18.

700 Linear Algebra- Georg Oberdieck

Roger Jin
September 6, 2017

Contents
1 Lecture 1- Vector Spaces 1
1.1 Administrivia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 Properties of Vector Spaces 2

2.1 Subspaces: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.2 Sum of Subspaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

3 Lecture 3: Bases 3

1 Lecture 1- Vector Spaces

1.1 Administrivia
1. Lecture MWF 10-11
2. Office hours M 12-1, T 11-12
3. Pset: Due Wednesdays in class.

4. Midterms 9/25, 10/30, 11/29

5. 10 Psets
6. 1 Final
7. Book: Axler Linear Algebra Done Right

1.2 Definition
1. We ignore the traditional arrow out of the origin componentwise vector picture and try to make
things more abstract.

2. The first thing we need to define is R.

3. Rn = {(v1 , . . . , vn )|vi )} R
4. v = (v1 , . . . , vn ), w = (w1 , . . . , wn )

5. v + w := (v1 + w2 , . . . , vn + wn )
6. Associativity - ( ) v = ( v)
7. Distributivity
8. R

9. v = (v1 , . . . , vn )

1
 10. Goal: Formalize this to be more abstract.
11. Def: A vector space over R is a set V together with operations
(a) Addition - A function + which assigns to any pair v V , w V an element v + w V .
(b) Scalar mult A function which assigns to any R, v V assigns an element V .
(c) Satisfying the following conditions:
i. Commutativity: v + w = w + v for all v, w V .
ii. Associativity: (u + v) + w = u + (v + w) for u, v, w V .
iii. zero element: There exists an element 0v V s.t. v + 0v = vfor all v V . (This means
the empty set is not a vector space).
iv. Additive inverse: For all v V , w V s.t. v + w = 0v .
v. Multiplicative identity: 1 v = v for all v V .
vi. Distributativity: ( + ) v = v + v. For all R, v V .

1.3 Examples
1. V = Rn is a vector space. 0v = (0, . . . , 0). What happens if n = 0? R0 has one element- the empty
list, and this is a vector space.
2. The empty set is not a vector space.
3. V = {(v0 , v1 , v2 , . . . )|vi R} is a vector space.
4. V = {f : R R} The set of functions is a vector space.

1.4 Remarks
1. Instead of R we can take any field K (German Korper) (set that behaves like real numbers). See
exercises for definition. Ex: K = R, K = C, K = Z/zz. To fix the definition, replace R by K.

2 Properties of Vector Spaces

Remark IF I want to learn set theory, I should learn Halmes- NAive Set Theory.

Lemma: There exists a unique element 0v s.t. v + 0v = v for all v V .

Proof: Assume 0v , 00v V with this property. Need to show 0v = 00v . 0v = 0v + 00v = 00v + 0v = 00v .

Lemma: For every v V a unique w V s.t. v + w = 0v . (W write v for w).

Proof: Assume w, w0 V with v + w = 0, v + w0 = 0. Note: We cant just subtract the two equations
because we havent defined subtraction.
w = w + 0v = w + (v + w0 ) = (w + v) + w0 = (v + w) + w0 = w0 + 0v = w0 . Proof by associativity.

Lemma: 0 v = 0v for every v V .

Proof: 0 v = (0 + 0) v = 0 v + 0 v Add 0v . Thus 0v = 0 v.

Lemma: 0v = 0v for all K. The proof is left as an exercise.

Lemma: (1) v = v. The proof is left as an exercise.

2.1 Subspaces:
1. Definition: A subspace of a vector space (V, +, ) is a subset u V which is a vector space for the
restrictions of +, to u.
2.
3. Prop: A subset U V is a subspace iff the following hold: U is not the empty set, u + v U ,
u U.

2
 4. Proof: If u subspace, then u is a vector space which implies 1-3. Other direction: Check properties
of vector space for U . (a, b, e, f ) conditions for V imply conditions for U . By (i) Let u U be an
element. Lemma 5 implies that u = (1) u U which implies 0v = u + (u) U .

2.2 Sum of Subspaces

Let U1 . . . Un be subspaces of V . Prop: ni=1 = u1 n . . . nUn is a subspace of V . Proof: Not empty becasue
0 ui for all i. Is U1 U U2 a subspace fo V ? The answer is sometimes. for example take the xy plane
and yz plane and the vecotr space to be 3D space.
Def: The sum of a subspace is the set of the sums of every sequence from ui .
Prop: U1 + U2 is a subspace.

3 Lecture 3: Bases
Let V be a vector space over K. Let v1 , . . . , vn V , a1 , . . . , an K. The vector v = a1 v1 + a2 v2 + +
an vn is called a linear combination of v1 , . . . , vn .
Def: The span of v1 , . . . , vn V is the set Span(v1 , . . . , vn ) = {a1 v1 + + an vn ka1 , . . . , an K} of
all lin comb. of v1 , . . . , vn .
Note: The empty list has span 0.
Prove: The span is a subspace. We first must show that the set is not empty. If we choose all
ai = 0, then we get a linear combination of 0. Thus the set is not empty because the 0 vector is
in the set. Next we must show that the span is closed under addition and scalar multiplication. Let
v, w Span(v1 , . . . , vn ). v = a1 v1 + + an vn for some a, b K, w = b1 v1 + . . . . + b + nv + n
v + w = (a1 + b1 )v1 + + (an + bn )vn Span(v1 , . . . , vn ). Now we must show that the span is
closed under scalar multiplication. v = a1 v1 + + an vn Span(v1 , . . . , vn ). And for any K,
a1 v1 + + an vn Span(v1 , . . . , vn ).

1. Let W V be a subspace of V with v1 , . . . , vn V . Let a1 , . . . , an K. Some W closed under

scalar multiplication a1 v1 W

Def: The list of vectors v1 , . . . vn V is called linearly dependent if there exists a1 , . . . an K,

not all 0, such that we can write a1 v1 + . . . an vn = 0. If such a1 , . . . , an exist, then v1 , . . . , vn are called
linearly independent.
Examples: The empty list is linearly independent. (e1 , . . . , en ) is linearly independent.
Prop: Let v1 , . . . , vn V , TFAE (The following are equivalent):
1. v1 , . . . , vn are linearly independent

2. No vi is a linearl combination of the other vectors in the list.

3. Every vector in Span(v1 , . . . , vn ) can be written in a unique way as a linear combination of the
v1 , . . . , v n .
Proof: Assume v1 , . . . vn V are linearly independent and let v Span(v1 , . . . , vn ) such that there exists
ai , bi such that v = a1 v1 + + an vn and v = b1 v1 + b2 v2 + . . . bn vn . Subtract the two equations.
0 = (a1 b1 )v1 + + (an bn )vn . By linear independence, ai bi = 0 for all i.