复杂网络学习笔记：01、networkx实现er随机网络

1、networkx实现er随机网络

import networkx as nx
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import random

'''
    （1）初始化：给定N个节点以及连边概率p∈[0,1]
    （2）随机连边：
	    ①随机选取一对没有边相连的不同的节点
        ②生成一个随机数r∈[0,1]
        ③如果r＜p，那么在这对节点之间添加一条边；否则就不添加
        ④重复步骤①~③，直至所有节点对都被选择一次
'''
class ERModel:
    EdgeNum = 0
    # def __init__(self, NodeNum, EdgeNum):
    #     self.NodeNum = NodeNum
    #     self.EdgeNum = EdgeNum
    def __init__(self, NodeNum, Probability):
        self.NodeNum = NodeNum
        self.Probability = Probability
    #生成er网络邻接矩阵
    def ERNetworkAdjancyMatrix(self):
        N = list(range(self.NodeNum))
        G_ER = nx.Graph()
        G_ER.add_nodes_from(N)
        AdjacencyMatrix = np.zeros((self.NodeNum, self.NodeNum), dtype=int)

        #无权无向图
        AdjacencyMatrix[0][0] = 0
        for i in range(1, self.NodeNum):
            for j in range(i + 1, self.NodeNum):
                probability = random.random()
                if probability < self.Probability:
                    self.EdgeNum += 1
                    AdjacencyMatrix[i][j] = AdjacencyMatrix[j][i] = 1
                else:
                    AdjacencyMatrix[i][j] = AdjacencyMatrix[j][i] = 0
        return AdjacencyMatrix
    #生成er网络图
    def ERNetworkGraph(self, AdjacencyMatrix):
        G_ER = nx.Graph()

        for i in range(len(AdjacencyMatrix)):
            for j in range(len(AdjacencyMatrix)):
                if AdjacencyMatrix[i][j] == 1:
                    G_ER.add_edge(i, j)
        return G_ER
er = ERModel(94, 0.1)
adj = er.ERNetworkAdjancyMatrix()

G_ER = er.ERNetworkGraph(adj)
# pos = nx.spring_layout(G_ER)
# 生成圆形节点布局
pos=nx.circular_layout(G_ER)
# # 生成随机节点布局
# pos=nx.random_layout(G_ER)
# # 生成同心圆节点布局
# pos=nx.shell_layout(G_ER)
# # 利用Fruchterman-Reingold force-directed算法生成节点布局
# pos=nx.spring_layout(G_ER)
# # 利用图拉普拉斯特征向量生成节点布局
# pos=nx.spectral_layout(G_ER)
# #使用Kamada-Kawai路径长度代价函数生成布局
# pos=nx.kamada_kawai_layout(G_ER)
nx.draw(G_ER, pos)
# nx.draw()
plt.show()