Imagine a piece of string that is twisted to fill a square grid. The resulting curve could be characterized as a Hilbert curve, a continuous fractal-space filling curve that enables mapping from a single dimension into multiple dimensions while preserving locality fairly well. While this may sound rather abstract, Hilbert curves have several applications such as encoding geographic information, allocating resources in large computing tasks, and compressing data warehouses. Here, Ryan Adams, has released a numpy-based implementation of Hilbert curves.