Chapter 4, General Vector Spaces Section 4.

1, Real Vector Spaces In this chapter we will call objects that satisfy a set of axioms as vectors. This can be thought as generalizing the idea of vectors to a class of objects. Vector space axioms: Denition: Let V be an arbitrary nonempty set of objects on which two operations are dened on which two operations are dened addition, and multiplication by scalars. By addition we mean a rule for associating with each pair of objects u and v in V an object u+v, called the sum of u and v; By scalar multiplication we mean a rule for associating with each scalar k and each object u in V an object ku, called scalar multiplication of u by k. If the following axioms are satised by all objects u,v,w in V and all scalars k and m, then we call V a vector space and we call the objects in V vectors. 1. If u and v are vectors in V, then u+v is in V. 2. u+v=v+u 3. u+(v+w) = (u+v)+w 4. There is an object 0 in V, called a zero vector for V, such that 0 + u = u + 0 = u for all u in V. 5. For each u in V, there is an object -u in V, called a negative of u such that u+(u) = (u) + u = 0 6. If k is any scalar and u is any object in V, then ku is in V. 7. k(u+v)=ku+kv 8. (k+m)u=ku+mu 9. k(mu)=(km)u 10. 1u=u Note: When the scalars are real numbers, V is called a real vector space. When the scalars are complex numbers, V is called a complex vector space. Addition and scalar multiplication do not need to be the operations we dened before.

Examples: 1. Zero vector space: V = {0}, 0+0=0, k0=0 2. The set of all n m matrices with real entries, with addition= matrix addition , and scalar multiplication= scalar matrix multiplication forms a real vector space. (a) If A and B are n m matrices then A+B is also a n m matrix (b) A+B=B+A (c) A+(B+C)=(A+B)+C (d) There is zero matrix 0, such that A+0=0+A=A (e) For each matrix A we have -A such that A+(-A)=0 (f) If k is any real scalar, then kA is in V (g) k(A+B)=kA+kB (h) (k+m)A=kA+mA (i) k(mA)=(km)A (j) 1A=A, here 1 is the scalar 1. These properties hold by Theorem 1.4.1 and Theorem 1.4.2. 3. Every plane through the origin is a vector space: Here our set is the set of points of the plane, addition and scalar multiplications are the usual ones. Then we have 0 in the plane, If ax + by + cz = 0 is the plane equation, then for any u = (u1 , u2 , u3 ), v = (v1 , v2 , v3 ) on the plane, then u+v will be on the plane as well. Other properties can easily be checked. 4. The set of all real-valued functions dened on the entire real line (, ) with addition (f + g)(x) = f (x) + g(x) scalar multiplication (kf )(x) = kf (x) is a real vector space. We have 1= identity function, 0=zero function Example: Let V be the set of positive real numbers. Show that V is a vector space with addition: u + v = uv, and scalar multiplication: ku = uk Solution: We need to check that vector space axioms are satised by the objects of V. 1. Let u and v be objects in V, this means u and v are positive real numbers, then u+v=uv must also be in V as uv will be positive real number. 2. u+v=v+u due to property of real numbers 2

3. u+(v+w)=(u+v)+w this also holds as u,v,w are real numbers 4. There must be a zero vector 0 in V, such that 0+u=u+0=u, Let us denote this zero vector by a, then we want u=a+u=au, dividing by u gives a=1. That is zero vector of V is 1. 5. For each u in V there must be an object -u in V such that u+(-u)=(-u)+u=0. In 1 our vector space 0=1, so we want (-u) such that u+(u) = u(u) = 1 u = u 6. if k is any scalar and u is an object in V, then ku must be in V. We have ku = uk and since u is positive real number uk will be also a positive real number. 7. k(u+v)=ku+kv; Note that k(u + v) = (uv)k = uk v k = ku + kv 8. (k + m)u = uk+m = uk um = ku + km 9. k(mu) = k(um ) = (um )k = umk = (km)u 10. 1u = u1 = u Examples that are not vector spaces: 1. Let V = R2 and dene addition as v + w = (v1 + w1 , v2 + w2 ) and scalar multiplication as kv = (kv1 , 0). Now consider 1v = (v1 , 0) = (v1 , v2 ). Hence V cannot be a vector space. 2. Consider the set of all 2 2 matrices of the form addition and scalar multiplication. Consider a1 1 a 1 a a2 2 + 2 = 1 1 b1 1 b2 2 b1 + b2 this sum is not in the set so this set with the standart matrix addition and scalar multiplication is not a vector space. Theorem 4.1.1: Let V be a vector, u a vector in V, and k a scalar; then: (a) 0u=0 (b) k0=0 (c) (-1)u=-u (d) If ku=0, then k=0 or u=0. a 1 with the standard matrix 1 b

