Basis & Dimension — Basis and Dimension - Linear Algebra

SECTION 1

Historical Context & Motivation

For centuries, mathematicians worked with numbers arranged in rows and columns, solving systems of equations one problem at a time. They noticed patterns: some collections of arrows (which we now call vectors) could be combined to produce every other arrow in a space, while others were redundant. The search for the smallest, most efficient set of building blocks led to two of the most important ideas in all of linear algebra: basis and dimension.

1637 • Descartes Introduces Coordinates

René Descartes shows that geometry can be described with numbers, laying the groundwork for representing points and directions as ordered pairs and triples.

1844 • Grassmann's Theory of Extension

Hermann Grassmann publishes a bold work describing how to build higher-dimensional spaces from basic directions, foreshadowing the modern idea of a basis.

1888 • Peano Defines Vector Spaces

Giuseppe Peano gives the first formal definition of a vector space, making the concepts of linear independence and spanning precise.

1900s • Basis & Dimension Become Standard

Textbooks by mathematicians like Halmos and Strang establish basis and dimension as the core vocabulary of linear algebra, used in physics, computer graphics, data science, and beyond.

The central question these mathematicians were asking is surprisingly simple: What is the smallest number of vectors you need to reach every point in a space? That question leads directly to the ideas of basis and dimension.

SECTION 2

Core Principles & Definitions

Before we define a basis, we need two prerequisite ideas. First, a set of vectors spans a space if you can reach every vector in that space by adding scaled copies of the vectors in the set. Second, a set of vectors is linearly independent if none of its members can be written as a combination of the others — no vector in the set is "redundant." A basis is the sweet spot where both conditions are true at the same time.

Spanning Set

A collection of vectors that, when scaled and added together, can produce every vector in the space. It may include extra, unnecessary vectors.

Linear Independence

No vector in the set can be made from the others. The only way to combine them and get the zero vector is to multiply each by zero.

Basis

A set of vectors that is both linearly independent and spanning. It is the smallest possible spanning set — remove one vector and you can no longer reach everything.

Dimension

The number of vectors in any basis of the space. Every basis of the same space has the same number of vectors, so dimension is a fixed property of the space.

✦

KEY TAKEAWAY

Think of a basis like a set of LEGO base pieces. With just the right pieces you can build any structure in your collection. If a piece can already be made from the others, it is redundant — throw it out. The number of base pieces you truly need is the dimension of your building world.

SECTION 3

Visual Explanation

Basis Vectors in 2-D

The diagram below shows two basis vectors, e₁ and e₂, in a two-dimensional plane. By scaling and adding them, you can reach any point on the grid. The green target vector (3, 2) is built by taking 3 copies of e₁ and 2 copies of e₂.

The violet arrow e₁ = (1, 0) points right, and the pink arrow e₂ = (0, 1) points up. Together they form the standard basis of 2-D space. The green vector (3, 2) = 3·e₁ + 2·e₂.

Notice that exactly two basis vectors are needed. One vector alone could only cover a single line; three vectors would be redundant because any third arrow in a flat plane can already be built from the first two. The number two is the dimension of the plane.

SECTION 4

Mathematical Framework

Let's make these ideas precise with a little algebra. A linear combination of vectors v₁, v₂, …, vₙ is any expression of the form c₁v₁ + c₂v₂ + … + cₙvₙ, where each c is a real number (a scalar). These scalars are like volume knobs — they control how much of each vector you include.

LINEAR COMBINATION

v = c₁v₁ + c₂v₂ + … + cₙvₙ

Here v is the resulting vector, each cᵢ is a scalar (any real number), and v₁, v₂, …, vₙ are the vectors being combined.

LINEAR INDEPENDENCE TEST

c₁v₁ + c₂v₂ + … + cₙvₙ = 0 ⟹ c₁ = c₂ = … = cₙ = 0

The vectors are linearly independent if the only way to combine them to get the zero vector is when every scalar equals zero. If any non-zero combination works, they are dependent.

DIMENSION FORMULA

dim(V) = n, where n = |basis of V|

The dimension of a vector space V equals the number of vectors in any basis. All bases of the same space have the same size.

💡

Quick Check

If you have a set of 4 vectors in ℝ³ (three-dimensional space), can they be a basis? No! A basis for ℝ³ must contain exactly 3 vectors. Four vectors in 3-D space are always linearly dependent — at least one is redundant.

