Spending about an hour on debugging I got really close to the source of the problem. The main problem is that DiGraph (and probably Graph also but I did not test this) is making the labels unique, even if the labels are not hashable. This is very wrong and might lead to really hard to find bugs as these.
The problem is demonstrated below: sage: G=DiGraph([[0,1,2]],loops=g.allows_loops(),sparse=True) sage: G.add_edge(0,1,[]) sage: G.add_edge(0,2,[]) sage: G.edge_label(0,1) is G.edge_label(0,2) True But note that: sage: [] is [] False So DiGraph actually discarded the second list and kept the first one and used that also for the label of the edge [0,2]. This leads to unpredictable results such as: sage: G.edge_label(0,2).append(3) sage: G.edge_label(0,1) [3] (notice the two edges above are different edges). The reason why you example breaks of this is because it uses the function: graph_isom_equivalent_non_edge_labeled_graph which depends on modifying the labels of a graph. I don't know if I will have much time to work on this further so it would be good if someone took it up from here. Kind Regards, Maarten Derickx ps. don't change the code in graph_isom_equivalent_non_edge_labeled_graph, although that is where the error is really coming from, i think the code there is correct and I blame the DiGraph for making non hashable things unique! -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