Section 4.2 Subspaces Denition: A subset W of a vector space V is called a subspace of V if W is itself a vector space under the addition and scalar multiplication dened on V. There is a quick way to decide if a subset of a vector space is a subspace or not. Theorem 4.2.1: If W is a set of one or more vectors from a vector space V, then W is a subspace of V if and only if the following conditions hold (a) If u and v are vectors in W, then u+v is in W (b) If k is any scalar and u is any vector in W, then ku is in W. If (a) is satised we call W to be closed under addition If (b) is satised we call W to be closed under scalar multiplication Example: 1. {0} is a subspace of Rn , since 0 + 0 = 0 (closed under addition), k0 = 0(closed under scalar multiplication) 2. Rn is a subspace of Rn {0} and Rn are called trivial subspaces of Rn 3. Let W be a plane through origin determined by two vectors v1 and v2 that is we can express any vector x in W as x = t1 v1 + t2 v2 . Then W is a subspace of Rn . Note: Every subspace of Rn must contain the vector 0. Why? Example: Determine whether the given sets are subspaces of R3 . If it is not indicate which closure properties fail. (a) All vectors of the form (a, 0, 0) (b) All vectors with integer coecients (c) All vectors (a, b, c) for which b = a + c Solution: For each set we need to check for any two vectors v, w in the set, if v + w in the set and kv is in the set for any scalar k R. (a) Let v = (a, 0, 0), w = (b, 0, 0) then v + w = (a + b, 0, 0), kv = (ka, 0, 0) and these are both in the set hence it is a subspace of R3 . (b) Let v = (a1 , a2 , a3 ), w = (b1 , b2 , b3 ) where ai , bi Z, then v + w = (a1 + b1 , a2 + b2 , a3 + b3 ), kv = (ka1 , ka2 , ka3 ). kv will not be in the set if k R Z. Hence this set is not a subspace of R3 . 4

(c) Let v = (a1 , b1 , c1 ), w = (a2 , b2 , c2 ) we have v+w and kv for any real number k are in the set. Hence this set is a subspace of R3 . Theorem 4.2.3: If S = {w1 , w2 , . . . , wr } is a nonempty set of vectors in a vector space V, then: 1. The set W of all possible linear combinations of the vectors in S is a subspace of V 2. The set W in part (a) is the smallest subspace of V that contains all of the vectors in S in the sense that any other subspace that contains those vectors contains W. Idea of the proof: Let us form a subspace from the vectors v1 , . . . , vr . For a subspace we need a subset that is closed under scalar multiplication and addition. So if vi is in the set where i = 1, . . . , r, then kvi must be in the set for any scalar k, so we need t1 v1 , t2 v2 , . . . , tr vr to be in the set. Hence t1 v1 + t2 v2 + . . . + tr vr must be in the set. Denition: If S = {v1 , . . . , vr } is a set of vectors in a vector space V, then the subspace W of V consisting of all linear combinations of the vectors in S is called the space spanned by v1 , . . . , vr and we say that the vectors v1 , . . . , vr span W. We denote this by W=span(S) or W=span{v1 , . . . , vr }. Example: 1. {0} = Span{0} 2. Rn = Span{e1 , e2 , . . . , en } 3. Line passing through origin and parallel to the vector v = Span{v} 4. Plane passing through origin and parallel to the noncolliner vectors v1 , v2 =Span{v1 , v2 } 5. Set of polynomials with degree less than or equal to n, denoted by Pn with ususal polynomial addition and scalar multiplication forms a vector space and Pn =Span{1, x, x2 , x3 , . . . , xn } Example: Determine whether the polynomials p1 = 1 x + 2x2 , p2 = 3 + x, p3 = 5 x + 4x2 , p4 = 2 2x + 2x2 span P2 ? Solution: If these polynomials span P2 then any arbitrary vector ax2 + bx + c in P2 can be expressed as a linear combination of p1 , p2 , p3 , p4 . That is ax2 + bx + c = k1 p1 + k2 p2 + k3 p3 + k4 p4 where k1 , k2 , k3 , k4 R Expressing this equation in terms of components we get a = 2k1 + 4k3 + 2k4 b = k1 + k2 k3 2k4 c = k1 + 3k2 + 5k3 2k4

Hence our question is a,b,c values. Form the 2 0 1 1 1 3

reduced to determining whether augmented matrix; 1 0 4 2 a Row operations 0 1 1 2 b 5 2 c 0 0

this system is consistent for all 2 1 1 1 0 0

a 2

b+ a 2 c2a3b 3

From the last row we see that linear system is consistent if c = 3b + 2a which gives restrictions on the scalars a,b,c. Hence p1 , p2 , p3 , p4 cannot span P2 . For example x2 +x+6 is in P2 but not in the span of p1 , p2 , p3 , p4 . Theorem 4.2.4: The solution set of a homogenous linear system Ax = 0 in n unknowns is a subspace of Rn . Exercise: Prove this theorem Example: Find a general solution of the linear system and list a set of vectors that span the solution space x1 + 6x2 + 2x3 5x4 = 0 x1 6x2 x3 3x4 = 0 2x1 + 12x2 + 5x3 18x4 = 0 Solution: 1 6 2 5 0 1 6 2 5 0 1 6 1 3 0 Row operations 0 0 1 8 0 2 12 5 18 0 0 0 0 0 0 x1 , x3 are leading variables, x2 , x4 are free variables, let x2 = s, x4 = t x3 = 8t, x1 = 6s 11t x1 6s 11t 6 11 x2 s = = s 1 + t 0 x3 0 8 8t x4 t 0 1 11 6 1 0 Solution space=Span{ , 0 8 } 1 0 Theorem 4.2.5: If S = {v1 , v2 , . . . , vr } and S = {w1 , w2 , . . . , wk } are two sets of vectors in a vector space V, then span {v1 , v2 , . . . , vr } = span {w1 , w2 , . . . , wk } if and only if each vector in S is a linear combination of those in S and each vector in S is a linear combination of those in S. Importance: Dierent number of vectors and dierent vectors can span the same space. We will see that one needs a minimum number of such vectors and that number is unique. For that we need linear independency. 6