I am currently implementing an existing networkx implementation in graph-tool and observed that the new implementation runs slower than the networkx one. I then found out that a simple node access is already slower. So I have the exact same undirected graph with 1000 nodes and avg. degree of 2 in nx and gt, and then run: def select(g): for i in range(1000): n = g.vertex(0) def select_nx(g): for i in range(1000): n = g.nodes(0) %time select(gt) CPU times: user 75.6 ms, sys: 69 µs, total: 75.6 ms Wall time: 72.6 ms %time select_nx(g) CPU times: user 3.82 ms, sys: 0 ns, total: 3.82 ms Wall time: 3.46 ms So I am wondering whether this is to be expected and that I can only see the drastic runtime improvements when working with calculations like shortest paths etc. -- View this message in context: http://main-discussion-list-for-the-graph-tool-project.982480.n3.nabble.com/... Sent from the Main discussion list for the graph-tool project mailing list archive at Nabble.com.