Baltimore City Crash Analysis

Overview

This Python application visualizes crashes involving non-motorists resulting in injury or death in Baltimore City. It employs Streamlit for web application framework, DuckDB for spatial data processing, and Folium for interactive mapping.

Installation

To run this application, you need to install the following Python packages:

• streamlit
• geopandas
• folium
• streamlit_folium
• janitor
• duckdb
• pandas

You can install these packages via pip:

pip install -r requirements.txt

Usage

After installing the necessary packages, run the Streamlit application:

streamlit run your_script_name.py

Features

• Spatial analysis using DuckDB.
• Interactive mapping with Folium in Streamlit.
• Filtering data based on date range and various map layers including city council districts, neighborhoods, and camera locations.
• Customizable heatmap settings for better visualization.

Data Sources

• Crash data from Maryland State Police Crash Reporting Dashboard.
• City council district and neighborhood boundaries from Open Baltimore.
• Red light and speed camera locations from ArcGIS servers.

For the crash data, download each table from the Tableau dashboard. Make sure you select all fields. Many are not shown by default

Application Structure

1. Data Loading: Load geospatial data using GeoPandas and process it using DuckDB for spatial queries.
2. Streamlit Interface: Utilize Streamlit to create an interactive web application.
3. Mapping: Implement Folium to generate interactive maps with heatmaps and various layers.
4. Session State Management: Manage UI state for interactivity and user inputs.

Code Snippets

Data Loading Function

@st.cache_data
def load_data():
    # Code for data loading and processing

Heatmap Configuration

heatmap_defaults = {
    "radius": 8,
    "blur": 6,
    "min_opacity": 0.3,
    "gradient": {0.2: "blue", 0.4: "lime", 0.6: "yellow", 1: "red"},
}

The heatmap defaults are used to configure the heatmap layer in Folium, but this app also provides in-app customization of the heatmap settings.

It can be a little tricky to find a good set of heatmap settings for a given dataset.

Some settings look better at the city-wide zoom-level, whereas others look better at the neighborhood-level.

The defaults above are the result of some trial and error to find a happy medium that highlights specific problem intersections.

Setting a higher radius and blur makes the heatmap look better when highlighting general areas of high crash density.

Contributions

Feel free to contribute to this project by submitting issues or pull requests.

License

This project is licensed under the MIT License.

Links

• Application GitHub Repository
• Original Data Source