def DPA(num_nodes, m):
"""
DPA function return a DPA graph with num_nodes
given num_nodes, and m : the number of existing nodes to which
a new node is connected in each iteration
"""
# make a m nodes completed graph
m_node_comp_graph = make_complete_graph(m)
current_graph = m_node_comp_graph
dpatrial = DPATrial(m)
for node in range(m, num_nodes):
# sum of the in-degrees of existing nodes
# choose randomly m nodes from V-current graph and
# add them to V', where the probability of choosing node j
# is (indeg(j)+1)/(totindeg+|V|)
# totindeg = 0.0
# in_degrees = compute_in_degrees(current_graph)
# for node in in_degrees:
# totindeg += in_degrees[node]
# probability = (in_degrees[idx]+1.0)/(totindeg+len(current_graph))
# using DPATrial class to encapsulate
# optimized trials for DPA algorithm
new_neighbours = dpatrial.run_trial(m)
# new node i is added to set V
# connect the new node to the randomly nodes
current_graph.update({node: new_neighbours})
return current_graph
def dpa_digraph_gen(target_node, step_node):
dpa_graph = DPATrial(step_node)
directed_graph = make_complete_graph(step_node)
for i in range(step_node, target_node):
directed_graph[i] = dpa_graph.run_trial(i)
return directed_graph
def dpa_digraph_gen(target_node, step_node):
'''
generate digraph use DPA Algorithm
Parameters
----------
target_node : int
step_node : int
Returns
-------
graph : dict
'''
DPA_GRAPH = DPATrial(step_node)
directed_graph = make_complete_graph(step_node)
for i in range(step_node, target_node):
directed_graph[i] = DPA_GRAPH.run_trial(i)
return directed_graph
