This vignette demonstrates a complete, reproducible workflow for analyzing NYC census tract demographics in R using spatial data. The analysis moves from importing a GeoPackage layer to producing tract-level racial composition measures (proportions) and concentration measures (densities). The workflow is designed to be defensible, meaning it includes checks that help prevent common sources of error such as missing columns, invalid numeric types, impossible values, and division-by-zero. NYC is a strong case for tract-level demographic analysis because the city contains extreme variation over short distances. Tracts can shift quickly from high concentration to high fragmentation, which makes it possible to observe how demographic patterns become spatial patterns. Importantly, these patterns are not only descriptive. They reflect how urban space is produced through institutions, housing markets, infrastructure, and neighborhood change processes. Measuring proportions and densities at the tract scale provides a way to operationalize these spatial outcomes. Before running the code, ensure the GeoPackage is available in the same folder as this project or update the path accordingly.