SECTION 5

Common Bases & Their Dimensions

Different vector spaces have different bases. The diagram below shows the standard bases for 1-D, 2-D, and 3-D real spaces. The standard basis always uses vectors that have a 1 in one position and 0 everywhere else.

As the dimension increases, the standard basis grows by exactly one vector. In ℝⁿ the standard basis has n vectors, and the dimension is n.

Space	Standard Basis	Dimension
ℝ¹ (a number line)	`{ (1) }`	1
ℝ² (a flat plane)	`{ (1,0), (0,1) }`	2
ℝ³ (3-D space)	`{ (1,0,0), (0,1,0), (0,0,1) }`	3
ℝⁿ (n-dimensional)	n unit vectors, each with a 1 in one slot	n
{ 0 } (zero space)	Empty set (no vectors needed)	0

Common vector spaces and their dimensions

An important point: the standard basis is not the only basis. For ℝ², the set { (1, 1), (1, −1) } is also a valid basis because those two vectors are linearly independent and span the entire plane. However, every basis for ℝ² will always contain exactly two vectors, confirming dim(ℝ²) = 2.

SECTION 6

Worked Example

Let's determine whether a given set of vectors forms a basis for ℝ³ and, if so, confirm the dimension.

Does { (1, 0, 0), (0, 1, 1), (0, 0, 1) } form a basis for ℝ³?

1
Step 1 — Count the Vectors

We have 3 vectors, and ℝ³ has dimension 3. So the count matches — a good start! If we had 2 or 4 vectors, we'd know immediately it couldn't be a basis for ℝ³.

Number of vectors = 3 ✓
2
Step 2 — Check Linear Independence

Set up the equation c₁(1, 0, 0) + c₂(0, 1, 1) + c₃(0, 0, 1) = (0, 0, 0). This gives three mini-equations: c₁ = 0, c₂ = 0, and c₂ + c₃ = 0. From the second equation, c₂ = 0, so the third equation gives c₃ = 0.

c₁ = c₂ = c₃ = 0 → linearly independent ✓
3
Step 3 — Verify Spanning

Because we have 3 linearly independent vectors in ℝ³, a theorem in linear algebra guarantees they span ℝ³. (Any n independent vectors in ℝⁿ automatically span the whole space.) But we can verify: take an arbitrary vector (a, b, c). We need c₁(1,0,0) + c₂(0,1,1) + c₃(0,0,1) = (a, b, c). This gives c₁ = a, c₂ = b, and c₃ = c − b. Every choice of a, b, c gives unique scalars, so every vector can be reached.

Every vector (a, b, c) can be reached → spanning ✓
4
Step 4 — State the Conclusion

Since the set is both linearly independent and spanning, it is a basis for ℝ³. The dimension of ℝ³ is confirmed as 3.

Yes — this is a basis. dim(ℝ³) = 3.

SECTION 7

Spanning Sets vs. Bases vs. Linearly Independent Sets

Students often mix up these three related ideas. The table below shows how they compare.

Property	Spanning Set	Linearly Independent Set	Basis
Reaches every vector in the space?	Yes	Not necessarily	Yes
No redundant vectors?	Not necessarily	Yes	Yes
Size compared to dimension	≥ dimension	≤ dimension	= dimension
Example in ℝ²	`{ (1,0), (0,1), (1,1) }`	`{ (1,0) }`	`{ (1,0), (0,1) }`

Comparing three key concepts

✦

KEY TAKEAWAY

A basis is like a perfectly packed suitcase. A spanning set might have extra clothes you don't need (redundant vectors). A linearly independent set might leave out essentials (doesn't cover the whole space). A basis is the Goldilocks set: everything you need and nothing you don't.

SECTION 8

Connections to Advanced Topics

The ideas of basis and dimension are stepping stones to powerful techniques you will encounter later in linear algebra and beyond. The table below previews some of these connections.

