The rank of a linear map (or matrix) is the dimension of its image (column space); equivalently, the maximum number of linearly independent columns or rows in any matrix representing the map.

Principle

Compute rank by finding the dimension of Im(A) or the number of pivot columns in a row-reduced form. For a linear map V → W with finite-dimensional domain, rank ≤ dim(domain).

Demonstration

For A = [[1,2,3],[2,4,6],[0,0,1]] the second row is twice the first and the third row is independent, so the column space has dimension 2 and rank(A) = 2.

Misapplication

Equating rank with the number of nonzero entries in a matrix, or assuming that a zero determinant alone quantifies exactly how many independent directions are lost without further analysis.

Consequence

Rank determines solvability properties of linear systems (consistent solutions, dimension of solution spaces), the dimension of images under linear transformations, and enters directly into the rank–nullity theorem.

Reversal

The complementary concept is nullity: while rank measures dimension of the image, nullity measures dimension of the kernel; together they sum to the domain dimension in finite dimensions.

Boundary

Typically stated for finite-dimensional vector spaces and matrices; for infinite-dimensional operators the notion of rank extends but may be infinite or require specifying finite-rank approximations. Rank concerns linear independence, not numerical magnitude of entries.

Semantic Tension

Tension with determinant: determinant indicates whether full rank (invertibility) holds for square matrices but does not quantify partial rank; tension with algebraic multiplicity arises when interpreting zero eigenvalues and their multiplicities.

Synthesis

Rank is the linear-algebraic measure of how many independent output directions a linear map produces; it is computed via images or pivots, controls solvability and dimensional trade-offs with nullity, and underlies many structural results about matrices.