A library for Graph-Based Learning in Python.

Provides several types of graphs container objects, with associated visualization, analysis, and embedding functions.

Requires recent versions of:

• numpy
• scipy
• scikit-learn
• matplotlib

Optional dependencies:

• python-igraph
• graphtool

Testing requires:

• nose
• nose-cov

from graphs.construction import random_graph

G = random_graph([2,3,1,3,2,1,2])

print G.num_vertices()  # 7
print G.num_edges()     # 14

G.symmetrize(method='max')
X = G.isomap(num_vecs=2)

G.plot(X, directed=False, weighted=False, title='isomap embedding')()