Detroit 911 Calls & Hospital Proximity Research

Exploring spatial correlation between 911 call density and hospital locations in Wayne County using MySQL × Python × R

SQL
MySQL
Python
R
Data Engineering
Spatial Analysis
Public Health
Author

Raameen Ahmed

Published

June 29, 2025

Research Question: Is there a statistical spatial correlation between 911 call density (2018–2025) and proximity to hospitals in Wayne County?
Approach: Transform 2M+ raw calls → clean database → spatial analysis → healthcare planning insights.

♡️ Research Framework

Hypotheses: - H₀ (Null): No significant spatial correlation between hospital locations and 911 call density - H₁ (Alternative): Positive correlation between 911 call density and hospital proximity

Data Sources: - Detroit Open Data Portal (911 calls, 2018–2025) - US Census Bureau 5-Year Estimates (population, demographics) - Wayne County ZCTA mapping data

♡️ Tech stack

Layer Tools & Notes
Database MySQL 8.0+ with centralized staging table
ETL Modular SQL scripts for cleaning, enrichment, modeling
Analysis Python (pandas, matplotlib, numpy) + R (planned)
Import/Export TablePlus for bulk imports, SQL SELECT … INTO OUTFILE for exports
Dev tools VS Code, Jupyter Notebook, TablePlus, Git & GitHub

♡️ Project structure

911Calls-ResearchProject/
├── Data Import:Export Process/
│   ├── Specialized Tables/
│   │   ├── 911CallLocations.csv
│   │   ├── mostCallTypesLocations.csv
│   │   └── mostShootingLocations.csv
│   ├── data_export_process.sql
│   └── data_import_process.md
├── Initial Attempts/
│   └── initial_sqlite_attempt.sql
├── MySQL/
│   ├── 01_schema_definition.sql
│   ├── 02_data_cleaning_and_enrichment.sql
│   ├── 03_data_modeling.sql
│   └── 04_analysis_queries.sql
├── Python Visuals/
│   ├── 911CallsAnalysis.ipynb
│   └── Datasets/
│       ├── 2022FiveYearEstimate.csv
│       ├── 2023FiveYearEstimate.csv
│       ├── 911CallLocations.csv
│       ├── mostCallTypesLocations.csv
│       └── mostShootingLocations.csv
├── R/
│   └── (future R scripts)
└── README.md

♡ Key highlights

  • Centralized Data Hub: MySQL database consolidates 2M+ 911 calls and Census demographic data for reproducible analysis
  • Modular ETL Pipeline: Automated data cleaning, enrichment, and specialized table creation via SQL scripts
  • Multi-Format Analysis: Python notebooks for exploration + planned R visualizations for spatial analysis
  • Research-Driven: Focused on demand-side factors in hospital location planning using emergency call data as proxy

♡ Current progress

Completed:

  • Research question and hypothesis formulation

  • Literature review on hospital location factors

  • Database architecture and data import pipeline

  • Python analysis setup with Census data integration

Next Steps:

  • Data cleaning and validation

  • Spatial analysis with hospital location mapping

  • Statistical correlation testing

  • Heatmap visualization of call density vs. hospital proximity