Project-2-Python-For-Data-Analysis-BadDrivers

End-to-end EDA project on the FiveThirtyEight Bad Drivers dataset | Analyzing fatal collision patterns across U.S. states using Python, Pandas, Matplotlib & Seaborn

🚗 Exploratory Data Analysis — Bad Drivers Dataset

A structured EDA project on the Bad Drivers dataset from FiveThirtyEight, completed as part of the Python for Data Analysis course.

📌 Objective

Explore fatal collision patterns across U.S. states and uncover key relationships between speeding, alcohol impairment, insurance premiums, and accident rates.

📂 Dataset

• Source: FiveThirtyEight — bad-drivers.csv
• Records: 51 U.S. states
• Features: fatal_collisions, pct_speeding, pct_alcohol, pct_not_distracted, pct_no_prev_accidents, insurance_premium, insurance_losses

🛠️ Tools & Libraries

Library	Purpose
Pandas	Data loading, cleaning, and exploration
NumPy	Numerical operations
Matplotlib	Base plotting
Seaborn	Statistical visualizations

📊 Analysis Structure

1. Imports & Setup
2. Load Dataset
3. Rename Columns
4. First Look (head / tail)
5. Data Structure & Info
6. Missing Values Check
7. Descriptive Statistics
8. Feature Engineering — Speed Level
9. Univariate Analysis — Numerical
10. Univariate Analysis — Categorical
11. Bivariate Analysis — Numerical vs Numerical
12. Bivariate Analysis — Numerical vs Categorical
13. Multivariate Analysis
14. Correlation Heatmap
15. Pairplot
16. Key Insights

💡 Key Insights

• Speeding alone is not the strongest predictor of fatal collisions — alcohol impairment shows a clearer impact
• North Dakota and South Carolina are clear outliers with the highest fatal collision rates
• Insurance premiums don't always reflect actual collision rates — pricing involves more complex factors
• Fatal collision distribution is right-skewed — most states fall between 10–20 collisions per billion miles
• Medium-speed states show the highest variance in collision rates

🚀 How to Run

git clone https://github.com/ammarelsayed-2a/Project-2-Python-For-Data-Analysis-BadDrivers.git
cd Project 2 Python-For-Data-Analysis-BadDrivers
jupyter notebook "Project 2 BadDrivers.ipynb"

👤 Author

Ammar Elsayed — Python for Data Analysis | 2026
LinkedIn