Welcome to the EV-Insight Market Analyzer. This project provides a robust, end-to-end solution for navigating the complex Electric Vehicle landscape. By combining Data Engineering, Inferential Statistics, and Object-Oriented Programming, we have built a pipeline that transforms raw technical specifications into actionable consumer recommendations.
Our aim is to peel back the layers of manufacturer claims, revealing the true relationship between Battery Capacity, Range, and Engine Performance.
The foundation of this analysis is a comprehensive dataset of over 500 Electric Vehicle models.
- Market Metrics: We focused on standardized WLTP (Worldwide Harmonized Light Vehicles Test Procedure) data, which provides the most accurate real-world figures for range and energy consumption.
- Technical Parameters: The dataset includes critical variables such as Battery Capacity (kWh), Engine Power (KM), Top Speed, and Acceleration.
[Data Overview Screenshot]
Raw data often contains noise that can skew recommendations. This phase focused on ensuring the "health" of our dataset:
- Cleaning and Formatting: Using Python and Pandas, we addressed inconsistencies in unit measurements and missing values.
- Statistical Filtering: We implemented the Interquartile Range (IQR) method to identify and remove anomalous energy consumption data. This ensures that recommendations are based on realistic performance averages rather than statistical outliers.
To add a layer of management-grade analysis, we moved beyond simple averages to perform hypothesis testing:
- The Comparison: We conducted a Two-Sample T-Test to benchmark the engine power of industry leaders Tesla and Audi.
- The Result: The test yielded a P-Value of 0.1068.
- The Insight: Because the p-value exceeded the 0.05 threshold, the analysis statistically proved that there is no significant difference in engine power between these two brands in this segment. This allows buyers to focus on other value-drivers like interior tech or charging infrastructure.
The culmination of our data engineering is the EVRecommender class. This is a reusable Python framework designed to solve "Analysis Paralysis" for the end-user:
- Constraint Filtering: The engine processes user-defined limits for Budget (under 350,000 PLN), Range (minimum 400 km), and Battery Size.
- Top-Tier Selection: Instead of a simple list, the tool identifies and returns the Top 3 Vehicles that provide the best performance-to-price ratio within the specified parameters.
[Recommendation Engine Output]
Data becomes intuitive through visualization. Using Matplotlib, we developed a series of insights to map the market:
- Efficiency Mapping: Scatter plots identify the correlation between Battery Capacity and Range, pinpointing "Efficiency Leaders"—manufacturers that achieve superior range with lower battery weight.
- Competitive Benchmarking: We visualized the distribution of battery sizes across manufacturers to help users understand where a specific model sits in the broader market landscape.
[Efficiency Scatter Plot]
To maintain a professional standard, the repository is organized as follows:
data/: Contains the standardized EV market metrics.notebooks/: Includes the Exploratory Data Analysis (EDA) and visualization scripts.src/: Houses the production-readyEVRecommenderclass and statistical testing logic.requirements.txt: Lists all necessary dependencies for reproducing the analysis.
- Language: Python 3.x
- Data Manipulation: Pandas, NumPy
- Statistical Modeling: SciPy (Stats Module)
- Visualization: Matplotlib
- Environment: Jupyter Notebook / Python IDE
By following this pipeline, the project successfully bridges the gap between raw Data Science and practical Product Management.