This project presents a visual and statistical analysis of traffic accident data to uncover patterns, trends, and high-risk zones. Through compelling visualizations, we aim to improve understanding of accident dynamics and support informed decisions in traffic safety, urban planning, and public awareness.
To visually explore traffic accident datasets and answer key questions such as:
- When are accidents most likely to occur?
- What locations are most accident-prone?
- How do weather, lighting, and road conditions affect accident rates?
- What vehicle types are involved most frequently?
- Python 🐍
- Pandas – data manipulation
- Seaborn & Matplotlib – advanced visualizations
- Plotly – interactive visual analysis
- Jupyter Notebook – exploratory environment
The dataset contains real-world traffic accident records with variables like:
- Time & Date
- Weather & Lighting conditions
- Location / Coordinates
- Vehicle type involved
- Casualties / Severity
🗂️ Dataset Source: A kaggle dataset, Same used for previous analysis
Here are some types of plots generated in the analysis:
- 📆 Heatmap of accidents by weekday and hour
- 🛣️ Choropleth maps or scatter maps showing geographic hotspots
- 🌧️ Bar plots comparing accidents by weather condition
- Accidents peak during evening rush hours on weekdays
- Fog, rain, and low light conditions are correlated with higher accident severity
- Urban intersections are frequent hotspots
- Create a dashboard using Streamlit or Dash
- Use clustering to identify accident-prone zones
- Integrate real-time weather or traffic data for predictive analysis
This project is open-source under the MIT License.
Feel free to explore, use, or contribute!
Linkedin: Yash Ghansham Thakare
Thanks to "Yuri Kashnitsky" for his notebook