Add functionality

[cdunn2001]

• [ ] Basic graph (maybe clone an existing Nim graph library?)
  • [x] clear()
  • Node
  • [x] add_node(v)
  • [x] has_node(v)
  • [x] remove_node(v)
  • [ ] g.node[v][key] = value
  • Edge
  • [x] add_edge(u, v)
  • [x] add_edge2(u, v)
  • [x] has_edge(u, v)
  • [x] remove_edge(u, v)
  • [x] remove_edge2(u, v)
  • [x] add_path(edges)
  • [x] add_path2(edges)
  • [ ] collections
  • [ ] all_nodes()
  • [ ] all_edges()
  • [ ] iterators
  • [x] nodes()
  • [x] edges()
  • [x] edges2()
  • API wrappers
  • [x] add_edge(u, v, weight/key) API?
  • [x] remove_edge(u, v, key=value)
• [ ] Directed graph
  • [ ] reverse() (user can clone first, if needed)
  • [ ] clone() (probably just deepCopy())
  • [x] add_path(path) (seq[Node]?)
  • [x] predecessors_iter(node)
  • [ ] in_degree(node) out_degree(node)
  • [ ] in_edges(node) out_edges(node) (with and without keys)
  • [ ] algorithms.dag.is_directed_acyclic_graph() (pypeflow)
• [ ] MultiDiGraph?
• Other functions
  • [x] shortest_path(g, u, v) unweighted
  • [x] shortest_path(g, u, v, weight)
  • Also, unzip calls this without the weight argument
  • [x] shortest_path_length(g, u, v, weight)
  • [ ] ego_graph(g, start, radius=1, center=True, undirected=False, distance=None) (aka neighbour graph)
  • See falcon_kit/mains/ovlp_to_graph.py and falcon_unzip/phased_ovlp_to_graph.py
  • Once with undirected=False
  • Once with radius=10
  • Also nx.ego_graph( ug2, n, radius = 10, undirected=False )
  • [ ] weakly_connected_component_subgraphs(g)
  • See falcon_unzip/graphs_to_h_tigs.py (unzip only)
  • [ ] init_reduce_dict() ?

[cdunn2001]

Ideas from RPG:

• https://github.com/superfunc/charlotte (disappeared!)
• https://gist.github.com/scriptum/34933a248dc22d0dd516 (actually this is not library but uses compact intsets)
• Alternative - bindings to igraph.