Linear Combinations, the Span and Bases
A linear combination multiplies each of several vectors by a scalar and then adds them:
where are arbitrary scalars (real numbers).
Example: v₁ = (1, 0), v₂ = (0, 1)
w = 3·v₁ + 2·v₂ = (3, 0) + (0, 2) = (3, 2)
The key point: the scalars may be any real number (including negative numbers, zero and fractions).
Definition
The set formed by all linear combinations of a set of vectors is called their span:
Intuition: let the scalars range over all real numbers, and you get every position the vectors can “reach”.
Interactive demo
What to observe:
- two non-collinear vectors → they span the whole plane (every point can be reached)
- two collinear vectors → they span only a straight line
- a set containing the zero vector → the zero vector contributes nothing, and the span is still determined by the remaining vectors
The Core Question about a Span
Given a set of vectors, how much space can they “cover”? The answer depends on whether the vectors are linearly independent.
Definitions
- Linearly dependent: there exist scalars , not all zero, such that .
- Linearly independent: the linear combination is the zero vector only when .
Intuition
| Situation | Dependent / independent | Explanation |
|---|---|---|
| Two collinear vectors | dependent | one can be obtained from the other by scalar multiplication |
| Two non-collinear vectors | independent | neither is a multiple of the other |
| A set containing the zero vector | always dependent | the zero vector is dependent with any vector |
| Three vectors in two dimensions | always dependent | two-dimensional space cannot hold 3 independent vectors |
Examples:
v₁ = (1, 1), v₂ = (2, 2) → dependent (v₂ = 2v₁)
v₁ = (1, 0), v₂ = (0, 1) → independent
Definition
A basis is a set of linearly independent vectors that spans the space, and it is “exactly enough” — no more, no less:
- Spanning: the linear combinations of the basis vectors can reach every point of the space
- Independence: there is no redundant vector (removing any one of them makes the span smaller)
The number of vectors in a basis = the dimension of the space.
The standard basis
The most commonly used basis of the two-dimensional plane is called the standard basis:
Any vector can be written as a linear combination of them:
A basis is not unique
The same space has infinitely many bases. For example is also a basis of the two-dimensional plane — as long as the two vectors are not collinear.
Example: verifying that (1,1) and (1,-1) are a basis
any (x, y) = a(1,1) + b(1,-1)
Solving: a = (x+y)/2, b = (x-y)/2 ← there is a solution for any x, y
So they span the whole plane, and they are independent → they are a basis
Core Conclusions
- Spanning answers “where can we reach”
- Independence answers “is there any redundancy”
- A basis = spanning + independence = a set of vectors that is “exactly enough”
- Dimension = the number of vectors in a basis
These are the three key steps in understanding a “vector space”, and the column space, the null space and the dimension theorem all build on them.
| Concept | In one sentence | Example (two dimensions) |
|---|---|---|
| Linear combination | scale and then add | 3(1,0) + 2(0,1) = (3,2) |
| The span | every position that can be reached | two non-collinear vectors → the whole plane |
| Linear dependence | there is redundancy (one can be expressed using the others) | (1,1) and (2,2) |
| Linear independence | no redundancy | (1,0) and (0,1) |
| A basis | spanning and independent | 1 |
Symbols Used in This Article
| Symbol | Type | Pronunciation/Explanation | Meaning in This Article |
|---|---|---|---|
| Mathematical symbol | v one through v n | The set of vectors taking part in the linear combination | |
| Mathematical symbol | c one through c n | Arbitrary scalars (real numbers), which may be negative, zero or fractional | |
| Mathematical symbol | linear combination of v’s | A linear combination: scaling each vector and then adding | |
| Mathematical symbol | span of v’s | The set formed by all linear combinations of the vectors (the positions that can be reached) | |
| Mathematical symbol | the reals | The field of real numbers, the range of the scalars | |
| Mathematical symbol | zero vector | The zero vector; the linear combination equals in the test for linear dependence | |
| Mathematical symbol | i hat, j hat | The standard basis: and |
