Basis and Dimension: Concepts, Formulas and Examples

A basis is a minimal toolkit for a vector space: linearly independent, so nothing is redundant, and spanning, so every vector can be built from it. Once you have a basis, every vector gets a unique address, its coordinates, turning abstract vectors into computable lists of numbers.

The dimension is the number of vectors in any basis, and the deep fact is that this number never depends on the choice of basis: every basis of a space has the same size. Dimension is therefore intrinsic to the space, telling you how many independent directions it truly has, which is why R cubed needs three coordinates and no fewer.

Key Concepts

• Basis: an independent spanning set; every v writes uniquely as c1 b1 + … + cn bn.
• Dimension dim(V): the number of vectors in any basis.
• Standard basis of R to the n: e1 = (1,0,…,0), e2 = (0,1,…,0), and so on.
• Coordinates: the unique scalars expressing v in a chosen basis.
• Dimension theorem: all bases of a space have equal size, so dimension is well defined.

Formulas

• Counting test: in an n-dimensional space, any n independent vectors form a basis, and any n spanning vectors form a basis.
• Too many / too few: more than n vectors are dependent; fewer than n cannot span.
• Standard dimensions: dim(R to the n) = n; polynomials of degree at most k have dimension k+1; m by n matrices have dimension mn.
• Subspace bound: dim(W) is at most dim(V), with equality only for W = V.

Worked Example

Find a basis for the span of v1 = (1,0,1), v2 = (0,1,1), v3 = (1,1,2) in R cubed. Check redundancy: v1 + v2 = (1,1,2) = v3, so v3 adds nothing and the set is dependent. Test v1, v2 alone: if v2 = c v1, the first component gives 0 = c, but then the second would be 0, not 1, contradiction, so they are independent. Every combination a v1 + b v2 + c v3 equals (a+c) v1 + (b+c) v2, so dropping v3 loses nothing. Hence (v1, v2) is a basis and the dimension is 2: geometrically the plane z = x + y, which each generator satisfies.

Common Misconceptions

• A basis must be the standard one. In R squared, (1,1) and (1,minus 1) are a perfectly good basis.
• Dimension counts your starting vectors. It counts basis vectors after removing redundancy; three vectors can span a mere plane.
• Bases are unique. Only the count is fixed; coordinates are unique relative to a chosen basis, the basis itself is not.

Real-World Applications

• Graphics: a camera is a basis of three direction vectors; changing basis moves the viewpoint.
• Compression: JPEG keeps only the most important basis directions and discards the rest.
• Quantum mechanics: dimension counts the distinguishable states of a system.

A basis is a vector space with no waste, and dimension is the honest count of directions. Remove dependent vectors until the set is independent, confirm it still spans, and count what remains.