Steja715/Sub-national-Hostpot-detection-of-malaria-in-tanzania-

Investigated the sub-national distribution of Plasmodium falciparum parasite rate (PR) in Tanzania, by utilizing data from the Malaria Atlas Project # 📍 Spatial Epidemiology & Hotspot Detection: Malaria in Tanzania ## 📋 Summary This project identifies statistically significant geographic clusters of malaria in Tanzania. By analyzing over 4,400 survey points, this study provides a data-driven framework for targeted public health interventions. --- ## 🔬 Core Analysis & Visualizations ### 1. Spatial Topology & Connectivity Before statistical inference, I constructed a **Queen Contiguity** weights matrix to define regional neighbors. This network (visualized below) is the foundation for calculating spatial autocorrelation. ### 2. Hotspot Identification This plot shows the raw point - prevalence. **Key Finding:** Yellow and Orange dot indicates the high prevalence which can be observed in the southern region of Tanzania. ### 3. Cluster Identification The final analysis utilizes **Local Moran’s I** to distinguish between random noise and statistically significant clusters ($p \le 0.05$). **Key Finding:** A significant **High-High (Hotspot)** cluster was identified in the Southern regions, indicating a geographic area where high prevalence is spatially persistent. --- ## 🛠️ Technical Workflow 1. **Data Sourcing:** API retrieval via `malariaAtlas`. 2. **Preprocessing:** Coordinate auditing and spatial joins using the `sf` package. 3. **Statistical Modeling:** Global and Local Moran's I tests conducted in `spdep`. 4. **Visualization:** Multi-scale mapping with `ggplot2`. ---