python类fast_gnp_random_graph()的实例源码

def generateGraph(nnodes, edgeprob, directed, pathtosave):
    if os.path.exists(pathtosave):
        matrix = np.loadtxt(pathtosave)
    else:
        shape = (nnodes,nnodes)
        G = nx.fast_gnp_random_graph(n=nnodes, p=edgeprob, directed=directed)
        matrix = nx.adjacency_matrix(G)

        if pathtosave is not None:
            np.savetxt(pathtosave, matrix.toarray(), fmt='%d',)

        print nx.info(G)
        matrix = matrix.toarray()

    return matrix

#######################################################################
# Main
#######################################################################

def generate_data(ds_name):
    graphs = []
    labels = []
    ni = np.random.binomial(n=300, p=0.5, size=100000)
    ei = np.random.binomial(n=3, p=0.3, size=100000)

    for i in range(2000):
        n = random.uniform(130, 180)
        e = random.uniform(0.15, 0.45)
        g1 = nx.fast_gnp_random_graph(n, e)
        graphs.append(g1)
        labels.append(1)
        n = random.uniform(130, 180)
        e = random.uniform(0.15, 0.45)

        n = random.uniform(130, 180)
        e = random.uniform(0.15, 0.45)

    pass

def generate_synthetic():
    import random
    max_nodes=200
    min_nodes=100
    community_num_nodes=10
    graphs=[]
    labels=[]
    com_1= nx.caveman_graph(1, community_num_nodes)
    com_2= nx.star_graph(community_num_nodes)

    for i in range(500):
        num_nodes= random.randint(min_nodes, max_nodes)
        graph= nx.fast_gnp_random_graph(num_nodes, 0.1)
        graph = nx.disjoint_union(graph,com_1)
        for i in range(num_nodes,graph.number_of_nodes()):
            for j in range(num_nodes):
                if random.random() > 0.9:
                    graph.add_edge(graph.nodes()[i], graph.nodes()[j])
        graphs.append(graph)
        labels.append(1)
        num_nodes = random.randint(min_nodes, max_nodes)
        graph = nx.fast_gnp_random_graph(num_nodes, 0.1)
        for i in range(num_nodes, graph.number_of_nodes()):
            for j in range(num_nodes):
                if random.random() > 0.9:
                    graph.add_edge(graph.nodes[i], graph.nodes[j])
        graphs.append(graph)
        labels.append(0)

    return graphs,labels

def generate_profile_file(iterations=9, d=3, steps_per_10_factor=2):
    code = ['"""THIS FILE IS GENERATED. COMPUTERS CAN WRITE CODE NOW TOO."""',
            '',
            'import timeit',
            'import networkx as nx',
            'import graphpca',
            '',
            'print "Timing graphpca(G, 3) on Erdos-Renyi Graph nx.fast_gnp_random_graph(n, p)"',
            'print "\t".join(("n", "p", "t (ms)"))']
    fcn_names = []
    s = steps_per_10_factor
    n_range = [int(pow(10.0, float(i)/s)) for i in range(s, iterations + s)]
    p_range = [0.2] * s + [2 * pow(10.0, -float(i)/s) for i in range(s, iterations)]
    for n, p in zip(n_range, p_range):
        fcn_name = 'profile_{}'.format(n)
        fcn_names.append(fcn_name)
        code.extend([
            '',
            'def {}():'.format(fcn_name),
            '    g = nx.fast_gnp_random_graph({}, {})'.format(n, p),
            '    tic = timeit.default_timer()',
            '    graphpca.reduce_graph(g, {})'.format(d),
            '    toc = timeit.default_timer()',
            '    print "\t".join((str({}), str({}), str((toc - tic) * 1000)))'.format(n, p),
        ])
    code.append('')
    code.extend(['{}()'.format(fcn_name) for fcn_name in fcn_names])

    with open('profile_graphpca_functions.py', 'w') as f:
        f.writelines([l + '\n' for l in code])

def profile_d():
    d_vs_t = [[], []]
    print 'Timing graphpca(G, d) on Erdos-Renyi Graph nx.fast_gnp_random_graph(1000, 0.02)'
    print '\t'.join(('d', 't (ms)'))
    g = nx.fast_gnp_random_graph(1000, 0.02)
    for d in range(1, 950, 30):
        tic = timeit.default_timer()
        graphpca.reduce_graph(g, d)
        toc = timeit.default_timer()
        print '\t'.join((str(d), str((toc - tic) * 1000)))
        d_vs_t[0].append(d)
        d_vs_t[1].append(toc - tic)
    return d_vs_t

def profile_n(iters=9):
    n_vs_t = [[], []]
    print 'Timing graphpca(G, 5) on Erdos-Renyi Graph nx.fast_gnp_random_graph(n, p)'
    print '\t'.join(('n', 'p', 't (ms)'))
    n_range = [int(pow(10.0, i/2.0)) for i in range(2, iters + 2)]
    p_range = [0.2, 0.2] + [2 * pow(10.0, -i/2.0) for i in range(2, iters)]
    for n, p in zip(n_range, p_range):
        g = nx.fast_gnp_random_graph(n, p)
        tic = timeit.default_timer()
        graphpca.reduce_graph(g, 3)
        toc = timeit.default_timer()
        print '\t'.join((str(n), str(p), str((toc - tic) * 1000)))
        n_vs_t[0].append(n)
        n_vs_t[1].append(toc - tic)
    return n_vs_t

def addChaos(di_graphs, k):
    anomaly_time_steps = sorted(random.sample(range(len(di_graphs)), k))
    for t in anomaly_time_steps:
        n = di_graphs[t].number_of_nodes()
        e = di_graphs[t].number_of_edges()
        di_graphs[t] = nx.fast_gnp_random_graph(n, e / float(n * (n - 1)),
                                                seed=None, directed=False)
        di_graphs[t] = di_graphs[t].to_directed()
    return di_graphs, anomaly_time_steps