1. What Are Vectors, Really?
| View | Strength | Weakness |
|---|---|---|
| Lists of numbers | Clear-cut; makes 4D or 100D concrete | Ties vectors to one choice of coordinates |
| Arrows in space | Matches the feeling that space exists independently of coordinates | Vague in high dimensions |
Core ideas like the determinant (area scaling) and eigenvectors (vectors staying on their span) are independent of the coordinate system — they’re spatial. So what does “space” mean in general?
2. Functions as Vectors
Definition 2.1 – Adding functions
Like adding vectors coordinate by coordinate — but with infinitely many “coordinates” (one per input).
Definition 2.2 – Scaling functions
Since adding and scaling are all vectors really do, linear algebra’s tools should carry over to functions.
3. Linear Transformations of Functions
Example – The derivative
takes a function and returns another function — a transformation (also called an operator).
Definition 3.1 – Linearity (formal)
is linear if it satisfies:
| Property | Statement |
|---|---|
| Additivity | |
| Scaling |
“Preserves addition and scalar multiplication.” The earlier picture (grid lines parallel and evenly spaced) is just what these properties look like for arrows in 2D.
The derivative satisfies both:
Consequence
A linear transformation is completely determined by what it does to basis vectors — true for functions as well as arrows.
4. The Derivative as a Matrix
Basis for polynomials
(infinitely many, since degrees are unbounded; each polynomial uses only finitely many).
| Polynomial | Coordinates |
|---|---|
The derivative matrix
Take the derivative of each basis function and record its coordinates as a column ():
Mostly zeros, with on the diagonal just above the main one.
Worked example. :
Matrix–vector multiplication and differentiation — seemingly unrelated — are members of the same family.
Dictionary
| Arrows / lists | Functions |
|---|---|
| Linear transformation | Linear operator |
| Dot product | Inner product |
| Eigenvector | Eigenfunction |
Example eigenfunction. → is an eigenfunction of with eigenvalue 3.
5. Vector Spaces and Axioms
Definition 5.1 – Vector space
Any set of “vector-ish” objects with sensible addition and scaling — arrows, lists, functions, or anything else — satisfying a checklist of rules.
Definition 5.2 – The eight axioms
For all in the space and scalars :
- (associativity of addition)
- (commutativity)
- There is a with (zero vector)
- Each has with (additive inverse)
- (compatible scaling)
- (identity scalar)
- (distributes over vector addition)
- (distributes over scalar addition)
Role of the axioms
An interface between whoever proves results and whoever applies them:
- Prove theorems using only the axioms → they apply to every vector space, even ones you never imagined.
- Someone with a new kind of “vector” just checks the axioms, then uses all of linear algebra.
That’s why textbooks define linearity by additivity and scaling instead of grid lines — more abstract, but fully general.
⚠️ Pitfall: The abstract definitions aren’t the best first way to learn — build intuition with concrete 2D arrows, then recognize that the tools apply far more broadly.
6. The Mathematician’s Answer
“What is a vector?” → It doesn’t matter. Anything that can be added and scaled according to the axioms is a vector — just as the number 3 abstracts every possible triplet of things.
Exercises: Build the 4×4 derivative matrix for polynomials of degree ≤ 3 by differentiating each basis function, apply it to , and check whether the set of polynomials with is a vector space (lesson page).