= Rank inequality for a composition
{wiki}
For <linear maps> $B:U\to V$ and $A:V\to W$ with finite-dimensional $V$, the <rank inequality for a composition> is
$$
\operatorname{rank}(AB)\ge\operatorname{rank}(A)+\operatorname{rank}(B)-\dim V.
$$
Restrict $A$ to $\operatorname{im}B$. The <rank-nullity theorem> gives $\operatorname{rank}(AB)=\dim\operatorname{im}B-\dim(\operatorname{im}B\cap\ker A)$, and the intersection has dimension at most $\dim\ker A$.
For example, the product of two rank-four <endomorphisms> of a five-dimensional <vector space> has rank at least three. The bound controls how much rank can be lost when the image of one map meets the kernel of the next.
Back to article page