Supplementary Material for review (Revealing the co-occurrence patterns of the group emotions from social media data)

<b>For detailed introduction, please download the zip and read the README.docx file</b>This dataset and code package supports the reproducibility of results from the above-titled paper. It includes raw and processed Weibo data, geographic boundary files, emotion classification outputs, and clustering results.The repository reproduces all tables and figures (Table 1, 3–6; Figure 1–9) and the main abstract metric (38.89%) using Python scripts and standard GIS tools. Code is organized for easy execution, with relative paths and full annotation.Main components:<code>data/</code>: Input and intermediate data<code>code/</code>: Scripts for data cleaning, emotion classification (BERT), clustering, and visualization<code>results/</code>: Outputs from analysis<code>README.docx</code>: Reproduction instructionsRecommended environment: Python 3.12 .<br>Run <code>main.py</code> for full pipeline or execute scripts individually.Citation for data source:<br>Liu, Z. et al. (2024). <i>Shifting sentiments: analyzing public reaction to COVID-19 containment policies in Wuhan and Shanghai through Weibo data</i>. Humanities and Social Sciences Communications, 11, 1104.<br>