Algorithms for Pattern Containment in 0-1 Matrices

We say a zero-one matrix $A$ avoids another zero-one matrix $P$ if no submatrix of $A$ can be transformed to $P$ by changing some ones to zeros. A fundamental problem is to study the extremal function $ex(n,P)$, the maximum number of nonzero entries in an $n \times n$ zero-one matrix $A$ which avoids $P$. To calculate exact values of $ex(n,P)$ for specific values of $n$, we need containment algorithms which tell us whether a given $n \times n$ matrix $A$ contains a given pattern matrix $P$. In this paper, we present optimal algorithms to determine when an $n \times n$ matrix $A$ contains a given pattern $P$ when $P$ is a column of all ones, an identity matrix, a tuple identity matrix, an $L$-shaped pattern, or a cross pattern. These algorithms run in $Θ(n^2)$ time, which is the lowest possible order a containment algorithm can achieve. When $P$ is a rectangular all-ones matrix, we also obtain an improved running time algorithm, albeit with a higher order.