4. Dashboard for Spatial Analysis of Homicides

---

This project involved the design and programming of an interactive dashboard to analyze homicides in Mexico City. The tool integrates spatial statistics, econometric models, and interactive visualizations, allowing users to explore homicide dynamics at multiple geographic levels. It was conceived as a platform for journalists, researchers, and policymakers to interact with spatial data and interpret complex patterns of violence. 

Explore the dashboard below

1. Data preparation

• Compiled and standardized homicide records at neighborhood and borough levels.
• Processed datasets to ensure geospatial compatibility.

2. Spatial and econometric analysis

• Implemented Kernel Density Estimation (KDE) maps, manipulable by the user.
• Calculated measures of spatial autocorrelation (Moran’s I, LISA).
• Programmed spatial regression models to identify correlates of homicide incidence.
   

3. Dashboard development

• Built an interactive platform using Python Dash and Plotly.
• Integrated filters by year, borough, and neighborhood.
• Designed visualization modules that combine maps, charts, and regression results.

4. AI-enhanced interpretation

• Added buttons powered by natural language processing (NLP) to provide plain-language explanations of statistical coefficients.
• Facilitated accessibility for non-technical users.