Jainam03/RESCAST-100k

RESCAST-100k-sharded是一个经过处理并分片的RESCAST-100k住宅建筑时间序列数据集。该数据集专注于时间序列预测任务，包含表格数据，涉及建筑能源、住宅建筑、HVAC（供暖、通风和空调）以及ResStock相关主题。数据以Parquet格式存储，包括时间序列文件、清单文件、建筑特征文件、基线结果和元数据文件。时间序列数据每15分钟记录一次，涵盖总电力负荷、室内条件空间温度、加热和冷却设定点、室外干球和湿球温度、相对湿度、风速、漫射和直接太阳辐射等变量，并包含建筑标识符。清单文件提供了建筑与分片文件的映射信息。数据集基于CC BY 4.0许可证分发，适用于住宅建筑能源分析和预测建模。RESCAST-100k-sharded is a processed sharded RESCAST-100k residential building time-series dataset. It focuses on time-series forecasting tasks, containing tabular data related to building energy, residential buildings, HVAC (heating, ventilation, and air conditioning), and ResStock topics. The data is stored in Parquet format, including time-series files, a manifest file, house features file, baseline results, and metadata files. The time-series data is recorded at 15-minute intervals, covering variables such as total electricity load, indoor conditioned-space temperature, heating and cooling setpoints, outdoor dry-bulb and wet-bulb temperatures, relative humidity, wind speed, diffuse and direct solar radiation, along with building identifiers. The manifest file provides mapping information between buildings and shard files. The dataset is distributed under the CC BY 4.0 license and is suitable for residential building energy analysis and predictive modeling.