α1 + α2 ) | α1 , α2 ∈ R} We can notice that this is a plane in R3 which passes through the origin. c) Let V = F [x], S = {1, x, x2 , x3 , x4 , . . . } (elements of S are called monomials). Then the span is: spanF (S) = {α0 1 + α1 x + · · · + αn xn | αi ∈ F, n ∈ Z≥0 } = F [x] Linear Algebra – Week 1 6 Theorem 1.4 — Span is a Vector Space Let V /F be a vector space, and S ⊂ V such that S ̸= ∅. spanF (S) is a vector space with addition and scalar product induced from V . Proof. Let W = spanF (S) ∈ V . To show W is a vector space, it suffices to show that: a) x + y ∈ W , ∀x, y ∈ W . b) αx ∈ W , ∀α ∈ F, x ∈ W . c) 0 ∈ W . Let x = α1 v1 + · · · + αn vn , y = β1 w1 + · · · + βm wm , where vi , wj ∈ S, αi , βj ∈ F . We now prove each of the identies: a) x + y = ni=1 αi vi + m j=1 βj wj ∈ W . Since by definition, it is a linear combination of vectors in S. P P b) αx = (αα1 )v1 + · · · + (ααn )vn ∈ W c) 0 = 0v, for all v ∈ V . ■ Linear Algebra – Week 1 7 Definition 1.5 — Generating Set Let V /F be a vector space, and S ̸= ∅ ⊂ V . We say S is a generating set of V , if V = spanF (S). The elements of S are called generators if S is a generating set of V . Definition 1.6 — Finite Dimensional Let V /F be a vector space. We say V is finite dimensional, if V is spanned by a finite subset S ⊂ V . Example. a) Let V = {(x, y, z, w) ∈ R4 | x + 2y − z + w = 0, 2x + y + z + 2w = 0} Then V = spanR {(−1, 1, 1, 0), (0, 1, 1, −1)}. Since V can be spanned by a set with two elements, V is finite dimensional over R. b) Let V = C/R. Then V = spanR {1, i}. Thus V is finite dimensional. c) Let V = C/Q, then V is not finite dimensional. d) Let V = F [x]/F , then V is not finite dimensional. Proof. Suppose V is spanned by S = {f1 , . . . , fn |fi ∈ F [x]}. Let N = 1 + max {deg fi | i = 1, . . . , n}, and xN ∈ V . Since V is spanned by S. We have the following property: xN = α 1 f 1 + · · · + α n f n αi ∈ F =⇒ xN − α1 f1 − α2 f2 − · · · − αn fn = 0 But we can see that the above equation has only N roots (making the equation not hold for every x ∈ F ), which contradicts the fact that xN can be spanned by S. This further indicates that our assumption that V can be spanned by a finite subset S is wrong. ■ Linear Algebra – Week 1 Definition 1.7 — Subspace Let V /F be a vector space. A subset W ⊂ V is a subspace of V , if: a) x + y ∈ W , for all x + y ∈ W . b) αx ∈ W , for all α ∈ F, x ∈ W . c) 0 ∈ W , or W ̸= ∅ Remark. Let V /F be a vector space. If W is a subspace of V , then W/F is a vector space, with addition and scalar product induced from V . Remark. Let W1 , W2 ⊂ V are subspaces, then W1 ∩ W2 is also a subspace. But in general W1 ∪ W2 is NOT a subspace, unless W1 ⊂ W2 or W2 ⊂ W1 . (Check this) Example. a) Let V = C/R, then R ⊂ V is a subspace of V . Note that spanR {1} = R b) Let V = R2 , W1 = {(x, y) | y = 3x} ⊂ V is a subspace in V . W2 = {(x, y) | y = 3x + 1} is NOT a subspace in V , as it contains no zero vector. c) W = {(x, y) | y = x2 } is NOT a subspace. Since (1, 1) + (2, 4) ̸⊂ W . d) Let V = F [x]. Then W1 = {f ∈ V | f (0) = 0} is a subspace, but W2 = {f ∈ V | f (0) = 3} is NOT a subspace, since the zero vector, f0 (x) = 0 is not in W . 8 Linear Algebra – Week 1 9 Corollary 1.5 — Redudant Vectors Suppose V is a vector space spanned by a generating set S. In general, S may contain "redudant" vectros. Example. If S = {v1 , v2 , v3 = 2v1 + 3v2 }. Then spanF (S) = spanF {v1 , v2 }. Since: spanF (S) = {α1 v1 + α2 v2 + α3 v3 | αi ∈ F } = {(α1 + 2α3 )v1 + (α2 + 3α2 ) v2 | αi ∈ F } = spanF {v1 , v2 } Definition 1.8 — Linearly Dependent Let V /F be a vector space, and {v1 , . . . , vn } ∈ V . We say {v1 , . . . , vn } linearly dependent, if ∃ non-trival α1 , . . . , αn ∈ F , such that: α1 v1 + · · · + αn vn = 0 Where non-trivial means "not all zeros". Example. n √ o a) Let V = C/R a vector space. Then 1, 2 are linearly dependent. Since: −1 √ 1·1+ √ · 2=0 2 n √ o b) Let V = C/Q a vector space. Then 1, 2 are NOT linearly dependent. Since if: √ α1 · 1 + α2 · 2 = 0, αi ∈ Q, α1 ̸= 0 or α2 ̸= 0 We have √ Which is a contradiction. 2=− α1 ∈Q α2 Linear Algebra – Week 1 10 Definition 1.9 — Linearly Independent Let V /F be a vector space, and {v1 , . . . , vn } ∈ V . We say {v1 , . . . , vn } linearly independent, if the equation: α1 v1 + · · · + αn vn = 0 only has the trival solution such that α1 = α2 = · · · = αn = 0. Corollary. {v1 , . . . , vn } is lineary independent ⇐⇒ {v1 , . . . , vn } is NOT lineary dependent. Example. a) Let V = R3 /R, v1 = (1, 0, −1) , v2 = (1, 2, 3) , v3 = (0, 1, 1). Then {v1 , v2 , v3 } are linearly independent, since: α1 v1 + α2 v2 + α3 v3 = 0 =⇒ (α1 + α2 , 2α2 + α3 , −α1 + 3α2 + α3 ) = 0 =⇒ α1 = α2 = α3 = 0