[byteshift]

Cool stuff! We actually built a similar system to analyze the NYC taxi dataset (or any other geospatial dataset). We use the (PostgreSQL wire protocol compliant) HyPerSpace database in the backend [1]. Let us know if you're ever experiencing performance problems with PostGIS.

[1] https://db.in.tum.de/downloads/publications/hyperspace.pdf

[billygoat]

Wow, I'll take a look. Looks awesome! This particular dataset was very small because the agency which did the data collection aggregated everything into neighborhoods before they gave it to me. So, performance hasn't been a problem at all yet. We'll be tackling some larger datasets soon, so we'll see!