Course contentsShow
← Mathematics
Lesson 2135 of 3,12144. Linear Algebra — Matrices and SystemsFree lesson

Rank of Sums and Products

Exploring properties like rank(AB) ≤ min(rank(A), rank(B)) and rank(A + B) ≤ rank(A) + rank(B).

Rank of Sums and Products

What you'll learn: How the rank of matrix sums and products relates to the ranks of the individual matrices.

The Core Idea

When you combine matrices through addition or multiplication, their ranks don't behave arbitrarily—there are predictable upper bounds. Think of rank as measuring the "information content" or "dimensionality" of a matrix. Just as you can't create more information than what you started with, these operations have natural limits.

Rank of Matrix Products

When you multiply matrices A and B, the result cannot have more rank than either factor:

rank(AB) ≤ min(rank(A), rank(B))

Why? Matrix multiplication transforms vectors. If A has rank r, it can only map vectors into an r-dimensional space. No matter what B does beforehand, the output is confined to that r-dimensional subspace. Similarly, if B has limited rank, it restricts what can even "reach" A.

Analogy: Imagine B as a filter that loses some color channels, and A as another filter. The final image can't have more color information than what survived the most restrictive filter.

Rank of Matrix Sums

When you add matrices A and B, the rank of the sum is bounded by:

rank(A + B) ≤ rank(A) + rank(B)

The sum's column space lies within the space spanned by combining both matrices' column spaces. Since those spaces have dimensions rank(A) and rank(B), the combined space has at most rank(A) + rank(B) dimensions. Often the actual rank is much smaller due to cancellation or overlap.

Example: If A and B have opposite effects in some directions, adding them might reduce the rank significantly.

Key Takeaway: The rank of a product is limited by the smaller rank of its factors, and the rank of a sum is limited by the sum of the individual ranks—multiplication is more restrictive than addition.