Check node identifiers and edge meanings before running a layout. Filter a large network to the question of interest, then compare statistics with visual impressions rather than treating the layout as proof. Record layout settings and weighting choices when exporting a figure so the visualization can be interpreted and reproduced.
This is best suited to exploring social, citation, or organizational networks. Before committing a larger project, account for these trade-offs: large graphs require substantial memory; attractive layouts can obscure methodological choices.
