CrimeHotspotSim 🔍🗺️

A spatio-temporal hotspot mapping tool that forecasts where crime is most likely to occur in Los Angeles and simulates the impact of preventative interventions.

---

Overview

CrimeHotspotSim is a research-driven project that leverages crime data (2020–present) to predict high-risk areas in Los Angeles. The project’s main contribution is the ability to simulate interventions (for example, adding 10 patrol cars to a grid, or increasing pedestrian activity by 5%) and generate counterfactual hotspot maps that estimate how predicted crime risk shifts under different scenarios.

This approach provides policymakers, law enforcement, and researchers with a way to explore strategies virtually before implementing them in real life.

---

Features

• Preprocess large-scale crime datasets for analysis and modeling.
• Categorize raw crime descriptions into standardized groups (e.g., THEFT, ASSAULT, FRAUD).
• Engineer spatio-temporal features (Year, Month, Hour, DayOfWeek).
• Visualize hotspots on an interactive Los Angeles map using Folium.
• Simulate preventative interventions (patrol allocation, pedestrian activity).
• Export cleaned datasets for further modeling or research.

---

Project Structure

CrimeHotspotSim/
│
├── data/ # Raw & processed datasets (ignored in Git)
│ ├── raw/ # Place original CSVs here (e.g., crime.csv)
│ └── processed/ # Outputs after preprocessing
│
├── notebooks/ # Jupyter notebooks (exploration, preprocessing)
│ └── preprocessing.ipynb
│
├── src/ # Python source code
│ ├── preprocessing.py # Cleaning & feature engineering
│ ├── modeling.py # Training & evaluation
│ └── visualization.py # Hotspot map generation
│
├── visualizations/ # Output plots & interactive maps
│
├── .gitignore # Ensures raw data is not uploaded to GitHub
├── requirements.txt # Python dependencies
└── README.md # Project documentation

---

Installation

Clone the repository and install dependencies:

git clone git@github.com:cOderZmAk/CrimeHotspotSim.git
cd CrimeHotspotSim

# Create virtual environment
python -m venv .venv
source .venv/bin/activate      # On Mac/Linux
.venv\Scripts\activate         # On Windows

# Install dependencies
pip install -r requirements.txt

Usage

1. Place the raw dataset (crime.csv) into data/raw/.
2. Open JupyterLab:

   jupyter lab
   

Run the notebooks in order: preprocessing.ipynb → cleans data and creates processed parquet file. hotspot_mapping.ipynb → generates LA crime hotspot visualizations. simulation.ipynb → runs counterfactual experiments (e.g., add patrols). Check visualizations/ for exported maps (.html) you can open in a browser.

---