Researchers use eigendecomposition to leverage MIT's Wi-Fi network activity data and analyze to the physical environment. We proposed a method to analyze and categorize wireless access points based on common usage characteristics that reflect real-world, placed-based behaviors. It uses eigendecomposition to study the Wi-Fi network at the Massahusetts Institute of Technology (MIT), correlating data generated as a byproduct of network activity with the physical environment. Our approach provides an instant survey of building use across the entire campus at a surprisingly fine-grained level. The resulting eigenplaces have implications for reseach across a range of wireless technology as well as potential applications in network planning, traffic and tourism management, and even marketing.