Dialects

NetworkX

NetworkX support is nearly complete; you can expect most networkx algorithms to run on a Grand graph:

import networkx as nx
import grand

# This one is a grand graph:
g1 = grand.Graph()

# This one is a networkx graph:
g2 = nx.complete_graph(3)

nx.algorithms.isomorphism.GraphMatcher(g1, g2)

Suggestions

Graph ingest is fastest if you use the non-networkx-standard edgelist importers rather than a set of add_edge calls.

Feature Parity

Method	Support	Notes
nodes	✅
edges	✅
adj	✅
add_edge	✅	Return values additionally contain backend-specific metadata.
add_node	✅	Return values additionally contain backend-specific metadata.
pred	✅	Valid on Graph(directed=True) only
succ	✅	Valid on Graph(directed=True) only

IGraph

IGraph support is currently rudimentary, with further development pending convincing use-cases.

Feature Parity

Method	Support	Notes
vs	✅
es	✅
add_edges	🤔	Return values additionally contain backend-specific metadata.
add_node	🤔	Return values additionally contain backend-specific metadata.

Networkit

For more details and documentation on this API, see the networkit docs.

Method	Support	Notes
addNode	✅
addEdge	✅
nodes	✅
iterNodes	✅
edges	✅
iterEdges	✅
hasEdge	✅
addNodes	✅
hasNode	✅
degree	✅
degreeIn	✅
degreeOut	✅
density	✅
numberOfNodes	✅
numberOfEdges	✅
removeEdge	🤔
removeNode	🤔
append	🤔
copyNodes	🤔
BFSEdgesFrom	🤔