This Lesson	Where It Leads
Choosing a basis for ℝⁿ	Change-of-basis matrices — switching between coordinate systems (used in computer graphics and robotics)
Dimension of a vector space	Rank of a matrix — the dimension of the column space tells you how many truly useful equations a system has
Linear independence	Eigenvalues & eigenvectors — special directions in which a transformation just stretches, used in data science (PCA)
Standard basis e₁, e₂, …	Orthonormal bases — bases where every vector has length 1 and they are all perpendicular, making calculations much easier

From basis & dimension to advanced linear algebra

Understanding basis and dimension gives you a lens for seeing the "true size" of any mathematical space. Whether you're compressing an image, training an AI model, or animating a 3-D character, the underlying question is always the same: how many independent directions does this data really have? That number is the dimension, and a basis gives you the directions themselves.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

In your own words, explain why a basis cannot contain the zero vector (0, 0, …, 0).

Show answer

The zero vector makes any set linearly dependent because you can assign it a non-zero scalar (say c₁ = 5) while keeping all other scalars at 0 and still get the zero vector as the result: 5·(0, 0) + 0·v₂ + … = (0, 0). Since a basis requires linear independence, including the zero vector automatically disqualifies the set.

Question 2: BASIC CALCULATION

Is the set { (1, 2), (2, 4) } a basis for ℝ²? Why or why not?

Show answer

No. Notice that (2, 4) = 2 × (1, 2), so the second vector is just a scaled copy of the first. That means they are linearly dependent. Together they only span a single line through the origin, not the entire plane. A basis for ℝ² needs two linearly independent vectors.

Question 3: INTERMEDIATE

Show that the set { (1, 1, 0), (0, 1, 1), (1, 0, 1) } is a basis for ℝ³ by testing linear independence.

Show answer

Set c₁(1,1,0) + c₂(0,1,1) + c₃(1,0,1) = (0,0,0). This gives three equations: c₁ + c₃ = 0, c₁ + c₂ = 0, and c₂ + c₃ = 0. From the first equation c₃ = −c₁. From the second, c₂ = −c₁. Substituting into the third: (−c₁) + (−c₁) = 0, so −2c₁ = 0, meaning c₁ = 0. Then c₂ = 0 and c₃ = 0. Since the only solution is the trivial one, the vectors are linearly independent. Three independent vectors in ℝ³ automatically span ℝ³, so this is a valid basis and dim(ℝ³) = 3.

Question 4: APPLIED

A video-game designer describes every color on screen using three sliders: Red, Green, and Blue (RGB). Explain why the set of colors forms a 3-dimensional vector space and identify a natural basis.

Show answer

Each color can be represented as a vector (r, g, b) where r, g, and b range from 0 to 255. You can add colors component-wise and scale them, so the set of all colors behaves like a vector space. A natural basis is { (1,0,0), (0,1,0), (0,0,1) } — pure red, pure green, and pure blue. Any color is a linear combination of these three. No single slider can recreate another, so they are linearly independent. The dimension is 3, matching our everyday experience that three sliders are both necessary and sufficient to produce any color on screen.

Question 5: CRITICAL THINKING

Suppose V is a subspace of ℝ⁴ spanned by { (1,0,1,0), (0,1,0,1), (1,1,1,1) }. Find a basis for V and state its dimension. (Hint: check if all three vectors are independent.)

Show answer

Test independence: c₁(1,0,1,0) + c₂(0,1,0,1) + c₃(1,1,1,1) = (0,0,0,0). This yields c₁ + c₃ = 0, c₂ + c₃ = 0, c₁ + c₃ = 0, c₂ + c₃ = 0. From the first two, c₁ = −c₃ and c₂ = −c₃. Set c₃ = 1: then c₁ = −1, c₂ = −1. Since (−1)(1,0,1,0) + (−1)(0,1,0,1) + (1)(1,1,1,1) = (0,0,0,0), the third vector equals the sum of the first two. It is redundant! A basis for V is { (1,0,1,0), (0,1,0,1) } and dim(V) = 2. Even though V lives inside ℝ⁴, the subspace itself is only a 2-dimensional plane.

SUMMARY

Lesson Summary

Summary

A basis is a set of vectors that is both linearly independent (no vector is redundant) and spanning (every vector in the space can be built from the basis by scaling and adding). The dimension of a vector space is the number of vectors in any basis — a fixed property that tells you the "true size" of the space.

The standard basis for ℝⁿ uses n unit vectors, but many other bases exist. To test whether a set is a basis, check that the number of vectors equals the dimension and that they are linearly independent. These ideas connect to advanced topics such as matrix rank, change of basis, and eigenvectors, making basis and dimension foundational to all of linear algebra.