Feature-Engineered Daily Dataset (2015–2025)

This file contains a feature-engineered daily dataset covering 2015 to 2025. It includes a wide set of derived variables such as temporal features, lagged values, rolling statistics, volatility measures, entropy metrics, jump indicators, and fractal-dimension estimates, created from multiple public data sources. All features are aligned by date and preserved without interpolation except for annual values that are broadcast to a daily frequency. This dataset is intended for reusable modeling, forecasting, and exploratory data analysis workflows.

本文件包含一份经特征工程处理的每日数据集，时间跨度为2015年至2025年。该数据集涵盖多类衍生变量，包括时间特征、滞后值、滚动统计量、波动率指标、熵指标、跳跃指标及分形维度估计值，所有变量均源自多个公开数据源。所有特征均按日期对齐，且未进行插值处理，仅针对年度值进行了每日频率的转换。本数据集旨在支撑可复用的建模、预测与探索性数据分析工作流。