Code and model-ready data for "From Resilience Assessment to Risk-Transmission Diagnosis: Identification and Spatial Mechanisms of Flood Adaptive Capacity in Chinese Counties"

This record contains the model-ready processed data and reproduction code for the manuscript "From Resilience Assessment to Risk-Transmission Diagnosis: Identification and Spatial Mechanisms of Flood Adaptive Capacity in Chinese Counties" submitted to Nature Cities. The upload includes one processed county-year analysis panel, two Python scripts for factor screening and reproduction of the manuscript result data, and example output tables/source data generated by the scripts. The uploaded dataset is a preprocessed, model-ready complete-case panel. It is not a raw third-party dataset. Raw precipitation products, raw flood-event products, administrative boundary files and hydrographic map layers are not redistributed because they are maintained by external providers or have separate access terms. The analysis panel starts in 2001 because the manuscript models use one-period lagged adaptive-capacity and control variables. The broader project panel contains 2000 observations, but those observations do not have preceding-year values for the lagged variables and are therefore not included in this model-ready upload panel.本数据集包含投至《Nature Cities》的论文《从韧性评估到风险传导诊断：中国县域洪水适应能力的识别与空间机制》所需的适配模型预处理数据与复现代码。本次上传内容包括一份处理后的县域-年度分析面板数据集、两份用于因子筛选及复现论文结果数据的Python脚本，以及脚本生成的示例输出表格与源数据。 本次上传的数据集为经过预处理、适配模型的完整样本面板数据集，并非第三方原始数据集。原始降水数据集、原始洪水事件数据集、行政边界文件与水文地图图层未在此处重新分发，原因是这些数据由外部机构维护或存在独立的获取条款。 本分析面板的时间跨度始于2001年，原因在于论文模型使用了滞后一期的适应能力变量与控制变量。本项目的完整面板共包含2000条观测样本，但其中部分样本不具备滞后变量所需的前一年度数据，因此未被纳入本次上传的适配模型面板数据集。