def iter_random_walk(
G: nx.Graph, n: Hashable, weight: Optional[str] = None
) -> Iterable[Hashable]:
"""
Given an input graph and a root node, repeatedly yield the results of a random walk starting with the root
node; if the node is disconnected then the walk will consist just of the node itself.
:param G: input graph
:param n: root node
:param weight: name of weight attribute to use, or None to disable, default None
:return: yields nodes in a random walk, starting with the root node
"""
def _next_node(node):
if len(G[node]) == 1:
return list(G[node])[0] # if there is one node, return it
elif weight is None:
return random.choice(
list(G[node])
) # if there is are weights, simply pick the next node at random
else:
nodes = []
weights = []
for _, to_node, to_weight in G.edges(node, data=weight, default=0):
nodes.append(to_node)
weights.append(to_weight)
weights = np.array(weights)
return nodes[
np.random.choice(np.arange(len(nodes)), p=weights / weights.sum())
]
if len(G[n]) == 0:
return # if there are no adjacent nodes to the start node, do not iterate
yield from iterate(_next_node, n)
def squareroot(N, eps):
def _improve(a):
return (a + N/a) / 2.0
def _within(gen):
prev = gen.next()
cur = gen.next()
while abs(cur - prev) > eps:
prev = cur
cur = gen.next()
return cur
return _within(iterate(_improve, 1))
def squareroot(N, eps):
def _improve(a):
return (a + N / a) / 2.0
def _within(gen):
prev = gen.next()
cur = gen.next()
while abs(cur - prev) > eps:
prev = cur
cur = gen.next()
return cur
return _within(iterate(_improve, 1))
def test_iterate():
assert list(itertools.islice(iterate(inc, 0), 0, 5)) == [0, 1, 2, 3, 4]
assert list(take(4, iterate(double, 1))) == [1, 2, 4, 8]
def test_iterate():
assert list(itertools.islice(iterate(inc, 0), 0, 5)) == [0, 1, 2, 3, 4]
assert list(take(4, iterate(double, 1))) == [1, 2, 4, 8]