The following graph visualizes co-attendance at the meetings of the Shanghai Rotary Club under President Fitch's term (1930-31). This one-mode network of persons-persons joined by events results from the projection of the initial two-mode network of meeting attendance, using the R package tnet (project_tm function) (Opsahl, 2008). In this network, nodes represent participants (persons who attended meetings), whereas edges represents events. Any edge links two persons who participate in the same events.
What we propose here is a clustered version of the network using Newman fast greedy algorithm (GLay algorithm in Cytoscape App Cluster Maker 2). We chose this method because it works well for weighted networks (Newman 1997) and returned the highest modularity score (we obtain the same results as with Louvain algorithm but slightly better ones than other hierarchical clustering methods). The algorithm detected 6 communities in this network, which means that the Rotary present a more cohesive structure than in its early years (Petit's term). For more detailed information on the communities detected (size of communities, color code and global characterization), please refer to the 'tables" section, which includes tables summarize the results of community detection, including the size of communities (number of members) and their global characterization.
The original dataset used to build this network is available in the "tables" section. An interactive version of this network is accessible on the CyNetShare platform (see URL below). The attached files include a snapshot of the graph and its legend.
N.B. In this network, edges represent co-attendance but not necessarily direct interaction. A major limitation of such one-mode projection is that it creates direct ties between persons without discriminating between mere co-presence and active interaction. Two persons may attend the same event without interacting directly. Yet they will be treated in the same way as two attendees engaged in active intercourse.