Abstract
A code city is a visual representation of a code repository using aspects of the repository to build and arrange the city; code files can become buildings, and a directory of files can become a city block. Until now, most code cities have consisted of a static depiction of code repositories. This thesis lays the groundwork to populate a code city with dynamic entities that represent bugs reported in a code repository. These dynamic entities are controlled via an evolutionary algorithm that allows the bug to adapt to its environment as well as the existence of other entities. By measuring fitness as a metric of bug lifespan, the evolutionary algorithm optimizes time spent alive given a bug's surroundings.