Yoad22/craigslist-used-cars-eda

该数据集是一个用于探索性数据分析（EDA）的Craigslist二手车和卡车列表数据集，源自Kaggle，包含美国各地的真实二手车广告。原始数据约有426,000行和26列，目标变量为价格（price），特征包括年份、制造商、车况、气缸数、燃料类型、里程表读数、变速箱、驱动方式、车型、油漆颜色、州等。数据集经过全面清洗：删除了无用列（如URL、描述、VIN等），处理了缺失值（通过智能恢复制造商和气缸数），过滤了不现实值（如价格在500-150,000美元之间、年份在1990-2026年、里程低于400,000英里），检测了重复项和异常值，并进行了特征工程（如计算车龄）。分析重点探讨了影响二手车价格的因素，包括价格分布、制造商频率、车况影响、里程与价格关系、燃料类型、车龄、驱动方式等，并通过可视化展示了关键洞察。最后，项目还提供了一个基于EDA结果的简单价格计算器。This dataset is a Craigslist used cars and trucks listings dataset for exploratory data analysis (EDA), sourced from Kaggle, containing real used vehicle advertisements across the United States. The raw dataset contains approximately 426,000 rows and 26 columns, with the target variable being price, and features including model year, manufacturer, vehicle condition, number of cylinders, fuel type, odometer reading, transmission, drive type, vehicle model, paint color, state, and more. The dataset underwent comprehensive preprocessing: irrelevant columns (such as URL, vehicle description, VIN, etc.) were removed, missing values were addressed via intelligent imputation of manufacturer and cylinder count, unrealistic values were filtered out (e.g., price ranging from $500 to $150,000, model year between 1990 and 2026, and odometer reading below 400,000 miles), duplicates and outliers were detected, and feature engineering was performed (e.g., calculating vehicle age). The analysis primarily examines factors influencing used vehicle prices, including price distribution, manufacturer frequency, the impact of vehicle condition, the correlation between mileage and price, fuel type, vehicle age, drive type, and more, and presents key insights through data visualization. Lastly, the project also provides a simple price calculator built based on the EDA results.