Arboviral Disease Cases in Recife, Brazil (2013–2025): Clean and Geocoded Dataset

This dataset contains over 157,000 geolocated arboviral case notifications from Recife, Brazil, covering the period 2013–2025. The dataset comprises 110,560 dengue cases, 38,849 chikungunya cases, and 7,728 Zika cases, representing one of the most comprehensive high-resolution urban arbovirus datasets available for a major tropical city. All records were harmonized across multiple surveillance schemas, cleaned, standardized, temporally normalized, and geocoded using a multi-stage pipeline combining address normalization, CEP harmonization, polygon intersection, and distance-based fallback methods. The dataset includes: Fully standardized English variable names following epidemiological conventions Consistent temporal variables with validated date formats (e.g., date_symptom_onset, date_notification) Verified spatial coordinates derived from IBGE street network data Dual spatial representations (street-level random point and centroid) for robustness Neighborhood-level geocoding provenance and hierarchical spatial identifiers Anonymized and LGPD-compliant records Clean and analysis-ready CSV and Parquet versions Complete data dictionary and reproducible processing pipeline documentation This dataset supports research in: Spatial epidemiology and disease diffusion analysis Geostatistics (e.g., Moran’s I, LISA, Getis–Ord statistics) Spatial regression models (SAR, SEM, CAR/BYM2, SPDE-INLA) Machine learning and spatiotemporal risk modeling Network-based and topological analysis of disease spread Public health surveillance and urban analytics Source data were originally obtained from SINAN (Brazil’s national disease notification system) and systematically processed, harmonized, and enriched by the author for scientific and public use.