GraphMath

Shape & rank

How matrix shape and rank determine what a matrix can and cannot do

Which facts follow immediately from the size of a matrix and its rank?

Shape and rank already determine a large part of the story. They control the dimensions of row space, column space, null space and left null space, and from those dimensions follow one-to-one, onto, solvability, determinant and inverse behavior. This chapter organizes the main cases side by side.

Key ideas

Matrix shape tells us the sizes of the domain and codomain, while rank tells us how many independent directions survive the transformation.

  • Null space dimension is n − r and left-null-space dimension is m − r
  • A matrix is one-to-one exactly when null space is trivial, which means r = n
  • A matrix is onto exactly when left null space is trivial, which means r = m
  • Square full-rank matrices are both one-to-one and onto, so they have nonzero determinant and an inverse
  • Tall full-column-rank matrices are one-to-one but not onto
  • Short full-row-rank matrices are onto but not one-to-one

The chapter compares these facts across the core shape-and-rank cases rather than treating each property in isolation.

Why do shape and rank already tell us so much?

Because once the dimensions of the four fundamental subspaces are fixed, many consequences become unavoidable. Free variables, unreachable outputs, uniqueness of solutions and invertibility all follow from how much dimension is lost and where that loss occurs.

Related chapters

Chapter contents

The PDF is a single document. The page links below are best-effort: most browsers support them, but some viewers may ignore the page hint.

Topic Pages
Square full-rank: m = n, r = n 1–3
Square non-full-rank: m = n, r < n 4–7
Tall full-column-rank: m > n, r = n 7–10
Tall non-full-column-rank: m > n, r < n 11–14
Short full row-rank: m < n, r = m 14–17
Short non-full row-rank: m < n, r < m 18–21

How do one-to-one and onto split across the main cases?

Full column rank gives one-to-one, because null space is trivial. Full row rank gives onto, because left null space is trivial. Only in the square full-rank case do both happen at once.

Was this chapter helpful?

Quick feedback helps us improve the site.