Three Large Real-World Growing Networks (2B+ nodes, 4B+ edges, 30+ Years of Evolution)

This Zenodo deposit is part of the supplemental materials and replication package associated with the preprint, "Empirical Growing Networks vs Minimal Models: Evidence and Challenges from Software Heritage and APS Citation Datasets", available on https://arxiv.org/abs/2501.10145 and http://ssrn.com/abstract=5191689. Deposit Description This deposit contains data files related to the construction and analysis of large evolving graphs. Notes:- All `.pkl` files are serialized Python objects (Pickle format).- All graph data are represented using the Compressed Sparse Row (CSR) format.- The derived graphs (BigO and L0) correspond to different rules for inheritance between nodes after temporal partioning. The resources are organized as follows: Main Graph Files Full O/RV/RL-O/RV/RL graph exported from the raw original export (2021/03/23, SWH). - `nodes_20240310.pkl`: Nodes array in the main dataset (CSR format).- `edges_20240310.pkl`: Edges array in the main dataset (CSR format).- `nodesad_20240310.pkl`: Node timestamp attributes.- `tsoutlier_20240310.pkl`: Timestamp outlier information. Total number of nodes = 2,224,378,840Total number of edges = 3,841,679,043 Derived Graph Files O-(RV/RL)-O graphs derived from the full *O/RV/RL-O/RV/RL* graph, after applying temporal partitioning, with inheritance (`BigO` rule) or without inheritance (`L0` rule). Derived Graph using the `BigO` Rule - `nodes_o_derived_O-RVRL-O_BigO_20240429.pkl`: Nodes array for the derived graph under the BigO inheritance rule.- `edges_o_derived_O-RVRL-O_BigO_20240429.pkl`: Edges array for the derived graph under the BigO inheritance rule.- `nodesadderived_derived_O-RVRL-O_BigO_20240429.pkl`: Node timestamp attributes for the BigO-derived graph. Total number of nodes = 139,524,533Total number of edges = 156,682,302 Derived Graph using the `L0` Rule - `nodes_o_derived_O-RVRL-O_L0_20240429.pkl`: Nodes array for the derived graph under the L0 inheritance rule.- `edges_o_derived_O-RVRL-O_L0_20240429.pkl`: Edges array for the derived graph under the L0 inheritance rule. Total number of nodes = 139,524,533Total number of edges = 149,732,521 Metadata and Additional Information - `id_unreal_origins_20210321.txt`: List of indexes of approximately 8,000 origin nodes identified as unreal (relevant for both the main graph and the derived graphs; see the Replication Package for detailed information). Documentation - `README.md`: Provides general information about this deposit, including a link to the associated preprint, copyright, and credit information.本Zenodo存档属于预印本《实证演化网络与极简模型：来自软件遗产（Software Heritage）与美国物理学会（APS）引文数据集的证据与挑战》的补充材料与复现包，该预印本可于https://arxiv.org/abs/2501.10145及http://ssrn.com/abstract=5191689获取。 存档说明 本存档包含与大型演化图构建及分析相关的数据文件。 注意事项： - 所有`.pkl`文件均为序列化Python对象（采用Pickle格式）。 - 所有图数据均采用压缩稀疏行（Compressed Sparse Row, CSR）格式存储。 - 衍生图（BigO与L0）对应时序划分后节点间的两类继承规则。 资源组织方式如下： ## 主图文件 从原始导出文件（2021/03/23，软件遗产）导出的完整O/RV/RL-O/RV/RL图。 - `nodes_20240310.pkl`：主数据集节点数组（CSR格式）。 - `edges_20240310.pkl`：主数据集边数组（CSR格式）。 - `nodesad_20240310.pkl`：节点时间戳属性文件。 - `tsoutlier_20240310.pkl`：时间戳异常值信息文件。 总节点数：2,224,378,840 总边数：3,841,679,043 ## 衍生图文件 基于完整*O/RV/RL-O/RV/RL*图，经时序划分后，通过继承（BigO规则）或不继承（L0规则）得到的O-(RV/RL)-O图。 ### 采用`BigO`规则的衍生图 - `nodes_o_derived_O-RVRL-O_BigO_20240429.pkl`：BigO继承规则下衍生图的节点数组。 - `edges_o_derived_O-RVRL-O_BigO_20240429.pkl`：BigO继承规则下衍生图的边数组。 - `nodesadderived_derived_O-RVRL-O_BigO_20240429.pkl`：BigO衍生图的节点时间戳属性文件。 总节点数：139,524,533 总边数：156,682,302 ### 采用`L0`规则的衍生图 - `nodes_o_derived_O-RVRL-O_L0_20240429.pkl`：L0继承规则下衍生图的节点数组。 - `edges_o_derived_O-RVRL-O_L0_20240429.pkl`：L0继承规则下衍生图的边数组。 总节点数：139,524,533 总边数：149,732,521 ## 元数据与附加信息 - `id_unreal_origins_20210321.txt`：约8000个被识别为非真实起源节点的索引列表（对主图与衍生图均适用；详细信息请参阅复现包）。 ## 文档说明 - `README.md`：提供本存档的通用信息，包括关联预印本链接、版权与署名信息。