Data & code for: Leveraging the Spatial Relative Risk Function (SPARR) for Disease and Crime Risk Analysis

This is the data & code repository for the paper titled "Leveraging the Spatial Relative Risk Function (SPARR) for Disease and Crime Risk Analysis", currently under review at Cartography and Geographic Information Science. Abstract: Spatial analysis of disease and crime data is essential for identifying areas of elevated risk and informing targeted interventions. The SPAtial Relative Risk function (SPARR), implemented in the sparr R package, offers a flexible, kernel-based approach to estimating continuous risk surfaces and identifying hotspots of arbitrary shape. sparr is widely applied, but the recently added hotspot report table functionality now enables systematic characterization of statistically significant risk areas. This study demonstrates the utility of SPARR for mapping and quantifying hotspots using two case studies: Lyme Disease in Baltimore County, Maryland, and Violent Crime in Manhattan, New York City. We illustrate both fixed and adaptive kernel smoothing approaches and leverage the new hotspot report table functionality to generate hotspot-level statistics, including risk, case counts, area, and case density. Our results reveal distinct spatial patterns in each context, including three localized Lyme disease hotspots with varying urbanicity and risk, and temporally dynamic violent crime hotspots in Manhattan. The findings underscore SPARR's utility for spatial analysis, highlighting its strengths, including its availability as an R package, its capacity to detect hotspots, and detailed, interpretable characterization of spatial risk.