Brazil-Dengue-Arboviral-Diseases-Dataset

This dataset provides municipal-level time series data on dengue hospitalizations in Brazil, covering 5,028 municipalities nationwide. The original data were harmonized and interpolated to a weekly temporal resolution, enabling their use in epidemiological forecasting and spatio-temporal modeling. The statistical validity of the weekly interpolation was assessed using a higher-resolution reference dataset provided by the São Paulo State Health Secretariat, allowing a direct comparison between interpolation strategies. Linear interpolation, jittering-based interpolation, and cubic spline interpolation were evaluated, with cubic spline interpolation demonstrating the best statistical adherence to the reference data. In addition to dengue hospitalization rates, the dataset includes a comprehensive set of explanatory variables commonly used in epidemiological models, such as Demographic Density Index, CH₄, CO₂ and NO₂ emissions, Poverty Index, Urbanization Index, Maximum Temperature, Mean Monthly Precipitation, Minimum Relative Humidity, as well as the geographic latitude and longitude of each municipality. This dataset is suitable for multivariate time series analysis, environmental health studies, and the development of machine learning and deep learning models for disease forecasting.本数据集提供巴西全国范围内5028个市级行政区的登革热住院病例时序数据。 原始数据经标准化规整与插值处理后，统一为周度时间分辨率，可支撑流行病学预测与时空建模相关研究。 本研究采用圣保罗州卫生秘书处（São Paulo State Health Secretariat）提供的高分辨率参考数据集，对周度插值方案的统计有效性开展评估，以此实现不同插值策略的直接对比。 本次评估覆盖线性插值、基于抖动（jittering）的插值与三次样条插值三种方案，其中三次样条插值与参考数据的统计契合度最优。 除登革热住院率指标外，本数据集还涵盖流行病学模型中常用的多类解释变量，具体包括：人口密度指数（Demographic Density Index）、甲烷（CH₄）、二氧化碳（CO₂）与二氧化氮（NO₂）排放数据、贫困指数（Poverty Index）、城市化指数（Urbanization Index）、最高气温（Maximum Temperature）、月平均降水量（Mean Monthly Precipitation）、最低相对湿度（Minimum Relative Humidity），以及各市级行政区的地理经纬度坐标。 本数据集可应用于多变量时序分析、环境健康研究，以及用于疾病预测的机器学习与深度学习模型开发工作。