binmat | A C library for efficient handling of 2D binary matrices

binmat is a C library typically used in Internet of Things (IoT) applications. binmat has no bugs, it has no vulnerabilities, it has a Strong Copyleft License and it has low support. You can download it from GitHub.

Some applications need to store and manipulate a 2D matrix of binary (ie. boolean) values. For example, an unweighted graph can be stored as an adjacency matrix - the (i,j)-th element is a boolean indicating whether or not the i-th node is connected to the j-th. Given such a data structure, k-connectivity (for each node, the other nodes that can be reached in at most k edges or "hops") can easily be computed by taking the k-th power of the adjacency matrix. However, for large matrices, storing each element as a byte (or worse, a 32 or 64 bit integer) is incredibly inefficient and wasteful, both in time and space. For example, a 10k square matrix requires 400Mb of RAM if stored as 32 bit int values. Binmat is a library that bit-packs these matrices, so that the above 10k square matrix requires just 12.5Mb (the minimum space possible to store such a dense matrix). More than that, binmat takes advantage of extremely fast bit-operations when multiplying matrices. The usual series of multiplications and additions required to compute each element are replaced by bitwise AND and OR operations. Furthermore, on 64-bit hardware each bitwise operation can replace up to 64 multiplications or additions, reducing operations that can take hundreds of clock cycles down to just a single cycle. This gives rise to some very considerable performance increases, especially when taking the power of a matrix. Binary exponentiation is implemented to further improve the performance of higher matrix powers.